Philippines - Second Irrigation Operations Support Project (IOSP II)

Document of The World Bank FOR OFFICIAL USE ONLY Report No: 22153 IMPLEMENTATION COMPLETION REPORT (CPL-36070; SCL-3607A) ONA LOAN IN THE AMOUNT OF US$ 51.3 MILLION TO THE REPUBLIC OF THE PHILIPPINES FOR A SECOND IRRIGATION OPERATIONS SUPPORT PROJECT (IOSP II) June 13, 2001 Rural Development and Natural Resources Sector Unit East Asia and Pacific Region 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.  CURRENCY EQUIVALENTS (Exchange Rate Effective for January-May 2001) Currency Unit = Peso (P) I Peso = US$ 0.020 US$ 1 = 50.0 FISCAL YEAR Government: January 1 - December I ABBREVIATIONS AND ACRONYMS AFMA Agriculture and Fisheries Modemization Act (Republic Act 8435) AO Administrative Order BAS Bureau of Agricultural Statistics CAS Country Assistance Strategy CIA Council of Irrigators' Associations DA Department of Agriculture ERR Economnic Rate of Return GOP Govemment of the Philippines IA Irrigators' Association ICR Implementation Completion Report IMIS Irrigation Management Information System IMT Irrigation Management Transfer IOSP I First Irrigation Operations Support Project IOSP II Second Irrigation Operations Support Project IPM Integrated Pest Management ISF Irrigation Service Fee LGU Local Government Unit MRIIS Magat River Integrated Irrigation System NDC-5 North Diversion Canal-5 in MRIIS (pilot for IMT and for modality of operation) NIA National Irrigation Administration NIS National Irrigation System NPV Net Present Value O&M Operation and Maintenance SAR Staff Appraisal Report Vice President: Jemal-ud-din Kassum, EAPVP Country Manager/Director: Vinay K. Bhargava, EACPF Sector Manager/Director: Mark D. Wilson, EASRD Task Team Leader/Task Manager: Syed Husain, EASRD  FOR OFFICIAL USE ONLY PHILIPPINES SECOND IRRIGATION OPERATIONS SUPPORT PROJECT (IOSP II) CONTENTS Page No. 1. Project Data 1 2. Principal Performance Ratings 1 3. Assessment of Development Objective and Design, and of Quality at Entry 2 4. Achievement of Objective and Outputs 4 5. Major Factors Affecting Implementation and Outcome 10 6. Sustainability 10 7. Bank and Borrower Performance 12 8. Lessons Learned 13 9. Partner Comments 14 10. Additional Information 15 Annex 1. Key Performance Indicators/Log Frame Matrix 16 Annex 2. Project Costs and Financing 18 Annex 3. Economic Costs and Benefits 20 Annex 4. Bank Inputs 27 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 29 Annex 6. Ratings of Bank and Borrower Performance 30 Annex 7. List of Supporting Documents 31 Annex 8. Beneficiary Survey Results 32 Annex 9. Borrower's ICR 46 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not be otherwise disclosed without World Bank authorization.   Project ID: P004589 Project Name. IRRIG OPER SUPP 11 _I_ Team Leader: Syed Husain TL Unit: EASRD ICR Type: Core ICR Report Date. June 13, 2001 1. Project Data Name: IRRIG OPER SUPP 11 L/C/TF Number. CPL-36070; SCL-3607A CountryIDepartment: PHILIPPINES Region: East Asia and Pacific Region Sector/subsector: Al - Irrigation & Drainage KEY DATES Original Revised/Actual PCD: 03/27/1992 Effective: 09/07/1993 10/15/1993 Appraisal: 11/16/1992 MTR: 12/05/1995 03/31/1997 Approval: 05/20/1993 Closing: 06/30/1999 12/31/2000 Borrower/lImplementing Agency: REPUBLIC OF THE PHILIPPINES/NIA Other Partners: STAFF Current At Appraisal Vice President: Jemal-ud-din Kassun A. Karaosmangolu Country Manager: Vinay K. Bhargava Rolando Arrivillaga Sector Manager: Mark D. Wilson Pamela Cox Team Leader at ICR: Syed Husain C. Gunasekara ICR Primary Author: Alvaro Bueno (FAO) 2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N=Negligible) Outcome: S Sustainability: L Institutional Development Impact: M Bank Performance: S Borrower Performance: S QAG (if available) ICR Quality at Entry: S Project at Risk at Any Time: No  3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective. The project's principal objective was to achieve sustainable improvement in the operational efficiency of national irrigation systems (NIS) and thereby help increase agricultural production (mainly rice) and small farmers' incomes, expand rural employment opportunities, and contribute to rural poverty alleviation. The project, to be implemented from mid-1993 to end-1998, was to finance the second phase of the program launched under the Irrigation Operations Support Project (IOSP I, Loan 2948-PH, approved on June 2, 1988). A distinctive feature of the project was its emphasis on participation of beneficiaries in identifying irrigation system improvement works, sharing in improvement costs through labor contributions, and assuming operations and maintenance (O&M) responsibility for laterals and sub-laterals after completion of improvement works. The main objective of improving irrigation efficiency was clear and in line with the Bank's Country Assistance Strategy (CAS) of restoring economic growth, while reducing poverty and improving equity, through, inter alia, accelerated and environmentally sustainable rural development and support to infrastructure. It was also consistent with the high priority given by the Government of the Philippines (GOP) to expanding irrigated agriculture as an instrument for increasing food security, raising small farmers' incomes and alleviating rural poverty. The project, as originally designed, was moderately demanding for the implementing agency. 3.2 Revised Objective: The project's principal objective remained unchanged during implementation. However, the scope of the systems improvement sub-component and the institutional development component was considerably expanded after the mid-term review in March 1997 (details in sections 3.5 and 4.2). 3.3 Original Components: The project had four major components: (a) improvement of 18 selected NIS; urgent repair of 22 structures in 14 NIS; construction of three pilot sedirnent exclusion structures and a few improved pilot water control structures on a selected lateral; and erosion prevention measures in critical areas within and in the vicinity of existing NIS; (b) support to sustain the improved system-level O&M achieved under IOSP I; (c) institutional development through support to the Irrigators' Associations (IAs) and the National Irrigation Administration (NIA); and (d) strengthening of agricultural support services. The components were well-related to the project objective and to the capacity of the implementing agency. 3.4 Revised Components: The major components remained the same during implementation. However, the number of NMS improved was eventually 17, as one system was financed under another program. 3.5 Quality at Entrv: Quality at entry is rated satisfactory in view of the clear main objective, which was consistent with the Bank's CAS and GOP policies and strategies for the sub-sector. Project design and components were coherent, incorporated valuable lessons from IOSP I (Loan 2948-PH), and gave due consideration to the implementing capability of NIA. For the five NIS where improvement works were to commence in the first year, NIA had prepared before negotiations a detailed program of work, supported by relevant engineering data and cost estimates. Although an increase in rice yield was not a project objective, the project did include a small agricultural support services component (1% of project cost). The component was appropriately kept small since, at the time of appraisal, the responsibility for agricultural extension was being transferred from the Department of Agriculture (DA) to the local government units (LGUs) under the Local Government Code of 1991, and there was considerable uncertainty about the organisation, staffing - 2 -  and funding of the devolved agricultural support services. Moreover, the Bank at that time was discussing with the Government a second agricultural support services project (ASSP II), which was expected to provide services to both irrigated and non-irrigated areas. Although the ASSP II did not materialise, the DA eventually implemented the component through a Memorandum of Agreement with the LGUs. In addition, the DA's larger national programs, such as the Grains Production Enhancement Program, provided support services in most of the project areas during the implementation period. One significant question relating to quality at entry that can be raised in retrospect, and was indeed raised during the Staff Appraisal Report (SAR) review process within the Bank, is whether project design was substantive and innovative enough, considering the long experience with rehabilitation and improvement of NIS and water users' involvement with system O&M in the Philippines. To address this issue, and to use the project as a vehicle for initiating long-term irrigation sector reforms in the Philippines, substantial additions to project content were introduced after the mid-term review. The additions to project content were made within a comprehensive framework: (a) The first step was to introduce technical change: rehabilitation and improvement of the existing NIS within the framework of an Irrigation System Improvement Plan (ISIP) prepared by NIA for each system included for improvement. The ISIP aimed at increasing the irrigated area, and, at the same time, ensuring equitable distribution of water by changing the modality of operation from the old labor-and-management-intensive gated water control and distribution system to a new system requiring little or no operation, based on long-crested weirs for water level control, and proportional dividers and/or reduced pipe sizes (in proportion to the area irrigated) at turnouts for water distribution. (b) The technical change was accompanied and followed by institutional change: empowerment of irrigation water users through forrnation or strengthening of the Turnout Service Area Groups, LAs, IA Federations, and Council of lAs (CIAs), and signing of an irrigation management transfer (IMT) contract by NIA with the IAs/CIAs, which transfers to them the responsibility for O&M of the secondary canals in larger irrigation systems (also mandated by the Agriculture and Fisheries Modemization Act (AFMA) of 1997), and O&M of the entire NIS for smaller systems (less than 3,000 ha), as agreed under the follow-on Water Resources Development Project (WRDP - Loan 41 10-PH). (c) The institutional change was accompanied by changes in cost recovery policies: the IMT contract replaced the earlier cumbersome system of sharing of the irrigation service fee (ISF) collection between NIA and the IAs/CIAs by a straightforward negotiated share (usually 50/50); this sharing system also has drawbacks (the lAs frequently complain of late remittance of their share by NIA). and is envisaged to be gradually replaced by a lump sum payment system, either on a per ha basis or based on volume of water supplied by NIA (volumetric pricing). The volumetric pricing pilot, which had been agreed under WRDP to have started by December 31, 1999, started in late 2000 on two pilot laterals, and will start on another pilot lateral in the next wet season (June 2001). (d) The above interventions were accompanied by efforts to establish self-sustaining systems for O&M and rehabilitation of irrigation schemes: this was aimed at overcoming the long-standing problem of deferred maintenance, which leads to deterioration of irrigation systems, which then require rehabilitation prematurely, thus consuming scarce resources which could have otherwise been used for expansion of the irrigated area. This element of reform has been partially achieved (establishment of trust funds), while discussions on earmarked O&M funds are continuing (details in section 6.1). (e) To adddress the problem of NIA staff redundancy resulting from the progress of the IMT program, a proposal for downsizing of NIA was formulated (details in section 4.5). The actual downsizing has not yet started due to lack of funds for early retirement of NIA staff. (f) The elimination of direct and indirect O&M subsidies to NIA was to follow NIA downsizing. Several elements of the reform program (a to d) made significant progress during project implementation. However, since the original IOSP II design had not addressed the NIA overstaffing issue (e above), discussions on this issue started only in year 2000 after the extent of redundancy resulting from the IMT - 3 -  program implementation was known. Neither NIA downsizing nor elimination of O&M subsidies had started by mid-2001 (see further section 4.5). 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective. The outcome of the project in terms of its contribution to food security and incomes of small farmers is rated satisfactory. However, there was a delay of about two years in project implementation, and the Bank loan closing date had to be extended from June 30, 1999 to December 31, 2000. This delay was due to a variety of reasons, including: (a) inadequate budgetary allocation in the first three years of the project (1993-95), when only 49% of the amount projected in the SAR and requested by NIA was allocated by the Government; (b) expansion in the scope of the systems improvement sub-component and the institutional development component (section 3.5), which necessitated a time-consuming process of consultation and agreement with the LAs; (c) a considerable delay in the recruitment of consultants for the design of modality of operation and the DMT program, primarily due to the long time taken by the Government to approve the use of Bank loan proceeds for TA (although it was already included in the Loan Agreement); and (d) the lack of knowledge and experience among the NIA design and field staff of the new technology promoted for water level control and modality of operation, and the consequent delays from trials and errors. Overall, at Bank loan closing, the project is estimated to have benefited 504,850 farm families (110% of the SAR estimate of 460,000), with about 45,300 families benefiting from rehabilitated areas, and about 10,700 from restored and generated areas in 17 improved NIS; and the rest in NIS provided with incremental O&M support. An additional 10,000 farm families are expected to benefit when the restored areas reach full development in year 2003. The main benefits accrued to farmers in restored and generated areas, as these were converted or will be converted from rain-fed to irrigated areas. The incremental rice production at full development of irrigated areas in year 2003 is estimated at about 94,000 t/year, compared to the appraisal estimate of 93,000 tons. The project has a highly satisfactory ERR (section 4.3) and the net financial return to beneficiaries is expected to be higher than that estimated at appraisal (section 4.4 and Annex 3). The project objective of increased operational efficiency of NIS was substantially achieved through a higher cropping intensity and technical changes in modality of operation in 17 NIS (see section 4.2). Moreover, greater equity in water distribution, not envisaged at appraisal, was achieved through changes in modality of operation. The project also supported formation and/or strengthening of lAs and CIAs, achieved participatory planning and implementation of improvement works, and initiated a major IMT program covering 17 improved NIS under the project and other NIS being improved under the follow-on WRDP (Loan 4110-PH). 4.2 Outputs by components: Improvement and Urgent Repair of Selected NIS. The performance of the component, which accounted for about 50% of project costs at appraisal and 52% at completion, is rated satisfactory overall, although its implementation was considerably delayed due to reasons explained in section 4.1. All major improvement works on the 17 selected NIS were completed by the Bank loan closing date. Since minor tertiary and farm-ditch works by farmers are in progress, full development of incremental irrigated areas is expected in year 2003. The SAR targeted improvements on 78,655 ha in 18 NIS with a service area of 95,944 ha. The improved - 4 -  area included both rehabilitated and incremental irrigated area, but the figures for the two were not specified. During implementation, as a result of parcellary mapping and exclusion of one NIS, the firmed-up service area of 17 NIS was determined to be 86,913 ha. Further, after agreement with the August/September 1999 Bank supervision mission, improvement targets were revised upwards to 84,200 ha in both wet and dry seasons (97% of the firmed-up service area). The improved area target included: (i) rehabilitation of about 61,580 ha in the wet season and 60,000 ha in the dry season; and (ii) incremental irrigated areas, consisting of restored areas of about 22,470 ha in the wet season and 24,000 ha in the dry season, and new/generated areas of about 200 ha in both seasons (rehabilitated, restored and generated areas are defined in Annex 3, para. 5). As some incremental irrigated areas were achieved in the early years of the project due to other funding available to NIA, the incremental area to be achieved through project funds alone was projected at about 16,900 ha in the wet season and 17,700 ha in the dry season. By the end of year 2000, about 73,000 ha in the wet season and 72,000 ha in the dry season had been improved in the 17 selected NIS. This represented about 92% of appraisal expectation, and about 86% of the revised higher target. The achievement of rehabilitated and generated areas was 100% of the respective revised targets. However, the restored area was about 11,700 ha in the wet season (52% of the revised target) and 12,200 ha in the dry season (51% of the revised target). The relatively low achievement of restored areas was due to the fact that many tertiary works by farmers were still in progress at the end of year 2000 when NIA completed the improvement works on main and secondary canals. Since, in addition to the system improvement and repair component, the project provided incremental O&M support to all existing 165 NIS, the SAR target for irrigated cropping intensity (increasing from 140% without project to 150% with project) referred to the entire service area of 620,000 ha. Although the irrigated area in year 2000 was about the same as expected at appraisal, the irrigated cropping intensity was 144%, lower than that projected at appraisal, since the service area had increased to 646,000 ha without a commensurate increase in irrigated area (see Annex 3 for details). However, for the 17 NIS included in the system improvement component, the irrigated cropping intensity increased from 149% in without project situation to 167% by the end of 2000, and is expected to reach 194% at full development in year 2003 (compared to 164% projected at appraisal). The results of the survey undertaken for the ICR show that 80% of the farmers and CLA/IA officials reported timely delivery and equitable distribution of water after the project, compared to only 40% before the project (a summary of the survey results is at Annex 8, and details are available in Annex 9 of the Borrower's ICR in Project Files). This and other findings in the survey clearly demonstrate the view held by an overwhelmning majority of beneficiaries that there were significant improvements in irrigation service in the 17 NIS brought about by the project. The positive views of the beneficiaries are corroborated by a technical evaluation, which shows that water level control through long-crested weirs and changes in the modality of operations (proportional weirs and modified pipe sizes, or a combination of the two), greatly simplified operation, thus reducing the labor and management inputs required from both NIA and CIAs/lAs, shortened the time required for land soaking and preparation, and improved timeliness, adequacy and equity of water delivery (details of the technical evaluation are in Annex 8 of the Borrower's ICR in Project Files). Further attention to improving the planting calendar and on-farm water management is needed to reap full benefits of the new technology. Appraisal targets for repairs to major structures were exceeded, as 28 structures in 17 NIS were repaired compared to 22 structures in 14 NIS envisaged at appraisal. Works included repair of diversion dams and sluice gates, additional sluices, river protection works, etc. Two of the three planned silt excluders were constructed; and erosion control measures (vetiver grass, gabion or rip-rap) were implemented in 52 NIS or - 5 -  163% of SAR target (see section 6.1 for further discussion on the siltation problem). Incremental O&M. All activities were substantially completed, and the output of the component (accounting for 36% of project costs both at appraisal and completion) is rated satisfactory. By the end of year 2000, the project had provided nation-wide incremental funding for O&M activities covering 181 NIS with a service area of about 646,000 ha, including canal maintenance (desilting and repair of embankment), maintenance of service roads, maintenance of control structures and gates, and incremental personnel costs for the Institutional Development Officers. Targets for physical improvements under this component were not specified in the SAR, but achievements included over 1,600 km of de-silting of canals and gravelling or grading of service roads, and repair/replacement of some 4,700 structures. The project also sought to close the gap between O&M expenditures and NIA's revenues from ISF collection and thus ensure sustainability of the improved system-level O&M services achieved under IOSP I (Loan 2948-PH). While NIA was expected to maintain in 1992 constant terms the O&M expenditure level (Peso 822/ha) achieved under IOSP I, sections of NIS were to be turned over to the lAs for O&M, and ISF collection was to be increased in tenns of efficiency (from 60% of collectibles in 1991 to 70% by project completion) as well as net collection (from Peso 317 million in 1991 to Peso 757 million in 1999 in current terms), so that at least 95% of the O&M cost in 1999 was met through ISF collection. NIA succeeded in maintaining the O&M expenditures in 1992 constant terms, and in implementing a major IMT program. However, the objective of increasing net ISF collection, which was being achieved until 1997, received a major setback when the then President announced "condonation" of ISF during his State of the Nation Address in July 1998. As a result, many farmers stopped paying ISF. To control the damage done by the Presidential announcement, a compromise solution was worked out, embodied in the Administrative Order No. 17 (AO 17) of August 31, 1998, establishing "socialised" ISF rates based on size of holdings, which were substantially lower than the previous rates. The net ISF collection in 1998 declined to Peso 316 million from Peso 455 million in 1997, and the collection efficiency to 49% from 63%. However, there was a marked improvement in 1999 and 2000, and net ISF collection improved to Peso 398 million and the collection efficiency to 67% (71% in the 17 improved NIS) by the end of year 2000 (unaudited figures). While the collection efficiency in year 2000 was thus close to the appraisal target of 70%, the net ISF collection was about 87% of the 1997 level and only 53% of the appraisal target in current terms. Moreover, the net ISF collection in year 2000 met only about 54% of the O&M expenses (compared to 95% expected at appraisal). The remaining O&M expenses in year 2000 were met from NIA's equipment rental income and government budgetary allocations (further discussion in section 6.1). Institutional Development. The project continued the institutional development efforts initiated under IOSP I for strengthening of lAs and NIA, through improvement in the ISF collection process (parcellary mapping and preparation and updating of irrigation fee registers), staff training, support to the Irrigation Management Information System (IMIS), technical assistance, and provision of equipment and materials. The achievement of the component (accounting for 13% of project costs at appraisal and 11% at completion) was satisfactory overall. The project initiated the implementation of a major IMT program in 17 NIS improved under the project. The institutional efforts focused on reorganization and strengthening of Turnout Service Area Groups and then LAs, and eventually organizing CIAs. This process was piloted in the North Diversion Canal (NDC)-5 of MRIIS, with signing of the IMT contract in December 1998, and later replicated in other 17 NIS. As of end-2000, O&M responsibility under the new IMT contract in these NIS was turned over to 23 CIAs and 17 lAs, covering an area of 44,715 ha and 36,726 farmers. In addition, management by lAs under existing Stage II and III contracts covered 9,571 ha (33 lAs) and 2,308 ha (3 UAs), respectively. The total area under CIA/IA management, as of end-2000, was, therefore, 56,594 ha, or 71% of the 80,240 ha targeted - 6 -  for farmer management (80,240 ha is derived by excluding from the total service area about 6,670 ha served by main canals in large NIS, which are expected to remain under NIA management). The IMT program in MRIIS was halted in mid-2000 due to resulting redundancies of NIA staff and lack of financing for their early retirements. However, negotiations for the signing of IMT contracts in the other systems are continuing, and the total area under farmer management is expected to increase to about 68,000 ha by the end of 2001 (85% of the area targeted for farmer management). As regards the overall institutional development program in 181 existing NIS, achievements for organising new lAs, registering them (as well as the backlog of lAs) with the Securities and Exchange Commission, and providing them with O&M contracts, fell short of appraisal expectations (Annex 1), primarily due to the decision to devote intensive efforts to the IMT program in the 17 NIS rather than extensive efforts in other NIS. However, about 80% of the total NIS service area is now under various forms of participatory irrigation management (compared to about 62% at appraisal in 1992). The number of farmers and IA officers trained (166,260) in all NIS far exceeded the appraisal target of 79,900. This training covered basic leadership, financial and system management, improved rice farming, and integrated pest management (IPM). Local training was provided to NIA staff, except for two NIA staff, who were sent on a study visit to Mexico to observe the IMT program in that country. One of the foremost authorities on the Mexico IMT program also provided consultancy assistance to NIA. His recommendations were consistent with the thrust of the on-going IMT program and improved O&M services. The IMIS did not make much progress as the IMT program and the changed modality of operation required a new set of indicators and monitoring arrangements. This had not been finalized at the time of Bank loan closing. However, data for most of the monitoring indicators specified at appraisal were provided by NIA to Bank supervision missions. The one exception was gender-disaggregated data relating to lAs and training activities. Although some benchmark information on IA women members and officers was collected, and some Gender and Development training provided to NIA staff, the efforts were not systematic. Moreover, targeting of training programs to increase participation of women in IA activities, envisaged at appraisal, was not implemented. Agricultural Support Services. The achievements of this component (accounting for about 1% of project costs both at appraisal and completion) were close to targets (Annex 1). Activities under the component consisted of demonstration trials of appropriate technologies; IA-based seed multiplication and distribution; and training of farmers in soil and water management, integrated pest mnanagement (IPM), improved rice production technologies, crop diversification, seed production, rice-fish culture, etc. All support provided by the component was channelled through the lAs, in coordination with the extension services of the concerned LGUs. Average yield (weighted by service area) of palay (paddy) in the 17 improved NIS, reported for the year 2000 in the survey for the ICR, was 3.92 tlha in the wet season and 4.39 t/ha in the dry season (a detailed discussion on yields is at Annex 3, para. 7). As regards fertiliser use, the survey results showed that, consistent with DA's recommendation, the farmers applied an average of 335 kg/ha (about 7 bags) of conmuercial mixtures (appraisal expectation was 4-6 bags). The effect of the training on IPM was reflected in the reduced expenditure on pesticides. The average cost of pesticide application reported in the survey was P 1,140/ha for the dry season and P 1,234/ha for the wet season, compared to P 1,920/ha and P 2,004/ha in 2000 constant prices, respectively, estimated during project preparation. - 7 -  4.3 Net Present Value/Economic rate of return: Following appraisal methodology, the NPV and ERR were re-estimated for the system improvement and urgent repairs component only, including 100% of costs for institutional development and agricultural support services. These costs account for 64% of total costs. For the incremental O&M component, which accounts for the remaining 36% of total costs, the NPV and ERR were not recalculated since, as stated in the SAR, the component was designed to maintain, and in fact did maintain, the O&M expenditures achieved under IOSP I in 1992 constant terms, and, its ERR is, therefore, the same as for IOSP I (the ERR for IOSP I was estimated in the SAR for IOSP II at 21%, but was re-estimated at 28-35% in the subsequent OED Performance Audit Report No. 18034 of June 19, 1998 for IOSP I). For the system improvement and urgent repairs, only the benefits of incremental irrigated areas in 17 NIS were included, and benefits of rehabilitation as well as urgent repairs, which are significant but difficult to quantify, were not included. Similarly, in the without project situation, there could have been a decline in the irrigated areas due to further system deterioration, but this possible decline has not been taken into account in the analysis. The re-estimated NPV and ERR are thus conservative. Incremental production benefits were estimated on the basis of weighted average yields of 3.92 t/ha and 4.39 t/ha for the irrigated areas in the wet and dry seasons, respectively, as reported in the ICR survey (compared to 4 t/ha for both seasons assumed in the SAR), and 2.2 t/ha for the rain-fed areas (compared to 2.8 t/ha in the SAR), as reported by the Bureau of Agricultural Statistics (BAS) for the year 2000. The weighted average yields for irrigated areas are heavily influenced by relatively high yields in MRIIS, which accounts for about 75% of the service area in the system improvement sub-component. The higher yields in MRIIS are mainly due to the fact that it is a reservoir-backed system with more stable water supplies, compared to the other NIS under IOSP II which are run-of-the-river systems (except for four pump systems with a relatively small total service area) with highly variable water supplies from year to year (details in Annex 3, para. 7). The average official exchange rate of Peso 50=US$I, which prevailed from late 2000 to mid-2001, and a Standard Conversion Factor of 1.0 for local costs were used for the base case ERR. Based on the above assumptions, the ERR is re-estimated at 21% (and NPV at Peso 1,251 million at a discount rate of 10%), compared to 17% estimated in the SAR. Separate ERRs calculated in the Borrower's ICR show about 26% for system improvement and 18% for urgent repairs. The major factor contributing to the relatively high ERR at completion is the farmgate economic price of palay (paddy), which was projected at P 3.78/kg for year 2000 in 1992 constant terms (Peso 6.8/kg in 2000 constant terms) in the SAR, but was estimated at P 8.7/kg or about 28% higher in real terms at end-2000, principally resulting from the steep exchange rate depreciation over the last few years (from Peso 25.5=US$1 in December 1992 to an average of Peso 44 in year 2000, and further to Peso 50 in the first half of 2001), which has made import substitution of commodities like rice economically very attractive. Sensitivity analyses show that the project is economically viable in several adverse scenarios. The ERR is most sensitive to the level of achievement of full development targets for incremental irrigated area and to world rice prices: if only 50% of the incremental irrigated area was achieved, the ERR would decline to 12%, and if world rice prices decline by 20%, the ERR would be 14%. With an assumed irrigated paddy yield of 3.75 t/ha for both seasons over the 30-year period of analysis (the assumption made in the OED Performance Audit Report on IOSP I), the ERR would still be 16%. If the incremental irrigated area starts declining at 3% per annum from year 2001 onwards due to inadequate O&M (historically, the decline has been much less even with inadequate O&M), the ERR would remain at 21%, although the NPV at 10% discount rate would decline a little to Peso 1,146 million (see Annex 3 for details of financial and economic analyses). 4.4 Financial rate of retutrn: It is estimated that, at full development of irrigated areas in the 17 NIS improved under the project (194% irrigated cropping intensity), net returns for share-tenants and owner-operators would be about Peso - 8 -  21,600/ha/year and Peso 44,100/ha/year, respectively (US$430 and 880, respectively). This represents increases of 19% and 21%, respectively, over the without project situation. These returns are also higher than those expected at appraisal (see Annex 3 for details). The increased farm incomes would have direct poverty alleviation effects, since the beneficiaries are small owner-operators and share-tenants. 4.5 Institutional development impact: Although significant institutional reform efforts were initiated under the project, beginning in 1998, and, in particular, considerable progress on IMT was made, several issues remain to be addressed relating to the financial viability of NIA, and the lack of political conmnitment to pursue irrigation sector reforms. The institutional development impact of the project at the time of Bank loan closing is, therefore, considered modest. The IMT program is a positive step towards empowerment of farmers, and a critical element of the irrigation reform process in the Philippines. It has encouraged more active involvement of beneficiaries in decision-making relating to irrigation systems management and CIA/IA operations. It has several desirable features, which distinguish it from the earlier turnover programs followed by NIA: (i) the IMT contracts are signed after improvement works are completed; this greatly improves sustainability of O&M after the tumover, and since the improvement works are aimed at simplifying operation, the CIAs/lAs can concentrate almost exclusively on maintenance after turnover; (ii) larger organisations (CIAs or Federations) are encouraged; this facilitates more professional management (like in Mexico); (iii) the earlier cumbersome arrangements for sharing of ISF between NIA and the CIAs/JAs are replaced by a simple sharing formula (section 3.5); and (iv) monitoring by NIA of CIA/IA performance in O&M during a two-year transition period after signing of the IMT contract, based on a set of technical, financial and organizational indicators, is built into the IMT contract. However, the IMT contract needs further refinement, since it does not have a self-sustaining mechanism to ensure proper O&M by the CIAs/IAs, and the sharing of ISF is still fraught with problems (see sections 3.5 and 6.1). Although the IMT program made good progress until mid-2000, it was halted in MRIIS as staff redundancy became an issue. At the Bank's urging, NIA established a Task Force in mid-2000 to assess the impact of IMT on staffing. According to the Task Force report, some Peso 143 million was needed in the first phase for about 300 staff willing to retire in the NIS covered by IOSP II, WRDP and the ADB-supported ISIP II. Unfortunately, the Government's acute budgetary problems since last year make it unlikely that funding for early retirement of NIA staff will be available soon. While serious efforts should continue to be made to secure funding from the Government for already-identified NIA staff willing to retire, the Task Force should expand its exercise to cover all the other NIS, Regional and Provincial offices, and the central office, so that a phased streamlining program can be formulated to ensure NIA's long-term viability. A Streamlining Plan, which did not consider the implications of the IMT program, was submitted by NIA to the Department of Budget and Management (DBM) in January 2000. The Plan proposed a 50% reduction in authorized NIA monthly positions (from 11,451 to 5,701), with the consequent reduction in personnel costs from Peso 1.64 billion to Peso 0.95 billion. However, since only 5,816 positions were filled at that time, the Plan implied an actual net reduction of 115 positions, with virtually no change in the actual personnel cost of about Peso 0.9 billion. No decision on the Plan has been taken by the DBM so far. Clearly, the reduction in NIA monthly positions needs to be much larger than that proposed in the Plan, particularly if the IMT program is taken into account. There should also be a significant reduction in the number of dailies (about 700 are charged to the NIA operating budget, and about 7,400 to various on-going projects). In addition to a significant downsizing of NIA, continued efforts to improve ISF collection would be necessary to overcome the setback to NIA's financial viabilitv caused by the Presidential "condonation" of -9-  ISF in 1998 (see further section 6.1). It is essential, for the Government to assist NIA in streamlining its staffing and organization, and thereby improving its financial viability. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors ouitside the control of government or implemnenting agency: Improvements and repair in some NIS were affected by occurrence of typhoons and flooding during the construction period, resulting in increased cost and delayed implementation. 5.2 Factors generally subject to government control: Implementation of the key components (system improvement and institutional development) was affected by delays in the release of counterpart funds, and the Government's reluctance to use loan funds for TA (see further section 7.5). 5.3 Factors generally subject to implementing agency control. NMA's efficiency in equipment procurement declined considerably in the last three years of the project. As a result, USS3 million of the Bank loan, which was planned to be used for procurement of pump equipment and accessories for MRIIS, was cancelled. NIA could also have made more concentrated efforts to accelerate implementation of improvement works and the IMT program. 5.4 Costs andfinancing: The total project cost at appraisal was estimated at US$69.6 million (Peso 2,081.3 million), and the Bank loan of US$51.3 million was to finance 74% of the project cost. At completion, the total project cost is estimated at US$68.9 million (Peso 2.2 billion), with Bank financing of US$47.3 million, or 69% of total costs (Annex 2). US$3 million of the Bank loan was canceled on January 5, 2001, and the remaining US$1.0 million on April 30, 2001. 6. Sustainability 6.1 Rationalefor sustainability rating: Sustainability, using the Bank's prevailing definition (the likelihood of the project generating a flow of benefits large enough to exceed the expected marg&nal costs of operation and maintenance), is considered likely. The sensitivity analysis reported in section 4.3 and Annex 3 shows that the economic returns from the project would be attractive in many adverse scenarios, including the scenario of insufficient O&M expenditures leading to a 3% per annum decline in irrigated areas (ERR of 21%). If the capital costs are excluded from the analysis (treated as sunk costs), and only the post-project O&M costs are included, the ERR and NPV would be very high. While sustainability is thus rated likely, a few steps should be taken by the Government and NIA to ensure proper O&M of systems and thereby maximise project benefits. The current average annual O&M expenditure per ha for the 17 MS improved under the project is estimated at Peso 1,500 (US$30). The desirable O&M expenditure per ha, as calculated in a recent consultant study (funded by ADB) on cost recovery mechanisms for NIS, is Peso 2,300 (US$46). The desirable expenditure level would be lower as the IMT program progresses, since the LAs/CIAs with IMT contracts would have their own staff whose salaries and wages are much lower than those of NIA staff. Even with the desirable average expenditure of Peso 2,300/ha, eight of the 17 NIS are not expected to face any problem in meeting the O&M requirements from the ISF income alone (a detailed system-by-system analysis is in section 6 of the attached Annex 9 - Borrower's ICR - and in Annex 7-Ga of the Btorrower's ICR in Project Files). For the others, a combination of increased ISF collection and reduction of O&M cost through IMT will be necessary. It * 10-  should be pointed out, however, that apart from ISF, NIA has other sources of funds for O&M, including equipment rental income, annual budgetary allocations under the locally-funded programs for repair and rehabilitation and drainage of NIS, and O&M subsidy of Peso 145 million per annum (Peso 220/ha) being received from the Government since 1998. The prospects of sustainability of O&M expenditures are, therefore, much better than they would appear if one considered ISF alone, although, in the ideal situation, ISF collection should at least provide for routine O&M, and the other sources of funds should be reserved for repair and rehabilitation. The annual O&M expenditures, a significant proportion of which is for silt removal, can be further reduced if a solution is found to the siltation problem. The silt reduction measures adopted under the project, and continued under the follow-on WRDP (Loan 41 10-PH), appear promising. A preliminary evaluation carried out by NIA last year in four selected NIS showed favorable technical and economic results. In the case of Tumauini, where a settling basin was constructed under IOSP II, the volume of sediments removed declined by about 35% (from the pre-project annual average of 9,997 cubic meters to 6,450 cm). However, substatial additional data is needed to arrive at a definitive judgement on the cost and technical effectiveness of sediment reduction measures. The data is expected to be generated under the on-going sediment monitoring sub-component of WRDP. Moreover, a complementary, broader watershed management program is included under WRDP, and would be continued under the proposed (FY04) River Basin and Watershed Management Program (likely to be an Adaptable Program Loan). To further improve sustainability of O&M expenditures on headworks and main canals in larger systems (>3,000 ha), which, as stipulated under AFMA, will remain with NIA even after IMT, NIA should continue its program of increasing ISF collection (including efforts to re-instate the pre-1998 ISF rates) and reducing its operating costs, particularly by persuading the Government to allocate sufficient funds for early retirements of NIA staff. The current costly practice of collecting ISF from hundreds of thousands of farmers (water retailing) should be replaced by direct billing to the lAs (wholesaling). Moreover, a move from the present system of sharing of ISF based on uniform national rates to a lump sum, system-specific, and negotiated payment (including volumetric pricing, if found feasible under the on-going pilots) should be seriously considered, as it will be advantageous to both NIA and the CIAs/IAs. To improve sustainability of O&M expenditures on smaller systems (<3,000 ha), which are to be completely managed by CIAs/IAs after IMT, and on laterals and sub-laterals of larger systems already transferred or to be transferred to the CIAs/lAs, monitoring by NIA of actual O&M expenditures during the transition period after signing of the IMT contract should be given due attention. Since continued monitoring by NIA after the transition period is uncertain, establishment of earnarked O&M funds from the share of the CIAs/lAs in ISF collection (such funds have already been established in some countries which have IMT programs) should be seriously considered. These earmarked funds are necessary because the IMT contract does not have any in-built mechanism or assurance that there will be proper maintenance in areas covered by IMT. The exact proportion to be earmarked would depend on the O&M cost, and is likely to differ from lateral to lateral and system to system. The earmarked portion could be supplemented, as necessary, by voluntary labor by the farmers or additional funds from the IA's share of ISF collection. The earmarking of O&M funds could be done by changing the bylaws of the lAs. The IMT contracts could also be modified accordingly. The modified IMT contract with an earmarked O&M fund could then combine the desirable features of earlier contract systems followed by NIA: it would retain the independence of CIAs/lAs, and, at the same time, provide for a self-sustaining mechanism for proper O&M. The earmarked O&M fund, to be available with the CIAs/lAs, would supplement the trust fund, currently kept by the Regional Managers of NIA, which was established for each NIS improved under IOSP II from - 11 -  the equity contribution of farmers for improvement works (25% of labor cost for improvement of secondary canals and 100% of labor cost for tertiary systems). Guidelines for the utilization of the trust fund were finalized by NIA and circulated to its field offices in April 2001. These should be disseminated to the CIAs/lAs immediately, so that they can access the funds for their priority needs. The CIAs/IAs should also consider setting their own ISF rates, since they are not bound by AO 17. A higher collection of ISF would help ensure proper O&M and capital build-up for their other needs. 6.2 Transition arrangement to regular operations. Since development of incremental irrigated areas and institutional processes required for IMT were still in progress in several NIS at the time of Bank loan closing, transition arrangements to regular operation are extremely important. The most important objective would be to ensure that conversion of rain-fed to irrigated areas proceeds as planned, and is completed by year 2003, since the ERR is highly sensitive to this development (scc section 4.3). NMA should monitor this development and should prepare six-monthly progress reports. Several steps should also be taken to ensure the sustainability of the improvement works and the IMT program: (i) close monitoring of the performance of CIAs/IAs as regards O&M in IMT areas is necessary; (ii) monitoring of the functionality and impact of changes in the modality of operation should be continued; (iii) the O&M plans for the systems and laterals transferred or to be transferred, which were being translated into local dialects at the time of Bank loan closing, should be delivered to the CIAs/IAs without further delay; (iv) the prototype manual for IMT, which was produced by the TA team, should be polished by incorporating improvements based on the experience gained in different NIS; (v) the cascading approach to technology transfer proposed by the TA team (training the IA officers, who would, in turn, train farmers) and followed in the IMT pilot area (NDC-5), should be implemented in other IMT areas to facilitate rapid capability build-up of ClAs/IAs; and (vi) efforts should continue to establish earmarked O&M funds (section 6.1). As regards performance indicators to help monitor future operations and impact of the project, current activities of NIA field offices and BAS in collecting yield data should continue. The monitoring format produced by NIA and regularly used by Bank supervision missions should also be maintained and updated every year. The format includes indicators for service area, irrigated area in wet and dry seasons, irrigated cropping intensity, benefited area in wet and dry seasons, ISF collectibles and gross and net collection, other income used for O&M, O&M expenses, and net income (deficit) for O&M. The monitoring table used during supervision showing the achievement of targets for rehabilitated, restored and generated areas under the project should be updated regularly until full development is reached. NIA should also have regular rather than sporadic gender-disaggregated data collection related to CIA/IA activities and leadership. 7. Bank and Borrower Performance Bank 7.] Lending Bank performance during the design, preparation and appraisal processes was generally satisfactory. However, there was scope for a more substantive and innovative project design (section 3.5). 7.2 Supervision: Bank supervision is rated satisfactory. The supervision missions had an appropriate skill mix and staff continuity, and reviewed project progress every six months. The Bank helped NIA initiate a reform program for the irrigation sector (section 3.5), not envisaged at appraisal. The Bank also played a decisive role in ensuring technological change in the 17 NIS (structures for water level control and improved modality of operation), expanding the IMT program coverage and content (including complete transfer of - 12-  small systems to the CIAs/lAs, not envisaged at appraisal), initiating the discussion on earmarked O&M funds, and helping initiate volumetric pricing pilots. The Bank also assisted the Government in resolving the problem of shortage of counterpart funds by establishing a Project Special Account (PSA) and a 90-day advance facility from the PSA to field offices to expedite implementation of works. An 1 8-month extension of the loan closing date facilitated completion of major physical works by NIA. The Quality Assurance Group (QAG) of the Bank reviewed this project in the third Rapid Supervision Assessment (FY99), and gave a satisfactory rating to focus on development impact, supervision of fiduciary aspects, adequacy of supervision inputs and processes, and overall quality of supervision, and a highly satisfactory rating to realism of project performance ratings. 7.3 Overall Bankperforrnance: The overall Bank performance is rated satisfactory in the light of discussion in sections 7.1 and 7.2 above. Borrower 7.4 Preparation: Project preparation by NIA was satisfactory. Detailed arrangements for first year civil works and other activities were in place before negotiations. A notable feature of project preparation was the high level of participation by the LAs in identifying and agreeing to the improvement works to be undertaken in the selected NIS. Equity participation by the farmers through contribution of labor in improvement works was a good idea as it led to accumulation of funds which can be used by the CLAs/LAs for priority activities. 7.5 Government implementation performance: The implementation perfornance of the Borrower is rated unsatisfactory due to several reasons: (a) insufficient budgetary allocation to the project in the first three years of implementation; (b) long delay in approving the use of Bank loan proceeds for TA; (c) "condonation" of ISF by the then President; and (d) lack of initiative to implement IMT (see further section 8). The above actions and inactions considerably delayed project implementation and benefits, and undermined irrigation sector reform efforts. 7.6 Implementing Agency: Overall, the performance of the implementing agency, NIA, was satisfactory. Detailed and regular monitoring of project progress was undertaken, and the draft ICR on behalf of the Borrower was submitted to the Bank in time. No significant problems relating to accounts and audits were encountered. However, equipment procurement performance was unsatisfactory in the last three years of the project (section 5.3). 7.7 Overall Borrower performance. Despite the problems mentioned in section 7.5, the overall performance of the Borrower is rated satisfactory since most project objectives were substantially achieved by the loan closing date. 8. Lessons Learned * Political commitment is essential for irrigation sector reform. (a) Although AFMA (Republic Act 8435) of 1997 mandated IMT and stipulated its completion within five years, that is, by the end of year 2002, no initiative was taken by central agencies to implement this requirement. The IMT program, which commenced under IOSP II, was based entirely on the agreement between the Bank and NIA, and the central agencies have so far not addressed the issue of NIA staff redundancy and financing requirement for early retirements. As a result, the IMT program remains stalled in MRIIS, and its future progress in MRIIS and other large NIS remains uncertain. (b) An equally serious evidence of the lack of political comniitrnent was the populist action taken in mid-1998 when payment of ISF by the farmers was publicly "condoned" by the then President of the country. As a result, many farmers stopped paying ISF, and ISF collection declined dramatically. This action seriously undermined the efforts which had been made in the Philippines over the - 13-  last two decades to improve cost recovery, increased budgetary subsidies to NIA in the short term but jeopardized NIA's long-term financial viability, weakened the lAs as their share of ISF declined with the overall decline in ISF collection, and diminished the capability of lAs and NIA to undertake proper O&M of irrigation systems. A strong political commitment is urgently needed if efficiency and equity are to be improved in the irrigation sector, and if the AFMA provisions relating to IMT are to be implemented. * A program approach is needed for irrigation sector reform and NIS improvement. The system improvement and IMT models developed under IOSP 11, and the sequence of IMT following improvement works, showed promising results in terms of improvement of both efficiency and equity in irrigation services and generation of greater farmer involvement in planning and implementation of improvement works and in O&M of irrigation systems. Moreover, rehabilitation, repair and improvement of existing NIS proved to be a cost-effective way to increase irrigated areas, as the average cost of incremental irrigated area under the project was only about US$1,500/ha equivalent, which is much lower than the cost/ha of new construction. However, only a small number of NIS were covered under IOSP II and other externally-supported projects, including WRDP (Loan 4110-PH). Since the gap between the service area and the actual irrigated area of NIS is over 200,000 ha, a big potential exists to close the gap at a relatively low cost by replicating the model developed under IOSP II in the remaining NIS. This, along with implementation of other elements of the irrigation sector reform program initiated under IOSP II (section 3.5), necessitates adoption of a 10-15 year program approach. The proposal recently submitted by NIA to the central agencies for improvement of the remaining NIS should be seriously reviewed by the central agencies, and they, along with NIA, and with the support of extemal agencies, if necessary, should ensure that the reform momentum is not dissipated. - 'The borrowing policy for TA should be decided early in the project cycle. Although the Loan Agreement included a disbursement category for TA, the Government, given its general policy to minimise borrowing for TA, took a long time during the implementation phase to approve the use of the loan proceeds for TA required for the design and piloting of modality of operation works and IMT. This seriously delayed project implementation and benefits. The borrowing policy for TA should be decided early in the project cycle, so that alternatives for project implementation can be considered before appraisal. Participatory system evaluation improves sustainabiity of improvement works. The NIS Assessment undertaken by NIA, with the participation of the [As, for each system prior to improvement, proved to be a valuable tool for identifying the performance deficiencies and improvement needs of the system. The participatory evaluation ensured acceptability by the lAs of the specific improvement works undertaken, particularly changes in the modality of operation, and thereby improved prospects for sustainability of the improvement works. 9. Partner Comments (a) Borrower/implementing agency: Comments from NIA We have no objection on the presentations in the ICR. In fact, they have provided a clearer picture of details that effectively substantiate the factors, constraints, lessons and items for sustainability. One minor thing we only want to elaborate pertains to the last paragraph of Institutional Development in - 14 -  section 4.2 "...targetting of training programs to increase participation of women ... was not implemented". We wish to explain that while this did not happen as envisaged in the SAR, this aspect was mainstreamed into our programs and strategies. We encouraged the adoption of LA policies to ensure that there are women BOD members. In the conduct of training programs, we ensured that there are women participants as much as possible. Further, and since the situation in our country is different from other countries, we realize that it is not increasing women participation that we need but providing the opportunities. The improved modality of operations gave more opportunities for women to occupy operations - related positions because the control structures are easy to operate and monitor. For more than six years span of Systems Improvement under IOSP II offered immeasurable opportunities for physical and institutional advancement on 17 NIS's. Achieving the goal at the highest level becamne difficult at a time when political climate is less cooperative. The needed strong political commitment has been disturbed by internittent changes in the agency leadership within the period of implementation. Nevertheless, what was achieved sufficiently justifies nurturing similar efforts in the future. With lessons leamed accumulating and political situation ripening, the needed efficiency and equity to improve irrigation sector is more appealing as ever. To ensure a successful replication of the physical improvements and improved modality of operation, design and construction of weirs and proportional dividers must be completed simultaneously before operations. (b) Cofinanciers: Not applicable (c) Other partners (NGOs/private sector): Not applicable 10. Additional Information The project's environmental category was appropriately rated B at appraisal. While there is no evidence of any significant increase in application of fertilisers due to the project, the survey undertaken for the ICR preparation shows that pesticide use has actually declined in project areas since appraisal (see discussion on Agricultural Support Services in section 4.2). - 15-  Annex 1. Key Performance Indicators/Log Frame Matrix Outcome / Impact Indicators: tndicatorMatrix Projecte in ta estPSR ActuaULatest Estimate Farmers benefited (No.) 460,000 504,850 Farm (rice) net income (P/ha/year) Share tenants 19,104 21,600 Owner-operators 38,209 44,100 Incremental rice production (t/year) 93,000 94,000 Output Indicators: IndiirlMatrix Projected in te PSR Actua/Latest EstImat* A - Improvement and urgent repairs Number of NIS for system improvement 18 17 Service area (ha) 95,944 86,913 Improved area (ha)* Rehabilitated - Wet season 61,577 61,147 - Dry season 59,970 59,644 Restored - Wet season 22,463 11,705 - Dry season 24,020 12,144 Generated -Wetseason 210 210 - Dry season 210 210 Incremental irrigated area (ha)- Wet season 22.673 11,915 Dry season 24,230 12,354 No. of NIS for urgent repairs 14 17 No. of structures for repair 22 28 Silt excluders (No.) 3 2 Erosion cortrol (No. NIS) 32 52 Pilot water control structure 1 1 B- Incremental O&M Numberof NIS 165 181 Irrigated cropping intensity* 150 144 ISF collection efficiency 70 67 Canal maintenance (km) Not quantified 1.640 Control gates and structures repaired Not quantified 4,707 C- Institutional development No. of new lAs organised (overall) 47 29 No. of lAs registered with the SEC (overall) 347 130 No. of lAs with O&M contracts (overall) 536 226 No. of IMT contracts in 17 NIS Not quantified 40 Ha under farmer management in 17 NIS Not quantified 56,594 No. of IA officerstfafners trained (overal) 79,900 166,260 0 Agricultural support services Number of demonstraton farms 421 345 Numberof IPM-Fanmer Field Schools 18 17 Farming systems demonstraton (No.) 11 12 Fertiliser demonstrations (No. sites) 120 102 Variety trials (No. sites) 7 7 Rice/fish cuture (No. sites) 18 9 Grain drying pavements (No.) 4 4 Seedling nursery (No.) 1 1 Field days (No.) 98 43 - 16 -  Seed production (No. 40 kg bags) Foundation seed procured 126 100 Registered seed procured 2,863 2,259 Foundation seed dstributed 126 100 Registered seed cistributed 1,908 1,395 Registered seed produdion 12,960 8,272 Certified seed production 284,600 124,000 Certified seed distribution 205 267 Farmers' training (No. participants) Rice production/pest management 261 189 Farmer field school 3,087 2,743 Crop diversffication 44 43 Seed producion 540 370 Farming systems technology 50 20 Capital build up 27 14 Rice/fish culture 50 25 End of project Note: The column 'Projected in last PSR' refers to projections in SAR, Mid-Term Review or last PSR, as the case may be. * For a definition of rehabilitated, restored and generated area, see Annex 3, para. 5. ** The incremental irrigated area is the total of restored and generated area. The achievement of restored and generated areas as of end-2000 was 11,915 ha in the wet season and 12,354 ha in the dry season, as shown above. Full development of the targeted 22,673 ha in the wet season and 24,230 ha in the dry season is expected by early 2003. Some incremental irrigated area was achieved through other funding sources in the early years of the project. The targets to be achieved through project funding alone are estimated at about 16,900 ha in the wet season and 17,700 ha in the dry season at full development in 2003. "** The figures refer to overall intensity for the entire service area of NIS. For the 17 NIS selected for improvement (18 at appraisal), the target irrigated cropping intensity was 164%, the achievement by end-2000 was 167%, and projected by full development in early 2003 is 194%. - 17 -  Annex 2. Project Costs and Financing Project Cost by Component. (in US$ million equivalent) ra i -sal Actual/Ls Percentage of one l~~Esimae Etm pria 000j:jPrw$ectL Cost Bytiti|X ECoponent0 US$ m ;0;0 0jU illion; 0 tUS$ millilon;0; Improvement and Urgent Repair of NIS 29.10 36.15 103 Support to System-level O&M 22.20 24.66 100 Institutional Development 7.90 7.25 82 Agricultural Support Services 0.80 0.80 89 Total Baseline Cost 60.00 68.86 Physical Contingencies 3.40 Price Contingencies 6.20 Total Project Costs 69.60 68.86 Total Financing Required = 69.60 68.86 Note: In the column 'Percentage of Appraisal', appraisal costs are inclusive of contingencies. Project Costs by Procurement Arran ements (Ap raisal Estimate) (US$ million equivalent) Eedttr atgV IB 0 0NCB Otet ;;.F MTd ta lgF Cos 1. Works 0.00 14.40 21.60 0.00 36.00 (0.00) (1311.40) (20.10) (0.00) (33.50) 2. Goods 3.60 1.00 0.00 0.00 4.60 (3.20) (0.90) (0.00) (0.00) (4.10) 3. Services 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 4. Incremental O&M 0.00 6.20 18.50 0.00 24.70 (0.00) (2.40) (7.00) (0.00) (9.40) 5. Institutional 0.00 0.00 4.30 0.00 4.30 Development & T.A. (0.00) (0.00) (4.30) (0.00) (4.30) 6. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (o.00) (0.00) (0.00) (0.00) Total 3.60 21.60 44.40 0.00 69.60 (3.20) (15.70) (31.40) (0.00) (51.30) - 18 -  Project Costs by Procurement Arrangements (Actual/Latest Estimate) (US$ million equivalent) Expenditure Category Procurement Method N.8.F. Total Cost N4C8 Other' _ _ _ _ _ _ _ _ 1. Works 0.00 15.30 21.65 0.00 36.95 (0.00) i(13.02) (18.42) (0.00) (31.44) 2. Goods 3.00 1.35 0.00 0.00 4.35 (3.00) (0.96) (0.00) (0.00) (3.96) 3. Services 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00) 4. Incremental O&M 0.00 4.20 20.46 0.00 24.66 (0.00) (1.53) (7.43) (0.00) (8.96) 5. Institutional 0.00 0.00 2.90 0.00 2.90 Development & T.A. (0.00) (0.00) (2.90) (0.00) (2.90) 6. Miscellaneous 0.00 0.00 0.00 0.00 0.00 (0.00) (0.00) (0.00) (0.00) (0.00 Total 3.00 20.85 45.01 0.00 68.86 (3.00) I(15.51) (28.75) (0.00) (47.26) Figures in parenthesis are the amounts to be financed by the Bank Loan. All costs include contingencies. v Includes civil works and goods to be procured through national shopping, consulting services, services of contracted staff of the project management office, training, technical assistance services, and incremental operating costs related to (i) managing the project, and (ii) re-lending project Rinds to local government units. Project Financing by Component (in US$ million equivalent) Pereeatage of Appraisal Component Appraisal Estimate Actual/Latest Estimate _ _ Bank I Govt. CoF. Bank Govt. CoF. Bank Govt CoF. Improvement and Repair 32.70 2.40 30.75 5.40 94.0 225.0 Incremental O&M 9.40 15.30 8.96 15.70 95.3 102.6 Institutional Development 8.40 0.50 6.85 0.40 81.5 80.0 Agric. Support Services 0.80 0.10 0.70 0.10 87.5 100.0 Total 51.30 18.30 47.26 21.60 92.1 118.0 -19-  Annex 3: Economic Costs and Benefits Pree;t ValueofFaws(2000 Con t Te Appraisal Latest Estimate Appra 0isa La \ test Estimate Economic rate of 17 21 return (%) Net return (P/ha/year) -Share-tenant 19,104 21,600 -Owner-operator _ 38,209 44,100 NVP at 1O % (Peso 292.1 1,251.4 million) I I Note. Netfinancial returns and NPVat appraisal have been converted to 2000 constant terms using the consumer price index. FINANCIAL AND ECONOMIC ANALYSIS for ICR A. Introduction 1. The analysis presented here re-estimates the economic rate of return (ERR) calculated in the Staff Appraisal Report (SAR), using updated prices, actual project costs, and latest information on crop production. The impacts of the project on agriculhtual production and farmers' incomes are also re-estimated. 2. The framework of the analysis follows, as far as possible, that of the SAR. Project costs and physical areas covered were based on project records and are considered fairly accurate. The crop yield estimates are based on a survey conducted towards the end of year 2000 by the consultants who prepared the Borrower's ICR. Since yield projections over a 30-year period of analysis for the ERR calculation are always subject to uncertainties, a sensitivity analysis, based on a lower yield level, was also undertaken to test project economic viability in an adverse scenario. B. Project Benefits 3. The project's major objective was to achieve a sustainable improvement in the operational efficiency of the National Irrigation Systems (NIS). The main benefit of the project was an increase in irrigated area or irrigated cropping intensity, which, in turn, has and would help increase palay (paddy) production, farm incomes and employment. 4. Since, in addition to the system improvement and repair component, the project provided incremental O&M support to all existing 165 NIS the SAR target for irrigated cropping intensity (increasing from 140% without project to 150% with project) referred to the entire service area of 620,000 ha. The wet season irrigated area was expected to increase from 446,000 ha to 483,000 ha, and the dry season irrigated area from 421,000 ha to 446,000 ha. The latest data provided by NIA show that the - 20 -  appraisal figures were not entirely accurate: although irrigated cropping intensity was 140% in 1991, it had already declined to 134% in 1992, the appraisal year, and was the same in 1993, the year of project commencement. The latest data also show that the irrigated area in year 2000 was about 478,000 ha in the wet season and 450,000 ha in the dry season. These figures are almost the same as expected at appraisal. However, the irrigated cropping intensity in year 2000 at 144% was lower than that expected at appraisal, since the service area had increased to 646,000 ha without a commensurate increase in irrigated area. However, for the 17 NIS included in the system improvement component, the irrigated cropping intensity increased from 149% in without project situation (1992) to 167% by the end of 2000, and is expected to reach 194% at full development in year 2003 (compared to 164% expected at appraisal). The SAR estimated the annual incremental rice production at about 100,000 tons (93,000 tons, to be precise, converting the 144,000 tons of paddy at 65%). This was projected on the basis of the overall NIS service area of 620,000 ha. For the ICR, the incremental production at full development in year 2003 is projected at about 94,000 tons of rice (145,000 tons of paddy). This projection is based on the incremental production from the 17 NIS improved under the project. The production projection is about the same as at appraisal, since the yields of irrigated paddy are assumed to be higher and of rain-fed paddy lower than those at appraisal (see para. 7 below for a detailed discussion). If the appraisal yields were assumed, the production projection would be somewhat lower in the ICR than that in the SAR. 5. For the 18 NIS to be improved under the project, the SAR targeted 78,655 ha out of a service area of 95,944 ha. The improved area included both rehabilitated and incremental irrigated area, but the exact figures for the two were not shown. During implementation, as a result of exclusion of one NIS and parcellary mapping, the firmed up service area of 17 NIS was reduced to 86,913 ha. Further, after agreement with the August/September 1999 Bank supervision mission, improvement targets were revised upwards at 84,200 ha in both wet and dry seasons (97% of the firmed-up service area). The improvement targets were disaggregated into rehabilitated areas (existing irrigated areas where irrigation structures and/or water availability are improved as a result of the project), restored areas (area within the firmed-up service area which was not irrigated before or not irrigated for the past few years, and is now irrigated as a result of the project), and new or generated areas (new areas over and above the firmed-up service area which are brought under irrigation as a result of the project, and thus lead to a higher firmed-up service area). The targets for rehabilitated areas were about 61,580 ha in the wet season and 60,000 ha in the dry season, for restored areas about 22,470 ha in the wet season and 24,000 ha in the dry season, and for new/generated areas about 200 ha in both seasons. However, part of the areas restored in the early years of the project were due to other funds available to NIA for the 17 NIS, and the incremental area targeted for achievement through project funds alone is about 16,900 ha in the wet season and 17,700 ha in the dry season. The August/September 1999 Bank supervision mission also agreed with NIA that project contribution would be measured primarily in terms of restored and new/generated areas, and the ERR would be estimated on the basis of these areas only. Benefits from the existing areas, which were rehabilitated under the project, were to be regarded as additional benefits but not included in the ERR calculation. The rationale for this methodology was that restored and generated areas could be clearly ascribed to the project, whereas benefits in rehabilitated areas, which would be mainly in terms of yield increases, were difficult to ascribe to the project alone, since many other factors, including weather, could have affected the yields, and yields would have fluctuated from year to year because of those factors. 6. By the end of year 2000, about 73,000 ha in the wet season and 72,000 ha in the dry season had been improved in the 17 selected NIS. This represented about 92% of the appraisal expectation, and about 86% of the revised higher target. The achievement of rehabilitated and generated areas was 100% of the respective revised targets. However, the restored area was about 11,700 ha in the wet season (52% of the revised target) and 12,200 ha in the dry season (51% of the revised target). The relatively low achievement of restored areas was, in general, due to the delay in the start of improvement works, and, in -21 -  particular, due to the envisaged introduction of improved structures, which were completely new in the irrigation sector in the Philippines (long-crested weirs for water level control and proportional dividers and reduced pipe sizes for modality of operation), and which started only in 1999 after a time-consuming process of consultation and agreement with the lAs and after technical trials and errors. The remaining restored areas are expected to be fully irrigated by early 2003 as tertiary works and farm ditches are completed by farners. C. Major Assumptions for Financial and Economic Analysis 7. Project benefits were computed on the basis of the incremental irrigated areas only. Estimates of outputs in the irrigated areas covered by the project were based on the survey of beneficiaries in all the 17 NIS conducted for the Borrower's ICR by consultants. Production costs for rain-fed areas were based on the 1999 Bureau of Agricultural Statistics (BAS) average regional data converted to 2000 prices using the consumer price index (CPI). The average yield (weighted by service area) of palay (paddy) for the rehabilitated, restored and generated areas in the 17 improved NIS, as reported in the survey carried out by consultants for the Borrower's ICR, was 3.92 t/ha in the wet season and 4.39 tlha in the dry season (see table below). This was 2% lower in the wet season and about 10% higher in the dry season than the yield levels estimated in the SAR (4 t/ha for both seasons). As expected, these yield levels were relatively higher than the average of 3.48 t/ha for irrigated areas reported by BAS for the year 2000, since BAS figures are the average of IOSP II and non-IOSP II areas, and the latter generally did not have the improved facilities provided to the 17 selected NIS. The survey yield figures show a huge variation among the 17 NIS, from 2.52 tlha in the dry season in Pongso and 2.62 t/ha in the wet season in Barit to 5.40 t/ha in the dry season in MRIIS District 3. The weighted average yield is heavily influenced by relatively high yields in the three Districts of MRIIS, which account for about 75% of the service area. The higher yields in MRIIS are mainly due to the fact that it is a reservoir-backed system with more stable water supplies, compared to the other NIS under IOSP II, which are run-of-the-river systems (except for four pump systems with a relatively small total service area) with highly variable water supplies from year to year. The relatively high yields in MRIIS for the year 2000, reported in the survey, are consistent with those reported by NIA field offices for MRIIS over 1992-2000 (average 3.96 t/ha and 4.34 t/ha in the wet and dry season, respectively). There were also significant annual variations in yields of all 17 NIS from 1992 to 2000, as reported by NIA field offices, and as would be expected from the vagaries of weather and other natural phenomena in the Philippines. The average yield used for the financial and economic analysis (in the base case) is 3.92 t/ha in the wet season and 4.39 t/ha in the dry season, as reported in the consultant survey. This implies a wet season incremental yield over rain-fed yield (2.2 tlha, as reported by BAS for the year 2000) of 1.72 t/ha and dry season incremental yield of 4.39 t/ha (as there was no rain-fed crop in the dry season). - 22 -  Yield Per Hectare (kg) - Survey Data for Year 2000 Bonga 1 4,805.64 4,805.64 Bonga 2 4,517.34 4,535.51 Bonga 3 3,780.00 3,780.00 Cura 2,801.78 2,801.78 IAAPIS 2,750.00 3,681.08 Baggao 3,484.16 3,572.74 MRI1S District I 3,828.95 4,254.06 MRIIS District 3 4,528.51 5,396.83 MRIIS District 4 4,425.36 5,017.29 Agos 3,054.50 3,829.19 Matogdon 4,473.25 4,076.62 Barit 2,620.04 3,135.87 Sta. Maria 3,081.11 4,228.42 Pongso 3,390.49 2,517.64 Maranding 3,807.51 4,065.12 Malasila 4,573.61 4,049.76 Cantingas 2,843.71 3,012.70 Weighted Average 3,920.29 4,387.46 8. The above reported yields in year 2000 are lower than the potential with good irrigation and drainage services, management, and adequate agricultural support services. However, for the financial and economic analyses, no further increases have been assumed over the 30-year period of analysis. Further, a sensitivity analysis for the ERR was undertaken, assuming an average yield of only 3.75 t/ha in both the wet and dry seasons over a 30-year period of analysis (the yield figure of 3.75 t/ha was assumed in the OED Performance Audit Report No. 18034 of June 19, 1998 for the preceding IOSP - Loan 2948-PH). Similarly, in the without project situation, there could be a decline in irrigated areas in both seasons due to further system deterioration. This possible decline has also not been taken into account in the economic analysis. D. Financial Analysis 9. Farmers' Income. The financial attractiveness of the project to fanners was evaluated using crop budgets and farrn models. To determine the impact of the project on the individual farrners, detailed financial crop budgets were prepared and typical farm models were analysed. The results of the survey by the consultants, who prepared the Borrower's ICR, show that the majority of the farmers in project areas are owner-operators, representing 64% of farmers, with an average holding of 1.5 ha. The rest are predominantly share tenants, cultivating about 1 ha per household. The most common arrangement for the tenants was equal sharing of the net harvest after deducting the expenses for inputs and planting costs. Based on crop budgets, and two indicative farm models with average farm size of 1 ha and 1.5 ha under different types of tenure (share-tenant and owner-operator, respectively), the full development annual incomes of farmer beneficiaries estimated at appraisal and completion are summarised below: - 23 -  Annual Income Projections for Farmers (Peso '000) , ICR with :roject X *.R with projec' 3 Without prjec 1 ha share-tenant 21.6 19.1 18.1 1.5 ha owner-operator 66.1 57.3 54.6 ** Adjusted to 2000 price using CPI At full development of irrigated areas, which is to be achieved by early 2003, net returns for a one hectare share-tenant and 1.5 ha owner-operator would have increased by 19% and 21%, respectively, compared to the without project situation. Assuming (as at appraisal) 82 person-days/ha required for farmning in irrigated areas, net returns per person-day of family labor per ha would be Peso 538 for an owner-operator and Peso 263 for a share-tenant. These returns are very attractive, compared to the prevailing rural wage rate of Peso 60-100/person-day. The higher farmer's income now projected at full development compared to that at appraisal is due to the higher cropping intensity (194%), higher farm-gate prices, and higher yields (4.39 ton/ha) for irrigated areas in the dry season, compared to the cropping intensity, farm-gate price, and yield estimated in the SAR. However, the net return figures in the ICR are not comparable to those in the SAR, because the SAR figures were calculated on the basis of an overall irrigated cropping intensity of 150% and average yields of 4 ton/ha over the entire project service area of 620,000, whereas the ICR estimates are based on the irrigated cropping intensity and yields on the 17 NIS irnproved under the project. 10. Poverty reduction impact. As mentioned above, most of the farmers benefiting from the project are smallholders/share-tenants. Project benefits, as reflected in increased farm incomes, would, therefore, have direct poverty alleviation effects. E. Economic Analysis 11. At appraisal, the ERR was calculated for the system improvement sub-component on the basis of costs and benefits of five NIS where improvement works were to commence in the first project year. The ERR for the sub-component was 19%. For the urgent repairs sub-component, the ERR (28%) was calculated for 20 structures in 14 NIS. The overall ERR for the project was 17%. At completion, the ERR for the system improvement sub-component covers all the 17 NIS under the project. For the urgent repairs sub-component, the ERR (calculated in the Borrower's ICR) covers 13 of the 17 NIS repaired. 12. Investment costs. Project costs used in the analysis were based on actual costs incurred. All costs were included in the analysis except incremental O&M costs. For the incremental O&M component, which accounts for 36% of total costs at completion, the NPV and ERR were not recalculated since, as stated in the SAR, the component was designed to maintain, and in fact did maintain, the O&M expenditures achieved under IOSP I in 1992 constant terms, and its ERR is, therefore, the same as for IOSP I (the ERR for IOSP I was estimated in the SAR for IOSP II at 21%, but was re-estimated at 28-35% in the subsequent OED Performance Audit Report on the project). For the economic analysis of IOSP II at completion, costs in current Pesos were converted to 2000 prices by applying the Consumer Price Index to the local costs and the MUV Index to the foreign costs. The foreign exchange component was estimated at 20% of the total costs. Project investment costs in constant financial prices were converted into economic prices by applying the Standard Conversion Factor (SCF) of 1.0 to local costs. The average official exchange rate of Peso 50=US$1, which prevailed from late 2000 to mid-2001, was used in the base case ERR. However, since there has been a considerable exchange rate volatility over the last few years, and the average exchange rate in year 2000 was Peso 44=US$1.00, a sensitivity analysis - 24 -  was also undertaken at that exchange rate with a SCF of 1.0. At appraisal, the SCF used was 0.83 (or a shadow exchange rate 20% above the market exchange rate), and the opportunity cost of unskilled labor was estimated at 60% of the prevailing market wage rate. At completion, the SCF applied to the financial labor costs (both hired and family labor) is 1.0, since there appear to be no significant distortions in the rural labor market. 13. O&M and Other Recurrent Costs for 17 NIS. For the restored and generated areas included in economic analysis, an 0 & M cost of Peso 1,500/ha, which is the estimated actual for year 2000, was used through the year 2001. From the year 2002 onwards, the desirable O&M expenditure level of Peso 2,300/ha, as estimated in a recent consultant study, was used. Economic costs of trade-able farm inputs were valued at border (parity) prices. For economic costs of non-tradeable farm inputs and labor, an SCF of 1.0 was used. Savings in cost of operation for pump systems, which were included in the ERR calculation in the Borrower's ICR, were not considered in the Bank's ERR calculations. 14. Benefits. Project benefits, as mentioned earlier, are based on the incremental irrigated (restored and generated) area in the 17 NIS. The areas restored due to other funding sources were excluded, and only areas attributable to project funding were included in the economic analysis (16,900 ha and 17,700 ha in the wet and dry season, respectively, at full development). The economic farm-gate price of paddy is based on import parity. In the Borrower's ICR, the ERR for the urgent repairs sub-component was calculated separately, and the benefits included were: reduction of future repair costs; restoration and increase of irrigated areas; prevention of loss of irrigated areas; savings in O&M costs; prevention of health hazards; and prevented upstream flooding and facilitated releases downstream. 15. Re-estimation of the ERR. The ERR of the project was based on all project costs (except that of the incremental O&M component) and incremental production benefits in restored and generated areas in 17 NIS. The ERR was re-estimated at 21% and NPV at Peso 1,251 million at a discount rate of 10%, as compared with the ERR of 17% in the SAR. The ERR of 21% is conservative, since no benefits of rehabilitation and urgent repairs were included in the economic analysis, even though the corresponding costs were included. The Borrower's ICR also estimates ERRs separately for the systems improvement sub-component (about 26%) and the urgent repairs sub-component (about 18%, compared to 28% estimated at appraisal), and individual ERRs for the 17 NIS with system improvement and 13 NIS with urgent repairs. The individual ERRs for the system improvement sub-component range from a low of about 7% for Bonga Pump #3 and Cura to a high of 115% for Malasila. Of the 17 NIS, only three had ERRs of less than 10%. For the urgent repairs sub-component, the individual ERRs range from a low of 4% for the Pagbahan intake to a high of 63% for the M'lang Diversion Dam. Of the 13 NIS analysed, only three had ERRs of less than 10%. The overall ERR for the urgent repairs sub-component (18%) was lower than that at appraisal (28%), principally because delayed and staggered fund releases led to increases in costs, slower implementation and the consequent reduction in actual and potential benefits. The details of individual system ERRs are in Annexes 7 and 11 of the Borrower's ICR in Project Files. 16. A relatively high ERR is to be expected for projects of this nature where incremental benefits are generated at a relatively low cost through improvement and rehabilitation. The cost of an incremental irrigated ha under the project at full development is estimated at only Peso 75,000 or US$1,500 (the total cost of system improvement works and associated costs of agricultural support services, IA/CIA development, technical assistance, parcellary mapping, NIA staff training, and equipment and materials, was Peso 1.279 billion in 2000 constant terms, and the expected full development incremental irrigated area in year 2003 is about 17,000 ha). Moreover, the steep exchange rate depreciation over the last few years has made import substitution of commodities like rice economically very attractive. The ERR at completion is also higher than at appraisal due to the following major factors: (i) the irrigated cropping - 25 -  intensity at full development in the 17 NIS is projected to be higher than that estimated in the SAR; (ii) yield in the irrigated areas in the dry season is assumed to be higher than that in the SAR; and (iii) the economic farmgate price of palay (paddy), which was projected at Peso 3.78/kg for year 2000 in 1992 constant terms (Peso 6.8/kg in 2000 constant terms) in the SAR, was estimated at Peso 8.7/kg or about 28% higher in real terms at end-2000, principally resulting from the steep exchange rate depreciation over the last few years (from Peso 25.5=USS1 in December 1992 to an average of Peso 44 in year 2000, and further to Peso 50 in the first half of 2001). 17. Sensitivity Analysis. Several sensitivity analyses were undertaken. (i) At an exchange rate of Peso 44=-US$1 and an SCF of 1.0 for local costs, the ERR would decline to 17% and the NPV at 10% discount rate to Peso 710 million. (ii) At an assumed paddy yield in irrigated areas averaging only 3.75 t!ha over the 30-year period of analysis, the ERR would decline to 16% and the NPV at 10% discount rate to Peso 642 million. (iii) If only 50% of the targeted incremental irrigated area is achieved at full development, the ERR would be only 12% and the NPV at 10% discount rate only Peso 163 million. (iv) If the incremental irrigated area starts declining at 3% per annum from year 2001 onwards due to inadequate O&M (historically, the decline has been much less even with inadequate O&M), the ERR would remain at 21%, although the NPV at 10% discount rate would decline a little to Peso 1,146 million. (v) With a 20% decline in world rice prices, compared to the currently projected level of US$250-270/ton over 2000-2010 in 2000 constant terms, the ERR would be 14% and NPV Peso 407 million at a discount rate of 10%. The major sensitivity results are shown below: With paddy With 50% achievement of full With irrigated With world rice price decline of yield of 3.75 development irrigated area area declining at 20% t/ha target 3% p.a. 16% 12% 21% 14% The sensitivity results show that the project is economically viable and relatively robust, although quite sensitive to the level of achievement of full development targets for incremental irrigated area and to world rice prices. - 26 -  Annex 4. Bank Inputs (a) Missions: _ Stage of Project Cycle No. of Persons and Specialty Performnce Rating (e.g. 2 Economists, I FMS, etc.) Implementation Development MontvYear Count Specialty Progress Objective Identification/Preparation Aug/1991 4 C(2), B, E Nov/1991 2 C(2) May/1992 2 C, B Appraisal/Negotiation Pre-Appr: 4 B(2), C, E Sepl 1992 Appraisal: 6 A, B(2), C(2), E Dec/1992 Supervision Jul/1993 I C S S Jan/1994 1 C S S Jun/1994 I C S S Feb/Mar/1995 4 A, B, C, D S S Mar/1996 3 A, B, C S S May/1996 I C S S MTR - 3 B, C, E S S Mar/Apr/1997 Jun/1997 2 B, C S S Oct/Nov/1997 3 B, C, G S S Mar/1998 2 B, C S S AuglSep/1998 5 B(2), C, F, G S S Mar/1999 2 B, C S S Sep/1999 3 B, C, D S S Mar/2000 3 B, C, D S S Sep/2000 4 B, D, E, F S S ICR Feb/2001 3 A, B, C Specialty: A: Agronomist B: Economist C: Engineer D: Financial Analyst E: Institutional Development Specialist F: Procurement Specialist G: Operations Officer - 27 -  (b) Staff | Stage of Project Cycle Actual/Latest Estimate No. Staff weeks US$ ('000) Identification/Preparation 110.00 243.9 Appraisal/Negotiation 33.10 80.4 Supervision 77.01 224.6 ICR 10.00 60.0 Total 230.11 608.90 Above includes Bank-financed and Trust Fund consultants. - 28 -  Annex 5. Ratings for Achievement of Objectives/Outputs of Components (H=High, SU=Substantial, M=Modest, N=Negligible, NA=Not Applicable) Rating O Macro policies O H OSUOM O N * NA [II Sector Policies O H OSUOM O N * NA O Physical O H *SUOM O N O NA O Financial O H OSUOM * N O NA n Institutional Development 0 H 0 SU * M 0 N 0 NA O Environmental O H OSUOM O N * NA Social El Poverty Reduction O H O SU * M O N O NA El Gender O H OSUOM * N O NA M Other (Please specify) * H OSUOM O N O NA Stakeholder participation O Private sector development 0 H O SU O M 0 N 0 NA n Public sector managementH 0 0 SU 0 M 0 N 0 NA C0 Other (Please specify) O H OSUOM O N O NA - 29 -  Annex 6. Ratings of Bank and Borrower Performance (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HU=Highly Unsatisfactory) 6.1 Bank perfornance Rating 0- Lending OHS OS OU OHU E Supervision OHS OS OU OHU Q Overall OHS OS O U O HU 6.2 Borrowerperformance Rating Q Preparation O HS * S OU O HU E Government implementation performance 0 HS O S 0 U 0 HU O Implementation agency performance O HS * S O u O HU Q Overall OHS OS 0 U O HU - 30 -  Annex 7. List of Supporting Documents Back-to-Office Report of the ICR Mission (April 9, 2001) Aide-Memoire of the ICR Mission (February 26, 200 1) Borrower's ICR and Annexes (March 30, 2001) Financial and Economic Analyses QAG Report on the Third (FY99) Rapid Supervision Assessment - 31 -  Additional Annex 8. Bene ficiary Survey Results 1. Although a beneficiary survey is not required for Core Accountability ICRs, a survey of 17 NIS improved under the project was carried out by the consultants as a part of preparation of the Borrower's ICR. The survey had respondents from CIA/IA officers, farmers and NIA staff. The detailed results of the survey are in Annex 9 of the Borrower's ICR in Project Files. Given the significance of the survey results for a proper understanding of project impact, a summary of the major findings relating to perceptions of CLIA officers and farmers is provided below. 2. A pre-determined sample of 1,400 farmers was taken from the 17 NIS improved under IOSP II. The sample was further stratified into farmers in IMT areas (those CIA/IAs who were under contract as of 1999; the year 1999 was chosen so that two cropping seasons after contract signing could be captured) and farmers in areas with IOSP II interventions but without IMT contract. Within the lAs, sample farmers were further stratified as farmers within the rehabilitated area and farmers within the restored and generated areas. The percentage sample was selected at random within the sub-stratum. Major Survey Findings 1) CIA/IA Officers a) Knowledge and Perceptions of IOSP Il/IMT Modality of Operation 3. When asked whether they were farniliar with how the modality of operation installed in their system operated, a big majority of 341 (97%) said yes. In terms of satisfaction of the members of the LA/CIA with irrigation service performance as a result of the new modality of operation, 251 (74%) gave a satisfaction rating of 7-10. The main reason given was that the system was functioning well and majority of members were satisfied, as mentioned by 136 (40%) of the respondents. Seventy one (21%) respondents gave a rating of 4-6 while 17 (5%) did not respond to the question. Among the reasons given for the lower ratings were: (a) unequal water distribution; (b) not all structures/facilities installed; (c) unrepaired damaged canals and structures; and (d) some farms not irrigated during the dry season (Tables I & la). - 32 -  Table 1. Knowledge and Perception of the Modality of Operation Installed 1. Knowledge of how the modality operates Yes 341 97 No 9 3 No answer 3 1 Total 353 100 2. Are you and the members of the LAJCIA satisfied with the modality? Yes 341 97 No 12 3 Total 353 100 If yes, rating given: I -3 2 1 4 - 6 71 21 7 -10 251 74 No answer 17 5 Total 341 100 Table la. Reasons for the Satisfaction Rating on Modality of Operation 1. System is fanctioning well and majority of members are satisfied 136 40 2. Adequate water is provided at the right time 19 6 3. Un-repaired damaged canals and structures 17 5 4. Not all structures/facilities have been installed 23 7 5. Some farms are not irrigated, especially during dry season 13 4 5. Unequal water distribution 32 9 6. Late water delivery 4 1 7. Target ISF collection of 90-100% not achieved 16 5 8. Some members are uncooperative and cannot accept the new system 7 2 9. Farm located downstream 16 5 10. Farm located midstream 12 4 11. Noanswer 46 13 Total 341 100 Impact of IOSP II/IMT on Irrigation Services 4. Table 2 presents the perceptions on the irrigation system performance before and after IOSP II in terms of timeliness, equity of water distribution and adequacy. An almost equal number of respondents - 33 -  indicated that timeliness of water delivery and adequacy of irrigation water had improved significantly. In terms of equity of water distribution, more than half (52%) of the respondents thought that with IOSP II, there was more equitable water distribution. Table 2. Impact of IOSP [ LMT on Irrigation Services 1. Timeliness Just right 130 37 235 67 Delayed 193 55 58 16 Advance 14 4 44 12 No answer 146 41 16 5 Total 353 100 353 100 2. Equity of Distribution Equitable 88 25 182 52 Just right 93 26 120 34 Not equitable 149 42 33 9 No answer 23 7 18 5 Total 353 100 353 100 3. Adequacy Adequate 134 38 236 67 Not adequate 167 47 86 24 More than adequate 29 8 19 5 No answer 23 7 12 3 Total 353 100 353 100 b) Institutional Development and Strengthening CIA/IA Strenjthenina Activities Implemented 5. The real challenge of IMT is the development of strong and self-reliant Irrigators' Associations with the capability to manage the irrigation system on a sustainable basis. The activities undertaken by NIA in preparation for CIA organization, and participation by CLA/IA officers and their perceptions, are detailed in Table 3. - 34 -  Table 3. Activities Undertaken for CIA Strengthening 1 Situation Analysis 242 69 18 5 190 54 7 2 2. Consultative Workshops 281 80 5 I 211 60 5 1 3. Information/Dissemination re: IMT 287 81 5 1 225 64 7 2 4. Setting of CLA/IA Vision, Mission, Objectives 265 75 12 3 204 58 5 1 5. Establislmuent of CIA/IA CBL 213 60 22 6 158 45 4 1 6. Preparation of CIA/IA O&M Plan and Budget 253 72 18 186 53 4 1 7. Assessment of Irrigation Structures 225 64 20 6 167 47 3 1 8. Training 250 71 8 2 184 52 8 2 Total 353 * 353 * 353 * 353 * *There were multiple responses 6. Training. Majority of the respondents attended one or more training conducted by NIA or by other government agencies such as the Department of Agriculture. Financial Management and Basic Leadership training was attended by 196 (5 6%) and 180 (51%) respondents, respectively. Table 4. Training Attended, Adequacy of Training, and Adoption of Lessons Learned Basic Leadership 180 5 1 1 |0 164 46| l Systems Management 73 21 2| 1 70 20 11| 0 Financial Management 196 56 4 l 185 l 52l 3 11 Operations and Management 25 7 22 6 Gender 45 13 40 11 Value Formation 62 18 56 16 Irrigation Management 99 28 2 1 90 25 Water Management 8 2 8 2 PMES 12 3 12 3 IPM 49 14 2 1 39 11 Rice Production 25 7 25 7 Modem Farming 15 4 12 3 Others ** 23 7 23 7 Total 353 * 353 * 353 * 353 * * There were multiple responses. ** Others include: Aquaculture, crop protection, cooperative, livelihood, grains classification, etc. - 35 -  c) ISF Collection and Payment 7. When asked whether the CIA/IA management believed that all the O&M expenses had to be recovered to be able to continue providing irrigation service, 85% said yes (Table 5). Only 12 (2%) respondents believed otherwise, while 42 (12%) did not give any answer. 8. Of the officers interviewed, 330 (93%) claimed that they paid their ISF voluntarily. They added that among the members, 55% voluntarily paid ISF, 37% waited for the collector and 17% did not pay ISF at all. A big majority of the respondents did not answer this question. These percentages, as reported by the CIA/IA officers, closely coincide with the responses obtained from the farmers where 75% reported that they paid the ISF and 14% said that they did not pay ISF. The reasons given for non-payment of ISF were: low yield, sale from harvest not enough for family needs, insufficient water and late water delivery, among others. 9. The ISF collected by the IA is remitted to the CIA who immediately remits the collection to the NIA. However, as reported by 64% of the respondents, NIA is not able to give the CIA share of the ISF even if all the required documents have been complied with. This situation seemed to have caused frustration and distrust among farmers in relation to NIA's sincerity in pursuing its commitment to the ClA/IA as expressed in the Memorandum of Agreement. 10. Collection rate improved after IMT according to 226 (68%) respondents (Table Sa). Fifty four (16%) did not think that there was any increase in ISF collection. AO 17, which was issued by the Philippine President in mid-1998, was identified by 33% as among the causes of the non-improvement in the ISF collection even after IMT. AO 17 is the adoption of a socialized irrigation service fee, which established new ISF rates based on farm size. Non-payment of ISF by some farmers is another reason cited by 19% of the respondents who claimed that ISF collection did not improve. 11. To improve ISF collection, a number of ways have been devised by the CIAIA. Foremost among the assistance provided by the officers is an intensive collection drive, as indicated by 159 (69%) respondents (Table 5b). Information dissemination on the importance of payment, and collection by group, were cited by 48 (21%) and 38 (17%) respondents, respectively. Another 16 respondents (7%) mentioned good irrigation service as a means of improving collection. Sharing Arrangaements 12. Less than half (47%) of the respondents expressed satisfaction with the present sharing arrangement for ISF, while 36% said they were not satisfied (Table 5). The reasons given for the dissatisfaction was the inability of NIA to give the CIA share immediately (35%) and the misunderstanding of the sharing basis indicated in the Memorandum of Agreement signed by both parties. - 36 -  Table 5. ISF Collection Does the CIA/IA Management believe that, to be able to continue irrigation service, all the O&M expenses have to be recovered? Yes 299 85 No 12 3 No answer 42 12 Total 353 100 Do you voluntarily pay your ISF? Yes 330 93 No 4 1 No answer 19 5 Total 353 100 What percent of the members: a.) Voluntarily pays the ISF 55 17 b.) Waits for the collector 37 11 c.) Does not pay ISF 17 5 d.) No answer 221 67 Total 330 100 Was the ISF collection remitted immediately to CIA or NIA? Yes 284 86 No 7 2 No answer 39 12 Total 330 100 Was the CIA/IA share given back by NIA? Yes 70 21 No 210 64 No answer 50 15 Total 330 100 Are you satisfied with the sharing arrangement? Yes 155 47 No 118 36 No answer 57 17 Total 330 100 If no, what do you suggest? 50-50 from total cash collection as under - stood from old MOA 35 30 IA share must be given back immediately 35 30 Give honorarium 1 1 Completion of canal lining 1 1 No answer 46 39 Total 118 100 - 37 -  d) Operation Adequacy of Water Supply 13. A majority (84%) of the respondents reported that at least 88% of members received adequate irrigation water. On the other hand, 52% said that about 16% of members did not get adequate supply of water in their farms. It was also claimed that fanns of nine percent (9%) of members did not receive any water at all. It may be because some farms are in high locations or are located downstream. It may also be because efficiency of irrigation water delivery through the lateral decreased in some areas since some structures remained incomplete or were damaged (1Table 6). Table 6. Adequacy of Water Supply Are all members given adequate water? - With adequate supply 295 84 88 - With inadequate supply 182 52 1 6 - With no w 150 42 9 Total 353 Why are there members not receiving enough water? - Farm located upstream 42 33 - Incomplete/damaged structure 27 21 - Farm located downstream 25 20 - Inadequate water 9 7 - No discipline among farmers 4 3 - No farm ditch 4 3 - Existence of illegal turn-out 3 2 - Canal not in proper level 2 2 - Irrigation water overflows 2 2 - Drainage problem 2 2 - Wrong placement of turn out 2 2 - Shallow canal 1 I - Water was given to other CIS 1 1 - Because of reduced pipe size 1 I - El Nino/dry season 1 1 - Forced irrigation 1 1 Total 127 100 * There were multiple responses. Participation in Systems Operation 14. According to a big majority of respondents (65%), there is a System Management Committee (SMC) for the NIS, with the CLA/IA as a member. As to whether NIA involved the CLI/IA in the - 38 -  preparation of the Operation Plan, 77% said that they did participate in this activity. Water Distribution and Monitoring 15. According to a majority (77%) of the respondents, the CIA/IA was in-charge of water distribution and monitoring in the areas covered by the lateral. However, almost one-half (47%) of the respondents were not aware what measuring device was used in monitoring the volume of water available for distribution. On the other hand, 40% mentioned staff gauge as the most common measuring device used. In terms of being able to follow the water distribution schedule, 273 (77%) respondents said that the organization was able to adhere to the schedule (Table 7). Table 7. Adherence to Water Distribution Schedule Is the water distribution schedule being followed? Yes 273 77 No 29 8 No answer 51 14 Total 353 100 If no, why? Farm located downstream 5 17 Late repair 4 14 Damaged canal 3 10 Water availability 2 7 Wrong schedule 2 7 Damaged equipment 2 7 Blocked passage 1 3 Schedule mo na inaagaw pa 1 3 No answer 9 31 Total 29 100 e) Repair and Maintenance Maintenance Activities 16. Vegetative clearing and canal clearing are the two activities that are participated in by majority of the members with an average attendance of 74% and 71%, respectively, according to the respondents. A big number of members also participate in canal reshaping and repair of service road (Table 8). - 39 -  Table 8. IA Members' Participation in Maintenance Activities 2 ) I What are the maintenance activities participated in by members? - Vegetation clearing 221 63 74 - Repair/cleaning of farm 164 46 73 - Canal clearing 306 87 71 - Canal reshaping 155 44 67 - Repair of service road 132 37 67 - Desilting 161 46 63 Walk thru inspection 206 58 61 Total 353 * * * There were multiple responses. 2) Farmers a) Participation of Family Members in Farming Activities 17. When asked if there were members of the family who were involved in farmning activities, less than half or 47 percent of the respondents reported that at least one male member was helping full time in farming. On the other hand, only 9 percent reported that a female member was involved full time in farming (Table 9). This low involvement of family members in farming activities could be attributed to the fact that some of them were still studying or were working outside of the farm. This also explained the present dependence of farmers on hired labor. As expected, the survey also revealed that women are actively involved in food preparation for farm labor, particularly during planting season, as reported by 69 percent of the respondents. More than 30 percent of the respondents also reported that women are involved in choosing seeds/variety to plant, planting, drying and selling of palay. Other activities reported were harvesting, threshing and winnowing. - 40 -  Table 9. Family Members' Participation in Farming Activities (1399 Farmers, 17 NIS) Family Members Involved in Farming Activities Full time: Male 658 47 1.20 Female 140 9 1.17 Part time: Male 261 19 1.42 Female 363 26 1.16 Participation of Women in Farming Activities Use of seeds 482 34 Food preparation 967 69 Threshing 179 13 Drying 539 39 Vegetables planting 492 35 Planting palay 449 32 Harvesting 358 26 Winnowing 273 20 Selling 563 40 Others (weeding) 30 2 b) IA Membership and IMT Participation in IA Activities 18. IA farmer members participate in three major activities of the IA: planning, meetings and irrigation canal maintenance. Results of the survey showed that farmer members participated in IA planning activities as indicated by 75% of the respondents. Those who did not participate were either not aware of the activity (18%), busy (19%) or believed that planning activities were only for officers (23%). With regards to maintenance activities, the CIA/IA officers interviewed reported that vegetative clearing and canal clearing were the two activities that were participated in by majority of the members with an average attendance of 74% and 71%, respectively. A big number of members also participated in canal reshaping and repair of service road as gleaned from the responses of the CIA,IA respondents when asked the same question. Knowledge and Perception of IOSP II/IMT 19. A majority (70%) of the farmer respondents claimed that they knew what IOSP II/IMT was (Table 10). However, when asked what they knew, 21% did not give any answer while the responses given by 22% indicated that they were not completely familiar with the project. Responses included: no more illegal turnout, NIA project, increased production, training/meeting, etc. This would imply that information dissemination about IOSP II/IMT to the farmer members was not adequate. 20. Those, who said that they did not know what IOSP II/IMT was, were asked if they were aware of any improvements in irrigation facilities and managernent. A majority (69%) acknowledged that there had been improvements such as : better irrigation servize, repair and/or lining of canals, improved control points, etc. (Table 10). More than one-third were not aware of any improvement done while a small minority did not give any answer. - 41 -  Changes in the IA Due to IOSP II/IMT 21. Generally, in all of the 17 IOSP II systems, the farmer members indicated their satisfaction with the improved irrigation service, particularly in terms of reliable and adequate water supply brought about by system improvement (Table 10). However, some noted that system improvement had not been completed, and, according to them, this could bring about maintenance problems. Some noted that members had become more cooperative, active and responsible, although there were also those who said that one of the problems of the IA was the lack of cooperation among the members. Table 10. Knowledge and Perception of IOSP II/IMT 1. Do you know IOSP II/IMT Yes 977 70 No 422 30 Total 1399 100 If yes, what do you know about it? Transfer of management of irrigation to IA 398 41 Irrigation improvement, i.e. adequate water 158 16 No more illegal TO 109 11 NIA Project 48 5 Increased Production 18 2 Training, meeting 10 1 Road construction, service road 13 1 IA collects ISF 6 1 Seed Production 3 1 Financial support 4 1 Good rules and regulation 2 1 Maintenance of farm ditch I I No answer 207 21 Total 977 100 If no, do you know of any improvements done in your irrigation facilities/management? Yes 256 61 No 166 39 Total 422 100 If yes, what are these Repair of canalUcanal lining 161 33 Better irrigation service 203 41 Improved control points 68 14 Increased Production 6 1 No more illegal turn out operation 33 7 No answer 21 4 Total 492 100 Impressions On IMT 22. Varied responses were obtained when respondents were asked about their impression on IMT. More than half (55%) said that IMT was beneficial to farmers. This may be related to improved water -42 -  supply as a result of improvement of irrigation facilities as articulated by 11% of the respondents. Others looked at IMT as a channel by which farmers could become more responsible and cooperative. However, a number of respondents did not look at IMT positively but rather believed that IMT may bring about maintenance problems because of the non-completion of system improvement. A small percentage (3%) was not satisfied with IMT and suggested that this was difficult to manage. Still others thought that even with IMT, NIA needs to provide support to the farmers (Table 11). Table 11. Impressions on IMT WI, 1. Beneficial to farmers 769 55 2. Systems improvement brought about improved water supply 150 11 3. May bring about maintenance problem because of incomplete 41 3 systems improvement 4. Farmers will become more responsible and cooperative 32 2 5. Improvement in ISF collection 16 1 6. Fair ISF payment 7 1 7. Good rules and regulations 4 1 8. Still needs NIA's support 10 1 9. Not satisfied/difficult to manage 46 3 10. Easy to manage 6 1 I1. No answer 297 21 c) Impact of IOSP Il/IMT on Irrigation Service 23. The farmers were asked if they observed improvements in the irrigation service vis-a-vis timeliness of water delivery, equity of distribution and adequacy of water supply after project implementation. As shown in Table 12, improvements were observed by a significant number of respondents. With regards to timeliness of water delivery, farmers in areas without IMT contract reporting timely delivery increased from 29 percent before project to 69 percent after the project. Simnilarly, in IMT areas, 70 percent of the farmers reported satisfaction after the project as compared to only 44 percent who said they were satisfied before the project. For all areas on the average, 39 per cent of farmers reported timely and advanced delivery of water before the project and 77 percent after the project. 24. As regards equity of distribution, improvement was reported by farmers both with and without IMT contracts. Satisfied farrners increased from 45 percent to 76 percent in non-IMT areas and from 62 to 84 percent in IMT areas. Overall, the satisfaction before the project was reported by 53 percent of farmers and that after the project by 80 percent of farmers. 25. As to the adequacy of the water delivered to their farms, the number of respondents who said it was adequate/more than adequate increased from 40 percent before the project to 76 percent after the project and from 54 to 84 percent for non-IMT and IMT areas, respectively. Overall, the proportion of respondents reporting adequate to more than adequate water supply increased from 46 percent before the project to 80 percent after the project. 26. These observations of the farmer respondents were consistent with the observations of the officers of the CIAs and LAs. The officers reporting that there was a timely/advance delivery of water increased from 41 percent before the project to 79 percent after the project. Similarly, those who observed an equitable distribution of water increased from 41 percent before project to 86 percent after the project, while those reporting adequate water supply increased from 46 percent before the project to 72 percent - 43 -  after the project. 27. The observations reported above clearly demonstrate the view held by a majority of beneficiaries that there have been significant improvements in irrigation service brought about by the project. However, considering that there are still about 20 to 30 percent of the farmers who did not report having benefited from the project, there is a need to continue improvement of irrigation facilities and to have better information dissemination relating to project interventions. -44 -  Table 12. Impact of the Project on Irrigation Service Areas without IMT Contract _ 1. Timeliness of Water Delivery . Just Right 235 29 552 69 Delayed 513 64 170 21 Advance 22 3 55 7 2. Equt of Distribution_ Equiitable 161 20 356 44 _Just Right 205 25 258 32 Not Equitable 396 49 131 16 3. Adequacy Adequate 256 32 555 69 Not Adequate 442 55 165 21 More than Adequate 66 8 59 7 Areas with IMT Contract 1. Timeliness of Water Delivery Just Right 263 44 414 70 Delayed 287 48 108 18 Advance 25 4 55 9 2. Equity of Distribution Equitable 214 36 343 58 Just Right 154 26 157 26 Not Equitable 200 34 73 12 3. Adequacy Adequate 306 51 481 81 Not Adequate 246 41 84 14 More than Adequate 15 3 18 3 All Areas 1. Timeliness of Water Delivery Just Right 498 36 966 69 Delayed _ _ 800 57 278 20 Advance 47 3 110 8 2. Equity of Distribution Equitable 375 27 699 50 Just Right ___ 359 26 415 30 Not Equitable 596 43 204 15 3. Adequacy Adequate 562 40 1036 74 Not Ad e 688 49 249 18 More than Adequate 81 6 77 6 - 45 -  Additional Annex 9. Bo rrower's ICR Borrower's Implementation Completion Report On The Second Irrigation Operations Support Project (IOSP II) (Prepared by NIA) 1. Assessment of Development Objective and Design, and Quality at Entry 1.1 Original Objective: The Project's primary objective is to improve and sustain the operational efficiency of the National Irrigation Systems (NISs) thereby, help increase agricultural production (mainly rice), expand small farmer incomes and rural employment opportunities and contribute to rural poverty alleviation. Specifically, the project aims to: 1) undertake intensive improvements for priority NISs, urgent repairs, improved water and erosion control measures; 2) support the improved system-level O&M services achieved under IOSP-I; 3) institutional development through: IA organization, training and development, strengthening of NIA through improvements in collection process, staff trainings on O&M and appropriate engineering design techniques, and provision of high priority O&M equipment and materials; and 4) improved agricultural support services. A distinctive feature of the project is its appraisal process. Emphasis was placed on participation of beneficiaries in identifying irrigation improvement requirements and increasing responsibility for O&M of the portions of the system serving them. The objectives were clear, realistic and in line with the Bank's Country Assistance Strategy (CAS) of restoring economic growth, while reducing poverty and improving equity, through accelerated and environmentally sustainable rural development and support to infrastructure. The objectives were likewise consistent with the Government's objectives for the agricultural sector of: (a) increased productivity and real incomes of small farming families, (b) ensuring productivity of the agricultural resource base, and (c) attaining self-sufficiency in rice and com for food security. 1.2 Revised Objective The project objectives remained unchanged during implementation. 1.3 Original Components: The project has four (4) major components: (a) System Improvement; (b) Incremental O&M; (c) Institutional Development; and (d) Agricultural Support Services 1.4 Revised Components: The project design remained unchanged throughout implementation. However, changes were made on specific NISs covered, for example, Hanagdong RIS was excluded in the coverage of NISs for system improvement in view of duplication of donor Agency for civil works in adjacent scheme. This, therefore, reduced to 17 the number of systems covered by IOSP HI. Additional three (3) NISs were covered by urgent repairs. These changes duly passed the approval process. 1.5 Quality at Entry: Quality at entry is rated satisfactory in view of clear and realistic project objectives and consistency with the Bank's CAS and GOP priorities for the sub-sector. Project design was coherent, incorporating lessons from the preceding Irrigation Operations Support Project (IOSP I, Loan 2948-PH), and was consistent with the capability of the implementing Agency. - 46 -  2. Achievement of Objective and Outputs 2.1 Outcome/achievement of objective: The outcome of the project in terms of its contribution to the Governnent's goals of poverty alleviation and food security is rated satisfactory. Overall, the project is estimated to have benefited about 504,850 farm families (110% of the SAR estimate of 460,000 families to be benefited within the systems included in IOSP II), with about 45,294 families benefiting fiom rehabilitated areas, 10,710 from restored and generated areas in the 17 systems and the rest from systems with incremental O&M. The incremental rice production under the project is about 66,101 tons/year or about 66% of the appraisal expectation of 100,000 tons/year. In terms of net revenue per hectare of palay production per year a farm family generated an average revenue of P28, 756 at an average cropping intensity of 163%. For the restored and generated areas, and factoring/considering cropping intensity, a total yearly net income of P27, 899 was realized. On the other hand, P27,570 was obtained from rehabilitated areas-P13, 570 and P20, 138 for the wet and dry seasons, respectively, which is still higher than the estimated net revenue at project preparation. The project also provided support aimed at increasing IA participation in identification of improvement needs, as well as increased level of participation in O&M. lAs already organized prior to the project were mobilized for participatory identification of improveiment works, implementation of works at secondary level and assumption of O&M responsibility to sustain improvement works. 2.2 Output by components: Systems Improvement and Repair. The output for civil works is rated satisfactory. The SAR envisaged intensive improvement of 18 NISs covering 95,782 ha but was reduced to 86,913 ha after firming up of SA and exclusion of Hanagdong RIS from the original IOSP II coverage. Based on the survey, improvements in the irrigation service vis-a-vis timeliness of water dlelivery, equity of distribution and adequacy of water supply was observed by a significant number of respondents. For all areas surveyed, farmers reporting timely and advanced delivery of water increased from 39 percent before the Project to 77 percent after the project. Similarly, CLA/IA officers reporting timely delivery of water increased from 41% before the project to 79% after the project. In addition, officers who observed an equitable distribution of water increased from 41% before the project to 86% after the project. These findings revealed that there were significant improvements in irrigation service brought about by the project. At completion, the total irrigated areas during wet season was 69,509 compared to 82,958 ha expected at appraisal, and 72,068 ha compared to 83,725 at appraisal stage during dry season for an average of 85% accomplishment. Cropping intensity nationwide decreased from 140 in 1992 to 138% in 2000. However, if we consider only the 17 NISs with improvements, the cropping intensity increased from 149 in 1992 to 163% in year 2000. Activities related to improved modality of operations wvere planned in order to improve equitability of water delivery and distribution. This led to the technical change in irrigation operation through adoption of water control structures that are easy to operate and require less labor. Proportional weir, modified pipe sizes, or a combination of the two (2) schemes have been tested in the 17 systems. As of ICR preparation date, an assessment of pilot areas for modality of operations adopting the different schemes showed initial successes on equitable water distribution. It was observed that there was a shorter required time for land preparation and wider areas irrigated. For pump irrigation system, a low energy cost/hectare was also noted. The operation, however, would need further attention on optimization of water supply through timely planting calendar scheduling and institutional measures for sustainability. Both the NIA field staff and IAs need to - 4,7 -  fully understand the interrelationships of technical adjustments and irrigation operation, and the necessity of strict maintenance of farm level facilities to ensure wider irrigated area (as in the case of IAAPIS). The project envisioned to provide for urgent repairs of 22 river diversion structures in 14 National Irrigation Systems (NISs). This involved repair of major structures which when not attended to will compromise stability of the structures and may ultimately result to their collapse. Aside from preventing total damage of structures, the intervention was deemed to prevent more costly repairs and to increase irrigated areas. At project completion, the component has accomplished repair/restoration works of 28 structures in 17 NISs despite the 3-year delay in the initial release of fumds as programmed. Works included repairs of diversion dams, sluice gates, protection works, provision of sluiceway and river protection works, and other interventions to restore and/or protect the major structures. A total of P1264.5 M was expended, representing 122% of total original SAR allocation of P1036.9. Changes in allocations were mainly due to force majeure that affected the initially accomplished works in the course of implementation, complemented with the intensification of the degree of physical damage before the repairs are completed as well as in delays in the flow of funds. A prominent example of this case is the Balanac RIS with a dramatic increase of POWs from P6.0 M when the substantially completed works were washed out by unusually strong flash floods in November, 1995. Among other NISs which have substantial changes in Programs were Cabadbaran, Pagbahan and Magasawang tubig. The interventions facilitated the restoration of the structures thereby caused savings in costs for future repairs, restored irrigated areas in some NISs and prevented loss of irrigated areas in other NISs. Economic analyses of 13 NISs show EIRR of ten (10) NISs ranging from 11.56% to 53.35%. In three NISs, however, EIRR fell below 10%. Overall, however, economic benefits prove to be satisfactory as compared to a scenario where repair of structure should have been delayed by 3 years and/or structures would have been totally damaged if repair works were not implemented. Rating of the sub-component is satisfactory in view of the actual interventions provided and the number of NISs covered and works done. Urgent repair works must be given more emphasis so as not to compound the damage. The project gave more emphasis on Incremental O&M but in the occasion of limited funds, this component should give way to urgent repair which could give more benefits on specific NIS basis, and would prevent higher cost of interventions. Incremental O&M. The rating is satisfactory. In all NISs the project supported the incremental O&M activities which included canal maintenance (desilting and repair of embankment), maintenance of service roads, control gates and structures, and incremental personnel costs (particularly IDOs). It is also envisioned to reduce the gap between O&M expenditures and revenues from ISF collections by turning over NIS sections to LAs and, thus, increase their participation in O&M, broaden the base for ISF collection, and improve ISF collection. The project succeeded in broadening the base for ISF collection and increasing IA participation particularly in the 17 NIS with system improvement. However, the nationwide target collection of 70% was not attained (actual CE was 67% in CY 2000) but this was because of factors outside the implementing Agency. Government pronouncement on non-payment of ISF during the State of the Nation Address of the President in July 1998 and the succeeding issuance of Administrative Order No. 17 (Socialized ISF rates) in September 1998 dampened the farmers' willingness to pay ISF and caused a drastic reduction in collection efficiency. A.O. 17 reduced the collectibles by 40% and significantly affected the collection efficiency. Institutional Development. Institutional development and strengthening activities varied in intensity among the systems included in IOSP II. Overall, rating is satisfactory in terms of facilitating the - 48 -  progressive turnover of NIS infrastructure to lAs for O&M, mobilizing lAs for self management of financially viable IA, improvements in the collection process and indication of increased participation of women in IA activities. The Technical Assistance component (whose implementation was delayed for about two years, due to factors external to the implementing agency) facilitated major institutional change through irrigation management transfer or IMT which was not envisioned at appraised. IMT provided for the transfer of O&M of NISs or parts of NIS based on hydrologic control point (per lateral) for NMSs of more than 3,000 ha and of the whole of smaller NISs (3,000 ha or less). As of completion date, 44,715 ha out of 86,913 ha have been turned over to lAs under the IMT or Stage 111 scheme. A major deterring factor, however, is the issue on staff redundancy and the reduced ISF collection base and efficiency resulting from the socialized ISF rate based on A.O. 17. A study on affected personnel has been done recently and support for acceptable retrenchment package is being pursued. IA development programs focused on IA and TSAG/FIG reorganization and strengthening, CIA organization and preparatory activities towards IMT which were part of the IMT Process developed in the IMT pilot area in NDC-5 of District III in MRIIS. Activities such as Consultative workshops, Planning workshop and Core Group Formation, orientation on IMT and presentation of CBL, Articles of Incorporation and IMT Contract were conducted in some systems, notably in MRIIS and Agos, but sadly lacking in other systems. The project also supported the regular IA activities such as BOD and General Assembly meetings, conferences, functionality survey and information dissemination. Training and upgrading of skills for both the CIA/IA officers and the general membership were similarly provided by the project. These are recognized by the members as primary interventions towards strengthening the CIA/IA. Among the trainings conducted were Basic Leadership Training, Financial Management, System Management, Values Formation, IA Collectors' Training, Rice Production and Integrated Pest Management. Based on available data, a total of 166,259 CIA/IA officers and members participated in the training at a total cost of P18,799,000. It is to be noted, however, that training continued to be provided by NIA rather than taken over by the CIA or IA as its responsibility. The cascading strategy for capability development as conceptualized and adopted in the IMT piloting activities in NDC-5 of District III in MRIIS through the Technical Assistance was envisioned to facilitate the learning process through the assistance of CIA and IA officers in the conduct of training activities for their members. This strategy is based on capability build-up by actually doing the activity and learning on-the-job. The available data on institutional development activities imply that this strategy was not operationalized in the IMT replication areas. As of December 2000, 22 of the 37 CIAs in MRIIS Districts I, III and IV with a membership of 26,422 fanners have signed the IMT contract, representing about 36,770 ha or 42.3% of the total area of 86,913 ha to be benefited under IOSP II. In the smaller systems with service area of less than 1000 ha (Bonga I, 2 & 3, Cura, Cantingas, Matogdon and Pongso) 12 IAs/CIA have signed the IMT contract covering an area of 2,957 hectares and farner-members of 4,638. Baggao, which has a total service area of 2,300 ha has one IA with IMT contract which covers 707 ha and 1218 farmers. In Agos, one IMT contract has been signed covering an area of 175 ha with a membership of 500. In Malasila four IMT contracts have been signed covering a combined area of 4106 ha and a combined membership of 3948 farmers. In IAAPIS, Sta. Maria-Mayor, Barit and Maranding IA contracts remain to be either Stage 3, Type 1 or Type I and 2 although preparatory activities towards IMT have been initiated and IMT contracts set to be forged with IA/CIA. Overall, 40 IMT contracts have been signed with a total area of 44,715 ha which is 51.4% of the target area (86,913 ha) and a total membership of 36,726 farmers (refer to Annex 10A). Of the 302 lAs in the 17 systems, 213 have existing contracts with NIA. Of this, 149 lAs have Type 1 and 2 contract, 24 have Type 2 contract and 33 with Stage 2 contract (refer to Annex lOB). - 49 -  NIA Staff Training and Workshops. Orientation and review and planning workshops were the major activities conducted within 1993 to 1996. As a preliminary activity to the preparation for system improvement and NIS performance, technology transfer sessions on NIS assessment were conducted by the PMO. These were aimed at providing the staff with the appropriate skills in identifying gaps and measures that should be considered in the design and implementation of improved modality of operation. As an offshoot of the Technical Assistance, which started in April 1997 and was completed in 1998, majority of trainings and allied activities centered on Modality of Operation, generally through on-the-job training (OJT) and Irrigation Management Transfer. However, replication of the OYT on modality in the other IOSP systems was not carried out as programmed. Delays in project implementation resulted to delays in system improvements, which consequently affected the implementation of the OYr scheme for the NIA staff. Project Review and Consultation Sessions were likewise given more emphasis in the later part of project life. By and large, trainings on O&M and Technical Design were conducted on-the-job (OJT). Two NIA staff were sent on a Study Visit to Mexico in the later part of CY 2000 to observe and study the Irrigation Management Transfer program in Mexico. To develop the intemal capability of NIA in project preparation, on-the-job skills training on the participatory project preparation procedures adopted in IOSP II was conducted through actual project preparation for NISPOP in 14 pilot NISs in 14 regions. Agricultural Support Services. The agricultural support services component of IOSP II aims to support only the essential interventions and farmer/extension staff training to provide farm families under the System Improvement Program the best available technologies in order to maximize yields and incomes. These consist of the setting up of verification and demonstration trials of new and appropriate technology packages, IA-based seed multiplication and distribution, training of IA members on soil and water management and agricultural production practices, and the conduct of IPM-Farmer Field Schools. Whenever possible, all the support provided by the program was channeled through the lAs. The lAs coordinated with the Municipal Agricultural Officer (MAO) and Agricultural Technicians (ATs) of the concerned LGUs through their Sub-Committee for Agricultural Production. During the first year of project implementation, the component was piloted in three schemes along with the implementation of systems improvement and based on the results, implementation was expanded with necessary modifications. A number of verification and demonstration trials were established covering all the systems under IOSP II. These consist of demonstrations on appropriate palay production technologies, farming systems technologies, variety trials and IPM. A total of 482 demonstration trials (81% of target) were established until project completion involving a total of 473 farmer cooperators. Limited agricultural support facilities were also provided consisting of a nursery and four (4) multi-purpose drying pavement (100% of target) in Region I benefiting 416 farmers. The seed production and distribution sub-component was able to distribute a total of 1,495 (73% of target) bags of foundation and registered seeds, respectively, producing a total of 132,272 bags (44% of target) of registered and certified seeds. A total of 1,402 farmers (75% of target) were benefited by this scheme. In addition, 267 bags of certified seeds (130% of target) were distributed directly to 279 farmers (104% of target). Overall this component is rated satisfactory. The institutional strengthening sub-component provided training on a wide variety of subject matter which include rice production technologies, pest management, crop diversification, seed production, rice-fish culture and other farming system technologies, assistance to organization of self-financed group involving a - 50 -  total of 3,464 farmner participants (75% of target). Improvements in yield levels for both the dry and wet season were observed only in 9 out of the 17 systems. Despite the decline in yield levels of the other systems, overall, the 17 systems posted an average yield level of 3.92 MT and 4.39 MT for the wet and dry season, respectively. During project preparation (1992) yield levels were 3.87 MT for the wet season and 3.97 MT for the dry season. This gives a yield increment of 50 kilos per hectare for the wet season and 420 kilos during the dry season. The NISs reporting relatively high yield levels include MRIIS Districts 3 & 4 and Bonga Pumps I & 2. As expected, these yield levels are also relatively higher than the average yield of 3.37 MT and 3.45 MT for the wet and dry season for irrigated areas, respectively, as reported by the Bureau of Agricultural Statistics (BAS) since BAS survey covers both IOSP and non-IOSP areas. On the other hand, the average estimated yields reported by the NIA field offices were lower than the projected yields and the survey results at 3.58 MT and 3.92 MT for the wet and dry season, respectively. Although on a per system basis, seven (7) systems were reported to have attained slightly higher yields than the survey results. For the rehabilitated areas, the average yield attained was 3.89 MT and 4.41 MT for the wet and dry season respectively. Compared to the 1992 level of 3.87 MT and 3.97 MT for the wet and dry season, the increase in the wet season of 20 kilos was compensated by the substantial increment of 440 kilos in the dry season. On the other hand, for the restored and generated areas, the average yield reported was 3.96 MT and 4.25 MT for the wet and dry season, respectively. This brought about an incremental yield of 1.82 MT for the wet season over the average yield of non-irrigated rice as per BAS data for 1999 of 2.4 MT. In terms of fertilizer usage, the farmers reported an average usage of 6.71 bags, which approximates the recommended package of the Department of Agriculture. It is interesting to note, that the use of organic fertilizer was reported in most of the systems, albeit in small quantities. The effect of the training on IPM, one of the inputs of the Agricultural Support Component, was reflected in the relatively small expenditure on chemical applications. The expenditure on chemicals as reported in the survey consists of chemical applications to control pest and diseases and snail infestation. The average cost of chemical applications for all the system was P1,140 for the dry season and P1,234 for the wet season as compared to 1992 levels of P1,920 and P2,004 (adjusted to 2000 price using the CPI). The bulk of the present chemical usage went to the high cost of chemical to control snail (P1,100/li) as reported by a majority of the respondents in all areas. 2.3 Net Present Value/Economic rate of return: A re-estimation of the EIRR was done the following the NEDA guidelines used in project preparation and using updated information gathered from the survey and reports from the systems. These include actual project costs incurred by the NIS and incremental areas by season (as of project completion), crop yields, input use and prices (based on the survey). For systems improvement component, despite the delays in project implementation, the resulting EIRR for all the 17 NIS of 25.57% is highly acceptable. It is significantly higher than the SAR estimate of 19.1% although lower that the FS estimate of 33.6%. This is understandable since the SAR estimate considered only the five systems and the assumptions used were different. The major reasons for the good project performance are the high incremental yield levels attained by MRIIS which accounted for majority of the area and the low investment but high incremental benefits of Malasila which more then compensated for the marginal performance of the other systems. The estimated NPV and BCR at 15% discount rate are P690 million and. 2.02, respectively. - 51 -  A sensitivity analysis of the project showed that even if only 25% of discrepancy between the target and accomplished incremental areas as of December 2001 could be attained by 2003, overall, the project will still be viable with a computed EIRR of 16.8%. For the urgent repairs component, EIRR re-estimation was done for 13 out of 17 systems repaired using IOSP funds. Despite the delays in implementation, which resulted to higher repair costs, all the resulting EIRRs are at acceptable levels, except for Nasisi Dam, Pagbahan and Lasang. The average (weighted) EIRR for the 13 systems is 18.21% which is lower than the 28% ETRR for the 20 systems as per SAR. 2.4 Financial rate of return: A cost and returns analysis per hectare of palay production reveals that irrigated palay production still generates attractive revenue for the owner operators. On the average, an owner operator generated a net return per hectare of P16,012 and P19,213 for the wet and dry season, respectively, or a total net revenue of P28,756 for one year at an average cropping intensity of 163% (0.83 for dry and 0.8 for wet). The imputed cost of family labor amounting to an average of P2,192 per year was included in the expenses. Adding back to the net revenue, the farm family has actually generated an average revenue of P30,948 from farming. For the rehabilitated area, the average net revenue reported was P13,570 and P20,138 for the wet and dry season, respectively, for an average net revenue of P27,570 per hectare per year and considering/factoring the effect of cropping intensity. For the restored and generated areas, the average net revenue generated was P15,898 for the wet season and P18,290 for the dry season for a total yearly net income of P27,899. The estimated net revenue of P14,638 and P21,196 (excluding family labor) for the wet and dry season, respectively, in the rehabilitated area is still higher than the estimated net revenue during project preparation of P13,104 and P20,331 for wet and dry, respectively (P7,268 for the wet and P1 1,276 for dry adjusted to Year 2000 using the CPI). The total incremental benefit for the year is P865 for the dry season and P1,534 for the wet season. For the restored areas, the incremental net return per hectare (over the rainfed crop of P4,496 (BAS 1999 regional data adjusted to 2000)) is about P11,650. For the 17 NIS, the average incremental net revenue generated was P3,183.72 for the wet season and P3,854.92 for the dry season. 2.5 Institutional development impact: Survey results showed that there was a high percentage of participation by the CIA/IA officers in organizational activities such as in the formulation of Operations Plan together with NIA, planning activities for improved ISF collection scheme and preparation of cropping calendar. The IA members have also taken on a more active participation in planning meetings to strengthen the CIA/IA as well as in planning for income generating activities. Results also indicated that the CIA/IA officers consulted with members before making decisions on certain issues such as on matters related to finance, water distribution schedule and cropping calendar. This implies that as an offshoot of the institutional development and strengthening activities, members are now more recognized as active partners and participants rather than mere recipients in the irrigation development and management process. With IMT there has been a more organized and focused management by the CIA/lAs at the lateral level. With the transfer of the O&M responsibilities as part of the IMT contract, the CIA/lAs have taken on an increased role in O&M particularly in the area of monitoring of water delivery, identification of irrigation facilities that need repair, canal clearing and vegetative clearing at the lateral level. Similarly, the officers - 52 -  have taken an active role in ISF collection. This is a positive step towards empowerment and a realization of the vision that as IMT progresses and farmers take on the responsibility to make decisions for the improvement of the irrigation service, not only in the area of O&M but in organizational and institutional areas as well, there is a corresponding decrease in interventions provided by the NIA. The project also facilitated increased capability through the participation of the lAs in system improvement and this provided the lAs the financial capability to put up their share for system improvement which is now kept as trust fund for future repair. Generally, in all of the 17 IOSP systems, the members have indicated their satisfaction with the improved irrigation service particularly in relation to reliable and adequate water supply. There were also indications that members have become more cooperative, active and responsible. Officers of the CiA/IA have also been observed to be more critical and wanting in information that will enable them to make informed decisions. They see their role in varied ways in relation to the success and sustainability of the project in terms of efficient performance of their duties and responsibilities as officers, punctual payment of ISF and assistance in ISF collection, fostering of cooperation with and among the CIA-IA officers and NIA as well as providing support in the implementation of programs and policies. 3. Major Factors Affecting Implementation and Outcome 3.1 Factors outside the control of government or implementing agency: The implementation in a number of NISs was affected by the occurrence of typhoons during the construction period and crop maintenance. The repairs to the damage increased cost and resulted to delays in implementation. 3.2 Factors generally subject to government control: Civil works implementation was affected by the delays in the release of support funds, particularly during the first three (3) years of implementation. These delays caused uncoordinated implementation of components and lowered the overall performance rating. Moreover, the sensitivity analysis that was conducted in the SAR showed that the Project's ERR will decrease from 16.7 to 12.3 with a three-year delay in implementation. The promulgation of R.A. 8435 (Agriculture and Fisheries Modemization Act) best enforces the objectives of the Project and plays a vital support to its achievements. However, presidential pronouncements and issuances (particularly, A.O. 17) resulted to difficulty in attaining the desired CE level and was cause for apprehensions by both the field staff and IAs. Furthermore, the Project and the Agency also faced difficulties in acquiring approval for the TA component in view of Government policies to undertake TA using grant funds only. Approval was acquired only in 1997 resulting in the delay of TA-dependent activities. 3.3 Factors generally subject to implementing agency control: Untimely release of funds was resolved by the Agency through the conscientious reprogramming of activities, which considered the interdependent aspects and implementation of action programs. With assistance of the Bank, the Agency was able to source grant funds for four (4) out of six (6) studies affected by the non-approval of the TA and further sought reconsideration for the remaining two (2) sub-components. 3.4 Costs andfinancing: The total project cost at completion is estimated at P2,211.6 million equivalent to US$ 67.86 million (SAR - 53 -  estimate was P2,081.3 million). Of this total, 71 percent came from the loan proceeds equivalent to P1,574.7 million (US$ 46.5 million) while government counterpart amounted to P636.9 million equivalent to US$ 21.3 million (see Annex 2 for details by component). The average cost for incremental irrigated area in NIS as of November 2000 was P63,146/ha (US$1,503.00). On the other hand, the rehabilitation cost/ha in NIS (which also included incremental irrigated area) was P 12,571 (US$299.00). 4. Sustainability The sustainability of the irrigation service hinges on the proper mnaintenance and operation of the facilities and the preparedness of the CIA/IA to manage the systems either fully or at the lateral level. The eventual transfer of management of the laterals to the CIA/1A carries with it the responsibility for operations and maintenance. The IMT contract between NIA and the CIAs/IAs defined the sharing anrangement when it comes to ISF collection and the responsibility on the operations and maintenance of the system but as pointed out by previous Bank review mission, there is no built-in mechanism or assurance that the lateral will maintained properly. The CIA/IA should be able to determine the level of expenditures for maintenance requirements of the lateral and set aside sufficient funds for the purpose. NIA records show that from 1993 to 2000, the average cost of O&M per hectare for the 17 systems under IOSP 11 is P2,297 per hectare. It ranges from a low of P79 per hectare for Cantingas to as high as P3,862 per hectare for Sta. Maria. This wide disparity in O&:M expenses necessitate a per system sustainability analysis. Given this average per hectare costs, the total O&M expenses per system was computed and compared to the ISF collectible under the old rate and the socialized rate. ISF collectible was further computed under four conditions: a) at presently irrigated area; and upon attainment of the balance of the target incremental areas at b) 25%; c) 50%, and d) 100%. It is encouraging to note that eight (8) NIS have more than enough collectibles even at socialized rate to sustain the O&M expenses with the present irrigated area. However, the analysis also reveals that the ISF collectible both at the old and socialized rates, even at the maximum area attainable for Agos, Barit and Pongso will not be sufficient to sustain the present O&M expenses. The rest of the systems can survive only if the ISF will be reverted to the old rate (See Amnex 7Ga). With the IMT program, it is envisioned that NISs with service area of 3,000 hectares and below shall be fully tumed-over to the lAs. In fact, some of these snrall systems have been turned-over to the lAs but are still being supported by NIA. At full turnover, the O&M expenses shall be fully shouldered by the lAs technical supervision by charging a management fee of 10-15% of the collectible ISF to cover the salaries of the personnel who will supervise the system. Full turnover will mean a lower O&M costs since it is known fact that government expenditures, particularly salaries and wages, are usually higher (40-70%) that if it will be performed by the lAs. Furthermore, the IAs are expected to be more effective in ISF collection and could also make adjustments in the ISF rates if the sustainability of the system will be affected. For pump systems with high operations costs, NIA has to provide financial assistance to LAs (soft loan) to sustain their operations until such time when they could be on their own. For Agos, Barit and Pongso, NIA has to seriously study all possibilities to reduce the O&M costs to make the tumover acceptable to IAs. This should be not only through the retirement or transfer of IMT affected personnel but also reduction of the direct costs of O&M. For example, for Agos O&M cost could be reduced if the affected personnel will be retired or transferred to other projects since almost all of the COB went to Iheir salaries and wages. Only two relevant position may be absorbed by the CIA/LAs (1 Engineer A and I WRFT) with a total basic salary of P274,056. Administrative costs to the IA will amount to P1.3M plus the maintenance costs of P1.13M. The total - 54 -  operation and maintenance costs will amount to P2.71M, which could sufficiently be covered by the potential ISF only at the old rate. This cost could be reduced further when the lAs take over the operations. Taking MRIIS District 1 as a case for partial turnover, the 0 & M cost could be reduced by as much as P4.28 M by retiring the willing affected personnel and by another P1.85 M with the expenses on canal maintenance contracts with lAs. This will reduce the COB from P23.92 to P17.8 M. Adding one-half of the expenses coming from other local funds of P4.1 M, the total 0 & M expenses to be shouldered by NIA will be P21.9 M. This could be sufficiently met by the potential ISF at socialized rate (at 50-50 sharing arrangement) of P23.43 million. Since MRIIS is collecting at socialize rate, collection efficiency should therefore not go below 93%(at the present irrigated area) to make it sustainable or at a lower efficiency of 84% if 100% of the target irrigated area under IOSP II could be attained. On the side of the CIA/lAs, the 0 & M costs to be shouldered is estimated at P9.5 M consisting of the following: a) the salaries and wages of displaced personnel at basic rate amounting to P2.46M; b) one-half of the expenses coming from other local sources of P4.1 M; c) the costs of IA contracts formerly being paid by NIA in the amount of P1.85 M; and d) an assumed administrative cost of 15% or P1.24 M. Of the 50% share from the ISF collectible equivalent to P23.43 M, the CIA need to earmark 40% for 0 & M. IMT therefore is favorable to the CIA/lAs in MRIIS since they could accumulate funds from their share after 0 & M which could be used for other income generating activities of the associations and for future rehabilitation of the system. It is therefore important that he CIAs/LAs understand that under the IMT program, the share in the ISF collection is actually intended to cover operation and maintenance responsibilities that will be turned over to them. The commitment and responsibility of the CIAs/IAs for the maintenance and operation the irrigation facilities has to be clearly defined not only in the IMT contract but should also be part of the by-laws of the association. The sustainability of the systems improvement under IOSP-II is likely, however, this would need further support to IMT in terms of the following: a) establishment of CIA/IA self-sustaining mechanisms in O&M. there is an indication that funds for maintenance and operation are allocated in the CIA/IA budget but strict operationalization of the plans has to be observed, monitored and evaluated. Furthermore, NIA and CIAs have to ensure that sufficient funds are allocated based on actual needs of the systems and strictly disbursed for the purpose. This has to be explicit in the NIA-CIA/IA Contract, in the CIA/IA By-Laws, and supported with intemal rules, procedures and mechanisms. b) Refinement of IMT policies in accordance with the provisions or AFMA (RA 8435) particularly those pertaining to affected personnel. Since the goal of IMT is sustainability of both NIA and [As, proper attention given this issue will assure long-term success. There is really a need for unrelenting enforcement of NIA's organizational streamlining program. Streamlining is an imperative to achieve corporate viability. Corollary to this, definite guidelines on IMT specifying manpower requirement vs NIA and CIA roles has to be finalized. This will conscientiously define the manpower requirement during IMT, such that O&M is given due attention. With successful IMT, assurance of sustainable O&M will also be attained. c) M&E of IMT performance should be done intensively during the transition period in order to ensure sustainability. Mechanism for M&E and IMT sustenance which have been introduced in pilot NISs, have to be assessed and if necessary refined and thereafter, replicated in all the areas. d) Implementation of other transition measures in order to complete the transfer of technologies to - 55 -  lAs, particularly for some aspects of O&M. This should adopt and on-the-job (OJT) approach. Likewise, the proper reorientation of NIA-ISO and staff roles, duties and skills is also necessary. e) Institutional strengthening measures should ensure that the capabilities and skills required to be able to effectively manage the system are installed among the officers and the necessary internal reorientation among the general membership is effected to ensure continued member participation and cooperation in the activities of the organization. 5. Bank and Borrower Performance Bank 5.1 Lending: The Bank lending performance was highly satisfactory. Despite the delays in the provision of counterpart funds and in the recruitment of consultants, the Bank continued to provide support to the Project. Moreover, the Bank even granted an extension of the loan closing date from June 30, 1999 to December 31, 2000 to allow the institutional component to have a significant impact on the Project beneficiaries. 5.2 Supervision: Bank supervision was satisfactory. The supervision missions reviewed the progress of work almost every six months. Apart from reviewing progress of work, technical details, factors, constraints and prospects were likewise probed through interactions and field visits. 5.3 Overall Bank performance: The overall Bank performance was satisfactory. The Bank was flexible in administering the loan and assisted the GOP in resolving shortage of counterpart funds. Upon the GOP's behest, the Bank established the Project special account (PSA) and later agreed to a request for the use of a 90-day facility to expedite works implementation. The Bank provided quick responses to queries and requests, including requests for specific modifications and an 18-month extension of the loan closing date. Borrower 5.4 Preparation: The preparation by the Borrower of the different components for project appraisal was satisfactory. The Borrower completed the participatory project preparations before the final negotiations, involving the active participation of beneficiary lAs. All components, particularly those in the 17 NISs with S.I. were implemented in a coordinated manner. 5.5 Government implementation performance: The performance of the Borrower may not be considered satisfactory with regards to availability of counterpart funds that made it difficult for the Project to conscientiously pursue programmed activities. However, the situation slightly improved in 1998 but this was still insufficient to meet the requirements in terms of volume and timeliness. 5.6 Implementing Agency: Overall, the performance of the implementing agency (NIA) was satisfactory. The investigations and surveys were adequate with the incorporation of improved participatory procedures that strengthened socio-technical coordination in system improvement. The accomplishments would not have been substantial if not for the efficient reprogramming in view of delays in support funds. Strict implementation of policies and guidelines for works implementation that ensured benefits was carried out. In addition, the implementation of joint civil works implementation at secondary level to assure IA participation with improvement share deemed to install ownership were also undertaken. This scheme, further, assured the - 56 -  availability of some funds for emerging repair works and are kept as trust funds for such. 5.7 Overall Borrower performance: Despite the delays in counterpart funds, overall borrower performance is rated satisfactory. This is in view of the ability to attain objectives substantially. 6. Lessons Learned * ,The technical change under modality of operations and all the underlying principles would warrant sustainability of project benefits. However, knowledge and skills have to be appropriately transferred to NMA field personnel and lAs. * Urgent repairs should be done at the soonest time possible, in order to realize its objective in the most cost-effective manner. Late start implementation could give way to further deterioration of the structures that make them prone to damage in time of calamities, hence, increases in cost requirement as in the case of Balanac. It is a recognized fact that NISs naturally deteriorate at a certain rate/year. And structures identified for urgent repair have more sensitive condition. Thus, fund support should be released as per program/schedule. * The IMT likewise caused institutional change as in the assumption of more responsibilities by ClAs/IAs. The immediate implication is the staff redundancy issue for which any implementing agency should have corresponding programs before vigorously pursuing the program. * Improvements on technical design and O&M are better facilitated by on-the-job trainings. As in the case of Technical Design Workshops, outputs were immediately used for the design of modality of operations. This enables immediate application and evaluation. Participatory system evaluation as in the case of NIS Assessment enables wholistic identification of performance gaps and ensures measures acceptable to both parties (NIA and lAs). The strategy ensures acceptability and provides better prospect of sustainability. The timing of all components is critical to project implementation. The bulk of activities were conducted after start of IMT development and more needs have been identified thereafter. This validates the fact that institutional interventions should have a lead time, closely coordinated with technical interventions. The appropriate timing of detailed interventions, however, boils down to factors extemal to the implementing Agency. (Late approval of the Technical Assistance Component on the part of the GOP). * The new rates promulgated by AO 17 resulted to significantly lower total ISF collectibles but much higher collectibles from big landowners. Many of those with big landholding (>5.0 ha) have, in the past, incurred ISF back accounts and have been difficult to convince to pay ISF. If this trend continues the projected collection efficiency will be affected. The cascading theory of technology transfer must be implemented to facilitate capability build-up and allow the participants to develop personal sense of confidence and openness to new learning. This is also meant to facilitate farmer empowerment by providing assistance to the limited NIA field personnel tasked with providing the support services necessary in effective group participation and change. There is no doubt that coordinative implementation of components abates the repercussion of delayed financial support for works. More benefits could have been attained if financial allocations were provided on time. - 57 -  While substantial activities have been conducted, the introduction of IMT, with inputs from other countries necessitate further interventions that were not covered in the original project design. There are activities identified but completion of project duration did not warrant its implementation. Among these are: Polishing of IMT manual, which cover: operations planning, design of modality of operations, technology transfer for both NIA and LAs, and monitoring and evaluation. A prototype manual was produced under TA but polishing of the manual would provide opportunities for incorporating improvements based on experience of the different NISs. Likewise, the manual should emphasize on the coordination of all components, which should be well contextualised at operations level. Thereafter, translation into the vemacular may be done and should be included in the documents to be transferred to lAs. * Policy Deliberation Workshops for IMT. The program has implications on IA policies - manpower, extent of transfer coverage, financial and institutional change towards sustainability. Series of policy deliberation workshops would enable management to intensify consultations with affected personnel and offices, such that policy issuances would expedite in more conscientious manner, maximizing synergy within the Agency. * IMT Advocacy Workshops. The implication of IMT on NIA personnel is a threat to Program success. However, the knowledge and experiences from the project would benefit other projects. IMT Advocacy Workshops would enable dissemination and sharing towards a unified direction in participatory irrigation development. NIA may use its experienced personnel and/or invite external resources, if necessary. 7. Additional Information Additional data and information are supplied in the supporting annexes of this Implementation Completion Report (ICR). - 58 -