Philippines - Upper Pampanga River Irrigation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No.3063 PROJECT PERFORMANCE AUDIT REPORT PHILIPPINES - UPPER PAMPANGA RIVER IRRIGATION PROJECT (Loan 637-PH) June 30, 1980 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. LIST OF ABBREVIATIONS ADB - Asian Development Bank ADCC - Agricultural Development Coordinating Council (at the Province level) ADD - NIA, Agricultural Development Division (now, Department) HYVs - high-yielding varieties ICB - international competitive bidding IOMP - Input-Output Monitoring Program IRRI - International Rice Research Institute MCM - million cubic meters MOA - Ministry of Agriculture NEDA - National Economic Development Authority NGA - National Grain Authority NIA - National Irrigation Administration (the executing agency) NPC - National Power Corporation 0 & M - operation and maintenance OSP - NIA, Office of Special Projects PDD - NIA, Project Development Division (now, Department) RA(No.)- Republic Act No..... SAR - A BaTk's Staff Aptraisal Report UPRIIS - Upper Pampanga River Integrated Irrigation System (the project after completion) UPRP - Upper Pampanga River Irrigation Project (the project during construction) USAID - US Government, Agency for International Development USBR - US Government, Bureau of Reclamation WCCC - UPRIIS, Water Control Coordination Center WMT - water management technicians EXCHANGE RATES Name of Currency (Abbreviation) Peso (M) Year: Appraisal Year Average Exchange Rate: US$1 = 3.90 Intervening Years Average US$1 = 6.80 Completion Year Average US$1 = 7.50 WEIGHTS and MEASURES Metric System, except: 1 cavan (of rice) = 50 kg FOR OFFICIAL USE ONLY Project Performance Audit Report PHILIPPINES - UPPER PAMPANGA RIVER IRRIGATION PROJECT (Loan 637-PH) Table of Contents Page Preface i Basic Data Sheet ii Highlights iii PROJECT PERFORMANCE AUDIT REPORT I. Project Summary 1 II. Main Issues 3 A. The Lack of Resettlement Planning 3 and its Consequences B. Water Management 6 1. Training 6 2. Planning and Operation 6 3. Operation and Maintenance; Water Charges 7 C. Optimistic Targets for Double Cropping 9 D. Labor Constraints 10 E. Allocation of Observed Benefits Among Different Projects 11 F. Farm Size and Number of Farmers 12 Attachment 1 - Philippines Watershed Management and Erosion 13 Control Project; Loan and Project Summary Attachment 2 - Comments Received from the Government of the 15 Philippines PROJECT COMPLETION REPORT 1. Summaryl/ 21 2. Background 21 3. Formulation 22 4. Implementation 23 5. Agricultural Impact 27 6. Institutional Performance 28 7. Economic Evaluation 31 8. Bank Performance 33 9. Conclusions 33 Tables Annex 1 1/ Included in the Audit Memorandum. 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.  Project Performance Audit Report PHILIPPINES - UPPER PAMPANGA RIVER IRRIGATION PROJECT (Loan 637-PH) Preface This is a performance audit of the Upper Pampanga River Irrigation Project, for which Loan 637-PH was approved in August 1969 in the sum of US$34.0 million and closed, after cancellation of US$0.12 million, in December 1977. The audit report consists of an audit memorandum prepared by the Operations Evaluation Department and a project completion report (PCR) dated April 12, 1979. The PCR was prepared by the East Asia and Pacific Regional Office on the basis of a draft PCR prepared by the project's executing agency (the National Irrigation Administration - NIA). An OED mission visited the Philippines in November 1979. The mission held discussions with officials of NIA, both at headquarters and in the field. The project area and some of the project features were visited; some farmers and farmer representatives were interviewed. The information obtained during the mission was used to test the validity of the analysis and conclusions of the PCR. The audit memorandum is based on these discussions, on interviews with Bank staff associated with the project, and on a review of the PCR, the President's Report (No. P-736) dated July 30, the appraisal report (No. PA-2a) of July 30, 1969, the Loan Agreement dated August 18, 1969, correspondence with the borrower, and internal Bank memoranda on project issues as contained in relevant Bank files. A copy of the draft report was sent to the Borrower on April 11, 1980 for comments. Comments received from NIA and NEDA are in Attachment 2 of the memorandum. Suggested clarifications have been incorporated. The audit finds most data and analyses in the PCR appropriate. The audit memorandum discusses the PCR's economic evaluation of project benefits and the resettlement operation, which has provided the Government and the Bank with valuable lessons, and highlights the difficulties encountered in maximiz- ing project benefits. The valuable assistance provided by the Government of the Philip- pines, by NIA and its staff, and by the farmers visited is gratefully acknowl- edged. o Project Performance Audit Report PHILIPPINES - UPPER PAMPANGA RIVER IRRIGATION PROJECT (Ln. 637-PH) BASIC DATA SHEET KEY PROJFCT DATA Appraisal Actual or Item Expectation Current Estimate Total Project Cost (US$ million) 62.5 127.94 Overrun (%) 105 Loan Amount (US$ million) 34.0 34.0 Disbursed ) 33.88 enal1ed up to March 31, 1980 3.0 Outstanding ) 30.29 a/ Date Physical Components Completed June 1975 December 19771. Proportion Completed by Above Date (%) 78 100 Proportion of Time Underrun or Overrun (%) 43 Economic Rate of Return (M) 13 14_1 Financial Performance Good Good Institutional Performance Good Good CUMMULATIVE DISBURSEMENTS (US$ million) FY70 FY71 FY72 FY73 FY74 FY75 FY76 FY77 FY78 Appraisal Estimate 1.0 5.4 11.2 18.3 27.0 34.0 Actual 03.4 6.1 12.3 16.8 23.9 30.2 33.1 33.7 33.9 Actual as % of Estimate 34 113 110 92 89 89 97 99 99 OTHER PROJECT DATA Original Actual or Item Plan Revisions Pat. Actual First Mention in Files 9/66 Government's Application 9/67 Negotiations 7/7-18/69 Board Approval 8/12/69 Loan Agreement Date 8/18/69 Effectiveness Date 10/15/69 10/10/69 Closing Date 6/30/76 6/30/77 12/31/77 Borrower Republic of the Philippines Executing Agency National Irrigation Administration (NA) Fiscal Year of Borrower July 1 - June 30 Follow-on Project Name Aurora-Peharanda Irrigation Project Loan/Credit Numbers 984/472-PH Amount (us$ million) 9.5/9.5 Loan/Credit Agreements Dates Hay 14, 1974 MISSION DATA Number of Number of Date of Missions Mtonth/Year Days of Persons Martweeksd/ Report Appraisal 9/69 21 4 12 7/30/69 Total 12 Supervision 1 1/70 10 1 2 2/12/70 Supervision 16 4/70 7 1 1 6/18/70 Supervision Ii o 11/70 11 2 3 2/18/71 Supervision IV 5/71 9 1 1 6/30/71 Supervision V 1I7 t 2 3 4/27/72 Supervision VI 9/72 14 2 4 12/11/72 Supervision VII 7/73 13 2 4 9/07/73 Supervision VIII 4/7Ae s 13 2 4 5/21/74 upervision IX o/Yer 25 3 11 6/26/75 Supervision I 3/76e 18 4 10 5/13/76 Supervision II 1/7 18 3 8 2/23/77 Supervision XII 10/77.5 14 3 3 2 Total 57 Completion 7/78 4/12/79 a/ Pl s an echange adjustment of US$4.13 million, for a total Borrower's obligation of 1 /734.41 million. / Most works were completed by December 1976. I/rTclu-es benefitS which bane also been credited to other projects. See par. 3X - 33. d/ This includes only man-weeks spent i the fielr. e/ Combined with other projects.  Project Performance Audit Report PHILIPPINES - UPPER PAMPANGA RIVER IRRIGATION PROJECT (Loan 637-PH) HIGHLIGHTS Loan 637 provided funds for constructing the Pantabangan Dam on the Pampanga River, rehabilitating and upgrading 46,000 ha of existing irriga- tion systems; building new irrigation works to serve about 31,000 ha and constructing feeder roads and operations and maintenance facilities. The loan also provided for a study of the operations, structure and financial situation of the National Irrigation Administration (NIA, the project executing agency); land classification of the project area, and a feasibility study of possible future power development at the Pantabangan Dam (which the Bank later financed under Loan 1034-PH of 1974). Also included in the project during implementa- tion was a feasibility study of the contiguous Aurora-Pefiaranda project (subsequently financed by the Bank under Loan 984-PH and Credit 472-PH, of 1974). The dam was completed 10 months ahead of schedule, despite sig- nificant changes in design. Construction of the irrigation facilities, in contrast, was completed some 30 months behind schedule. Resettling the population displaced by the reservoir proved an unexpectedly difficult task. Cost overrun was 105%, of which about 86% was due to inflation and 19% to changes in dam design (in particular, the addition of an auxiliary embankment to avoid a poor left abutment), population resettlement and higher than expected engineering and supervision costs. The project serves some 78,400 ha, an area similar to appraisal expectations. While dry season cropping rose rapidly, appraisal targets for cropping intensity proved overoptimistic. In contrast, paddy yields achieved are higher than anticipated, largely due to the near universal adoption of high yielding rice varieties. In view of these achievements, the appraisal estimate for the period to full development was reduced from 13 to 7 years. Increase in project costs has been compensated by higher rice prices and the faster rate of development. The economic rate of return of the project is now calculated at 14%, slightly above the 13% estimated at appraisal. The following points may be of special interest: - lack of appreciation at appraisal of the magnitude and complexity of population resettlement. The ensuing lack of resettlement planning led to protracted negotiations; a ten-fold increase in costs, lack of economic viability of the resettled farmers; additional inflow of compensation-seekers and poor choice of resettlement site (PPAM, paras. 4 and 10 - 17; PCR, paras. 4.06 - 4.08 and 8.01); - NIA still lacks a policy setting the appropriate course of action for those years when total water inflows will be insufficient to meet all requirements (PPAM, para. 19);  - iv - - collection of water charges is still not comprehensive but is continuously improving; an important feature which has ensured that water charges are inflation-proof is that rates were set in kind rather than in monetary units (PPAM, paras. 21 - 24; PCR, paras. 6.06 - 6.07); - insufficient allowance was made at appraisal for the difficulties of shifting from single to double cropping (PPAM, paras. 25 - 26; PCR, para. 5.01); - despite some required adjustments, labor availability was suffi- cient to meet increased requirements of expanded production (PPAM, paras. 27 - 29); - methodological problems in evaluating two or more concurrent and mutually reinforcing projects - as this particular project illustrates - are still being solved (PPAM, paras. 30 - 33); and - the project was a major factor in developing NIA's technical and managerial capabilities (PCR, paras. 6.01 - 6.04, and 9.02).  Project Performance Audit Memorandum PHILIPPINES - UPPER PAMPANGA RIVER IRRIGATION PROJECT (Loan 637-PH) I. PROJECT SUMMARY-! 1. The Upper Pampanga River Irrigation Project (UPRP) was appraised in September 1968; the loan was signed in August 1969 and became effective in October 1969. The project works were completed in December 1977, some 30 months behind the scheduled date of June 1975. The loan was closed on Decem- ber 31, 1977. 2. The project was designed to provide a reliable water supply for irrigation of a service area of about 77,000 ha in Central Luzon, some 120 km north of Manila. The aim of the project was to raise production of irrigated rice through higher wet season yields and by a very substantial increase in dry-season cultivation. The project works included construction of an earth and rockfill dam and a reservoir at Pantabangan on the Pampanga River, rehabilitation and upgrading of 46,000 ha of existing irrigation systems and construction of new irrigation to serve about 31,000 ha. During detailed design the dam crest was lowered by 3.5 m to avoid submerging the nearby town of Caranglan; an auxiliary dam was added to reach a better left abutment. The dam, as built, consisted of a main embankment across the Pampanga river, an auxiliary embankment across the Aya creek, and a saddle dam connecting the two embankments. These changes resulted in 30% increase in embankment material and a reduction in reservoir storage capacity from 3,500 million m3 to 3,000 million m3. The service area built under the project totalled 78,400 ha, consisting of 46,000 ha of rehabilitation and 32,400 ha of new construction. The relatively intensive system of ditches and drains, accompanied by a good road network, was new to the Philippines. It permitted the introduction of rotational irrigation to 10 ha units, compared to the previous practice of continuous flow irrigation, with field to field flooding of relatively large areas of land. The system has since been adopted as standard on all subse- quent Bank-assisted projects built by the National Irrigation Administration (NIA). 3. The construction of the dam was completed 10 months ahead of sched- ule, despite significant changes in design. However, there were delays in construction of the irrigation facilities due to procurement problems, poor performance of the contractors2/, and interruptions of work to allow delivery of irrigation water to farmers. This delayed completion of work in the service area to the end of 1977, some 30 months behind schedule. At project 1/ Adapted from the PCR. 2/ Contracts were awarded to firms which did not have previous experience in this kind of on-farm facilities - actually, nobody had such experience in the Philippines (see PCR, para. 4.05). Thus, they turned out to be incap- able of successfully completing the items on conditions stipulated in the contract. NEDA suggests that more care and' a more thorough counter- checking or a more rigid system of screening should be made before pre- qualifying contractors to participate in public bidding (see Attachment 2, page 6). -2- inception, neither the NIA nor the Bank realized the magnitude or complexity of the task involved in resettling the population displaced by the reservc4r formed by the dam. None of the preappraisal studies addressed the resettl-, ment issue, and its impact was not evaluated at appraisal, although a modest provision for relocation was made in the appraisal cost estimate for the dam, A total of 2,300 families, or 13,000 people, was displaced by the dam. Of these, 187 families were resettled on irrigated lands in the service area and the remaining 2,116 families were, by their own choice, resettled in four newly developed sites on the hills along the southeastern rim of the reservoir. 4. Relocation became a very emotional issue and was magnified and exploited for political purposes. To obtain agreement of the affected popula- tion, NIA complied with the demand that resettlement be within the municipal boundaries of the flooded town, so as to preserve its political identity. Although the settlers were supplied with new housing on easy terms, potable water, electricity, schools and excellent roads, the hilly terrain chosen for resettlement was not capable of providing a livelihood. The cost of resettle- ment amounted to US$11.7 million, compared to the appraisal estimate of US$1.1 million. 5. The total cost of the project at completion was US$127.9 million compared to US$62.5 million at appraisal. The cost overrun was 105%, of which about 80% was due to inflation during the implementation period and the balance to changes in design of the dam, population resettlement and higher than expected engineering and supervision costs. 6. Prior to implementation, the area now served by the project con- sisted of 46,000 ha of existing run-of-river irrigated rice lands and 32,400 ha of rainfed rice lands. In the dry season only about 5,000 ha of the irrigated lands received water and were cropped to rice. Annual cropping intensity in the project area was about 106%. At appraisal it was projected that dry season irrigated rice would increase to 72,000 ha within three years of completion of construction for an annual cropping intensity of 193%. As built, the project serves 78,400 ha, but 2,900 ha are low lying and are not cropped in the wet season, while 4,200 ha cannot be irrigated in the dry season. Target cropping intensities are, therefore, 96% in the wet season, 95% in the dry season, and 191% annually. While dry season cropping rose rapidly from less than 10,000 ha in 1974, when water from the reservoir first became available, to 55,000 ha in 1976, the appraisal estimate of attainment of designed cropping intensity in three years proved overoptimistic. Insuffi- cient allowances were made for the difficulties of shifting from single to double cropping or for the disruptive effect of typhoons on timetables. NIA now expects the target cropping intensity for the wet season crop to be reached in 1979 and for the dry season crop in 1981. 7. While the rate of increase in cropping intensity has failed to meet appraisal expectations, the paddy yields achieved in the project are higher than anticipated. At appraisal, full development was estimated to be reached 13 years after completion of construction, with a mean paddy yield of 3.8 ton/ha/crop. The yields for the 1977 and 1978 wet season crops averaged 3.4 ton/ha/crop, or about 90% of target in the first years after completion of construction. The good yield performance is largely due to the near-universal adoption of high yielding rice varieties, which had only just come on the -3- scene at the time of appraisal. As a result, project yields at full develop- ment have been revised to 3.8 ton/ha in the wet season and 4.2 ton/ha in the dry season. In view of the yields currently being achieved, the appraisal estimate of a period of 13 years after completion of construction to attain full development has been reduced to 7 years. 8. The increase in project cost has been compensated by a substantial rise in the world market price of rice, the principal commodity produced by the project, and by a faster rate of development, with the result that the economic rate of return of the project is now calculated at 14% compared to 13% at the time of appraisal. 9. The National Irrigation Administration, an independent agency of the Government responsible for irrigation development throughout the Philippines, implemented the project. This was the first major project undertaken by NIA and, at the same time, the largest irrigation project in the Philippines. The legislation authorizing construction of the project also permitted the estab- lishment of a project office as a discrete entity within NIA. Personnel attached to the office enjoyed improved pay and conditions of service, thus allowing the project to attract and retain the best staff available. The management of NIA was quick to seize the opportunity and used the project office as a vehicle to develop capabilities aimed at wider and longer term objectives, not immediately associated with construction of the project. NIA's current, excellent project preparation and farmer support capabilities were created and nurtured within the project organization. The project office provided all the key staff for the Office of Special Projects (OSP) established in 1975 to handle the large foreign assisted irrigation program that followed on NIA's successful performance on the UPRP. The procedures and monitoring systems developed on the project have been adopted throughout the OSP and are currently being implemented in the other sectors of the NIA. II. MAIN ISSUES A. The Lack of Resettlement Planning and its Consequences 10. The PCR recounts the problems encountered while trying to relocate the population living within the reservoir area (PCR, paras. 4.06 to 4.08). Since this is one of the Bank-supported projects that has encountered serious resettlement problems, some additional discussion seems to be warranted. 11. The PCR mentions some of the problems that developed due to the lack of analysis of the resettlement component at appraisal: protracted negotiations with the population to be resettled; ten-fold increase in reset- tlement costs and lack of economic viability of the resettled farmers. In addition, two other problems are worth mentioning: (a) When resettlement was first contemplated, about 1,500 fami- lies (some 8,000 persons) were to be relocated; eventually, 2,300 families (13,000 people) were relocated. On first sight, this may have been just a difference between previous estimates and actual figures, but several NIA -4- officials told the audit mission that additional families had moved into the reservoir area after it became known that the project would be launched in order to become eligible for expected compensation. (b) It was important for NIA and the contractors to have the reservoir area cleared of inhabitants within the shortest possible time. As a result, NIA had no leverage on the population; its negotiating stance vis-a-vis the population weakened as the reservoir filling date approached. Since they had to be moved out of the reservoir area, they dictated the condi- tions. Eventually, NIA had to accept almost any condition. The Government decided to fill the reservoir 10 months earlier than originally scheduled!!. But instead of weakening NIA's leverage even further, the earlier dam closure helped in convincing the Pantabangan inhabitants the Government was serious indeed regarding area inundation and population resettlement. As a result, they became more amenable to discussing the main issues and a workable solution was produced in time. 12. The most serious of the conditions involved in the solution con- cerned the choice of the resettlement site. NIA had planned to resettle the evacuated population in the service area, where relocated farmers would benefit from the project's irrigation facilities. But when they turned down the offer and insisted on being resettled adjacent to the old urban center (Pantabangan, for which both the dam and the reservoir were named), NIA had to give in and resettle them beside the inundated old town. 13. This situation is an interesting contrast to cases where agencies have imposed an inappropriate plan on the resettled population; in this case the Government's proposal, which was sound and well conceived, was turned down by the people involved. Actually, the people resettled cannot earn a liveli- hood in the reservoir's catchment area. Up to now, they have earned income mainly by working on project construction. Since most cultivable soils upstream of the dam have been inundated, there is no basis for further agri- cultural development. No important mines or quarries are found in the area; no industries are likely to be established. Consequently, the resettled population has turned to logging and other forest exploitation (making fire- wood and charcoal)1/2/. Some forest or bush land is being cleared for agri- culture. Cropping is usually carried out on hilly, erosionprone terrain. Grasslands are regularly burned at the end of the dry season, exposing the steep slopes to erosion during the rainy season. Neither agriculture nor forestry activities generate adequate incomes, forcing the population to expand the ecologically destructive activities continuously. This is con- tributing to a progressive degradation of the reservoir's catchment area_'. Effects of soil erosion are already becoming apparent and some land slides have occurred at the lake sides. 1/ The reservoir's catchment area is 90,000 ha, including Canili and Diayo. Fifty-nine per cent of it is currently under primary and secondary forests; 43% is in open grassland and only 9% lends itself to cropping. 2/ Commercial logging also exists in the area. 3/ Commercial logging seems to be much more important a factor in degrada- tion. - 5 - 14. Since it is too late now to move the population out of the catchment area (although the restricted economic possibilities may accelerate emigration), reforestation seems the only viable alternative to protect the watershed and provide resettlers with employment and income. A study of watershed reforestation in the Pantabangan, Canili, Diayo and Magat reser- voirs!/ was included as part of an Erosion Control Study financed under the Chico River Irrigation project (supported by Loan 1227-PH of 1976). As a result, a first watershed management project (involving a total cost of US$75 million and a loan of US$38 million) was prepared. It aims at establishing an economically viable and technically sound system for preserving and enhancing the productive capacity of the catchments of the two largest multipurpose dams in the Philippines (Pantabangan and Magat) and would provide a permanent and improved livelihood for the local population. Negotiations were substantially completed in late April; Board presentation is currently scheduled for July 1980. A summary of the proposed project and loan is attached (see Attachment I to the PPAM). 15. Coordinating problems at institutional level compounded the pro- blems created due to the hasty resettlement. NIA is responsible for operation and maintenance of the dam and reservoir and, by implication, also in charge of watershed management2/. But reforestation is the responsibility of the Bureau of Forestry Development, while the area around the reservoir lake is the responsibility of the Ministry of Agrarian Reform (because of the resettlement project). Entrusting administrative control of catchment areas of dams under construction - or already completed - to one agency would be more desirable. This issue should be discussed with the Government in connec- tion with irrigation projects under preparation_. 16. The lessons learned from the belated resettlement have been put to a good use. In projects recently started, the population subjected to re- location had been contacted and a census4/ was undertaken at an early stage. Possible resettlement sites are now jointly inspected and alternatives dis- cussed. Resettlement plans are prepared along with the engineering plans. But no matter how well conceived plans are, they need to have the people's support in order to be successful5/. 1/ The Canili and Diayo reservoirs provide the Pantabangan reservoir with additional water, which is used to irrigate the Penaranda area (Loan 984/Credit 472-PH, Aurora-Penaranda project). The Magat project is assisted by another Bank loan (Loan 1567-PH, Magat River Multipurpose II project). 2/ Most of the watershed area is publicly owned land but squatters have occupied large tracts of it. The Government has no enforcing capacity to keep the catchment area clear of intruders. 3/ An Inter-Agency Committee, chaired by NIA, has been established for the Magat project. It is still too early to judge its impact, performance and effectiveness. 4/ I prevent later population "increases" by the influx into the project aroa of "compensation-seeker" families. 5/ NEDA has suggested that from the planning stage of projects requiring compulsory resettlement, the people concerned be consulted and involved (see Attachment 2, page 5). -6- 17. The Bank, too, has learned lessons from the Upper Pampanga experi- encel'. Current operational procedures require resettlement studies and plans as part of every agriculture, power or other project involving com- pulsory resettlement. However, an opportunity for further understanding the sociological aspects of this project was missed. To expand its understanding of the sociological dimensions of involuntary resettlement the Bank should consider ways for systematically reviewing past project experience. Partic- ularly - but not exclusively - it seems advisable to include in future project completion missions dealing with resettlement issues appropriate expertise. B. Water Management 1. Training 18. NIA is training farmer group leaders: 3-day courses have been attended by 160 of them (about 5% of the 3,000 existing leaders2/). Selec- tion of trainees is made on the basis of proposals submitted by the water management technicians. There has been no formal follow-up evaluation of this training program. NIA field technicians, however, informed the audit mission that in their judgment units with trained leaders have better maintained canals and higher 0 & M collection rates; more problems are already resolved at the unit level; NIA instructions are better followed; the units are more open to cooperation with NIA officials; and farmer meetings show better farmer attendance and are more effective. This experience should be followed up and evaluated to test the subjective findings and draw the necessary conclusions for other projects. 2. Planning and Operation 19. Up to now, water available has been more than adequate to satisfy all farmer requirements. But the reservoir has not been designed to secure full water supplies in the dry season for each and every year. Thus, sooner or later a year will come when total water inflows will be insufficient to meet all requirements. No policy or criterion has been defined for this certain event - a serious ommission in the audit's view because it would be much easier to define allocation criteria prior to an emergency, when nobody is yet affected, rather than when the urgency of the problem would not allow adequate time for considering various alternatives. There seems to be no consensus among NIA officials on the recommended course of action because different solutions were mentioned to the audit mission: (i) reduce water allocation to every farmer on an equal basis; this would allow all farmers to have some area under crops and, thus, have some employment and income. 1/ As well as from similar cases, like the Sobradinho dam in Brazil (a component of the Paulo Afonso IV Power, supported by Loan 1008-BR). 2/ Farmer organizations start with Irrigation Units of 10 ha, led by a unit leader. Five units (50 ha) constitute an Irrigators' Group, which commands a rotation area; the chairman of the rotation area is elected by and from the five unit leaders. About ten groups constitute a Federation of Irrigators' Groups, with about 200 - 250 farmers and covering 500 to 600 ha. -7- (ii) allocate all water available to farmers located in the upper reaches of the system canals; this would minimize water losses (maximizing the system's conveyance efficiency) and thus maximize the area to be irrigated per cubic meter of water released from the dam. (iii) allocate all water available to, first, crops other than rice and, then, to the best paddy lands, on the basis of the available land classification; this would maximize production and, most certainly, the added value per cubic meter of water actually delivered to the fields. (iv) some areas - particularly at the lower reaches of the canals - are permanently inundated during the wet season, and only during the dry season can farmers get a crop (as compared with two crops per year elsewhere); these areas should get full water allocation first to allow these farmers to have at least a single crop, the remaining water being allocated elsewhere. This approach would cause resistance by the upstream farmers, who would object to water bypassing their lands while their fields remain dry. Each possible solution has merits, either on economic or social grounds. This will become an important issue which needs to be discussed and settled at the earliest, well ahead of the first "rationing" year.!V 3. Operation and Maintenance; Water Charges 20. The system has been effectively operated up to now, but water con- veyance and use efficiency is still low (PRC, para. 6.06). Water losses in the river are about 10% between the Pantabangan dam outlet and the Rizal diversion dam (about 5 km downstream, at the head of Diversion Canal No. 1, which carries about 60% of the total water released to the service areas2/), and 15 - 20% between there and the Pampanga-Bongabon diversion dam (about 25 km further downstream, at the head of Diversion Canal No. 2, which carries the remaining 40% of the irrigation waters). Canal conveyance efficiency (from the main diversion points down to the final farm outlets) is about 65%3 . Overall conveyance efficiency, including losses in the river, is 55%. Farm level efficiency is estimated at 78% in the dry season and lower in the wet season. Thus, overall water use efficiency of the system is about 43% during the dry sqason and can be estimated to be between 33 and 40% during the wet season. 1! The Upper Pampanga project has a feature which will make an eventual year of water shortage less painful for the farmers: the water shortage will be known well ahead of starting planting in a dry season - even before nurseries have been prepared. Then, a dry season crop will be missed, but no expenditures and work will be lost. 2/ Upper Pampanga project water only; it excludes water for Penaranda. 3/ Results of NIA studies show conveyance losses ranging from 1.5 to 13.6%/km at major canals, and from 2.4 to 28.5%/km at laterals and sub-laterals. - 8 - 21. Regular canal maintenance (weeding, trimming slopes, etc.) is carried out manually by the ditch tenders. Desilting is done by heavy equip- ment. However, equipment is in short supply: only 50% of needed gradalls, graders and cranes and about 30% of dump trucks required are available. There is no shortage of equipment operators. Regular gate maintenance and repair is carried out by the ditch tenders.1/. 22. Operation and maintenance (0 & M) is currently estimated to cost the equivalent of about 5 cavans of paddy (250 kg) per hectare, of which about 70% is for operation and 30% for maintenance. Water charges have been grad- ually raised over the years. Currently they stand at 6 cavans (300kg)/ha (PCR, para. 6.07) or its equivalent in pesos, thus more than covering 0 & M costs. It is interesting to note that in the Philippines water charges are legally set in cavans of rice and not in monetary units. This has been a very important factor in keeping water charge amounts in real terms in an in- flationary period. 23. Collection of water charges continues to improve. Before project completion, available water could irrigate only 10 - 15% of the dry-season area. Even during the wet season, water was sometimes insufficient for the fields along the canals' lower reaches. Narrow canals plus the continuous flooding system also worked against downstream farmers. Now that farmers receive assured water supplies almost directly from canal outlets, NIA is in a much better position to impose and collect water charges. Further, double cropping has generated additional farmer incomes, thus making water charges less burdensome. And, in addition, NIA did not have any workable billing system before; paying was, to a large extent, left to the farmers' discretion. NIA had no way of knowing who had received water, let alone how much, and payment was the responsibility of - frequently absentee - landowners. The Agrarian Reform Act has made actual operators responsible for paying 0 & M charges and NIA has an updated roster of irrigators. As a result of the management consultant's recommendations, formal procedures for billing and collection have been set up (PCR, paras. 4.04 and 6.07). Bills are prepared at district level (one billing clerk per 2,500 - 3,000 ha) and hand-delivered to all resident irrigators; bills are mailed to absentee farmers. A payment schedule is agreed upon at bill delivery. Collectors visit farmers who did not keep up with the agreed payment schedule-I. 1/ Most damage to gates is caused by accidental hitting. There are, however, some cases of intentional damage. Typically, the stem of an open outlet gate is bent to prevent the gate from being closed. With plenty of water, current incidence is about 10% (1 case per section per year), but the rate of occurrence is likely to increase whenever water will become scarce and rationed. No problem has been reported on the main gates. 2/ Each district has only one or two full time collectors, but NIA assigns water masters, WH1Ts and ditch tenders to this task during the 2-month collection period. - 9 - As a result, the collection rate has improved; it was 60% in the 1979 dry- season (ending April 30, 1979)1/. This percentage is still too low and NIA is making efforts to improve it2/. 24. The law gives farmers the option of paying water charges in cash or kind. Up to 1978, most payments were in cash, because market prices for paddy were above the official support price (at which paddy is received for water charge payment). But in 1979 market prices fell to P0.80/kg, well below the support price of rl.30/kg, and most farmers started paying in kind. Nobody had anticipated this switch to paddy payments when the law was enacted. NIA lacks scales, moisture meters, sacks, silos or other storage facilities to handle delivered paddy. This makes collection an almost insurmountable and unmanageable problem. Some farmers deliver wet paddy to NIA, which is then stacked in storage rooms, office space, meeting rooms, corridors, even in the open air. NIA is supposed to transfer the paddy to the National Grain Author- ity (NGA), which has its own problems handling paddy under the strong selling pressure from farmers. In November 1979, when the wet-season collection started, some 500 tons (10,000 cavans) of paddy out of the 2,600 tons (52,000 cavans) collected during the dry season were still with NIA. NIA was antici- pating the delivery of some 7,500 tons (150,000 cavans, 60% of total collect- ibles) of paddy before the end of January 1980. C. Optimistic Targets for Double Cropping 25. Appraisal targets for achieving double cropping were over-opti- mistic (PCR, para. 5.01). Two factors were overlooked: (i) Most farmers had insufficient working capital to switch to double cropping immediately and it took some time for the credit system (formal as well as informal) to provide them with the additional working capital required. (ii) Double cropping requires adherence to a strict calendar, but farmers were following no crop- ping discipline. Therefore, farmers had to be compelled to normalize their cropping schedule, and this also took some time. 26. First experience in 1975 and 1976 with schedule normalization was quite poor. The 1976 wet season was extended for almost 10 months, instead of the hoped for 6 months3/. Since the period left for dry season was sub- stantially shortened, a second crop was no longer possible. NIA was forced to take a bold action: it shut down the whole irrigation system for three months (throughout the rest of what would have been the 1977 dry season), compelling all farmers to start the 1977 wet season planting at the same time. Farmers were informed in advance that water supplies would be suspended, but they 1/ Up from 23% in 1976 and 30% in 1978, but still below the 76% reached in 1977. These figures refer to dry and wet seasons. 2/ Downstream farmers usually have better payment records; they are actually acknowledging NIA's water management which prevents water monopolization by upstream irrigators. 3/ Two factors led to this protracted cropping season. One was the lack of farmer discipline, since they were still used to deciding by themselves when to transplant. The other was heavy typhoon damage during the second and third month of the wet season; a majority of farmers made a desperate effort to replace damaged crops by replanting -the paddies with new seed- lings (see Attachment 2, item 3). - 10 - were not consultedl/. Foregone benefits were substantial and led to farmer complaints and protest rallies. No compensation for foregone production was granted. Water was released again in June for the 1977 wet season crop, which was grown over a seven months period, still a bit excessive but considerably less than the ten months of 1976, permitting a dry season crop in 1978. System operations and double cropping have been smooth up to now; indications are that farmers pay increasing adherence to the required planting and harvesting schedules. D. Labor Constraints 27. In contrast with other irrigation projects elsewhere, labor avail- able was sufficient to meet increased requirements of expanded production. This has not been without some adjustments, though. Prior to the project, house and machinery maintenance and repair, off-farm jobs, sports and other recreational activities, vegetable cropping, etc. took up most of the farmers' time. It was not easy for farmers to re-schedule or abandon all these activi- ties and work in the fields during a second season. There were some labor- management problems, too. Manpower for planting as well as for harvesting had to be mobilized for a shorter and more intensive period; threshing had to overlap with land preparation, etc. 28. The rigid cropping calendar also led to peaking labor demand at harvesting time. Even though there is no overall labor shortage, supplies are insufficient to permit every farmer to harvest his crop at the same time. Ten percent of the farmers have com lained about labor shortages at harvest time due to schedule normalization- . They had to delay harvesting/thresh- ing or extend it over longer periods, relying more heavily on family labor. Another side-effect of tight cropping schedules has been the strengthening of the position and power of local "cabecillas" who are middlemen (2 or 3 per barangay - the smallest administrative unit in the Philippines) who provide farmers with labor teams. Harvest postponement has substantial costs: some plants lodge; some grain overripens and shatters; rice quality deteriorates. Thus, yields and income are lower than they would have been with a more timely harvest1'. 29. Although less important, there are also some labor problems during transplanting time. Seven percent of the farmers complained about that. Lack of working capital to pay the hired hands is more serious at this time than at harvest, when they can be paid in kind (a share in the crop harvested). Yields are also affected by late transplanting: older seedlings show lower yields. 1/ Actually, not even the normalized cropping schedule was ever discussed with the farmers. NIA discussed it with several Government agencies in the agricultural field, but not with people actually affected, because farmers were not yet fully organized (see Attachment 2, item 3). 2/ Information obtained through the Input-Output Monitoring Program (IOMP). For details on this M & E system see OED, Built-in Monitoring and Evalua- tion Units: Third Review, currently under preparation. 3/ All harvest labor problems discussed in this paragraph refer to the wet season only. Dry season harvesting has been mostly free of labor problems: the overall area is slightly smaller; fields can be easily drained; labor productivity is higher (longer days, no stoppage for rains and firmer ground, which makes transportation easier). - 11 - E. Allocation of Observed Benefits Among Different Projects 30. The Upper Pampanga project poses an evaluation problem: how to allocate benefits, which are generated jointly or concurrently by two or more projects, to their different causal factors. The solution adopted in the attached PCR is appropriate in that it makes every effort to avoid double counting of benefits: the "without-project" situations (both before and after the project) try explicitly to take into consideration the effects of the other projects. Only the incremental residue is attributed to the project and incorporated in the cash flows used in computing its rate of return. However, some inconsistencies -- and a kind of "collective" double counting of bene- fits -- occur when several concurrent projects are being evaluated (i.e., the addition of all those "residues" adds up to more than the total benefits to be attributed to the different factors involved). This problem is further discussed, though not resolved in the following paragraphs. 31. The Upper Pampanga project provides expanded or improved irrigation to 78,400 ha. The project benefits are the expanded production resulting from increased cropping intensities, higher yields and - to a minor extent - higher crop security. But expanded production is also partly attributable to other projects or programs: two Bank-supported credit projects; a Government- financed support program, and the development of high-yielding rice varieties by the International Rice Research Institute (IRRI - an international insti- tute partly financed by the Bank). 32. All these projects and programs have been evaluated by the Bank and others - and each has been found successful. (i) The performance audits of the Second and Third Rural Credit projects recalculated the ERR at 32% and 44%, respectivelyl/. Several of the rural banks participating under each project are located in the Nueva Ecija province and, particularly in the UPRP's service area (e.g. Cabanatuan, Talavera, Rizal, Santa Rosa). (ii) In 1973, the Government launched the Masagana-99 Program, setting up a delivery system for HYV seeds, credit, fertilizer and other inputs. An official Govern- ment evaluation gave credit to the Program - which operated also in the UPRP service area - for substantial yield increases of paddy, with fertilizers being singled out as the most significant factor contributing to the higher yields2/. The evaluation concluded that the Masagana-99 Program could be directly identified for attaining and sustaining rice self-sufficiency in the Philippines since 1976.2/ (iii) The same could be said of IRRI breeding the high yielding new rice varieties. All evaluations of the international research effort concluded that the HYVs have been extremely successful. The high yields of the Upper Pampanga project have been largely attributed to the very rapid adoption by farmers of HYVs, which had just been released at the time of appraisal (PCR, para. 5.02). 1/ PPAR - Philippines: Second Rural Credit Project, Report No. 1277, August 18, 1976; and PPAR - Philippines: Third Rural Credit Project, Report No. 2784, December 27, 1979. 2/ J. C. Alix - The Impact of the Masagana-99 Program on Small Farmer Produc- tion and Income in the Philippines, Ministry of Agriculture, Bureau of Agricultural Economics, Research Report No. 11, November 1978. 3/ However, the Masagana-99 program peaked in 1974 and 1975. Inputs dis- tributed and the credit volume have been declining since, while benefits of the Upper Pampanga project are starting to materialize. - 12 - 33. Computing the "correct" rate of return for each of the four programs or projects - the Upper Pampanga project included - would call for multifactor methods of economic evaluation, instead of the mono-dimensional method cur- rently applied by the Bankl. CPS has issued some instructions - and warn- ings as well - on the methodologies to be utilized when separate project com- ponents are interdependent (CPN No. 2.01 of June 27, 1977; paras. 34 to 38). It appears worthwhile for CPS to explore further now the case of inter- dependent and mutually reinforcing projects - some of which may not be Bank- supported - and issue appropriate guidelines for this particular case2./. F. Farm Size and Number of Farmers 34. There is no solid data base to determine the exact number of farmers and the average farm size in the project area since there has been no census recently. The number of "farms" billed for water use is not a reliable indication either, because of fragmentation of farm holdings. The best estimates are based on data provided by MOA's Bureau of Agricultural Economics and by NEDA's Bureau of Census and Statistics. In the audit's view, the best estimate is 35,900 farmers in the project area, with an average holding of 2.1 ha. That number is 50% larger than the appraisal estimate (24,000 farmers), and the increase is almost fully explained by 3.5% annual "vegeta- tive" increase in the number of farmers as a result of splitting up larger farms through inheritance, sale or succession at the time of farmer retirement or decease. The figure quoted in the PCR (45,500; para 5.03 and Table 6) seems to refer to number of farms instead of number of farmers. Agrarian reform led many farmers to subdivide their holdings between relatives, thus increasing the number of farms, but it had little or no effect on the number of farmers in the area. Actually, all agrarian reform beneficiaries were already there as share-croppers, tenants, lease-holders, etc. 1/ Similar problems were discussed in the audit reports on other irrigation projects. See, for instance, PPAR, Indonesia: (First) Irrigation Re- habilitation Project (Credit 127-IND), Report No. 2047, May 4, 1978, paras. 24 to 34, and PPAR, Sri Lanka: Lift Irrigation Project (Credit 121-CE), Report No. 2801, December 28, 1979, para. 24. Another case calling for simulation or partial equilibrium models for evaluating water regulation projects is mentioned in PPAR, Thailand: Sirikit Dam and Chao Phya (First) Irrigation Improvement Project. Report No. 2850, March 10, 1980. 2/ As long as no guidelines are provided, PCRs should explicitly mention that the benefit stream utilized for re-estimating the project's rate of return include (or may include) benefits already credited to another project 3/ Foregone production formerly obtained in the area inundated by the reservoir should have been deducted from the project benefits (PCR Tables 7 and 8). - Attachment-1 Page 1 PHILIPPINES WA771'SHED MANAGE!ENT AND EROSION CONTROL PROJECT Loan and Project Summary Borrower: Republic of the Philippines Amount: $38 million equivalent Terms: 20 years, including 5 years of grace, with interest at 8.25% per annum Project Description: The project addresses the problems of the watersheds of the two largest multipurpose dams in the Philippines. It aims to establish an economically viable and technically sound system for preserving and enhancing the productive capacity of the catchments and provides a permanent and improved livelihood for the local population. Project components include: (a) reforestation in the two watersheds; (b) con- struction and maintenance of forest roads; (c) construction 'of charcoal ovens; (d) provision of equipment and vehicles for forest protection and fire prevention; (e) a feasibility study of the Magat watershed; (f) pilot programs in range management and smallholder agroforestry in the Magat water- shed; (g) initiation of forest protection pilot programs in two provinces in Northern and Central Luzon, and (h) provision of consulting services for (a), (e), (f) and (g) above. In the Pantabangan area, the project would provide full- time employment to 4,710 people by 1985, increasing to 6,780 people by 1995. Labor employment, together with revenue- sharing, would increase per capita incomes of these families from about $75 in 1979 to about $565 in 1990. In the Magat area, the project would provide full-time employment to about 1,200 people by 1985, increasing to 1,990 people in 1992. Employment provided by the project would increase the per capita income of these families from an average of $100 in 1979 to about $275 by 1985. The forest protection pilot component would improve the lives of about 300 forest occupant families. No major technical risks are anticipated as the project includes a wide range of tree crops and end products to minimize dependence on any single product. NIA has already demonstrated substantial capabilities in the S14 - Attachment 1 Page 2 development of nurseries, tree planting operations and protection of seedlings from fire in both project water- sheds. Two factors that could affect the success of the project are: the ability of NIA's project management to adapt to a commercial viewpoint, and the full involvement of the local population in project implementation. These factors are being addressed in the project by the provision of technical assistance to strengthen the commercial and production aspects of the project and to strengthen NIA's already good public information services and by providing incentives for participation in the project. Risks associated with weaknesses in the administrative capacity of the Bureau of Forest Development (BFD) are being reduced by the estab- lishment of a detailed operation plan and an intensive monitoring and evaluation mechanism for the Forest Protection Pilot component. Estimated Local Foreign Total Cost: /1 -- ($ million) ---- Plantation establishment 15.7 7.7 23.4 Forest roads construction 3.3 3.8 7.1 Forest roads maintenance 0.5 0.8 1.3 Forest protection & fire prevention 3.2 1.2 4.4 General administ-ation 4.7 3.6 8.3 NIA support services 1.1 - 1.1 Magat feasibility study 0.8 1.1 1.9 Magat pilot programs 2.0 1.0 3.0 Forest protection pilot component 2.1 1.2 3.3 Total Base Cost 33.4 20.4 53.8 Physical contingencies 3.1 1.6 4.7 Expected price increases 11.5 5.0 16.5 - - Total-Project Cost 48.0 27.0 75.0 Financing Plan: Local Foreign Total ------ ($ million) IBRD 11.0 27.0 38.0 Government 37.0 - 37.0 48.0 27.0 75.0 /l Excluding customs duties and taxes. Attachment 2 -15 - Page 1 Comments Received from the Government of the Philippines Republika ng Pilipinas abnr anig pIanga5ffiuann Ng Rltuhig (NATIONAL IRRIGATION ADMINISTRATION) (TANGGAPAN NG TAGAPANGASIWA) Lungsod ng Quezon May 16, 1980 Mr. Shiv S. Kapur Acting Director General Operations Evaluation Department The World Bank 1818 H. Street, N.W. Washington D.C. 20453 U. S. A. Dear Mr. Kapur, We thank you for furnishing us a copy of the OED Project Performance Audit Report on the Upper Pampanga River Project (UPRP). After reviewing the report, our staff has recommended that further clarification of the following issues be mades 1. There are actually other factors that contributed to the delay in the construction of irrigation facilities in the UPRP service area which are not mentioned in the report. Among these, the most significant was the frequent suspension of work in the construction of terminal facilities, to give way for the delivery of irrigation water to farmers. UPRP authorities have to revise construction schedules quite often to accommodate this contingency, as they do not want to deprive farmers in the UPRP service area of their main source of livelihood. Paragraph 3 (under Project Summary) of the Audit Report should include this important fact regarding the delay in the com- pletion of the UPRP irrigation facilities. 2. In paragraph 6.04 of the attached PCR, the delay in the construction of irrigation works was attributed to less attention given by NIA and the consultants to this phase of the work. There seem to be no clear basis for this, as after the completion of the dam in 1974, the attention of UPRP authorities were shifted to work in the service area and resettlement sites. Attachment 2 -16 - Page 2 NIA has concentrated four (4) construction divisions (one in San Jose, two in Cabanatuan and one in Gapan aside from an agricultural division in Mufioz compared to only two divisions in the damsite) to oversee cons- truction in the service area. However, these efforts to bring about the early completion of irrigation facilities (and possibly duplicate the construction feat at the damsite) had to be relaxed, as construc- tion activities were rescheduled for a much longer period to allow for construction shutdowns during delivery of water. Poor contractor performance and delays in procurement have also contributed to the delay in the completion of the terminal facilities. For the sake of accuracy, we suggest that paragraph 6.04 be corrected accordingly. 3 Another issue is the shutting down of the whole irrigation system in 1977 (paragraph 26 of Audit Report) which resulted in substantial losses in foregone benefits and complaints and protest rallies by farmers. During the wet season of 1976 (June to November) Central Luzon was visited by numerous typhoons which caused widespread destruc- tion to paddy plants. Majority of the farmers made a desperate effort to replace damage crops by replanting the paddies with new seedlings although the wet cropping season was already on its second or third month. This resulted in the extension of the wet season to nine (9) months, instead of the usual 6 months. Since the period left for thedry season crop was substantially shortened, a second crop was no longer possible as it would go beyond June and disrupt cropping schedules for 1977. NIA stopped the delivery of water for three months and took advantage of this lull to complete the construction of terminal facilities in the service area. The loss of one cropping season to farmers could not be attributed to the shutdown as this was dictated by adverse weather conditions in 1976. Consultation with the farmers regarding the shutdown was not possible as they were not yet fully organized into Samahan Nayons and therefore, had no representatives who could discuss with NIA the problems of irrigation Attachment 2 17 - Page 3 water delivery. In the absence of representatives, practical approach was to inform the farmers thru notices. Except for the above issues, we would like to express our full accord to the report. We therefore hope that the clarifications cited will be reflected in your final conclusions. We also would appreciate it very much if you could furnish us a copy of your final conclusions. Very truly yours, IO0 ESTU A tin dmi trator EGF/ils - 18 - Attachment 2 Page 4 N'A "WAL 1A _)V\ 1( IN'\; N I W i1 A 4 June 1980 Mr. Shiv S. Kapur Acting Director-General Operations Evaluation The World Bank 1818 H. Street, N.W. Washington, D.C. 20433 U. S. A. Dear Mr. Kapur: Re: Project Performance Audit Report * on Philippine-Upper Pampanga River Irrigation Project (Loan 637-PH). We thank you for the opportunity in reviewing the Project Performance Audit Report which you have prepared for the Upper Pampanga River Irrigation Project as this provides us with useful insights which we could apply in our post-evaluation activities. We have reviewed the report and found its contents generally acceptable. There are however a number of comments and observations which we would like to mention for your consideration in drafting the final version of the report. Attached is a copy of our comments on the resettle- ment issue, the reservoir problem, the economic analysis prepared, and the performance of contractors. Very truly yours, ANTONIO M. LOCSIN Deputy Director-General Attachment 2 Page 5 Comments on "The World Bank Project Performance Audit Report: Upper Pampanga River Irrigation Project (Loan 637-PH)" A. RESETTLEMENT ISSUE It was pointed out that though the National Irrigation Administration had planned to resettle the evacuated population in the service area the people turned this down and insisted on being resettled ad- jacent to the old urban center which could not pro- vide them with economic opportunities. As such the people turned to logging and other forest exploita- tion activities thereby causing and/or contributing to the progressive degradation of the reservoir's catchment area with actual cases of soil erosion and land slides. Therefore, from the above, it may be gathered that the resettlement issue is not only a sociological problem. The consequence effect of resettlement may be forest destruction and therefore a future siltation problem, like that encountered in the Chico River Irrigation Project. The resettlement issue experienced in this project has made the Bank to require the preparation of resettlement studies and plans as part of every agricultural, power or other project involving com- pulsory resettlement. However, we would like to point out here that no matter how well conceived plans are, they need to have the people's support in order to be successful. And in this case it is suggested that from the planning stage of projects requiring compulsory resettlement, the people concerned be consulted and involved. B. RESERVOIR PROBLEM The water management issue points out that the reservoir has not been designed to secure full water supplies in the dry season for each and every year and that the over-all water conveyance efficiency is a low 55 per cent (paragraphs 19, 20). As this pro- ject was conceived to provide water for two crop pro- ductions annually, if these problems are not settled Attachment 2 - 20 - Page 6 at the earliest time, then definitely production out- put may decrease at some point in time in the future. There is a question then, as to why this prob- lems were not foreseen at planning stage, or even during the time the project was implemented in order that changes in design may have been done to rectify the problem. This thus points out to the need for more planning for project operation and maintenance when the project's feasibility study and detailed engineering are prepared. C. ECONOMIC ANALYSIS No sensitivity analysis was prepared for in- crease in production cost (Project Completion Report). With inflationary trends, there may be unusual in- creases in production cost which increases in the price of rice may not be able to offset in order to derive a constant rate of return. It is also suggested that the economic analysis consider the shift in the payment of irrigation fees from money to cavans of rice as there are economic costs in inventory. D. PERFORMANCE OF CONTRACTORS The poor performance of contractors caused considerable delay (30 months behind schedule) in the completion of irrigation facilities. This problem can be traced to the period when contracts were awarded to firms which turned out to be incapable of successfully completing the items on conditions stipulated in the contract. More care, therefore, and much thorough counter-checking or a more rigid system of screening should be made before pre- qualifying contractors to participate in public bidding. - 21 - PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT 1/ 1. SUMMARY- 2. BACKGROUND 2.01 In the mid-1960s the Government launched an all-out effort to achieve self-sufficiency in food production, particularly of rice and corn. The main emphasis of the program for increasing rice yields was on the introduction of high-yielding varieties and modern agricultural techniques. However, an expansion of the area under improved varieties was dependent on the availability of a reliable water supply. Most of the existing irri- gation systems were incapable of delivering an assured water supply and except for the Angat system of 32,000 ha, none had storage facilities for dry season irrigation. The urgent need to expand the irrigated area was recognized by the Government which assigned high priority to irrigation development under the four-year program for the country, covering the period ending in June 1970. 2.02 The irrigation development program was based on two approaches. The first consisted of the rehabilitation and upgrading of existing systems to permit greater control of water deliveries and higher utilization effi- ciencies, thus allowing a firming up of wet season supplies and some increase in dry season cropping without providing additional water. The second approach involved the construction of storage to significantly increase 1/ Included as a part of the audit memorandum. -22 - the area under dry season irrigation. The first major step in developing dry season irrigation was the construction of the Angat dam in Central Luzon and rehabilitation of the 32,000 ha irrigation system served by the dam. Construction of the dam was financed with Bank assistance (Loan 297-PH) as part of the Second Power Project and was completed in 1967. Rehabilitation and improvement of the service area was carried out initially with USAID support and subsequently expanded and completed with Asian Development Bank (ADB) assistance. 2.03 The Upper Pampanga River Irrigation Project (UPRP), is located in Central Luzon, the largest rice producing region in the Philippines, which contains about 20% of the area planted to rice annually and accounts for about 27% of total national production. The Government's drive for increased rice production gave priority to development of three traditional rice bowls, Central Luzon, the Cagayan valley and Mindanao, with primary emphasis on Central Luzon. The project was the first Bank-assisted irrigation project in the Philippines and has since been followed by a further nine irrigation projects, as well as by two rural development projects with substantial irrigation components. Recent projects in Central Luzon, besides the one discussed in this report, are the Aurora-Penaranda Irrigation Project (1974) which utilizes the storage reservoir and main canal system of the UPRP, and the Tarlac Irrigation Systems Improvement Project, a rehabilitation and expansion project along the western edge of the UPRP. 3. FORMULATION 3.01 In order to provide a basis for expanding irrigated agriculture the Government initiated comprehensive studies of seven of the country's major river basins in 1962. Technical assistance in carrying out the studies was provided by the US Bureau of Reclamation (USBR). A study of Central Luzon, produced in 1966, listed a series of water development projects in the basins of the Pampanga and Agno rivers and recommended that priority be given to the multipurpose development of the Upper Pampanga river. The recommendation was accepted by the Government, and the National Irrigation Administration (NIA) with USBR assistance prepared a feasibility study advocating the construction of the Pantabangan dam on the Pampanga river for irrigation and power. The Government requested a Bank loan for the project in 1968. 3.02 In order to reduce the costs of the project, the Government proposed several modifications to the preliminary plans shown in the feasibility study. The National Power Corporation's (NPC) review of the power aspects of the feasibility study cast some doubts on the benefits of power generation. It was agreed at appraisal that provisions would be made in the project for a more detailed evaluation of the power potential of the project. The addi- tional study, completed in 1970, supported the original findings and recommended the immediate installation of a 100 MW power plant. The recommen- dation was eventually implemented under the Sixth Power Project (Loan 1034-PH) in the wake of the 1973 oil crisis. 3.03 Two nontechnical difficulties were apparent at appraisal. Under the laws prevailing at the time, international competitive bidding (ICB) would have been precluded. Prior to negotiations, legislation was approved - 23 - by the Congress of the Philippines waiving the restrictive provisions and allowing ICB. The second problem concerned the frequent difficulties experi- enced by development projects in obtaining adequate funding. Before negotia- tions the Congress authorized the financing of the local costs of the project through the sale of Government bonds and releases out of gen ral revenues. To ensure a timely and continuous flow of funds for the project, the estab- lishment of a special fund, into which the Government deposited every month the local currency requirement for the next three months, was made a con- dition of loan signature. This provision proved very successful and was a feature of nearly all subsequent irrigation projects. 4. IMPLE4ENTATION Project Description 4.01 The project as planned included provision of a dependable water supply for year round cultivation of 72,000 ha, and an additional area of 5,000 ha during the wet season only and flood protection to lands lying below the dam site. At appraisal the project included: (a) construction of an earth and rockfill dam and reservoir at the Pantabangan site; (b) rehabilitation of existing irrigation systems serving about. 46,000 ha; (c) construction of new irrigation systems to serve about 31,000 ha; (d) construction of feeder roads and operation and maintenance facilities; (e) a study of NIA's operations, structure and financial operation by management consultants; (f) land classification of the project area; and (g) a feasibility study of possible future power development at Pantabangan dam. Design Changes 4.02 At appraisal, the Pantabangan dam was designed as an earth and rockfill embankment rising 110.5 m above the riverbed, with a crest length of 920 m. In addition to the main embankment, seven small embankments were proposed to fill deep ravines along the left abutment ridge. The dam would create a reservoir with a total storage capacity of nearly 3,500 MCM. During construction, the river was to be diverted through two tunnels, one of 4.25 m diameter, and the other of 2.25 m diameter. These tunnels were to be converted to an irrigation outlet and a power tunnel respectively after completion of the dam. The spillway, consisting of an inclined shaft and a 10 m tunnel, was designed to pass 1,960 cu m/sec, the outflow resulting from a peak reservoir inflow of 12,000 cu m/see, assuming 1,650 MCM of storage - 24 - would be available at the time of the flood. During detailed design, the above proposals were changed. The crest of the dam was lowered by 3.5 m to avoid flooding the nearby town of Caranglan. The dam, as finally designed, consisted of a main embankment across the Pampanga river, an auxiliary embankment across the Aya creek, and a saddle dam connecting the two embank- ments. These changes resulted in a 30% increase in embankment material and a consequent increase in the cost of the dam. Lowering of the crest by 3.5 m reduced the reservoir storage capacity from 3,500 MCM to 3,000 MCM. The spillway was moved to a downstream location with more favorable geologic conditions and it was redesigned as an open chute, fed by three 8 x 10 m gated bays and one 15 m wide ungated bay. The spillway capacity was increased to 4,200 cu m/sec to cope with a peak design flood inflow of 12,000 cu m/sec routed through a full reservoir. During construction of the dam, the river was diverted through two 9 m diversion tunnels. One of these tunnels was later converted into an outlet tunnel, and the second into a power tunnel. 4.03 In order to meet the increased irrigation requirements in the service area and to allow for future expansion, the diversion capacities of the existing Pampanga-Rizal and Pampanga-Bongabon diversion weirs were increased. The crests of these weirs and river training banks were raised, canal inlets and sluiceways were increased and old gates were replaced. The irrigation service area of the project, as built, totaled 78,400 ha, consis- ting of 46,000 ha of rehabilitation and 32,400 ha of new construction. This constituted a small increase of 1,400 ha over the appraisal target. A detailed study of water supply and demand, carried out in conjunction with reservoir operations studies for power generation at Pantabangan, showed that an area substantially larger than estimated at appraisal could be served, assuming a gradual increase in irrigation efficiencies. In the course of construction, NIA provided capacity in the main canals and laterals and built a number of structures which would allow some 5,000 to 6,000 ha to be even- tually added to the project area. All canals with discharges in excess of 5 cu m/sec were provided with 6 m wide gravel-surfaced 0&M roads on both banks, while smaller canals were provided with a 3.5 m wide gravel road on one bank. Gates and water measuring devices were incorporated in all headstructures, and orifice type turnouts were furnished for each 30-40 ha. Farm ditches were constructed to deliver water from the lateral canals to areas of about 10 ha each. This relatively intensive system of ditches and drains, subsequently adopted in all future Bank-financed irrigation projects constructed by NIA, was new to the Philippines. It permitted the introduction of rotational irrigation to 10 ha units, compared to the previous practice of continuous flow irrigation, with field to field flooding of relatively large areas of land. 4.04 The construction of the dam was completed nearly 10 months ahead of schedule, despite significant changes in its design. There were delays, however, in the construction of irrigation facilities due to procurement problems, poor performance of the contractors engaged on the irrigation works, and interruptions of work to allow delivery of irrigation water to farmers. This delayed completion of these works to the end of 1977, two and a half years behind schedule. NIA's organization was modified a number of times to deal with the implementation and operational requirements of the project on the recommendation of Sycip, Gorres, Velayo and Co. (SGV), the management consultants employed under the loan. NIA, with assistance from - 25 - the USBR, completed in 1971 a land classification study of 135,000 ha in the project service area. The feasibility study of power generation, carried out in 1970 by NIA and its consultants (EDCOP-ECI), recommended the installation of two 50 MW generating units, construction of the Masiway re-regulation dam, and a 230 kV transmission line connecting the power plant with the National Power Corporations's (NPC) grid at Munoz, Nueva Ecija. It was implemented under Bank loan 1034-PH. Also included in the project during implementation was a feasibility study of the contiguous Aurora-Penaranda project, subse- quently financed by the Bank Group in 1974 (Loan 984-PH and Credit 472-PH). Procurement 4.05 Procurement of equipment and all civil works, except for construction of on-farm facilities, was subject to international competitive bidding. On-farm facilities were built by NIA. The contract for construction of the dam (US$27.8 million) was awarded to a consortium of Philippine construction companies, Hydro Resources Contractors Corporation, who successfully completed the job nearly one year ahead of schedule. At the request of NIA, the Bank agreed to an amendment of the loan agreement to allow for disbursement for the full foreign exchange cost of initial purchases of imported equipment by the contractor up to a total of US$6 million. This innovation proved very successfully in accelerating contractor mobilization and has become a standard feature in all subsequent Bank-assisted irrigation projects in the Philippines: Contractor performance on civil works in the service area was uneven. Large, monolithic works, such as diversion dams and major canals, were completed satisfactorily and on time, while smaller, more dispersed works were poorly handled. Contractors had no previous experience in the successful organization of such works and were more interested in the familiar and easier road construction work which was readily available at the time. Four contracts for construction of irrigation works had to be terminated due to excessive delays, in spite of NIA's efforts to screen out inexperienced contractors in their prequalification procedures. The work was completed by force account. Resettlement 4.06 At project inception, neither NIA nor the Bank realized the magnitude or complexity of the tasks involved in resettling the population displaced by the reservoir formed by the Pantabangan dam. None of the preappraisal studies addressed the resettlement issue, and its impact was not evaluared at the time of project appraisal, although a provision of US$1.1 million was made for relocation in the appraisal cost estimate for the dam. A total of 2,308 families, or 13,000 people, was displaced by the dam. Of these, 187 families were resettled on irrigated rice lands in the dam's service area and the remaining 2,116 families-were, by their own choice, resettled in four newly developed sites on the hills along the southeastern rim of the reservoir. 4.07 The affected population, mainly located in the small town of Pantabangan, complained that they had not been informed of the need for relocation prior to promulgation of the laws authorizing the project and did not agree with NIA's plans to resettle them downstream of the dam in the irrigation service area. The issue generated a great deal of emotion and was exploited for political purposes. To obtain agreement of the population, NIA complied with the demand that resettlement be within the municipal boundaries of the flooded town, so as to preserve its political identity. - 26 - Despite setbacks caused by bureaucratic delays in payment of compensation for flooded land, and the need to hastily evacuate about 1,000 families due to partial dam closure 10 months ahead of schedule, the resettlement program was physically successful. Settlers were provided with new houses on easy terms, potable water, electric light, schools, community buildings and excellent roads. NIA attempted to introduce cottage industries, such as weaving, pottery and mushroom culture, to supplement the poor farming potential of the hilly terrain, but these ventures did not succeed. World Food Program aid was obtained as a transitional measure and NIA provided training to the settlers to facilitate their employment on construction of the Pantabangan dam and of the nearby Canili and Diayo dams (Loan/Credit 984/472-PH) and in the current reforestation program in the watershed. 4.08 The total direct cost of resettlement, exclusive of administrative overheads, ultimately amounted to US$11.7 million, with a cost per relocated family of US$5,070. Both NIA and the Bank failed to appreciate the size of the reservoir resettlement problem and thus made no adequate plans to deal with it. The initial problem was compounded by the mistaken, if well inten- tioned, decision to accede to the population's wishes to be resettled in the nearby hills, which were not capable of providing a satisfactory source of livelihood. Final Costs and Cost Overruns 4.09 The actual construction costs of the project amounted to nearly US$128 million compared to an appraisal estimate of US$62.5 million. A more detailed comparison is given in Table 1. The appraisal estimate included a provision of 20% for physical contingencies, but no provision was made for possible price increases during implementation. In order to separate infla- tionary effects from cost overruns caused by increased quantities or higher unit prices, actual construction costs have been converted into 1969 prices to permit comparison with the appraisal estimates. This comparison (Table 2) shows that the cost overruns for the project as a whole were largely due to inflation, but it appears that the cost of the dam, and the cost of engineering and supervision were underestimated at the time of appraisal. Reporting 4.10 NIA was made aware of the Bank's reporting requirements and furnished reports acceptable to the Bank. These reports included quarterly engineering progress reports on construction of the project, semi-annual agricultural progress reports as of 1973 and input-output monitoring reports on both wet and dry season crops as of 1976. In June 1977, NIA submitted a completion report which gave a good description of the project as appraised and as built. Allocation of Loan Proceeds 4.11 The original proposal for allocation of loan proceeds was mod- ified during the course of implementation in response to the borrower's request as shown below: - 27 - Allocation of Proceeds of the Loan Category Item Planned Actual ---- US$ million ---- I Equipment 3.5 3.277 II Civil works 19.1 22.657 III Consulting services 1.5 1.862 IV Interest 5.0 6.088 Unallocated 4.9 - Cancelled 0.116 Total 34.0 34.000 5. AGRICULTURAL IMPACT 5.01 At appraisal it was envisaged that the project would have a service area of 77,000 ha, all of which would be under irrigated rice in the wet season, while only 72,000 ha would be cropped to irrigated rice in the dry season, as 5,000 ha in the Talavera area would not be able to get water from the reservoir as they were located upstream of the main canal. The projected cropping intensity of 193% was to be achieved three years after completion of the project works. The project, as constructed, serves 78,400 ha, but 2,900 ha are lower lying and subject to flooding and are not cultivated in the wet season crop, while the Talavera area that can not be cropped in the dry seaon totals 4,200 ha. Target cropping intensities are, therefore, 96% in the wet season, 95% in the dry season, and 191% annually. While dry season cropping rose rapidly from less than 10,000 ha in 1974, when storage water first became available, to 55,000 ha in 1976, the appraisal estimate of achievement of designed cropping intensity within three years proved over optimistic. Insufficient allowances were made for the magnitude of organi- zational problems involved in going from one to two crops of rice a year on such a large area and for the destructive effects of typhoons on physical structures and timetables. Thus, NIA had to forego a 1977 dry season crop in order to get the farmers back on to a manageable schedule (Table 3). A typhoon in October 1977 caused damage to irrigation works affecting some 9,000 ha, which were not repaired in time for the subsequent crop. NIA now expects to achieve the full wet season cropping target of 75,500 ha only in 1979. The dry season cropping target of 74,200 ha is expected to be attained in 1981. 5.02 Although the rate of increase in cropping intensity has been below appraisal estimates, the paddy yields achieved in the project are higher than anticipated. At appraisal, full development was estimated to be reached 13 years after completion of construction, with an average paddy yield of 3.8 ton/ha/crop and a total annual production of 566,000 tons. The paddy yield for the 1977 and 1978 wet season crops averaged 3.4 ton/ha, or about 90% of target in the first year after completion of construction (Table 4). On the other hand, only about 47% of the total target production has been achieved - 28 - to date, due to the difficulties encountered with the implementation of double cropping. The relatively good yield performance is largely due to the very rapid adoption of HYVs, which had only just made their first appearance at the time of appraisal. In view of this, the yields of HYVs at full development have now been revised from 3.8 ton/ha to 3.8 and 4.2 ton/ha for the wet and dry season crops respectively. The revised projections of yields and cropped areas are compared to appraisal projections in Table 5. The revised estimates would result in an annual increase of about 30,000 tons of paddy. The appraisal estimate of a period of 13 years after completion of construction to attain full development appears too conservative in view of the yields already being achieved and has been reduced to seven years. 5.03 The greatest change in the project area since appraisal is the increase in the number of farm families, the dramatic decrease in the size of farm and the changes in land tenure brought about under the Government's program of Agrarian Reform. At appraisal, the project area held some 24,000 farm families, with an average farm size of 3.2 ha. About 84% of the farms were between 1 and 5 ha and accounted for 70% of the land. About 70% of the farmers were tenants, the great majority being sharecroppers. By 1978 the number of farm families had risen to 45,500, the mean size of holding had decreased to 1.65 ha, about 98% of the farms were under 4 ha, covering 90% of the land, and 48% of the farms on 50% of the land had become owner operators. The sharp decrease in farm size resulted to a large extent from subdivision of farms above 3 ha to comply with the Agrarian Reform limit on farm size in irrigated areas (Table 3). 6. INSTITUTIONAL PERFORMANCE Organization for Project Implementation 6.01 The appraisal report discussed at some length the origins, functions and organization of the National Irrigation Administration (NIA), but did not go into any detail on organization of the unit within NIA responsible for construction of the project. The report stated that a Project Manager had been appointed and that he would be assisted by two Assistant Project Managers; the first would be in charge of construction in the field and the second would be responsible for administrative and technical supporting services. In the course of implementation, the project organization underwent a number of changes that reflected NIA's increasing awareness of the complexity of the job. In 1970 the UPRP organization consisted of two construction divisions, one technical division and one administrative division. By the end of 1973, when the project organization achieved its final form, there was a total of 10 divisions, of which 5 were engaged in construction. 6.02 The legislation authorizing the construction of the project (Republic Act 5499) also permitted the establishment of the UPRP office as a discrete entity within NIA. Personnel attached to the project office enjoyed better pay and conditions of service than the rest of the staff in NIA, thus allowing the project to attract and retain the best manpower available. The higher management of NIA was quick to seize the opportunity and used the UPRP office as an umbrella to develop capabilities aimed at wider and more distant objectives, not immediately associated with the construction of the project. A case in point was the creation within the UPRP organization of the Project Development Division (PDD) which was assigned the function of preparing - 29 - feasibility studies for future irrigation projects. The PDD, which has since become a Department in NIA, has prepared, or been closely associated with the preparation of nearly all of the 11 irrigation projects financed by the Bank since 1973. A further example was the establishment of an Agricultural Development Division (ADD) which was originally responsible for land classi- fication, water use studies and some farmer training. The scope of work has increased enormously to include training of water management staff, coordi- nation of agricultural supporting services and input-output monitoring. The ADD has also become a Department and its responsibilities have spread far beyond the limits of the UPRP. 6.03 A further illustration of the "umbrella" functions of the UPRP office was the assignment to it in 1974 of responsibilities for construction of the Bank-assisted Aurora-Penaranda and Tarlac projects. At the same time, the NIA charter was revised under Presidential Decree No. 552 which gave the agency much broader powers and authority over a wide field of activities. An immediate result of the Decree was the creation of the Office of Special Projects (OSP) which was given responsibility for all foreign-assisted projects being implemented by the agency. The UPRP office provided all the nucleus staff for the OSP and the procedures, methodologies and monitoring systems first worked out at UPRP have been adopted throughout the OSP and are currently being implemented in the other offices of the NIA. 6.04 The main focus of interest of NIA and the consultants was on the construction of the dam and appurtenant works, which were completed ahead of schedule. Construction of the irrigation works, particularly the smaller terminal facilities, received less attention. Besides the delays mentioned above (see paragraph 4.04), quality of work was less than satisfactory. An inventory of terminal facilities carried out by the ADD in 1976 showed that a considerable amount of work was required to bring them to design standards. The work included a large number of farm turnout gates, laterals and drains which needed to be completed or rehabilitated. Remedial measures were completed in mid-1977 at a total cost of P 11.8 million (US$1.6 million), of which 58% was spent on minor canals and laterals and the balance on farm delivery facilities. The staff of the ADD, not being responsible for either construction or maintenance, but having a keen interest in operations, proved to be well suited to make the inventory and to perform a useful quality control function. Agricultural Extension 6.05 The appraisal report called for the complement of extension workers servicing the project area to be raised from 50 at the time of appraisal to about 390 by 1975. The increased emphasis on Agrarian Reform in 1972 and the Masagana 99 campaign in 1973 to foster increased rice pro- duction resulted in a faster than anticipated expansion in the number of extension workers in the UPRP. By 1977 some 420 people were engaged in extension. In 1976 an Agricultural Development Coordinating Council (ADCC) was established under the chairmanship of the Provincial Governor to coor- dinate the work of the various agencies providing supporting services to the project area farmers. NIA supplied the secretariat of the council from ADD staff. The ADD staff was also instrumental in getting the council to prepare a detailed five year agricultural development plan for the UPRP, which has - 30 - proved very useful in setting targets. Under the provisions of Loan 1080-PH training for Water Management Technicians (WMT) was set up in the ADD-operated training center at Munoz and on three 500 ha pilot training areas in the UPRP. The aim is to have a trained WMT, an agriculture or agricultural engineering graduate, to every 500 ha of irrigated rice. The WHT would be primarily concerned with water control, but would also help with extension and with promotion of farmers' organization. Over half the WMTs required for operation of the UPRP have been through the one year training and are ope- rational. A system of monitoring inputs and outputs was instituted by the ADD in 1976 with consultant assistance under the provisions of Loan 1367-PH. The program has had considerable success in identifying potential problems before they became critical, as well as pinpointing the main themes for the extension effort. Operation and Maintenance 6.06 At the completion of construction the project was renamed the Upper Pampanga River Integrated Irrigation System (UPRIIS) and placed under an Operations Manager, responsible to the SPO. The project service area was divided into three irrigation districts, each with an appropriate complement of engineers, water management technicians and canal and ditch tenders. This organization operates and maintains irrigation and drainage systems from headworks to turnouts serving areas of 30 to 40 ha. Supplementary farm ditches serving areas of 10 ha were built under the project and their operation and maintenance is the responsibility of the irrigator groups. NIA has been active in organizing the groups, but the degree of response has been uneven and one of the main functions of the newly trained WMTs is to generate greater farmer interest in the irrigator groups to achieve better understanding and a degree of partnership between the irrigators and the agency. NIA has established a Water Control Coordinating Center (WCCC) located at Cabanatuan City, which is responsible for preparation of the irrigation programs and coordination and monitoring of water releases from the reservoir, taking into account the irrigation, power and flood control requirements. The WCCC, together with the Input-Output Monitoring and the ADD, has already had a considerable degree of success in improving the discipline of operations in the service area in response to the more critical timing required by the objective of double cropping. Water efficiencies, particularly in the wet season, remain low. Both the NIA staff and the farmers are still familiarizing themselves with the rotational irrigation system and it will be some time before they become fully at ease with the technique. Cost Recovery 6.07 Republic Act No. 3601 establishing NIA gave it the authority to "collect from the users of each irrigation system constructed by it such fees as may be necessary to finance the continuous operation of the system and reimburse within a period of not less than 25 years the cost of construction thereof". The appraisal report indicated that prevailing rates were in- sufficient to cover O&M costs, collection rates were very poor and that all such collections were paid in to Government general revenue funds and did not accrue to the agency. The Act establishing the UPRP allowed water charges to be retained by NIA for O&M expenses. The management consultants employed - 31 - under the project made recommendations for improved billing and collection procedures, which resulted in a higher rate of collection of water charges. The appraisal report suggested that irrigation charges be raised gradually to P 100/ha per annum within five years of completion of constrt tion and ultimatel, to F 150/ha within 10 years of completion of construction. The loan agreement gave no target figures or dates for water charges, but adopted the RA 3601 minimum cost recovery period of "not less than 25 years" without any maximum limit. In 1975, NIA obtained the President's approval to a nationwide rise in irrigation fees and to their denomination in terms of cavans (50 kg) of paddy. The rates were set at 5 cavans/ha annually for double cropped land, compared to the 6.1 and 10.7 cavans/ha suggested in the appraisal report. During negotiations for the Aurora-Penaranda project the Government agreed to raising of water rates on the UPRP within five years of project completion to 7.9 cavans/ha/annum for double cropped land. Since July 1, 1975, NIA has been charging the beneficiaries of UPRP at an interme- diate rate of 6 cavans/ha/annum. Consultants 6.08 The project englaeering consultants were of great assistance to NIA, particularly on the dam component of the project and on the feasibility studies for the follow-on Aurora-Penaranda project and the power aspects of Pantabangan dam. The consultants developed excellent working relations with NIA and other Government agencies and contributed significantly to the smooth and relatively problem free implementation of the project. A notable feature of the consultants' work was the degree of confidence and competence developed in the project counterpart staff. 7. ECONOMIC EVALUATION Project Benefits 7.01 At full development in 1984, the annual net value of production in the project is expected to reach US$78.2 million before costing labor. After deducting the imputed labor cost and an estimated net value of production for the without project situation, the incremental net value of production would be US$50.7 million. Expected future yields, cropping intensity and production are discussed in Chapter 5 and summarized in Table 7. Farmgate prices and production costs, expressed in mid-1978 values, have been based on current prices for the period 1974 to 1978 and thereafter on the Bank's commodity price forecasts for the period 1979 to 1990. Project Costs 7.02 The total economic cost of the project is US$188.6 million. Annual expenditures in pesos were converted to dollars using the average exchange rate for the year. The costs were then adjusted to mid-1978 values using - 32 - the index of international inflation. All local costs were multiplied by the Standard Conversion Factor (SCF) of 0.9. The annual operation and maintenance cost, estimated at US$30 per hectare, is US$2.5 million. 7.03 On the basis of the labor requirements per hectare for the conti- guous Penaranda irrigation project (SAR No. 348a-PH), the incremental annual labor demand at full development is expected to be 7.32 million man-days, equivalent to 30,500 full-time jobs. With 85,670 land holding and landless families in the project area in 1975, an expected rate of 2.5% annual growth in population, and an available 40 man-days/month/family, the monthly labor supply at full development is expected to be 4.3 million man-days. The peak labor demand of 2.0 million man-days, occuring in November when harvesting and land preparation overlap, would be 47% of the available labor supply. For the economic analysis, labor has been shadow-priced at P 2.7/man-day with the project and P 2.4/man-day without the project, based on the Magat project, where cropping intensity is comparable though labor is scarcer (SAR No. 2150-PH). This would amount to an incremental labor cost at full development of US$3.0 million. Economic Analysis 7.04 Using the foregoing costs and benefits and with a project eva- luation period of 50 years, the economic rate of return is 14%, compared to 13% at appraisal in 1969. The project's increased cost has been offset by the higher price of rice and by the expected achievement of full development seven years after completion of construction rather than the 13 years assumed at appraisal. Sensitivity of the rate of return was tested to reduction in yields, prices and cropping intensity and to delay in achievement of full development yields. In no case did the rate of return fall below 12% (Table 8). Farm Incomes 7.05 Farm budgets for owner-operators and leaseholders were prepared for farms of 1.0 ha, representing 44% of the farmers and 30% of the area, and 3.0 ha, representing 21% of the farmers and 35% of the area. For a 1.0 ha farm, per capita income, expressed in mid-1978 prices, would increase from an average US$60 in 1969 to US$165 in 1984 for an owner-operator, and from US$50 to US$115 for a leaseholder. For a 3.0 ha farm, per capita income would increase from an average of US$95 in 1969 to US$420 in 1984 for an owner- operator, and from US$70 to US$255 for a leaseholder. At full development about 60% of the farmers would have incomes above the per capita absolute poverty level of US$190. While the incomes of the remaining 40% would increase as a result of the project, they would remain below the poverty level due to the extremely small size of holdings (Tables 9-10). - 33 - 8. BANK PERFORMANCE 8.01 Appraisal of the project was carried out over ten years ago. In evaluating the Bank's performance consideration must be given to the fact that this was the first Bank-assisted project of its type in the Philippines, and was also the most extensive and ambitious irrigation project planned up to that time by a relatively unproven agency of the Government. Looking back, it is apparent that both the Bank and the Government did not realize at the time of appraisal the magnitude and long-term implications of resettlement of the population displaced by creation of the Panbabangan reservoir. An indication of this is the absence of any mention of the problem in the appraisal report, except for the inclusion in the appraisal estimates of an item for relocation of US$1.1 million, compared to an actual cost of US$11.7 million. 8.02 Supervision of the project was timely and satisfactory. During the initial stages of the project, the Bank concentrated primarily on the major engineering works, because these works were critical to the success of the project. However, the issue of resettlement was picked up by the first supervision mission in 1971 and became a major subject of attention thereafter. In the later stages of the project the Bank turned its attention increasingly to the problems of operating the project and integrating its agricultural and engineering aspects. The series of follow-on irrigation loans beginning in 1974 proved excellent vehicles for the introduction to the project of such operations and production oriented innovations as water management training and input-output monitoring which were first identified in the course of supervision. The Bank was instrumental in persuading NIA to review its procurement procedures for civil works in order to attract the interest of small and medium contractors. The resDonse to the initiative started under the project has been most encouraging and has shown up clearly in the greatly increased participation of small contractors in the implementation of the subsequent projects. 9. CONCLUSIONS 9.01 This first major irrigation project in the Philippines was success- ful from the agricultural, engineering and institution building aspects. The rapid expansion of dry season cropping, although slower than expected at appraisal, is still very gratifying, as is the accelerated attainment of levels of yield well in advance of the anticipated timetable. As the single largest irrigated area in the country, the project area has been one of the principal contributors to the recent attainment of the national goal of self- sufficiency in rice. Through construction of the dam and major structures and advances for equipment purchase the project had made a substantial and significant contribution to the development of large scale contracting capability in the Philippines, while at the same time laying the foundation for the strengthening of small local engineering contractors by providing numerous small civil works contracts in the project area in small packages. - 34 - .02 The project was a major factor in developing NIA's technical and anagerial capability to undertake large-scale engineering works, with alatively limited short-term consultant assistance in specific technical reas. In addition, the project has contributed substantially to a widening f NIA's perceptions of its role in irrigation development from a rather imited engineering function, to a more active interest in farmers' organization, gricultural production, input mobilization and watershed protection. The gency has changed from a narrow engineering organization to an imaginative, ide ranging development institution. - 35 - Table I PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Final Cost and Appraisal Estimate (US$ million) Appraisal Estimate /a Final Cost Total Local Foreign Total Local /b Foreign I. Pantabangan dam and reservoir 1. Dam and structures 23.6 8.3 15.3 43.58 15.33 18.25 2. Access road and structures 0.8 0.2 0.6 1.46 0.37 1.09 3. Power line 0.6 0.1 0.5 0.07 0.01 0.06 4. R.O.1. 5.7 5.7 - 4.54 4.54 - 5. Relocations 1.1 1.1 - 11.70 11.70 6. Reservoir clearing 0.5 0.5 - 1.09 1.09 - 7. Service facilities 0.6 0.4 0.2 2.78 1.85 0.93 Subtotal 32.9 16.3 16.6 65.22 34.89 30.33 II. Irrigation systems and roads 1. Modif. existing div. dams and headwork structures 0.5 0.3 0.2 1.12 0.67 0.45 2. Div. canal #2 and irrigation Area III 1.2 1.2 - 0.30 0.30 - 3. Completion of irriS. system Area III 1.3 0.8 0.5 5.34 3.29 2.05 4. Completion of div. canal #2 0.6 0.4 0.2 0.93 0.62 0.31 5. Div. canal #1 2.3 1.1 1.2 1.91 0.91 1.00 6. Rehab. irrig. system Areas I, II and IV 4.9 2.5 2.4 16.07 8.20 7.87 7. Constr. irrigation system Area V 3.4 1.9 1.5 7.87 4.40 3.47 3. Roads 4.8 2.2 2.6 4.02 1.84 2.18 9. Perm. proj. buldings 0.2 0.2 - 2.21 2.21 - 10. Irrig. proj. serv. fac. 0.4 0.2 0.2 0.73 0.37 0.36 Subtotal 19.6 10.8 8.8 40.50 22.81 17.69 III. Engineering, supervision & administration 5.8 5.8 - 17.96 17.96 - IV. Operation and main- tenance equipment 2.0 0.2 1.8 1.62 0.16 1.46 V. Consulting services 2.2 0.4 1.8 2.64 0.48 2.16 Total 62.5 33.5 29.0 127.94 76.30 51.64 /a Including 20% physical contingencies. No provision made for prtce contingencies. /b The division into local and foreign currencies based on the proportion used in appraisal estimates. - .16 - Table 2 PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Comparison of Appraisal Estimate with Actual Costs Costs Cost Overrun Appraisal In constant Quantities & Item estimate Actual 1969 $ Inflation unit prices --- (US$ million)----- (%) ------- 1. Pantabangan dam and reservoir 32.9 65.23 41.17 73 25 2. Irrigation system and roads 19.6 40.50 19.92 105 2 3. Engineering supervision and administration 5.8 17.95 10.15 134 75 4. O&M equipment 2.0 1.62 1.16 - - 5. Consulting services 2.2 2.64 1.81 - - Total 62.5 127.95 74.21 86 19 - 37 - Table 3 PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Actual and Projected Areas Cropped (ha) Wet season Dry season Rainfed Irrigated Irrigated Actual 1969-73 29,500 46,000 5,000 1974 29,500 42,510 9,315 1975 28,400 47,100 26,900 1976 9,560 58,550 50,100 1977 11,400 64,100 - 1978 2,260 23,000 65,000 Projected 1979 - 75,500 68,000 1980 - 75,500 71,000 1981 onwards - 75,500 74,200 - 38 - Table 4 PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Actual and Projected Paddy Yields (ton/ha) With project Without project Wet season Dry season Wet season Dry season Rainfed Irrigated Irrigated Rainfed Irrigated Irrigated Actual 1974 1.6 2.2 2.8 1.6 2.2 2.5 1975 1.6 2.3 2.7 1.6 2.1 2.4 1976 1.7 2.7 2.1 1.7 2.3 1.8 1977 1.7 3.5 - 1.7 2.3 2.6 1978 1.7 3.3 3.4 1.7 2.3 2.6 Projected 1984 - 3.8 4.2 1.9 2.6 2.9 - 39 - Table 5 PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Paddy Yields and Areas at Full Development/1 Appraisal Estimate Revised Estimate Area Mean Yield Total Yield Area Mean Yield Total Yeild (ha) (ton/ha) (ton) (ha) (ton/ha) (ton) Wet Season 77,000 3.8 292,600 75,500 3.8 286,900 Crop Dry Season 72,000 3.8 273,600 74,200 4.2 311,640 Crop TOTAL 149,000 566,200 149,700 598,540 /1 Full Development estimated at 13 years after completion of construction at appraisal now revised to 7 years after completion of construction. - 40 - Table 6 PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Farm Size Distribution in 1978 Farm Size Farms Area (ha) No. % ha. % < 0.5 5,191 11 1,885 2 0.5-1.5 20,074 44 22,287 30 1.6-2.0 9,835 22 17,883 24 2.11-4.0 9,499 21 26,063 35 4.1-7.0 773 2 4,058 5 > 7.0 178 75,026 100 Total 45,550 100 75,026 100 Mean Farm Size 1.65 PHILIPPINES UPPER PAMIPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Net Value of Production at Full Development - 1984 /a Farm-gate Gross value Production Net value of Net value of Area Yield price of production . costs production production (ha) (ton/ha) (P/ton) ---------------- (P/ha) ---------------- (P million) Wet Season Irrigated rice W 75,500 3.8 1,285 4,885 1,230 3,655 276.0 W 46,000 2.6 1,285 3,340 840 2,500 115.0 Rainfed rice W - - - - - - - W 29,500 1.9 1,285 2,440 675 1,765 52.1 Dry Season Irrigated rice W4 74,200 4.2 1,285 5,395 1,320 4,075 302.4 W 5,000 2.9 1,285 3,725 915 2,810 14.1 - P million - Total net value of production (before costing labor) 181.2 578.4 Less imputed cost of labor /b 18.4 40.4 Total net value of production 162.8 538.0 Net incremental value of production 375.2 (US$50.7 m) /a W - With project W - Without project lb Labor requirements estimated at 100 man-days/ha with project and 95 man-days/ha without project. Labor shadow priced at P 2.7/man-day with project and P 2.4/man-day without project. -42- Table 8 PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Economic Costs and Benefits (US$ million) Project Costs Benefits Year Capital O&M 1969 1.15 - 1970 4.85 - 1971 30.24 - 1972 31.93 - 1973 42.08 - 1974 35.39 1.6 1975 23.21 10.1 1976 11.24 13.0 1977 5.61 2.2 8.4 1978 2.87 2.4 21.1 1979 2.5 32.7 1980 2.5 36.3 1981 2.5 39.9 1982 2.5 43.5 1983 2.5 47.1 1984 2.5 50.7 1985 2.5 51.5 1986 2.5 52.3 1987 2.5 53.0 1988 2.5 53.8 1989 2.5 54.6 1990-2019 2.5 55.4 Economic rate of return: 14% Sensitivity Analysis Alternative Rate of Return (%) Base line 14 10% decrease in projected yields 12 20% decrease in world market price of rice 12 15% decrease in projected cropping intensity 13 Five-year delay in reaching full development 13 - 43 - Table 9 PHILIPPENES UPPER PAMPANCA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Farm Incomes in 1969 (Mid-78 Prices) Rainfed Irrigated 1.0 ha 3.0 ha 1.0 ha 3.0 ha Arable area (ha) 1.0 3.0 1.0 3.0 Transplanted rice (ha) 1.0 3.0 1.0 3.1 Cropping Intensity (%) 100 100 104 104 Total Paddy Production (ton) 1.2 3.6 1.8 5.5 Gross value of production (P-) 1,320 3,960 1,980 6,050 Production cost excluding labor (P)/a 345 1,035 585 1,755 Cost of hired labor (P) /b - 370 - 400 Net value of production before water charges (P) 975 2,555 1,395 3,895 water charges (P) /c - - 120 360 Net value of production (F) 975 2,555 1,275 3,535 (a) Owner-Operator Net crop income (P) 975 2,555 1,275 3,535 Per capital family income (US$) /d 60 85 65 105 (b) Leaseholder Annual Payment for land (?) Ie 290 865 440 1,320 Net crop income (F) 685 1,690 835 2,215 Per capita family income (US$) /d 50 65 55 75 /a Based on costs in Philippine Rural Infrastructure Project where yields are of same magnitude as in 1969 (SAR No. 1884a-Pf). /b Based on a maximum of 40 days/month and a total of 480 man-days/year of family labor; hired labor priced at a wage of *? 8.0/nan-day. /c Based on charge of 2.0 cavans/ha (PI 110/ha) in the wet season and 3.0 cavans/ha (P 165/ha) in the dry season. /d Based on a family size of 5.8 and includes other income of " 1,500 for a 1.0 ha farm and I 1,000 for a 3.0 ha farm. /e Based on 25% of gross value of production minus seed, harvesting and threshing costs. -44 - Table 10 PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT Farm Incomes at Full Development in 1984 (mid-78 prices) Irrigated 1.0 ha farm 3.0 ha farm Arable area of farm (ha) 1.0 3.0 Transplanted rice (ha) 1.9 5.6 Cropping intensity (%) 191 191 Total paddy production (ton) 7.6 22.9 Gross value of production (P) 9,730 29,190 Production costs, excluding labor (P) 2,905 8,715 Cost of hired labor (P)/a - 1,435 Net value of production before water charges (P) 6,825 19,040 Water charges (P)/b 470 1,415 Net value of production (P) 6,355 17,625 (a) Owner-Operator Net crop income (P) 6,355 17,625 Per capita family income (US$)/c 165 420 (b) Leaseholder Annual payment for land (P)/d 2,190 6,375 Net crop income (P) 4,165 11,050 Per capita family income (US$)/c 115 255 /a Based on a maximum of 40 days/month and a total of 480 man-days/year of family labor; hired labor priced at a wage rate of 8.0 P/man-day. /b Based on future charges of P 220/ha (3.5 cav/ha) in the wet season and P 275/ha (4.4 cav/ha) in the dry season. /c Based on a family size of 5.8 and includes other income of P 800 for a 1.0 ha farm and P 300 for a 3.0 ha farm. /d Based on 25% of gross value of production minus seed, harvesting and threshing costs. - 45 - ANNEX 1 PHILIPPINES UPPER PAMPANGA RIVER IRRIGATION PROJECT PROJECT COMPLETION REPORT List of Main Reports and Studies Prepared During Implementation 1. Report on the Study for the Integral Management of the National Irrigation Administration, Sycip, Gorres, Velayo and Co., 1970. 2. Agronomic and Economic Correlation Study on the UPRP Service Area, NIA and Bureau of Agricultural Economics, 1971. 3. UPRP Power Study, ECI-EDCOP, 1971. 4. Land Classification and Supporting Studies on the UPRP, NIA and USBR, 1971. 5. Feasibility Report on Aurora Transbasin Diversion and the Penaranda River Irrigation System, NIA, ECI-EDCOP, 1973. 6. Evaluation Report on Irrigation Planning Procedures and Scheme of Implementation for Upper Pampanga River Integrated Irrigation System, NIA, 1976. 7. UPRP Completion Report, NIA, 1977. s LBRD 10921 j.FEBRUARY 19 -~~5 ýA7-C CA SAN I~ AAN TIAN SANEASSEN AAT RRIS DIVVESSI OA M E TRIS AREA LTRS A I, c, SANTO DOMINGO ARE U DPERA S DVERSION AM - AllAGA AREA AURORA-PEÑARANDA IRRIGATlON PROJECT UPPER PAMPANGA RIVER PROJECT SERVICE AREAS EýPSTING MAIN CANAL ----PRDPOSEDMVAIN CANAL EXISTING LATER AL CREE - ---- PRDPOSED LATERA. SERVICE ARA FPRISDIVERSI 0AM SA - RSA11SE RVI CE A REAS 01F PBRIS CIVER1ONDA PAMPANGA RIVER AREA SANTO DOMINGO EXTENSION AREA BR S EXTENSION RA - S Y U E T RI0 3 0 0G T I D 1 0 0 0 03 U 1 I E L - - - VE-5 SDIETIN A