Sri Lanka - Irrigation Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 13835 PERFORMANCE AUDIT REPORT SRI LANKA VILLAGE IRRIGATION REHABILITATION PROJECT (CREDIT 1160-CE) THIRD MAHAWELI GANGA DEVELOPMENT PROJECT (CREDIT 1166-CE) DECEMBER 29, 1994 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. CURRENCY EQUIVALENTS Currency Unit = Rupees (Rs) Appraisal Years (1980/81) = US1.00 = 17.5 RS Average 1981-1990 = US$1.00 = Rs 28.2 Completion Years (1990/91) = US$1.00 = Rs 40.8 FISCAL YEAR January 1 - December 31 ACRONYMS AMDP - Accelerated Mahaweli Development Program Anicut - Diversion Channel chena - Slash and Burn, Shifting Cultivation CPS - Central Projects Staff DAS - Department of Agrarian Services ERR - Internal Economic Rate of Return FAO/CP - Food and Agricultural Organization/World Bank Cooperative Program FO - Farmers' Organization ha - Hectare ID - Irrigation Department IDA - International Development Agency IMI - International Irrigation Management Institute KRDP - Kurunegala Rural Development Project LHGs - Low Humic Gley Soils Maha - Long Rainy Season Mahaweli III - Third Mahaweli Ganga Development Project MASL - Mahaweli Authority of Sri Lanka, formerly MDA M&E - Monitoring and Evaluation MDA - Mahaweli Development Authority MDB - Mahaweli Development Board MDP - Mahaweli Development Program MEA - Mahaweli Economic Agency MECA - Mahaweli Engineering and Construction Agency, formerly MDB MMD - Ministry of Mahaweli Development MUV - Index of Manufactured Goods Unit Value MW - Mega-Watt NGO - Non-Government Organization NIRP - National Irrigation Rehabilitation Project O&M - Operations and Maintenance OEA - Office of Environmental Affairs PCR - Project Completion Report RBEs - Red Brown Earths Rp - Rupee SAR - Staff Appraisal Report SLCC - Sri Lanka Cashew Corporation Tank - Small or medium scale dam USAID - United States Agency for International Development UNDP - United Nations Development Program Vel vidane - Traditional Village Water Manager VIRP - Village Irrigation Rehabilitation Project VTR - Village Tank Rehabilitation WFP - World Food Program Yala - Short Rainy Season FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Offoin of Diretor-Gnwal Opwatom Evaluation December 29, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Sri Lanka Village Irrigation Rehabilitation Project (Credit 1160-CE) Third Mahaweli Ganga Development Project (Credit 1166-CE) Attached is the Performance Audit Report on Sri Lanka-Village Irrigation Rehabilitation Project (Credit 1160-CE) and Third Mahaweli Ganga Development Project (Credit 1166-CE) prepared by the Operations Evaluation Department. Both projects sought to raise the incomes of small farmers by the provision of irrigation water, predominantly for rice production. The Village Irrigation Rehabilitation Project was designed to repair 1,700 small scale village irrigation schemes, in the dry zone. Civil disturbances hindered implementation in seven of the eighteen designated districts after 1984; however the geographic focus of the project was shifted to the south-east without having to reduce its scope. The Third Mahaweli Project was an "integrated development" project, which built a 26 km extension to the Right Trans-Basin Canal, cleared jungle to provide irrigated farms to 16,136 farm families as well as jobs for 8,000 off-farm families. It included the provision of roads, urban centers, schools, water supply, police, etc. Both projects took twice as long as projected to complete, in part due to the impact of the Accelerated Mahaweli Development Program as a whole, which Government acknowledged would stretch the nation's irrigation implementation capacity. The intended economic impact of both projects was adversely affected by the unforeseen, and in the Audit's view unforeseeable, dramatic drop in the world price of rice. Attachment 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.  FOR OFFICIAL USE ONLY PERFORMANCE AUDIT REPORT SRI LANKA VILIAGE IRRIGATION REHABILITATION PROJECT (CREDIT 1160-CE) THIRD MAHAWELI GANGA DEVELOPMENT PROJECT (CREDIT 1166-CE) TABLE OF CONTENTS Page No. PREFACE ............................................................... BASIC DATA SHEET .....................................................ii EVALUATION SUMMARY ................................................ vii 1. BACKGROUND AND SETTING ......................................... 1 A. Introduction ................................................ 1 B. Economic Development ........................................... 2 C. Setting and Context .............................................. 2 2. IRRIGATION AND MAHAWELI GANGA DEVELOPMENT ................... 3 A. The Mahaweli Ganga Development Program ............................ 3 B. The iest ui the irngation Sztor ..................................... 4 3. VILLAGE IRRIGATION REHABILITATION ................................ 5 A. Project Formulation .............................................. 5 B. Project Implementation ........................................... 7 C. Project Administration ............................................ 8 D. Project Impact .................................................... 9 Sustainability .................................................. 9 The Economic Rate of Return .................................... 13 4. THIRD MAHAWELI GANGA DEVELOPMENT ............................ 16 A. Project Formulation ............................................... 16 B. Project Implementation ............................................ 17 C. Project Administration ............................................. 19 D. Project Impact ................................................... 19 Sustainability ............................................. 20 Environmental Impact ......................................... 20 Economic Rate of Return ...................................... 21 This report was prepared by Wilfred Candler, Task Manager. Pilar Barquero provided administrative support. 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. 5. MAIN FINDINGS AND ISSUES ........................................... 22 A. The Role of Farmer's Organizations .................................. 23 B. Low Internal Economic Rates of Return ............................... 23 C. Special Situations ................................................ 24 D. Erroneous and Different Internal Economic Rates of Return ................. 25 TABLES 1.1. Bank Assistance to Sri Lanka ........................................... 2 1.2. Donor Contributions to the Irrigated Sector (1967-1988) ........................ 3 2.1. Land Use in Sri Lanka, 1982 ............................................ 4 3.1. Input Costs in Kilograms of Paddy, 1980 and 1993 ............................ 14 3.2. PCR Benefit Cost Ratios for Representative Farms and the Project (FY 1991) ....... .15 5.1. Comparison of PAR and PCR Ratings .................................... 22 BOX 1. An Exceptional Problem with Riparian Rights ............................ 12 BOX 2. Rehabilitation at its Best ........................................... 13 ANNEXES I. Bank Lending for Irrigation .............................................. 27 II. C ashew s ............................................................. 29 III. Economic Cash Flow for Mahaweli III .................................... 31 IV. Revised Annex I for Part III of the Project Completion Report for the Village Irrigation Rehabilitation Project .................................. 33 V. Revised Annex I for Part II of the Project Completion Report for Mahaweli III ... ............................................... 45 VI. ERR Methodology Comparisons for Mahaweli III ............................ 51 VI.1. Bank, Borrower and Audit Views on ERR Calculations .................... 51 VI.2. Paddy Prices used by Bank and Borrower .............................. 52 VI.3. Paddy Yields used by Bank and Borrower .............................. 52 MAPS IBRD 24287 IBRD 13477R2 IBRD 15154R1 1 PERFORMANCE AUDIT REPORT SRI LANKA VILLAGE IRRIGATION REHABILITATION PROJECT (CREDIT 1160-CE) THIRD MAHAWELI GANGA DEVELOPMENT PROJECT (CREDIT 1166-CE) PREFACE This report presents the results of the Performance Audit Report of the Village Irrigation Rehabilitation and Third Mahaweli Ganga Development projects for which Credits 1160-CE and 1166-CE in the amounts of SDR 24.5 million (US$30 million equivalent) and SDR 73.3 million (US$90 million equivalent) respectively were approved on June 16 and 23, 1981. The credits were provided to the Democratic Socialist Republic of Sri Lanka. The Credit Agreements were signed on July 15, 1981 and November 5, 1981. The credits became effective on September 24, 1981 and February 8, 1982. The credits were closed on December 31, 1990 and May 12, 1992, after four and five year extensions respectively. Remaining amounts of credits which were canceled were SDR 6.9 (28.2 percent) and SDR 14.2 (19.3 percent) million. The Performance Audit Report (PAR) was prepared by the Operations Evaluation Department (OED), which has reviewed the PCR, the Staff Appraisal and President's Reports for the two projects, the legal documents, Bank files and the transcripts of the meetings of the Executive Directors at which the credits were approved and discussed. The projects have been discussed with the Bank staff involved. An OED mission visited Sri Lanka in May 1994 and discussed the project experience with the borrowers, and other involved agencies and beneficiaries. The cooperation and assistance of these officials is gratefully acknowledged. The Audit found that the PCRs gave an accurate account of the projects implementation and experience; however a computational error was found in the internal economic rate or return (ERR) in the Village Irrigation Rehabilitation Project, and the wrong estimate of the ERR had been included in Part II of the Third Mahaweli Ganga Development Project. The PCR for the Mahaweli project questioned the sustainability of the Accelerated Mahaweli Development Program. The Audit takes a more optimistic view of the Government's political commitment to the Program and the progress in getting farmers to contribute labor to system maintenance. The Audit agrees with the PCR for Village Irrigation on the key role of Farmer's Organizations in the sustainability of rehabilitated schemes, but cautions against assuming that this can be easily achieved. Following standard procedure, the draft PAR was sent to the Borrower for comment. No comments have been received.  iii PERFORMANCE AUDIT REPORT SRI-LANKA VILLAGE IRRIGATION REHABILITATION PROJECT (CREDIT 1160-CE) BASIC DATA SHEET KEY PROJECT DATA item Appraisal Actual or Actual as % of Expectation Current Estimate Appraisal Estimate Total Project Costs SDR million 35.60 22.02 62 US$ million (Equivalent) 43.60 31.49 72 Credit Amount SDR million 24.5 US$ million (Equivalent) 30.0 Cancellation 6.9 SDR million 8.5 US$ million (Equivalent) Disbursed 17.6 72 SDR million 24.7 82 US$ million (Equivalent) Modest Institutional Performance CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS Fiscal Year 82 83 84 85 86 87 88 89 89 90 Appraisal Estimate SDR 2.5 5.3 9.4 15.1 21.6 24.5 US$ million (Equivalent) 2.8 5.5 9.2 16.6 26.4 34.8 Actual SDR 1.0 2.6 3.5 6.1 9.5 12.3 14.6 15.8 16.8 17.6 USS million (Equivalent) 1.1 2.7 3.4 6.7 11.6 17.5 19.7 20.8 24.0 25.1 Actual as % of Appraisal (%) 39 49 36 40 43 50 60 64 69 72 Date of Final Disbursement: May 16, 1991 Comments: The last disbursement was made on May 16, 1991, and SDR 6.92 million was cancelled. Slow and low disbursements were due mainly to slow implementation resulting from budget constraints, depreciation of the Rupee against the US dollar and SDR, and shortage of experienced accountants. PROJECT DATES Item Date Planned Date Revised Actual Date Identification 1979- October 1979b Preparation Early 1980 May 1980 Appraisal August 1980 September 1980' Credit Negotiations March 1981 March 1981 Board Approval May 1981 June 16, 1981 Credit Signing n.a. July 15, 1981 Credit Effectiveness n.a. September 24, 1981 Project Completion December 31, 1985 December 31, 1990d Credit Closing December 31, 1986 (four times) December 31, 1990 Credit Completion June 30, 1987 May 16 1991 a. Sri Lanka Aid Group Meeting. b. FAO/CP identification mission. c. Follow-up mission in February 1981. d. Substantially completed. iv sTAFF INPrIS (STAFFWEEIS) Flueal Year 80 81 82 83 84 85 86 87 88 89 90 91 Total Preappraisal 1.0 3.5 4.5 Appraisal 0.1 79.2 0.1 79.4 Supervision 0.4 17.8 7.2 12.9 7.5 17.5 13.0 11.7 15.0 8.3 18.2a 129.5 Total 1.1 83.1 17.9 7.2 12.9 7.5 17.5 13.0 11.7 15.0 8.3 18.2 213.4 MISSION DATA Mission Type Month/ Number of Days in SpeciaUlization Performance Rating' Type of Problemsd Ye Pasons Field Represente Identification (FAO/CP) November 79 3 18 IE, E, A Preparation (FAO/CP) March 80 3 23 IE, E, A Pre-appraisal Aug/Sept 80 4 21 IE, IE, E, A Appraisal January 81 2 14 E, W Supervision I July 81 1 4 E 1 Supervision 2 January 82 3 14 E, A IE 1 Supervision 3 November 82 2 10 E, IE I Supervision 4 December 83 3 14 IE, IE, E 2 F Supervision 5 February 85 2 10 A IE 2 FT Revised Rating Available Project Development Overall Fumds Management Impact Rating Supervision 6 November 85 3 12 E, A, IE 2 2 1 2 Supervision 7 July 86 2 10 E, IE 2 2 2 4 Supervision 8 February 87 2 10 E, IE 1 2 2 2 Supervision 9 February 88 3 11 E, IE, IE 2 2 1 2 Supervision 10 September 88 3 7 E, IE, IE 2 2 1 2 Supervision 11 March 89 3 5 A, IE, IE 2 2 1 2 Supervision 12 December 89 3 10 E, IE, IE 2 2 1 2 Supervision 13 July 90 4 22 E, E, IE, IE 1 2 1 2 Supervision 14 December 90 2 18 E, IE 1 2 1 2 a. Date of return to headquarters. b. A = Agriculturalist; E = Economist; W - Water Management Specialist; IE = Irrigation Engineer. c. I = Problem-free or minor problems; 2 = Moderate problems; 3 = Major problems. d. F = Financial; T = Technical. V PERFORMANCE AUDIT REPORT SRI-IANKA THIRD MAHAWELI GANGA DEVELOPMENT PROJECT (CREDIT 1166-CE) BASIC DATA SHEET KEY PROJECT DATA Item Apprisal Actualor Actua as % Expectation Curret Ftnmate Appraia Estimate Total Project Costs SDR 162.9 133.6 82 US$ million (Equivalent) 200.0 192.7 96 Credit Amount SDR 73.3 US$ million (Equivalent) 90.0 Cancellation SDR 14.16 US$ million (Equivalent) 17.39 Disbursed SDR 59.14 81 US$ million (Equivalent) 82.77 92 Institutional Performance Modest CUMULATIVE ESTIMATED AND ACTAL DISBURSEMENTS Fiscal Year 82 83 84 85 86 87 88 89 90 91 92 Appraisal Estimate SDR 16.7 35.8 45.2 53.4 61.5 69.6 73.3 US$ million (Equivalent) 20.5 44.0 55.5 65.5 75.5 85.5 90.0 Actual SDR 11.77 19.73 28.45 37.06 37.91 40.04 43.47 46.03 49.78 54.40 59.14 US$ million (Equivalent) 13.10 21.77 30.79 39.29 40.23 42.89 47.43 50.74 55.51 61.95 68.48 Actual as % of Appraisal (%) SDR 70.5 55.1 62.9 69.4 61.6 57.5 59.3 62.8 67.9 74.2 80.7 US$ million (Equivalent) 63.9 49.5 55.5 60.0 53.3 50.2 52.7 56.4 61.7 68.8 76.1 Date of Final Disbursement: May 12, 1992 PROJECT DATES Item Date Planned Date Revised Actual Date Identification 1965-1968 19651968' Preparation December 1979 December 1979b Appraisal March 1980 March/April, 1980C Credit Negotiations April 28-May 6, 1981 April 28-May 6, 1981 Board Approval December 16, 1980 June 23, 1981 Credit Signature November 5, 1981 November 5, 1981 Credit Effectiveness February 8, 1982 February 8, 1982 Project Completion December 31, 1986 n.a.d Credit Closing December 31, 1987 December 31, 1991 May 12, 1992e a. UNDP/FAO plus Sri Lankan engineers. b. Report funded by ODA. c. SAR dated May 26, 1981. d. Completion is now expected by 1994. e. With cancellation of SDR 14,163,980.96. The closing date was extended four times, and the credit account was kept open four months beyond the final closing date. Reasons for extension were: (i) security problems in the project area during 1988/89; and (ii) poor implementation performance of contractors. Vi STAFF INPUTS (STAFFWEEKS) Pre- FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 FY89 FY90 FY91 FY92 FY93 Total FY81 Preappraisal 96.0 96.0 Appraisal 97.5 85.7 183.2 Negotiations 17.7 17.7 Supervision 3.1 6.9 5.1 5.4 16.2 22.0 13.3 14.5 8.5 9.4 13.2 4.6 - 122.2 (excl. PCR) PCR 0.1 19.8 19.9 Total 193.5 106.5 6.9 5.1 5.4 16.2 22.0 13.3 14.5 8.5 9.4 13.2 4.7 19.8 439.0 MISSION DATA Mission Type Month/ Number of Days in Specialization Performance Ratingc Type of Problemsd Years Persons Field Representedb Identification (PAO/CP) May 74 Identification (FAO/CP) October 79 1 7 Identification December 79 3 17 E, AE, AC Preparation (FAO/CP) April 80 3 22 E, AC, PA Appraisal September 80 5 25 EC, E, AC, E, Ex Supervision 1 April 82 1 4 E 1 P Supervision 2 December 82 2 8 E, AE 1 P Supervision 3 January 84 1 9 E 2 M Supervision 4 February 85 2' 14 E, AE 2 M Revised Rating Available Project Developm Overall Funds Manageme ent Rating at Impact Supervision 5 November 85 3 10 E, AE, FC 1 2 1 2 Supervision 6 July 86 2 8 AE, AC 1 2 1 2 Supervision 7 January 87 2 8 AE, B 2 2 1 2 Supervision 8 October 87 2 12 AE, E 1 2 1 2 Supervision 9 June 88 3 10 AE, E, E 1 2 1 2 Supervision 10 March 89 3 10 AE, E, E 2 2 1 2 Supervision 11 March 90 2 11 AE, E 3 2 2 3 Supervision 12 October 90 2 11 AE, E 1 2 2 2 Supervision 13 June 91 4 13 AE, 2xE, BC 1 3 2 3 Supervision 14 November 91 2 11 AE, E 1 3 2 3 a. Date of return to headquarters. b. AC = Agronomist Consultant; AE = Agric. Economist; E = Engineer; EC = Economist; FC = Financial Consultant; PA = Project Analyst. b. I = Problem-free or minor problems; 2 - Moderate problems; 3 - Major problems. d. 1 = Improving; 2 = Stationary; 3 = Deteriorating. e. F - Financial; M = Managerial; P = Political; T = Technical. f. Jointly with representatives of Kuwait Fund and OECF; Ex - Extension. vi PERFORMANCE AUDIT REPORT SRI LANKA VILIAGE IRRIGATION REHABILITATION PROJECT (CREDIT 1160-CE) THIRD MAHAWELI GANGA DEVELOPMENT PROJECT (CREDIT 1166-CE) EVALUATION SUMMARY Introduction 1. This is the Project Audit Report for the ticable to implement government programs in Village Irrigation Rehabilitation and Mahaweli the affected areas. Ganga Development Projects (Credits 1160-CE and 1166-CE), approved by the Board in June 4. The Mahaweli Ganga Development Pro- 1981. They followed respectively the Tank Irri- ject (Mahaweli III) provided for (a) completion gation Modernization and Kurunegala Rural De- of the Right Bank Trans-Basin Canal (26 kin), velopment Projects (Credits 666-CE and 891- (b) construction of main and branch canals, (c) CE); and Mahaweli II and Mahaweli Technical land clearing, and leveling to settle 24,100 farm Assistance Projects (Credits 701-CE and 979- families, (d) road and social infrastructure cons- CE). The Mahaweli Ganga Development Pro- truction, (e) settlement of 7,200 non-farm fami- ject was cofinanced with the Kuwait Fund (22 lies and (f) establishment of 3,000 ha of fuel- percent) and the Japanese Overseas Economic wood plantation and 2,000 ha of cashews. This Cooperation Fund (22 percent). IDA's ap- was the archetype of an "integrated" develop- proved contribution was SDR 73.3 million (45 ment project, with the Mahaweli Authority start- percent). ing from raw jungle to provide both the productive investments and all social Objective infrastructure and even services, for a population designed to reach about 180,000. The project 2. Both projects aimed to improve small was expected to produce 186,000 additional tons farmers' access to irrigation, and thereby im- of paddy, with significant volumes of other crops. prove incomes and social conditions. Most importantly, it would provide land to the landless, and relieve population pressure. 3. The Village Irrigation Project (VIRP) provided for the rehabilitation and moderniza- Implementation Experience tion of about 1,700 small scale irrigation schemes. Some ancillary strengthening of the 5. Both projects experienced implementa- Irrigation Department and Department of A- tion delays, but achieved close to their appraisal grarian Services was also provided. Appraisal targets. The re-focussing of the VIRP, to avoid identified about 3,000 schemes in need of im- areas of civil disturbance explains a large mea- provement; 1,700 individual schemes were to be sure of that project's delay. Delay in Mahaweli selected, designed and appraised as an ongoing III can be explained in large part by the imple- part of the project activities. The initial project menting organization being overstretched, with focus on the drier Northern and Eastern areas many newly recruited staff and working with had to be changed to the wetter South and contractors from a recently emancipated private West, due to civil disturbance making it imprac- sector. In the circumstances, the SAR time es- vm timates may have been unrealistically ambitious; 10. The PCR for Mahaweli m has widely dif but they nevertheless provided useful targets. ferent estimates of the ERR as estimated by the Over-stretching of the Mahaweli Authority and Bank (4 percent)and the Borrower (14.28 per- its implementation units was implicit in the cent). Due to some mix-up in the submission of Government's political decision to accelerate the Part II, the ERR calculations used were "inter- Mahaweli program; a decision which was mediate results." Even the Borrower's final esti- endorsed by donors in their generous provision mate (11 percent) leaves a very wide gap. of assistance. 11. The ERR estimate (15.7 percent) in the Project Results PCR for VIRP involved computational errors. Calculated correctly, it is 5.6 percent. This 6. Civil unrest resulted in a major shift in throws into question the estimated ERR of 31 the geographical focus of VIRP. Despite this percent for the follow-on National Irrigation Re- unrest, 1,654 schemes were modernized or habilitation Project (Credit 2260-CE), (NIRP). rehabilitated (versus an SAR projection of 1,700), and 46,500 families were benefitted Findings and Lessons versus a projection of 25,000. However, lack of involvement of project beneficiaries in planning 12. The Environment. A draft environmental of the sub-projects, has resulted in significant impact study of the Mahaweli described Zone 4 difficulty in ensuring that they will accept as a critical habitat for several endangered and responsibility for maintaining the rehabilitated endemic fauna, one of which was confined to projects. Zone 4, and recommended that this area be in- corporated into a National Park. On learning of 7. Mahaweli III provided irrigated farms to this during Appraisal the Bank told the Govern- 16,136 settlers, versus 18,500 estimated in the ment that deletion of this area would require re- SAR (87 percent). It took twice as long to examination of the economic viability of the pro- implement as projected, and cost 30 percent ject. After further requests for reassurance from more in real terms than expected. the Bank, the consultants final report made no recommendation for the protection of Zone 4. Project Ratings Concerns expressed by the Bank's Environ- mental Advisor were not addressed. 8. Soundly conceived at appraisal and well, if slowly, executed these two projects were 13. The Role of Farmer's Organizations. unsatisfactory ex post as economic development There is now a consensus on the importance of projects due to an unforseen, and in the Audit's well-run Farmer's Organizations (FOs) to main- view unforeseeable, fall in the world price of tain distributory and field canals, and manage rice. Both projects assisted large numbers of water releases. The follow-on NIRP requires poor farmers. If such farmers were to be given that a FO be in existence and in agreement be- access to land and more reliable irrigation, then fore rehabilitation work can commence. This is despite their "unsatisfactory" categorization, plainly helpful. However, it is no assurance that these projects may well have been cost-effective. an FO will run smoothly and efficiently after the works are completed. The more project design 9. Institutional development was modest in increases the role of FOs, the less is implemen- both projects. For VIRP sustainability is tation of the project under government control. uncertain due to uncertainty with respect to the Increased recognition of the vital role for FOs, operations of FOs. Government commitment implies increased recognition of the riskiness of makes the sustainability of Mahaweli III, likely. such projects. This is not to argue against an important role for FOs; but it is to recognize Pat-I-te-R-clultin-ue-wre"ntr ix that the government (and iso facto the Bank), mers in regaining their nparan rights, or obtain loses an element of control as the role of full compensation. independent FOs expands. 17. The above two case illustrate the need 14. In the light of the changed price ratios, for continued review of project operations (not it would be difficult to design any new invest- necessarily by the Bank) even after project cos- ment or national project which would yield a 10 ing. In practice this would go beyond the percent ERR on the basis of increased irrigation monitoring of operations and development im- water for paddy production. The ongoing NIRP pact called for in the context of the newly intro- should be reviewed in the light of the lower ERR duced Implementation Completion Report, to now attributed to the VIRP. If continued, then taking remedial actions. No recommendation is each sub-project in the NIRP should have a pro- made on the basis of these two projects, but this perly calculate4 and farmer-validate4 Net is an issue which should be watched by Opera- Present Value of Benefits, prior to construction. tions and OED. 15. Special Situations. Two special situations 18 Erroneous and Different Internal Econo- encountered by the Audit, deserve comment and mic Rates of Return. An erroneous ERR was re- lead to recommendations. The appraisal seems ported in the PCR for the VIRP, and different to have misjudged the suitability of unirrigated ERRs were reported for Mahaweli In. Even land in Mahaweli III for cashews, due to the where these estimates are prepared by a consul- high degree of water-logging experienced by tant, the final product is the responsibility of the even these highly permeable soils in the Maha Operating Region (and Divisional Task Mana- (long rainy) season. The Bank should re-examine ger), who should ensure that the documentation the appraisal recommendation of sole cropping in the PCR is complete and calculations can read- cashews on unirrigated areas, giving particular ily be followed attention to the possibility of mixed cropping to take advantage of the different soil moisture re- 19. While the Bank and Borrower may well gimes on the ridges, intermediate slopes and differ, as for Mahaweli m, in their views as to gully bottoms. how the ERR should be calculated it should be possible to define these differences. Were the 16. In another case a small hydro-electric Bank and Borrower differ in their estimates of the plant had deprived a rehabilitated anicut (water ERR; an agreed Reconciliation Statement indicat- diversion channel) of waters that they had used ing the differences in methodology or data which "since the time of the kings." The Bank should lead to the different outcomes, should be included examine what it can do to hlp these small1far- in the Implementation Completion Report.  1 PERFORMANCE AUDIT REPORT SRI LANKA VILLAGE IRRIGATION REHABILITATION PROJECT (CREDIT 1160-CE) THIRD MAHAWELI GANGA DEVELOPMENT PROJECT (CREDIT 1166-CE) 1. BACKGROUND AND SETTING A. Introduction 1.1 Starting in the 1930's the Colonial Government began a low cost program of rehabilitating and repairing the ancient irrigation tanks and canals, some built over two thousand years ago. With independence this program was augmented, and overshadowed by, a more capital intensive program of new investments, often involving significant donor support. The first of these schemes was the Gal Oya multipurpose scheme, and the largest was the Mahaweli development scheme. With independence, donor also began to support the rehabilitation of existing tanks (i.e. dams) and anicuts (diversion canals). 1.2 A separate "integrated" authority, the Mahaweli Authority of Sri Lanka (MASL) was set up to develop and operate the Mahaweli scheme (paras 2.01 and 2.02). Mahaweli Ganga Development Project (Mahaweli III) was one of a series of projects designed to assist MASL. The balance of the irrigation sector, based on maintenance, improvement and extension of the smaller, but quite complex, network of village tanks and diversion canals (anicuts) was under the control of the Irrigation Department (ID) for civil works, and the Department of Agrarian Services (DAS) for operations. Independence saw an over-expansion of Government involvement with the irrigation sector. In place of self-governing village organizations, with a designated leader the Vel Vidane, the Government provided extension workers for water management, and also took on the responsibility of system maintenance. With literally thousands of village tanks and anicuts, it has become evident that the Government did not have the resources to provide proper maintenance of village irrigation systems. The Village Irrigation Rehabilitation Project (VIRP) was designed to help catch-up with some of the resulting maintenance back-log. 1.3 As shown in Table 1.1, lending for crop and livestock production, including irrigation, accounted for about 27 percent of Bank lending to Sri Lanka. Within this irrigation accounted for just under half; and within irrigation Mahaweli accounted for two thirds. The Mahaweli scheme involved hydro-power and four major dams (which were supported entirely by other donors), irrigation and social infrastructure. The Bank' was a relatively small contributor to the Mahaweli scheme: having, by 1990, provided only 6.2 percent of the funding. Put another way, donor assistance to the Mahaweli approximated total Bank assistance to all sectors. 1. Unless otherwise specifies the "Bank" is used to refer to both The International Bank for Reconstruction and Development (IBRD) and The International Development Agency (IDA). 2 Table 1.1. Bank Assistance to Sri Lanka (US $ Million) Percent Total Lending to Sri Lanka $1,863 100.0 Irrigation' $244 13.1 Mahaweli $157.5 8.5 Other $86.4 4.6 Other Agriculture $267 14.3 a. For individual projects, see Annex 1. 1.4 This audit focusses on only two of the eleven irrigation projects in Sri Lanka (see Annex I). As such it is not directly concerned with the large questions of the Bank's lending strategy,' or indeed priorities within agriculture or the irrigation sectors. B. Economic Development 1.5 From independence on, high rates of unemployment of the literate and well educated labor force have been a major concern for economic policy makers in Sri Lanka. The government's enlightened social policies, and a certain donor myopia on the efficiency implications of a overly large public sector, meant that donors were initially receptive to requests for developmental assistance, largely focussed on the public sector. Up until 1977 the Government attempted to accommodate the unemployment problem by twin policies of using irrigation to expand the agricultural frontier, and an import replacement industrialization policy with heavy public sector involvement. This proved to be a capital intensive strategy, in part disguised by the level of donor assistance. Despite this assistance, this strategy was unable to keep pace with the population increase: the resulting rising unemployment contributed both to social instability, and a change in Government. C. Setting and Context 1.6 The election of a Government which repudiated the earlier policies and was desirous of restructuring and opening the economy, was welcomed by donors, who had become frustrated at the growing inflexibility and declining productivity of the economy. The upsurge in donor assistance after 1977 is illustrated in Table 1.2. 2. See "Sri Lanka. A Break with the Past: The 1987-90 Program of Economic Reform and Adjustment," World Bank, Report 7220-CE, 1988. 3 Table 1.2. Donor Contributions to the Inigated Sector (1967-1988) 1986 Rs Million Donor Period New Const. Rehab. O&M % 1986 Rp 1993 US$ Million Million 1967-70 4,329 0 244 5 228 8 1971-74 3,848 0 442 59 2,531 89 1975-78 5,244 81 499 22 1,281 45 1979-85b 17,846 2,195 1,010 68 14,315 502 1986-88 5,995 1,441 440 63 4,961 174 a. Grants and Concessional Credit. b. Data for 1979 and 1980 is incomplete. Source: "Irrigation Investment Trends in Sri Lanka: New Construction and Beyond," IIMI, 1991, pages 31 and 60. 1.7 With hindsight, it can be seen that this was to some extent at the expense of the rest of the economy, and in particular agriculture and the rest of the irrigation sector, since other sectors did not have similarly well prepared projects "on the shelf." 2. IRRIGATION AND MAHAWELI GANGA DEVELOPMENT A. The Mahaweli Ganga Development Program 2.1 Development of the Mahaweli Ganga has been a center-piece of Sri Lankan development thinking for thirty-five years. It has been expected to provide, and in large measure has provided, employment in construction, land for the landless, food and power. Together with providing an appealing focus for foreign assistance programs. Formal work on the Mahaweli Ganga Development Program (MDP) started in 1958 with a three-year study of the water resources of the Mahaweli Ganga River, with bilateral support. A Bank mission made recommendations in 1961, and a full Master Plan was formulated by the Government in 1964-68 with assistance of a UNDP/FAO-CP team. This Master Plan envisaged irrigation of 650,000 acres of undeveloped land and 250,000 acres of existing irrigated land. As originally proposed, this development would have been in three phases, and have taken place over a 30-year period. 2.2 Implementation started in 1970. The new, more market-oriented government elected in 1977 decided to accelerate the development of the Mahaweli. This new, foreshortened Accelerated Mahaweli Development Program (AMDP) was initially expected to be completed in six years. It was the archetype of an integrated development program, with the MASL required to provide roads, schools, police force, all social services in addition to the civil works for irrigation, extension services and settler support. The Bank was a relatively modest contributor to the AMDP. Nevertheless its contribution was important for its own sake, and perhaps even more as the Bank's "seal of approval" of the Government's development strategy. 4 2.3 In these circumstances the particular portion of the AMDP financed by the Bank was not very important, since other donors could almost certainly have been found for any elements omitted from the project. Indeed, the project itself was cofinanced with the Kuwait Fund (22 percent), Japan (22 percent), IDA (45 percent) and the Government (11 percent). B. The Rest of the Irrigation Sector 2.4 Large though the Mahaweli project is, it is in total dwarfed by the myriad of small and medium sized schemes. The last agricultural census was in 1982, before Mahaweli had been fully developed. The overall land use pattern which emerged from the census is given in Table 2.1. Table 2.1. Land Use in Sri Lanka, 1982 ('000 hectares). Total Area 6,477 Agricultural Holdings 2,010 Arable 752 Asweddumized Paddy' 557 Other Annual Crops 195 Permanent Crops 975 Tea 207 Rubber 171 Coconut 392 Other 28 Pasture 20 Forest 54 Other 208 a. Land prepared specifically for cultivating paddy (bunded and levelled), but not necessarily irrigated. Source: 1982 Agricultural Census. 2.5 In 1993, the Mahaweli Scheme contributed 122,000 hectares of paddy (allowing for double cropping) grown on something under 100,000 hectares of land. Other field crops were grown on 24,000 hectares, mostly in Yala (the short rainy season) on land cropped with paddy in Maha (the long rainy season). The 100,000 or so hectares contributed by the Mahaweli scheme is new land formerly in forest; it represents a major expansion of the land available for paddy production, and is for the most part generously supplied with irrigation water. Independent studies have warned that Sri Lanka cannot afford to provide significant additional employment through schemes as capital intensive as the Mahaweli. 3. The Mahaweli produced power as well as food; and the Project Monitoring Unit (PMU) of MASL estimates the benefit-cost ratio as 1.02. "Mahaweli Statistical Handbook, 1993," PMU/MASL, Colombo 10. See also "Irrigation Investment Trends in Sri Lanka: New Construction and Beyond," IIMI, 1991. 5 2.6 Responsibility for irrigation outside the Mahaweli Project area is split between the Irrigation Department (ID) and Department of Agrarian Services (DAS). Construction and rehabilitation of all but the smallest tanks is the responsibility of the ID. Water management, operations and maintenance (O&M) and construction of very small tanks is the responsibility of the DAS. 2.7 Organizationally, this is a very untidy system, where the same function, such as determining water release policies, can be carried out in each of the three organizations: in MASL if the decision relates to the Mahaweli; in ID if it is a rehabilitated scheme that has not yet been handed over to the DAS; and, by DAS for other schemes. There is general agreement that at the distributory and field canal level, water release decisions should be made by the farmers themselves. 2.8 In practice, having more than one organization within the public service with very similar functions is a way of introducing competition. The frequently remarked difference in efficiency between unregulated public sector monopolies and a competitive private sector, involves both a change in ownership and market structure; and the change in market structure may be the more important. Certainly, in the development of Farmer Organizations (FOs), MASL and DAS have taken quite different approaches, and provide a comparison which would have been unlikely to emerge under a single monopolistic bureaucracy. 3. VILLAGE IRRIGATION REHABILITATION "Operation and maintenance of minor irrigation schemes in Sri Lanka were traditionally the responsibility of the farmers themselves, with Government support confined largely to arranging pre-cultivation meetings, resolving water disputes and undertaking major repair works. Over the past two decades, however, increased Government intervention and changes in agency responsibility for minor irrigation schemes have tended to undermine the traditional system without replacing it with an effective alternative. Furthermore, political and administrative intervention has not - at least until recently-been associated with any significant increase in resources. As a result, and given the inherent difficulties of ensuring adequate maintenance and water use practices in minor schemes, it is perhaps not surprising that physical facilities have deteriorated, that many such schemes are no longer operating and that many other are working at levels well below their agricultural potential."' A. Project Formulation 3.1 Earlier Bank lending to the irrigation sector is listed in Annex I. The immediate predecessors to the Village Irrigation Rehabilitation Project" (VIRP), Credit 1160-CE, were "Tank Irrigation Modernization Project," Credit 666-CE and "Kurunegala Rural Development Project" (KRDP), Credit 891-CE. Neither was based on a comprehensive analysis of the investment priorities within the sector, but rather each identified a set of rehabilitation investments which had acceptable internal economic rates of return (ERRs). This continued with VIRP, where the first description in the files says to the FAO/CP,..." In a nutshell the project we request your assistance is as follows: 4. Village Irrigation Rehabilitation Project, Issues Paper, 17 October 1980, para 1.02. 6 aaa. There are about 13,000 (thirteen thousand) village tanks in Sri Lanka ninety percent of which are in dry zone in Northwestern, North Central and Southeastern part of the country only fifty percent are working at present providing water at low efficiency to about four hundred thousand acres .... ."s 3.2 Initial project design was based on the tank component of the earlier KRDP. A line of credit would be established to be drawn against for the rehabilitation of 1,700 of the 3,000 tanks and anicuts' meeting project criteria, which were based on the criteria used for the KRDP. The Staff Appraisal Report (SAR) for the earlier project reported ERRs in the range 25 percent to 50 percent for its various components. Since the KRDP had only become effective in August 1979, insufficient time had elapsed for significant lessons to be incorporated into the design of the VIRP. 3.3 Three thousand tanks satisfying these criteria were to be identified, and a sample of six would be subject to detailed feasibility studies. Given the experience in the design of the KRDP and that all tanks would be appraised by the ID prior to rehabilitation, the examination of only six tanks in detail for project preparation does not seem unreasonable.' 3.4 Back-to-office reports, the Issues Paper and Project Brief all emphasize the importance of proper water management and O&M for completed projects. Indeed one of the six components described in the Project Brief was water management. However this was the only component for which not even a skeletal outline was available, "(f) Support of an unspecified water management component, of which the details would be prepared at appraisal." 3.5 The Aide-Memoire for the Appraisal Mission had elevated water management to one of three main components (the others being civil works (rehabilitation) and minor irrigation activities of the DAS). An annex dealt with "Irrigation and Water Management." However, this was devoted primarily to bureaucratic issues concerning the respective roles of ID, DAS and the Extension Service. Having observed that "Farmers tend to cultivate and sow their paddy lands late in the Maha season after the tanks have sufficiently filled ... to ensure a safe Maha crop," and that ..." farmers particularly in the lower reaches of the command area are most reluctant to commence early sowing, due to a limited water supply in the tank would not reach their fields... ," the Annex states that ..." a water plan will be incorporated into the project. Its basic principle would be to distribute water equitably to each farmer in the command area..... ." No details of how this was to be achieved were provided; nor the strategy to be employed in seasons of lower rainfall. Risk was not addressed, even illustratively, and supporting budgets to demonstrate the increased profitability of what was being proposed were not provided. The Aide Memoire notes that "in principle such a program could be extended to all operating tanks and anicuts while in practice it may not be possible..... due to their geographical distribution," thus indicating that whilst rehabilitation would facilitate improved water management it was not a necessary condition. The relative returns to water management versus civil works does not seem to have been examined. 5. Cable 16 July 1979. 6. Unless specifically distinguished, the term "tank" will be used to include both tank and anicut investments. 7. Had this small sample been found to be infeasible, and/or uneconomic, then this would have been reason to re-eamine the project concept of VIRP. In the event, a technical survey of 50 tanks was undertaken. 7 3.6 A second annex on "Selection of Tanks for Agricultural Potential" sounds a salutary note of caution: "3. The evaluation of the irrigation potential of village tanks requires a more complete knowledge of their hydrology. Reasonably complete data are available for the last four seasons for only one tank (Walagambahuwa). This tank is located in the dry zone a short distance from the Maha Illuppallama Research Station. Improved crop practices to make efficient use of available rain water and tank storage have also been developed in the command area for this tank. It is mainly the experience of this one tank that has provided the model which is being extended to other tanks in the dry zone under the proposed VTR project. 4. The rather favorable characteristics of the Walagambahuwa (large catchment, small command area and storage area of some ten acre ft./acre cultivated area) are, however, not representative of a large proportion of the approximately 8,000 working minor tanks. Many of these tanks are relatively small in size, shallow (average depth of about five ft. with ten ft. of water above the sluice) with a much less favorable command to tank size (capacity) ratio."' 3.7 Ironically, even though the Issues Paper was able to say "In some ways, the operating system proposed is a return to the traditional system, ..." (para 2.07) it was a "return" which left the control of water releases from the tank firmly under the control of the Tank Supervisor, a DAS employee.' The design thus failed to address the key issue identified in the Issues Paper and quoted at the head of this section," ... increased Government intervention and changes in the agency responsibility for minor irrigation have tended to undermine the traditional system without replacing it with an effective alternative." B. Project Implementation 3.8 The PCR provides a good summary of project implementation. The VIRP was approved by the Board in June 1981, and it became effective in September. Quick compliance with the conditions for effectiveness, was followed by relatively slow initial implementation of the project. Problems included (a) shortage of recurrent funds, (b) civil disturbance in the North and East, and (c) a decision by the Bank not to implement VIRP in districts served by an Integrated Rural Development Project. This latter decision affected four districts and reflected a desire not to over-stretch the ID and DAS, by asking them to implement two projects in the same district. Restriction of the project to eleven districts after June 1986 (and indeed effectively before that) limited the rate of implementation, since district staffs could not be expanded to compensate for the increased resources available per district. In the event the project was extended by four years. It was closed with 72 percent disbursed. 8. Annex II, Aide Memoire, 19 September 1980, emphasis added. 9. Predictably the SAR endorses increased Government guidance ".... though a number of attempts have been made in the past to assist farmers in one way or another. Uttle impact has been made, however, and farmers continue to utilize an overall farming system which suits them best irrespective of efficient use of land and water resources...... Since it is difficult for one individual farmer to change the farming system, increased Govemnent guidance would be required." (SAR, Annex C, Para 1, emphasis added). 8 3.9 This slow start-up may have been a blessing in disguise, since there were initial problems with design and implementation of some structures. Staff training, and conscientious supervision from experience ID staff, largely remedied this problem as implementation proceeded. These problems were exacerbated by an absence of inflow and outflow data for small schemes. Project selection thus had to be based on iso-yield curves, based on flows in major rivers, which are not necessarily highly correlated with flows in minor watersheds. 3.10 The SAR called for a relatively modest involvement of the affected farmers in the design and implementation of individual tank rehabilitation: an initial meeting at the beginning of the rehabilitation design was to be followed by a ratification meeting on completion of the design. In practice even this modest involvement was frequently honored in the breech. (Only 70 percent of initial, and 50 percent of ratification meetings were held in a sample of 210 schemes studied after project completion.) Lack of farmer "ownership" of the rehabilitate schemes, and hence maintenance of these schemes was a subsequent problem with a significant number of tanks and anicuts (para 3.19). C. Project Administration 3.11 As envisaged in the SAR a Project Steering Committee was established, and met at two to three monthly intervals. An agenda was circulated before meetings, and minutes were recorded. Coordination in Colombo, and between Colombo and the Bank appears to have operated well. Significant resource constraints and difficulties stemming from civil disorder were encountered. These were coped with in an orderly manner, and the Bank was kept fully informed of problems and proposed remedies. 3.12 The good communication in Colombo does not seem to have been fully reflected in the field. The split between the ID's responsibility for rehabilitation, and the DAS's (and farmer's) responsibility for operating the rehabilitated scheme presented real problems. The ID found it difficult to fully involve the DAS and farmers in the finer points of the design of the rehabilitation; and there was a mismatch between the views of the ID and DAS as to when many of the rehabilitated schemes were ready to be transferred. The DAS often required repairs or supplementary works, which necessitated the ID to return to works which were thought to have been completed. Provision of DAS with a fund to carry out small supplementary repairs went some way towards easing this problem. 3.13 In Moneragala, the rehabilitation criteria were ignored in many cases so that new schemes were undertaken."o In these cases hand-over from ID to DAS was further complicated by the need to complete land alienation for the new settlers, which often proved to be a very drawn-out undertaking. 3.14 Clearly design and supervision of individual sub-projects presented a major challenge to ID and DAS; especially since the sub-project sponsors within ID or DAS could well feel more "loyalty" to the farm community being served, than to the intent of agreements negotiated in Washington or 10. About 20 percent of projects have zero or no recorded irrigated area before the ID rehabilitation schemes. In some cases this reflected an entirely new tank being constructed (in violation of the project's intent), sometimes rehabilitation of a tank which had ceased to function at all, and sometimes under-reporting to increase the funds available for the rehabilitation. The costing formula was such that the smaller the before project area, the more could be spent to achieve any given after-project area. 9 Colombo. There were a total of 1,271 rehabilitation projects initiated by ID, and 670 modernization projects initiated by DAS. Thus the number of these sub-projects to be supervised (or canceled) within this one project, substantially exceeded the total number of operations under supervision by the Bank. Not all of the sub-projects were accessible by four-wheel drive vehicle. In the Audit's view the leakage of resources to tank construction not strictly within the project guidelines (or mis- reported to achieve higher expenditures per project) was not a serious problem; certainly not compared to the problem of civil works complying with project size and expenditure guidelines which turned out to be unsuited to the needs of the intended beneficiaries. Even where new tanks were built they were relatively small, and served the needs of poor small farmers. 3.15 Finally, as a result of the separation in responsibilities between ID and DAS, water management plans were generally only prepared by DAS when hand over from the ID had been completed; whereas ideally, the water management plan would be prepared simultaneously with decisions as to the scope of the civil works to be undertaken. D. Project Impact 3.16 According to the PCR, the project actually rehabilitated 1,027" schemes and modernized a further 627 schemes, covering a total command area of 45,555 hectares, compared to an SAR estimate of 31,500 hectares. Families benefitted also exceeded SAR estimates, 46,500 versus 25,000. However, something like half the rehabilitated and modernized schemes are not operating as expected (if at all). The PCR counts only 26,080 hectares as "benefitted areas." It is not clear how much of this slippage is due to (a) civil unrest, (b) inappropriate design of sub-projects, or (c) delayed hand-over of completed schemes to DAS. The decision to switch the project from the disturbed North and East to the tranquil, albeit often wet, areas of the South and West was then exactly the sort of "adaptive management," which is needed, in such circumstances, to allow project benefits to be realized. 3.17 Central monitoring of this project was restricted to quite basic data on sub-project size, location and cost, so that many of the questions which an audit might naturally ask, cannot be answered. An expost survey carried out under a subsequent Bank funded project (Credit 2260-CE), indicated significant improvements in irrigation and cropping intensity, and higher paddy yields in operating schemes. These estimates have been included in the re-estimated ERR which was reported as 15.7 percent in the PCR versus about 20 percent in the SAR (para 7.05). Properly calculated, however the PCR estimate should have been 5.6 percent (paras 3.28 and 3.29). 3.18 Sustainability. The key issue in project sustainability seems to be the existence of a farmer's organization (FO). A healthy FO, leads to willingness to provide the labor needed for channel maintenance and facilitates dispute resolution between farmers. Both of which lead to improved system performance, and a virtuous circle by which farmers are re-inforced in their willingness to maintain the system. The PCR reports 880 schemes (or 55 percent of total sub-projects) had FO's, the majority of which were active. When it is remembered that only 56 percent of the total area treated is counted as the benefitted area, this suggests that a high proportion of the benefitted area 11. The discrepancy in these numbers from those reported to the audit (see para 3.14), may be due to a few projects which were canceled; and a much larger number which had still not been handed over from ID to DAR, even though the originally specified civil works had been completed. As brought out earlier, often the ID was ready to hand over projects, which the DAS refused to accept for a number of reasons. 10 has a responsible FO. Though continued work is needed to complete the engineering of initiated sub-projects, there is an equal need to form FOs where they do not yet exist, and to support their effective operations once organized. If these steps are taken, the prospects for sustainability look good. In the meantime sustainability is uncertain. 3.19 Both Parts I and II of the PCR makes several reference to the importance of farmer or community ownership or management of their scheme, especially for scheme sustainability, and how community involvement in the planning and implementation of the scheme can foster such ownership: "The outlook for sustainability ... is promising for about half or all schemes, where functioning farmer organizations (FOs) have been established..." (Para I 7.01) a decline in farmers sense of responsibility for their schemes in cases where they were not fully participating in the planning, design and implementation of rehabilitation works." (Para 17.02) further public support to achieve FO self-reliance is indispensable." (Para I 7.03) the formation of farmers organizations should come before physical rehabilitation." (Para I 9.04a) "Sufficient emphasis has not been given at the time of project design to the participation of farmers in investigation, planning and implementation of the project......... Strong emphasis should have been given at the time of project design to establish a strong farmer's organization before a scheme is selected for rehabilitation." (Para 11 1.2) "One of the important lessons learnt from the VIRP is : 'The establishment of a stable farmer organization is a prerequisite for the successful rehabilitation and management of minor irrigation schemes." (Para II 3.2.1) 3.20 Thus this lesson has been learnt and internalized: FOs are regarded as an integral part of the implementation of the NIRP follow-on project. The SAR also recognized an important role for farm groups in water management, however in the VIRP design of the physical structures and associated water management plan was seen as essentially a technical problem, with only the irrigation headman (vel vidane) involved (SAR para 5.11). On completion of the works, "the actual management and distribution of water would be decided by farmers," (SAR para 5.14). In part this intended involvement of the farmers and their representatives in the design and management phase of VIRP seems to have been honored in the breech, and where involvement did occur, it was too late in the process for farmers to take real "ownership" of the scheme: hence the very great emphasis the Bank and Borrower place on this issue in the PCR. 3.21 All tanks and anicuts visited in the field had farmer organizations. However, it was clear that though some were working well, others were not. In th: latter cases, individual farmers did not feel they had been consulted by the FO, and the FO had great difficulty in mobilizing farmers to contribute needed labor for channel cleaning and maintenance. This warns that it is not merely the existence of an FO that matters, but rather how well this organization represents and involves farmers in decision making. Fostering effective FOs will require great sensitivity by the ID and DAS, since what officials may regard as "consultation," farmers may recognize as "being dictated to;" similarly though officials may have had substantive discussions with FO representatives, this does not ensure 11 that these representatives provided adequate feed-back and involvement of their members. Thus whilst there is a consensus that healthy FOs are essential to the economic operation of small irrigation schemes, there remains significant uncertainty as to how to bring them about. (See para 5.02 for the Audit's reservations as to the willingness of FOs to fund O&M). A particularly egregious example of the powerlessness of a FO, and its failure to get Government's attention to its problems, is described in Box 1. As a partial antidote to the unsatisfactory experience described in Box 1, a particularly successful rehabilitation is described in Box 2. 3.22 Even though there was provision in the SAR for consultation with the vel vidane, it appears that in many cases this was rapidly obscured by other apparently more immediate problems of implementation. Not only were farmer's not substantively involved in scheme design, but the ID who had charge of construction, frequently failed to discuss their designs even with DAS who were designated to take over schemes for purposes of O&M. Note that this handing over of schemes from one government department to another, differs markedly from a model of handing it over to the beneficiaries. Indeed, despite the quoted priority for farmer involvement, it appears to the Audit that the DAS's preference would be to have sufficient staff to provide direct management of all village tanks. In this view, it is staff shortages which primarily explain the willingness of DAS to foster and work with farmer's organizations. 12 Box 1. An Exceptional Problem with Riparian Rights A particularly extreme example of the difficulty DAS can have in working with FOs, was shown to the Audit. This concerned a 70 acre scheme where "there had been an anicut (diversion weir) from the time of the Kings". The project had slightly upgraded and repaired this ancient anicut. Subsequently, a donor supported project had installed a hydro-electric scheme so that only half the acreage could be irrigated in any one year. The hydro station design simply ignored the needs, and rights, of the farmers. It has a head of about 200 feet, if they had used only 170 feet, then the water would be delivered above the anicut (albeit in a four-hour surge per day, instead of the historic constant flow). For the first time in centuries only half the bunded area was under irrigation; and this was subsequent to investments made under the VIRP. Several issues arise from this case. Firstly, this is, understandably something of a cause celebre, which has resulted in a Parliamentary Question to the Minister of Irrigation, but the DAS did not know of this problem until the farmers mentioned it to the Audit. This raises a question as to how effective is the communication between the FOs and DAS. Secondly, under Sri Lankan law, as explained to the Audit, the Farmers have riparian rights, and no one is allowed to take water that they have used historically. Yet this has been done, and now the Electricity Department is negotiating with them as to what portion of their historical rights will actually be delivered to them. If this is correct then a court case would simply result in the Electricity Department having to deliver the farmers' water entitlement, regardless of whether the donated power station, which even now only has water for four hours a day, turned out to be a white elephant. There would seem to be an onus on DAS to support the farmers, in Court if necessary. Ideally, the DAS project monitoring activities should have brought this problem to the Bank's attention, while the project was still under supervision. Formal Bank support for the project having ceased, it is the Government of Sri Lanka which bears primary accountability for follow-up actions. Since no action was being contemplated by the Borrower at the time of the audit, and since the audit happens to have identified a problem, it seems to the Audit that the Bank has at least the responsibility to monitor the situation, until resolved. 13 Box 2. Rehabilitation at its Best Padalangala Tank: Rehabilitation started and completed in 1987, with hand-over to Department of Agrarian Services in 1989. Command area increased from 70 acres before rehabilitation to 85 acres after project completion. Farmer Organization (FO) formed in 1986 and has been very active. In particular farmers were consulted in planning and design and were involved in the construction process. Design and quality of work are good. Kanna (work planning) meetings are held regularly and the system is managed by the FO; which also extended the supply canal by 2,700 ft. to irrigate and additional 25 acres (making a total command area of 110 acres). Cropping intensity is 200 percent and rice yields have increased from 45 bushels/acre before project to 80 bushels/acre after project (and 110 bushels/acre in 1991). No additional works are required. (Scheme 174 as reported in "Village Irrigation Rehabilitation Project Impact Assessment Study," Vol. III, TEAMS (Pvt) Ltd. P.O. Box 262, Colombo, February, 1992). 3.23 For the moment, DAS resources look to be sufficient to assure sustainability for those tanks with effective FOs. However true sustainability of decentralized village tanks would seem to be best assured by direct management by the farmers concerned. Only if they know that the Government (or indeed the Bank) will not provide O&M resources, is it economic for farmers to do the work. The vagaries of Government finances are just too great, to be relied on for long term sustainability. The need is thus for DAS to hand O&M over to the FOs, with DAS simply providing technical advice, as needed. The Economic Rate of Return 3.24 The basic problem for the economic viability of this project, as for Mahawheli m, was the unforseen' and dramatic fall in the price of rice. Using a constant technology, in 1980 production costs accounted for 40 percent of the revenue from paddy production; by 1993 these same (physical) production costs accounted for 80 percent of the revenue from paddy production.' By 1980 most of the yield benefits of the Green Revolution had been reflected in farmers' fields. There has been little improvement in yields over the past decade, so that the above price squeeze has translated into 12. Unforseen by the Bank's Commodities Department; and if unforseen by commodity specialists, "unforeseeable" as far as FAO/CP, Bank and Sri Lankan production erperts were concerned. 13. Cost Series maintained by the Bank's Colombo Office. 14 a revenue, and indeed economic, squeeze. This cost squeeze is expressed in Table 3.1, in terms of the quantity of paddy needed to purchase production inputs. Table 3.1. Input Costs in Kilograms of Paddy, 1980 and 1993 1980 1993 Hired Labor 425 985 Fertilizer 219 523 Agro-Chemicals 89 256 Farm Power 492 572 3.25 We do not need to go beyond Table 3.1 to diagnose that rice production would look a lot less attractive er post in 1993, than it looked ex ante in 1980. 3.26 Two surveys were commissioned to study the project's impact. The first carried out by the University of Paradeniya under the project, and the second by a consulting company under the successor NIRP project. Neither survey was able to identify major increases in production between the pre- and post-project situations. The second survey had the larger change, a seven percent increase in cultivated area, and 16 percent increase in total output. Neither survey could, of course, provide data on the counter-factual of the without-project situation. 3.27 ERR calculations are frequently as revealing in what was assumed about the counter-factual without-project situation, as in the number which results from these assumptions. In the present case the PCR assumed a 20 percent decrease, over a ten year phase-out perio4 in irrigated area" in the without project situation. Since there are about 13,000 village tanks and there has not been a significant decline in production from the 11,000 untreated tanks, nor was there a pre-project production trend which would have justified this assumption,' the assumed decline strikes the Audit as highly suspect. 3.28 In the present case the PCR reported an ERR of 15.7 percent (PCR, Annex I, Table 7). However, when benefits and costs were calculated for the project as a whole (PCR, Annex I, Table 7) and for representative farms (from PCR Tables 4 and 9), markedly different results are obtained, as shown in Table 3.2. 3.29 We would of course expect the benefit/cost ratio to be the same regardless of the scale used to calculate it. When this discrepancy was drawn to the attention of the authors of the PCR, they reworked their figures and discovered an error with the result that using exactly the same assumptions as the PCRA the ERR was shown to be 5.6 percent. The revised Annex I for Part M of the PCR has been incorporated as Annex IV of this audit. A knowledgeable staff member has commented: "Let 14. Largely compensated for by increased rain-fed area, albeit at a lower yield. 15. Note that the SAR (Annex 5, Table 1) assumed the same output for the pre-project and without-project situations. 15 me quickly add that I don't believe for one moment that the ERR for VIRP was as low as OED' tries to tell its readers. OED brands as 'highly suspect' the assumption that production would decrease without the project since: 'there has not been a significant decline in production from the 11,000 untreated tanks.' No kidding!! The project, of course, took on the schemes which were most at risk. During my visits I have seen so many schemes where at relatively modest costs earlier damage was remedied or future damage prevented. It is a misconception to believe that remedial works need to be justified by increases in future production. There won't be much production increase as a result of rehabilitation (except that project - promoted better water management practices may lead to some improvements). Stabilization of production at current levels is the main benefit of projects such as VIRP. OED appears to have problems with the notions outlined above. Let me direct their attention to para 3.1 (page 9) of the PAR where FAO/CP's project preparation files are quoted: 'only 50 percent (of the tanks) are working at present providing water at low efficiency to about four hundred thousand acres.' Now, how did that come about? And would VIRP (and its successor NIRP) not have helped arrest, or perhaps reverse, that trend? Take it from me: VIRP definitely did so." Table 3.2. PCR Benefit Cost Ratios for Representative Farms and the Project (FY 1991) Per Farm (0.75 ha) Project (26,800 ha) (Rs) (Rs M) Without Project Benefits 12,326 383.0 Costs 6,068 306.3 B/C Ratio 2,031 1.250 With Project Benefits 18,081 629.9 Costs 12,258 391.9 B/C Ratio 1,475 1.607 3.30 Three comments are in order: i) Much more justification for the change from the SAR assumption with respect to the without project irrigated area should have been provided in the PCR (Para 6.04); especially, since this change tended to boost the ERR. In the Audit's view, the SAR has the more plausible assumption. ii) The 15.7 percent ERR estimate should, in the Audit's view, have been suspect simply on the basis of field visits, both to the authors of the PCR, and the project Task Manager. There was no way that the sub-projects inspected by the Audit were yielding anything like 15.7 percent; and it is implausible that the Audit was shown a sample chosen specifically for their poor performance. 16. i.e. As calculated by FAO/CP, on the basis of the PCR assumptions. 16 iii) Reporting this relatively high e post ERR estimate only a year after approval of the follow-on NIRP, can only have had the effect of providing a false sense of security as to the economics of the NIRP (the SAR reported an ERR of 31 percent), especially given its greater emphasis on farmer participation." is 3.31 Despite the lack of overlap between SAR intentions and actual project achievements, the Borrower's verdict is that this is "one of the most successful projects in comparison to minor irrigation development projects launched in the past." (PCR, Part II, para 6). The Audit believes that this judgement may well have been colored by the erroneous estimate of the ERR included in Parts I and IlI of the PCR. 4. THIRD MAHAWELI GANGA DEVELOPMENT A. Project Formulation 4.1 System C is an area of about 78,400 ha on the right bank of the Mahaweli Ganga. Prior to development, the northern half had virtually no resident population, and even in the southern half only 4,000 hectares were irrigated. Population pressure from the wet zone was leading to increased immigration into the area with associated shifting cultivation (chena). However, for the most part the jungle provided a habitat for a wide range of wildlife. 4.2 In 1978 consultants for a bilateral donor carried out an initial survey of System C, as part of a preliminary feasibility report on the Victoria Dam and Hydropower Project. This survey concluded that System C development would be viable in terms of agricultural production and would also offer substantial social and employment benefits. In 1979 the consultants were again commissioned to carry out a full feasibility study of System C. 4.3 The consultant's report recommended the development of about 10,400 ha of double-cropped rice, a 4,000 ha sugar nucleus estate and 6,900 ha of out-grower sugar production. At full development 17,300 farm families were to have been settled. The Project Brief based on this proposal raised issues with respect to (a) managerial and implementation capacity of MASL, (b) the high cost of the proposed sugar estate, and (c) the marginal ERR estimated to be 10.7 percent On the latter, it was noted that the Bank was "aware of the economic marginality of the Mahaweli 17. The Region has commented: "Mis-calculation of the ERR for the Village Irrigation Rehabilitation Project (VIRP) in the PCR. In 1992, the draft PCR prepared by FAO/CP was reviewed by two economists in SA3AG and cleared by the Project Adviser, SA3DR. The recalculated ERR (of about 16 percent, versus 20 percent in the SAR) and the underlying assumptions on the "without project" situation were found reasonable. Based on field experience and previous discussions with the long-term task manager, it was felt that in the "without project" case production was bound to decline without adequate maintenance. We still believe this result not to be unreasonable, also in view of the ERR of about four percent reestimated in the PCR for the capital-intensive Third Mahaweli Ganga Development Project (Mahaweli III), which the draft PAR accepts. If Mahawdi III gives a rate of return of about four percent, a rehabilitation project like VIRP based on Sri Lanka and Bank experience should produce a higher rate than the six percent shown in the draft PAR. Regarding the follow-on National Irrigation Rehabilitation Project (Credit 2260-CE), its performance to date is not satisfactory. We do not think that the development impact as originally envisaged will be achieved because of implementation delays. This will be addressed during a planned mid-term review, when the project will be restructured." 18. The NIRP ERR estimate was based on technical assumptions made jointly by the appraisal mission members on site at six tanks in need of rehabilitation. - A thoroughly professional approach. The ERR gap suggests some problems in implementation. 17 Program but has supported it actively in view of its significant employment, settlement and other social benefits." 4.4 The appraisal mission recommended a project which would construct 26 km to complete the Right Bank Trans-Basin Canal, clearance and irrigation of 21,000 ha currently in jungle, provision of associated infrastructure and social services, and technical assistance with implementation. At completion 20,400 farm families would be settled and another would be 8,000 provided with employment in associated support services. The proposed sugar estate was rejected due to the high cost of sub-surface drainage which would have been required, and paddy production, with some additional minor crops (tomatoes, potatoes, chilies and onions), was substituted. Subsequently, 2,000 ha of cashews, and 3,000 ha of fuelwood plantation were added, without the genesis of this addition being very clear from the files. Environmental concerns were assuaged by the promise of a Government plan to minimize negative environmental impacts, (see however paras 4.18 to 4.22). B. Project Implementation 4.5 This project was part of the accelerated Mahaweli development program, and it showed. In particular a major part of the cost over-run can be attributed to unexpectedly large amounts of rock as encountered in the construction of the Trans-Basin canal and a lack of fill material." In the circumstances, it reflects well on project management that supervision missions diagnosed predominantly "minor problems." Overall, project implementation took four years longer than expected and only 81 percent of the credit (US$64.5 million equivalent) was disbursed. This delay may however have reflected more on over-optimism in the SAR, than under-performance in the field. The project had to be implemented by an already overstretched organization, facing new technical and organizational issues. Technical issues involved the degree of land leveling to be incorporated with land clearing and the merits of pre-cast versus in situ structures. Organizational issues involved the optimum size of contract, and the classes of work best suited to large Sri Lankan contractors, small local contractors, and to construction by the settlers themselves. In addition, though project design had anticipated priority for farmers re-settled from areas flooded by other parts of the Mahaweli scheme, it had not foreseen that these involuntary re-settlers (and also refugees from the civil disorder) would arrive in System C without regard to the supply of irrigated land ready for distribution. The result was that some settlers had to be accommodated for two years before they could be assigned land. The World Food Program (WFP) had a major role in feeding these intended settlers in the interim. Similarly, delay in provision of water to cleared land lead to unforseen problems with (a) spontaneous settlement, and (b) jungle regrowth. 4.6 Parts I and II of the PCR provide a good description of the implementation problems encountered, and they will not be repeated here. Suffice it to note (a) good responsiveness of the project management to advice provided by Bank supervision missions, (b) the realism and relevance of this Bank advice, (c) a heavy use of training, both of project staff and farmers, to upgrade their technical competence, and (d) that the result was that both the hardware and human capital needed to use it effectively were created. The project took longer to implement than expected, but it did more than described in the SAR: it created human as well as physical capital. 19. I.e. the O.M. 2.28 on State of Project Preparation had not been fully observed. 18 4.7 The one area where implementation left much to be desired was in the transfer of "ownership" to the farmers being served. Ownership, not in a legal sense, but in terms of planning water use and cooperating to maintain the distribution system. This is a vexed and "classic" problem of irrigation projects; one which is easier identified than solved. 4.8 Project conditionality called for an Environmental Action Plan to be submitted for IDA's review and concurrence by December 31, 1981; to be followed by implementation not later than March 31, 1982. In the event a very unsatisfactory document was submitted in March 1983. The needs of Zone C being presented as a two page, generalized Annex to a sketchy project proposal asking for bilateral support for a Mahaweli-Wide Environmental Project. The total proposed budget was $6.9 million, of which $5.0 was proposed to be provided by bilateral funding. This document made no quantitative estimates of the wildlife population and habitat destruction of Zone C. Nor did it relate the proposed actions to projected changes in wildlife populations. Nothing resembling "with and without" scenarios was presented. It was sent within the Bank to the Environmental Advisor with the bromide: "Given the importance of environmental protection and the role this proposed program can have in charting specific action towards this goal, we would appreciate your detailed technical review..." (memorandum March 8, 1983). There was nothing on which a technical review could be based. In the Audit's view the Annex should have been rejected out of hand, as not responsive to the Credit Agreement. 4.9 The files contain a draft telex submitted by the Environmental Advisor asking for very substantial elaboration. There is no evidence that this was sent; and the next supervision (December 1983) makes no reference to the environmental action plan, beyond noting, against the covenant to commence implementing a plan that "6.0 Action program initiated with bilateral assistance." No mention that this was for the Mahaweli as a whole. The Audit can find no evidence that the Bank ever fulfilled its part of the conditionality, to give its concurrence to the Environmental Action Plan. Indeed, unless an entirely different plan was submitted it is hard to believe that the reservations of the Environmental Advisor could have been met. 4.10 Cashews were introduced during appraisal in recognition of a large area of poor, sandy and unirrigable land. It proposed that 500 hectares be planted each year for four years. The appraisal mission identified cashews as being able to thrive under similar adverse conditions. Initially a nucleus estate of 400 hectares with outgrowers was proposed. The cashew area was first handed over to the Sri Lanka Cashew Corporation (SLCC) to develop, but after 44 million rupees had been spent on its development, only 67,000 trees were found to have survived (at 90 trees a hectare, enough for 745 hectares).' Mahaweli took back responsibility for cashew establishment in 1991, and after a period of poor results due to fire, and wildlife (elephant) damage to the trees decided to establish 400 settlers, in the hope that they would control fires, and chase the elephants away. This policy has met with mixed results. Two years of food support under the WFP is due to come to a halt, and cashews are far from having been satisfactorily established; in part due to recurring elephant damage. Fires have been reduced. 4.11 An electric fence around System B is due to be completed within a few months, which is expected to eliminate the problem of elephant damage. Problems remain unfortunately with the soil, which though sandy overlies rock or impervious clay at about three feet. The result is that during 20. A subsequent survey suggested that only 30,000 trees, enough for 333 hectares, survived. 19 the rainy season (Maha), the water table rises and the soil becomes water logged for from three to four months. Cashews cannot stand wet feet, accordingly there are high tree losses in the waterlogged areas. The conclusion by the Audit is that the appraisal misjudged the suitability of the rainfed area for cashews (see Annex II). C. Project Administration 4.12 The Ministry of Mahaweli Development (MMD) has overall responsibility for the development and operation of the irrigation system within the Mahaweli scheme; and for economic development. Acting under the authority of MMD, the Mahaweli Authority of Sri Lanka (MASL) acts as implementing agency. Key functions within the MASL are split between: i) The Mahaweli Development Board/Mahaweli Engineering and Construction Agency (MDB/MECA)" is responsible for physical works, including planning, design and construction of irrigation and social infrastructure. ii) The Mahaweli Economic Agency (MEA), is responsible for post-construction implementation of projects. This includes (a) settlement and O&M down to the farm turnouts, (b) on farm development, (c) agricultural development and water management, (d) community development, (e) assistance in input/product marketing, (f) agricultural and other advisory services, (g) land allocations, and (h) promotion of secondary industries. 4.13 Thus the structure of the MASL provides for the same institutional dissonance between MECA (construction) and MEA (operations) which characterized the VIRP as between the ID (construction) and DAS (operations). Though not nearly as marked as in the VIRP, there were minor problems as to how much leveling was to be included in forest clearing. Forest clearing being the responsibility of MECA, whilst subsequent field leveling was the responsibility of MEA or the settlers. Operating within the same Agency, incipient disputes could be resolved without having to involve Cabinet Ministers. 4.14 Uncertainty as to career prospects for MECA staff on project completion, together with withdrawal of incentives towards the end of the project, led to some loss of staff morale and decline in the desire of staff to complete the project. D. Project Impact 4.15 The project substantially met its irrigation and settlement objectives, but took twice as long to do so as was projected in the SAR. This, together with a halving of the world price for rice, resulted in a very substantial drop in the estimated discounted net benefits (see para 4.23). Both the cashew and fuelwood plantations performed significantly less well than projected, but these were minor components. The project cost twice as much in Rupees (30 percent more in real terms) as projected, but this was very largely due to local inflation. Inflation which was one of the costs of the Government's decision to accelerate the development of the Mahaweli scheme. 21. MECA is the successor organization to MDB. Unless otherwise specified MECA will be used interchangeably for both organizations. 20 4.16 Irrigated farms were provided to 16,136 settlers versus 18,500 projected in the SAR (87 percent) at a total cost of $280 million, or $17,350 per irrigated farm. This gross cost includes forest clearing, irrigation works, roads, social infrastructure, supporting towns, and over 2,000 families settled without irrigation.' Additional rice production is estimated to be about 145,000 tons versus an SAR estimate of 153,000 tons (95 percent). 4.17 Sustainability. Whilst the PCR concludes that overall construction standards were acceptable, it is clear that more will be required in the way of rehabilitation, repairs and maintenance, than was envisaged in the SAR. If this is not to burden MEA with insupportable costs it will be essential that farmers be involved in water management, maintenance of tertiary canals, and rectification of irrigation defects. MASL has a very active program designed to promote and support farmer organizations (FOs), which are expected to take responsibility for water management and O&M at the distributory and field canal levels. Without this active farmer involvement in water planning and management at the field level, severe degradation of system performance would be likely. Government remains fully committed to the scheme. 4.18 Environmental Impact. Almost from before project inception, the issue of environmental impact was of concern. Never as a central issue for Operations, it was important for Central Projects Staff (CPS), the Environment Office, and individuals both within and outside the Bank. At the inception of the AMDP, the question of environmental impact was raised by the Bank and others. USAID financed a $1.0 million environmental impact study referred to as the TAMS report. The preparation of the Mahaweli III project proceeded in tandem with the TAMS study. 4.19 An early preparation mission observed that "Another issue that may not be receiving adequate attention is the value of the forest both present and future without the project........ It was not clear that anyone had considered and evaluated receiving financial return through systematic harvesting of the forest without the project or the benefits of leaving some areas for wildlife habitat." (Memo 12/26/79, para 31). This is the only reference the audit found in the files to the need to properly define and evaluate the "without project" scenario. 4.20 The final draft of the TAMS report recommended incorporation of "Zone 4 of System C into the Wasgomuwa National Park. Zone 4 is categorized by TAMS as critical habitat for several endangered and endemic fauna, one of which is confined to Zone 4." The Bank commented to the Government that "the deletion of this area would require a re-examination of the economic viability of System C development," and expressed the assumption that "Government is working with TAMS to finalize environmental solution, to be reflected in TAMS final report" (telex 8/6/80). An internal Bank report noted that Government had not yet seen the report officially, but "The proposal affecting Zone C is completely unacceptable to Government and a committee has been working .... The most likely alternative is to create a protected catchment and animal migration area between the Maduru Oya and Gal Oya Parks. .... Whatever the solution, it will in no way, as far as Government is concerned, affect the viability of System C" (telex 8/12/80). On the 18 of August the Bank again sought clarification: "Regarding TAMS report on environmental protection in Mahaweli Program. For purposes of the appraisal report drafting, we would be most grateful if you could urgently telex us Government's current thinking or decision..." (telex 8/18/80). This elicited the reassurance that 22. Comparable costs for the Indonesian Transmigration projects, which also cleared farms from jungle were $12,000 per ha for irrigated and $6,000 per ha for rainfed. These Indonesian projects have, on occasion been criticized for being excessively expensive. 21 Government "does not agree with several recommendations regarding wildlife in the TAMS draft report. After discussion of alternative proposals made by us the team leader has agreed to modify the draft final report. The modified draft does not recommend deletion of Zone 4 of System C nor does it recommend a permanent corridor across System C," (telex 8/19/80). This allowed the SAR to report that: "The general view of the TAMS (final) report, however, is that with prudent action, net environmental degradation under the Mahaweli Program would be considerably less than is expected to occur were the project not implemented" (SAR, para 3.49). 4.21 The Yellow Cover Staff Appraisal Report (SAR) was panned by both CPS and the Bank's Environmental Advisor on environmental grounds. "The project would destroy the habitat of wildlife including eight endangered species. Although the (appraisal) mission included a representative of OEA (Office of Environmental Affairs), the report contains no recommendations to avoid adverse environmental project effects and merely calls for the submission of a plan by end 1981....... the only environmental proposal is to preserve a 1,000 ha block of Sri Lankan Dry Mixed Evergreen Forest. This seems a remarkably impractical suggestion as a sop to the environmental community. Will such a small area .... have any real value? Can it be preserve anyway, or will it be gradually eroded by surrounding cultivators ...?" (memorandum September 18, 1981)." 4.22 No substantive response was made to these concerns. Operations in advising the Vice- President wrote: "Environmental Issue. Both the Environmental Advisor and AGR have expressed concerns .... The Region shares these concerns, but believes that there is no realistic alternative to proceeding with the project,' while at the same time trying to minimize the adverse environmental effects...... paras 3.50 and 7.01 (h) (should) be strengthened by requiring that the Plan .... be acceptable to the Bank" (memorandum October 1, 1981). Suffice it to say that the record does not bear out the Region's claimed concern. Not only had the Bank implied the need to change the draft TAMS report, but eighteen months later no significant progress had been made on the preparation of the promised environmental plan. Reliance on a yet to be formulated plan, reflected either indifference to environmental issues or an unnecessarily incomplete project preparation and appraisal.' 4.23 Economic Rate of Return. Parts II and III of the PCR have widely different estimates of the ERR. In Part II the Borrower estimates the ERR to be 14.28 percent, in Part III the Bank estimated it to be four percent.6 Due to some mix-up in the submission of Part II, the ERR calculations used were "intermediate results," whereby benefits in terms of paddy produced were used at constant 1991 23. This same memo ended with a reflection on declining Bank professional standards. ".... it is unfortunate that mission members cannot be made available even for Yellow Cover Review. It used to be that the a mission had to see its project through Green Cover. While I realize the increasing workload on key Bank technical staff, there seems to be little point in a CPS technical review if there is no one to answer the question." 24. In a more reflective mood, the PCR suggests that: "One step that could have helped might have been to confine the project to development of Zones 3-5 and leave Zone 6 for a later phase. This would have kept part of the elephant habitat and provided a connection between Madura Oya Park and the areas to the north." (para 1.6.12). 25. The Region has commented: "Environmental Concerns under Mahaweli III not Addressed - Since the early 1980s, the Bank including the Asia Region has become more sensitive toward environmental concerns. In the course of preparing the Mahaweli Restructuring Project (1992-1993), a joint USAID/Bank environmental assessment was undertaken based on which a component in the proposed Mahaweli Restructuring Project was designed to improve environmental management in the Mahaweli areas." 26. This low ERR did not come as a complete surprise to the Bank, see para 4.3. 22 prices, whilst costs to produce this rice were held at the (lower) current level at which they were actually incurred. When benefits and costs were held at current levels (a debatable procedure, see para 4.26 to 4.28) the Borrower's estimate of the ERR dropped to 11 percent (See Annex V). Even 4 to 11 is a very wide gap, when it is remembered that the judgement as to whether a project has been satisfactory or unsatisfactory (and hence the Bank's reputation as an effective developmental agency) is very heavily influenced by the estimated ERR. As discussed in Annex VI, the Audit agrees basically with the methodology used by the Bank. 5. MAIN FINDINGS AND ISSUES Ratings 5.1 The low ERRS obtained in these projects, can be attributed more to the decline in real rice prices (see Table 3.1) which has occurred in the face of stagnant production technology, than to identifiable weaknesses of project design or implementation. Certainly, implementation could have been improved, as has been recognized in the follow-on NIRP, particularly with respect to the involvement of all stake-holders in the design of VIRP sub-projects; and some aspects of contracting and construction of the distribution and field canals under Mahaweli III. But these deficiencies were not, in the Audit's view, a major cause of the disappointing ERRs. Audit and PCR project ratings are compared in Table 5.1. Though the Audit classifies these projects as "unsatisfactory," parentheses could usefully be added to explain "(for reasons beyond the control of the Bank or Borrower)."" The Audits "unsatisfactory" rating for the VIRP is due to the revised ERR of 5.6 percent. In the case of Institutional Development, "modest" is a synonym for "partial" (the latter adjective no longer being available for project rating). Similarly "uncertain" and "marginal" are synonyms for sustainability. The Audit's upgrading of the Sustainability of Mahaweli III from "uncertain" to "likely" reflects a perception that Government has the political will to find the resources needed to maintain the investment in System C. Table 5.1. Comparison of PAR and PCR Ratings Village Irrigation Mahaweli III PAR PCR PAR PCR Overall Unsatisfactory Satisfactory Unsatisfactory Unsatisfactory Institutional Modest Partial Modest Partial Development Sustainability Uncertain Marginal Likely Uncertain 27. Given the decline in world rice prices which occurred over the period of project implementation, it is doubtful whether any project to provide new irrigated land to the poor and landless farmers in Sri Lanka could have been designed. Thus the projects may well have been "cost effective" in that more economic alternatives for providing irrigated land were not available. With hindsight, economic ways of helping poor and landless farmers would have required a radically different assistance strategy. 23 A. The Role of Farmer's Organizations 5.2 The key Bank policy applicable to this project was OPN 2.10 pertaining to cost recovery. Project compliance was mixed. VIRP had no conditionality pertaining to cost recovery, whilst Mahaweli III called for charges to recover at least the average O&M costs by 1991. As different from the PCR, the Audit found that any farmer financial contribution to O&M costs was minimal in both projects. However, major efforts and progress seemed to be being made in persuading farmers to provide the labor required for distributory and field canal maintenance. There is now a consensus on the importance of well run Farmer's Organizations to maintain distributory and field canals, and manage water distribution, "well run FO's" are not items which can be bought off the shelf. The NIRP requires that a FO be in existence and consulted before rehabilitation work can commence; indeed the FO has to sign-off on the proposed works. This is plainly helpful,' however, it is no assurance that an FO will run smoothly and efficiently after the works are completed. Project design, and Government administration of its services, can influence how effective FOs are likely to be; but it is village dynamics and the relations amongst the members which ultimately determines how effective a FO will be. The more "governance," cost, efficiency and other considerations lead to an increased role for FOs, the less is implementation of the project under government control --- under government influence "yes," under government control "no". Increased recognition of the vital role for FOs, implies increased recognition of the riskiness of such projects. This is not to argue against an important role for FOs;' but it is to recognize that the government (and ipso facto the Bank), loses an element of control of project performance with this transfer of responsibility for project success. The Audit was impressed with the apparent willingness of FOs to provide the labor for cleaning and minor O&M of distributory and field canals; but their almost total unwillingness to provide even small amounts of cash for needed hardware. Given the savings, and improved system performance, which would flow from this labor input from farmers, the Audit believes that it would be counter productive to also try to get farmers to pay directly the cash costs of system maintenance (this despite OPN 2.10). Despite the lip-service which officials pay to the Bank's preoccupation with the "user pays" principle, the Audit got the impression that there was quite widespread political will for taxpayers to pay the cash cost of system maintenance; in part for income redistribution purposes. B. Low Internal Economic Rates of Return 5.3 To reiterate, in the Audit's view both projects turned out to be "unsatisfactory," not because of poor design as approved by the Board, and not because of poor implementation; but because the price regime moved against the irrigation sector. This price change was unforseen by the Bank's Commodity Division, and hence may be treated as "unforeseeable" by technically oriented project officers or irrigation officials. In part the price decline came from the increased supply of small grains which has resulted from the Green Revolution, the Common Agricultural Policy and farming subsidies in other OECD countries, reinforced in Sri Lanka's case by a more open economy which has lead to an increase in real wages. 28. As has been the MASL's appointment of Institutional Organizers to foster creation of, and provide support, to FOs. 29. Clearly, there are gains in equity, and likely efficiency, as the views of beneficiaries are taken seriously into account. Furthermore, the option of "full central control" is really not financially feasible. Farmers are all too willing to let the government bear the full cost of system maintenance, if it so wishes. 24 5.4 In the circumstances, the mis-calculation of the ERR for the VIRP in the PCR was most unfortunate, since with an ex post ERR of 15.7 percent for the VIRP, and an NIRP with improved design reflecting lessons learned from the VIRP, an ERR of 31 percent for the NIRP appeared plausible. As brought out in para 3.31, recalculation yielded an ERR of only 5.6 percent. 5.5 It would be wrong to treat any ERR calculation as sacrosanct. These calculations, resting as they do on an assumed "without project" scenario and often using borderline information on even the "with project" facts, are inevitably indicative rather than precise. However, in the present case it is the Audit's view that the "without project" scenario is already excessively generous to the project (para 3.27); and two surveys agree on the quite limited (less than 20 percent) production increases associated with project investments. The ERR calculated for the NIRP was 31 percent, was based on technical assumptions made jointly by the appraisal mission on site for six tanks in need of rehabilitation and is predominantly explained by the rapid decline in both paddy area and yield projected in the "without project" situation. If the 31 percent estimate for a new project, is right, the reported 5.6 percent for a very similar, recently completed, project suggests problems in project implementation. 5.6 In the light of the changed price ratios (see para 3.24), it may now be difficult to design any new investment or national project which would yield a ten percent ERR on the basis of increased irrigation water for paddy production.' The ongoing NIRP should be reviewed in the light of the lower ERR now attributed to the VIRP. If continued, then in addition to the maximum expenditure per hectare of command area each sub-project in the NIRP have a properly calculated, and farmer- validate4, Net Resent Value of Benefits, prior to construction. The performance of rejected sub- projects should be carefully monitored, to check the assumptions used for the "without sub-project" scenarios. 5.7 How can small and impoverished farmers be assisted other than in the provision of additional irrigation water? Three directions suggest themselves (a) agricultural research (locally produced or imported) to increase the genetic merit of rice plants; (b) diversification into higher valued crops; and (c) provision of better educated and more entrepreneurial young farmers. 5.8 Even though the Mahaweli Agricultural and Rural Development (MARD) Project on crop diversification has not yet had the demonstrable successes initially hoped for, yet it seems to the Audit that this is a direction of public sector research which should continue to be supported." However, education, at least to the diploma level of one of two young farmers per village, would seem to hold the highest promise for increasing the dynamism of smallholder agriculture. 30. Obviously, there may be isolated cases where huge sunk costs, or recent but easily repaired tank failure would yield an acceptable return; however, these are unlikely to justify a significant investment program and would need to be carefully budgeted on a case by case basis. 31. Operations is currently in the process of completing an A&icultural Sector/Crop Diversification Policy Study. This should clarify the reasons for the lack of demonstrable success by MARD, and the relative roles of technology, policies and institutions in explaining the limited diversification which has taken place to date. 25 C. Special Situations 5.9 Two special situations encountered by the Audit, deserve comment. As described in para 4.11, the appraisal seems to have misjudged the suitability of unirrigated land for cashews, due to the high degree of water-logging experienced in the rainy Maha season. The Bank should re-eramine the wisdom of the appraisal recommendation of sole cropping cashews on unirrigated areas, giving particular attention to the possibility of mixed cropping to take advantage of the different soil moisture regimes on the ridges, intermediate slopes and gully bottoms. 5.10 The other situation, described in Box 1, was where subsequent installation of a small hydro- electric plant had deprived a rehabilitated anicut of waters that they had used "since the time of the kings." With this particular problem identified, the Bank should examine what it can do to help these small farmers in regaining access to their riparian rights, or obtain full compensation. 5.11 The above two case illustrate the need for continued review of project operations (not necessarily by the Bank) even after project closing. In practice this would go beyond the monitoring of operations and development impact called for in the context of the newly introduced Implementation Completion Report, to taking remedial actions. No recommendation is made on the basis of these two projects, but this is an issue which should be watched by Operations and OED. 5.12 The data-base used to support contact with Farmer's Organizations was quite inadequate, both for the MASL and ID, but most particularly for the DAS. It was not possible to tell from headquarters, what production had been in recent seasons, what problems farmers had identified, or the status of ongoing rehabilitation work. It should, for instance, have been possible to identify that there was a major problem for the anicut deprived of much of its water, by the resulting halving of the area under paddy. The Bank should consider supporting the development of a data base which would allow government to identify problems, and successes, of individual small scale irrigation schemes. 5.13 This would at least allow the Government to efficiently monitor and support project sustainability, after the completion of construction; and would inform the Bank's continuing subsectoral dialogue with the government. D. Erroneous and Different Internal Economic Rates of Return 5.14 Finally, the erroneous ERR reported in the PCR for the VIRP, and different ERRs reported for Mahaweli III, should not be forgotten. In this case the PCR was prepared by FAO/CP and not by Bank staff, however it should be emphasized that the final product is the responsibility of the Operating Region (and Divisional Task Manager), who should ensure that the documentation in the PCR is complete and calculations can readily be followed. 5.15 While the Bank and Borrower may well differ, as for Mahaweli III, in their views as to how the ERR should be calculated it should be possible to define these differences. Where the Bank and Borrower differ in their estimates of the ERR, an agreed Reconciliation Statement indicating the differences in methodology or data which lead to the different outcomes, should be included in the Implementation Completion Report.  27 ANNEX I Bank Lending for Inigation Mahawell Credit No Title Approved US$ Million 174-CE Mahaweli I January 1970 15.5 701-CE Mahaweli II April 1977 19.0 979-CE Mahaweli T/A January 1980 3.0 1166-CE Mahaweli III June 1981 90.0 1494-CE Mahaweli IV June 1984 30.0 Total 157.5 Other Credit NO Title Approved US$ Million 121-CE Lift Irrigation June 1968 2.2 168-CE Drainage & Reclamation October 1969 2.6 666-CE Tank Irrigation Modernization November 1976 5.0 891-CE Kurunegala Rural Development March 1979 20.0 1166-CE Village Irrigation Rehabilitation June 1981 30.0 1537-CE Major Irrigation Rehabilitation December 1984 17.0 2260-CE National Irrigation Rehabilitation June 1991 29.6 Total 86.4 Now. 1. Amounts approwd, not disbursed. 2. Cr 1494-CE, (Mahawell IV) was canceled without being disbursed. 3. Kurundgala Rural Development was an integrated project, other Rural Development projects not included.  29 ANNEX II Cashews 11.1 In the light of the subsequent difficulty in implementing the project's proposed cashew plantation, it is worth quoting the SAR's proposal in its entirety: "3.31 Cashews. For optimal utilization of out-of-command areas and in support of the country's policy of increasing the exports of high-valued commodities, about 2,000 ha of unirrigated land would be planted to cashew trees. Land clearing would be carried out under the project. About 500 ha would be planted to cashew each year over a four-year period beginning in 1983 either by private entrepreneurs or by the Sri Lanka Cashew Corporation. Establishment costs have been estimated at about Rs 2,400/ha with maintenance costs averaging about Rs 920/ha each year thereafter, all to be borne by the planters". 11.2 In Annex 4, Table 2 of the SAR Cashews are credited with a project benefit of Rs 79.6 million (1.7% of project benefits), but on what basis is not clear. Apart from occasional passing references to the cashew plantation, that is the whole treatment of cashews in the SAR. 11.3 As the PCR (para 4.7) says ".... The SAR provided little information on each (i.e. cashew and phalloid plantations) and was unspecific about implementation responsibility. Hence Cashew development could be either by private entrepreneurs or the Sri Lanka Cashew Corporation (SLCC).....No indication of how the benefits of these developments would be shared". And, one could add, even the location of the plantation was left in doubt. 1.4 Given this rather superficial appraisal, of what was admittedly a minor component, it should not be surprising that difficulties were encountered in implementation. 11.5 In the event, in 1983 MASL asked SLCC to plant 2,000 ha to be paid for by MEA, with the intention of reserving 400 ha for a nucleus estate to be run by SLCC, and the remaining 1,600 ha to be split into 3 ha lots for individual settlers. SLCC planted 197,510 trees (100 trees/ha) over the period 1984-87; a careful survey at the end of 1990 could find only 30,000 surviving plants. Problems overlooked by the appraisal or specific site selection decisions were thin soils subject to water logging in Maha, fire, wildlife (especially elephant), weed and insect damage (PCR para 5.20). II.6 In January 1991, MASL took over the area from SLCC. Despite the poor survival of the original cashew trees it was decided to go ahead with settling 378 families on 7.5 acre blocks, in the hope that their presence would reduce the damage from wild life, and lead to better fire control. II.7 The farming system envisaged for these settlers was mono-crop cashews, with a small garden area. Since the surviving trees were quite insufficient to support the settled families, MASL provided a subsidy of Rs 50/- per cashew seedling planted and surviving one year, and a further Rs 25/- if it survived a second year. One estimate is that this has cost about Rs 28,000/- per settler family. In addition, the World Food Program provided food for two years. There is no recent data on tree survival, though field inspection suggests that it continues to be poor; no doubt reflecting the basic ANNEX 1I 30 unsuitability for cashews in Maha of the large areas of water logged or field capacity soils. Fires continue to plague the young cashew trees, and elephants continue to do widespread damage to both cashew trees, and settlers gardens. An electric fence around System B is expected to be completed shortly, at which stage the depredations of wild elephants should cease. Total assistance (including food aid and cashew planting subsidy) is thought to approximate Rs 100,000/- per family; and self- sufficiency is nowhere in sight. 11.8 MASL and the Bank, on the doubtful authority of the SAR, still have their sights set on proving monocrop cashews are a viable farming system in the selected area. 11.9 In 1992 a small NGO, supported by a bilateral donor, was provided with a representative 7.5 acre plot in the cashew area. Guided by an experienced Sri Lankan agronomist, the NGO has shown that mixed cropping can return a much higher income than mono-cropping with cashews. The key to an improved farming system seems to be capitalizing on the undulating nature of the topography, and resulting different moisture status of the soil during Maha. In the valley and gully bottoms the soil is water logged for up to four months, allowing a good (30 bushels per acre) crop of paddy to be grown, without irrigation. On the slopes, the soil though not water logged is too wet for cashews, due to water draining down from the ridges; a wide range of water tolerant crops, banana, coconut, corn, chilies, big onions, groundnuts are being experimented with. On the dryer ridges, cashews can be grown. Eucalyptus and teak are being grown as hedgerow crops for poles, phalloid and possibly cash sale. Small scale chicken and dairy production provide a small cash income and the basis for a bio-gas plant. Though not yet proven, such a mixed farming system, seems to hold promise of giving farmers both a more secure and higher income than could be obtained from mono-cropping with cashews. 31 ANNEX III Table 1. Economic Cash Flow for Mahaweli III (Rs Million) Project Year---) 1 2 3 4 5 6 7 8 9 10 11 Date - ) 1982 1983 1984 198 1986 1987 1988 1989 1990 1991 1992 A. PROJECT BENEFITS Paddy 0.0 3.3 2.3 84.b 209.4 295.5 497.4 596.4 612.0 751.0 729.4 Uther Field Crops 0.0 0.0 1.6 4.0 6.5 4.8 21.6 39.5 30.0 46.5 65.4 Fuelwood 0.0 0.8 0.0 0.0 0.8 0.0 0.0 a .0 0.0 0.0 0.8 Cashew 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.3 0.7 Homestead Plot 0.6 0.0 1.9 3.9 6.9 11.1 13.1 15.5 17.2 19.4 28.1 Dairying 0.0 1.8 3.8 s.6 7.4 9.3 11.2 13.6 16.1 18.6 21.0 TOTAL BENEFITS 0.0 5.1 39.6 98.0 230.2 320.7 543.3 665.0 675.4 835.8 836.6 B. PROJECT COSTS B. Capital Costs Right Bank Transbasin C 126.7 231.1 226.2 0.0 7.8 -0.6 0.3 16.7 -1.4 0.0 0.0 Main & Branch Canals 0.0 ;10.3 83.4 229.2 193.9 -655.6 372.7 617.9 0.0 0.0 0.0 Irrig. & Dr. Dev. 23.8 68.0 186.2 362.2 243.3 166.3 191.4 131.3 107.4 136.9 18M.5 Soc. & Admin. Infrast. 5.8 28.0 16.5 1b4.1 31.7 41.0 32.5 23.6 "9.4 40.7 23.1 Equipment 20.8 26.8 2.6 0.0 27.7 2.2 80.1 t5.8 83.5 3.9 .7 Roads 0.0 0.0 0.0 16.4 8.5 4.6 5.4 7.6 2.4 14.7 14.5 Cashew & uelwood 1.4 8b.7 0.0 :3.8 43.5 101.5 8.4 21.6 -21.1 9.3 8.7 Tech. Ass. & Mon. 14.7 1.3 29. 1 31.2 34.9 59.6 75.7 1 4.5 54.7 99.5 106.8 SUBTOTAL 1 193.2 665.2 544.0 816.1 590.5 1830.2 '66.5 999.0 54.9 30 854.3 8. Hecurent Costs Staff Sala. & Trav. 0.0 19.3 3.6 119.3 76.1 108.6 97.3 114.0 12.5 184.8 64.5 Yehicles 0P 18.7 0.0 0.0 1 . 3 b7,9 41.7 31. 1 11.0 17,1 11.4 12,8 Equipment OP 0.0 0.0 0.0 16.8 2.2 1.1 36.6 8.5 b.9 -0.4 3.3 Materials 0.0 1.1 0.0 0.2 1.1 5.6 2.1 2.8 3.6 3.,3 3.1 Other UP 0.0 0.0 0.0 2.3 26.5 30.8 4.1 7.2 14.3 7.6 27.4 %UBTOTAL 2 18.7 21.0 3.6 150.9 163.8 187.8 171.4 143.5 143.4 126.7 111.1 83. Crop Production Cos Paddy Prod. Costs 0.0 4.7 32.6 44.4 131.3 161.7 244.9 249.5 319.4 354.0 364.2 UFC Prod. Costs 0.0 0.0 0.7 1.8 2.8 2.1 9.4 17.2 13.2 20.2 28.4 Dairying Prod. Costs 0.0 0.6 1.2 1.9 2.5 3.1 3.7 4.5 5.4 6.2 7.o SUBTOrAL 3 0.0 5.3 34.5 48.1 136.6 166.9 53.0 271.2 338.0 380.4 399.6 34. NET BENEFITS FOR6ON 14.1 14.1 17.2 18.4 15.3 16.5 28.9 34.8 23.9 d9.0 24.1 0TAL CUSTS 226.0 705.6 599.3 1033.5 906.2 1401.4 1;24.8 1448.5 /60.2 841.1 789.1 C. TOTAL NET BENEFITS -226.0 -700.5 -b59.7 -935.b -676.0 -1080.7 -681.b -783.5 -84.8 -5.3 47.b Memo Item Paddy (000tons) 3-6 0.0 0.3 3.5 10.8 31.9 45.4 63.3 71.3 85.2 98.0 103.6 Paddy (MB'tons) 1 0.0 0.0 0.0 4.6 5.0 5.1 5.4 J.1 5.7 b.0 5.2 PAR Total Benefits 0.0 5.1 39.6 133.9 262.8 353.8 585.8 707.4 716.1 874.2 873.4 PAR Total Cost 226.0 705.6 599.3 1052.4 926.6 1419.5 1245.7 1466.3 /81.4 859.2 807.5 D. PAk TOTAL NET BENEFITS -226.0 -700.5 -559.7 -918.5 463.9 -106b.7 -659.9 -758.8 -65.3 15.0 65.9 ANNEX I1 32 Table 2. Economic Cash Flow for Mahaweli III (Rs Million) Project Ypar --) 12 13 14 15 16 17 18 19-23 24-50 Date - ) 1993 1994 1995 1996 1997 1998 1999 200-04 2005-31 A. PROJCT BENEFITS Paddy 772.1 841.8 913.4 886.0 904.4 904.4 92.8 922.8 869.8 Other Field Crops 68.7 68.7 68.7 68.7 68.7 68.7 68.7 68.7 68.7 Fuelwood 0.0 0.0 0.0 0.0 0.8 0.0 6.0 6.0 6.0 iashew 1.3 2.2 3.2 4.2 5.0 5.8 6.2 6.3 6.3 Homestead filot 23.6 25.2 26.1 26.1 21 2 26.1 26.1 26.1 26.1 Dairying ?1.8 21.0 41.8 21.8 41.0 21.0 21.0 21.0 21.0 TOTAL BENEFI1S 886.7 958.9 132.4 1006.0 WS25.2 1k6.0 15.8 150.9 997.9 8. PHOJECT COSTS Bl. Capital 1Cost5 Right Bank Transbasin L 0.0 0.0 0.0 0.0 0.0 - 0.0 0.0 0.0 0.8 Nain & Branch Canals 0.0 0.0 0,0 0.0 0.0 0.0 0.0 0.0 0.0 Irrig. & Dr. Dev. 42.9 21.8 0.0 0.8 0.e 0.e 0.8 0.0 0. Soc. & Adnin. Infrast. 24.6 9.5 8.8 8.8 8.8 8.8 . 0. 0.0 Equipment 0.7 0.7 0.8 0.0 0.0 0.0 0.0 0.8 .0 Roads 8.1 1.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Cashew & fuelwood 5.8 3.9 0.0 0.0 0.0 0.0 0.0 0.4 0.0 Tech. Ass. & Mon. 91.6 1,5 0.0 0.0 0,0 o.o 0.9 0.0 0.0 sU8TOTAL 1 173.7 38.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 B2. Recurent Costs Staff hala. & Trav. 79.8 61.4 61.4 61.4 61.4 61.4 61.4 61.4 61.4 Vehicles OP 15.1 15.1 15.1 15.1 15.1 15.1 15.1 15.1 15.1 Equipeent OP 0.8 0.0 0.0 0.8 0.0 0.0 0.0 0.0 0. Materials 5,6 8.0 8.0 8.0 8.0 8.0 8. 8.8 8 Other OP 24.2 28.4 20.4 20.4 20.4 20.4 20. 4 20, 4 20.4 SUBTOTAL 2 125.5 104.9 104.9 104.9 104.9 104.9 104.9 104,9 104.9 83. Crop Production Los Paddy Prod. Losts 412.4 454.4 475.9 4/5.9 4-/5.9 475.9 4·/5.9 4/5.9 475.9 OFC Prod. Costs 29.8 29.8 29.8 29.8 9.8 29.8 29.8 29,8 29,8 Dairying Prod. Costs 7.0 7.0 7.0 7.0 7.0 7.0 7.0 7.0 7.0 !UBTOTAL 3 449.2 491,j b 12. 7 j1,7 12.7 512, 7 51;. 7 12.7 512.7 84. NET BENEFITS FORGN 18.8 18.5 20,0 18.3 18.0 18.0 18.0 18.0 18.0 TOMAL COT 767.;<,2 . 637.6 63i,9 63b.6 b.Si.6 63å.6 63j.6 63.6 C. TOTAL NET BENEFIls 119.5 305.4 394.8 370.1 369.6 390.4 415.2 415.3 362.3 memo Iten Paddy (MIe'tons) ó-6 118.7 1å.5 19.2 19.2 142.1 142.1 145.0 145.0 144.9 Paddy (OS8'tons) 1 5.4 5. 4 i.4 b 4 .4 i 4 ý.4 n 4 5.4 PAR Total Benefits 921.7 993.6 1067.7 1048.2 1859.4 1060.2 108b.O 185.1 1030.1 PAR fotal Cost 785.9 672.2 656.3 654.3 603.6 633.6 653.2 653.2 653.2 D. PAR TO1AL NET BENEFITS 135.8 321.3 411.7 385.9 405.8 416.6 431.8 431.9 376.9 33 ANNEX IV REVISED ANNEX I FOR PART LU OF THE PROJECT COMPLETION REPORT FOR THE VILLAGE IRRIGATION REHABILITATION PROJECT Project Completion Report Sri Lanka Village Irrigation Rehabilitation Project (Credit 1160-CE) Table 1 FY Pre-project- Cropped Arealha) Annual Command Area ................... Annual................... Generated Maha eason Yala'Seson Rainted Irngated Irrigated irrigated Cropping paddy paddy paddy chillies intensity(%) 82 80 24 56 24 0 130% B3 1030 309 721 309 0 130% 84 3060 918 2142 918 0 130% 85 3340 1002 2338 1002 0 130% 86 2800 840 1960 840 0 130% 87 4035 1211 2825 1211 0 130% BB 4535 1381 3175 1361 0 130% 9 2180 654 1526 654 0 130% 90 2140 642 1498 642 0 130% 91 2880 864 2016 864 0 130% cummutativ FY91 26080 7825 18257 7825 0 130% ;,，。，。，,，。。；i ！要 謬 &.“。。。：。。。，! 濺k 【。“懿龍弱：！氤， 【。’一”&&&lh誰：喜 弓l認J洶、澀 論矓叫｝酌龍暮i& 騙州練州啊霉“ ＿禹訪磚痲 ~＿挺煙認牌認 &&‘伯編洛’‘蘊萬‘!!I-: _＿蘿露一―審 。”。”&&&&&&”一‘變 ,,,,,,,,,,,l身豐 l&&&&f&& 籐 。蘊 ‘藝“‘盤“藝“-i 話藝 鬢。他才 i必弓邊毒露喜弓認易森―b言常 I―賽11：喜臘 ！讓，·，”一！!―〕！一‘｛霄i 蠱蘿露籐禁呈要益靈必，遝織―量召． 。，謹笞錢 !&&&&&“】；!I; ！〕！！〕！〕〕！！〕；;-! 華離離細頸華華華離華果麾騷二、 i 要 ANNEX IV 36 ~ject Compietion Repon Sri Lanka Vidage Imgarion Rehabduatzon Project (Credir 1160-CE) Tabk 4. Per Ha Financial Fann Budget Pre ana WiIMOU? PrOICC! Ramntec mana wrig. Yata kmp Chilkes Paacry paacy pacm Farm Poworih~ RsO) 570 SM SM ew 1590 too 15W 19e0 Secd-*9 110 110 110 1 - pa 1045 10" 1045 200 20 55 70 90 198 -11M am ma 20 55 60 80 -85 194 533.5 582 778 MOP- kg 10 35 55 85 -R* 91 318.5 5= 591.5 ch*mcw4-p4 300 9W 9m 2= Mhem 300 m imo Latmtr-nred(mdw 25 75 70 135 -R* 1500 4m0 4200 8100 Total Ca&h Cosn-Rs 4198 9308 9254 15010 Yledsturla) 1.3 2.6 2.3 1 Grogg produccon va~RO 9100 8= 16100 325X Femdy Lahourlmd) 20 50 45 90 1) Average at ac= 25% Ot farm POwfir recunmBryt 2) Averagt at aciout 60% 01 t~ ImbOu r reaLpremort 37 ANNEX IV Project Compledon Report Sli Lanka Vilage Imgarion Rehabiliuatwn Project (Credit 1160-CE) Table 5. Per Ha Financial Farm Budget Witn Pmtect Raintem Mana irig. Yata ing Chilles Poady Pay Paddy Farm Pown-ed-RalI 570 570 570 80 AD 1590 150 1N0 1S S.ea-cg 110 110 110 1 -A 1045 1045 1045 200 Ur«e-kg 20 70 80 90 -A 196 65 784 8U TSP-kg 20 60 75 80 -A 194 582 727.5 776 MOP-kg 10 50 60 65 -A 91 45 548 591.5 Chemuas-Pa 3® 975 90 2=0 Oxho 300 9®0 m 1100 LaoLr-nrea(md)2/ 25 90 75 135 -Ra 150 5400 450 8100 TotmlCas Coat-RAs 4196 10613 9973 15010 Yil0s/a) 1.3 3.1 2.8 1 Grsaa proaucaon Vaium44 9100 21700 19600 32500 Family Labourimc4 20 40 45 90 a) AMwage at anut 25% of =tal trm Dwe renurummen b) Aveago a aau 60% of to~ sbour roqum ANNEX IV 38 Project Completion Report Sri Lanka Vdlage Irigation Rehabilitation Project (Credit 1160-CE) Table 6. Farm Incomes - 0.75 Ha Farm Present Without i Cropped Area (ha) NMha Season: - Rainted Paddy 0.23 0.33 0.20 - Irrigated Paddy 0.53 0.42 0.55 Yala Season: - Irrigated Paddy 0.23 0.18 0.25 - Chillies 0.00 0.00 0.04 Total 0.99 0.93 1.04 Gross Production Value (Rs) 15,440 13,545 19,955 Production Cost (Rs) 8,030 6,965 9,770 Net Farm Income (Rs) 7,410 6,580 10,185 39 ANNEX IV Project Completion Report Sri Lanka Viflage Irrigation Rehabilitation Project (Credit 1160-CE) Table 7. Economic Analysis (Rs Million) Prooct O&M Costs Without Project With Project FY Costsl/ WOP21 WP3/ Prod Costs Benefits Prod Costs Benetits Net Benefits 4/ 4/ 81 20.5 0.0 0.0 0.0 0.0 0.0 0.0 -20.5 82 101.9 0.0 0.0 1.2 3.0 1.3 3.0 -102.0 83 101.8 0.2 0.6 16.7 36.0 18.1 39.6 -100.0 84 94.7 0.8 2.3 61.9 116.9 68.5 130.8 -88.9 85 115.1 1.3 4.2 109.5 177.0 123.0 205.2 -103.2 86 164.8 1.8 5.8 148.2 235.5 169.3 349.1 -76.4 87 146.6 2.5 8.1 203.6 322.2 234.6 374.9 -130.5 88 68.1 3.3 10.7 264.7 411.4 307.3 512.4 -17.1 89 63.7 3.6 12.0 290.4 448.9 342.5 572.9 -0.2 90 77.5 3.9 13.3 314.6 477.5 375.7 620.7 -4.9 91 0 4.3 14.9 351.1 461.2 422.5 608.7 65.5 92 0 4.2 15.0 347.0 462.4 426.0 627.2 75.0 93 0 4.1 15.1 342.9 463.6 427.9 641.6 82.0 94 0 4.0 15.1 339.5 465.4 429.3 653.4 87.1 95 0 3.9 15.1 336.5 467.9 430.0 663.4 90.9 96 0 3.9 15.1 334.2 470.9 430.7 672.7 94.0 97 0 3.8 15.1 332.4 468.0 431.2 672.7 94.6 98 0 3.8 15.1 331.6 473.0 432.0 681.9 97.3 99 0 3.8 15.1 331.1 471.7 432.5 681.9 97.4 00 0 3.8 15.1 331.0 477.7 433.2 691.1 99.8 01 0 3.8 15.1 330.9 471.0 433.0 681.9 97.4 02 0 3.8 15.1 330.7 457.5 432.8 663.4 92.5 03 0 3.8 15.1 330.7 450.8 432.7 654.2 90.0 04 0 3.8 15.1 330.6 437.4 432.5 635.7 85.1 05 0 3.8 15.1 330.4 430.6 432.3 626.5 82.7 Economic Rate of Return = 5.6% 1/ See Annex 1. Table 8. 21WOP: estimaia at Rs 200lta. adjusted by CF of 0.90. 3/ WP: estimatso at Rs 600/ha. adjusted by CF of 0.87. 4/ Incluoing famifiy iaboour and owned farm power. ANNEX IV 40 Project Completion Report Si Lanka Vilage Inigation Rehabilitation Project (Credit 1160-CE) Table 8 Project Costs (In Rs Million) Project Cost in Project Cost in Curent Prices Constant 1991 Prices FT1 Iotal a/ Excl. 7 Dist., Adjust. b/ MEM KE Financial Economic c/ 1981 12.5 - 2.6 9.9 1.41 2.70 23.5 20.5 1982 65.2 -14.0 51.2 1.43 2.56 116.6 101.9 1983 84.4 -23.6 60.8 1.47 2.05 115.8 101.8 1984 93.6 -26.6 67.0 1.50 1.63 107.0 94.7 1985 100.9 -29.1 71.8 1.49 1.93 130.7 115.1 1986 151.0 -46.5 104.5 1.26 1.98 188.1 164.8 1987 150.3 -45.8 104.5 1.15 1.75 167.2 146.6 1988 75.8 -19.9 55.9 1.07 1.49 77.4 68.1 1989 76.4 -20.4 56.0 1.08 1.36 72.2 63.7 1990 113.5 -33.0 80.5 1.02 1.11 87.5 77.5 &/ See Attachzent 1. b/ Excluding 25% of ID's and 10% of DAs' civil works costs, equivalent to 35% of total civil vorks costs. c/ About 25% foreign exchange and 75% local costs. Local costs adjusted by CF of 0.85. 41 ANNEX IV Project Completion Report Sri Lanka Village Irrigation Rehabilitation Project (Credit 1160-CE) Table 9. Summary of Financial and Economic Prices' unt FVurJai Economic (19o1) 1980 1985 1990 1991 2000 205 Outs: Paddy Rsc 7 15.8 8.2 7.4 6.4 7.1 6.4 CrIues Rst 32.5 27.6 27.6 27.6 27.6 27.6 27.6 Inut: Lrea RsMCO 9.8 21.0 13.5 10.6 10.5 11.4 10.9 TSP Rag 9.7 18.8 13.8 9.3 9.6 11.0 11.1 MOP RsACQ 9.1 13.4 10.2 8.2 7.8 8.3 8.3 UrkWed labour Ahn 60 54 54 54 54 54 54 Seeds: Paddy RaLag 9.5 8.1 8.1 8.1 8.1 8.1 8.1 Chilles RCg 200 170 170 170 170 170 170 Farn Power Raf 2500 2125 2125 2125 2125 2125 2125 1/ From SAR on National itation Rehabwftatlon Prolect. May 1991. Arnex 12. Table 2: Prces ot non-traded goods ad sted by CF of O.85 and unskilled labour by 0.90.  REVISED ANNEX I FOR PART II OF THE PROJECT COMPLETION REPORT FOR MAHAWELI III Table 1. Calculation of Import Parity Price of Paddy im99 cons. Prevailing Exchange C.I.F. value Port, Tr. Farm gate rice Year price 1/ Foreign price F & I rate Colombo Local Rice Paddy US$/ton Inflator US$/ton US$ USs=Rs US$ Rs Inflator FsI (Rs) (Rs 1983 208 1.47 208 1701 2500 2250 56252 201 447 8 60729 41296 1984 197 1.50 197 1667 2628 2137 56152 1 60 562.5 61777 42008 1985 188 1 49 188 1681 2741 2048 56138 1 89 4762 60900 4141 2 1986 160 1.26 160 1984 2852 1798 51291 1 94 4639 5593.0 38032 1987 198 1.15 198 2174 3076 2197 67592 1 71 526.3 7285.5 49541 1988 261 1.07 261 2336 3303 2844 93926 1 45 6207 10013.2 68090 1989 269 1.08 269 2315 4000 292.1 116859 1 33 6767 12362.6 84066 1990 258 1.02 258 2451 40.24 282.5 113682 1 09 825.7 12193.9 8291.8 1991 31; 1.00 314 2500 42.58 339.0 14434.6 100 900.0 15334.6 10427.5 1992 286 286 2500 4258 311.0 132424 90Q0 14142.4 9616.8 1993 262 262 2500 42.58 287.0 12220.5 9000 13120.5 8921.9 1994 260 260 25.00 4258 285.0 12135.3 9000 13035-3 88640 1995 265 265 25.00 42.58 290.0 12348.2 9000 13248.2 9008.8 2000 256 256 25.00 42.58 281.0 11965.0 9000 12865.0 87482 2005 onward 240 240 25.00 42.58 265.0 11283.7 9000 12183.7 8284.9 I/ The prices until 1991 were the actual prices Table 2. Paddy Production and Value of Zone 3-6 Product. Price Value Year (tons) (Relon) (Rs Mill) 1983 340 4129 6 1.40 1984 3524 42008 1480 1985 10797 4141.2 44.71 1986 31916 38032 121.38 1987 45409 4954.1 224.96 1988 63304 6809.0 431.04 1989 71304 8406.6 59942 1990 85206 8291 8 706.51 1991 98038 104275 1022.29 1992 103585 9616.8 99616 1993 118659 8921 9 105866 1994 130494 88640 1156,70 1995 139170 90088 125375 1996 139170 90088 125375 1997 142065 9008 8 1279.83 1998 142065 90088 127983 1999 144960 90088 130591 2000 144960 8748.2 1268141 2005 144960 8284.9 1200 98 Table 3. Paddy Production and Value of Production in Zone 1 Cultivated Extent 1/ Harvested Extent 2/ Average Yield 3/ Total Year (ha) (ha) (tons/ha) Produc. Value 4/ Maha Yala Maha Yala Maha Yala (tons_ (Ps Mill) 1985 395 1185 335.75 1007.25 3.582 3.361 4588.02 19.00 1986 395 1185 33575 1007.25 3.885 3632 4962.72 1887 1987 395 1185 33575 1007.25 4.569 3531 5090.64 2522 1988 395 1185 335.75 1007.25 4.068 4010 5404.90 3680 1989 395 1185 335.75 1007.25 3.525 3861 5072.51 4264 1990 395 1185 33575 1007.25 3.951 4307 5664.77 46.97 1991 395 1185 335.75 1007.25 3.824 3.701 5011.74 5226 1992 395 1185 33575 1007.25 3.556 4 5222.93 5U.23 1993 395 1185 33575 1007.25 4 4 5372.00 4793 1994 395 1185 33575 1007.25 4 4 5372.00 47.62 1995 395 1185 335.75 1007.25 4 4 5372.00 4840 1996 395 1185 335.75 1007.25 4 4 5372.00 4840 1997 395 1185 335.75 1007.25 4 4 5372.00 48.40 1998 395 1185 33575 1007.25 4 4 5372.00 4840 1999 395 1185 33575 1007.25 4 4 5372.00 4840 2000 395 1185 335.75 1007.25 4 4 5372.00 4700 2005 onwar 395 1185 335.75 1007.25 4 4 5372.00 44.51 1 / 750%1 in Yala and 25% in Maha out of total 1,580 ha 21 0.85 of the cultivated extent 3/ Source: Dept. of Census and Statistics until 1991/92 Maha season 4/ Production multiplied by the price given in Table 1 Table 4. Calculation of Cost of Fertilizer thea [SP moll 1991 con Post. Tr. Exchange Farm gate price 1991 cons Farm gate rice 1991 cons FarM gate price Year Foreign price 1! F & I Local & Proce. ate (Jrea N price TSP P price MOP K Inflator US$/ton uS$ InflatOr ils/ton US$=Fls Rsiton Rs/kg JSS/lton Fsiton Rs/kg USSIton [isiton l1s/kg 1983 1.47 114O 17 01 201 4478 2500 37229 09 114 00 3722 93 7.92 112 3672,93 612 1984 150 1084 1667 160 5625 2628 38493 837 10845 385057 819 126 431178 719 1985 149 1040 1681 189 4762 2741 37875 823 10398 378699 806 125 4363 14 727 1986 126 999 1984 194 4639 28.52 38789 843 9993 387979 825 86 348251 580 1987 115 927 2174 171 5263 30.76 40465 880 9265 404493 861 79 362505 604 198 107 1105 2336 145 6207 3303 50422 1096 11051 504256 1073 94 449724 750 1989 108 913 2315 133 6767 4000 52546 1142 9125 525262 11 18 107 588262 980 1990 102 1938 2451 109 8257 40.24 96105 2009 23733 11362.12 24 17 100 583596 973 1991 100 2419 2500 1.00 9000 4258 122646 2666 24190 1226460 2609 109 660572 1101 1992 1500 2500 9000 4258 8351.5 1816 12200 720926 1534 109 660572 1101 1993 1550 2500 9000 42.58 8564.4 1862 12600 7379.58 15.70 110 664830 11 08 4 1994 1610 2500 9000 4258 88199 1917 13400 772022 1643 112 673346 1122 1995 1650 2500 9000 42.58 8990.2 1954 13700 784796 1670 110 664830 1108 2000 1740 25.00 900.0 42.58 93734 20.38 134.00 7720.22 1643 106 647798 1080 2005 onward 171.0 25.00 900.0 42.58 9245.7 20.10 131 00 7592.48 16.15 106 6477 98 10 0 1/ The prices given until 1991 are the actual prices Table 5. Cost of Production of Paddy for the Revised Economic Analysis 1983 1984 1985 1986 1987 1988 quantity Rate Coat Rate Cost Rate Cost Rate Cost Rate Cost Rate Cost Seed 209 5 16 107586 5 25 109746 5 18 1081 88 4 75 993 59 6 19 1294 27 8 51 117885 Ferillizer N 138 809 1116 8 637 1154 78 8 23 113626 8 43 116360 860 121394 1096 151267 -P 34 792 26932 a 19 27055 606 27395 825 28067 861 29261 10 73 364 78 -K 50 6 12 30608 7 19 35932 727 36360 580 29021 604 30209 7 50 374 77 Chemical 4 77lit 138000 138000 138000 138000 138000 138000 0.45 kg tabour - Hired 109 1 5265 5744 12 5265 5744 12 5265 5744 12 5265 5744 12 5265 5744 12 526b 5741 12 - Fangy 1222 5265 6433 83 5265 6433 83 52.65 643383 52.65 6433 83 5265 6433 83 5265 6433 83 Animal 213 22970 489261 22970 489261 22970 489261 22970 489261 22970 489261 22970 489261 Tractor 62 651.50 403930 651 50 403930 651 50 403930 851.50 4039 30 651 50 403930 651 50 403930 Misc. 1621 00 1621 00 1621 00 1621 00 1621 00 1621 00 Total 2688199 2700096 2696655 2683901 2721376 2814193 Pet hectare 14298.93 14362 21 1434391 1427607 1447 40 14969 11 1989 1990 1991 1992 1993 1994 Quantiy Rate Cost Rate Cost Rate Cost Rate Cost Rate Cost Rate - Cost Seed 209 1051 219622 1036 2166 24 1303 272420 1202 251239 11l1 233085 1108 231572 Fedrifizer - N 138 11.42 157639 2089 288314 2666 367938 18 16 250545 1862 2569.32 19 17 264596 - P 34 1118 37998 2417 82194 2609 887.23 15 34 52152 1570 53384 1643 55848 - K 50 980 49022 973 48633 1101 55048 11.01 55048 11 08 554 03 11 22 561 12 Chemical 477iht 138000 138000 138000 138000 138000 138000 0.45 kg tabour - ired 109 1 5265 5744 12 5265 5744 12 5265 5744 12 5265 5744 12 5265 5744 12 5265 5744 12 - Famly 122.2 5265 643383 5265 643383 5265 643383 52.65 6433 83 5265 643383 5265 643383 Animal 21 3 22910 489261 229.70 489261 22970 489261 22970 489261 22970 4892 61 22970 489261 Tractof 62 651 50 403930 651 50 403930 651.50 403930 651 50 403930 651 50 403930 651 50 403930 Misc. 1621 00 1621 00 1621 00 1621 00 1621 00 1621 00 Total 2875365 3046851 31952.13 3020070 3009809 30192 15 Per hectare 1529450 1620666 16995.82 1606420 1601005 1605965 1995 1996 1997 1998 1999 2000 2005 onwards Quantity Rate Cost Rate Cost Rate Cost Rate Cost Rate Cost Rate Cost Seed 209 1 i26 235354 1 26 -235354 11 26 235354 1 i26 235354 1126 0235354 094 228546 1036 216443 Fertilizer - N 138 1954 269706 1954 269706 1954 269706 1954 269706 1954 269706 2038 281203 2010 277370 - P 34 1670 56772 1670 56772 1670 56772 1670 56772 16.70 56772 1643 55848 16 15 549 24 - K 50 1108 55403 1108 55403 1108 55403 1108 55403 1108 554 03 1080 53983 1080 53983 Chemical 4 77lit 138000 138000 138000 138000 138000 138000 138000 045 kg labour - tred 109 1 52 6b 5744 12 52tt 5744 12 52 65 5744 12 52 65 5114 12 5265 5744 12 521 5 51 12 5rb ,il4 12 - FaIngy 122 2 5265 6433 83 5265 643383 5265 643383 5265 6433 83 5265 5433 83, 5265 643383 5265 643383 Animal 21 3 22970 489261 22970 489261 22970 489261 22970 489261 22970 489261 22970 489261 22970 489261 Tractor 6 2 651 50 403930 851 50 4039 30 651 50 403930 651 50 4039 30 651 50 4039 30 651 50 4039 30 651 50 4039 30 Misc. 1621 00 1621 00 1621 00 1621 00 1621 00 1621 00t 1621 00 Total 30283 21 3028321 3028321 30283 21 30283 21 3030666 3013807 Per hectare 1610809 1610809 1610809 1610809 1610809 1612056 1603089 Table 6. Economic Cash Flow 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 A. BENEFITS (Rs Million) Paddy (Zone 3 to 6) 1 40 14.80 44.71 121.38 22496 431.04 599.42 70651 102229 996 16 Paddy (Zone 1) 19.00 18.87 25 22 36.80 42.64 46.97 52.26 50 23 OFCs 1 2.12 3.35 2 81 14.9 29.7 27.52 46.5 654 Fuelwood 0.3 0.7 Cashew 0.09 19.4 20.1 Homestead 1.19 2.06 3.56 6.49 9.03 11.65 15.78 18.6 21 Dairing 2.37 2.96 3.81 5.44 7.72 10.23 14.77 18.6 21 Total 0.0 1.4 19.4 70.9 151.0 264.9 499.5 693.6 811.6 1178.0 1174.6 B. COSTS (Rs Million) Capital cots 98.07 382.30 329.70 482.90 369.13 720.49 608.41 829.05 241.52 30500 254.3 Recurrentcost 9.49 12.07 2.18 89.29 102.38 131.33 136.03 119.09 135.92 126.70 111.1 Paddy (Zone 3 to 6) 4.10 27.88 39.22 120.99 151.25 228.22 240.90 328.65 378.29 373.67 Paddy (Zone 1) 23 23 23 24 24 26 27 25 OFCs prod. cost 0.44 0.90 1.44 1.23 6.48 12.93 12.11 20.20 28.4 Dairy prod. cost 0.30 0.75 1.01 1.29 1.81 2.55 3.38 4.95 6.20 7 Net benefit forgone 5.62 7.01 10.57 9.74 7.89 9.65 19.93 26.17 21.93 29 24.1 Total 113.2 405.8 371.5 _ 645.7 625.7 1038.6 1025.3 1255.7 770.7 892.2 824.0 Table 6 (Continued) 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005-2031< 105866 115670 125375 1253.75 127983 127983 130591 1268.14 1268.14 126814 126814 1268 14 120098 47.93 47.62 48 40 48.40 48 40 48.40 48 40 47.00 47.00 47.00 47 00 47 00 44 51 68.7 68.7 68.7 68.7 68.7 68.7 68.7 68.7 68.7 68.7 68.7 68 7 68.7 13 22 32 42 5 58 62 63 63 63 63 63 63 23.6 25.2 261 26.1 261 26.1 26.1 26 1 261 26.1 26.1 261 26 1 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 1242.2 1342.4 1442.1 1443.1 1470.0 1470.8 1497.3 1458.2 1458.2 1458.2 1458.2 1458.2 1388.6 1737 38.9 125.7 104.9 1049 104.9 1049 104.9 104.9 1049 1019 104.9 104.9 104.9 1049 422.67 465.73 487.80 487.80 487.80 487.80 487.80 488.18 488.18 488.18 488.18 488.18 485.46 25 25 25 25 25 25 25 25 25 25 25 25 25 29.8 29.8 29.8 29.8 29.8 29.8 29.8 29.8 29.8 29.8 29.8 29.8 " 7 7 7 7 7 7 7 7 7 7 7 7 7 18.8 18.5 20 18.3 18 18 18 18 18 18 18 18 18 803.0 690.2 675.0 673.3 6730 673.0 673.0 673.3 673.3 673.3 673.3 673.3 670.5 Economic Rate of Return Discount rate 10.00% 1.00% 11.00% PV of project benefit. 6914.15 44915.57 5931.77 PV of project cost 655082 26919.77 5918.92 BCR 1.06 1.67 1.00 ERR 49 ANNEX V Table 7. Farm Budget SAR BMS 1/ BMS 2/ Total inflow 54298 58157 58157 Operating costs Seed 765 1668 1668 Fert N 2060 P 538 K 392 Total 2990 4444 4444. Compounaing and other costs (21%) 635 Sub total of fertilizer 3625 Pesticides 1004 1393 1393. Labour 260 7095 4147 Animal power 2042 5436 5436 Tractor 4039 4039 Miscellaneus 1019 1801 1801 OFCs operating cost 277 277 Sub total per season 8714 Total operating costs + 26153 23205 10% contingencies 19172 Crop insurance 894 Water charges 1255 30 30 i Land payment 2806 Total Outflow 24126 26183 23235! Net benefit before financing 30172 31974 34922! BMS 1 Given as same in the PCR BMS 2/ Only hired labour considerea (63.8 labour units)  51 ANNEX VI ERR Methodology Comparisons for Mahaweli III VL1 The PCR for Mahaweli I contains two very different estimates of the ERR (para 4.23). This difference was one reason for choosing to audit this project. As discussed below and summarized in Table VL1, the major differences between the Borrower and the Bank revolve around methodology rather than data. Table V.1. Bank, Borrower and Audit Views on ERR Calculations Issue Bank Borrower Audit Rice Price Deflate 20% due to Don't deflate. Sri Deflate 20%. See higher level of Lankan paddy para 431. brokens in imports. quality equal to US bench-mark. Transport Farm to Subtract Add Subtract. Extra Market production sold in Colombo. Benefits to Zone 1 Exclude Include Include Cost Basis Constant Historic Constant VL2 The Borrower's justification for using historic prices is stated, ".... since the investment cost of Mahaweli System C is a 'sunk cost' the past benefits and costs should be valued at the economic prices which prevailed during each year (at historic values), and only the future benefits and costs should be estimated at base year (1991) economic values" (PCR II 3.0). VIL3 The reference to "sunk costs" reminds us that there are two ERRs which could be calculated. The first is for the project as a whole, and the second for any further required investments (Le. ignoring all costs incurred, and resulting benefits, and asking only what is the extra return on extra costs?). We are not engaged in answering this latter question, since this would involve calculation of firther benefits from further expenditures, and this clearly has not been attempted. VI.4 The internal economic rate of return, ERR, refers to the break-even interests rate in constant dollars. Accordingly, we can either use prices in constant dollars, obtained from say the Bank's Commodities Division, as Operations did, or work in actual prices in current dollars but then correcting them to constant dollars by some accepted deflator such as the index of Manufactured Goods Unit Value (MUV). The Borrower's calculations used Sri Lankan prices in current dollars without the required correction. VL5 The Borrower believes that the benefits from making extra water available to the already settled Zone 1 should be included in the benefits. This undoubtedly is a benefit, however the Bank excluded 20 percent of the costs of main roads and half the cost of the Right Bank Trans-Basin Canal to reflect the unmeasured benefits that would be obtained from these investments outside the project ANNEX VI 52 area. However, to provide an additional sensitivity type estimate of the ERR, it has been r-zalculated including the Zone 1 benefits, (Annex M). This increases the Bank's calculated ERR from 3.98 percent to 4.24 percent VI.6 In addition to these issues of principle, there are also some disagreements with respect to the data itself: i) Paddy Price. The Bank worked from the Commodity Division's estimate of Thai rice export prices, whilst the Borrower used actual average c.iL import prices where available. Comparison of these prices for 1983 and 1988 (Table VL2) suggest that if anything the Bank under estimated the discount (whether based on quality or good bargaining) actually achieved by Sri Lanka. For future years both calculations are based on the Commodity Division's projections. The difference being primarily due to adding or subtracting the cost of transport to Colombo to obtain farm price. The Audit believes that by using actual import price, the Borrower under-estimated the value of project paddy, since this is probably of better quality than is imported. ii) Yields. As shown in Table VI.3, the yields used by the Bank, are noticeably higher than used by the Borrower, with the exception of 1985. Table VI2. Paddy Prices Used by Bank and Borrower (Rp/ton) Year Bank Constant Borrower Current Borrower 1991 to 1991 Constant 1991 Constant Later 1983 9,744 4,129 6,960 1988 7,858 6,809 7,338 1993 6,507 8,922 8,922 2000 6,366 8,748 8,748 2005 on 6,000 8,285 8,285 Sourc. PCR III, Annex 1, Table 3, and Audit, Anna IV, Table 1. Table VI3. Paddy Yields Used by Bank and Borrower (tons/ha) Maha Yala Year Bank Borrower Bank Borrower 1985 1.24 3.58 NA 3.36 1988 4.62 4.07 4.02 4.01 1992 4.80 4.56 4.50 4.00 1993 4.90 4.00 4.50 4.00 iBRD 24287 化 DECEMB〔R 1992  l BRD 13477R2 79o060 8,'O ol 30 SRI LANKA MAHAWEI GANGA TECHNICAL ASSISTANCE PROJECT Existing and Proposed Irrigation Schernes . Project boundaries Jaffna Exishng canals -- - Proposed canak Irr,gated areas Proposed irrigahon areas Existing dams and reservoirs -3Proposed dams ond reservoirs Existing weirs and reservoirs • Proposed wetrs and reservoirs SExistng power station 0 Proposed power stations 0 Proposed pumping station Rivers Roads L -- Raif.oys Mannar A Irrigaton service areas 9.00- 2 Natural reserves and sancruares NOR H C :RAL - -NCRB Complex RIV R ,NS S"' ya - Proposed tunnels - Area boundores wocch,ya Trincomolec 4 5 1 15 2 25 Anuradhapj,a MILES kl 0 10 20 30 40 N AH KJROM*ETERS PuttolomY - 00*3 -00. Dombul a 2 Bttcaloa Kurunegala .- rna ra .f fl, e n e C ran ra Nuwaa ". oo IPyy ~aaa.-* 00, p~~Nci tt a 0aa.-f "COLOMBO Hottn PAKCSTANA lN DI A A RE A- OJF MAP SRI LANKA -830 JULY 1985  IBRD 15154R1 Lu uE z0 z0 > -~~L T Z0 t' E Z - z 0 li n --. Z -~ 0 I/o *- .o-5* - E~O DE~MBER 199