Sri Lanka - First and Second Mahaweli Ganga Technical Assistance Projects

Document of The World Bank FOR OFFICIAL USE ONLY Repon No. 7336 PROJECT PERFORMANCE AUDIT REPORT SRI LANKA MAHAWELI GANGA DEVELOPMENT PROJECT II (CREDIT 701-CE) MAHAWELI GANGA TECHITCAL ASSISTANCE PROJECT (CREDIT 979-CE) June 30, 1988 Operations Evaluation Department This document bas a resticted distibuion and may be used by recipients only in the perfonnance of .~ ~ .. . ., . . . .-S  WEIGHTS AND MEASURES Metric System CURRENCY EQUIVALENTS Appraisal Year Ave. (1973) US$1.00 - (Rupees) Rs 6.40 Intervening Year Ave. US$1.00 - Rs 14.72 Completion Year Ave. (1987) US$1.00 - Rs 30.62 ABBREVIATIONS AND ACRONYMS AMP = Accelerated Mahaweli Program CECB = Central Engineering and Consultancy Bureau CMD = Bank's Commodity Markets Division EEC = European Economic Community FAOICP - Food and Agriculture Organization/World Bank Cooperative Program Gibb = Sir Alexander Gibb and Partners (UK consulting firm) GOSL - Government of Sri Lanka HTS = Hunting Technical Services, Ltd.1/ ICB - International Competitive Bidding IDA = International Development Association ISS Implementation Strategy Study JVMTD = Joint Venture Mahaweli Transbasin Diversion2/ LCB = Local Competitive Bidding MASL = Mahaweli Authority of Sri Lanka (previously Mahaweli Dev. Auth.) MDB = Mahaweli Developnent Board MEA = Mahaweli Economic Agency MECA = Mahaweli Engineering and Construction Agency MGDP = Mahaweli Ganga Development Program NCRB = North Central River Basin NEDECO = Netherlands Financed Consulting Firm NWDZ = Aorthwest Dry Zone nlED = Operations Evaluation Department OFC = Other Field Crops O&M = Operation and Maintenance PCR = Project Completion Report PPAM 5 Project Performance Audit Memorandum PPAR 5 Project Performance Audit Report RPM 5 Resident Project Manager RVDB = River Valley Development Board SEDZ Southeast Dry Zone WMP = Water Management Panel WMS 5 Water Management Secretariat FISCAL YLAR January 1 - December 31 )/ United Kingdom consulting firm, in association with Sir Alexander Gibb and Partners and Huszar Brammah and Associates. 2/ Consortium comprising Electrowatt Engineering Services Ltd., Zurich, Salzgitter Consult GmbH, Salzgitter, and Agrar-und Hydrotechnik, Essen.  FOR OMCAL USE ONLY THM WORLD BANK Washington. D.C 20433 1 U.S Oibe of OlDt.cuCwatI Opswatm lvaht.n June 30, 1988 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report Sri Lanka Mahaweli Ganga Development II Project and Mahaweli Ganga Technical Assistance Project (Credits 701- and 979-CE) Attached, for information, is a copy of a report entitled 'Project Performance Audit Report: Sri Lanka - Mahaweli Ganga Development II Project and Mahaweli Ganga Technical Assistance Project (Credits 701- and 979-CE)" prepared by the Operations Evaluation Department. The PCRs were prepared by the Asia Regional Office on the basis of PCRs prepared by the Borrower. Yves Rovani by Graham Donaldson This document has a restncted distribution and may be used by rocipients only in the performance of their official duties. Its contents may not otherwise be discksed without World Dank authorization.  FOR OMICUL USE ONLY PROJECT PERFORKANCE AUDIT REPORT SRI LANKA MAaAWELI GANGA DEVELOPMENT PROJECT II (CREDIT 701-CE) MNAHWELI GANGA TECHNICAL ASSISTANCE PROJECT (CREDIT 979-CE) TABLE OF CONTENTS Page No. Preface ~~.... *.* ** *......*.................... i Basic Datsahet $ Evaluation Suiay vom . .ooo .. Proy . . .. v$$ PROJECT PERFORMANCE AUDIT MEMORANDUM I. BACKGROUND AND SETTING ...... ... ..................... ..... 1 Ao Introduction ooo... 1 B. Economic Development ...*... ...........00 9-............. 2 C. Setting and Context .. ***......*** * *0.....*.*...0.0000 3 II. IRRIGATION AND MAHAWELI GANGA DEVELOPMENT ... o....*........ 4 A. Irrigation Sector ................o..oo.*.*..........*.*.... 4 B. The Mahsaveli Ganga Development Program..*#........... 6 III. PROJECT IMPLEMENTATION - MAHAWELI GANGA DEVELOPMENT II ... 8 A. Project Formulation ........... ...... 8 B. Project Implementation ....... ... . ............ *. 9 C. Project Administration ............................. .. 10 D. Project Impacto........................................ 10 IV. PROJECT IMPLEMENTATION - MAHAWELI TA PROJECT ..........o... 11 A. Project Formulation .... 11 B. Project Implementation and Impact ................... 12 IV. MAIN FINDINGS AND ISSUES... .. ........................... .. 13 A. The Second Dimension of Sustainability ............... 14 B. Bank Supervision ................. ..... .. ........ .. 15 C. Civil Works Contractors .. . ........... ..... ...... . 16 PPAM Annexes 1. Sri Lanka Lending Program ........................... . 19 2. OED Project Evaluation ............................ .. 22 3. Borrower Comments.......................... 23 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 authofiation.  PROJECT COMPLETION REPORT (MAHAWELI GANGA DEVELOPMENT II) I. lackgroundo- .................................... 33 II. Project Processing ........... *.*..* ....... .. 34 III. Implementation ............................................... ........ 41 IV. Organization and Manrgement ................................. 53 V. Project Impact ......................................................... 57 VI. Economic Analysis..*.................................. ..... 62 VII. Suimmary and Conclusions. .................................... . 63 PCR Tables 1. Cultivation Loans in H4 Area ....................... 65 2. Project Cost ..............................a . .... .... 66 3. Crop Areas in H4 and H5 ...5....................... 67 4. Alternative Estimates of Areas, Yields and Production.... of Paddy in all New Areas of System H .............. to 68 5. Estimated Crop Areas and Crop Production in H4 and H5 Areas ................. .. ..... . 69 6. SAR and Revised Estimates on Crop Areas and Crop Production in H4 and H5 Areas .................... 70 7. Crop Inputs per Acre ........... ............. . 71 8. Financial and Economic Prices ...................... 72 9. Budget for Farm of 2.5 aC ......................................... . 73 10. Economic Cash Flow, H4 and H5 Areas ..... .................. 74 PROJECT COMPLETION REPORT (MAHAWELI GANGA TECHNICAL ASSISTANCE) I. Project Data ......... *.......... . .......... .................to... 77 II. Project Background ........... . . *. ... ............ . .......... . 71 III. Project Costs and Financing t..........................to. 79 IV. Project Undertakings and Results .... ..................... 80 Transbasin Diversion Study ........................ 80 Minipe Right Bank Canal Review ........................... 81 Design of Infrastructure in System C ..................... 81 Support for Other Studies/Investigations ................. 82 Training .............. ....... .... ......... ................. 82 Performance of Consultants ............... ................ 83 V. Legal Undertakings and Compliance.o.......... ............ 83 VI. IDA Supervision o*......... ....... 00..... ;-.... 84 VIIo Conclusions to.................... ....0000 .............. ............ 84 Maps IBRD 14076R Agro-Ecological Regions IBRD 13477R2 Mahaweli Ganga Scheme IBRD 11908R1 Mahaweli Ganga Development Project II  PROJECT PERFORMANCE AUDIT REPORT SRI LANKA .AHAWELI GANGA DEVELOPMENT PROJECT II (CREDIT 701-CE) MAHAWELI GANGA TECHNICAL ASSISTANCE PROJECT (CREDIT 979-CE) PREFACE This is a performance audit of two Mahaweli development projects in Sri Lanka. Credit 701-CE in the amount of US$19.0 million was approved in April 1977 for the Mahaweli Ganga Development Project II (Mahaweli II). The last disbursement was made on February 6, 1986 followiug extensions of the closing date from June 30, 1983 to June 30, 1985. A total of US$7.69 million was cancelled on February 6, 1986. Credit 979-CE in the amount of US$3.0 million was approved in January 1980 for the Mahaweli Technical Assistance Project (Mahaweli TA Project). It was originally expected to be comple':ed in 1982 and closed in September 1982 but the closing date was extended three times, to September 1985, and the credit account was left open thereafter until June 1986. Extensions were granted to finance additional urgent studies requested by Government. The final disbursement was made on May 14, 1986 and the remainder of the IDA credit for a total of US$1.1 million cancelled. Both Mahaweli projects benefitted from cofinancing. Mahaweli II was cofinanced by Canada (an interest-free loan of US$6.0 million equivalent), the Netherlands (US$5.0 million equivalent), UK (a grant of US$7.2 million equivalent) and USA (US$5.0 million). The contributions by Canada, UK and USA were for tied procurement of equipment and for construction of rural works under parallel financing arrangements. The Netherlands contribution was untied in joint financing with IDA. Subsequently, a further contribution of 2 million Units of Account (about US$2.4 million equivalent at the time) was granted by the Commission of the European Communities to help the Government of Sri Lanka (GOSL) defray the cost of local expenditures. The Mahaweli TA Project was cofinanced with an EEC Special Action Credit of US$2.0 million equivalent. Like the IDA Credit it was extended three times. About US$280,000 equivalent of the EEC Special Action Credit was ultimately cancelled. The audit report consists of an audit memorandum prepared by the Operations Evaluation Department (OED), a Project Completion Report (PCR) on the Mahaweli II Project dated March 1987 prepared by the former Irriga- tion I Division in the South Asia Projects Department on the basis of a draft project completion report dated June 1984, which was prepared by the staff of the Planning and Monitoring Unit of the Mahaweli Authority of Sri Lanka and a PCR on the Mahaweli TA Project dated June 30, 1986 prepared by the same division also on the basis of a draft PCR prepared by the Borrower. The audit memorandum is based on a review of the Staff Appraisal  - ii - Report for the Mahaweli II Project (Report No. 1487a-CE dated March 31, 1977); the President's Reports (Mahaveli II: No. P-2021-CE dated April 7, 1977; Mahaweli TA: No. P-2686-CE dated January 10, 1980); the Credit Agreements (for Mahaweli II dated June 27, 1977; and for Mahaweli TA dated April 16, 1980); and the PCRs. Correspondence with the Borrower and internal Bank memoranda on project issues as contained in relevant Bank files have also been consulted, and local representatives of cofinancers and Bank staff associated with the projects have been interviewed. An OED mission visited Sri Lanka in July'August 1987. Discussions were held with project and Government officials and project beneficiaries who were interviewed during visits to project sites. The information obtained during the mission was used to test the validity of the conclusions of the PCRs. The audit concurs with the principal pcints made in the two PCRs. Points discussed in the audit memorandum have been selected because of their relevance and importance for these and other Bank-supported projects. The draft report was sent to the Borrower and cofinanciers for their comments on May 24, 1988. Corments received have been attached to the PPAK as an annex. The valuable assistance provided during the preparation of the report by the Government of Sri Lanka and the executing agencies and the hospitality received during the field trips is gratefully acknowledged.  - LU. - PROJECT PERFORMANCE AUDIT REPORT SRI LAKA VAHAWELI GANGA DEVELOPMENT PROJECT It (CREDIT 7M1-CE) BASIC DATA SHEET KEY PROJECT DATA Actual or Acul as X Appraosal Current Appraisal Estimate Estimt Estmate Total Project Cost (US mlillton) e9.e 84.9 195 Credt Amount (US$ milloin) 109 11.08 a Co-financing (USS million) Canad 6.0 0.0 leg The Netherlands 5. 4.7 95 United Kingdom 7.2 6.4 90 United States 6.0 4.2 e6 European Economic Community - 2.2 1N L to Physical Componenta Completed 96/92 06/88 Proportlon then Comploted (X) 100 100 Economic Rate of Return (M) 21 -1 Number of Direct Benfticieloro 76.9 790,96 STAFF IWrTS (Staff Weeks) FY72 FY7 FY77 FY7Y FY79 FY69 FY01 FY82 FY68 FY6 FYj FY86 FY87 Pr.apprateal 62.8 25.0 1.1 Appraisal 86.6 70.8 Negotiation 6.7 Supervision .2 0.4 88.6 20.9 24.6 21.0 15.8 12.8 18.9 15.6 5.6 .2 Other .5 .6 1.1 .8 Total .2 62.8 25.0 70.2 00.1 U8.6 20.0 24.8 22.1 16.1 12.8 18.6 15.6 5.8 .2 CUULATIVE DISSURSEMENTS FY76 FY79 fY8 FY1 FY82 FYS0 FY84 FY8 FY8 Appraisal Estimte (US$ million) 1.6 6.4 12.6 16.6 19.0 Actual (USS million) - 9.1 1.6 4.9 6.0 0.2 9.6 10.4 11.8 Actual a ot Approial 0 2 12 29 82 48 86 66 6f Date of Final Disbursemnt February 6, 1906  - 1V - PROJECT DATES On ginal Plan Actual Identification R port 09/72 Negotiations 08/77 Board Approval 04/21/7 Crilt Slgning 06/27/7 Credit Eft.ti veness 10/89/77 12/20177 Credit Closing 0J80//oS 63/89/35 MISSION DATA Cofinancelor Participating In Su4ervlsion Misslons Specialitr- Perform- Types Date No. of tlona Nether- once of Mission (mo.1yr.) Persona Reporsented Canada lands UK USA Rating Trend Probloms /lb Lc Id Project Prep. 11-12/78 1 Eng Project Prep. Wr.4 1 Eng Project Prop. 11-12/74 4 Eng, Ag, Ec(2) Project Prep. 98/76 6 Eng, Ag(4), Ec Firet Appr. 04-65/75 8 Eng(8), Ag(4), Ec Post Appr. 61/76 2 Ea, Ag Second Appr. 46-0/70 4 Ec(2), Eng, Ag Supervision 1 68.-4/77 2 Eng, Ag No No yes No 1 Supervision 2 is/m 2 Eng, 0 Yes Yee Yes Yes 1 2 P Supervision 8 9840/76 2 Engs Fin Yes Yes Yea Yes 8 2 U,T, P Super-vision 4 94-45/79 1 Eng Yes Yen Yes Yes 2 1 M,T Supervlsion C 11/79 2 Eng, Ag Yes Yes Yes Yes 2 1 M,T Supervision 6 69/60 1 Eng Yes No Yes Yes 2 2 M,F Supervision 7 8-04/61 2 Eng, Ea Yes No Yes Yes 2 2 ,L Supervilsion 8 18/32 2 Eng, Ee Yes You Yes Yes 2 1 L Supervision 9 02/81 2 Eng, Ag Ye Yes Yes Yoe 2 2 H Superviston 10 12/O8 8 Ag, Ec, Eng No Yes Yes Yes 2 2 H Supervision 11 61/85 2 Ag, Ee No Yes Ye No 2 2 U OTHER PROJECT DATA Borrower Gove rnmn t of Sri Lanka (GOSL) Executing Ageney Mahael1 Developmont Board (MDO) Precedins Pr--lect Follow-on Projeets Name MahnreI Mahaw lI TA Mahae IlI II Mhaw 1I IV Credit Number Cr. 174/Ln. 68 Cr. 979 Cr. 1166 4r. 1494/Ln. 2487 Amount (US8 millions) 29.0 8.0 90.0 8O.0/12.1 Appreval Date 61/89/70 91/15/79 66/28/81 96/67/34 !/ Eng a Irrigation Engineer; Ag = Agronomist; Ec * Economist; Fin a Flnancial Analyst; 0 a Other. k/ 1 a Problem free or minor problems; 2 a Moderate problema; = Major problems. cj 1 a Improving; 2 a Stationary; 8 a Deteriorating. y/ P u. Political; M a Managerial; T a Technical; F a Flnancial.  PROJECT PERFORMANCE AUDIT REPORT SRI LNKA MAAUnELU OANGA TEC04ICAL ASSISTANCE PROJECT (IDA CREDIT 9T9-CE AND EEC SPECIAL ACTION CREDIT 40-CE) BASIC DATA SHEET KEY PROJECT DATA Actual as Actual or Porcentage Appraisal Current Appraisal Etmato Estimate Estimte Total Project Coot (USS1 million) 0.O 4.4 78 IDA Credit Amount (USS million) 8.0 1.66 62 EEC Credit Amount (USS million) 2.0 1.72 86 STAF- INPUTS (Statt Weeks) FY80 FY61 FY62 FY80 FY84 FY8S FY86 FY8? TOTAL Preappralual 8.2 8.2 Appraisal 8.0 8.e NK otiatlon .6 .5 Supervision 6.1 5.6 4.0 .6 8.2 2.6 5.1 .2 26.7 Other .8 2.8 8.1 TOTAL 21.9 6.4 6.8 .6 8.2 2.8 5.1 .2 48.6 CUMULATIVE DIS8URSEMENTS IDA Credit (USS million) fY80 FY1 FY62 FY88 FY84 FY85 Ff86 Appraisal Etimate (USS million) 0.8 2.0 8.6 - - - - Actual (USS mlillon) a 0.29 1.87 1.61 1.61 1.66 1.86 Actual as S of Estmte a 15 48 54 64 65 02 EEC Credit (US1 million) Appraisal Estimat (US8 million) 6.5 1.5 2.6 - - - Actual (USS million) 0 1.04 1.11 1.11 1.11 1.11 1.72 Actual as X of Estimate 69 66 56 60 5s 86  - vi - PROJECT DATES Original Plan Revision Actual T.A. Credet Proposed te Govt. - 06/22/70 Requet free Government - - _0/10/79 Negobletlone 11/79 12/20/79 12/27/79 Board Approvel 1/29/89 91/15/79 19/29/fi Credit Agromnt (IDA A EEC) -- -- 14/16/89 Credit Effectiveneas (IDA A EEC) - -- O7/67,f Closing IDA Credit 09/80/92 -- 0/9/U65/ EEC Credit 09/80/62 09/0/8/52/ MISSION DATA Month/ No. of Specialls. Performance Types of tsar Porronu Rer nte d n . Rat/n_ _ _ / Trend ___Problo / Preperatlon/Followup 11/79 a Eng,Ag,Chlef - - - Supervision 09/80 1 Eng 1 2 - Supervislon II 12/80 2 Eng, Eng 1 2 - Supervision III U4/81 1 Eng 1 2 - Supervision IV Ot/81 1 Eng 1 2 - Supervision V 94/82 1 Eng 1 2 - Superviston VI 01/85 1 Eng 1 2 U OTHER PROJECT DATA Borrow r ThO Deoeratic Socialist Republic of Sri Lanka (OOSL' Executing Aency Mshoseli Authority of Sri Lanka (MASL) Preceding Prolects Follow-on ProJects Name Mahaowol Nbhawe lI I Uaheweli III Mahawell IV Loan/Credit Number Cr. 174/Ln. 65 C. 791 Cr. 1106 Cr. 1494/Ln. 2487 Amount (US$ million) 290. 19.0 90.0 80.0/12.1 Approval Dote 01/80/79 94/21/77 06/28/81 96/07/84 l/ Although formal closing dat was 09/80/86, credit account was left open to 06/20/U8. fr/ Eng a Irrigetion Engineer, Ag a Agriculturist, Chlot = Projects Division Chief, ~/ 1- Problem free or minor problms. i/ 2 a Stationary. 2/ U a Managerial.  - vii - PROJECT PERFORMANCE AUDIT REPORT SRI LANKA MARAWELI GANGA DEVELOPMENT PROJECT II (CREDIT 701-CE) MAHAWELI GANGA TECHNICAL ASSISTANCE PROJECT (CREDIT 979-CE) EVALUATION SUMMARY Introduction 1. Sri Lanka is a tropical island with distinctly livided wet and dry zones (see map). The vet zone encompasses the southwest quadrant of the island, or about 302 of the land area. At the time of project appraisal, the wet zone supported more than three-quarters of the nation's population and accounted for about 822 of the cultivated land. The dry zone was relatively undeveloped although it wso there that the an.ient Singhalese kings developed a highly advanced and ingenious irrigation system. 2. During most of the project implementation period, Sri Lanka has experienced an unprecedented economic growth. Per capita incomes were much higher and unemployment much lower than a decade ago (before 1977). The expansion of economic activity also bonefitted the population below the poverty line. However, the Government is now confronted with the most difficult economic (slowing growth, rising unemployment, precarious fiscal and balance of payment situation) and political problems since it took office in 1977. 3. About 751 of Sri Lanka's population resides in rural areas, with some 45Z of the total active labor force directly employed in agriculture. While agriculture has declined as a percent of GDP it continues to make a significant contribution to both export earnings and Government revenue. 4. Bank lending for agriculture started in 1968 with three Irrigation projects; it accelerated in the second half of the 1970s, and also became more diversified. Apart from more Irrigation projects, It included treecrop rehabilitation, rural development, dairy, forestry and agricultural support services. As of December 31, 1987, lending for agriculture amounted to about US$550 million, or Jus% under half of total lending to Sri Lanka. 5. The largest share of agricultural lending was for Irrigation, US$222 million or 441 of total lending for agriculture; about 751 of it in support of the Mahaweli program. Five Mahaweli projicts were financed by the Bank. The Bank's earliest irrigation projects primarily provided the physical infrastructure necessary to distribute irrigation water, or to protect and reclaim irrigated areas. Increased food production, In support of the Government's self-sufficiency policies, was the main project objective. 6. The five loanslcredits in support of the Mahaweli Gangs program. were for discrete elements of the program. Except for the first project,  - viii - which included a power station, the Bank has only supported the downstream development of the Mahaweli program (irrigation and related supporting services, as well as technical assistance). 7. The Bank's first financed Mahaweli project (1970) is completed. Despite implementation delays of about two years, the project was regarded as economically and socially successful. The main reasons for success were considered to be sound project design, continued support from government agencies, and the ability of farmers to adapt to the rigorous time frame imposed by double cropping. However, certain problems remained, such as cost recovery, monitoring and evaluation, effective water management, and illegal occupation of land. Although these problems were addressed in follow-up projects, they have remained critical issues until today. Obiectives 8. The main objectives of the overall Mahaweli program are: (i) the reduction of unemployment by creating short-term and long-term job opportunities; (ii) the increase of agricultural production, particularly rice, to achieve food self-sufficiency; and (iii) the alleviation of power shortages. The Mahaweli I: Project would complete the irrigation and development in Mahaweli system H area (see maps), and provide productive support and social services throughout the system area. In this context the project would provide employment and substantial incremental production of rice, as well as other output (pulses, chilies, Vegetables and oilseeds). 9. The Mahaweli TA Project aimed to support the Government's efforts to accelerate implementation of the Mahaweli development program by assisting with studies for conveying and utilizing surplus Mahaweli Ganga water to develop land in any of three alternative areas in order to select the best plan for a future transbasin diversion project. Technical assistance was also to be provided for detailed engineering and design for downstream development of head works already committed to be constructed. Finally, support for unspecified studies and designs, to be identified by the Mahaweli Development Authority, were also;"included in the Mahaweli TA project. Implementation Experience 10. Work on the Mahaweli II Project was started by the Government without prior consultation with the Bank. The Government decided at the end of 1977 to accelerate the project ahead of schedule but this resulted in poor quality of works, excessive water conveyance losses and operation and maintenance costs higher than anticipated. Project implementation, however, improved in 1979/80 and progress as well as quality of works were satisfactory. Implementation slowed down again in 1981, mainly due to budget restrictions. In the end, instead of completing the project ahead of schedule the Credit had to be extended twice. 11. The Government's strong commitment to the project was reflected in adequate levels of funds and staff allocated to the project during most years. Management and organization of the Mahaweli Ganga development scheme were excellent.  - ix - 12. The Mahaweli TA Project was successful and all studies and activities identified at time of approval were executed. Consultants were selected expeditiously for the Transbasin study and fielded just two months after Credit effectiveness. Equally fast was the contracting of consultants for the Minipe Right Bank Review which led to the successful completion of the largest irrigation canal in Sri Linka. Due to financial and other considerations, funds for studies not identified during project preparation remained uncommitted despite Credit extensions. In retrospect, the envisaged implementation period of two years was unrealistic. Results 13. The Mahaweli II Project should be considered successful in regard to attaining major objectives such as land development, settlement, and crop production but there is scope for further development. The project created employment for a large number of unemployed and inched Sri Lanka close to becoming self-sufficient in food. However, the project ERR has been recalculated at -11 mainly due to much lower actual and forecast prices for rice than the price forecasts at the time of appraisal. 14. The Mahaweli TA Project made a valuable contribution to the accelerated implementation of the Mahaweli program (AMP). It also provided valuable on-the-job training of counterpart staff by the consultants during their work assignments. Sustainability 15. The Mahaweli program has been considered a cornerstone of development efforts by all Sri Lanka Governments. The projects under review have been succeeded by two more Bank-assisted projects. From the side of Government there is full commitment to the program and the development authorities have proven their management capabilities. Farmer participation is also assured in light of lacking employmentifarming opportunities elsewhere in the country. The real question is how this commitment can be transformed into realistic and lasting sustainability. Findinzs and Lessons 16. The principal findings, which show relevance for other project situations as well, that emerge from the evaluation of these two projects are: - As shown by the problems (negative impact on costs, poor quality) caused by the accelerated startup of the Mahaweli II Project and the inability of Government to select the unidentified studies in the Mahaweli TA Project, the Mahaweli Ganga projects once again demonstrate the need for proper planning prior to starting work (PPAM, paras. 28 and 41, Mahaweli II PCR, paras. 3.02 and 7.02-7.03, and Mahaweli TA PCR, paras. 4.06-4.07 and 7.01). - Reliance on small contractors is unjustified without a proper assessment of overall capacity of the contras ting industry and competing demands for its services at tbe time of project  appraisal. Coordination between various Baik activities is also warranted in this context (EPAM, paras. 55-61); Technical supervision was of high quality and staff continuity had a positive impact. However, there are limits to what technical supervision can be expected to solve in relation to "macro" issues such as cost recovery and water charges (PPAMI, paras. 51-54); Two years for a technical assistance project is unrealistic, especially if it includes components that have to be defined during implementation (PPAM, para. 42 and Mahaweli TA PCR, para. 7.02); Project sustainability depends on sustaining the interest of participating farmers. This requires better consultation and dialogues with farmers and training of farm leaders (PPM, paras. 45-48); and While planning at the "micro" level (day-to-day management and execution) was well taken care of, "macro" planning received less attention. However, to ensure sustainability would require more focus on long-term changes and the development of alternative scenarios to meet such changes (PPM, paras. 49-50).  PROJECT PERFOR1MANCE AUDIT MEMORANDUM SRI LANXiA MAHAWELI GANGA DEVELOPMENT PROJECT II (CREDIT 701-CE) MAHAWELI GANGA TECHNICAL ASSISTANCE PROJECT (CREDIT 979-CE) I. BACKGROUND AND SETTING A. Introduction 1. Bank lending to Sri Lanka (as of December 31, 1987) for 54 proj- ects amounted to about US$1.2 billion (see also Annex 1). The lending program is largely concentrated on agriculture and power with the remainder scattered over other sectors as Table 1 below showst Table 1: BANK LENDING TO SRI LANKA No. of Loan/Credit Percentage Subsector Proiects Amounts of Total (US$ million) Agriculture 19 364.9 31 Mahaweli 5 184.1 16 Power 9 218.7 19 Transportation 4 103.4 9 DFCs 4 24.5 2 Industry 5 111.0 10 Training & Industrial 3 41.5 3 Development Water Supply & Sanitation 3 76.2 6 Other a/ 2 45.0 4 54 1.169.3 100 8/ One telecommunications project and a program credit. 2. Bank lending for agriculture started in 1968. Three irrigation projects were approved in 1968-70, which fitted the Goverament's objective of increasi'g food pr._uction, but they were designed without a clearly defined Bank scracegy for the agricultural sector as a whole. Sector know- ledge was subsequently updated through an agricultural policy and programs review mission which visited the country in 1974 to identify lending oppor- tunities. Lending for agriculture accelerated in the second 'nalf of the 1970s, and also became more diversified. Apart from more irrigation proj- ects, it included treecrop rehabilitation, rural development, dairy, for- estry and agricultural support services. As of December 31, 1987, gross lending for agriculture amounted to US$550 million, or just under half of total lending to Sri Lanka.  - 2 - 3. As expressed in various documents, the main element of the Bank's strategy since 1974 has been to help alleviate the country's foreign exchange constraints by supporting measures to increase agricultural output and exports. Furthermore, the Bank saw its agricultural program as a vehi- cle to obtain policy changes in the agricultural sector, including cost recovery in irrigation. The Bank has also supported the Government's rural development program to raise productivity, employment and living standards of the rural population. More recently, the Bank has attached high prior- ity to the maintenance and rehabilitation of the existing capital stock in all productive activities. 4. To date OED has evaluated 15 projects in Sri Lanka; evaluations of another 7 are in progress (details in Annex 2). In addition, OED has undertaken an impact evaluatlon of the Lift Irrigation Projectl/ and reviewed the World Bank/Sri Lanka relationship.2/ The Mahaweli Ganga Devel- opment Project II (hereafter referred to as Mahaweli II) was the Bank's sixteenth lending operation in Sri Lanka. The Mahaweli Ganga Technical Assistance Project (hereafter referred to as Mahaweli TA) was the third lending operation in support of the Mahaweli program (see PPAM, para. 20). B. Economic Development 5. Policies pursued by various governments have changed over time, and as a result so did socio-economic development. Four periods can be distinguished: 1956-1965; 1965-1970; 1970-1977; 1977 to present. All governments in Sri Lanka have been committed to welfare, equality and the alleviation of poverty and have seen economic growth as an important objec- tive. However, strategies to achieve growth have differed. The periods 1956-65 and 1970-77 were characterized by a less open economy, emphasis on import substitution, a diminished role for foreign aid and advice, nation- alization, public enterprises and an expansion of government control. The periods 1951-55, 1965-70, 1977 to the present relied more on private enter- prise and price incentives, opening up the economy, stressing export devel- opment, and attracting foreign resources--aid, foreign private investment and foreign loans--with more emphasis on agriculture.3/ 6. Since 1977, Sri Lanka has experienced an unprecedented economic growth of 5.6Z per year. Per capita GDP is now 402 higher, in real terms, and the rate of unemployment is much lower than a decade ago. Albeit at a lower rate than the average, the population below the poverty line has also benefitted from the expansion in economic activity. There are, however, 11 Impact Evaluation Report, Sri Lanka Lift Irrigation Project (Credit 121-CE), OED Report No. 5634, dated May 8, 1985. 2/ The World Bank and Sri Lanka, A Review of A Relationship, OED Report No. 6074, dated February 24, 1986, later published as a World Bank Operations Evaluation Study. 3/ These changes in economic development in Sri Lanka have been reviewed in much more detail in OED's Review of a Relationship, OD cit.  - 3 - clear indications of a reversal in economic performance and the Government is now confronted with the most difficult economic problems it has had to face since it took office in 1977. Growth is slowing, unemployment is rising, the fiscal situation is precarious, and a balance of payments crisis is not unlikely. The reasons for the deterioration of the economy are complex. They can be related to both external and internal adverse developments. 7. Prices of Sri Lanka's traditional exports, tea, rubber and coco- nuts, have been at their lowest level since World War II. The deteriora- tion in Sri Lanka's terms of trade (about 30% since 1978), thk 1uotas set by developed countries on Sri Lanka's exports of garments, the more recent stagnation in workers' remittances from the Middle East as a result of the decline in the price of oil, the reduced flow of tourists, and higher for- eign exchange expenditures related to national security have all contrib- uted to further weakening of the balance of payments, and the Government's capacity to respond vigorously to the deteriorating economic environment.4/ 8. While these developments have serious implications for the sustainability of the projects under review, they were not prevalent during most of the periods these projects were executed. Neither is it likely that they could have been predicted at that time. C. Setting and Context 9. Sri Lanka has an estimated population of about 16.1 million, 80X of which is rural. Of the total land area of 16.2 million acres (about 25,300 sq. mi.), about 5 million acres are used for agriculture. The climate is characterized by nearly constant temperatures and large varia- tions in rainfall. The island can be divided into two distinct precipita- tion zones (see also map 14076R attached to this report)s the west zone (average annual rainfall over 75 inches), situated in the southwest quad- rant and covering about 301 of the land area, and the dry zone (average anrual rainfall ranging from 35 to 75 inches) covering the remainder of the island. The wet zone supports more than three-quarters of the total popu- lation and accounts for about 702 of the cultivated land. 10. Rice is Sri Lanka's most important staple and export crops such as tea, rubber and coconuts form the backbone of the country's economy. How- ever, while export crops were largely and satisfactorily produced on estates, production of fooderops was left to small farmers and was far from satisfying the island's needs, necessitating substantial imports of basic foods. For example, in 1968, 451 of the total value of imports was for food: of this expenditure, rice accounted for about 351 and sugar about 101. Although there was considerable room for increasing agricultural yields, the main prospect for attaining self-sufficiency in food supply existed in the expansion and intensification of cultivated areas. Since 41 Details on the latest economic development can be found in the Bank's report: Sri Lanka--Issues in Macro-Economic and Industrial Development Policy, Report No. 6701-CE, dated April 20, 1987.  there was a comparative scarcity of land for further cultivation in the vet zone, the expansion needed had to take place in the dry zone. Sri Lanka also needed more electric energy. In 1968, the effective generating capa- city of the national system was 261 MW, including 191 MW of hydro capacity. The firm capacity was about 220 MW. Since the projected demand for 1973 was already 296 MW, there was a need to increase the nation's generating capacity quickly. The development of the Mahaweli Ganga, the island's largest and almost unexploited water resource which cuts through the east- ern portion of the dry zone, therefore, was seen by the Government as the cost logical source of irrigation water for arable lands in the dry zcne, ginerating more electric power to meet increasing demand, and creating em2loyment opportunities. 11. Because of its difficult physical environment and the prevalence of malaria, the dry zone has developed slowly in modern times. Yet it was here, rather than in the more inviting wet zone, that the ancient Singha- lese kings developed a highly advanced civilization, which flourished by means of ingenious and extensive irrigation modes (see also PPAM, para. 18). But various wars and neglect took their toll on these irrigation systems. However, it was again in this zone that the major Mahaweli Ganga Development Program (MGDP) was executed as Sri Lanka's bold attempt to iMrove its eco*omy and way of life. II. IRRIGATION AND MAHAWELI DEVELOPMENT A. Irrigation Sector 12. The Bank's involvement in irrigation in Sri Lanka has been sub- stantial. Not only has irrigation been the object of Bank funding from the earliest Bank projects in the agricultural sector, but the Bank has played an important role in shaping the MGDP (see PPAM, paras. 16-22). It has been able to mobilize substantial cofinancing, and cooperation with other donors in project preparation, appraisal and supervision has been good. Although project implementation generally has been satisfactory, serious challenges remain with respect to cost recovery, water management, and the rehabrlitation of existing irrigation infrastructure. Also, with self- sufficiency in rice approaching, crop diversification in irrigated areas presents a formidable challenge. 13. The largest share of agricultural lending was for irrigation, US$241 million or 44 percent of total lending for agriculture; about 75 percent of it in support of the Mahaweli program, mainly downstream devel- opment. Five Mahaweli projects were financed by the Bank (see PPAM, para. 20). Another five projects covered other types of irrigation in Sri Lanka.  - 5 - Table 2: BANK-FINANCED IRRIGATION PROJECTS La Approval Credit Proiect Name Year (FY) Number Amount (US$ mil) Lift Irrigation 1968 121 2.00 Drainage and Land Reclamation 1970 168 2.64 .ank Irrig. Modernization 1977 666 5.00 Village Irrig. Rehabilitation 1981 1,160 30.00 Major Irrig. Rehabilitation 1985 1,537 17.00 56.64 la Excluding Mahaweli projects. 14. The Bank's earliest irrigation projects primarily provided the physical infrastructure necessary to distribute irrigation wattr, or to protect and reclaim irrigated areas. Increased food production in support of the Government's self-sufficiency policies were the main project objec- tives. The Lift Irrigation and the Drainage and Land Reclamation projects encountered substantial implementation delays, owing to civil unrest, the change in Government, and procurement problems. Overall, the Lift Irriga- tion scheme was a failure,51 while the Drainage and Land Reclamation Project6l was judged to have fared much better. 15. The later irrigation projects focussed on the rehabilitation of existing irrigation infrastructure, and addressed water management issues and irrigation practices in order to achieve a more efficient use of water. The Tank Irrigation Modernization Project: aimed at improving water distri- bution systems, infrastructure and servtices in five major tank schemes in the northern part of Sri Lanka. Implementation took three years longer than expected owing to management and funding problems, but at completion7/ project performance was judged to be satisfactory. The Village Irrigatlon Rehabilitation Project supports the rehabilitation of 1200 village irriga- tion schemes and modernization of about 500 other small schemes. Progress has fallen behind appraisal targets partly due to security problem., but the quality of works is generally satisfactory. Under the Major Irrigation Rehabilitation Project work remains suspended at three subprojects, but on the four other subprojects good progress is being achieved. However, col- lection of O&M charges from beneficiaries continues to be poor. Suitable 5/ Project Performance Audit Report, OED Report No. 2801, dated December 28, 1979, and Impact Evaluation Report, OED Report No. 5634, dated May 6, 1985. 6/ Project Performance Audit Report, OED Report No. 2798, dated December 28, 1978. 7/ Project Completion Repoit, OED Report No. 5986, dated December 23, 1985.  remedial action will be planned after study proposals on a feasible col- lection system have been evaluated. B. The Mahaweli Ganaa Development ProRrsm 16. The MGDP has been the centerpiece of Sri Lanka's development efforts for the last 25 years. Formulation of the program was initiated during 1958-61 and implementation of the Master Plan commenced in 1970 (Mahaweli II PCR, paras. 1.01-1.03). Late in 1977, the new government decided to accelerate the implementation of the program. At that time work on the first phase of the MGDP was in progress, with part of the funds provided by the Bank. The projects of the Accelerated Mahaweli Program (AMP) included five major reservoirs and hydroplants with a total potential power capacity of 500-600 MW, and downstream development of about 112,000 ha. After consultation with the Bank, in late 1977, GOSL agreed to undertake an implementation strategy study (funded by the Netherlands). While that study was going on, GOSL approached bilateral donors for financing. They responded favorably, and funding for the construction of four dams and reservoirs was obtained (see also Mahaweli TA PCR, paras. 2.01-2.07). 17. As noted above (PPAM, para. 10), the main objectives of the Maha- weli program related to three of the major problems that were facing Sri Lankas (a) reducing unemployment by creating short- and long-term job opportunities; (b) increasing agricultural production, particularly rice, to achieve food self-sufficiency; and (c) alleviating power shortages. 18. The acceleration of this large investment program has also to be looked at in its historical and political context. Politically, the deci- sion to accelerate was to a certain extent to provide compensation (through increased employment and production) for the costs of substantial economic reforms. Historically, the program pays tribute to the ancient civiliza- tion of Sri Lanka with its magnificent irrigation works (see also PPAM, para. 11). Every Sri Lankan knows of the philosophy of the ancient mon- archs, most clearly expressed by the 12th century King Parakramabahu I who said, 'Let not one drop of water reach the sea without first serving man'. Large-scale irrigation works were built as early as the 1st century A.D. High standards of irrigation technology enabled the monarchs to extend waterworks throughout the dry zone by the 6th century A.D. So thorough was their work that many of those systems are still in use today. 19. From its inception, the Bank has been closely involved with MGDP. Over the last 15 years, there was substantial interaction at the technical level, mainly through supervision of Bank projects. A close relationship developed between the Bank and the Mahaweli Authority, and at the technical level the Bank's advice has always carried substantial weight. In addition to supervision, the Bank played a role in designing the overall development program, and in coordinating foreign assistance. A full assessment of MGDP and the Bank's involvement in it can be found in OED's special study: Sri Lanka and the World Bank: A Review of a Relationship, cited earlier.  20. The Bank has made five loans/credits (see Table 3) in aupport of the MGDP for a total of US$184 million, or 16 percenc of total Bank Group lending to Sri Lanka. The projects were for discrete elements of the program. Except for the first project, which included a power station, the Bank has only supported the downstream development of the MGDP: irrigation and related supporting services, as well as technical assistance. Table 3: BANK-FINANCED MAHAWELI PROJECTS a! Approval Loan Credit Proiect Name Year amount amount (US$mil) (US$mil) 1. Nshaweli Ganga 1970 14.50 15.46 2. Mahaweli Ganga II 1977 - 19.00 3. Mahaweli Ganga TA lb 1980 3.00 4. Mahaweli Ganga III 1981 - 90.00 5.. Maxaweli Ganga IV le 1984 12.10 30.00 26.60 157.46 la For details of individual projects see Table 1. lb Technical Assistance Project. /c Not yet effective. 21. The Bank's first project (1970) has been completed and was audited bw OED in 1981.81 Despite implementation delays of about two years, the project was regarded as economically and socially successful. The main reasons for success were considered to be sound project design, continued support from government agencies, and the ability of farmers to adapt to the rigorous time frame imposed by double cropping. However, certain prob- lems remained, such as cost recovery, monitoring and evaluation, effective water management, and illegal occupation of land. Although these problems were addressed in follow-up projects, they have remained critical issues until today. 22. Implementation of the second project (Mahaweli II - 1977) and the Mahaweli TA Project (1980) are the subject of this report. Implementation of the Third Mahaweli Project (1981) was delayed and completion of the project is now scheduled for December 1989. The Minipe Right Bank Transbasin Canal has been in operation since 1984. Construction of the main and branch canals is more than two years behind the original schedule due to long hauling distances of suitable materials, adverse weather and larger than expected rock formations. Downstream construction work and settlement activities have been rescheduled accordingly. The Fourth Mahaweli Project was approved in July 1984. Loan and Credit agreements have been signed, but have only become effective in early 1988 due to delays in resolving questions concerning cofinancing. 8/ Project Performance Audit Report, OED Report No. 3730, dated December 28, 1982.  III. PROJECT IMPLEMENTATION - MAHAWELI II PROJECT A. Project Formulation 23. Bank staff was concerned upon receipt of the eight volume feasi- bility report that i.nvestment costs of about US$1,800 per ac were high and that the proposed implementation rate of 14,000 ac per year was too opti- mistic. An FAOICP preparation mission was mounted in 1974 to prepare a much smaller project. The GOSL was unhappy with the Bank's decision to support only a small project. An early 1976 post-appraisal mission discov- ered that GOSL had already initiated work on a large scale, i e., covering about 71,000 ac and recommended that Government's efforts be encouraged and increased pace maintained (Mahaweli II PCR, paras. 2.01-2.10). Under these circumstances, a revised project was proposed. The Bank decided that the changes required a new full appraisal which focussed on construction methods proposed, procurement procedures, the implementation schedule, and organizationlmanagement aspects (see Mahaweli II PCR, paras. 2.11-2.17). 24. Negotiations took place in March 1977. Subjects discussed included procurement, audit arrangements, technical assistance and manage- ment appointment. However, the major debate centered around cost recovery issues (see also Mahaweli II PCR, paras. 2.18-2.20). The Board approved the project in April 1977 but fin'lizing the cofinancing arrangements resulted in delays in effectiveness. The Credit agreement beeame ulti- mately effective in December 1977 (see Mahaweli II PCR, paras. 2.21-2.25). 25. The project as approved included the following components (quanti- fied and tabulated in Mahaweli II PCR, para. 3.01)s (a) Irritation and Land Development Works included completion of Kala- vewa Right Bank Main Canal and improving four major tanks along the canal; construction of branch and distributory canals, field channels and other structures; improving natural drains (about 10 miles) and construction of secondary and tertiary drains; clearing and land preparation of about 40,300 acres; and provision of vehi- cles and equipment and a construction workshop for the construc- tion, operation and maintenance of the works. (b) Production Support included reorganization and strengthening of agricultural extension services in the project area; provision of storage, marketing and processing facilities; provision of vehi- cles, equipment, farm implements and sprayers; and improvement of some existing roads and construction of new roads. (c) Social Infrastructure included provision of schools, additional medical facilities and community development facilities as well as construction of about 1,235 village wells and about 200 miles of village roads.  (d) Otiher Proiect C.mponents included the settlement of abouc 15,300 famllies in the main project area; and provision of consultants' se-vices, studies and surveys. 26. Project costs were estimated at US$80.0 million, and funding arrangements included: IDA, Canada, the Netherlands, UK, USA, EEC, and GOSL. For further details on planned and actual financing arrangements see Mahaweli II PCR, paras. 3.23-3.24. B. Proiect Implementation 27. The decision to accelerate the MGDP (PPAM, para. 16) took place shortly after Board apiroval of this project. While it had an important bearing on the shape on the Mahaweli program it also had a rather immediate and chaotic impact on the implementation of the Mahaweli II Project (see also Mahaweli II PCR, paras. 2.23-2.25). 28. Work on the Mahaweli II Pro4ect was started by GOSL prior to Bank involvement. Following the Government's decision at the end of 1977 to accelerate the MGDP, the army and other public corporations were mov.'ed into the project area to clear jungle and construct canals. However, this wa done without due regard for proper planning and work standards. The result was poor quality of works, excessive water conveyance losses in the canals and operation and maintenance costs much higher than anticipated. The Bank rated the project as a problem project in 1978. Project implementation, however, improved in 1979180 and progress as well as quality improved. Implementation slowed down again in 1981, mainly due to budget restriction. In the end, instead of completing the project ahead of schedule as originally intended by Government, the credit had to be extended twice, until June 30, 1985, due, to a large extent, to the need for additional canal lining required as a consequence of poor design and standards prior to the Bank's involvement. However, ultimately the project successfully attained its major land development, settlement and production objectives (for further details, see Mahaweli II PCR, paras. 3.02-3.22). 29. The Mahaweli Agencies' procurement actions for civil works caused major problems throughout the project implementation period. Contrary to the agreement reached before the project's approval, MDB expressed the view that the private sector was not able to even undertake the agreed 502 of the construction. In the end, all civil works were undertaken by 5 public sector agencies, including MDB, the &Army and the River Valley Development Board (RVDB). The September 1978 supervision mission observed that no contracts were bid competitively, none were submitted to IDA for approval and that some of the works were proceeding on the basis of oral agreements. RVDB had been assigned the design, construction and supervision of irriga- tion and drainage works in over one-half of the project area and was oper- ating without a contract. With GOSL's decision to accelerate the MGDP, MDB had no power to enforce proper planning or scheduling of work or to enforce construction quality standards. Later RVDB was excluded from further contracts because of poor quality of its construction work and because of non-completion of some of the work. From 1980 onward the public sector agencies were phased out of the program and private contractors were  - 10 - awarded remaining contracts. Likewise, a relatively large procurement of 1,050 two-wheel tractors was cancelled because IDA objected to MDB's award process and because it was decided that farmers could buy tractors directly from local suppliers. Other contracts for other items of equipment were awarded late in 1984 following satisfactory ICB action by MEA (Mahaweli II PCR, paras. 3.25-3.29). 30. Cost recovery for land improvements was to be accomplished by the sale of the farm and house plots given to each settler, wherein the direct costs of development would be charged and the settlers would receive clear title to the land. However, at completion no payments had been made by the settlers and ownership still rested with GOSL (Mahaweli II PCR, paras. 3.32-3.34). Collection of water charges began in 1984 after a publicity and education campaign regarding recovery of O&M costs was carried out in 1983 (Mahaweli II PCR, paras. 3.35-3.36); while this was an encouraging start, much remains to be achieved. C. Project Administration 31. The Mahaweli Authority of Sri Lanka (MASL) was set up by an Act of Parliament in March 1979 to be the strong central planning and decision- making body that was needed to meet the challenges of the ambitious Accel- erated Mahaweli Program. Its two principal executing agencies were the Central Engineering and Construction Bureau (CECB), responsible for the large head works project3 and the Mahaweli Development Board (MDB), charged with engineering and construction of downstream irrigation development projects. Subsequently MASL set up a Settlement Division in July 1980 and by Cabinet decision the Mahaweli Econcamc Agency (MEA) was established in 1981 to be responsible for downstream development and settlement. In 1983, MDB5s functions were taken over by its successor, Mahaweli Engineering and Construction Agency (MECA), further streamlining the MASL organization and improving the coordination between engineering, construction and settlement (for further details see Mahaweli II PCR, paras. 4.01-4.05). 32. Water management was controlled by the Water Management Secretar- iat (WMS) in accordance with an agreed plan, in close consultation with MEA's Resident Project Manager (RPM). Present imbalances between water delivery and off-take by the system are now reduced to some degree by the available storage capacity of the tanks. Improvements have been made to reduce water wastage by users, modifications to canal linings provide better seepage control, measurement structures improve reliability and accuracy of water issues, and training of operators and farmers in effec- tive and efficient water use has helped to improve water management in the System (Mahaweli II PCR, paras. 4.06-4.11). D. Prolect Impact 33. Revised crop production estimates confirm that the production of paddy exceeded the SAR estimate--not only in exceptionally good years, but in average years as well. On the other hand, production of other field crops has remained below the appraisal estimates, with the exception of  - 11 - chillies which has become the second most important cash crop, after paddy (Mahaweli II PCR, paras. 5.01-5.08). 34. Farm i_comes range widely (with a median of about Rs 4,900 and a mean income of about Rs 8,500), according to a sample survey conducted by MEA in 1985, derived mainly from cultivation of fruits and vegetables, plant nurseries and livestock produce, inclvding hiring out of draught animals. Most families also have some non-farm income, including tractor hiring, leasing-in and cultivation of farm land, and trading (Mahaweli II PCR, paras. 5.09-5.12). 35. The project financed a full package of investments in part of the H system and supporting services throughout the system area. Economic analysis has been limited to the full package cost and benefits. The reestimated rate of return is -12 compared to 21X estimated at appraisal. Depressed rice prices in recent years and low forecast are the dominating reasons for the low rare of return: using appraisal price forecasts instead would give a 162 rate of return (for further details see Mahaweli II PCR, paras. 6.01-6.03). IV. MAHAWELI TA PROJECT A. Prolect Formulation 36. The MGDP was originally based on a Master Plan prepared jointly by UNDPIFAO and Sri Lanka engineers in 1965-68 which envisaged a three-phase development program of about 900,000 ac of irrigable land in the dry zone, as well as about 500 MW of hydropower. The new Government which came to power in 1977 decided to accelerate the program. During the early months of 1978 the Government consulted with IDA and contacted several donors for assistance in the preparation of five major projects, including dams and hydropower plants with an installed capacity of about 600 MW, and development of about 290,000 ac of new lands, as well as downstream development and other works and studies (Mahaweli TA PCR, paras. 2.01-2.03). 37. As part of the general agreement to accelerate the program, IDA undertook to advise on the scope of studies. To this end, an Implementa- tion Strategy Study was carried out, financed by the Netherlands, done in 1979 by consultant*a, to determine a specific action plan; its findings differed signific*e.tiy from the original UNDPIFAO Master Plan. Ar.an the important findings were that the system as proposed would have excess water which could be used to irrigate lands elsewhere and that water diversion northward could be done at substantially lower cost as compared -ith the Master Plan proposals. This brought out the need for further feasibility studies of the transbasin diversion plans. There was also an Immediate need for detailed engineering and design in the command area to be served by the Victoria Project, since its construction was to start in early 1980 and the downstream development should be in place by the time the headworks were to be completed in 1984. Accordingly, at GOSL's request, the proposed  - 12 - technical assistance project was identified and jointly prepared by GOSL with IDA assistance (Mahaweli TA PCR, paras. 2.04-2.06). 38. Negotiations took place in Sri Lanka in December 1979 and the project was approved by the Board In January 1980. The credit agreements, including an EEC Special Action Credit, were signed In April and became effective in July that same year. Total project costs were estimated at US$6.0 million. IDA would fund US$3.0 million and EEC would provide US$2.0 million equivalent for a Special Action Credit to be administered by IDA (for further details see Mahaweli TA PCR, para. 3.01). 39. The project's primary objective was to finance consr.ltants tos (a) carry out reconnaissance level studies of transbasin diversion plans to convey surplus water developed under the Accelerated Mahaweli Ganga Program (AMP) to three alternative basins/areas for irrigation use and power production, determine the viability of each plan, and identify possible viable subprojects; (b) review detailed designs of the Minipe Right Bank Canal (planned to serve Systems B and C) and prepare terder documents suitable for international competitive bidding and possible financing of this canal by the Bank Group; (c) assist the Mahaweli Authority in the priparation of final designs and tender documents for irrigation and drainage systems and social infrastructure in part of System C, which was being consid- ered for Bank Group financing; (d) carry out other studies/investigation plans or designs required during the planning and implementation of the AMP; and (a) provide training to counterpart staff and locally-hired personnel engaged in the work. B. Project Implementation and Inviact 40. Most of the project objectives were achieved (project implementa- tion details are provided in the PCR, paras. 4.01-4.09). The project helped Government in defining development options for possible use of sur- plus Mahaweli water being developed by the various headworks projects although the final report was not completed until the end of the extended project period. It also contributed substantially to the timely completion of the Minipe Right Bank Canal. The canal was financed by IDA (under the Mahaweli III Project) and coAmir3ioned in April 1984. The assistance given by consultants financed under the Mahaweli TA Project in the review/preparation of designs and tender documents for civil works in System C enabled the Mahaweli III Project to be implemented as planned. The training and guidance given to MASL staff in the design of irrigation and social infrastructure and in the preparation of tender documents was not only instrumental in completing such work on a timely basis in System C  - 13 - but has contributed to the preparation of similar work in other project areas by MASL staff trained by the consultants. 41. The project was least successful in providing support to other studies/investigations in the AMP. The main reason was that as a result of a scaling back of the AMP, such studies were not identified by the Govern- ment in time to utilize the funds available. Some assistance was provided near the close of the credits in reviewing designs and tender documents for the System B Right Bank project (also appraised by the Bank and approved by the Board as Mahaweli IV). Because additional studies were not taken up as envisaged, 38 percent of the IDA credit (US$1.14 million) and 14 percent of the EEC credit was eventually cancelled. 42. In retrospect, the two years provided for project implementation was not realistic. Particularly the transbasin study, where various alter- natives or options kept surfacing, continued to require further study. Also, more time was needed to identify the other studies for AMP which had not been identified at the time of project preparation. The credits were extended three times for a total of three years, making this a five-year project. Nevertheless, and this is unfortunate, there was a period of nearly two years of project inactivity (during the third and fourth years of the project), due mainly to the scaling back of development plans in the Mahaweli Program and to a lack of consensus within MASL/GOSL on the studies/work to be undertaken. V. MAIN FINDINGS 43. Government's commitment to the implementation of the Mahaweli Ganga Development Program has and continues to be strong and it succeeded in attracting substantial external financing for the accelerated program from foreign donors. However, the decision to speed up was not without risk. Mahaweli must be reviewed within its overall development context and rather than looking at individual components, the program should be evaluated as a whole. Obviously it is still too early to determine the full development impact of the program although a first assessment has been attempted in the OED study: Sri Lanka and the World Bank--A Review of A Relationship, cited earlier in this audit report. 44. On the other hand, the project completed one of the earliest systems (System H) of the Mahaweli program. As such what happens here has been and can continue to be replicated elsewhere in the program. There- fore, it is rather important to ensure a continued evaluation of the devel- opment impact of this project. Serious consideration is therefore given to an impact evaluation to assess sustainability and development impact of this project in more detail.  - 14 - A. The Second Dimension of Sustainabilitv 45. The 1985 OED Report Sustainability of Projects: First Review of Experience defines project sustainability as 'the maintenance of an accept- able net flow of benefits throughout the economic life of the project".9/ The report lists a number of factors which are necessary to achieve sus- tainability such ass (i) Borrower commitment which determines if the policy environment will be positive and whether the project will be ener- getically maintained and efficiently operated; (ii) continuity in project organization and management; (iii) project design which must provide the correct balance between the desirable and the feasible in its technical, institutional and socio-economic dimensions. 46. The Bank has been concerned with project sustainability for many years and project design, appraisal and supervision are encouraged to focus on the sustainability issue. However, in most projects in general, but in the Mahaweli Ganga projects in particular, two factors enhancing project sustainability did not receive the required emphasis: first, the socio- economic dimensions and second, the planning of future developments. 47. It is generally recognized that continuous successful operation of any agricultural development project will depend on sustaining the interest of participating farmers. There is general agreement that consultations with farmers are necessary prior to planning or deciding on policies since the ultimate success or the sustainability of any project depends on the cooperation of the farmers. While such consultations are relatively easy in developed countries because of the presence of active and often powerful farmer organizations, these 'farmer lobbies' are usually non-existent in the Third World. To ask for views from all farmers would be unrealistic but there must be a way to listen to farmer representations and take cogni- zance of their needs and expectations. Frank discussions with farmer groups before policies are decided and plans formulated are in support of project sustainability. 48. Since dealings/discussions with all project farmers are not feasi- ble, special emphasis on training of farmer leaders/speakers becomes imper- ative. A recent OED auditlO/ commented on a successful program introduced in Malaysia which focuses specifically on training farm group leaders. Some thought to this approach has been given not by the Bank but in Sri Lanka by the Mahaweli Ganga Authority in a reportll/ which states: 'The end result of an irrigation or rural development project should not be the provision of land and water for cultivation. 9/ See OED Report No. 5718, dated June 14, 1985, Executive Summary, page i. 10/ See Project Performance Audit Report, Malaysia: National Extension Project, (Loan 1493-MA), OED Report No. 7287 dated June 13, 1988. 11/ Sustainability of Irrigation and Rural Development Projects in Sri Lanka, July 15, 1987, by Dr. Abhaya Attanayake.  - 15 - It should not only provide the much required water for culti- vation and all the other social, religious and medical needs but also post settlement needs of the farmers. It is neces- sary to organize training programmes in agriculture, water management enterprise and community development for farmers to build a rural leadership which would give the necessary impe- tus for settlers in development projects to take up the responsibility of ensuring their own welfare. The ultimate goal of a settlement would be to make the maA4mum use of land through an all round development effort and farmer training. It is important in that it would facilitate the organization of farmers who would point out their own needs and wants and also be responsible for the smocth functioning of the project. In short, it would pave the way for the creation of a self- contained community based on agricultural, religious and other social needs and a well organized management system." 49. The second aspect of enhancing the chances of project sustaina- bility is proper planning of activities. While planning at the "micro level", i.e. planning at management level to cope with day to day opesra- tions and adjusting to short term trends is well taken care of in the proj- ects under review, a different picture emerges when looking at the "macro economic" aspects which may have an impact on project sustainability and where some indicative planning, providing for alternative solutions under differing scenarios would be required. As far as Mahaweli is concerned the issue is not Implementation but sustainability. It is almost sure that the high level of support and management cannot be continued forever. However, no planning to face such realities is being undertaken even though sustain- ability experience with earlier basin developments in Sri Lanka has been far from satisfactory. 50. In the audit's view it will be necessary to give more attention to analyzing long-term markets, production, price trends, etc., and preparing plans which could be implemented if certain anticipated trends materialize. B. Bank Supervision 51. The PCRs emphasize the high standard of the Bank's project super- vision, especially the continuity of staff assigned and their positive impact on project implementation. This view is also shared by the Sri Lanka authorities associated with the project and the audit agrees with this assessment. However, some observers within the Bank have reservations about the comprehensiveness of the supervision activities in regard to the projects under review and the audit paid special attention to the super- vision aspects. 52. The above mentioned reservations were based on the fact that the "continuity" of project supervision of the Mahaweli Ganga projects was assured by assigning a highly qualified and motivated engineer and other dedicated staff to this task. Due to the specialization of the supervisors in charge Bank staff feared that supervision missions may have tended to  - 16 - emphasize engineering/agro-technical aspects to the detriment of economicl administrative problems. 53. Their criticisms are based on the fact that supervision reports of the Mahaweli Ganga projects mention that 'implementation has proceeded without proper planning and coordination' or mention the 'failure to look at trends in relation to social development and inequality, and to water management". Concern was also expressed that the issue of water charges remained unresolved for many years. 54. In reviewing the supervision aspects the audit found no indication that supervision missions were one-sided, i.e. emphasizing only technical aspects. The examples listed in the Mahaweli Ganga projects, not tackling problems such as water charges and failure to recognize socio-economic trends cannot be considered to fall into the competence of supervision missions. The mentioned issues are highly political and sensitive and the audit finds that by bringing these issues to management attention and by inviting management action the supervision missions adequately covered their terms of reference. C. Civil Works Contractors 55. The Mahaweli II Project had a sizeable civil works component as can be seen from Table 4 below: Table 4: CIVIL WORKS UNDER MAHAWELI II Total Private Contract (US$ millions) (USS millions) Kalawewa Main Canal, etc. 3.7 1.85 Distributory Canals 13.8 13.80 Roads, Social Infrastructure, Buildings 9.2 4.60 Other 9.1 8.25 Total 35.8 28.50 56. This massive work program has to be seen in context with the over- all turnover of the country's construction sector which amounted to about 'JS$120 million in the appraisal year (1977).121 The difficulty of coping with the envisaged construction program was further exacerbated by the Bank's recommendation that the major share of civil works contracts should go to the "small contractors", whose share of the annual US$120 million volume could be considered small in light of the work done by government- owned or parastatal construction companies. 57. It is difficult to understand why the Bank insisted on getting the sa11 contractor sector involved since even the appraisal had reservations 121 See Sri Lanka Recent Economic Development and Policies for Growth, Report No. 5628-CE, dated May 14, 1985.  - 17 - (SAR, para. 5.05) which states: 'The private sector in Sri Lanka has very limited capacity for civil works construction.' 58. The apparent lack of analysis of the contractors sector led to delays in project implementation and contributed substantially to the cost overruns. The audit agrees with the statement of the Mahaweli II PCR (para. 3.25, 3.28) that 'procurement actions for civil works caused major problems throughout the project implementation period' and that contracts to small private enterprises" caused many problems until the very end of the project." 59. There were two shortcomings in the appraisal approach of the proj- ects' civil works component. First there was never a proper assessment of the country's construction sector, its weaknesses and specifically its annual capacity. Second, a lack of interaction not only between the Bank's different technical departments but even within the Region's agricultural divisions. While the audit agrees that a developing country's small contractor Internrise sector needs to be developed there is little appreci- ation in the Bank on what the major constraints are that confront the sector. The situation has also been encountered in other projects.131 60. Small contractors as found in some--not all--villages tend to employ very few workers. Due to transportation as well as housing prob- lems, the circle of work activities is rather narrow--it hardly reaches beyond a village. Furthermore construction techniques may not be up to required standards, a fact which cannot be overcome by providing these contractors with additional machinery. Since adherence to local bidding procedures is the minimum condition regarding the Bank's procurement proce- dures, many of these contractors are not capable of submitting cost esti- mates in writing but present them orally which the Bank in turn would not accept (Mahaweli II PCR, paras. 3.26). 61. The interaction between Bank divisions for any project is impor- tant and highlighted in this case by the, relatively speaking, small size of the country's construction sector. Interaction is missing if looking at projects individually--without seeing the Bank-promoted construction volume in the aggregate. Since there is always a chance that the education, pub- lic utilities and transport sectors will also require local contractors, _.n annual assessment of these activities would be warranted prior to determin- ing the appropriate civil works contract arrangements. 13/ See also Ethiopia Second Livestock Project (Credit 365), OED Report No. 3977, dated June 21, 1982.  - 19 - Table 1 Pope 1 PROJECT PERFCRMANCE AWIT REPORT SRI LANKA MAIAUEU GANGA DEVELOPMW PROJECT I (CREDIT M91-CE) MWAAEU GANGA TEC04ICAL ASSISTANCE PROJECT (CREDIT 979-CE SRI LANKA LENDING PROGRAM (as of December 81, 1987) Loan/ Loan Credit Amount La ProJect Nam Credit No. Credit Loan Aprovel Dnate SOR US11 lb US Apricultue. Lift Irrigation Cr. 121 - 2.20/d - June 1968 Drainage & Land Reclamation Cr. 168 - 2.64Ld - Oct. 19s9 DaIry Dove lopoent Cr. 504 - 9.eeLd - July 1974 Agricultural Devolopment Cr. 696 - 2S.00Ld - Doe. 1975 Tank Irrigation Modernization Ce. 666 - s.ned - Nov. 1976 Tree Crop Divereittcation Cr. 819 - 4. SOld - June 1978 Tree Crop Rehabilitatlon Cr. 818 - 21msoed - June 1978 Kurunegalu Rural Dovelopmnt Cr. 691 - 20. W 0d - March 1979 Agricultural Extension and Cr. 981 - 15.6 5d - Juno 1979 Adaptive Resaarch SmalIholder Rubber Rehabilitation Cr. 1017 - 16.00 - may 1989 Second Rural Developmnt Cr. 1079 25.00 88.60 - Dec. 19890 Village Irrigation Rehabilitation Cr. 1160 24.50 89.80 - June 1981 Tea Rehabilitation & Diveret- Cr. 1240 17.809L 2.00d - may 1982 fication Forest Resoure. Developmnt Cr. 1817 8.60 9.06 - Jan. 1988 Third Rural Development Cr. 1863 21.40 28.00 - may 1988 Major Irrigation Rhabilittation Cr. 1587 17.00 17.00 - DOc. 1984 Fourth Tree Crops Cr. 1662 66.60 65.00 - March 1985 Dairy Dovelopment II Ln. 2778L. - - 88.0J June 1985 Agricultural Research Cr. 177lf 1S.40 18.60 - April 1987 186.70 826.94 88.00 Mahawelt Canoe Proprom Mahawel G ango DOv. I Ln. 658 - 15.48Ld 14.S01d Jan. 1970 Cr. 174 Mahawell Cango D-v. II Cr. 701 - 19.00/d - Apri l 1977 MahaweII Gangs T. A. Cr. 979 - 8.0D/d - Jan. 1980 Mahawelt Gangs Dev. III Cr. 1106 78.89 90.60 - June 1981 MahaweII Gangs Dev. IV Lo. 24871o 28.U8 80.00 12.10 June 1984 Cr. 1494 101.00 157.48 26.80  - 20 - Tabl- 1 Pag 2 PROJECT PERFORMANCE AUDIT REPORT SRI LANKA MAHAIW.I GANA DOEVELOPMENT PROJECT T (CREDIT 701-CE) MAHAnELI QANGA TECHNICAL ASSISTANCE PROJECT (CREDIT 97t-CE) SRI LANKA LENDIN4 PROGRAM (as ot Decber 81, 1987) Loan/ Loan Credit Amount 1 Project Nam Credit No. Credit Loan Asoroval Date SOR US /b USS Power Aberdb n-Laksapana Power Ln. 16l - - 19.11/e Jun. 1904 Grandpass Thermal Power Ln. 209 - - 7.40/e July 1905 Norton Bridge Thermal Ln. 208 - - 15.6900 April 1#51 Power VI Ln. 687 - - 21.05od July 1989 Power V Cr. 872 - 6.665d - April 1978 Power VI Cr. 1048 - 19.5 - June 1980 Power VII Cr. 1210 S1.80 86.00 - Feb. 1982 Power VIII Ln. 2187 - - 42.70/d June 1982 Power IX Cr. 1786 48.20 52.00 - Oct. 1984 74.S0 118.53 106.21 Tranuortatlon Highway Ln. S69 - 5.C d 4.0kLe Oct. 1#68 Cr. 188 Road Maintenance Cr. 900 - 1C6 SOj - May 1979 Road Passenger Transport Cr. 994 - 6800/d - March 1980 Road Maintenance T Ln. 2517 - 24.00 Aprll 1905 - 74.58 28.90 pFC Develop men t Flnance Corp. I Ln. S20 - - 4.066& Nov. 1967 Development Finance Corp. IT Ln. 684 - - *.005 July 1#90 Developmnt Finance Corp. III Cr. Sa6 - 4.5COd - June 1975 Develop men t Finance Corp. IV Cr. 742 - 8.001d - Sept. 1977 - 12 W 12.00 Industry Small and Mdium Industry Cr. 942 - 16.05Ld - June 1979 Small and dium IndustryII Cr. 1162 26.76 80.00 - Oct. 1to1 Industriol Development Cr. 1401 28.10 25.05 - July 1908 Second Industrial Development Cr. 1C92 17.40 20.05 - may 1986 Third Smll and Medum Industrls Cr. 16k lS70 20 00 - Dec 1907 82.90 111.05 -  - 21 - Takla I Peg. 8 PROJECT PERFORMANCE AUIT REPORT SRI LWNmA MANAWELI GANGA OEVELOPMENT PROJECT IS (CREDIT 7M1-C