Sri Lanka - Agricultural Extension and Adaptive Research Project

Document of The World Bank FOR OFFICIAL USE ONLY MIC.OFICHE COPY Report No. 10632-CE Type: (PPR) ReportNo. 10632 SARMIENTO,/ X31680 / T9032X/ OEDD1 PROJECT PERFORMANCE AUDIT REPORT SRI LANKA AGRICULTURE EXTENSION ANP ADAPTIVE RESEARCH PROJECT (CREDIT 931-CE) MAY 13, 1992 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. ABBREVIATIONS AEARP Agricultural Extension and Adaptive Research Project AGA Assistant Government Agent AI Agricultural Instructor ASSP Agricultural Support Services Project AO Agricultural Officer CCB Coconut Cultivation Board CF Contact Farmers DAPH Department of Animal Production and Health DEA Department of Export Agriculture DFP Department of Food Production DOA Department of Agriculture DS Divisional Secretary FAO Food and A 'riculture Organization FF Follower Farmer FPO Food Production Overseers GN Grama Niladhari (all-purpose village-level-worker) GOSL Government of Sri Lanka ICB International Competitiva Bidding IDA International Development Association IMF International Monetary Fund JSP Janasaviya Program (Poverty Alleviation Program) KVS Krushikarma Vyaptha Sevaka (village level extension worker) HARD Ministry of Agricultural Development and Research MRID Ministry of Rural Industries Development OED Operations Evaluation Department PCR Project Completion Report PPAR Project Performance Audit Report RTWG Regional Technical Working Grcups SAR Staff Appraisal Report SMO Subject Matter Officer T&V Training and Visit (an agricultural extension system) UNDP United Nations Development Program FOR OFFICIAL USE ONLY THE WORLD BANK WashinCton, D.C. 20433 U.S.A. Office of Directw-General Operatons Ivaluaion May 13, 1992 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on Sri Lanka - Agricultural Extension and Adaptive Research Proiect (Credit 931-CE) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on Sri Lanka - Agricultural Extension and Adaptive Research Project (Credit 931-CE)" prepared by the Operations Evaluation Department. 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 PROJECT PERFORMANCE AUDIT REPORT SRILAA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (CREDIT 931-CE) TABLE OF CONTENTS Page No Preface . . . . . . . . . . . . . . . . . . . . . . . . . . i Basic Data Sheet . . . . . . . . . . . . . . . . . . . . . . . . . . v Evaluation Summary . . . . . . . . . . . . . . . . . . . . . . . . . . v I. INTRODUCTION ................... .*** 1 II. PROJECT EVOLUTION, OBJECTIVES AND DESIGN . . . . . . . . . . . . 2 Project Evolution . . . . . . . . . . . . . . . . . . . . . 2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . 4 Design . . . . . . . . . . . . . . . . . . . . . . . . . 5 III. IMPLEMENTATION ......................... 6 Implementation Experience....... . ... . 6 Poet Implementation: Recent Development Affecting Agricultural Extension . . . . . . . . 8 IV. PROJECT OUTCOME . . . . . . . . . . . . . . . . . . . . . . . . 10 Agricultural Imp atc. . . . .......... 10 Institutional Impact . . . . . . . . . . . . . . . . . . . 11 Sustainability . .* 12 V. FINDINGS AND LESSONS . . . . . . . . . . . . . . . . . . . . . . 13 ANNEXES Annex I Table 1: A. Civil Works; B. Procurement of Vehicles . . . . . 17 Table 2: Total Project Staff . . . . . . . . . . . . . 18 Table 3: Technical Assistance ................ 19 Table 4: Overseas Training . . . . . . . . . . . . . . . . . . 20 Table 5: Indicators of the Monitoring Survey on T & V Extension System 9 . 9 . . . . . . . . . . . . . 21 Annex 2 Table 1: Paddy Yields 1975-1989 . . . . . . . . . . . . . . . 22 Table 2: Paddy Cultivated Area and Production: Rice Availability, Procurement and Distribution, 1975-89 . 23 Table 3: Cultivated Area and Production of Subsidiary Food Crops, 1975-89 . . . . . . . . . . . . . . . . . 24 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. J'ROJECT PERFORANCE AUDIT REPORT SRILAA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (CREDIT 931-CE) PREFACE This is a Project Performance Audit Report (PPAR) of the Agricultural Extension and Adaptive Research Project in Sri Lanka involving IDA credit in the amount of US$15.5 million. The credit was approved on June 12, 1979 and signed on July 24, 1979. US$7.96 mil ',on of the credit was disbursed and the undisbursed balance of US$7.54 million was canceled. The credit vas closed on June 30, 1986, one year behind schedule. Final disbursement of the credit was on December 31, 1986. The PPAR is based on the Project Completion Report (PCR)1' of the project, the Staff Appraisal' and President's Report,-' the Credit Agreement,!' the transcript of the Executive Directors' meeting at which the project was considered, a study of project files and discussions with Bank staff who were associated with or knowledgeable about the project. An ORD mission visited Sri Lanka in May-June 1991 and discussed the effectiveness of the Bank's assistance with the Government of Sri Lanka (GOSL) officials in the Department of Agriculture and with some provincial authorities. The kind cooperation and the valuable assistance of the GOSL in the preparation of this report is gratefully acknowledged. The PCR provides a full account and assessment of the project's implementation experience. The PPAR elaborates on specific aspects such as project design, agricultural and institutional impact, and post-implementation developments affecting agricultural extension. Following standard OED procedures, copies of the draft PPAR were sent to the relevant GOSL officials for comment. However, no comments were received. I' Project Completion Report: Sri Lanka Agricultural Extension and Adaptive Research Project (Cr. 931-CE), Report No. 7684, March 31, 1989. l Staff Appraisal Reports Sri Lanka Agricultural Extension and Adaptive Research Projecte Report No. 2396a-CE, May 17, 1979. Al President's Report: Sri Lanka Agricultural Extension and Adaptive Research Project, Report No. P-2569-CE, Nay 31, 1979. It Development Credit Agreement: Sri Lanka Agricultral Extension and Adaptive Research Project, Credit No. 931-CE, July 24, 1979. - ii:i - PROJECT PERFORMANCE AUDIT REPORT SRI LANKA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (CREDIT Y31-CE) BASIC DATA Key Proiect Data A:tual Z of Appraisal Estimate Actual Appraisal Estimates Total Project Cost (US$ million) 22.4 12.2 54 Credit Amount (US$ million) 15.5 7.96 51 Cancelled (US$ million) ---- 7.54 ---- Institutional Performance ---- Fair ---- Agronomic Performance ---- Fair ---- Cumulative Estimated and Actual Disbursement (in US$ million) FY80 FY81 FY82 FY83 FY84 FY85 FY86 FY87 Appraisal Estimates (US$ million) 0.7 4.5 7.8 11.0 14.0 15.5 -- -- Actual (US$ million) -- -- 1.4 2.2 3.2 4.7 6.0 7.96 Actual as I of Appraisal -- -- 18 20 23 30 39 51 Date of Final Disbursement December 31, 1986 Proiect Dates Orisinal Actual First Mention in Files ---- 11/03/77 Negotiations ---- 05/03/79 Board Approval ---- 06/12/79 Credit Signing ---- 07/24/79 Credit Effectiveness 10/24/79 10/04/79 Credit Closing 06/30/85 06/30/86 Physical Components Completion 12/85 12/86 Staff Input (In Staff Weeks) FY78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 Total Preappraisal .1 9.8 -- -- -- -- -- -- -- -- -- 9.9 Appraisal -- 59.7 -- -- -- -- -- -- -- -- -- 59.7 Negotiation -- 4.8 -- -- -- -- -- -- -- -- -- 4.8 Supervision -- 2.0 30.4 16.4 19.1 15.3 19.5 12.9 7.4 13.2 4.0 140.1 Other -- .6 -- -- .4 -- -- -- -- -- -- 1.0 TOTAL .1 76.9 30.4 16.4 19.5 15.3 19.5 12.9 7.4 13.2 4.0 215.6 - iv - Mission Data Date No. of Man Days Specfica- Perf. Rating Type of o/Yr Persons in Field tion- Statuall Trend- Problems/ Appraisal 10/78 4 A,B,C,D -- -- -- Supervision I 07/79 4 19 A,C,D,F 1 2 M Supervision II 01/80 2 9 B,E 2 1 M Supervision III 7-8/80 1 11 B 2 2 M,T Supervision IV 11/80 1 6 B 2 1 M,T Supervision V 07/81 4 15 A,B,C 2 1 M,T Supervision VI 02/82 2 10 A,B 2 1 M,T Supervision VII 09/82 3 12 B,C,E 2 1 M Supervision VIII 7-8/83 4 17 B,C,D,E 2 1 M Supervision IX 03/84 2 9 A,B 2 1 N,F Supervision X 9-10/84 1 7 A 2 1 F Supervision XI 03/85 1 5 A 2 1 M,F Supervision XII 9-10/85 1 13 A 2 1 M,.W Supervision XIII 02/86 1 12 A 2 2 M,F Supervision XIV 04/87 1 1 A 2 2 -- Other Prolect Data Borrower i Government of the Democratic Republic of Sri Lanka Fiscal Year : January 1 to December 31 Name of Currency a Rupees (Re) Exchange Rate Appraisal Year a 1978 US$ 1 - 15.80 Intervening Years/Average a 1979 US$ - 15.57 1980 US$ a 16.58 1981 US$ - 19.25 1982 US$ - 20.80 1983 US$ - 23.53 1984 US$ * 25.44 1985 US$ - 27.16 Average US$ * 2!.19 Completion Year 1986 US$ - 27.95 Post Implementation Years 1987 US$ - 30.76 1988 US$ - 33.03 1989 US$ - 40.00 1990 US$ - 40.24 A a Agriculturist; B - Agricultural Economist; C a Extension Specialist; D a Rural Development Specialist; E - Agricultural Extension/Research; F = Engineer. 1 u Problem freei 2 - Satisfactory 3 * Deteriorating. I * Improving; 2 - Stationary; 3 * Deteriorating. F * FInancial; It w Managerial; T a Technical; D a Others. -v - PROJECT PERFORMANCE AUDIT REPORT SRI LANKA AGRICULTURAL EXTENSION AND IDAPTIVE RESEARCH PROJECT (CREDIT 931-CE) EVALUATION SUMMARY Introduction earlier statewide T&V extension and adaptive research projects in India. 1. The Agr. _ltural Extension and The objective of AEARP was to assist Adaptive Research Project (AEARP) is Sri Lanka to apply the T&V system of the Bank's first and only investment extension to the entire country, to on a national scale in the agri- strengthen adaptive research and to cultural extension sub-sector in Sri improve and expand the training of Lanka. The Bank to date has approved extension staff. The project in- credits for 24 projects amounting to cluded funds to strengthen extenaion US$549 million for the agriculture services, establiah an adaptive re- sector. These include Mahaweli Ganga search network, improve agricultural Development, irrigation, treecrop, training institutions, provide over- irea development, dairy development, seas training for staff, provide forestry, input supply and agri- salaries and allowances for incremen- cultueal research projects. tal staff, con&truct offices, labora- tories and staff housing, purchase 2. In 1V90 agriculture accounted and maintain vehicles and eqaipment for about 24 percent of GDP, 43 and provide 138 man-months of techni- percent of employment and 32 percent cal assistance. of exports. In recent years agri- culture's importance to the economy Implementation Experience has declined due to its poor per- formance, partly caused by the ethnic 4. The Credit became effective on conflict in the North and East, civil October 4, 1979 and closed on June disturbances in the South and 30, 1986, a year behind schedule. unfavorable weather conditions. US$7.96 million of the original Treecrops mainly (tea, rubber and credit amount of US$15.5 million was coconut) account for 40 percen* and disbursed and the balance was can- paddy accounts for 33 percent of the celed. Actual. project costs (US$12.2 major areas of permanently cultivated million) were only 54 percent of SAR land. About 73 percent of agri- estimates (US$22.4 million) in US cultural land is under smallholdings; dollar terms and 81 percent in local 50 percent of the smallholders currency terms. This difference was operate 85 percent of the cultivable due to a local currency devaluation land, 39 percent have home gardens, of 80 percent relative to the US and about 11 percent are landless. dollar between 1979 and 1986. Local currency costs of all project Obiectives expenditure categories equalled or exceeded SAR estimates despite the 3. The design and objectives of substantial scaling down of the AEARP was almost identical to the project, except expenditures for - vi - operating costs which were cut eystem during imrlementation and substantially. Not only was the after project completion according to Bank's actual disbursement for the estimates of DOA staff. This meant loan about half of what was origi- that the extension system operated nally approved, b4t GOSL's contri- below its optimum as one of the key bution was also less than planned ingredients -- staff mobility -- was (about 61 percent). reduced to a minimum by scarce oper- ational support funds. 5. The civil works component was drastically reduced. Only about 43 Post Implementation percent of the istimated number of units were constructed. This was 7. First, a finacial crisis largely due to start-up delays, confronted the new *. tient soon design changes, poor supervision, and after it came to off i in early most of all, a sha-p increase in 1989. The fiscal deficit became so construction costs. Procurement of large that the Goverment becam goods waq also delayed due to DOA's unwilling to provide the level of lack of familiarity in dealing with recurrent expenditures required to international competitive bidding operate an efficient extension (ICB) procedures. In the end system. The costs of operating the though, most of the equipment and extension system had become prohibi- machinery were bought, but shortfalls tive over he years, and although in the purchase of jeeps and vans staff were not reduced or relocated, adversely affected the adaptive re- expenditures for operating the system search program since mobile units were drastically cut. could not be formed, and supervision wa lacking. 8. Second, although an even more serious set-back to the extension 6. The Treasury approved 1,568 system came in late 1988 as a result positions for incremental staff out of a Constitutional Amendment (which of a total of 1,789 proposed in the involved devolution of administrative SAR; at the time of project closing authority to the Provinces), the 1,549 positions had been filled. change is viewed as a blessing in Staff shortages were most serious at disguise as it is helping to reshape the KVS, AI,.A0 and SMO levels, and a more cost-effective extension sys- for technical staff assigned to tem. The devolution of agricultural adaptive research activities. The extension has meant that the entire project ws subjected to a number of 2,400 KVS cadre was transferred to other changes during implementation. the Poverty Alleviation Program. Of note: i.) extension for livestock Under this program, the KVS joined development was excluded from the the other village level all-purpose project activities and remained in workers known as the Gra-A Niladhari the DAPH; ii) the 138 man-months of (GN). The GN, including the V.S9 and expatriate consultanciss in the the Al are now placed under the Divi- fields of extension, animal hus- sional Secretary who is in charge of bandry, water manag6ment and crops all Government agencies at the Divi- were reduced to 50 man-months since sioaal level. At present, there are DOA expressed serious reservations about 14,550 GNs who collect data and regarding the hiring of such a large perform regulatory functions at the number of consultants; and iii) staff local administrational levels. Under salaries accounted for about 85-90 this system, the ex-KVS cannot per- percent of the cost of operating the form any extension work and the link- - vii - age with DO&, including technical supportive Government policies such linkage has more or less dissipated. as guaranteed price and marl"t have An Agricultural Support Services 01 contributed to increasing yield Project (ASSP) appraised by the Bank and production. It would be diffi- in July 1991 is expected to resurrect cult to isolate the cmntribution of an extension system based on the any single factor in this package of lessons learnt from AEARP. complementary factors. 12asulte 11. T&V extension appears to have played sn Important role essentially 9. Agricultural production in- to maintain, for a period of time, creased during the period of project the gains of *mproved technology and implementation, but the impact of incentives with the result that wide- AEARP, like similar projects else- spread adoption did take place in Sri where, is hard to measure and diffi- Lanka within a relatively short cult to attribute. In reality, paddy period of time. The strength of T&V yields began tc increase even before extension during the 1979-84 period the introduction of T&V in Sri lanka. was that it vas able to focus on a Paddy yields increased from about few aimple messages which happened to 2.3 tons/ha in 1975 (maha season) to be the relevant and important ones at 2.9 tons/ha in 1979/80 (Naha season). the time, and with emphasis on dis- Yields kept on increasing after T&V seminationg substantial gains from extension was introduced in 1980, paddy yields were achieved. However, reaching a high of about 3.4 tons/ha after 1985, the very strengths of T&V in the 1984/85 Naha season, most extension proved to be a liability. patily cultivation having switched to The reservoir of simple messages were high yielding varieties after T&V exhausted as paddy yields reached a extension was introduced in 1980. plateau, and as the system expanded However, after 1984/85, there has to cover subsidiary food crops and been no increase in yields, although minor export crops. The T&V ex- the T&V system was in full operation- tension system began to breakdown as al force between then and 1989. consideration of all these crops Yields have been constant and produc- required a whole farm approach in tion actually dropped in two of the contrast to the simple technical three years since 1986. messages associated with the "single crop" approach of the past. 10. The above are important con- siderations for assessing how ef- SustainabilitZ fective T&V extension has been in increasing paddy yields in Sri Lanka. 12. From the start, sustainability The general increase in paddy yies was dependent on the level of funding and production in the country bet,en provided by the Government and the 1975 and 1984 are attributable to a political and bureaucratic support number of complementary factors which for such funding. In 1988, two years affected productivity, sometimes as after project completion, agricultur- initiating and catalytic factors and al extension cost the Goverment sometimes as additive and facili- about Rs. 145 million (about US$4.4 tating factors. The availability of million in 1987 US dollar). This was high yielding paddy varieties, subsi- roughly 0.88 percent of agricultural dised fertilizer, improved quality GDP. Actual staff salaries of the ***ds, new and rehabilitated irriga- extension system alone accounted for tion facilities, &.panded credit and about 85-90 percent of the operating - viii - costs of the system. This compares centives favoring the agriculture with SAR estimates that incremental sector and is likely to have speeded operating cost at full developme. ? up the ratc of adoption of improved would amount to 30 percent of DOA's rice technology. current annual operating budget. Given the perceived and actual incre- 15. On the negative side, the ex- mental benefits of maintaining the tension system did not adapt to new extension system, these costs were extension challenges after 1984. grossly out of proportion and too This and its high fixed personnel high for the Government to maintain. costs led it to be perceived as The Government became especially co6tineffective in its existing form. concerned once paddy yields levelled The extension system was eventually off, production began to stagnate an' dismantled in the 1989 Constitutional new problems began to arise when the Amendment, which derolved adminis- extension systen attempted to diver- trative responsibility to the sify to crops other than paddy. Provinces, avd a new extension system is only now bein,? redeveloped in a 13. Perhaps as early as 1986, the more appropriete form with Bank as- supporr for maintaining the extension sie!"nce. Even during project system had weakened. Of course, Sri period a fully unified extension lanka'*s overall financial troubles service was not achieved and this and the subsequent pressures to cut issue also is being addressed in the the overall public sector did not design of the new project. help matters. However, despite these external preasures, what had also 16. Overall, despite its serious become clear was that the staff-in- deficiencies and inability to quanti- tensive extension system in place was fy physical benefits, the project is no longer cost-effective. This considered to havet a) had a posi- resulted in the transfer of KVS out tive impact on the agriculture sec- of extension to other duties in the torl b) eubanced the awareness of the Government's Poverty Alleviation Governent's relevant agencies of Program (para. 8), where it continued having a cost effective extension to cost the Government an equal system designed to meet the real amount of funds except under a dif- demands of the sector; and c) provid- ferent guise. With these changes the ed a valuable experience for redesign austainability of AEARP became moot. of the system. Findings and Lessons 17. The project experience provides a number of lessons: 14. On the positive side, AEARP was able to establish an extension system First, and most important, was for field crops that was r'e fo- the shortage of operating funds cussed than the extension ,ervice to run the extension system that existed prior to the project. effectively. During implemen- However weak, the project also initi- tation# operating expenditures ated an adaptive research program and in real terms (excluding sala- research-extension linkagie; and ries and allowances) were only training provided through the project 32 percent of appraisal tar- was useful. The extension system gets. After project comple- established played an important role tion, staff salaries absorbed between 1979-84 in explviting the much of the extension system's gains of improved technology and in- recurrent costs, leaving only about 10-15 percent for opera- ter-off farmers. It would have tionalizing the system. This been better to have paid great- emphasizes the need to develop er attention to defining groups an extension system which is of fArmers with similar values cost-effective and which sould and concerns (a reference be sustainable under projected group) and to use the groups as budgetary allocations to the the extension target of the agricultural sector. KVS. The extension approach would be farmer-centered rather * Second, an extension stem than message-centered, a re- that was less staff-intensive duced programmed visit fre- and more geared to quality quency would be required and staff with greater analytical fewer KVS would have been able and diagnostic skills is likely to reach larger number of to have been more cost-effec- farmers. tive than the one supported by the project. Fewer, better- * Fourth, the research-exte.tsion educated, better-trained and linkage did not work effec- better-salaried staff would tively, largely because of have made the extension system insufficient dialogue between more dynamic and given it much extension staff at the field more flexibility once the de- level and resarch staff at the mands on the system had changed research stations. The RTWG after 1984 with the switch to a were hampered in providtng the "single crop" approach. An al- desired resoarch-extension ternative system would have re- linkage due to inappropriate quired a greater emphasis on deployment of SMOs, the training component, es- pecially of SMOs, and may have * Finally, the adaptive research necessitated a phased introduc- component based on on-farm tion of extonsion services into trials was weak and was not the various districts in accor- sufficiently responsive to the danc. with the availability of needs of the farmers. It could trained personnel to promote have been considerably improved and lead the improved service. if a farming system approach was adopted which would con- * Third, the CF role as the main sider farmer profitability in agent of technology transfer using specific technology, re- was quite limited. CF in most source situation, availability cases were not typical repre- of complementary inputs and sentatives of the areas' farm- services and associated pro- ers and tended to be the bet- duction and marketing risks. PROJECT PERFORMANC._ADWIT_bRBE SRILAA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (CREDIT 931-CE) I. INRODUCTION 1.1 This is the audit of the Agricultural Extension and Adaptive Research Project (AEARP) in Sri Lanka, the Bank's first and only investment on a national scale in the Agricultural Extension sub-sector in the country. The audit eveluates the project in terms of its stated objectives of establishing the Training and Visit (T&V) system of agricultural extension nationally, strengthening adaptive research and improving and expanding training of extension officers. 1.2 The Bank to date has approved credits for projects amounting to US$549 million for the agriculture sector. These include five Mahaweli Ganga Dtevlopment projects (including a Teckh.-ical Assistant project), five other irrigation projects, three rural development projects, five treecrop projects (tea and rubber), two dairy developuint projects, one forestry project, one agricultural development project (providing foreign exchange for purchase of imports) and an agricultural research project. 1.3 Sri Lanka is an island country east of the southern tip of India. Its land area is 6.5 million ha. and its population was 17 million in 1991. Cultivated land per rural family is quite small in some parts of the count-ry since only a third of the land area is under cultivation. Agriculture plays an important role in the country's economy -- in 1990 it accounted for about 24 percent of GDP, 43 percent of employment and 32 percent of exports. Tea is the important export crop followed by rubber and coconut products. In recent years agriculture's importance to the economy has declined due to its relative poor performance. At the same time the urban-based sectors of the economy have registered higher growth rates. 1.4 The country is divided into three agro-ecological zones -- the dry zone, where annual rainfall averages between 600-900 am; the intermediate zone, where annual rainfall averages between 1,200-1,500 am; and the vet zone, where annual rainfall averages over 2,000 m. The vet zone covers about 30 percent of the cultivated area and 70 percent of the rural population live there. Rainfall over the island follows a bimodal pattern, the north-east monsoon prevails from November through February (the main season, called the "Maha" season) and the south-vest monsoon prevails from May through September (the off-season, called the "Yala" season). Variety of soils and vegetation, and temperature changes with elevation and topography add to the varied ecological conditions. 1.5 About 73 percent of agricultural land is under smallholdinga, but there are also a large number of plantations of tea, rubber and coconut. The average size of small holdings is about 0.8 has 50 percent of the smallholders operate 85 percent of the cultivable land, 39 percent have home gardens, and about 11 percent are landless. Many of the smallholders have lowland plots where -2- rice is usually grown and upland plots where subsidiary food crops and treecrops are grown. Treecrops (40 percent) and paddy (33 percent) account for major areas of permanently cultivated land. Other crops include minor export crops such as cinnamon, cocoa, coffee, cardamon, cloves, ginger, pepper etc. and also subsidiary food crops such as maize, groundnuts, pulses, chillies, onions etc. 1.6 Before 1977, Government intervention in the economy was extensive and Sri Lanka pursued inward looking economic policies. Despite considerable achievements in the social sectors in the 1960s and 70s, overall economic growth was slow. This led the elected Government in 1978 to liberalize the economy and undertake major new investment programs. In particular, the agricultural strategy after 1978 favored greater liberalization and strengthening of essential services. Agricultural growth rate picked up between 1978 and 1986. However, between 1987 and 1991 agriculture performed poorly. The negative growth rate of agricultural GDP in the past four years is explained by the disruptions daused by ethnic conflict in the North and East, civil disturbances in the South and unfavorable weather conditions and perceived climatic changes. 1.7 The new Government which came to office in early 1989 was faced with a serious financial crisis. The Government, with assistance from IDA and the IMF initiated an adjustment program in September 1989. The aim was to stabilize the economy and address key constraints to growth by reducing the fiscal deficit and by strengthening the balance of payments. In the agricultural sector, this meant eliminating major subsidies on wheat, flour, rice and fertilizer, rationalizing welfare vrograms, fraezing public sector hiring and reducing recurrent budgetary expenditras. The Government also sought to emphasize a diversification away froam .y the to increasing production of higher-value export crops. II. PROJECT EVOLUTION, OBJECTIVES AND DESIGN rS?_Lf,tEvolution 2. Government involvement in agricultural extension in Sri Lanka existed to Linited extrit long before the country's independence. The Department of Agriculture (D!.) was set up in 1921 absorbing the staff of the Ceylon Agricultural Society which had until then attempted to reach smaller farmers through wealthier land owners. Food shortages developed during the early 1920s. This led to increased importance of DOA and resulted in redefining its functions into research, extension and education. By 1938 there were six agricultural Divisions headed by Agricultural Officers and Instructors who served as field level extension workers with direct contact with farmers.11 More food shortages in the 1940s led the Government to revise its agricultural development policies. Agricultural Int :ructors were placed under the Assistant Government Agents (AGA) of the districto and became directly involved in food production activities. After the country's independence in 1948, rice production took on increasing importance which led the Government in the major colonization schemes to S. Wirasinghes Agricultural Extensicn in Sri lanka. Keynote paper at a Seminar held by the Sri Lanka Association for the Advancement of Science, October 22, 1989. -3- establish research facilities, seed production and demonstration sites. The Agricultural Instructors (AI) began to function as field extension staff. 2.2 In 1952, a Department of Food Production (DFP) was established in the Ministry of Agriculture. This resulted in the DOA playing a less important role. However, in 1957 the DFP was abolished and the staff absorbed into the DOA. About 500 such staff were absorbed into DOA. They were called Food Production Overseers (FPO). They did not have any agricultural background. The FPO was later renamed as Krushikarma Vyaptha Sevaka (KVS), the village level extension worker. Thus the new cadre of extension agents below the Al became the interface between the government's extension service and rural households. For the first time the Government established direct contact with farmers at the village level. In 1963 two important changes took place. The first was the creation of a Division of Agricultural Extension within the DOA; and the second was with regard to the extension system which had (thus far emphasized only rice) slowly began to include other crops such as vegetables, chillies and onions. A few other changes took place prior to AEARP. In 1967, with the rehabilitation of some major irrigation schemes, intensive extension efforts resulted in substantial diversification away from rice. And in 1970, some organizational changes took place in agricultural extension through the creation of Agriculture Service Centers that included support for production, fertilizer and other services. 2.3 In sum, at the time of AEARP preparation in 1977 the management of agricultural extension was diverse and uncoordinated. The extension Division of DOA was the main agency involved with extension with a ,taff strength of about 2,300 field officers at the district, division and village levels. However, under the same parent Ministry (Ministry of Agricultural Development and Research) the Department of Minor Export Crops provided support for spices, tea and coffee. Also, the Mahaveli Development Authority had an arrangement in which Project Extension Officers reported directly to the headquarters of the Extension Division of the DOA rather than to the District Agricultural Extension Officers. Livestock had its own extension system; and coconut also had its own extension systems along with responsibility for providing credit, input supplies and subsidies. Thus there was a wide variety in the quality and delivery system of extension by type of crop; and the existing system for delivering messages to small farmers was clearly inefficient. 2.4 During the 1970s, the Bank was actively promoting the T&V system of extension in South Asia to develop the agricultural potential of rainfed areas'. It was introduced in neighboring India in the early 1970s, first through three Bank-assisted irrigation projects in Rajasthan, Madhya Pradesh and West Bengal; the T&V extension approach was then repeated for rainfed areas!/ statewide in the various states of India. The design of the AEARP was almost 31 T&V is a hierarchically organized and time bound method of managing extension, designed to deliver selected and timely technology to farmers with strict regularity. The system encourages links with agricultural research institutions, emphasizes a professional approach to extension, and requires an exclusive devotion to extension work. In principle, all extraneous activities such as those related to the physical handling of inputs and credit applications are avoided. See OED's Performance Audit of six extension and research projects in India (Report No. 8808, June 29, 1990). -4- identical to the earlier statewide T&V extension projects in India that also included an adaptive research component. As in India, T&V extension was first introduced in Sri Lanka on a pilot scale as a component of the Tank Irrigation and Modernization Project (CR. 666-CE) financed by IDA in 1976 in Anarrdhapura district. The initial positive impressions of T&V extension on irrigated agriculture led to its replication in the rainfed areas. Obiectives 2.5 The project vas prepared by a FAO/UNDP funded project entitled "Reorganization of Research, Extension and Training Resources for Agriculture" for GOSL. It was initiated in January 1977, and after completion GOSL submitted it to IDA in June 1978. 2.6 The objective of the project was to assist Sri Lanka to apply the T&V system of extension to the entire country, to strengthen adaptive research and to improve and expand the training of extension staff. The project would include the following components: a) strengthen extension services by the establishment of a unified T&V system for all crops (except tea and rubber), and animal husbandry for the entire country; b) establish an adaptive research network on the basis of the country's agro-ecological regions; c) improve and expand agricultural training institutions of DOA1 d) strengthen the Economic Unit of DOA; e) strengthen the building construction capability of DOA; f) provide funds for salaries and allowances for a total of 1,789 incremental extension, adaptive research, training, economics and civil engineering staff; g) construct 1,203 living quarters for incremental staff and 120 buildings for training quarters, offices, laboratories and workshops; h) provide appropriate vehicles (jeeps, cars, buses, motorcycles and bicycles) for incremental staff, and equipment for farms, laboratories, workshops offices, extension and training; i) provide incremental operating and maintenance costs for project works, buildings, vehicles and equipment; j) provide overseas training of project staff from DOA, DAPH and CCB, including six-month training for 3 Senior Training Officers, 28 Research Officers and Subject Matter Officers (SMO); short term visits to India of 120 village level Extension Officers (KVS), 48 Agricultural Instructors (AI), 24 SM0s and 24 Agricultural Officers and; k) provide 138 man-months of consultancy to assist in implementing the project. Design 2.7 The design of AEARP was such that it supported establishing a "top- down" system through an organized chain of command within the Government bureaucracy. The project focussed on the management aspects of extension rather than systemic issues affecting agricultural operations at the farm level. The project did not focus on availability, profitability and acceptance of specific technical packages; it did not consider the ability of KVS to correctly identify farmer constraints and as such the adaptive research-message formulation system was weak. Some implications of project design are discussed below. 2.8 First, the design of AEARP was a break from past extension services in Sri Lanka in that the T&V extension system sought the full-time commitment of the KVS, devolving him of non-agricultural and non-extension responsibilities. The organization of a single line of command between full-time KVS and extension headquarters in a unified extension system for all crops and animal husbandry for the entire country would involve technical and administrative supervision by the extension headquarters of DOA. From the beginning, GOSL had some difficulty with this organizational set-up. During negotiations, the Sri Lankan team wanted to entrust the responsibility for extension work in the Mahaveli project area to the Mahaweli Development Authority by having DOA extension staff seconded to the Mahaveli Authority, while DOA would provide technical guidance. The Bank argued that such an arrangement would cut across the concept of the T&V extension system and would not be desirable. In the end, however, even though the Sri Lankan negotiation team agreed to the.Bank's proposal, project files suggest that there was considerable opposition to separating extension from input supply within the Ministry of Agricultural Development and Research (MARD). 2.9 Second, during internt1 review of the project it was pointed out that the project would incur budgetary liabilities on the Government and an assessment of such liabilities ought to be reflected in the SAR. In response, a paragraph .was added in the SAR (SAR, para. 4.10) which stated that incremental operating costs at full development would amount to 30 percent of DOA's current annual operating budget. This proved to be quite inaccurate -- an issue that is discussed later. 2.10 Third, internal review also expressed concern that the project would not yield expected benefits unless input supply was simultaneously ensured. These included fertilizer distribution, pricing adequacy of improved seed supply and water management. 2.11 Fourth, the staffing level proposed in the SAR was excessive. It assumed that extension would service all farm families except tea and rubber and those with only a house lot -- about 1.4 million farm families. The 1982 Sri Lanka Census revealed that usino the criteria proposed in the SAR would imply that 0.9 million farm families ought to have been the main extension target. -6- Thus using a rate of one extension worker for 750 farmers suggests that the staffing level proposed in the project was higher than necessary. In addition, the project assumed that staff from the other agencies could be easily transferred and trained in extension matters. Also, it underestimated the DOA's difficulty in establishing new agencies to take over programs such as seed distribution and other activities previously undertaken by the KVS. 2.12 Fifth, it assumed that the contact farmer (CF) approach would result in a rapid diffusion of technology throughout the project period and the research-extension linkage would operate smoothly. It was expected that new technology would be passed on to the farmers relatively easily and that feedback from farmers could indeed be forthcoming. III. IMPLEMENTATION Implementation Experience 3.1 The project's physical implementation is adequately described in the PCR 1'. The achievements and shortfalls of the physical implementation targets are shown in Annex 1, Tables 1-4. The implementation section of this PPAR summarizes the project's physical implementation and then updates some recent developments related to and affecting agricultural extension following AEARP's implementation. 3.2 The Credit Agreement was signed on August 24, 1979 and the credit became effective on October 4, 1979. The credit was closed on June 30, 1986, a year after the original closing date. US$7.96 million of the original Credit amount of US$15.5 million was disbursed and the balance of US$7.54 million was cancelled. Project costs and financing at appraisal and completion are shown in Table 1. Report No. 7684, March 31, 1989 (paras. 3.01-3.30) -7- TABLE I: Comparison of Appraisal and Actual Prolect Cost and Financing A. PROJECT COSTS ACTUAL AS 2 ACTUAL AS 2 APPRAISAL ACTUAL OF APPRAISAL APPRAISAL ACTUAL OF APPRAISAL ----------- (US$ million) -------- ----------- (Re million) --------- Buildings 5.4 4.0 74 86.1 95.0 110 Vehicles 2.8 1.8 64 44.3 42.2 95 Equipment 1.2 1.2 100 18.6 27.3 147 TA and Overseas Training 1.6 1.2 75 24.8 27.9 113 Salaries and Allowances 4.2 3.3 79 65.6 77.9 119 Operating Costs 2.2 0.7 32 34.5 17.4 50 Physical Contingencies 1.7 --- --- 27.4 --- --- Price Contingencies 3.3 --- --- 53.2 --- --- TOTAL 22.4 12.2 54 354.5 287.9 81 B. FINANCING IDA 15.5 7.96 Government 6.9 4.24 TOTAL 22.4 12.2 3.S Actual project costs were only 54 percent of SAR estimates in US dollar terms and 81 percent in local currency terms. This difference was due to a local currency devaluation of 80 percent relative to the US dollar between 1979 (project approval) and 1986 (project closure). Although overall costs were 81 percent of SAR estimate in local currency, if base costs only are considered, it appears that local currency costs of all project expenditure categories (except operating costs which were financed by GOSL) equalled or exceeded SAR estimates despite the substantial scaling down of the project. Expenditures for operating costs were cut substantially. 3.4 Not only was the Bank's actual disbursement for the loan about half of what was originally approved, GOSL's contribution was also less than planned (about 61 percent), as shown in Table I. Operating expenditures was the component most seriously affected. Three years into the project the Treasury imposed restrictions on allowances for travel and vehicle operating costs and these were set not by needs, but rather by grade - vels of the officers and extension staff. This meant that staff who needed to travel the most, such as the KVS, obtained little funds to do so. According to the PCR (PCR, para. 3.24), actual amounts allocated for travel allowance and vehicle operating costs were one-third of that required for efficient implementation of the extension program. Thus, the incremental operating costs of making the extension system fully operational was grossly underestimated in the SAR and there appears to have been little constituency outside the DOA for supporting such an expensive system. -8- 3.5 The civil works component was drastically reduce. Only about 43 percent of the estimated number of units were constructed (Annex 1, Table 1A). This was largely due to start-up delays, design changes, poor supervision, and most of all, a sharp increase in construction costs. 3.6 Procurement of goods was also delayed due to DOA's lack of familiarity with inte%national competitive bidding (ICB) procedures and lack of clear understanding& as to what items were to be purchased. In the end though, most of the equipment and machinery were bought (Annex 1, Table 1B), but shortfalls in the purchase of jeeps and vans adversely affected the adaptive research program since mobile units could not be formed and supervision was lacking. 3.7 At the time of appraisal it was expected that 1,789 incremental staff at all levels would be required to successfully implement the extension and adaptive research program. Staff were to be hired by DOA, but the positions had to be first approved by the Treasury. The Treasury approved 1,568 positions, and at the time of project closure 1,549 positions had been filled (Annex 1, Table 2). Staff shortages were most serious at the KVS, AI, AO and SHO levels, and for technical staff assigned to adaptive research activities. 3.8 The project was subjected to a member of other changes during implementation. Of note: i) extension for livestock development was excluded from the project activities and remained with the Department of Animal Production and Health (DAPH) after a decision by the Cabinet and the Ministry of Rural Industries Development (MRID); ii) the 138 man-months of expatriate consultancies i. the fields of extension, animal husbandry, water management and crops were reduced to 50 man-months since DOA expressed serious reservations and was unwilling to hire such a large number of consultants (Annex 1, Table 3). 3.9 The Monitoring Survey data reported in Annex 1, Table 5 for the 1985/86 Maha season is quite revealing. It shows that 77 percent of the CF fully or partially adopted the technical recommendations; and only 47 percent of other farmers fully or partially adopted such recommendations. The CF tended to receive information and technical recommendations much more easily than other farmers. This is not surprising. Experience in other South Asian countries suggest that the CF have tended to be more progressive farmers who seek out information more freely and attempt to remove existing constraints and adopt appropriate new practices more easily than do the less innovative farmers. Other data provided by the Monitoring Sutvey was scanty. Unfortunately, the survey data does not reveal much information about non-adoption, specifically whether non-adoption was due to lack of knowledge of a specific technology or due to other reasons. However, the audit mission's interviews of many farmers in Kurunegala district suggest that other reasons such as financial factors, unavailability of seeds, fertilizer and credit, and water availability often affect the decision for non-adoption. Post Imnlementation: Recent Developments Affecting Agricultural Extension 3.10 Since the PCR was written in early 1989, some changes have taken place at the macroeconomic and national administrative levels that have affected the extension system set up by the project. -9- 3.11 First, was the Government' a worsening fiscal crunch already felt at the time of writing the PCR. A financial crisis confronted the new Government soon after it was formed in early 1989. Among other problems the fiscal deficit became so large that the Government was prompted to embark on a major economic restructuring in mid-1989. One of the important goals of the stabilization program that followed was to cut the fiscal deficit. This translated for agriculture into eliminating some important subsidies such as on wheat, rice and fertilizer; also undertaken was structural reform with emphasis on public enterprise and public expenditure reform which translated into plans for cutting the civil service by 20 percent during the next 4-5 years with proportional effects on agricultural expenditures, especially current expenditures. 3.12 However, even before the crisis developed fully, and soon after AEARP's completion in 1986, the Government became unwilling to support the required level of recurrent exnenditures. The costs of operating the extension system became prohibitively high for the Government, especially in view of stagnating benefits. Paddy yields had reached a plateau by 1986 and the Government began to seriously question whether the extension system was cost- effective any more. At that time, although staff was not reduced or relocated, expenditures for operating the system such as travel allowances, vehicle replacement, repairs, maintenance etc. were drastically cut. Staff salaries accounted for about 85-90 percent of the cost of operating the system according to estimates of DOA staff. This meant that the extension system that was set up by the project was no longer effective, as one of the key ingredients of the extension system -- staff mobility -- was reduced to a minimum. 3.13 Second, although an even more serious set-back to the extension system came in late 1988 as a result of the Thirteenth Amendment of the Constitution which involved devolution of administrative authority to the Provinces, this change has been a blessing in disguise in helping to reshape a new cost-effective extension system. The devolution included agriculture and agrarian services and involved a shift from the old concept of the district as the administrative focal point. In addition to the responsibilities of the Center, two levelt were created at the Provincial and Divisional levels. At the Provincial level, the essential responsibilities are formulating policies, strategies, coordination and monitoring for the Province. At the Divisional level, the emphasis is on implementing developmental activities. The Divisional Secretary (DS) is also the Assistant Government Agent (AGA) and performs two functions. When dealing with devolved functions he acts as the DS, under the Chief Secretary of the Provincel and when dealing with non-devolved activities he acts as the AGA. Agricultural extension is a devolved function under the office of the PraAincial Director of Agriculture in most of the Provinces. 3.14 Given agricultural extension's devolved functions, the entire 2,400 EVS cadre was transferred from the agricultural program to the Ministry of Home Affairs' Poverty Alleviation Program, called the Janasaviya Program (JSP). Under the JSP, the KVS joined the other village level all-purpose workers known as the Grama Niladhari (GN). The GN, including the KVS, and the AI have been placed under the Divisional Secretary who is in charge of all Government agencies at the Divisional level. At present, there are about 14,550 GNs who collect data and perform regulatory functions at the local administrational level. Under this system, the ex-KVS act in their new roles and cannot undertake any extension - 10 - work; the linkage with DO&, including technical linkage has more or lose dissipated. An Agricultural Support Services Project (ASSP) appraised by the Bank in July 1991 is expected to resurrect an extension system based on the lessons learnt from AEARP. IV. PROJECT OUTCOME Agricultural Impact 4.1 It is not surprising that a quantitative assessment of the direct impact of the project is not available in the PCR. The agricultural impact of AEARP, like similar extension projects in other South Asian countries, is hard to measure because it is difficult to separate out the contribution of extension and adaptive research components to the overall increase in yields and production. Many other complementary inputs also affect productivity, sometimes as initiating and catalytic factors and sometimes as additive and facilitating factors. 4.2 Paddy yields began to increase significantly even before the introduction of T&V in Sri lanka. Paddy yields increased from about 2.3 tons/ha in 1975 (maha season) to 2.9 tons/ha in 1979/80 (Maha season). After T&V extension was introduced in 1980, yields kept on increasing, reaching a high of about 3.4 tons/ha in the 1984/85 Maha season, most paddy cultivation having switched to high yielding varieties (Annex 2, Table 1). After the 1984/85 season, there has been no increase in yields, although the T&V system was in full operational force between then and 1989. Yields have been constant and production actually dropped in two of the three years since 1986. Some of this may have been due to lack of seasonal rainfall and water availability during the past four or five years and some due zo the disruptions caused by the ethnic conflict in the North and East and civiJ disturbances in the South. However, the general trend has been one of unsustained improvement in paddy yields. 4.3 The above are important considerations for assessing how effective T&V extension has been in increasing paddy yields in Sri Lanka. The general increase in paddy yields and production in the country between 1975 and 1984 are attributabla to a number of complimentary factors such as the availability of high yielding paddy varieties, availability of subsidized fertilizer, improved quality seeds, new and rehabilitated irrigation facilities, expanded credit and supportive Government policies such as guaranteed price and market. It would be difficult to isolate the contribution of any single factor in this package of complementary factors. 4.4 T&V extension appears to have played an important role essentially to maintain, for a period of time, the gains of improved technology and incentives with the result that widespread adoption did take place in Sri Lanka within a relatively short period of time. The strength of T&V extension during the 1979-84 period was that it was able to focus on a few simple messages which happened to be the relevant and important ones at the time, and substantial paddy yield increases were achieved. - 11 - 4.5 However, after 1985 the very strengths of T&V extension proved to be a liability. The reservoir of simple messages were exhausted as paddy yields reached a plateau and as the system expanded to cover subsidiary food crops, minor export crops, vegetables, spices and coconuts. The T&V extension system began to breakdown as consideration of all these crops required a whole farm approach in contrast to the simple technical messages associated with the "single crop" approach of the past. The farmers required advice on a range of crops, cropping system, profitability and financial incentives. However, the established system was unable to adapt from the simple, single crop approach to the more complex demands. Institutional Impact 4.6 On the whole, the project was able to set up an extension system for field crops that was successful in achieving a more focused extension service than that which existed before the project, but the established system was not a unified system and other agricultural activities such as horticulture, minor export crops, and coconut extension did not receive enough attention and animal husbandry was excluded. 4.7 The project expected that contact farmers (CF) would be the main channel of technology dissemination. CF woulA spread the messages to other farmers and thus the target of 1.4 million farmers would be reached in a relatively short period of time. It turned out that the effectiveness of the CF on other fa2mers was quite limited. CF in most cases were not typical representative of the areas' farmers. They tended to the better-off farmers; often chosen because KVS found interaction with them somewhat easier. However, this was also the reason why there was such a gap between the CFs and other farmers in their attitudes to new technology adoption and innovation. Some results are reported in a Monitoring Survey undertaken at the time of project completion (Annex 1, Table 5). The main weakness appears to be that the KVS failed to visit the other farmers. Only a third of the follower farmers knew that the KVS should visit them once in 14 days; only 12 percent knew of the regularity of fortnightly visits; and only 9 percent knew of the KVS visit on a fixed day of the week. As a result, only 47 percent of such farmers adopted the messages fully or partially. About half of the other farmers perceived the extension system to be useful; and farmers in general appear to have met infrequently as a group. 4.8 The adaptive research component of the project based on on-farm trials was weak. The adaptive research component did not consider a farming system orientation to research and as such was not responsive to problems and needs of the farmers. It failed to consider farmer benefits and costs in using a specific technology, production and marketing risks, resource situation and access to complementary inputs and services available in the community. Extension-research linkages did not work effectively largely because of insufficient dialogue between extension staff at the field level and research station staff. The SMOs were wrongly deployed as teaching staff in the Education Training and Information Division rather than being used as the real link to transfer technology between research station staff and field extension staff. The Regional Technical Working Groups (RTWG) did not provide the linkage between - 12 - research, training and extension as expected, again due to insufficient extension staff deployed as SMOs. Sustainability 4.9 From the start, sustainability was dependent on the level of funding provided by the Government and the political and bureaucratic support for such funding. Even the potential problems associated with staff transfere were less of a key factor than the availability of funds to operationalize the extension system set up by the project and to make it sustainable. At appraisal, the project was expected to finance the capital costs and a high percentage of the recurrent costs. For example, 65 percent of incremental staff salaries were expected to be paid by the project during the implementation years. After project- implementation, the Government was expected to finance the incremental salaries from its own resources; the Government was also expected to provide for other operating expenditures to run an effective extension sstem including providing funds for travel allowances, vehicle replacement, repairs, maintenance etc. 4.10 In 1988, two years after the end of project implementation, agricultural extension was costing the Government about Re. 145 million (about US$4.4 million in 1987 US dollar). This was roughly 0.88 percent of agricultural GDP.!' Agricultural research cost the Government another Re. 91 million (about US$2.76 million) -- about 0.55 percent of agricultural GDP.!l Staff salary costs of the extension system alone was about 85-90 percent of the operating costs of the system. These staff salary costs as a proportion of the operating costs of the extension system is similar to that of some of the states in India, which are also facing operational fund constraints.- Given the perceived and actual incremental benefits of maintaining the extension system, these costs were seen to be too high in Sri Lanka. The Government became especially concerned with the high costs once paddy yields levelled off, production began to stagnate and new problems began to arise when the extension system attempted to diversify to crops other than paddy. 4.11 Perhaps as early as 1986 (at the time of project completion) the support for maintaining the extension system set up had weakened and run into serious trouble. Of course, Sri lanka's overall financial troubles and the subsequent pressures to cut the overall public sector did not help matters. But despite these external pressures, what was also becoming clear was that the current staff-intensive, face-to-face extension system was not cost-effective. The support for a relatively expensive extension system had weakened and the sustainability of the extension system had been seriously jeopardized by then. 1 S.Wirasinghe: Agricultural Extension in Sri Lanka. Key-note paper at a Seminar held by the Sri Lanka Association for the Advancement of Science, October 22, 1989. For comparison, India spends about 0.3% of its agricultural GDP on agricultural research; and developing countries as a whole spent about 0.42% of agricultural GDP on Agricultural research between 1980-85 (ISNAR: Research Expenditures as a percentage of Agricultural GDP, 1980-85). Z/ Project Performance Audit Report on Extension and Research Projects in India, Report No. 8808, June 29, 1990, (paras. 94-97). - 13 - 4.12 The final resolution came with the administrative changes of early 1989 (para. 3.13) which included the devolution of administrative responsibility of agriculture and agrarian services to the Provinces. As mentioned earlier, the KVS were transferred out of extension to the Government's Poverty Alleviation Program as all-purpose workers and have continued to cost the Government equal amount of funds except under a different guise. With these recent changes the sustainability of AEARP has become moot. V. FINDINGS AND LESSONS 5.1 The audit notes the considerable similarity of the experience of AEARP in Sri Lanka with that of the Indian Extension projects which were evaluated in a cluster audit in 1990. This is not surprising given the similarities in project design and rural cultural practices. However, the intensity of the implementation problems and post-implementation issues are somewhat different and as such the priorities and challenges are somewhat different as well. 5.2 On the positive side, AEARP was able to establish an extension system for field crops that was more focussed than the extension service that existed prior to the project. However weak, the project also initiated an adaptive research program and research-extension linkages; training provided through the project was useful. Also, when construction costs escalated sharply, the correct decision was made to cut back on the number of buildings constructed. Although the extension system established by the project was, in effect, dismantled in early 1989 with the davolution of administrative responsibility to the Provinces, the experience with AEARP provides a basis for building on past strengths and correcting previous mistakes. The extension system established played an important role between 1979-84 in exploiting the gains of improved technology and incentives favoring the agriculture sector and is likely to have speeded up the adoption of improved rice technology. 5.3 On the negative side, the system did not adapt to new extension challenges after 1984. This and its high fixed personnel costs led it to be perceived as cost-ineffective in its existing form. The extension system was eventually dismantled in the Constitutional Amendment of 1989 which devolved administrative responsibility to the Provinces, and a new system is only now being redeveloped in a more appropriate form.Y A unified extension system for all crops (excluding tea and rubber) and animal husbandry which was one of the important objectives of the project was never established. Even when the The Bank has been contemplating a follow-up to AEARP for the past five years. An Agricultural Support Services Project (ASSP) was appraised in July 1991. A major part of ASSP is to revamp agricutural extension on the basis of lessons learnt from AEARP; the project also includes compt %ent for seed industry development, pesticide management, manpower development, support for agro-enterprise investment and technical assistance. The extension component appears to address some of the important shortcomings of AEAFP. Of note is the support for a holistic approach based on the establishment of an integrated extension service incorporating the Department of Export Agriculture (DEA), CCB, DOA and DAPH, fewer and more qualified extension staff, attention to staff and operating cost questions, dissemination through groups rather than CF and a stronger link between extension and research. - 14 - extension system was fully operational, the DAPH maintained its own extension service for animal husbandry; extension service for coconut and minor export crops was included only in the last two years of the project but was not effective. This issue is being addressed in the design of the new project. 5.4 Overall, despite its sarious deficiencies and inability to quantify physical benefits, the project is considered to have: a) had a positive impact on the agriculture sector; b) enhanced the awareness of the Government's relevant agencies of having a cost-effective extension system designed to meet the real demands of the sector; and c) provided a valuable experience for redesign of the system. 5.5 The project experience provides a number of lessons: i) The shortage of operating funds to run the extension system effectively was a critical problem. Even during implementation, operating expenditures (excluding salaries and allowances) were only 50 percent of appraisal targets in local currency terms (32 percent in LS dollar terms). After project implementation, staff salaries absorbed much of the extension system's recurrent costs, leaving only about 10-15 percent of the operating budget for operationalizing the system. As a result, the mobility of extension workers was drastically reduced and the extension system was rendered largely ineffective. It would be tempting to suggest that the government budget and allocate funds for recurrent costs, but perhaps that is not the reality given the many demands on its limited resources. Availability of funds for recurrent costs depends on political and bureaucratic support and since resources are often fungible between various expenditure items it would have been desirable to have established an extension system which was less staff-intensive and able to allocate a greater proportion of funds to operate the system. The SAR's assessment of incremental operating funds required (para. 2.9) was unrealistic and off-base. This emphasizes the need to develop an extension system which is cost-effective and which should be sustainable under projected budgetary allocations to the agricultural sector. ii) An extension system that was less staff-intensive and more geared to quality staff with greater analytical and diagnostic skills is likely have been more cost-effective. Fewer, better-educated, better-trained and better-salaried staff would have made the extension system more dynamic and given it much more flexibility. When the demands on the system had changed after 1984 with the shift away from a "single crop" approach to include subsidiary food crops, minor export crops and animal husbandry, a leaner, better qualified and trained staff would have taken on the challenges more effectively. The implication is that a better-educated and better-trained extension worker wou.d not be simply "parroting" the simple technical messages but responding to the needs of the farmers - 15 - on the basis of a whole farm approach (including livestock), weighing options and even advising on economic and financial aspects. This alternative system would have required a greater emphasis on the training component, especially of SM0s, and may have necessitated a phased introduction into the various districts in accordance with the availability of trained personnel to promote and lead the improved service. iii) The CF's role as the main agent of technology transfer was quite limited. It would have been better to use groups of farmers to ei.Aeminate technology and use them as extension targets cf tha VS. It is important that the groups are not mere additive collection of farmers as was unsuccessfully tried in some p-oject areas, but rather, groups of farmers with aimilar values, resources and concerns (a reference group). Itteraction with a reference group facilitates an understan(iing by the extension staff of the real problems, constrauint and potentials faced by farmers, promotes group "ownership7 of mini trials/demonstrations, and enables consensue o.t appropriateness of measures designed to attain -objectix.- all of these factors are conducive to an effective transfer oA appropriate technology. The extension approach would be farmer-centered rather than message-centered, a reducee rcogrammed visit frequency would be required and fewer KVS wor1d be able to reach larger number of farmers. iv) -ho zesearch-extenaion linkage did not work effectively, largaly because of insufficient dialogue betwaen extension staff at the field level and research staff at the research stations. Also, inappropriate deployment of SM0s in the Education, Training and Information Division rather than in the RTWG adversely affected the linkage between research and extension. v) The adaptive research component, which was not sufficiently responsive to the needs of the farmers, could have been considerably improved if a farming system approach was adopted which would consider farmer profitability in using specific technology, resource situation, availability of complementary inputs and services and associated production and marketing riaize. - 17 - Annex 1 - Table 1 PROJECT PERFORM NCE AUDIT REPORT SRI LANKA AGRICULTURAL EXTENSION AND AI)APTIVE RESEARCH PROJECT (Credit 931-CE) A. Civil Works Appraival estimates Achievements ------------ No. of Buildings---------- Extension 40 27 Education & training 23 16 Adaptive research 27 30 Economics, monitoring & education office --- 1 Staff quarters 1.035 4081/ 1,125 482 B. Procurement of Vehicles Appraisal Estimates Actuals ------------- Units ---------------------- 5 9 Jeep/van 169 115 Audio visual unit 17 1 Minibus 8 2 Motorcycle 461 410 Bicycle 1,530 1,710 Sources PCR ' Includes 30 staff quarters purchased In Kandy (Hantana Complex). - 18 - Annex 1 - Table 2 PROJECT PERFORMANCE AUDIT REPORT SRI LANKA AGRICULTURAL EXTENSION AND ADAPTVE RESEARCH PROJECT (Credit 931-CE) Total Proiect Staff Approved by Actual as Post SAR Treasury Actual % of SAR Agricultural Officers 71 66 66 93 Subject Matter Specialists 39 28 28 72 Agricultural Instructors 275 267 257 93 Krishikarma Viyaptha Seveka (KVS) 1,111 923 923 83 Research Officer 40 35 36 90 Research Assistant -- 38 38 -- Laboratory Assistant 30 10 10 33 Economist 7 7 7 100 Economic Assistants 26 17 17 65 Computer Programmer 1 1 -- 0 Programmer Assistant -- 4 4 -- Agricultural Engineer 3 4 4 133 Farm Mechanical Inspectors 11 8 6 55 Assistant Engineer 2 2 -- 0 Work Inspectors 5 5 5 100 Electronic Foreman 1 1 -- 0 Roneo Operator 4 1 1 25 Artist 8 7 7 66 Cinema Operators 5 3 4 80 Photographer 1 2 2 200 Draughtsman 1 2 2 200 Accountant -- 1 1 -- Clerk/Typist 61 72 71 116 Driver 17 18 18 106 Labor (Stores) 4 4 -- 0 Total 1,789 1,568 1,549 87 Source: PCR -19- Annex 1 - Table 3 PROJECT PERFORMANCE AUDIT REPORT SRI LANKA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (Credit 931-CE) Technical Assistance Apraisal Estimate Actuals Specialization man-months man-months Team Leader: Management 24 -- Animal husbandry 18 -- Coconut intercropping 18 -- Goose grain 12 -- Grain legume 12 -- Root/tubes crops 12 -- Vegetable 12 -- Water Management 15 -- Plant Protection 15 6 Extension Management -- 2 Training -- 6 Adaptive research -- 2 Agronomy -- 2 Farm Economics *-- 12 Agricultural Engineering -- 12 (tools and equipment) Monitoring and evaluation -- 6 Total 138 50 Source: PCR -20- Annex 1 - Table 4 PROJECT PERPORMANCE AUDIT REPORT SRI LANKA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (Credit 931-CE) Overseas Trainin No. of Person Study Tours SAR Actual KVS 120 72 Al 48 68 AO 24 11 SMS 24 13 Economit -- 6 Sub-total 216 170 Short-term Trainins (6-month)'1 Training Officer 3 14 Research Officer (Agricultural Officer) 14 14 SMS 14 10 Economist -- 6 Accountant 4 1 Sub-total 31 45 Long-term Training (12-month) Extension -- 8 Research -- 3 Economic -- 2 Education & Training -- 3 Sub-total -- 16 TOTAL 247 231 Source: PCR Five places on Animal Husbandry at CIAT and four places for Coconut Production at 180 Vere dropped. -21- Annex 1. Table 5 PROJECT PERFORMANCE AUDIT REPORT SRI LANKA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (Credit 931-CE) Indicators of the Monitoring Survey on T & V Extension System (Maha 1985/86) 1. Preception of Role & Designation CFs who knew their role 76% FFe who knew their role 25% CFs who knew themselves as CFs 91% FFs who knew their CFs as CFs 56% CF FF ----- ----- 2. Awareness of Visiting Schedule Knew KVS should visit once in 14 days 75 33 Regularity of the fornightly visits 59 12 Visiting of a fixed day of the week 58 9 3. Adoption of Information (a) Fully 52 28 (b) Partially 15 19 (c) Not adopted 33 53 100 100 4. Perceived usefulness of the Extension Service Reported useful 77 51 Reported not useful (includes % who did not respond) 23 49 100 100 5. Farmers meeting as a Group CF9 meeting as a group 45% FFs meeting as a group 16% 6. Information flow via CFs CFe said to have discussed technical messages with FFs 65% FFs said to have dicussed same with their FFs 41% Source: PCR -22- Annex 2. Table 1 PROJECT PERFORMANCE AUDIT REPORT SRI LANKA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (Credit 931-CE) Paddy Yields, 1975-1989 (ton per hectare) 3.7 3.6 3.5 3.4 3.3- 3.2 3.1 3 2.9 2.9 2.7 2.6 2.5 2.4 2.3 2.2 195 1957 19 9 1981 1983 19 497 989 YEAR /a/ Harvested yields Sourcet Department of Censup and Statistics. SRI LANKA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (C~11 1.-CE) PADDY CULTIVATED AREA AND PRODUCTON: RICE AVA' 'BILTY, PROCUREMENT AND DISTRIBUTION, 1975-89 1975 1978 1979 190 1981 1ge2 1983 1984 1985 1986 1987 1988 a/ 1989 b/ CMLivated Arna ('000 hectron) Fatent A~edmied,/ 22 6 64 59 684 687 699 703 706 717 725 728 779 ofib;c e.# r@ation sch~m 217s l1120424) (8)(12 38 l3 mfaic: inor IrrigatioA 5c~. (283 175 (1> <177) {~< r18 I) (8) (8) 15 11 12 (99 Groms Are Somn 696 870 639 845 877 848 828 990 862 895 781 68 727 mat Ame. Harvemted ('00 ha) mg 724 697 728 740 681 689 787 768 740 297 725 812 maha 302 471 494 498 501 424 495 451 498 489 300 443 392 Ya 207 253 208 232 239 237 194 8a6 270 271 217 262 220 Paddm Production and Yield. Paddy Production <'000 ton) 1.154 1.891 1,917 2,133 2,280 2.156 2,484 2,420 2.661 2.588 2.128 2.477 2.063 %,ha 719 1,286 1,393 1.488 1,52 1,383 1,786 1.380 1,751 1.686 1.393 1.525 1.342 Tol 435 605 524 680 707 793 698 1.060 910 900 735 982 721 Averåge Yield (Kg per hectaro harvestod) 2.270 2.513 2.750 2.927 3.014 3.260 3.606 3.076 3.464 3.500 3,564 3.413 3.374 Rico Availabilit& ('000 tonö) wie* Produchumn 806 1,286 1,04 1,450 1.516 1,466 1,689 1,641 1.809 1.760 1.447 1,684 1,408 Nat Domo&ic s~1% 711 1.132 1,148 1,276 1,34 1.290 1,486 1.448 1.592 1,549 1.278 1,482 1,285 uponing stock da.C6 178 116 129 68 105 118 97 14 75 18 75 77 Dogning Soc arPM 2/ 49 188 128 65 38 82 84 98 43 29 sa 6 38 Ric. Importg h/ 457 160 212 190 157 160 128 26 182 220 99 184 288 Total Con 1pi/ 1.134 1.320 1.420 1,394 1.521 1,487 1.560 1,611 1.722 1.658 1.506 1.637 1,504 Procur-mant and Digtribution Rico Procured by PM ('000 tons> 169 469 878 142 85 87 220 118 69 105 43 71 3 wice Digtributod by FC ('000 tona) 629 671 447 327 159 161 140 18 180 174 245 215 276 ffmomrndum Itom (as a percentage) Importo/Rice Con uion 40.8 12.1 14.9 13.6 10.3 10.8 8.0 1.6 10.8 13.3 6.6 11.2 18.9 1Zpeto/Rico Distri neen by FC 72.7 28.8 47.7 58.1 98.7 99.4 87.9 15.5 101.1 126.4 48.0 85.6 103.3 Procurmant/Pre.dtion 21.5 36.7 29.0 9.8 5.6 8.9 13.0 7.0 3.8 6.0 3.4 4.3 0.2 FC Oialbribution/Ric Con uption 55.5 50.8 31.5 28.4 10.4 10.8 9.0 10.4 10.4 10.5 16.3 13.1 18.4 ----------------------------------------------------------------------------------------------------------------------------------------------- Provigional Lovoled and bunded land Guitable for paddy cultivation. ad on converaion factor of 1 motric ton of paddy - 0.68 mtric ton* of rico. Equalm rico production loe aed and vastag., eatiatad at 125 of total production. FC - oCamiaalonar PMB . ed, maroUng oa.,d. Cha ot on emt o n de 0ot rice oupply plu importa, adjut for *lach, pluo chang* in the *tocka hold by FC and Pnd. Noto: Thor* bad not been any rico importo in 1984, bu% a ohipmen% of 26,000 HT intended In 1981 arrived In 1984. Sourca: Food Conelseloner; Agricultural Implementation Program; tho Central Bank of SrI Lanka; and World Bank Country Mport on Sri Lanka. Report No. 8951-CE ~Stambor 26. 1990> 1) SRI LANKA AGRICULTURAL EXTENSION AND ADAPTIVE RESEARCH PROJECT (Credit 931-CB) CULTIVATED AREA AND PRODUCTION OF SUBSIDIARY FOOD CROPS, 1975-89 -------------- - --- -------------------- ----------- 1975 1976 1979 1980 1981 1982 1968 1984 1965 1986 1987 1988 1969 Cultivated Area ("000 hectares) Manloc (cassava) 79.2 74.3 58.6 51.0 66.1 69.1 565.8 56.8 62.9 49.3 48.6 50.0 45.6 Maize 40.1 24.7 19.4 19.4 24.0 26.6 28.2 82.8 83.1 81.2 84.7 86.8 29.1 Chillies 88.0 50.2 85.9 88.8 40.9 87.1 84.7 29.4 80.6 35.7 25.1 27.1 24.4 Red Onions 6.8 8.2 9.0 8.7 8.7 9.0 9.6 8.9 5.6 6.6 6.9 7.5 9.1 Groundnuts 7.6 8.8 9.9 12.0 14.2 14.6 14.1 10.6 10.2 10.1 7.8 10.6 11.2 Green Gram 9.8 12.1 12.6 18.4 15.5 17.9 19.8 22.8 28.6 25.0 25.6 28.5 27.1 Sorghum 8.6 2.0 0.7 0.5 0.5 0.8 0.4 0.6 0.5 0.8 0.1 0.1 0.0 Soyabean 1.1 - - - - - - 5.2 4.0 4.0 8.5 1.6 1.6 Potatoes 2.0 2.8 4.1 4.6 5.8 6.1 6.8 6.0 7.1 7.8 7.0 7.2 7.0 Sesame (gingelly) 12.7 * - 80.2 26.0 82.1 24.8 14.4 14.6 14.8 12.6 12.6 8.8 Cowpea 6.7 18.4 13.9 22., 26.7 81.8 82.6 28.0 29.8 27.7 27.8 24.2 22.8 Black Oram 2.0 - - - - - - 6.6 9.9 7.8 9.9 10.6 7.8 Ohal 0.8 0.7 1.4 0.8 0.5 0.5 0.8 0.3 0.1 0.0 0.1 0.2 0.0 Sweet Potatoes 21.2 20.6 16.8 14.8 16.6 16.2 14.4 18.4 14.6, 18.1 12.6 12.7 12.3 Production ('000 tons) Maniec (cassava) 767.0 685.8 584.5 499.5 528.1 672.6 722.1 682.5 597.7 485.9 427.5 491.7 420.8 Maize 84.6 19.6 22.2 22.6 23.8 28.9 80.7 87.7 43.6 40.7 41.7 88.6 80.6 Chillies 16.4 88.6 48.4 51.0 37.5 86.5 40.6 78.6 98.7 105.8 78.5 82.7 67.9 Red Onions 72.8 58.6 67.9 66.9 69.1 67.5 96.8 86.7 41.7 67.1 56.2 59.2 71.9 Groundnuts 7.6 4.0 6.1 7.1 7.4 6.6 6.8 6.8 5.9 6.6 4.8 6.1 7.4 Oreen Gram 6.0 7.6 8.5 9.2 10.0 11.6 16.8 16.1 18.8 18.7 17.6 19.5 20.4 Sorghum 6.8 2.8 1.1 0.5 0.6 0.8 0.4 0.6 0.4 0.2 0.1 0.0 0.0 Soyabean 1.2 - - - - - - 8.1 3.4 8.8 3.7 1.7 1.5 Potatoes 27.8 29.1 87.9 51.1 66.0 78.9 89.8 68.8 69.0 82.4 81.1 87.5 88.5 Sesame (gingelly) 6.4 - - - 10.8 18.2 16.0 7.7 8.7 8.4 6.2 6.7 4.6 Coupes 7.6 12.4 11.0 16.6 19.4 21.7 80.6 21.8 28.9 24.6 22.2 19.9 19.0 Ohal 1.0 - - - - - - 6.0 9.1 7.5 9.8 9.4 6.6 Sweet Potatoes 0.2 0.7 1.9 0.6 0.8 0.4 0.2 0.1 0.1 0.0 0.1 0.2 0.0 142.2 188.2 148.7 127.5 158.6 172.6 117.6 147.0 117.5 98.7 80.5 87.2 86.0 Source: Depafunat of Cnsus and SatMcs; and World Bank Couny Rpt on Sri Lana Reprot No. 8951-CE