India - Agricultural Research and Extension Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 8808 PROJECT PERFORMANCE AUDIT REPORT INDIA WEST BENGAL AGRICULTURAL EXTENSION AND RESEARCH PROJECT BIHAR AGRICULTURAL EXTENSION AND RESEARCH PROJECT KERALA AGRICULTURAL EXTENSION PROJECT MANARASRTRA AGRICULTURAL EXTENSION PROJECT TAMIL NADU AGRICULTURAL EXTENSION PROJECT NATIONAL AGRICULTURAL RESEARCH PROJECT (CREDITS 690-IN, 761-IN, 1028-IN, 1135-IN AND 1137-IN AND 855-IN) JUNE 29, 1990 Operations Evaluation Department This document has a resIded distrbutlan and my be usd by IenlQy In the performfanc'o thei offdal dolles. Its coantsb 'm not oftmis he dsclsed wwAWedd Bwan oelaion GLOSSARY ADO Agricultural Developmeat Officer AEO Agricultural Extension Officer AICRP All India Coordinated Research Project BDO Block Development Officer CDO Community Development Officer CDP Community Development Program CEO Chief Executive Officer CF Contact Farmer (in the T&V system) DA Director of Agriculture DOA Department of Agriculture FAO Food and **'tculturai Organization of the United Nations FSR Farming :ems Research GOI Governme-n. of India HYV High Yielding Variety IAAP Intensive Agricultural Areas Program IADP Intensive Agricultural District Program ICAR Indian Council of Agricultural Research IDA International Development Association ISNAR International Service for National Agricultural Research JAO Junior Agricultural Officer 1&E Monitoring and Evaluation NAEP National Agricultural Extension Project NARP National Agricultural Research Project NES National Extension Service OED Operations Evaluation Department, World Bank OF Other Farmer, or Non-contact Farmer PAO Principal Agricultural Officer PCR Project Completion Report PPAR Project Performance Audit Report SAR Staff Appraisal Report of the World Bank SAU State Agricultural University SDAO Sub-divisional Agricultural Officer SMS Subject Matter Specialist T&V Training &nd Visit (an extension methodology) VEW Village Extension Worker VLW Village Level Worker ZRS Zonal Research Station TO* WOUD BANKOR OF* U ISt ONLY THE WORLD DANK Washinton. D.C. 20433 US.A. Opeatims IVAta June 29, 1990 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECTs Project Performance Audit Report on the India West Bengal Agricultural Extension and Research Project Bihar Agricultural Extension and Research Project Kerala Agricultural Extension Project Maharashtra Agricultural Extension Project Tamil Nadu Agricultural Extension Project National Agricultural Research Project (Credits 690-IN, 761-IN, 1028-IN, 1135-IN and 1137-IN and 855-IN) Attached, for information, is a copy of a report entitled .aProject Performance Audit Report on the India West Bengal Agricultural Extension and Research Project, Bihar Agricultural Extension and Research Project, Kerala Agricultural Extension Project, Maharashtra Agricultural Extension Project, Tamil Nadu Agricultural Extension Project, National Agricultural Research Project (Credits 690-IN, 761-IN, 1028-IN, 1135-IN and 1137-IN and 855-IN)" prepared by the Operations Evaluation Department. Attachment This documnt has a restricted distribution and uy be wed by recipients only in the perforanace of their official duties. Its contents may not otherwise be disclosed without World Bank authrloation. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT WEST BENGAL AGRICULTURAL ITENSION AND RESEARCH PROJECT BINAR AGRICULTURAL EXTENSION AND RESEARCH PROJECT RERALA AGRICULTURAL EXTENSION PROJECT MARASA AGRICULTURAL EXTENSION PROJECT TAMIL NADU AGRICULTURAL EXTENSION PROJECT NATIONAL AGRICUL'URAL RESEARCH PROJECT (CREDITS 690-IN, 761-IN. 1028-IN, 1135-1, 1137-IN AND 855-IN) TABLE OF CONTENTS Page No. PREFACE ....................... BASIC DATA SHEETS .. ..... ............................. iii EVALUATION SUM2ARY ........................................ ..xvii PROJECT PERFORMANCE AUDIT I. INTrRODUCTION.................... Agricultural Extension and Research in India Prior to the T&V Extension Projects............ 2 Bank Involvement in Agriculture Extension and Research in India............................ 4 II. PROJECT FORMULATION, OBJECTIVES AND DESIGN........ 6 Extension Projects. ........ ..... .. .. ... .. .. ..... . 6 National Agricultural Research Project (NARP) ..... 9 NARP and the Extension Projects: Research- Extensitn Linkage.............................. 10 III. IMPLEMENTATION*............... ............... 10 Expected and Actual Staffing Requirements......... 11 The Extension Service Staffing .............. 11 Monitoring and Evaluation Staffing............. 14 Staffing of Research Components................ 14 Project Costs and Financing....................... 14 Costs......................... 14 Financing............................ 16 Staff Utilization ............... ................. 18 Educational Levels of VEWs........................ .19 Selection of Contact Farmers and Regularity of Visits...................................... 20 Monitoring and Evaluation Performance............. 22 National Agricultural Research Project (HARP) ..... 23 Physical Implementation........................ 23 Institutional Implementation................... 25 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. TABLE OF CONTENTS (continued) Page No. IV. PROJECT OUTCOME................................... 26 Impact............................................ 26 Changes in Productivity................. 27 Impact on Adoption of Technology.....oe........ 28 Institutional Achievements........................ 28 National Agricultural Research Project (KARP).. 28 Research-Extension Linkage..... ....... ..... 29 Extension Projects...... ...... ..... 30 Sustainability ........... 0.................0. 33 National Agricultural Research Project (NARP).. 32 Extension Projects...... ... ...... . ....... * ... 34 V. ISSUES AND LESSONS..................... ....... ... 36 Annexes 1. Recent Developments in Extension System Organization and Management 2. Comments on a Case Study of T&V Extension Impact 3. Monitoring and Evaluation (M&E) Data on Project Impact 4. Use of Reference Group Concept Attachments I - Project-Wise Observations/Comments on Project Performance Audit Report of the World Bank - Ministry of Agriculture, Government of India II - Comments from Kerala State Government's Department of Agriculture PROJECT PERFORMANCE AUDIT REPORT WEST BENGAL AGnICJLTURAL EXTENSION AND RESEARCH PROJECT BIHAR AGRICULTURAL EXTENSION AND RESEARCH PROJECT RERALA AGRICULTURAL EXTENSION PROJECT MAHARASHTRA AGRICULTURAL EXTENSION PROJECT TAMIL NADU AGRICULTURAL EXTENSION PROJECT NATIONAL AGRICULTURAL RESEARCH PROJECT (CREDITS 690-IN, 761-IN. 1028-IN, 1135-IN AND 1137-IN AND 855-IN) PREFACE This is a Project Performance Audit Report (PPAR) of the West Bengal Agricultural Extension and Research Project, Bihar Agricultural Extension and Research Project, Kerala Agricultural Extension Project, Maharashtra Agricultural Extension Project, Tamil Nadu Agricultural Research Project and National Agricultural Research Project (NARP) in India involving IDA credits in the amounts of US$12.0 million, US$8.0 million, US$10.0 million, SDR 18.9 million, SDR 22.9 million and US$27.0 million, respectively. The Credits vere approved on March 22, 1977, December 22, 1977, May 1, 1980, April 21, 1981, April 23, 1981 and October 10, 1978 respectively. The full amounts of the Kerala and Maharashtra project were disbursed and US$5.1 million of the West Bengal project, US$3.83 million of the Bihar project, SDR 0.06 million of the Tamil Nadu and US$7.5 million of NARP were cancelled. The Closing Dates weres West Bengal project March 31, 1985, Bihar project March 21, 1985, Kerala project June 30, 1986, Maharashtra project June 30, 1987, Tamil Nadu project June 30, 1987 and NARP September 30, 1985. Final disbursements were: West Benga' project November 18, 195, Bihar project August 23, 1985, Kerala project January 31, 1987, Maharashtra project April 12, 1988, Tamil Nadu project January 22, 1988 and KARP February 1986. The PPAR is based on the Project Completion Reports (PCRs) of the projects,1/ the Staff Appraisal and President's Reports, the loan 1/ Project Completion Reportsa i) India: West Bengal Agricultural Extension and Research Project (Cr. 690-IN), Report No. 6654, February 27, 1987; ii) India: Bihar Agricultural Extension and Research Project (Cr. 761-IN), Report No. 6800, May 29, 1987; ii) India: Kerala (Cr. 1028-IN), Maharashtra (Cr. 1135-IN) and Tamil Nadu (Cr. 1137-IN) Agricultural Extension Projects, Report No. 8258, December 29, 1989. iv) India: National Agricultural Research Project (Cr. 855-IN), (Draft) January 24, 1989. The West Bengal, Bihar and Kerala PCRs were prepared by FAO,'World Bank Cooperative Program and the PCRs for Maharashtra and Tamil Nadu Projects were prepared by the respective Continued on next page - ii - documents, the transcripts of zhe Executive Directors' meeting at which the projects were considered, on a study of project files and discussions with Bank staff in Washington, D.C. and New Delhi. An OED mission visited India in November-December 1989 and discussed the effectiveness of the Bank's assistance with GOI officials in the Ministry of Agriculture and its Directorate of Extension, Project Unit of NARP in New Delhi, State Government officials in Maharashtra, Kerala, Tamil Nadu and West Bengal and a number of zonal research stations and State Agricultural Universities. The mission met with farmers and field staff personnel in the respective states. The kind cooperation and extremely valuable assistance of GOI and the State Governments in the preparation of this report is gratefully acknowledged. The PCRs provide a full account and assessment of the project experience, including the performance of the Bank and the project executing agencies. The PPAR e.aborates on particular aspects such as projer't design, concept, the variable performance of the states, institutional achievements and shortcomings, research-extension linkage, sustainability and implications for replicability. Following standard OED procedures, copies of the draft PPAR were sent to GOI, the State Governments and Indian Council of Agricultural Research. The comments received from GOI's Ministry of Agriculture and Kerala State Government's Department of Agriculture are reproduced as attachments to the PPAR. Continued from previous page State Governments. The NARP PCR was prepared by the Indian Council for Agricultural Research with an Overview prepared by Bank staff. - it - PROJECT OOPLTIO REPORT WEST BENGAL AGRICULTURAL EKTEUSION AND RESEARCH PROECT (CREDIT 690-IN) BASIC DATA SHEST KEY PROJECT DATA Appraisal Actual or Actual as 2 of Item Estimate gsti_ated Actual Appaisal Estimate Total Project Costs (US$ million) 28.0 18.22 65 Total Project Costs (Rs Million) 253.0 188.9 75 Credit Amount (US$ million) 12.0 6.9 58 Disbursed (US$ Million) - 6.9 Cancelled (US$ Million) - 5.1 Institutional Performance Poor Number of DL.ct Beneficiaxies (million farmers) 5.0 2.5 50 j CUMULATIVE zSTIMATED AND ACTUAL DISBURSEMENTS (in US$ Million) EI Y79 FYI2 an F I 82 Y83 F84 US FY86 Appraisal Estimate (US$ million) 1.6 5.0 8.3 10.7 12.0 - - - - Actual (US$ million) - - - - - 1.3 4.68 5.66 4.85 Actual as I of Appraisal (1) 0 0 0 0 0 10.8 39.0 47.17 57.0 Date of Final Disbursement November 18, 1985 - tv - First mention in File& 04/75 Negottations 01/77 Board Approval- 03t22/77 Sgntna 06101rn Meottiveneus 05/30177 08130177 Closing Date 09130/82 03/31/85 <staff weks) Z? YZ7 L U 8 S ZII 2 MSAZ t 8 MS AE TIOTA Identlfication/ Preparation 0.5 - - - - - - - - - 0.5 Appralsal 39.8 - - - - - - - - - 39.8 Negotiations 4.0 - - - - - - - - - 4.0 Supervision : 16 ,U 1-5 154 14 1 1 - :3 9-.Li 8 ,,9 104.1 Total 46.9 16.2 15.5 15.4 14.8 11.0 12.3 9.6 2.8 3.9 148.4 MISSION DATA Spe*1alsa- Date No. of tions Perfomnce Types øf Mission (.lyr.) Persons Reresented Rati rd Problems Identlficatton 11/75 3 ES(2),E Preparation 03/76 3 AG,R,E Appraisal 05/76 7 Supervision I 08178 2 ES 3 3 p Supervision II 12/78 2 S,R 3 3 p Supervision III 10179 4 ES(3),R 3 1 K,P Supervision IV 03/80 2 R - - - Supervision V 01/81 5 ES(4),R 3 1 m Supervision VI 07/81 2 ES 3 1 K Supervision VII 11/81 2 ES 3 1 m Supervision VIII 02/82 2 ES 3 1 x Supervision IX 08/82 1 ES 3 1 N, Supervision X 02/83 2 ES 2 1 M,T,j Supervision XI 08/83 3 ES,AG,HE 2 1 M,T,? Supervision XII 02/84 2 ES,R 2 2 m Supervision XIII 08/84 2 AG,ES 2 2 N,T Supervision XIV 02/85 3 AGE,S(2) 2 3 M,T Completion 03/85 2 OTHER aJECT DATA Sorrowest Government of ladia Executing ASency: Government of West Bengal Fiscal Year of the Borrovert April 1-March 31 Follow-oa Projectt None jg Project covered 502 of the total area envisaged at appraisal. 4_ Input as staff weeks, Ic AG - Agriculturalists E - Economist R - Research Experts ES a Extension Specialistj ME = Monitoring and Evaluation Specialist. 4L 1 - Problem-free or minor problemst 2 * Moderate problemst S - Major problems. 1g 1 - Improvings 2 - Stationaryt 3 - Deteriorating. U M = Managerialt P - Politicalt T - Technical; F w Financial. * vi.- SIM AMGICULTURAL EXTENSON AND SEARCH PROJECT (CROEDIT 761-xv BASIC DATA SHUT EY PROJECT DATA Appraisal Actual or Actual as I of Item Estimate Estimated Actual Asaraisal Est%agte Total Project Costs (US$ million) 16.0 32.7 204 Total Project Costs (Ra million) 144.0 297.0 206 Credit Amount US$ million) 8.0 4.17 52 Disbursed (US$ million) 4.17 Cancelled (US$ million) 3.83 Inst itutional Performance Pair Number of Beneftiariea (million farm families) 7.6 3.8 50 CU-OLATIV MSTIAra AND ACIMA DISMIRSMNNTS (in VS$ Million) WI Ui EI1 Ell M IM MA 2M ElM Appraisal Estimate (US$ million) 0.3 1.6 2.8 4.5 6.8 8.0 - - Actual (US$ million) - 0.61 0.89 1.33 2.19 2.78 3.93 4.1, Actual as I of Appraisal (Z) - 38 32 so 32 35 49 52 Date of Final Disbursement A gust 23, 1985 Orisinal Plan First mentlon n FLe.- 0976 sesottatløns -1077 Board Approval 12/77 12122177 Sinug 01/06178 Effettveness 03/31/78 05/02/78 C1os1ng Date 10/31/83 03/21/83 STAP IXPUTS /a (stt ff Veeks) 1LZ Z Y78 9j FYb0 ZYi FY82 YB EY84 EL Y86 T2TI Identificatton/ Preparatton 1.2 2.1 - - - - - - - - 3.3 Appralsal - 4.7 - - - - - - - - 4.7 Nesotlations - 4.8 - - - - - - - - 4.8 Supervision - 1.6 11.4 5.8 4.1 1.0 8.4 16.3 6.3 15.6 79.5 ISSIONWDA ia Specializa- Date NC . of ti~ns Performance Types of MIssion (Eo./r.) Persons Reresented Rating Tggd Problems IdentlficationLb 06/76 Preparation fC Appraisal 04/77 4 5,R,ES(2) Supervision I 06/78 2 E,ES 3 2 p Supervision II 12/78 2 E,ES 3 1 p Supervision III 06/79 2 E,S 3 2 p Supervision IV 10/79 2 E,ES 3 2 p Supervision V 05/81 2 E,ES 3 2 p Supervision VI 11/81 3 AG,ES,R 3 2 p Supervision VII 05182 2 ES(2) 3 1 K,P Supervision VIII 01/83 2 ES(2) 3 2 M,T Supervision IX 09/83 3 ES,AG(2) 3 1 M,T Supervision x 03/84 3 AG,ES<2) 3 2 14,T Supervision XI 09/84 3 AG(2),ES,F 3 2 1,T Supervision XII 02/85 2 AG,ES 3 2 M,T Completion 03/85 2 - viii - OTHER PROJECT DATA Borrower: Government of India Executing Agency: Government of Bihar Follow-on Project: Special Sub-Project under National Agricultural Extension Project III Credit No: 1754-IN Amount (US$ million): 9.0 out of total credit of 85.0 Daic of Approval: November 20, 1986 14 Input as staff weeks. lb as a result of Agricultural Sector Survey of Eastern Region States of India (Report No. 1440-IN). Ic Prepared by Government of Bihar assisted by NDO. L_d AG - Agriculturist; E = Economists R = Research Expertt ES - Extension Specialisti F = Financial Analyst. /e 1 - Problem-free or minor problems; 2 = Moderate problems; 3 - Major problems. If 1 - Improvings 2 - Stationary; 3 = Deteriorating. LZ M - Managerials P = Political; T = Technical. - Ix - REM&TA AGRuTCULT1UAL nIIYI!TRNRTO PUn3TR"I (Credit 1029-IN) PRO1ECT COM6Ltd0yLREPORT BASTC~ DAL~TN SHEE U NSEY PROJCT DATA Asais ~al J Atual as & of Estiae Antnal..LL Apraisal Entimate. Total Project Cost (US$ million) 15.3 25.1 1641 21 Credit Amount (US$ million) - Disbursed 10.0 10.0 100% Proportion of Physical Targets Met 1 - Housing: for VENs 488 453 93% AEOs 94 102 109% * Dept. Vehicles: Sedan 15 18 120% 4-WD 95 88 104% Trucks 11 13 118% Motorboats 4 4 100% Bus 1 1 100% - Staff Owned: Motorcycles 417 429 103% Bicycles 11560 60 44 Date Physical Components Completed Sept. 1985 June 1984 Institutional Performance Satisfactory EOIT.ATV r.TRu yRme,MPNFR E.1~ EX...8 Ez..A £L.. F.L..8 EX..3A EX..A Appraisal Estimate (US$ million) 0.3 1.0 2.4 5.6 9.0 10.0 10.0 Actual (US$ million) - 0.4 1.0 2.2 3.9 6.2 10.0 Actual as I of Appraisal Est, (8) 0 40 42 39* 43 62 100 Date of Final Disbursementi 31 January 1987 PROJECT DATES Preparation June 1979 Appraisal August 1979 AL Negotiations April 1980 Board Approval 1 May 1980 Credit Signing 25 June 1980 Credit Effectiveness 18 August 1980 Credit Closing 30 June 1986 *X- STArF iEDTS 2 . .8 . . .A F.XAt F8A 2 TOTAL Identification/ Preparation 4.0 9.1 13.1 Appraisal 15.5 15.5 Negotiations 3.1 3.1 Supervision 1.4 11.6 10.0 12.3 21.4 9.7 5.3 0.4 72.1 Subtotal 4.0 29.1 11.6 10.0 12.3 21.4 9.7 5.3 0.4 103.8 MISSTnO DATA No Mn andayI S A rating Tyunes of M120im DArZ*. inFedRo nmee qAtAMAI.2 Probane (4) (5) (6) (7) Appraisal IL Aug. 1979 3 15 extn, e, Supervision 1 Jan. 1981 2 14 extn, e 3 1 P Supervision 2 Sep. 1981 3 15 a, e, extn 3 2 P. M Supervision 3 Apr. 1982 2 10 extn. 3 2 P, M Supervision 4 Jan. 1983 2 10 extn. 2 1 M, Supervision 5 Jun. 1983 3 15 extn. 2 1 M, Supervision 6 Jan. 1984 2 12 a, 2 2 m, Supervision 7 Jun. 1984 3 15 a, 2 2 MIF, Supervision 8 Dec. 1984 2 12 a, 2 2 M,F, Supervision 9 June 1985 2 10 a, 2 2 M* Supervision 10 Dec. 1985 1 3 a, 2 Supervision 11 Apr. 1986 1 5 a, 2 OT9ER DATA Borrower: Government of India State Government of Kerala Fiscal Year of Borrower: 1 April - 31 March Name of Currency (Abbreviation): Rupee (Rs) Exchange Rate (Rs/USS): - Appraisal Year Average (1979) Rs 8.40 - Intervening Years (1980 - 1985) R210.00 - 1986 Average Rs12.60 Follow - on project: None --****---------------------------------------------------------------------------* L As at 31 December 1986. 2L Not directly comparable with appraisal estimate, as records of amounts spent on staff and operating costs have not been maintained on an incremental basis (as in SAR) over the entire project implementation period. IL The project was prepared by GOR with assistance from staff of the World Bank's Resident Mission in India. AL a- Agriculturalist, e- Agricultural Economist, extn.- Extension Specialist. - 1 Problem free or Minor Problems, 2- Moderate Problems, 3- Major Problems. AL 1- Improving, 2- Stationary. 3- Deteriorating. 2L T- Technical, N- Managerial, P- Political, F- Financial, 0- Other. AL Follow up visits were made in December 1979 to finalize staffing details. - x- - MAHARASTRA AaRIULTURAL rXTENSION (Credit 1135 - IN) PRJECT COMPLETION REPORT BASIC DATA SHrMT KeVY 99JR flATA Appraisal Actual as % of EAtiM= Atn 'Amoraisal Pstimat Total Project Cost 39.2 40.8 107 (USS Million) Credit Amount (SOR Million) 18.9 18.9 100 Physical Targets Met -Housing for:VEW 1,200 748 62 AEO 150 91 61 -Dept.Vehicles-4WD 173 225 130 Bicycle 3,090 3,160 102 Motorcycle 762 766 101 Date Physical Components Completed: 30th June 1986 30th June 1987 Institutional Performance: Satisfactory PROJCT DATES Project Appraisal October 1980 Board Approval 21st April 1981 Credit Signing 7th May 1981 Credit Effectiveness 22nd July 1981 Credit Closing 30th June 1986 30th June 1987 CtUMULATIVE ISRURSEMENT* Appraisal Estimate 0.3 2.4 7.0 14.0 20.0 23.0 23.0 (US$ Million) Actual (US$ Million) 0.6 2.0 5.0 10.4 15.7 18.5 20.5 Actual as 0 of Appraisal Estimate 200 83 71 74 79 80 89 Date of Final Disbursement: 12th April 1988 * Note that the Credit (nominated in SDR) was fully disbursed. - xit - STAPPTNUT (Staff weeks) E,MI £.8 EXA USA L%SA X.. E=81 EI ZTOAL identificationj Preparation 0 Appraisal 6.0 6.0 Negotiations 1.5 1.5 Supervision 2.1 9.0 13.9 17.3 8.5 14.3 4.6 1.3 71.0 Subtotal 9.6 9.0 13.9 17.3 8.5 14.3 4.6 1.3 83.5 MISSION nATA nare nf T andayse neialiiationv f Miainn InEsA F-lld Ran ented SAtus Problems (1) (2) (3) (4) (5) Appraisal 10/80 5 45 A.A.,E.EX Supervision 1 09/81 3 18 A,E,EX 1 2 0 Supervision 2 03/82 3 18 A.E,EX 1 1 M Supervision 3 10/82 2 8 A.EX 1 2 N Supervision 4 07/93 3 18 A,ME.EX 1 2 M Supevvision 5 01/84 2 10 A.EX 1 2 M Supervision 6 07/84 3 12 R.EX.EX 1 2 H Supervision 7 01/85 3 15 A,EX,EX 1 2 F Supervision 8 08/85 3 15 EX,EX,A Supervision 9 02/86 3 15 LEX Supervision 10 09/86 2 10 E'EX Supervision 11 02/87 1 3 OTR .ATA Borrower: Government of India State Government of Maharashtra Fiscal Year of Borrower: 1st of April - 31st March Nam of Currency: Rupee (Rs) Exchange Rate (RS/US$): - Appraisal Year average Re 8.00 - Intervening Years average (1981-1986) Rs 10.72 - 1987 average Ra 12.92 Follow-on project None -------------------e---e *****---**-----------** * (1) About 27 days were spent in field to appraise 3 projects (2) A: Agriculture. E: Economics, EX: Extension: ME: Monitoring & Evaluation; F: Forestry; and R: Agriculture Research (3) 1: Problem Free or Minor Problems; 2: Moderate Problems; 3: Major Problems (4) 1: Improving; 2: Stationery; 3: Deteriorating (5) T: Technical; K: Managerial: P: Political; F: Financial: 0: Other - xthi - (Credit 1137-IN) PROJECT..OMPLETTIN REP6T Appraisal Actual as 0 of Es1! IAti Auttimat KEY PROJECT DATA Total Project Cost (USS Million) 45.6 37.8 83 Credit Amount (SOR Million) 22.9 22.94 99.7 Physical Targets Met Rousing for: AEO 300 300 100 VEN 1,450 1,450 100 Vehiclest 4WD 115 134 116 Motorcycles 830 692 82 Bicycles 2,500 2,053 84 Date Completeds 30th June 1986 20th June 1987 Institutional Performances Satisfactory PROJECT DATES Appraisal October 1980 Board Approval 23rd April 1981 Credit Signing 7th May 1981 Credit Effectiveness 22 July 1981 Credit Closing 30th June 1987 30th June 1987 Appraisal Estimate (US$ million) 0.4 3.9 11.1 19.2 24.6 28.0 28.0 Actual (USS million) 1.1 4.0 6.2 8.9 13.5 20.1 26.5 Actual as 0 of Appraisal Est.. 275 103 56 46 55 72 S" Date of Final Disbursement: 22nd January 1988 * Note that the Credit (nominated in SDR) was almost fully disbursed. - Xiv - (Staff Weeks) TQ,Ag £ggi £,8ag £gg2 £X,.L ,,ag f,a T,, ZfM ZAra Identification/ Preparation 0.4 2.3 2.7 Appraisal 22.3 22.3 Negotiations 1.3 1.3 Supervision 0.8 19.2 11.5 16.3 6.1 12.9 13.5 5.7 85.9 Subtotal 0.4 26.7 19.2 11.5 16.3 6.1 12.9 13.5 5.7 112.2 DTAUTO NDATA Date_of gg,f gg Mandav Soeializa1tin Yvnso Migsion PeSons= In Flold Repkresented MAt"g Problemsi (1) (2) (3) (4) (5) Appraisal Oct-80 5 9 A,A,E,E,EX Supervision 1 Oct-81 2 15 EX,EX 1 1 0 Supervision 2 Jan-82 1 6 EX 1 1 0 Supervision 3 Jun-82 3 15 EX,EX 1 1 0 Supervision 4 July-83 3 15 AX.EX ME 1 2 4 Supervision 5 Jan-84 2 8 EX,R 1 1 N Supervision 6 Jun-84 3 15 A,ZE EX 1 2 M supervision 7 Feb-85 3 15 A,E.EX 1 2 M supervision a Sep-85 2 10 EIEX Supervision 9 Mar-86 2 10 E1EX Supervision 10 Oct-86 4 20 APA,E,EX supervision 11 Feb-87 4 20 A.E.EX,F OTHER DATA Borrower: Government of India State Government of Tamil Nadu Fiscal Year of Borrower: 1 April - 31 March Name of Currency (Abbreviation): Rupee (Rs.) Exchange Rate (Ra./US$): - Appraisal year average (1980). Rs. 8.00 - Intervening years (1981-1986) Rs. 10.72 - 1987 Average Rs 12.92 Follow-on project: None -----------------***------------------------------------------------------**-------* (1) About 27 days were spent in the field to appraise 3 prijects. (2) A. Agriculture Z: Economicss EX: Extension; MB:Monitoring and Evaluation: F:Forestry; and R: Agricultural Research (3) 1: Problem Free or Minor Problems; 2: Moderate Problems; 3: Major Problems (4) 1: Improving: 2: Stationary; 3: Deteriorating (5) T: Technicals K: Managerial: P: Political; F: Financial: 0: Others - xv - NATIONAL AGRICULTURAL RESEARCH PROJECT (Credit 855-IN) PROJECT COMPLETION BEFORT BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Actual as % of Ttem Expctation Current Estimate Appraisal Estimate Total Project Costs (USS million) 54.0 44.0 81 Credit Amount (US$ million) 27.0 19.5 72 Physical Targets Background Papers 15 22 147 Research Reviews 15 16 107 Research Subprojects 56 86 154 Consultant Man-months 160 79 49 Date Physical Components Completed 30 Sept. 1983 30 Sept. 1985 - Institutional Performance Satisfactory CUMULATIVE ESTIMATED AND ACTIL L DISBURSEMENTS .L2 EMSA EY.1 Z= Z Xf Y8 EXF8A E X.. L1. Appraisal Estimate (US$ million) 0.9 3.8 9.2 17.5 27.0 27.0 27.0 27.0 Actual (US$ million) - 0.4 1.6 4.7 7.0 11.4 15.0 19.5 Actual as % of Appraisal (%) - 10.5 17.4 26.9 25.9 42.0 55.5 72.0 Date of Final Disbursement: 2/86 Revised Estimate A/ - - - - - 11.5 18.5 20.0 Actual as % of Revised Estimate - - - - - 99.0 81.0 97.5 PROJECT DATES Original Plan Revision Actiia3 Identification (Project Brief) 5/76 3/77 and 9/77 11/77 Preparation 2/78 - 2/78 Appraisal 3/78 - 3/78 Negotiations 8/78 - 8/78 Board Approval 10/78 - 10/78 Signing (Credit Agreement Date) 12/78 - 12/78 Effectiveness 2/79 - 1/79 Closing Date 9/83 9/85 9/85 a/ Project was extended for two years and disbursement estimates revised from 1984 onwards. STAFFTINPUTS (Staff weeks) EY7 .12 FY2 EM Er EM81 Z= XFA E8= Ttal Identification/ Preparation 38 38 Appraisal 140 140 Negotiations 3 3 Sup.ervision 11 13 5 19 20 18 32 118 Subtotal 38 140 14 13 5 19 20 1 32 299 MISSION DATA Date No. of Specializations Performance Types of (Molvr) Perons Represented_aL Ratina b/ Prohlem a/ Identification/ 8/76-8/77 7 AE, AP, EC, ED, - Preparation PB, AG2 Appraisal 3/78 5 AG2, EC, AE, E - - Negotiations 8/78 - - - - Board Approval 10/78 - - - - Supervision I 12/78 2 AG2 1 M Supervision II 3/79 3 AG2, EC 1 M Supervision III 9/79 1 AE 2 M Supervision IV 3/80 3 AE, AG, RES 2 M Supervision V 10/80 1 AG 2 M Supervision VI 3/81 2 AE, RES 2 0 Supervision VII 10/81 3 AE, AG2 2 M Supervision VIII 4/82 3 AE, AG 2 T, M Supervision IX 10/82 4 AG4 2 T Supervision X 6/83 2 AG2 2 T Supervision XI 3/84 4 AG4 2 T Supervision XII 9/84 5 AG4, AE 2 T Supervision XIII 6/85 4 AG4 2 T, M A/ AE = Agricultural Economist, AP = Animal Production Specialist, EC Economist, ED = Education Specialist, PB - Plant Breeder, AG - Agriculturalist, X Engineer, RES - Research Administrator. 1/ 1 - Problem free, 2 = Minor problems. a/ M - Management, T = Technical, 0 - Other. OTHER PROJECT DATA Borrower: Government of India Fiscal Year of Borrower: April 1 - March 31 Name of Currency: Rupee (Ra) Follow-On Project: Name National Agricultural Research Project II Credit Number 1631-IN Amount (US$ million) 72.1 Approval Date October 22, 1985 - xvii - PROJECT PERFORMANCE AUDIT REPORT INDIA WEST BENGAL AGRICULTURAL EXTENSION AND RESEARCH PROJECT BIHAR AGRICULTURAL EXTENSION AND RESEARCH PROJECT KERALA AGRICULTURAL EXTENSION PROJECT MHARASHTRA AGRICULTURAL EXTENSION PROJECT TAMIL NADU AGRICULTURAL EXTENSION PROJECT NATIONAL AGRICULTURAL RESEARCH PROJECT (CREDITS 690-IN, 761-IN, 1028-IN. 1135-IN AND 1137-IN AND 855-IN) EVALUATION SUMMARY Introduction This is the third audit by OED of the "Training and Visit" (T&V) system of agricultural extension implemented on a statewide or near state- wide basis in India.11 This audit, in addition to covering five statewide extension projects, also covers the National Agricultural Research Project (NARP) which included research support in four of the states. The inclusion of NARP provides the opportunity to specifically address a number of issues related to the linkage between research and extension services. Since these projects closed between 1985 and 1987, the audit examines the continued commitment of the states to the T&V extension and research systems through adequate funding and provision of qualified staff. The projects under audit cover more than 10 years of implementation experience of Bank-supported extension and research activities in five states in India. Extension has a long tradition in India. In the 1950s and 60s it was a significant component in community and agricultural development programs. In the early 19,70s, extension was implemented through both single crop schemes and mvltipurpose rural development programs whose extension activities overlapped. These services appeared fragmented and uncoordinated, resulting in reduced effectiveness and serious inefficiencies. At the same time, extension staff in the states increased sharply in number, some under Community Development Programs (CDP), others under the Department of Agriculture (DOA) and still others under DOA but administratively responsible to the CDP. Most extension staff belonging to CDP worked part-time on extension and were poorly trained in agriculture and extension. During the same period, research remained removed from extension. State governments, with support from Government of India (GOI), began to 1/ Throughout this report agricultural extension is referred to as extension; and agricultural research as research. - xviii - establish autonomous State Agricultural Universities (SAUs) with the hope of slowly taking over research from the DOAs. The establishment of SAUs led to increased isolation of researchers from farmers' problems. New extension sections were created at SAUs and mandated to undertake direct extension activities, but the impact on research and extension was very limited. Linkage between research and extension was extremely poor at the time of initial Bank involvement in the sector. Bank involvement started in the early 1970s and, to date, T&V extension has been supported in four stages: the first in 1974 on a pilot basis in Rajasthan, Madhya Pradesh and West Bengal; the second with the statewide introduction of the system in 1977 in six states and including a research component; the third in 1979, on a statewide basis but with extension as the sole project component in three states; and the fourth in 1984 with the multi-state National Agricultural Extension Project . (NAEP). T&V extension to date has been introduced in 17 major states of India through 15 Bank-assisted projects. Bank lending for these 15 projects amounts to US$373 million and total costs at appraisal were expected to amount to US$660 million. Bank lending for the five extension projects under audit amounts to about US$81.0 million and total costs at appraisal were US$143 million but at completion amounted to US$155 million approximately. In addition, GOI and the various State Governments continue to incv- large expenditures to f£aance the increase in staff and operating costs needed to support the institutional changes brought about by these projects. NARP vas initiated in 1978 with a US$27 million credit to support applied research through the SAU system. Obiectives The five extension projects aimed at achieving early and sustained increases in agricultural production through the reorganization and strengthening of the extension service of the DOAs in their respective states in line with T&V principles. The common characteristics of the projects were to: appoint full time village extension workers (VEWs) to work exclusively on extension; establish a single line of command between VEWe and extension headquarters in a unified extension system; select contact farmers (CFs) to disseminate information; establish a fixed and regular cycle of fortnightly visits by VEWs; use simple, practical, relevant messages and concentrate on the most important crops; provide regular in-service training to extension staff at all levels; initiate a system of feedback from farmers via extension staff to researchers; and develop monitoring and evaluation (M&E) procedures. In addition, the West Bengal and Bihar projects also included a component to develop and strengthen adaptive research. NARP provided support for the research system in the other three states and aimed at strengthening the capabilities of the SAUs for conducting location-specific, production- oriented research on the basis of identified agro-climatic zones within each state. Since NARP was to strpngthen need-based research it emphasized a closer link between research and extension at the grassroot level. - xiX - The T&V extension system embodied in the extens-on projects differed from the earlier approaches in India mainly in the responsibili- ties assigned to the front-line field officers working in villages and the organizational structure of the state's extension services. While in the past the village worke., had to undertake many non-agricultural and non- extension activities, these projects sought the full-time commitment of a VEW, relieving him of other responsibilities in his primary task of disseminating technology to the farmers. Organizationally, the institution of a single line of command, whereby the VEW would be both technically and administratively supervised through a chain of command under the Extension Headquarters, was also a major change from the past. Implementation Implementation performance among the six projects varied widely. The five extension projects fall into two groups. The first group, West Bengal and Bihar, covers two extension/research projects with severe imple- mentation problems. Despite the fact that the Closing Date of the West Bengal project was delayed 30 months and that of the Bihar project 7 months, neither project achieved tae expected statewide coverage and 43% of the former and 48% of the latter credit had to be cancelled. The second group covers Maharashtra, Kerala and Tamil Nadu which were successfully implemented and fully disbursed, although the Maharashtra ?roject closure was delayed by 12 months. Appraisal and actual costs, credit amounts and cancellations for the extension projects are shown in Table 3.2 of the text. Actual project costs in Rupee terms were higher than appraisal estimates in all states except West Bengal where the project was scaled down. The increased costs were largely due to the higher salary bill (resulting from both increased unit costs and employment of more staff than anticipated), higher civil works costs and slower implementation. There is some indication that the amount of credit allocated to civil works was excessive, particularly the allowance for VEW housing. The civil works component proved difficult to implement in Bihar, West Bengal and to some extent in Maharashtra. The use of government housing was not financially attractive to VEWs so that their occupancy rates after construction were low. In general, each of the states attempted to undertake the programs agreed to at appraisal and adopt the key principles of T&V extension during the early years of the project. However, a number of problems hindered implementation and adversely affected impact as a result of which the states modified the original approach to varying degrees during the later years and subsequent to the project. Following are some observations regarding the implementation experience. a) At the time of appraisal of the extension projects, it was expected that the majority of extension staff positions would be filled by reallocation of existing staff and that the number of new posts would be relatively limited, except in Maharashtra. In practice, staff transfer from CDF and other departments to extension posed major difficulties, more - xx - in some states than others. What was misjudged was the states' willingness and/or ability to enforce a reform when it did not have full political and administrative support, and also how strongly entrenched were the staff and management of existing programs. The inability of state governments to transfer staff into extension led to large scale recruitment of new staff and restricted implementation of the programs in West f- nal and Bihar. The need for a larger number of new staff and the need t% :rain them had a significant impact on DOA and state budgets. Instead of transferring the salary budget between departments along with staff, new positions were created and budgeted for the new extension system. This had not been anticipated at appraisal or included in project costs. b) The diversion of extension staff at all levels to non-extension duties, strictly against T&V principles, was a permanent source of concern during project implementation in all states. This diversion, for instance, was worst in Maharashtra where staff at district level were under the tech- nical guidance of DOA but remained under the administrative control oi the Chief Execuzive Officer (CEO) of the CDP. c) The involvement of extension staff in non-agricultural and agricultural input supply services was not the only diversion of their time. There were a multiplicity of specialized schemes, whoze staff had been transferred to extension, which had to be implemented under pressure for target achievements. Much of the time of SMS staff at district head- quarters was taken up in the administration of these schemes and in other routine non-extension duties, rather than in acting as technical specialists. d) The educational qualifications of VEWs, especially those trans- ferred from community development work, was a matter of concern. Very few had agricultural qualifications and even fewer were agricultural graduates. Minimum qualifications for new VEWs were supposed to be matriculation plus a two year agricultural diploma, but many VEWs with lesser qualifications were recruited partly because state level training institutes had inade- quate capacity to produce the required number of diploma holders. Training in the projects was usually unable to overcome the deficiencies inherent in the unacceptably low basic education level Vhere this existed; low educa- tion level of VEWs also inhibited using them for a more responsive form of extension which was developing in some states at the end of the projects. e) The extension "messages" concentrated on important items for major crops. However, towards the end of project implementation management of the programs generally agreed that the "messages" had to be responsive to all the significant technical constraints in the common production systems, including those dealing with soil and moisture conservation, agro- forestry and livestock as part of the system. f) Extension "messages", in general, did not have adequate eco- nomic focus and were often not cognizant of the resource circumstances of the farmer groups in the communities being serviced; many "messages" were, therefore, not appropriate to the farmers' situations. - xxi - g) The deficiencies in the extension "messages" were partly asso- ciated with the system of defining messages for a very large area, and many SMS were unable to adapt the messages in training sessions to local circum- stances faced by the VEWs. This was due to lack of experience and exposure to real farming conditions and of specialized knowledge of the SMS in some cases. The poor standard of many SMS also inhibited communicatinn with researchers. h) CF xelection was commonly deficient in that the selected indi- viduals often had little influence on their neighbors. No attempt was made to identify existing social groups of farmers and to select an individual w .hin that group as a contact farmer; usually the relatively progressive --ndividuals within the various farm size strata were selected. However, this was partly a project design problem as the concept of understanding and using the social characteristics of communities was not given attention at appraisal. i) An M&E capability was developed in all the projects. Regular surveys and special studies led management of the extension services to undertake some changes in their programs. However, with few exceptions, adequate use has not been made of the information generated by M&E by management. j) Changes in the organization of the programs has occurred in varying degrees in all the projects. In Krala, the duties of the VEW have been considerably broadened to cover most functions of the DOA at the local level, but with allocation of a specific VEW at an "extension center" to handle administrative and related agricultural services, leaving "field" VEWs to manage the scheduled farm visit program; VEWs now address groups rather than CFs. In Tamil Nadu, the extension, input supply and special scheme responsibilities have been amalgamated, while in Maharashtra and West Bengal considerations are being given to how the extension service can be involved in the input supply system without detracting from the exten- sion functions. In Tamil Nadu and Maharashtra, some districts are starting to experiment with the use of groups rather than CFs. In all states, the initial exclusive attention to field crops has been or is being changed to include a much broader range of issues including fodder production and livestock, horticulture, agroforestry and soil and moisture conservation as relevant to the farming systems in an area. In Kerala and Tamil Nadu, the fraquency of technical training has been reduced to one per month, with a session in the other fortnight devoted to coordination of input supply and other services. MWAR. Implementation of NARP got off to a slow start and by the time of the original Closing Date the Indian Council of Agricultural Research (ICAR) commitment of funds was only 53Z and IDA disbursement 26% of appraisal estimate. A two year extension of the Closing Date resulted in disbursements in local currency nearly equalling appraisal estimates. US'7.5 million of the approved IDA credit amount of US$27.0 million was cancelled due to currency adjustments. Final accomplishments in initiating subprojects were impressive and instead of 15 participating SAUs there were - xxii - 221 there were 21 subprojects for strengthening research administration at SAUs as against 15 planned; and 86 research projects were started as against the 56 planned. These additional subprojects were able to be financed because the actual sanctioned costs per research subproject were nearly one-half the appraisal estimate. The zonal research concept was the key element in NARP. During implementation, the zonal idea was further developed by requiring each state to produce a zonal status report which identified sub-zones, the specific problems of sub-zones and research priorities for them. The con- cepts of seasonal zonal workshops, monthly training sessions with SMS staff, and joint scientist/extension staff field visits and trials took about three years to be generally accepted. This acceptance was strongly influenced by a series of workshops at SAUs organized by ICAR to explain the new multi-disciplinary approach to adaptive research directed at solv- ing farmers' problems. As the reorganized extension system began to put pressure on research to develop technologies suitable for local agro- climatic conditions, the need for this approach also became more apparent. Impact The agricultural impact of these projects is very difficult to assess and no economic rates of return were calculated either at appraisal or completion. Since research and extension are only two of the many factors involved in the production process, it was virtually impossible to either separate out their effects on production and yields or attribute causality to these projects. To properly assess project impact M&E surveys should have included data from project as well as some non-project areas. However, data from non-project areas were not collected in any of the states as a result of which project impact measurement was not possible. M&E surveys in the project areas pointed to two important results on the adoption of technology: i) CFs in general, and others who are visited regularly by extension staff, tended to adopt technology sooner and achieved higher yields. However, the positive results were neither exclusively attributable to the projects nor to T&V extension embodied in them because of the process of CF selection which is in favor of relatively progressive individuals- and ii) lack of knowledge was rarely cited as a reason for non-adoption of a specific technology. The most common reasons cited were economic considerations, climatic factors, unavailability of inputs and lack of irrigation, with the implication that lower adoption was obtained where the recommended technology was either not relevant to the farmer's situation or was not appreciated as such by the farmer. Despite the inability to link incremental productivity to incremental investments, the overall conclusion of the audit regarding the extension projects is that relatively intensive and improved technical extension services have been provided to the farming communities in Maharashtra, Kerala and Tamil Nadu, and partial coverage has been obtained in West Bengal and Bihar. While recognizing implementation problems which - xxiii - are noted under "Issues and Lessons" in this Summary, the audit generally supports the relatively positive conclusions of the PCRs concerning the former three extension projects (although with less enthusiasm than the PCRs) and shares the reservations concerning the latter two projects. Under NARP, the zonal research concept was successfully introduced in the participating states and a good basis for research-extension linkages was provided. The project introduced an effective central funding mechanism for research in the SAUs which promoted and provided incentives for an improved research environment and a research program review system designed to justify expenditures. Sustainability The five extension projects required a substantial allocation of professional staff and financial resources. The recurrent cost element was substantially higher than initially anticipated due to problems of trans- ferring staff from other programs. These costs now have to be met by the states. The overall conclusion is that the states are continuing to support the extension schemes with varying degrees of operational diffi- culties and efficiency. Tamil Nadu has adopted a rational attitude of ensuring that its extension staff are adequately supported by having a non- salary budget at around 25% of total service costs. However, Maharashtra, despite its generally favorable attitude to extension, has reduced its non- salary budget to 7%-10% of total extension costs; this reflects the large increase in the salary budget associated with the project and is creating operational difficulties. In West Bengal and Bihar, the level of non- salary support for the services appear to have been maintained at a bare minimum. The allocation of a greater amount of the States' recurrent budgets to operating costs of the extension services usually competes with such items as input subsidy in the crop subsector, and irrigation services, soil conservation and animal health services, which are major consumers of annual recurrent expenditures in other subsectors. All these competitors are conspicuous programs which usually demand strong political and official support. Consequently, as adequate recurrent cost funding is essential to maintain an effective extension program, the sustainability of the system will depend significantly on the ability of the extension service to demon- strate its benefits to farmers, farmer organizations, political representa- tives and budgetary officials. Attaining and sustaining the level of financial commitment needed tor efficient operation of extension services is likely to occur only if the state governments are convinced that extension of this scale and type is necessary, that funding of the extension service has priority over some other aspects of the State's recurrent agricultural budget, and that the public sector must continue to bear a large part of the financial burden for extension. The on-going modifications to the original system are a reflection of an attempt to improve the system to meet these requirements. - xxiv - Sustainability for NARP will depend on the level of funding in general, and on the attitudes of the research establishment to zonal research. Recent analysis in India indicate that the states now share greater responsibility for research expenditures than in the past. Ten years ago state expenditures on research were only half that of 001, but now commitments by the state governments exceed GO commitments. NARP supported this transfer of responsibility for research and its funding to the states. Continued funding for operational cost support will depend on the level of transfers from the centrally-funded ICAR programs, allocations in state budgets and the apportionment to zonal programs within SAU bud- gets. However, the reluctance of some state governments to provide ade- quate operational costs for subprojects which had been completed under NARP is becoming evident. Indeed, the size of the system is such that there will be continued pressure on the research institutions to become more efficient and cost effective. Although the concept of zonal research is well established and ahould continue to develop, it behooves the SAU net- work to maximize the use of zonal research facilities, institute a trans- parent system for prioritizing research which demonstrates potential economic benefits and present its accomplishments in a convincing manner to the funding authorities. Issues and Lessons Extension. The audit supports the relatively positive conclusions of the PCRs on the extension projects in Tamil Nadu, Maharashtra and Kerala (although with less enthusiasm than the PCRs) and shares the reservations regarding the extension projects in West Bengal and Bihar. A number of significant implementation problems are noted in varying degrees in all projects. The audit also has reservations about some aspects of the T&V extension concept used in these projects which it believes have limited the impact of the extension investment. Some concerns about the Bank's approach in supporting extension projects are highlighted here. The implications of the implementation problems and issues which are elaborated in the text of the report can be summarized into a number of lessons: - Adequate operational expenditures to support salaried staff in the services are essential if effective use is to be made of extension investments; this requires a clear commitment to a realistic program which is consistent with anticipated resource levels by not only the responsible implementing agencies but also the funding institutions. - In instituting an extension system with considerable government investment, the whole gamut of public sector programs which provide technical services to farmers should be reviewed and rationalized; without this, there is likely to be a waste of resources in overlapping services. - xxv - Existence of a dual responsibility of the VEW to separate tech- nical and administrative superiors is likely to lead to dilution of the extension services and reduced effectiveness. While it is highly desirable to have personnel devoted to extension who are free of regulatory functions and who have minimum administrative responsibilities, realistically this may not be feasible in some situations where alternative administrative and political arrangements are deeply entrenched; in any case, extension staff should fully cooperate with other public and private sector entities servicing the farming community. - Allocation of funds for civil works, especially housing, in exten- sion projects should be carefully reviewed to ensure that not only is there a specific need but that full use will be made of such investments. Although a relatively low education level in extension staff can facilitate rapid recruitment and reduce salary costs, this can have trade-offs in the form of poor communication between exten- sion and research staff and weak analytical or responsive approach in dealing with farming problems and interactions with farmer groups. - The selection process of CFs has a very marked influence on the effectiveness of the CF as a diffusion agent; unless a system is instituted to define existing groups of farmers with similar resources, values and attitudes ("reference groups") and select an influential individual within that group as CF, the effectiveness of this approach will be substantially reduced. The use of a reference group rather than a CF as the focal contact point of extension is likely to have major advantages in dealing with communities with more traditional values in risk-prone environments ar- when dealing with more complex, interacting technologies; t... reference group approach allows the use of sociological principles and techniques to assist in technology adoption, emphasizes programming extension visits around nominated problems rather than technical messages, and increases the importance of the regular training/discussion sessions between VEWs and their technical support and training staff. Extension recommendations should not be developed or promoted on the basis of improved production, but must always be based on an analysis of the costs and benefits with full account taken of production and marketing risks. Extension recommendations can rarely be effectively programmed at the mac.o (state or provincial) level and must be adapted to the circumstances of the farming systems in each agroecological zone; this implies adjustments, at least, at the district and sub-divi- sional level. - xxvi - The recommendation which is finally presented to farmers should cater to the various circumstances of farmers with respect to resources and access to complementary services rather than to a standard package; this implies that an extension system which promotes uniform messages over large areas is only applicable to communities or farming systems which have low-risk production environment, similar production resources, and easy access to the complementary services required for profitable adoption. The perception by farmers of the front-line extension worker as a problem solving agent is of paramount importance for effectiveness of the extension program. Basic qualifications and skills of the VEW and the capability of the SMS staff charged with providing technical back-up and training for the VEWs affect farmers' perceptions. Thus the development of an effective SMS cadre needs much greater attention during project implementation where major deficiencies have tended to occur in the numbers of SMS, their training and experience. M&E will not be cost effective unless it is used by management; it should demonstrate substantial returns on the incremental extension investment to gain continued funding support; in-depth analysis of timely M&E reports at all management levels is necessary, and the data should be collated into progressive series to indicate changes in performance or results over time; wherever possible, specific areas should be selected for reliable, medium- term assessment of costs and benefits of the new extension investment compared to traditional or alternative forms. Apart from the direct operational lessons cited above, the experience of the extension projects in India and elsewhere suggest that it would be desirable to consider the following in supporting future extension projects: - The level of economic benefits from investment in extension services will continue to be debated because of the difficulty in linking the investment to specific production benefits. It is generally agreed, however, that extension services can enhance the rate of adoption of productivity increasing technology if that technology exists. - Heavy investment in extension services should not necessarily receive higher priority than other services. The extent of investment in extension should relate to (a) the importance of agriculture in the economy and social stability; (b) the current agricultural productivity relative to its potential considering the resources and technology available; (c) the relevancy, of extension services as an overall constraint to increased agricultural productivity; (d) the existing level and efficiency of complementary public and private sector services in the - xxvii - agricultural sector (credit, input supply, marketing and processing); (e) the public funding resources likely to be available on a sustained basis to support sectoral services; and (f) a logical apportionment of public sector funding amongst the sectoral services, including extension, in a complementary support program which is designed to ensure that the impact of one service is not seriously limited by a deficiency in another. There is no single management system which is most appropriate for extension services in all circumstances. This will depend on the characteristics of the production environments, the farming communities and the farming systems, the availability of relevant technology for each and the status of production services, infrastructure and funding prospects as mentioned above. It is unlikely that a rigid management system with fixed procedures and staff-farmer ratios will be the most relevant or justified on a national or statewide basis at a given point in. time. - There are a number of measures which are conducive to successful extension programs regardless of the details of the management system used. These include inter alia means to define technical and socio-economic constraints in farming systems in agroecological zones; effective research-extension linkages to ensure the availability of relevant technology; programming of extension activities to meet local circumstances and enable analysis of effectiveness for management purposes; constant updating of front-line technical staff on practical technical issues; training of staff in simple farm economics, risk cons iderat ion.s and the social sciences as they affect farmer behavior and adoption response; and adequate operational funding arrangements. However, Bank staff should resist the easy option of adopting a standard package in assisting governments to develop improved extension services; much greater emphasis should be given to the identification and preparation phases for projects concerning agricultural services, so that the most appropriate and sustainable systems are used for each set of circumstances. Special care should also be taken in designing and interpreting the replicability of "pilot" extension projet-ts because of the possible differences in circumstances outside the pilot area. Continued availability of appropriate technology is necessary to sustain an intensive extension program. This principle not only has relevance for extension design in particular agroecological zones, but also implies that an effective, applied and adaptive research system should be in place to support any substantial extension investment. Research and Research-Extension Linkaxe. The NARP has helped provide a framework for more effective agricultural research in India. However, a number of issues need to be addressed to consolidate this progress; many of these issues have relevance for other national research projects: - xxviti - The zonal stations should be accepted as an integral and key part of the agricultural research system. Staffing of zonal stations has been a persistent problem. There is a need to (i) rationalize the sanctioned research positions on zonal stations and substa- tions. (ii) institute a substantial hardship post allowance for facility-deficient areas; and (iii) recognize staff appointments to zonal stations as an important component of their performance, and career. Improvements in directing research to be more responsive to the requirements of the major farming systems are required. Increased emphasis also needs to be given to a better understanding of socio-economic constraints in farming systems and to non-tradi- tional research areas of water management, animal husbandry, agroforestry and farm management. The introduction of a transparent system for prioritizing research which takes into account potential economic impact is likely to enhance the importance of directing research to meet farmers' needs. Adopting a farming system orientation does not always necessitate the adoption of all the stages of discreet farm system research (FSR) projects, but implies that applied and adaptive research must be responsive to the real technical constraints which have been diagnosed in the important farming systems. The recent zonal status reports, which are an improvement on earlier reports, need to complement the technical description with an elaboration of the important socio-economic constraints in farming systems. With the development and increasing importance of the zonal research stations, ICAR should review the role of the 166 regional substations attached to the central and national institutes pro- gram, as it would appear logical to make more use of the SAUs and their zonal stations to carry out some of the functions currently allocated to the ICAR substations. Similarly, it would appear advantageous to transfer most of ICAR's extension programs to the State DOAs now that the research- extension linkage and the zonal research capacity of the SAUs are improving. A more appropriate system to help ensure the relevance of research by ICAR to farmers' problems could be to: (i) transfer the extension programs to the states; (ii) assist the SAUs in building up their FSR capacity, especially in the survey- diagnostic phase, and make use of these findings in the ICAR research planning and review system; and (iii) develop a FSR capability out of selected national institutes. ICAR could beneficially assume a greater role in developing formal systems to ensure improved scientific information flow to and among researchers. The effectiveness of research-extension linkage during implementa- tion was variable among the states. Improvements can be made in the four project components directed at improving this function -, seasonal zonal workshops, monthly training sessions, joint research-extension field visits and farm adaptive trials. PROJECT PERFORMANCE AUDIT INDIA WEST BENGAL AGRICULTURAL EXTENSION AND RESEARCH PROJECT BIHAR AGRICULTURAL EXTENSION AND RESEARCH PROJECT KERAIA AGRICULTURAL EXTENSION PROJECT MAHARASHTRA AGRICULTURAL EXTENSION PROJECT TAMIL NADU AGRICULTURAL EXTENSION PROJECT NATIONAL AGRICULTURAL RESEARCH PROJECT (CREDITS 690-IN. 761-IN, 1028-IN. 1135-IN AND 1137-IN AND 855-IN) I. I1RNT0DUCION 1. This is a combined audit of the West Bengal Agricultural Extension and Research (Cr. 690-IN), Bihar Agricultural Extension and Research (Cr. 761-IN), Kerala Agricultural Extension (Cr. 1028-IN), Maharashtra Agricultural Extension (Cr. 1135-IN) and Tamil Nadu Agricultural Extension (Cr. 1137-IN) Projects and the National Agricultural Research Project (NARP) (Cr. 855-IN). The West Bengal and Bihar projects, approved in 1977, were primarily extension projects but also included substantial expenditures for research activities. However, the more recent projects in Kerala, Maharashtra and Tamil Nadu states, approved in 1980 and 1981, did not specifically include research components; the NARP, approved in 1978, was expected to provide research support in these states. The projects under audit cover more than 10 years of implementation experience of Bank supported extension and research activities in several states in India.1 2. This is the third audit by OED of the "Training and Visit" (T&V) system of extension to be implemented on a statewide or near statewide basis in India. The first, the Rajasthan Agricultural Extension and Research Project (Cr. 737-IN) was audited by OED in 1985.2 The second audit undertaken in 1986, was a cluster that included the Orissa Agricultural Development (Cr. 682-IN), Madhya Pradesh Agricultural Extension and Research (Cr. 712-IN), Assam Agricultural Development (Cr. 862-IN) and the 1/ The West Bengal and Bihar projects were implemented between 1977 and 1983; the NARP was implemented between 1978 and 1985; the Kerala project between 1980 and 1986; and the Tamil Nadu and Maharashtra projects between 1981 and 1987. 2/ Project Performance Audit Report: India - Ralasthan Agricultural Extersion and Research Project, OED Report No. 5618, April 26, 1985. This project was approved in August 1977 and implemented between 1977 and 1983. -2- Composite Agricultural Extension Projects.3 Through the current audit, OED can build on recent evidence and develop some issues normally not treated in a single project audit. The report is designed to complement themes treated under the most recent cluster audit. Furthermore, the inclusion of the NARP in this cluster provides the opportunity to specifically address the design and implementation issues related to the linkage between research and extension activities. Since these projects closed some time during the past three to six years, the audit will also examine the continued commitment of the states to T&V extension and research through adequate funding and provision of qualified staff. As a result the audit will address the issue of sustainability of the extension and research investments supported by the Bank in India. Agriculture Extension and Research ini India Prior to the T&V Extension EK21egts 3. Agricultural extension has a long tradition in India. However, in the evolution of Government involvement in the agriculture sector extension did not always receive the focus that it does today. Prior to the initiation of the Community Development Program (CDP) in 1952, the Departments of Agriculture (DOA) in the States were entrusted with the responsibility of undertaking a range of agricultural activities that included research, extension and procurement/production and distribution of seeds, fertilizers and pesticides. Extension did not receive exclusive attention. The CDP was initiated with the hope that public participation in agricultural development would be strengthened and provide the development emphasis of the administrative and political structure already in place at the district, block and village levels, known as the Panchayati B-a. In 1953, the CDP was supported by the creation of the National Extension Service (NES) block program where extension at the field level was made the responsibility of the Community Development (CDO) and Block Development Officer (BDO). This system developed during the next 8 to 9 years with strengthening of field level extension staff. A substantial number of Agricultural Extension Officers (AEOs) who were graduates, and field level workers (VLWs) were appointed. The VLWs were assigned to each block uader the CDP cadre, whereas AEOs although administratively attached to the CDP, were under the DOA cadre in most states. The VLWs devoted 80% of their time to general agricultural activities, but because of the multipurpose development nature of local administration, VLWs were involved in non-agricultural activities that included farm input distribution. This evidently left the VLWs with very little time to undertake extension. I/ Proiect Performance Audit Report: India - Orissa Agricultural Development, Madhya Pradesh Agricultural Extension and Research. Assam Agricultural Development and Composite Agricultural Extension Projects, OED Report No. 6281, June 25, 1986. The Composite project included the states of Haryana, Gujarat and Karnataka. The first three projects were approved in 1977, the Composite in 1978. Madhya Pradesh and Assam projects were implemented between 1977 and 1983, Orissa between 1977 and 1984 and the Composite project between 1978 and 1985. -3- 4. In 1960, the GOI with the support of U.S. Ford Foundation introduced a new program on a pilot basis in 15 districts. It was called the Intensive Agricultural District Program (IADP). This was in response to the growing realization in the 1950s that India's population growth was outpacing incremental food production. IADP focussed on extending technology to farmers, but was not strictly an extension program. It was based on two main assumptions:4 i) that relevant technology was available, but that farmers were not using it because of ignorance or non- accessibility of inputs; and ii) that the known relevant technology was best extended by pushing a "package of practices" as opposed to single interventions. In order to provide the administrative support to IADP, the CDP extension staff was again strengthened. Each block within the district was provided with more AEOs and VLWs. Programs were undertaken to improve farmer training, provide inpat supplies, develop minor irrigation and expand field demonstration. In each pilot district, a pilot project officer (from DOA) was assigned as Project Leader and he was supported by 3-5 subject matter experts in implementing the program. This meant working closely with CDP personnel, interacting with the relevant agricultural institutions and servicing farmers. The main interaction with farmers was through the development of a crop plan worked out with every cultivator who was interested in cooperating. The Project Leader, in turn, was responsible to the District Collector who ensured coordination between DOA and CDP. 5. The impact of the IADP pilot schemes began to show that, where irrigation or rainfall was assured, sizable production and yield increases were possible through extension and input supply. It also became clear that apart from irrigation and fertilizer availability, factors such as shift in cropping pattern, marketing facilities and changes in relative prices were important contributors to increasing agricultural production. As a result of this experience, in 1964 GOI initiated the Intensive Agricultural Areas Program (IAAP) in 114 districts of the former National Extension Service (NES) blocks. An additional AEO was added to each block, but unlike IADP no Pilot Project Officer or subject matter specialists (SMS) were provided at the district level.5 6. In addition, extension activities were part of a number of other programs and schemes during the mid-1960s. GOI undertook a large number of centrally-sponsored commodity schemes that contained their own extension activities. Most of these are still in operation today. The crops schemes included: minikit input distribution for foodgrains, rice, wheat, millets, pulses; intensive cotton development program; intensive jute development program; national oilseed and pulses programs; and special program for rice production in the eastern states. These programs were meant to ensure that V Charles H. Antholt: Strategic Issues for Pakistan's Agricultural Extension: Looking Back to Look Ahead, January, 1990 (draft). / India Agricultural Extension Sector Review, India Agriculture Operations Division, World Bank, Report No. 6950-IN, January 15, 1988 (Annex 1). funds would be made available by GOI to activities that were identified as national priorities. These schemes include separate steff, seed production facilities, subsidies for pest control, demonstration, minikits, etc. 7. Rural cooperatives became more popular in India during this period. Cooperatives increasingly began to handle input distribution, particularly fertilizer distribution. This allowed DOA staff in the States to give more time to extension activities. New extension staff (AEOs and VLWs) were provided within each state DOA on the basis of area of crop covered, with the number of extension staff varying according to whether the land was irrigated or rainfed. Other beneficiary-oriented programs such as the Small and Marginal Farmers' Development program (1969-70) were also initiated during this period. The staff of each one of these programs worked independently of the others. In a number of cases an array of functionaries existed at the village level, often with overlapping responsibilities. 8. In sum, by the early 1970s agricultural extension activities undertaken by the state DOAs appeared to be fragmented and uncoordinated and their extension activities overlapped with one another. Staff participation in extension was through both single crop schemes as well as multipurpose rural development programs. This resulted in reduced effectiveness and serious inefficiencies. At the same time, extension staff in the states increased sharply in number, some under CDP, others under DOA and still others under DOA but administratively responsible to the CDP. Most staff belonging to CDP worked part-time on extension and were poorly trained in agriculture and extension. 9. Through all of this development, research remained removed from extension. It was undertaken mostly .n state colleges and DOA research institutes. Over time State Governments, with GOI support began to establish autonomous State Agricultural Universities (SAUs) with the hope of slowly taking over research from the DOAs. The establishment of SAUs led to increased isolation of researchers from farmers' problems. Researchers focused mainly on basic research and less on field level problems. New extension sections were created at SAUs and mandated to undertake direct extension activities. Linkage with research was extremely poor at the time of early Bank involvement in extension in India. Bank Involvement in Agriculture Extension and Research in India 10. T&V extension was initially introduced in India in the early 1970s, and Bank involvement to date has come in four stages.6 The first / 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. -5- was through three Bank-assisted irrigation projects on a pilot basis in Rajasthan and Madhya Pradesh.7 The initial positive results of this approach for irrigated agriculture led to replication of the approach for rainfed areas, but first on a limited basis.8 This was followed in the second stage when the statevide introduction of T&V extension began in earnest in 1977. That year, T&V extension projects were introduced in six states. These were in the eastern states of West Bengal, Assam, Orissa and Bihar as well as in Rajasthan and Madhya Pradesh. All six projects included research zomponents. 11. The third stage started in 1979 with the introduction of the statewide system in Haryana, Gujarat and Karnataka through the Bank- financed Composite Agricultural Extension project. These were the first statewide T&V extension projects in which extension was the sole component.9 In 1980, Kerala was the next state to have a T&V extension project. In 1981, three more projects were approved in the states of Maharas.tra, Tamil Nadu and Madhya Pradesh, the last being a follow-on project. In 1982, a T&V extension project was approved in Andhra Pradesh. The fourth stage began with three recent National Agricultural Extension Projects (NAEP) approved in 1984, 1985 and 1987. The first, NAEP I, provided second phase support for Rajasthan and Orissa and third phase support for Madhya Pradesh. The second, NAEP II, included the introduction of T&V system in Jammu and Kashmir and second phase support for Gujarat, Karnataka and Haryana. The third, NAEP III, included introduction of the system in Himachal Pradesh, Uttar Pradesh and Punjab and second phase support for Assam and Bihar. A fourth proposed NAEP has not been approved to date. 12. With the approval of NAEP III in 1987, the T&V system of agricultural extension has been introduced in 17 major states of India through 15 Bank-assisted projects. In this serse the Bank has made a strong commitment to the T&V approach in India. Of the 17 states, nine have had a first phase, seven have had a second phase and one has had a third phase project. Four of the five T&V extension projects under audit are in states which have had a first phase project and only one state (Bihar) has had support for a second phase project. Bank lending from the Z/ The Rajasthan Canal Command Area Development (Cr. 502), the Chambal (Rajasthan) Command Area Development (Ln. 1101) and the Chambal (Madhya Pradesh) Command Development (Cr. 562) projects. ./ An earlier project in West Bengal (Agricultural Development Project, Cr. 541-IN) mainly for minor irrigation included a component to reorganize and strengthen the extension system along the T&V lines in six districts. The initial positive results led GOWB to extend the reorganization to most of the other districts in July 1975, West Bengal becoming the first Indian state to introduce the system on a near statewide bh3is. q/ The National Agricultural Research Project (NARP) approved in 1978 was expected to provide research support to all T&V extension projects approved after 1978. - 6 - 15 projects amount to US$373 million.10 Total costs from these projects at appraisal were expected to total US$660 million. In addition to these very large project costs, GOI and the various State Governments continue to incur further expenditures in order to finance the increase in staff and incremental operating costs after project completion. II. PROJECT FORMULATION, OBJECTIVES AND DESIGN 13. The five extenston projects were an outgrowth of a number of factors that included the initial impressive results of T&V extension in irrigated areas of Rajasthan, Madhya Pradesh and West Bengal, a shift in Bank strategy for India away from multicomponent integrated package projects towards more specialized designs such as extension and research, and the urgency of developing the agricultural potential of rainfed areas. In the decade prior to the approval of these extension projects growth in agricultural production had come mainly from irrigated agriculture (which accounted for about 25% of the cultivated land) while growth from rainfed agriculture (which accounted for 75% of the cultivated land) remained stagnant. Thus increasing agricultural production and yields in rainfed areas provided a unique challenge requiring a strong emphasis on extension with feedback from responsive and effective adaptive research. That feedback was not expected to be useful if it came from the main campuses of the SAUs since they were removed from field level and location-specific problems. NARP was designed to provide feedback from research undertaken at the zonal levels of the SAUs which were defined on the basis of agro- climatic zones identified in each state. Extension Projects 14. In spite of the very different administrative structures and the variations in agroecological zones between the five states, the overall objective of the extension projects was rather similar. All five projects souyht to achieve early and sustained increases in agricultural production in the respective states through the reorganization and strengthening of the extension service of the DOA. Two of the projects - West Bengal and Bihar - also had a component to develop and strengthen adaptive agricultural research, while the later projects in Maharashtra, Tamil Nadu and KeXala relied on NARP to provide the research support at the state level. 15. The common characteristics of the extension projects formulated along T&V lines were: i) the appointment of full-time village extension workers (VEWs) who were to work exclusively on agricultural JO/ Part of this was directed at the development of a complementary research capability in the earlier combined Extension and Research projects. -7- extension, with each VEW being responsible for about 800-1200 farmers divided into 8 to 10 groups; ii) the establishment of a single line of command between full-time VEWs and extension headquarters in a unified extension system, responsible for all extension advice to farmers; iii) the selection of contact farmers (CFs) for the dissemination of information from the VEWs to the farming community at large; iv) the establishment of a fixed and regular cycle of fortnightly visits by the VEW which is supervised by the VEW's supervisor; v) the use of messages that are immediately relevant to the farmers' operations, and which are simple, practical and concentrate on the most important crop in the locality during each VEW visit; vi) regular in-service training of the VEWs at fortnightly intervals in programmed messages and the provision of specialized, technical higher-level staff (SMS) to provide the required technical input, in conjunction with the experience available from the research institutions; vii) improvement of linkages between extension and research through continuous feedback from the VEWs to the technical experts and researchers of the farmers' circumstances and technical production constraints; and viii) development of monitoring and evaluation procedures. 16. The T&V extension system embodied in these projects differed from the earlier approaches to extension in India mainly in the responsibilities assigned to the field level worker and the organizational structure of the State's Extension Services. While in the past a VLW had to undertake many non-agricultural and non-extension activities, these projects sought the full-time commitm'gnt of the VEW, devolving him of other responsibilities in his primary task of disseminating technology to the farmers. Organiza- tionally, the institution of a single line of command, whereby the VEW would be both technically and a4ministratively supervised by the Extension Headquarters, was also a significant departure from previous approaches. A number of underlying assumptions implicit in the design of these extension projects are discussed in the following paragraphs. 17. First, the projects assume that appropriate technology was available within India at the time and that the reservoir of untapped -8- proven technology simply needed to be disseminated. However, there was no attempt made in any of the SARs to define any of the technical packages that would be extended to the farmers, and ipso facto no attempt to demonstrate their profitability. The projects essentially assumed that there was enough technical innovations in stock to initiate the extension projects on a statewide basis, and that more would be forthcoming through the research components of the West Bengal and Bihar projects and NARP. 18. Second, implicit in the design of these extension projects was the assumption that complementary services such as fertilizer supply and improved seeds were readily available, that small farmers had ready access to credit and that existing marketing serv*ces were adequate; the underlying assumption was that lack of knowledge was the only binding constraint. 19. Third, it was assumed that the CF approach, a crucial element of T&V extension, would enable a rapid diffusion of technology to other farmers. Acceptance of the CF approach implied that this particular approach, rather than other techniques such as working with groups, mass media dissemination and commercial services, was the most effective way to disseminate technology. 20. Fourth, it was assumed that a single line of command between full- time VEWs and extension headquarters in a unified extension system could be established at reasonable cost in terms of time and money. This implied that staff from the community development and other rural schemes could be painlessly transferred and easily trained in extension operations. This required that the administrative structure of the state governments at the village/block level, which had evolved out of the political process, could be changed by a bureaucratic decree to suit the single line of command requirement. 21. Fifth, it was assumed that the recommended system of linkage with research would ensure that technology would pass through to the farmers and that the reverse feedback from the farmers through the extension staff to researchers would be effective. 22. Sixth, it was assumed that to the extent staff time was reallocated to extension, the tasks given up were less valuable than the new extension tasks. The State Government (with assistance from GOI) would have the political will and financial commitment to continue with the new system once the funds from the Bank-assisted projects had been spent. 23. The design and formulation of the projects focussed on the management aspects of extension rather than on systemic issues. Issues such as the availability, profitability and acceptance of specific technical packages, the ability of VEWs to effectively undertake the technology transfer process and to correctly identify farmer constraints and feed this back into the adaptive research-message formulation system, and the possible variable response to the T&V methodology according to the - 9 - type of production environment and farming community were not addressed. As a result, the projects supported the introduction of a largely top-down approach, in which the government bureaucracy was expected to be the principal actor in passing technology to the farmer through an organized chain of command. 24. The Staff Appraisal Reports (SARs) are largely silent regarding the varying capability and willingness of the states to implement the proposed restructuring of the administration (especially the DOAs). The SARs are repetitive, except for some variation in the number of farmers to be serviced by a VEW in each state. National Agricultural Research Project (NARP) 25. NARP was expected to strengthen the capabilities of the SAUs for conducting location-specific, production-oriented research on the basis of identified agro-climatic zones.11 NARP was to strengthen need-based research and thus emphasize a closer link between research and extension at the grass root level. It had no specific physical targets other than two sub-components to strengthen the training facilities of the Indian Council of Agricultural Research (ICAR), although project costs were based on an assumed 56 sub-projects. The IDA Credit was a financing facility which was available to any State Agricultural University (SAU) which met the project criteria. These were that the SAU: (a) complied with GOI/ICAR standards and procedures; (b) produced a background paper containing a critical analysis of the SAU and its research responsibilities; (c) was willing to establish mechanisms for research administration and recruit adequately experienced staff; (d) established satisfactory linkages with state extension systems; and (e) agreed to participate in a review of research in the SAU service area. 26. The credit proceeds were to be disbursed against sub-projects prepared by the SAU and approved by a Project Funding Committee (with representation by ICAR, Central DOA and two SAUs), as well as some administration and training costs. There was no restriction on the number of sub-projects that any one SAU could implement once the research review was completed. Those sub-projects were to cover: (a) research administration and basic research at SAU campuses; (b) strengthening Zonal Research Stations (ZRS); (c) strengthening zonal sub-stations; and (d) ad hoc research. 11/ Each state was divided into a number of agroclimatic zones in accordance with physical resources and climatic characteristics. - 10 - It was expected that 75% of the credit would be used for applied research, particularly item (b), and not more than 10% be for basic research, item (a). Each sub-project would be financed for five years. To facilitate the review process, a project unit under the Deputy Director General (Education) of ICAR was to be established as the secretariat to the Project Funding Committee. In addition, an Interdisciplinary Scientific Panel was established to provide technical advice to the Committee. C. NARP and the Extension Projects: Research-Extension Linkage 27. The research-extension linkage was built around the agroecological zone.12 This made use of the following mechanisms: (i) pre-season zonal workshops in each zone involving SAU scientists, extension staff and farmers to review the previous season's research results and field experiences and decide on future research proposals and priorities and the program for field testing; (ii) monthly training workshops involving SAU scientists and SMS and other senior extension staff, where they could also analyze priorities, results of experiments and their relevance and plan VEW training; (iii) joint field visits by diagnostic teams of research and extension staff for on-the-spot discussions with farmers to study their field problems and reasons for adoption or non-adoption of recommended practices; and (iv) jointly conducted farm trials which were the last stage of research verification and the first stage of technology transfer. III. IMPLEMENTATION 28. Implementation performance among the six projects varied widely and is adequately covered in the PCRs. The audit concentrates on selected implementation issues which are common to the extension projects; these include expected and actual staffing, project costs and financing, staff utilization, VEW educational levels, contact farmer selection and monitoring and evaluation performance. The NARP project is analyzed principally in terms of its interrelationship with the extension projects. 29. The five extension projects fall into two groups. The first group, West Bengal and Bihar, covers two extension/research projects with severe implementation problems. Despite the fact that the Closing Date of the West Bengal project was delayed 30 months and that of the Bihar project seven months, neither project achieved the expected statewide coverage, and 43% of the former and 48% of the latter credit had to be cancelled. The second group covers Maharashtra, Kerala and Tamil Nadu, which were successfully implemented and fully disbursed, although the Haharashtra project closing was delayed by 12 months. 12/ There was usually also a formal state level meeting between SAUs and DOA once or twice a year to review ongoing and future research and field trials and extension recommendations. - 11 - Expected and Actual Staffing Requirements 30. The Extension Service Staffing: At the time of appraisal of all the extension projects except Maharashtra, it was expected that the majority of extension staff positions would be filled by reallocation of existing staff and that the number of new posts would be relatively limited. Reallocation of these existing staff was a condition of negotiations in all the projects except West Bengal. The following table provides a comparative analysis by states of the number of staff estimated to be required and the achievement which varied greatly between states at project completion. Table 3.1 EXTENSION STAFF REQUIREMENTS AND ACHIEVEMENTS West Bihar Kerala Maharashtra Tamil Total Bengal ____ ___Nadu_ Total staff required: La Grade A & 4,000 8,980 1,952 6,000 4,777 25,709 Grade B 650 1,508 569 1,035 1,098 4,860 Grade C 77 229 55 69 82 512 Incremental Needs (Percent)1L Grade A nil nil 32 49 28 19 Grade B 23 24 28 55 29 32 Grade C 70 48 47 64 72 57 Project Accomplishment (Percent) Z4 Grade A 48 37 100 96 84 49 Grade B 66 32 89 91 100 69 Grade C 62 18 116,L 107,L 112,L 62 La Staff Appraisal Report assessment. /b Grade A - Village extension worker (VEW). Grade B - Block and sub- division staff, as Agricultural Extension Officers (AEO) and Subject Matter Specialists (SMS). Grade C - District and Headquarters staff, mostly as SMS. /c Difference between expected staff availability and requirements as estimated at appraisal. Lg Ratio of staff in position at project completion to original estimate of staff requirements. /e Additional districts were created during project implementation, thus increasing the number of district level staff required. - 12 - 31. In West Bengal, it was intended that all VEW posts would be filled by the transfer of staff, a process already started in 1975, either from the Community Development Department or from the special crop schemes implemented by DOA. Adequate sub-divisional staff (other than SMS) were also available. Early in project implementation, however, there was widespread staff objection to transfer from existing positions and the state government was obliged to curtail its program for transferring staff to the VEW posts. An overriding staffing problem was the raising of court injunctions by some public servants who considered they would be disadvantaged by the transfer program. Many of these appeals were still unresolved at the time of audit, thereby affecting the appointment process for the affected positions. As a result 1600 new VEW positions had to be sanctioned and virtually all staff needed to be trained. By the time of project completion only 61% of the expected VEW positions had been sanctioned and of these only 79% were filled, resulting in virtually no coverage in some areas and low coverage in others. At the sub-divisional level also, despite the appraisal expectation that adequate staff were available, one-third of posts were still vacant at project completion. Over half the district level SMS positions were also unfilled. 32. In Bihar, a similar situation developed despite the transfer of staff from the Community Development Department being a condition of negotiations. The transfer order was passed but not implemented. Ten months after effectiveness, village level staff in five of the 31 districts in the state were transferred from community development work to extension, although as they remained under the administrative control of the District Magistrate (called CEO elsewhere) they continued to perform many non- extension duties. Two years later, some one-third of these staff were transferred back again to community development work leaving the five districts very short-staffed. A further 11 districts were added in the third project year but the remaining 15 districts were never covered. Many posts remained vacant in the 11 districts because of the slow administra- tive procedures for appointing staff. At project completion, only about one-third of the expected number of village and sub-division staff were in position. Even in the 16 districts covered, 32% of the VEW posts and 47% of the AEO posts were unfilled. 33. In Kerala, although there were major delays in getting VEW staff into position, and many of them were only on a temporary basis, all VEW posts were filled by the fourth project year. Less than 10% of the new VEW positions were filled by transfer from community development work. There were difficulties in filling the AEO posts because of a shortage of new graduates and frequent transfers and promotions of existing staff, but by project completion 83% of the AEO posts were filled. All the higher level posts were filled. 34. In Maharashtra, the state government made strong efforts to requisition staff from other departments and recruit new staff so that it could succeed in establishing the new extension system throughout the state. This was indeed achieved within 16 months of project start-up, - 13 - compared with the three years expected at appraisal, although at a slightly lower farmer:VEW ratio than planned. Nearly half the new VEW posts were filled by assigning community development staff to extension. The rapid creation of this specialized work force, however, necessitated a major training program which was carried out in the remaining years of the project and is still continuing. 35. In Tamil Nadu, as in West Bengal and Bihar, the transfer of staff from community development work posed difficulties. When some staff refused to transfer the government decided not to attempt to fill the new VEW positions by transfer but increase the farmer:VEW ratio and fill the other VEW vacancies by new recruitment. By the third project year, this revised target was met, representing just over 80% of the VEW positions expected at appraisal. This ratio remained more or less unchanged until project completion. The recruitment of graduate level staff was satisfied by transfers within the department and the turnout of new graduates from the agricultural university. 36. This very varied experience of five states implementing what was essentially an identical administrative reform enables the following observations. (a) In terms of numbers, the magnitude of the reform required to set up the new extension system was not so great that it should have been beyond a state's capacity to implement. The Maharashtra experience proves that it could be done. The transfer was apparently accomplished without problems and faster than the SAR schedule, although the VEWs remained within te jurisdiction of the community development system and continued to be employed on some non-extension activities. The bhar PCR's conclusion (summary, para. 3) that "the project formulaticn seriously underestimated the magnitude of administrative problems related to changes within the civil service ... did not recognize the time it would take to introduce and consolidate a major reorganization", appears misdirected. What was underestimated was the state's willingness and/or ability to enforce a reform when it did not have full polriical and administrative support, and also how strongly entrenched (and valued) were the staff and management of existing programs. The main staffing problems arose from the inability of state governments to transfer staff into extension.13 (b) As community development staff did not transfer to the extent anticipated, a much bigger intake of new staff than foreseen at appraisal 13/ A similar situation is reported in the PPAR of the Rajasthan Agricultural Extension and Research Project (Report No. 5618, April 26, 1985). Community Development staff were transferred on "reverse deputation" which meant they could (and many did) return to their parent department if they were dissatisfied with their work as VEWs. The cluster audit on four other extension projects in India (Report No. 6281) found that similar problems of reverse deputation were experienced in the states of Gujarat, Haryana, Karnataka, Andhra Pradesh, Madhya Pradesh and Orissa. - 14 - was required (Table 3.1), all of whom needed training. As a result, the five projects underestimated the amount of training that would be needed and thus made inadequate provision for it. This need for greats-r numbers of new staff also had a significant impact on DOA and state budgets. Instead of transferring the salary budget between departments, along with staff, new positions had to be created and budgeted for the new extension system. This had not been anticipated at appraisal nor included in project costs. (c) In some cases Bank supervision put excessive emphasis on states to fill vacant posts, regardless of the quality of the staff that were available. This was especially true at the VEW level, but was also repeated at higher levels. In Kerala, for instance, VEW vacancies were often filled by untrained temporary staff who had to be replaced every six months in a temporary status, while the Public Service Commission went about its slow process of official recruitment of those with adequate qualifications.14 37. Monitoring and Evaluation Staffing: The staffing pattern of the monitoring and evaluation units in the five states closely paralleled the pattern for the extension service. In West Bengal and Bihar, the units were not formed until project implementation was well underway. They remained short of qualified staff and few reports were published. In the other three states, the units were formed in the first year of implementation although it took about three years to staff them adequately. 38. Staffing of Research Components: In the West Bengal Project, research was the responsibility of the DOA. The project provided funds for 159 incremental research positions of which or,-y 32 (20%) were filled during implementation. In addition, over half r'e 112 professional posts already sanctioned remained vacant. In Bihar, daptive research was also the responsibility of DOA. The scope of the research program was reduced from appraisal expectations in line with the reduction in the extension coverage. Out of the 156 additional professional posts to be provided, only 84 (54%) were filled. At the field assistant level, exactly half the 432 posts were filled. Project Costs and Financing 39. Appraisal estimate vs. actual project cost comparisons and credit approvals and final cancellations for the five extension projects are shown in Table 3.2. In West Bengal, the smaller actual cost shows credit 1_4 In the case of Bihar and West Bengal, however, when faced with staffing difficulties the DOAs preferred to raise the number of farm families per VEW while supervision missions continually pressed to keep the VEW:farmer ratio intact, even if this meant that other groups had to be left uncovered. -15 - cancellation from a scaled down project; the smaller actual cost in US$ was due to the devaluation of the Rupee between early 1977 and 1983. In Bihar, the much higher actual cost was caused entirely by larger than anticipated increase in salaries due to inflation and the extended project period. In Y4ga1a, the much higher actual cost figures reported in Table 3.2 are not %eflective of the true cost overrun. Data are not comparable because actual figures for salaries and operating cost are total rather than incremental. In Maharashtra, the cost overruns were mainly due to larger than anticipated staff cost, whose phasing was faster than planned, and higher civil works cost due to slow implementation. In Tamil Nadu, the higher actual costs were due to construction of larger than planned number of buildings for information-cum-training centers and the purchase of more vehicles and equipment than SAR targets. Table 3.2 PROJECT COST AND CREDIT AMOUNT La Credit Appraisal Estimate Actual ABmBt Cancellation (US$M) (Rs M) (US$M) (Rs M) US$ M US$ M West Bengal 28.2 195.2 18.2 180.9 12.0 5.15 Bihar 16.0 144.0 32.7 297.0 8.0 3.83 Kerala 15.3 128.4 25.1A& 284.5 10.0 - Maharashtra 38.2 306.1 40.8 463.5 23.0 - Tamil Nadu 4.6 364J. 3.8 431. 2 = Total 143.3 1138.5 154.6 1657.2 81.0 9.06 LA All costs and credit amounts are in current values. L Actual not comparable to SAR because actual figures for salaries and operating costs are not on incremental basis as in the SAR. 40. Table 3.3 shows that about 60-70% of project costs consisted of incremental salaries and civil works. Of this, incremental salaries at appraisal were expected to be between 27-50% of project costs, but turned out to be even higher at project completion. In Maharashtra, incremental salaries were as much as 60% of project costs at completion. This higher share of incremental salaries together with the higher project cost figures shown in Table 3.2 indicates a very high cost overrun for the incremental salaries component. A considerable proportion of the civil works costs, ranging from 36% in West Bengal to 100% in KeXala., consisted of staff housing. The share of incremental operating costs were between 13-17% of project costs at appraisal and were marginally smaller at completion. There was no significant cost overrun in rupees for incremental operating costs. - 16 - Table 3.3 EXPECTED AND ACTUAL PROJECT COSTS BY CATEGORY OF EXPENDITURES (PERCENT) /a Vest Tamil Beggal Bihar Erala faharashtra Nadu S Actual SA& Actu SAR Agual SAR Actual LM Actual Incremental Salaries 27 NA 31 NA 41 NA 50 60 41 49 Land and Civil Works 40 NA 30 NA 28 NA 17 17 27 24 Vehicles and Equipment 14 NA 24 NA 11 NA 10 8 11 9 Training and Conaultancy 6 NA 2 NA 5 NA 6 2 8 7 Incremental Operating Costs 13 NA 13 NA 15 NA 17 13 13 11 Total 100 NA 100 NA 100 NA 100 100 100 100 LA For West Be1al Bihar and Cerala comparable breakdown of SAR and actual costs by category are not available. Actual staff salaries and operating costs are not comparable vith corresponding SAR cost estimates because disaggregated actuals vete not maintained on an incremental basis as was done in thn SAR. 41. Disaggregated actual project costs are not avalable for West Beneal, Lii_L or Kerala because no separate account of incremental staff salaries was kept. In Maharashtra and Tamil Nadu, where these costs were kept separate, actual incremental salary costs were much higher than expected at appraisal. In rupee terms, they were 89% hi her than the SAR estimate in Maharashtral5 and 47% higher in Tamil Nadu.1b The effect of these large overruns on the sustainability of the extension system is discussed in Section IV. Financing 42. IDA credit disbursement was very closely tied to the ability of the state to recruit and retain staff. This was reinforced by the fact that, except for West Bengal where the nroject did not finance salaries, 70-75% of credit proceeds were allocated to salaries and civil works. This is shown in the Table 3.4, where actual disbursements are also compared. j5/ Project Completion Report, Maharashtra Agricultural Extension Project (Report No. 8258, December 29, 1989), Table 2. 16/ Project Completion Report, Tamil Nadu Agricultural Extension Project (Report No. 8258, December 29, 1989), Table 2. - 17 - Table 3.A IDA FINANCING - CREDIT AGREEMENT ALLOCATION OF PROCEEDS AND ACTUAL ALLOCATION (PERCENT) L West Tamil Aenal Bihar Kgra1a Maharashtra Nadu Salaries 0 24(119) 49(108) 48(120) 39(92) Civil Works 59(86) 36 (19) 27(124) 22(124) 30(124) Vehicles and Equipment 21(26) 30 (47) 13 (80) 14 (90) 13(108) Training and Consultancy 9 (7) 3 (78) 6 (55) 8 (42) 11(114) Unallocated 11 (0) 7 (0) 5 A 8 fi 7 A Total 100(57) 100 (52) 100(100) 100(100) 100(100) a figures in brackets are percentages of credit allocation eventually disbursed. A disbursement of unallocated amounts are included under the category to which amounts were reallocated. 43. The pattern of credit allocation varied from project to project. In the case of West Bengal and ihar, the credit amount was set at about 50% of expected project costs. In -he other three states it was set at 70% of expected project costs. In the case of West Bengal and Bihar, civil works and vehicle and equipment costs were expected to utilize most of this 50%; only 10% of eligible salary costs were financed in the B1ha Project. In the other three projects, the lower proportion of civil works and vehicles and equipment, and the higher proportion of total costs being financed, enabled a higher proportion of salary costs to be financed, the actual proportion being adjusted to make the total credit amount equal to 70% of total costs. In both West Bengal and Bihar, nearly 50% of the credits had eventually to be cancelled, in spite of an extension of Closing Dates, largely because of the slow progress on civil works. In all four states, however, where there were disbursements against salaries the category "staff salaries and allowances" was fully or almost fully disbursed. It is likely that the higher external support for the recurrent salary costs partly contributed to the better implementation and performance in Kerala, Maharashtra and Tamil Nadu. 44. There is some indication that the amounts allocated for civil works were in fact excessive, particularly the allowance for VEW housing.17 It was assumed in all projects that, in order to get VEWs and AEOs to live close to their work that housing should be provided. Project costs 2/ This point has also been brought out by the Bank's Agricultural Extension Sector Review for India (Annex 3, page 3) of January 15, 1988 (India Agriculture Operations Division). It comments that extension operations were not greatly affected by the provision of housing. This is based on the similarity in performance between states which housed many extension workers and those that have housed few. * 18 - therefore included an allowance to provide housing for a certain proportion of total VEW/AEO staff; 25% in the case of est Bengal, Biha and Kerala and 20% in Maharashtra and Tamil NadM. During project implementation this proved to be one of the more troublesome components. The works were small and widely scattered and it proved difficult to find contractors willing to build them. Building materials, particularly cement, were scarce. Costs were usually considerably underestimated and the states were reluctant to make the necessary budgetary allocations. No sites were selected prior to implementation and proved very difficult to find; many of those eventually selected were inappropriate for their intended use. Actual accomplishment ranged from 5% of the target in Bihag through 30% in West Bengal, 62% in Maharashtra and 95% and 100% in Kerala and Tamil Nadu, respectively.18 The occupancy rate was low; either the sites lacked basic amenities or were too remote or defective in construction, or they were being used for other purposes such as input stores. They were also financially unattractive to the VEW as moving into a project house meant the VEW lost both the existing housing allowance and was then charged a rent. Staff Utilization 45. The diversion of extension staff at all levels to non-extension duties, although strictly against T&V principles, remained a permanent source of concern during project implementation in all states. This diversion was worst in Maharashtra where staff at district level were under the technical guidance of DOA but remained under the administrative control of the Chief Executive Officer (CEO) of the Rural/Community Development Department. Here VEWs continued to be put to work on non-extension and even non-agricultural duties in some instances such as family planning and small savings programs. This continued despite a letter from the Additional Chief Secretary instructing all CEOs to refrain from requisitioning extension staff to participate in general campaigns. 46. In West Bengal, despite the limited transfer of community development and crop scheme staff, extension staff had to spend considerable time on input supply matters. This was particularly acute at the middle supervisory levels (i.e. at block, sub-division and district level) where responsibility for the full range of DOA services still remained, involving program administration, routine reporting, attending meetings, etc., thus reducing the time available for field supervision of extension. Here too, the concepts of full-time VEW and single line of command broke down, leaving the project as a "modified T&V". 18/ Other audits have found similar problems. For instance the audit on the Rajasthan Agricultural Extension and Research Project (Report No. 5618) found that only 62% of VEW houses were built. The Bank's Agricultural Extension Sector Review (for reference see last footnote) also notes that "the need for staff housing is widely doubted and many cases were cited where, once built, they remained unoccupied." (para. 27) . 19 - 47. In Bihar, the VEW's appointment was only to a district cadre under the district administrative head who was an Indian Administrative Service Officer, rather than to the DOA. The PCR comments that "for historical reasons, and because other branches of government do not have village level staff, VEWs are still considered by farmers to be general as well as technical representatives of Government, to whom all problems should be addressed... it is inevitable that this conflict of interest will continue until other supporting services become stronger" (para. 4.18). In Kerala, very few staff transferred from community development work. A parallel administration system for input supply was set up alongside the extension system within DOA. Despite this, there was strong agitation within the state against setting up the new extension system on a statewide basis on the grounds that there would be inadequate provision of input services when the project was implemented. This was resolved by increasing the number of input supply points threefold. Project implementation, by increasing the demand for inputs, increased the demands on the input supply system, particularly for HYV rice seed. 48. The involvement of extension staff in non-agricultural and agricultural input supply services was not the only diversion of their time. There was also a multiplicity of "special agricultural schemes" to be implemented. Before the creation of a unified extension system each of these schemes had their own staff. Many of these staff were transferred when the new extension service was established but the schemes still had to be implemented. These are time-bound schemes with pressure for target achievements. In West Bengal, for instance, there are about 50 such schemes. Typically they involve the selection of smallholder farmers to whom minikits (40,000 to 80,000 per sub-division) are given, demonstrations are set up and subsidized inputs are administered - all of which have to be reported on. None of these activities was integrated with the normal T&V work schedule of the VEW/AEO and were therefore disruptive.19 49. Much of the time of SMS staff at district and headquarters was taken up in the administration of these schemes and in other routine non- extension duties, not in acting as technical specialists. Most of the VEW training was conducted by SMSs from the sub-division and, where the research extension linkage was strong, these SMSs went direct to SAU staff when they needed help, by-passing the district or headquarters SMS staff. Educational levels of VEWs 50. Most of the general purpose VLWs who were transferred from community development work (including extension) or from special schemes to the reformed extension system were matriculates (10th grade) with either two years of professional training at a rural training institute or 6-12 12/ These single crop-oriented, subsidy-based programs are highly visible and attractive to politicians, policy makers and farmers, particularly as part of their cost is borne by the Central Government. Thus there is pressure on state DOAs to give priority to their implementation. - 20 - months at an agricultural school. Very few had higher qualifications and even fewer were agricultural graduates. Minimum qualifications for w VEWs were supposed to be matriculation plus a two year agricultural diploma, although many VEWa with lesser qualifications were recruited because state-level training institutes had inadequate capacity to produce the required number of diploma holders; the Bilb.r PCR comments that "additional qualification above the matriculation level is seldom a major factor in selection." In Tami Iadu, 20% of VEWs had not reached 10th grade.20 However, in aharashtra 11% of VEWs were graduates (from the old agricultural demonstration cadre), 80% had an agricultural diploma and only 9t did not satisfy the eligibility criteria.21 51. All the projects included training arrangements, with a series of pre-service induction training courses (varying from one month in Maharashtra to six months in Kerala, Bihar and West Bengal), refresher courses, special short courses, twice yearly pre-season training courses and fortnightly training at the sub-division level. For the most part these training programs were satisfactorily carried in terms of numbers, with the exception that plans to give a limited number of VEWs the opportunity to take a BSc. were never met because VEWs usually lacked the qualifications required by SAU. In terms of quality, however, the training suffered badly from lack of adequately trained and specialized SMSs and insufficient participation of senior officers in West Ben-al and Bihar. In the other states, the quality of training, though poor to start with, gradually improved during project implementation, although there was a tendency for the training to become routine without adequate emphasis given to local conditions. Selection of Contact Farmers and Regularity of Visits 52. Studies carried out by the M&E units on the characteristics of CFs indicate that, while the selection by size of farm, age, literacy level and type of irrigation was reasonably representative, many of the farmers selected were unsuitable as CFs. They knew little or nothing about their intended role, and even where they did know about it they did not pass on messages to other farmers. The proportion of CFs whose primar occupation was not farming ranged from 37% in Mrala to 8% in Maharashtra. 2 Some of 20/ Special Study on the Role of Assistant Agricultural Officers (T&V). M&E Unit, Tamil Nadu, 1986. I/ A Report on Evaluation Study of Training and Visit Scheme ir. Maharashtra. Special Study, Report No. 8 M&E Unit, Maharashtra, 1988. 22/ Although it is possible that persons with diversified interests are influential in agricultural technology adoption, a selection of a large number of this type of person suggest a bias towards selection of innovative and entrepreneurial characteristics rather than persons with identified influence potential, which need not be correlated. - 21 - the CFs in Maharashtra were blind or deaf, and even dead.23 According to M&E surveys a remarkably high proportion of them were unaware that they were contact farmers at all; 41% in Tamil Nadu, 35% in West Beneal,24 20% in Ker"lA25 and 12% in &barashtra. Even if they were aware that they had been selected, a proportion did not know what was expected of them and did not disseminate the VEWs' messages to other farmers; 57% in erala, 54% in Maharashtra, 39% in West fetal. Many indeed were not even aware who the VEW was; 15% in Tamil Naduz6 for instance and 32% in West Bengal. 53. These studies were mostly carried out at the midpoint of project implementation and did alert managers to the probzlem. In some cases CFs were changed and VEWs were instructed to make sure that all of them knew what was expected of them, and to pay more visits to non-contact farmers. By the time of project completion many of those figures had improved significantly; for instance, almost all CFs in Kerala, Maharashtra and Tamil Nadu knew who their VEW was by the end of the project and about two- thirds of them in West Bengal and Bihar. 54. The seasonal Monitoring and Evaluation (M&E) studies provide a useful insight into the regularity of the VEWs' visits to farmers. The following table summarizes the available information for the latest available project year: 2/ "A Report on a Special Study of Formation of Farmers' GrouDs and Selection of Contact Farmers", Monitoring and Evaluation Cell, Maharashtra, November 1984. 24/ "Special Study on -baracteristics of Contact Farmers", Socio-Economic and Evaluation Branth, West Bengal, June 1985. 25/ "A Study on the Characteristics of Contact Farmers", Monitoring and Evaluation Unit, Kerala, January 1985. Zk/ "A Special Study on the Selection and Role of Contact Farmers", Monitoring and Evaluation Unit, Tamil Nadu, December 1984. - 22 - Table 3. 5 RE1ATIONSHIP OF VEW AND FARMER (Percent) (latest available project year) West Mahar- Tamil LA B.nsal Bhar Kerala ashtra Nadu Knowledge by farmer CF 68 67 95 94 98 of VEW OF 46 40 57 70 92 Two or more visits by VEW CF 38 51 57 70 81 in previous four weeks OF 19 28 6 27 42 One visit by VEW in CF 24 22 20 15 12 previous four weeks OF 21 18 8 28 31 Knowledge of regular CF 31 46 56 75 81 visiting day of VEW OF 15 20 6 41 53 Visits by VEW reported CF 28 40 99 79 98 as being largely in field OF 18 18 92 43 65 Farmer's opinions as to VEW CF n.a. 88 78 100 being main source of technical OF 44 n.a. 26 48 78 advice /a CF - Contact Farmer; OF - Other Farmers or non-contact farmer. Source: Monitoring and Evaluation Reports (see footnotes 19-23). These figures show a big variation between states in the regularity of VEW visits. In Tamil Nadu, 81% of contact farmers had the intended two scheduled visits in the four weeks previous to the survey, with another 12% getting one visit. In the worst state (West Bengal) the respective figures are 24% and 31% (i.e. nearly half the contact farmers were not visited at all in the previous four weeks). Except for Kerala, visits reported to non- contact farmers were surprisingly high, with 40-73% of the sample reporting at least one contact with the VEW in the previous month. 55. As could be expected, the farmers' knowledge of the regular visiting day of the VEW and the extent to which the visit took place in the field, correlate closely with the frequency of the visits. Monitoring and Evaluation P#rformance 56. M&E units were es.: -olished in all the five states with the objective of conducting two broad types of surveys on a regular basis: (i) those monitoring the operational implementation of the project in order to - 23 - provide feedback to m-nagement and (ii) those concerned with the adoption of technology which also provided information on project impact. In addition special studies on topics/problems of particular interest were carried out as the need arose; the cited studies on the characteristics of CFs are examples. 57. The monitoring studies were regularly carried out in all the states once the M&E units were adequately staffed. However, the use made of them as a source of feedback varied greatly between states. This was partly due to delays in publishing results (often up to 12 months) and partly due to lack of appreciation by management of their value. All the states, except West Bengal, now produce summary findings ("quick reports") focussing on key findings within about two months of survey completion. Only in Tamil Nadu do these appear to be regularly circulated and discussed from state-level plan review meetings, through zonal workshops down to the fortnightly training sessions of the SMS, with their findings used by management in project implementation. Discussions occur intermittently in all the other states and the reports are generally circulated at least down to the sub- divisional level but they are not regularly ised as a management tool. 58. Some of the special studies have been valuable in highlighting problems. The cited 1985 Kerala special study on the characteristics of contact farmers alerted DOA (and later the state government) to the fact that the CF approach to the rapid dissemination of extension messages was inappropriate in Kerala. The election of a new state government in 1987 has led to the development of a new group farming approach (see Annex 1, paras. 9-13). This in turn has led to a modification of the ME questionnaire with a new set of questions now being used. National Agricultural Research Project (NARP) 59. Two aspects of the implementation of NARP are corsidered. The first is concerned with the physical implementation of project works and the * disbursement of IDA funds. The second is the extent of attainment of the basic project objectives of improving the national extension-research system through decentralization to make it responsive to local farm needs and strengthening the research-extension linkage. 60. Physical Imiplementation: Implementation got off to a slow start and by the time of the original project Closing Date ICAR commitment of funds was only 53% of the appraisal estimate (reflecting slow preparation and approval of sub-projects) and disbursement of IDA funds was only 26% of the appraisal estimate (reflecting the slow implementation of each sub- project). The project Closing Date was thus postponed for two years to allow project objectives to be met. By the new Closing Date both commitments and disbursements in local currency terms were very close to appraisal estimates. In US dollar terms, because of currency adjustments, US$7.5 million of the approved credit amount of US$27.0 million was cancelled (only 72% of the credit was disbursed). - 24 - 61. Final accomplishments in initiating subprojects were impressive. Instead of 15 participating SAUs there were 22, each of which produced the required background paper (item b in para. 25), and 16 of which produced a statewide research review. There were 21 sub-projects for strengthening research administration at SAUs as against 15 planned; and 86 research projects were started as against the 56 planned. The progress of sub- project approval is shown in the table below. The project was able to finance the additional sub-projects because the actual sanctioned cost per research sdo-project was only Rs 6.6 million compared with the appraisal estimate of Ra 12.5 million. Table 3.6 * RATE OF SUB-PROJECT APPROVAL PLANNED AND ACTUAL Administrative Research Planned Actual Planned Actual GOI Fiscal Year 1978/79 3 1 6 3 1979/80 4 3 11 2 1980/81 5 4 17 9 1981/82 2 7 13 22 1982/83 1 5 9 12 1983/84 - - - 19 1984/85 - 1 - 19 Total 15 21 56 86 62. Implementation of individual sub-projects was very slow in the first three years of the project, particularly for civil works due to delays in land acquisition, selection of architects, preparation of design and estimates. However, as SAUs gained experience with the assistance of the Project Unit in ICAR, formalities were completed more quickly and progress has been faster. Nevertheless, the appraisal report was overoptimistic about the speed with which the sub-projects could be prepared. Actual completion of all the sub-projects begun in NARP will not be until 1990, compared with the original expectation of 1986. Only 11 of the research sub-projects were fully implemented by the project Closing Date. The slow progress meant that during the five years in which NARP financed each sub- project, little more was done than complete the basic framework within which a better research program could be developed. 63. One major problem, which has still not been resolved, is the posting of staff to zonal research stations (ZRS) located in remote areas - 25 - where health, education and other facilities are scarce. There are few special incentives to staff to move to those unattractive places and vacancies were hard to fill. For instance in mid 1985, out of some 1,000 incremental staff positions created, 25-30% of them were not filled.27 Almost four years later the Mid-Term Review of NARP 1128 (para. 2.25) noted that vacancies at ZRSs ranged from 20-30%, and that there was a dearth of staff in certain disciplines, with a recruitment bias towards cereal breeding and a conspicuous deficiency in agricultural economist posts. It also noted (para. 2.26) that in some sub-projects, once IDA/ICAR funding ceased, the state governments refused to fill vacancies in order to save money, and drastically cut operating expenses. 64. Institutional Implementation: NARP involved not only upgrading the facilities available to SAUs for conducting research (the objective of the sub-projects) but, more importantly, it also involved reorienting the focus of research to make it responsive to local farmers' needs by building a strong linkage between research and extension and undertaking zonal adaptive research. Past research planning had been largely 'top-down', partly due to the limited exposure of researchers to field conditions and the failure of the extension system to influence research. This resulted in much research which had limited relevance to farmers' production constraints in many agroecological zones. 65. The zonal research concept was the key element in the project. The initial background paper of each SAU identified the agro-ecological zones within the state. These zones became the starting point for the detailed state research review. By the time these reviews were completed, 127 agro- ecological zones had been identified in the country (of which 6 were in the North Eastern Hill Complex not covered by NARP). Sub-projects for each zone were then planned, of which 81 were started under NARP, with the remainder to be completed under the follow-on NARP 11.29 During project implementation the zonal idea was further developed by requiring each state to produce a zonal status report which identified sub-zones, the specific problems of sub-zones and research priorities for them. 66. The concepts of seasonal zonal workshops, monthly training sessions with SMS staff, and joint scientist/extension staff field visits and trials took about three years to be generally accepted. This acceptance was strongly influenced by a series of workshops at SAUs organized by ICAR to explain the new multi-disciplinary approach to adaptive research directed at solving farmers' problems. As the reorganized extension system began to put pressure on research to develop technologies suitable for local agro- climatic conditions, the need for this approach also became more apparent. 27/ SAR, NARP II, para. 2.15. 28/ Mid-Term Review, National Agricultural Research Project II, ICAR, April 1989. 29/ NARP II was approved by the World Bank's Board in October 1985. - 26 - IV. PROJECT OUTCOME Inpact 67. The agricultural impact of these projects is very difficult to assess for two reasons. First, the contribution of extension and adaptive research activities to increased production is hard to measure since extension and research services are only two of the many inputs used in the production process. Attributing yield and area increases solely to extension and research activities as such would be inaccurate. Second, even if it was possible to separate out the effects of extension and research, a "without" project scenario is now non-existent in the project areas given the near statewide adoption of the extension system in the five states and the national coverage by NARP. 68. In view of the measurement problems, it is not surprising that a quantitative assessment of the direct effects of extension and research service investments on increased airicultural production is absent in most evaluation studies on the subject. An empirical study of the impact of and returns on extension investment in India was carried out, however, in two adjacent districts of Haryana (T&V) and Uttar Pradesh (non-T&V).31 The study found no statistically significant yield differences in the kharif (wet season) crops, but attributed an increase of about 6-7% in yields for HYV wheat in the rabi (dry) season to T&V extension, and estimated an economic rate of return of 15%. However, this difference in yields for HYV wheat (rabi) may have been influenced by less reliable irrigation in the non-T&V district (Annex 2). Despite the difficulties involved, a number of attempts have also been made to quantify the return on successful implementation of research investments in various countries, including India (but not NARP); these have generally estimated economic returns to be high (between 40% and 70%).32 69. An economic rate of return was not estimated at appraisal, and a quantitative assessment of incremental benefits at completion was not possible for any of the projects under audit. The extension projects, however, did provide some indicators of positive impact and achievements 20/ What is surprising is the absence of any assessment of the rate and profitability of adoption of new technology from whatever cause. 2J/ G. Feder, L. J. Lau and R. H. Slade. "The Impact of Agricultural Extension - A Case Study of the T&V System in Haryana. India", World Bank Staff Working Paper No. 756, 1985. 1/ Evenson, R. E., "Technology Generation in Agriculture," Proceedings of a Conference on Agriculture in Development Theory, Bellagio, Italy, May 1973 (mimeo, Economic Growth Center, Yale University). Schuh, E. and Tolini, "Costs and Benefits of Agricultural Research", World Bank Staff Working Paper No. 0360, 1979. - 27 - through their monitoring facilities, although how much of this is a result attributable to project activities cannot be ascertained. Despite the inability to attribute causality, some reported yield changes over time for each state are presented in the following paragraphs. Changes in Productivity 70. West Bental. Five-year moving average of statewide yields for the period 1975 to 1989 indicate annual increases of three percent or more for potatoes and oilseeds only, but a decline for pulses. The oilseed increase (about four percent annual average) has occurred at the same time as an area increase, while the decline in yield of pulses has occurred with a decrease in area planted. The better performing crops are both cash crops, which suggests that improved market conditions such as higher relative prices and profitability were likely to have been major factors in the increased farmer attention required for higher yields. The extent to which adaptive research and extension contributed to this improved productivity cannot be ascertained. 71. Bihar. The PCR reported that there has been an apparent impact through the introduction of more profitable cropping patterns. A third crop (usually short-season pulses) has been introduced in the annual program in irrigated areas. An increase in the production of maize and pulses in rainfed agriculture were quoted. However, five-year moving averages of statewide yields for the main crops between 1978 and 1984 did not indicate a significant impact, suggesting that production constraints other than knowledge of potential yield-increasing technology were limiting factors. Drought or flood conditions during project implementation, and relatively poor input supply and access to credit (25% of farmers being ineligible) were reported as limiting factors to increased production. 72. Kerala. The PCR reported that there has been significant yield increases for rice, millet and pulses during the project period. Statewide moving averages from 1981 to 1986 indicated increased yields at the rate of at least two percent annually for 10 crops, a neutral response for five crops and declining yields for another five crops. Five of the crops in the first group were annual crops and included the three mentioned above. Coconut which is an important perennial crop was in the neutral category. 73. Maharashtra. In view of the inability to reliably attribute causality to any yield responses, the PCR did not attempt to measure the impact of the project on yields, nor did it present any statewide yield changes. 74. Tamil Nadu. Five-year moving averages of statewide yields from 1982 to 1987 show a substantial increase - annual increases of 6.2 percent for pulses and 9.9 percent for cotton and 3.2 percent for rice. There have been no significant changes in the area covered under most crops, although pulses showed an unusually sharp increase in 1986/87. - 28 - Impact on Adoption of Technology 75. Some important results of the impact of the extension projects on adoption of technology are presented in Annex 3. The main implications of these M&E survey results are discussed below. 76. The first observation is that CFs in general, and others who are visited regularly by extension staff, tend to adopt technology sooner and achieve higher yields. However, since the process of CF selection is biased in favor of relatively progressive individuals, it is not surprising that the M&E data indicate faster adoption rates and higher crop yield achievements for that group. In any farming community progressive farmers will seek out information in removing constraints and adopting appropriate new practices before less innovative farmers do. Due to its relatively intensive coverage of farmers, T&V extension could claim that a greater proportion of farmers are able to realize the opportunity to seek and obtain advice and, therefore, speed up the adoption process. However, the positive results are neither exclusively attributable to the projects nor to T&V extension embodied in them. 77. A second observation from the M&E results presented in Annex 3 is that, in the case of non-adoption of a specific technology, lack of knowledge of the innovation was rarely the reason for non-adoption. Important reasons cited for a non-adoption were economic considerations, climatic factors, unavailability of inputs and lack of irrigation. Most states had an acceptable level of adoption in those innovations or practices which are easily perceived as advantageous, such as new adapted varieties, nitrogen fertilization and specific plant populations under conditions of adequate moisture control. However, for those practices which had with lower adoption, such as seed treatment and plant protection, the specific technological innovation was either not relevant to the farmer's situation or was not appreciated as such by the farmer. In either case, a detailed analysis of economic considerations including associated production risk in the farming systems in micro-zones, and of access to resources and inputs by farmers, could have explained the underlying reasons for lower adoption rates. This sort of analysis was virtually non- existent in the projects and could have led to modifications in recommendations or in the system of their delivery so that they were more applicable to farmers' circumstances. Institutional Achievements 78. As a group, the projects succeeded in institutionalizing zonal adaptive research within the agricultural research system in the states covered and in developing an extension service based on T&V principles, albeit all with modifications to adjust to local circumstances. 79. National Agricultural Research Project (NARP). An effective system of allocation of central research funds to the SAUs as state agricultural research institutions was established, and the zonal research - 29 - concept was successfully introduced in the participating states under NARP. However, at project completion, it is apparent that a number of problems have remained. These are: (a) Staffing of zonal stations, especially those with scarce facilities in education and services, has been a persistent problem so that many sanctioned positions have not been filled. (b) There is a tendency to have most resources directed to traditional fields of research although a start has been made in directing research to be more responsive to the requirements of the major farming systems. An adequate understanding of socio- economics constraints in farming systems and attention to non- traditional research fields of water management, animal husbandry and agroforestry are lacking. kc) The recent zonal status reports, which are an improvement on earlier reports on technical aspects, still fail to address socio-economic constraints in farming systems. (d) Although there has been some attempt to develop a ranking system to prioritize research expenditure, this function is not well developed in most SAUs. ICAR and Bank supervision missions could have paid more attention to instituting prioritization systems which take into account not only national and state production goals but also include benefit and cost estimates in major farming systems. 80. Research-Extension Linkage NARP provided a good conceptual basis for this research-extension linkage. The effectiveness of this linkage has varied between states. During implementation the four instruments (para. 28) had varying degrees of problems. (i) The interaction between scientists and extension staff in the seasonal zonal workshops was a significant achievement. However, these sessions were usually not structured to ensure that the maximum feed-back was obtained from the extension staff. In some cases, because extension staff did not have the educational background or social status, they were unable to confidently interact with the researchers. In addition, M&E findings were not sufficiently taken into account, and the multidisciplinary approach needed to address problems in a farming system context was not fully appreciated and utilized by researchers. (ii) The result of the monthly training sessions between researchers and extension SMS provided a useful forum for interchange of knowledge, although they tended to become repetitive and monotonous at times. Effectiveness varied - 30 - according to how well these were structured and the quality of leadership in the zones. (iii) The joint diagnostic field visits were generally less well developed than the other aspects of the research-extension linkage program. This function may not be properly developed until the SAUs become more involved in the diagnostic phase of farming system research (FSR), through which the importance of the many factors affecting adoption of technology will be appreciated. (iv) The design of i-daptive and on-farm trials improved over the project period and should continue to improve as the farming system orientation to research is advanced. 81. The Bihar Extension and Research Proiect strengthened the Department of Agriculture's adaptive research program while NARP strengthened the SAU applied research system in the same state. Although this duality did not apparently cause serious problems in project implementation, it would have been more efficient to have one institution responsible for the adaptive research in the state or, at least, one institution per region. In the West Bengal Extension and Research project, the Department of Agriculture had primary responsibility for research at the time of project implementation and continues to do so, despite the inclusion of West Bengal in NARP II in which the SAU was to assume responsibility for research. Staffing problems reduced the impact of the research component in West Bengal, especially in the area of adaptive research in the agro-ecological zones which was supposed to have supported the extension program. 82. Extension Projects. Apart from Bihar and West Bengal, the coverage of the districts by the T&V services within the states was satisfactory. However, the Lcale of the programs were so extensive that they led to recruitment of many staff with relatively low education levels in all five states. There were frequent observations that the message transmission became routine, that the VEW was unable to relate the message to the circumst._ ces of an individual farmer, that he lacked credibility with educated farmers and the ability or confidence to communicate farmers' problems back to the SMS.33 There is obviously much variation amongst VEW staff, but a higher average education level should have been able to reduce this variability and improve overall performance. JI/ The Bank's Extension Sector Review, (Annex 6, para. 6), 1988 (footnote 6) reports that sub-divisional and district staff estimated that the performance of one-third of VEWs was poor. The report, however, points out that the best of the VEWs were capable of going beyond what they learnt at the fortnightly training sessions and had developed effective communication with farmers on their own initiative (para 2.17). - 31 - 83. The educational level of VEWs has attained more significance as the original T&V extension has evolved during and subsequent to the project (Annex 1). The responsibilities of the VEWs are generally being expanded to cover more aspects and more complex issues in the farming systems; horticulture, soil and water conservation and management, agroforestry, cropping pattern changes, livestock as a part of the farming system, and a greater emphasis on cost/benefit analysis of technology are assuming greater importance. The latest recommendation of the GOI is that the minimum entry qualifications for new VEWs should be a degree in agriculture, and that incumbent non-graduate be retrained.34 The retraining program is inhibited, however, by many incumbents (who are permanent employees of DOA) not having sufficient schooling to allow them to enter B.Sc. programs, or being too old to be considered for retraining. The trend to expand the role of the VEW to be more responsive to most constraints in the farming systems reflects a departure from the original T&V concept of delivering predetermined but timely messages concerning the most important crops. 84. Currently, there is a surplus of unemployed agricultural graduates in many states of India. However, it would probably not have been possible to meet the VEW recruitment targets using only graduates within the time frame envisaged at appraisal. There would also have been a further burden of incremental staff costs to the states because of the ability to transfer even less of the VLW and special crop program staff who were on permanent employment status. In retrospect, it may have been more appropriate to proceed more slowly in implementing the VEW system in accordance with availability of suitable manpower, while thoroughly developing the adaptive zonal research capability at the same time. 85. Despite the general adoption of the T&V system, the principle of a single responsibility for technical agricultural extension by VEWs was not universally applied. In states where the VEW falls under the administrative control of Chief Executive Officer responsible for all services (e.g., Maharashtra), there were instances where the VEW was diverted from regular duties to non-agricultural activities (paras. 46-50). This diversion is undesirable, and the problem is likely to be minimized only if an exception is made in the administrative structure and have the extension service solely under the responsibility of DOA, or if the executive branch of local government is convinced of the need to have the extension service work exclusively on agricultural extension. 86. Institutional programs and arrangements in the Departments of Agriculture did not fully adjust to the central role of the new extension system in technology transfer to farmers. The existence of numerous concurrent national and special programs, which are usually single crop and subsidy-based schemes operating on target implementation figures, was not W Screening Committee Report, Ministry oi Agriculture, GOI, 1988. Four states - Haryana, Punjab, Uttar Pradesh and Jammu/Kashmir - have recruited VEWs who are agricultural graduates. - 32 - reviewed to determine how these programs would fit into the new approach to extension. In many states these programs absorbed substantial staff time at the VEW and SMS/AEO levels. Such a review is still wanting, and it appears that the crop demonstration-subsidy programs under these schemes could be appropriately converted to support real demonstration needs for the extension program. 87. The T&V principle of concentrating on the most important messages for the major crops in a production zone was initially adopted. However, the principle of limiting messages to key crop practices was queried in many states. In the later stages of the projects there was a widespread appreciation of the need for the VEW to be able to respond to all significant problems and constraints in the farmers' production systems (Annex 1). This has led to broadening of the responsibilities of the VEW (Annex 1, paras. 17-20), and an understanding that there needs to be more flexibility at the district level in determining the topics for programmed visits. The audit believes that it is important for the extension service to be capable of responding to significant problems in the local farming system.35 However, this does not imply that "messages" should be programmed to cover all of these problems. The response -apability should be developed through a competent SMS network and progressive training of extension staff by these SMS in a farming system context. The audit's views on programming of topics for regular visits is elaborated in Section V, and focuses on delivery of technology to solve problems as perceived by farmers. 88. The message delivery system incurred other criticisms. Common complaints were: (a) too general, rather than specific to the production microzones; (b) inadequate for many rainfed and more risk-prone production areas; (c) lacking a partial budget analysis; and (d) often unaware of the constraints faced by farmers lacking access to cash or credit for recommended levels of input expenditure. The inadequate experience and training of many SMS has obviously contributed to these problems. It is the opinion of the audit that these criticisms will persist until there is a greater appreciation within the entire applied research/extension system of the importance of economics and a farming system approach in defining research requirements and adapting research results to extension messages. 89. A more contentious issue is the extent to which VEWs should be involved in the organization of inputs and/or short- and long-term credit for agriculture - an issue that is being debated in some states (Annex 1). On the one hand, involvement in these services can increase the prestige and importance of the VEW in the community since he would be performing a tangible service for farmers. On the other hand, such involvement is _1/ A furthet complication occurs in broadening of the responsibilities of the VEW in some states, in that animal production and agroforestry may pertain to separate ministries (e.g., in West Bengal); this problem is not insurmountable, but does involve an extra burden of coordination at senior management level and at district SMS/AEO level. - 33 - likely to lead to an inordinate amount of time being occupied in administrative work at the expense of technical advisory duties. The audit believes that VEWs could participate without negative consequences for extension in non-extension activities which define requirements for inputs or production credit, and which feed this information to the relevant body which coordinates the supply of these services, if such assistance is warranted. Such an intervention should be directed at developing a capacity in the farming community to organize itself for such purposes, and the VEWs should not actually become involved in the administration of input or credit delivery. This particip. 'ion would still be within the objectives of an extension service dedicated to technology transfer.36 90. One of the most important failings of the program was the relatively poor selection of CFs in many cases (paras. 52-55). The audit believes that the projects' guidelines could have been substantially improved by placing more emphasis on contacting representatives of existing farmer groups, rather than selecting farmers from various strata in the village. The result was that the potential influence of CFs on other farmers was much less than intended. 91. The M survey system has been institutionalized, albeit with some resource problems, but the interpretation of results and use of the findings needs substantial improvement in most States (paras. 56-58). The data produced on the implementation of the extension system, the adoption of technology and the reasons for non-adoption provide useful information for analysis and action by extension management at the district and state levels. The adoption data also has relevance to the research/extension linkage program. With few exceptions, these functions have not been well developed and a valuable resource is underutilized. Sustainability 92. NARE. India spent about 0.3% of its agricultural GDP on agricultural research during 1981-85.37 The importance of agriculture for generating growth and for social stability is so great that the level of support currently accorded to agricultural research should at the least be 36/ Section V of the report discusses prioritization of the type and extent of public sector production services, including extension, for specific regional conditions. It is likely that region-specific approaches to services would involve cases in which the DOA representatives should have dual extension and other agricultural service functions for a period of time in accordance with availability of total services. 17/ This percentage is less than the corresponding figure for developing countries as a whole (0.42%), although the Indian system does have the largest number of full time equivalent agricultural scientists in the world, and total research expenditure is exceeded only by Brazil amongst developing countries. ISNAR: Research Expenditures as a Percentage of Agriculture GDP. 1980-85. * 34 - maintained. However, the size of the system is such that there will be continued pressure on the research institutions to become more efficient and cost effective. The mid term review of WARP II by ICAR indicates a reluctance of came State Governments to provide adequate operational costs for subprojects which had been completed under NARP. 93. Sustainability will largely depend on the levels of funding for research in general, and on the attitudes of the national agricultural research establishment to zonal research. The professional support for zonal research appears to be slowly increasing, and it is the audit's opinion that the concept is well established and should continue to develop. The biggest variable is likely to be state budgetary support. The most recent analysis of agricultural research expenditures in India38 indicates that commitments by the State Governments now exceed GOI commitments (which are mostly through ICAR programs), whereas 10 years ago the state expenditures on research were only half that of GOI. NARP supported this transfer of greater responsibility for research and its funding to the States, with particular emphasis on zonal applied and adaptive research at SAU stations and substations. Continued funding for operational cost support will depend on the level of transfers from the centrally-funded ICAR programs (currently covering about 30% of SAU expenditures), on the State budget authorities, on earmarking funds for zonal programs within SAU budgets and on the SAU's allocation of its discretionary resources. It behooves the SAU network to (a) use the zonal research facilities to increase the relevance and accessibility of their research, (b) institute a transparent system for prioritizing research which takes into account the potential economic impact of the research findings, and (c) present the accomplishments of the state agricultural -research and the zonal stations in a convincing manner to the funding authorities. 94. Extension Projects. The five extension projects required a substantial allocation of professional staff and financial resources. The projects financed the capital costs and, to varying degrees, the recurrent cost of running these operations. The recurrent cost element was substantially higher than initially anticipated due to problems of transferring staff from other programs (para. 37). These costs are now the responsibility of the states. 95. The overall conclusion is that the states are continuing to operate the extension schemes with varying degrees of difficulties and 1/ R. Slade et al., "Agricultural Research in India: Prologue, Performance and Prospects." Draft, December 1989. Research by the private sector has also increased substantially in the past 10 years, but is only about 4 percent of the expenditure by GOI and the States, and is concentrated in areas where gain in knowledge can be legally protected: plant hybrids, pest control chemicals, poultry, fisheries, agricultural machinery and selected plantation crops. - 35 - efficiency. Although the states of West Bengal and Biha did not reach the targeted staffing levels during the project, neither state has reduced the staffing level since the credit was cl.sed. The level of non-salary support, on the one hand, appears to have been maintained at a bare minimum operating level. Tamil Nad has adopted a rational attitude of ensuring that its extension staff are adequately supported by having a non-salary budget at around 25% of total service costs. However, Maharashtra, despite its generally favorable attitude to extension, has reduced its non-salary budget to 7% - 10% of total extension costs; this reflects the large increase in the salary budget associated with the project and is creating operational difficulties. Attaining and sustaining the level of financial commitment needed for efficient operation of the services is likely to occur only if the state governments are convinced that: (a) extension of this scale and type is necessary; (b) funding of the extension service has priorit over some other aspects of the State's recurrent agricultural budget; and (c) the public sector has to continue to bear a '.irge part of the financial burden for extension. 96. Many States have already begun to adapt and modify the original T&V concept envisaged at appraisal (Annex 1). It is likely that the system will continue to evolve away frcm "pure T&V" in accordance with the constraints and requirements of the states. In most cases, the need to cover more than a few key items in the extension program has been accepted, and there is even a move in some states (e.g., Kerala) to work more with farmer groups rather than concentrate on contact farmers.41 The audit considers these developments to be generally positive. 97. The allocation of a greater amount of the States' recurrent budgets to operating costs of the extension services should not, in theory, present a major problem. Increasing this c-mponent from say, 10% to 20% of the extension budget would represent only about one to two percent of the total agricultural recurrent budget in a state. However, for that to happen extension would have to compete with input subsidy in the crop subsector, and irrigation services, soil conservation and animal health services which are major consumers of annual recurrent expenditures in other subsectors. All these competitors are conspicuous programs which usually demand strong political and official support. Consequently, as adequate recurrent cost funding is essential to maintain an effective extension program, the sustainability of the system will depend significantly on the ability of the extension sarvice to demonstrate its benefits to farmers, farmer organizations, political representatives and budgetary officials. The modifications to the original system reflect this need. A/ Extension costs in States which have adopted T&V are commonly between 10% and 16% of the total State recurrent expenditure for agriculture. In Tamil Nadu the figure was 12% in 1986/87. 6/ Annex 1 describes the recent changes and modifications to the T & V approach in the states. - 36 - V. ISSUES AND LESSONS 98. The audit provides evidence of the varying performance of agricul- tural extension projects and aspects of the T&V system of extension embodied in them in five states of India. It also evaluates the validity and feasibility of the zonal research concept as a complement to the extension services introduced through NARP in India. 99. The overall conclusion regarding the extension projects is that relatively intensive technical extension services have been provided to the farming communities in laharashtra, Kerala and Tamil Nadu, and partial coverage has been obtained in West BenIal and Bihar. While recognizing implementation problems, the audit generally supports the relatively posi- tive conclusions of the PCRs concerning the former three extension projects and the reservations concerning the latter two projects. Similar observations have been made and tssues raised in the two previous audits of T&V extension projects in India (para. 2). These include: (a) the need for greater attention to the cultural and socio-economic characteristics of the communities when selecting CFb and organizing extension schedules; (b) the necessity to have the full support of the bureaucracy if the extension and supporting research programs are to be sustained; (c) apparent differences in the adoption rate for various types of technology; (d) the inadequacy of the complementary services in some districts; (e) the problems in making the research-extension linkage more effective; and (f) greater emphasis being required on a farm system orientation to applied research. 100. Unlike the PCRs on the five extension projects and the two previous audits, this report has somewhat stronger reservations about some aspects of the T&V extension concept used in these projects which have limited the impact of the extension investments. First, it points to the varying acceptance of the T&V system in the different states. Second, it addresses the source of the implemeatation probl,ms in the states and in this it focuses on funding, staffing and staff quality problems that also have implications for sustainability. Third, it addresses the difficulty of having a single line of command and the diversion of extension staff at all levels to undertake non-extension duties. Fourth, the zonal research concept and the research-extension linkage are more fully evaluated. Fifth, this audit has stronger reservations about some of the conceptual issues such as the T&V system's reliance on information transfers to attain rapid adoption and the selection of a contact farmer as the main medium of technology dissemination (paras. 108-113). And finally, it examines the conditions of replicability much more closely. 101. The main conclusions with regard to NARP were that an effective method of rational allocation of research funds from the center to the SAUs was established, the zonal research concept was successfully introduced in the participating states, and a good conceptual basis for research- extension linkage was initiated. However, various problems surfaced during - 37 - implementation and the effectiveness of research-extension linkage was variable among the states. A number of improvements have to be made if the zonal research system and its linkage with extension is to realize the full potential of the concept in developing and transferring relevant technology. 102. The audit acknowledges the difficulty of measuring the agricul- tural impact of these projects. Since research and extension are only two of the many factors invo'ved in the production process it is virtually impossible to either separate out their effects on production and yields or attribute causality. Recognizing these measurement problems, the audit emphasizes the institutional aspects of performance. The main issues and recommendations addressed in this section relate to the implementation problems covered earlier and the conceptual issues that emerge from that experience. Extension Proiects 103. The original T&V extension approach in all states has been considerably modified and adapted subsequent to project completion (paras. 83, 87, 89 and 96 and Annex 1), the most extreme being in Kerala which has abandoned the CF principle altogether and adopted a group farming approach for rice as the major annual crop. All the other states have made adaptations to meet their own circumstances. The audit considers these adaptive responses by the states to economic signals, socio-political circumstances and administrative structures as positive developments in the evolution of the extension systems. Certainly, there is insufficient evidence to warrant complete adherence to the original T&V concepts introduced in these projects. However, care must be taken by the states to maintain the emphasis which technical agricultural extension has attained through the projects, albeit with much greater attention being given to farmer benefits and ccsts than has been practiced so far in introducing technology (para. 88). 104. There has been a general recognition by the states of the importance of developing technology which is relevant to the farmer's situatien, transferring this knowledge from researchers to farmers and adapting extension methodologies to the circumstances in the farming communities. These have been positive contributions of the projects. However, a number of concerns have remained. 105. The most serious concern for the extension projects as a group is the guarantee of continued adeauate funding, especially for operational cost support for extension staff (paras. 94-98). This concern is likely to persist unless the extension service is seen as (i) being efficient; (ii) producing results in keeping with the relatively heavy allocation of resources to the program; (iii) operating in cooperation and coordination with other government and private sector services which affect farmers; and (iv) meeting its extension function in the eyes of the farming community and local political bodies. - 38 - 106. Another major concern during implementation has been the education level of VEWs and SMS (paras. 82-84). During project design and appraisal, educational level and experience were not given much emphasis; capability was to be achieved through retraining and continuous technical updating programs. This became a problem during implementation and resulted in many deficiencies in the technical messages provided to farmers (para. 88). Most states have subsequently elected to improve the basic education level of staff recruited as VEWs. This tendency has been accentuated by a deviation from the oxiginal T&V concept in the transmission of messages to farmers. The principle of addressing a restricted number of important messages for the major crops was generally not well received, and most states are attempting to broaden the coverage to address all serious constraints in the production system (para. 86). This implies a greater analytical ability on the part of the VEW and SMS and favors a higher basic education standard combined with intensive practical in-service training with a farming system orientation.42 Although a relatively low basic education level can be satisfactory for an extension system geared solely to the delivery of technical "messages", this can become an impediment if it is intended to provide a more responsive service for farmers. The basic educational level for a VEW would be less of a constraint if the VEW came from a practical farming background and was backed up by a competent SMS network; these two conditions, however, were seldom satisfied in the evaluated projects. 107. Yet another area of concern has been the existence of dMal responsibility in some states of the VEW to separate technical and administrative superiors, and a dilution of his exclusive role in technical extension (para. 85). With the growing emphasis on decentralized government, it is likely that pressures will increase to broaden the duties of VEWs from extension and, occasionally, even from agricultural duties, to service the local government requirements. It is feasible to have the VEWs involved in some aspects of coordinating the supply of complementary services and inputs, without sacrificing his primary role as an extension officer. It is important, however, that the VEW operating in the field does not assume regulatory duties, which might impinge on his reputation as an agent of technical assistance, or undertake administration duties that limit the implementation of the extension program. Where functions other than technology transfer are included in a DOA extension service, this issue may be addressed by having separate personnel for the non-technology transfer functions in the operating units. In the event that neither public nor private sector services provide adequate input supply and marketing channels necessary for profitable use of the main technologies to be proposed, extension project investment would best be restricted to diagnostic surveys and analysis rather than programmed extension _/ In practice, the development of more responsive services will depend heavily on the capability of the SMS and AEOs in working with VEWs in the regular training sessions to establish an appreciation of the whole farm operation and where technological change can profitably be introduced. - 39 - activities; mechanisms to supply these complementary services should be in place before intensive extension programs are commenced. 108. The provision of public housing in villages in order to encourage extension staff to operate close to the target communities ,as one of the more troublesome project components (para. 45). Allocation of funds for public housing in future projects should be carefully reviewed on a case by case basis to ensure that there is a need for such housing and that it is going to be a critical component for project success. 109. The multiplicity of programs providing extension services to farmers should be analyzed and rationalized before deciding on a major extension investment. In many states, parallel special schemes were still in existence at project completion (paras. 31-34), and the impracticality of transferring staff from existing programs led to significant initial and recurrent cost increases (paras. 42, 94 and 95).43 110. The projects demonstrate that the relevance of recommendations is of paramount importance for a successful extension system (paras. 78, 82 and 88). Critical areas are regard for farmer benefits and ccsts in using a specific technology, production and marketing risks, resource situation and access to complementary inputs and services in the community. This requires a farming system orientation in both developing technology and interpreting it into practical and profitable recommendations. It also emphasizes the importance of quality in the regular training/discussion sessions between SMS, AEOs and VEWs, and raises the question of the extent to which VEWs should be trained to carry out priority messages on time or to be more responsive to farmer pioblems (para. 106); the latter aspect is discussed further in the following paragraphs. 111. Practical MEg systems were developed in all projects, but the collected data, with few exceptions, has not been adequately analyzed and used by program management (paras. 57-59). More emphasis should also be given to presentation of the seasonally-collected data into time series to indicate progressive performance and changes in factors affecting performance. Unfortunately, opportunities for objective measurement of the performance of the T&V extension methodology against other forms of extension in a very similar production environment are rare in India because of the widespread adoption of T&V in a relatively short period. However, the original T&V concept has and will continue to be modified, and opportunities for comparative analysis between different extension systems are likely to arise. AJ/ In projects designed under a more rigorous investment-return analysis, the extent of the incremental recurrent costs associated with incremental staff recruitment would have prompted a re-examination of the economics of the project when this became apparent. - 40 - 112. With respect to the T&V extension concept in the projects, the audit has concerns about two related issues: (i) the reliance on information transfers to attain rapid adoption in all types of communities; and (ii) the selection of a CF as the main medium to influence his neighbors. 113. Reliance on information transfer to attain rapid adoption. The T&V system is designed to provide relevant technology44 through messages and demonstration to convince the farmer that he or she should use the technology to increase crop production. This information transfer process is most effective when dealing with simple, highly profitable innovations in a low-risk environment in a relatively progressive community. In these circumstances, the technology could be easily appreciated as relevant and quickly diffused through the community. However, when dealing with more traditional farming communities, especially when recommended practices become more complex, information alone tends to lose its pre-eminent character. Traditional farmers may not immediately see the relevancy and benefits of a specific technology and other factors which influence the adoption process must be taken into account if the rapid uptake of the relevant technology is to be obtained. The very small percentage of farmers who indicated lack of knowledge as a reason for non-adoption in M&E surveys support the contention that information alone is insufficient to promote quick adoption of many of the recommended technologies. 114. In undertaking the technology analysis-adoption process, some individuals, especially the more educated in progressive communities, tend to decide and act without much attention to the opinions of other members in the community. A large proportion of smallholder farmers, however, are members of groups within a village community which influence their behavior as individuals. Group members have similar values, concerns and attitudes which encourage them to use each other's opinions, and especially of those respected as having made good use of similar agricultural resources, in arriving at a decision on whether or not to reject, adopt or partially adopt a recommended technology - a "reference group" for behavior. The attitudes and behavior of a reference group may range from progressive to very traditional and conservative. These sociological principles in village communications are well established and documented. The audit believes that the T&V system would benefit from modifications which make A/ As determined by the researchers and technicians managing the program, based on their assessment of required technology, through a proper feed-back system from farmers. - 41 - use of these principles in defining target groups, establishing contact point with these groups, and in communicating with groups.45 115. Selection of a Contact Farmer as the Main Medium of Technology Dissemination. The focus should shift from CFs and the farmers in the vicinity of that CF to existing reference groups of farmers with similar resources, concerns and problems. The contact point then becomes the group itself for some contact sessions with the VEW, while nominated representatives of the group become the contact point for other meetings. The approaches in dealing with the groups should accordingly be changed from "message-centered" to a more "farmer-centered" approach in which echnology is introduced as part of a structured "problem-solving" procedure. This approach may not have significant advantages over a properly applied T&V or most other forms of extension when dealing with the more readily-acceptable forms of innovations in commercial operations in progressive communities. However, the audit believes that the reference group approach would have substantial advantages when dealing with more traditional, resource-poor or risk-averse communities, and in most communities when dealing with the following farming aspects: i) more complex changes (e.g., cropping pattern or integration of crops and livestock); ii) less obvious technologies, (e.g. integrated pest control); iii) technologies which require adaptation to particular input resource circumstances (e.g., crop varieties and fertilizer regimes for potentially flooded areas); iv) technologies which would benefit from group action with respect to resource use or marketing arrangements to reduce risks (Chart I). 116. Important elements in the group rather than the CF focus are that interaction with a group allows the VEW and his supporting SMS and AEO to more easily become aware of (a) the farmers' perception of their main farming problems and constraints to higher productivity, and (b) of the farming practices and technology being used by the majority. A key factor in obtaining adoption of technology in the more traditional communities is to "start from what the farmer thinks", rather than assuming that a particular innovation should be adopted by farmers because it provides higher yields in field trials. Farming practices are normally inter- related and inter-dependent, and it is common for smallholders to adopt improved practices sequentially, with the more profound changes following the successful adnption of other lesser innovations. Appreciation of this sequential adoption phenomenon and a knowledge of where farmers are in this technological chain can, therefore, facilitate the organization of the technology discussed with groups and their representatives in visits by the A/ The exi3tence of very defined social and reference groups in some states in India makes it even more important to use sociological principles in designing and managing the extension program. In one case study in a village in the Thanjavur district in Tamil Nadu, 33% of farmers and 25% of the CFs were "Untouchables", but the VEW had not visited any of the "Untouchable" CFs and these CFs did not even know the name of the VEW. As expected, technology adoption was low in this group. - 42 - VEW. Technology should be adjusted to the current practices, attitudes and concerns of the reference group.46 117. The adoption of a reference group approach can have implications for the educational level and training of staff in courses and regular training sessions, staff-farmer ratios, the frequency of programmed visits by the extension staff, the focus of the extension visits and the structuring of the communication-demonstration process during visits. These and other aspects involved in the adoption of these principles are elaborated in Annex 4. Zonal Research Concept and Implications for Research-Extension Linkage 118. The zonal research concept, which has been firmly established, is already starting to yield results and the research-extension linkage has been formalized. Despite these, a number of improvements are required and some issues which concern the relationship of state and zonal research to other components in the total research system appear to need attention if maximum benefits are to be achieved. These are discussed below: (a) The zonal stations should be accepted as an integral and key part of the agricultural research system. Staffing of zonal stations, especially those with scarce facilities in education and services, has been a persistent problem so that many sanctioned positions have not been filled (para 80). This suggests that: (i) the sanctioned research positions on zonal stations and substations should be rationalized so that a minimum core of researchers is established in accordance with the known and anticipated priority zonal requirements (implying that some specialty functions will be combined in a single agronomic research post); (ii) a substantial hardship post allowance should be instituted for facility- deficient areas; and (iii) staff appointments to zonal stations and their performance should be recognized as an important component in the career path of researchers. (b) Improvements in directing research to be more responsive to the requirements of the major farming systems are required. Although a start has been made in this direction, there is still a tendency to have most resources directed to traditional fields of research (para. 80). Further improvements in the research-extension linkage should increase the pressure on the traditional research body to adjust their interests and priorities. Increased emphasis also needs to be given to a better understanding of the A_/ This approach, however, does not detract from the need to develop and present technology in accordance with its impact on farmer benefits and costs in typical farming situations. The reference group-problem census-problem solving approach permits the most effective dissemination of this knowledge by introducing it in a medium in which the majority of farmers will more quickly appreciate the relevance of an appropriate technology to their own situations. - 43 - management and socio-economic constraints in farming systems and to non- traditional research areas of water management, animal husbandry, agroforestry and farm management. Adopting a farming system orientation does not necessitate sophisticated (and expensive) FSR through the diagnostic, component and whole farm research phases, but implies that the applied and adaptive research must be responsive to the real techuical constraints in the important farming systems. The introduction of a transparent system for prioritizing research, which ranks programs and projects in accordance with government's production objectives, economic cost/benefit analysis, and contribution to meeting important constraints in major farming systems would do much towards maximizing the benefits from research. (c) The recent zonal status reports, which are an improvement on earlier reports, need to complement the technical description with an elaboration of the important socio-economic constraints in farming systems (paras. 66 and 80). The appointment of agricultural economists to interact with researchers in the programming and analysis of research and economists and social scientists in farm system analytical work are obvious areas of improvement. (d) With the development and increasing importance of the zonal research stations, ICAR should review the role of the 166 regional substations attached to the central and national institutes program (AICRP, which assumes about 25% of the ICAR budget), as it would appear logical to make more use of the SAUs and their zonal stations in this function. (e) Similarly, it would appear advantageous to transfer most of ICAR's extension programs to the SAU and State DOAs now that the research- extension linkage and their zonal research capacity are improving. The belief that maintaining these programs with ICAR can ensure contact with zonal farming problems is very difficult to sustain, as coverage of production environments is necessarily very limited, and there tends to be a cadre of specialized "extension" professionals in the ICAR extension programs limiting the potential impact on research professionals. A more appropriate system to help ensure the relevance of research by ICAR to farmers' problems could be to: (i) transfer the extension programs to the states; (ii) assist the SAUs in building up their FSR capacity, especially in the survey-diagnostic phase, and make use of these findings in the ICAR planning and review system; and (iii) develop a FSR capability out of selected national institutes. (f) ICAR could beneficially assume a greater role in developing formal systems to ensure improved scientific information flow to and between researchers. This is suggested because the existing system is characterized by a relatively weak flow of information between researchers in the different states and the central research bodies, a poor library and research journal service especially of internationally reputable publications at most research stations, and a tendency for SAUs to recruit researchers and academic staff from within their own state system. ICAR - 44 - could also promote overseas fellowships, especially in the basic research fields; although this would not have a direct impact on the adaptive research capacity in production zones, it would have a trickle-down effect through improved quality in the overall research system. (g) The effectiveness of research-extension linkage during implementation was variable among the states. Improvements can be made in the four project components directed at improving this function - seasonal zonal workshops, monthly training sessions, joint research-extension field visits and farm adaptive trials (para. 80). Adoption of the reference group approach (paras. 110-113 and Annex 4) would also ensure an enhanced feedback from the farmer situation to researchers. A Set of Principles for Extension Prolects 119. Apart from the direct operational implications discussed above, the experience of the projects in India support or complement a number of general principles which the audit considers could form a framework for the Bank in considering support for future agricultural extension projects. (a) Relevant technology must be available: Investments in the intensive promotion phase of extension services should be undertaken only in areas with farming systems for which there is adequate verified data available to improve productivity and profitability; elsewhere, emphasis should be on undertaking the diagnostic phase of research-extension programming and on developing applicable technology (paras. 80, 88 and 114). (b) Adontion of teghnology will depend not only on the farmer's knowledge of the technology, but on the most limiting of the complementary factors needed for adoption: Availability of other inputs required to adopt the technology and access to markets for surplus production must be ensured before a reasonable return on extension investment can be expected (paras. 78, 88 and 109). (c) Continued improvement in productivity and response to extension services will depend on the availability of new. relevant technology which will, in turn, depend on adaptive research for the agroecological conditions which occur in each extension area. An adaptive research program which is responsive to the problems farmers face in their production systems should be in place to support extension which is designed to provide the required feedback to research. (d) The rate of adoption of technology in response to an extension service will be affected by the production risks inherent in the agroecological zone: For areas with favorable climatic conditions or infrastructure which largely removes environmental risks (e.g., irrigation and/or drainage), it will be easier to generate applicable technology and the adoption rate of profitable technology will also be faster. More marginal production environments for crops or livestock entail strong consideration of risk factors in developing technology and in its adoption by farmers; a higher - 45 - expected (average) payoff is required for an innovation to be attractive to farmers in general, and for resource-poor subsistence farmers the risk factor may lead to legitimate rejection. This implies a need for a more analytical approach that takes into account socio-economic as well as technical factors in developing and promoting technology (para. 88). (e) Different adoption responses can be expected to extension methodologies in accordance with the characteristics of the farming community: At one end of the scale, a progressive community with a relatively high level of education in which individuals are continually seeking the latest technology will be responsive to knowledge of any relevant practices. Any face-to-face extension system which introduces applicable technology can be expected to generate quick initial adoption and diffusion within this community. Even mass media communication, especially if supported by some accessible demonstrations, can be expected to yield good results. At the other end of the scale, there is the very traditional community which has well-established farming practices accepted by the majority and a strong group identity and respect for the opinions of group members. At its extreme development, this group would distrust innovations introduced from outside. In practice, there will be a whole range of variations between the two extremes in any country. However, in situations where the tendency is towards the more traditional end of the spectrum, an extension methodology geared only to information transfer will be less successful. Here, there are considerable advantages in dealing with identified farmer reference groups with emphasis given to farmers' perception of problems in order to obtain acceptance of the technology by the group, rather that concentrating on progressive individuals (as CFs) who may not be considered a part of the reference group. When traditional attitudes are combined with limited production resources and high production risks, the advantages in working with farmer groups become even more pronounced, and the concept of only providing timely technical messages become less appropriate. Regardless of the extension organization to be used in a project, whether it is a newly introduced methodology or an attempt to improve an existing service, it is important to make use of experience in the behavioral sciences in developing techniques which are designed to change behavior in the form of adoption of improved technology. (f) Since regional variations exist in the availability of appropriate technology, in the provision of production services, and in the mix of common farming systems- investment in intensive extension service may not be warranted in all circumstances. Furthermore, different types of farming systems and production environments are likely to respond differently to particular extension methodology. In some cases, conditions could be such that an intensive extension service based on a regular program of key messages should be a high priority; in other cases, priority could be to improve complementary services of input supply and credit with parallel attention to working with village groups to discuss and address production constraints; in still other cases, emphasis could be on applied and - 46 - adaptive research and on the diagnos -ic phase of farming system analysis. These diverging requirements imply a different mix and intensity of extension staff with varying budgetary consequences for particular situations. Apart from initial capital costs, intensive extension services invariably represent an increase in the recurrent budget, at least in operating cost and permanent salary budget. Thus incremental investment in extension should logically be based on an analysis of the circumstances of agricultural regions and in accordance with an immediate and longer term plan to provide for these circumstances. (g) I ay extension system the quality of the service should receive as much attention as the methodologa: Even if there exists a well-developed adaptive research program, there will always be a need to further adapt many technologies to the situation of individual farmers. Although in the end the degree of adoption will obviously depend on the indiv dual farmer's decision, the extension officer can facilitate the process by being very knowledgeable about all factors pertaining to the technology and its implications in the farming systems in the community. This requires training not only in technical theory, but also in practical aspects of its implementation, in farm economics and in the social sciences as they affect adoption. This applies to both the extension supervisory and training staff as well as the VEWs. Training in these aspects also has relevance for an effective and informative extension-research linkage systems. The ability to absorb the required training will depend to some degree on the basic education level of staff. (h) A substantial investment in extension will only result in a sustainable system if it has the suport of the administrative and funding authorities: Annual recurrent expenditures for "non-essential" services tend to be sacrificed in times of fiscal restraint and the working hours of government servants diverted from duties such as extension to other priority functions unless the extension service is seen as "essential". This implies that not only should the service be efficient and provide good returns on government investment, but that it must be perceived as such and meet the needs of the farming community. A major modification to an existing extension service should normally be introduced in a replicable model on a substantial pilot basis for a number of years before extending it statewide or nationally. In selecting replicable pilot areas and in interpreting the results of the pilot schemes, the principles noted in subparagraphs (a) to (e) should assume high importance. A slower pace would enable verification of the system or permit adaptation or rejection in accordance with performance, and build the regional and/or national consensus required for funding and administrative commitment to an agreed system. This would also permit a realistic recruitment and training program in accordance with the availability of suitable manpower thus allowing the true benefits of the new system to be demonstrated. Development of a practical and cost effective M&E system and appropriate use of the results for management and promotion in the bureaucracy, as well as eliciting the involvement and support of formal farmer organizations and village and district level politicians, should be considered as an integral part of extension management. - 47 - (i) A public sector extension service is not the only source of technical information, and advantae should be taken of comolementary private sector technoloZy and information resources: Fertilizer and pesticide companies, agroprocessing firms, marketing cooperatives, and major traders are examples of participants in the technology transfer process which occurs with or without a public sector servic%. The Government extension services should have a program to maximize the use of these resources in technology generation and as informed technical resources without jeopardizing the integrity of the service whose only interest must be the farmers' welfare. - 49 - Annex 1 Page 1 RECENT DEVELOPMENTS IN EXTENSION SYSTEM ORGANIZATION AND MANAGEMENT Introduction 1. Since the completion of the projects in Maharashtra, Kerala and Tamil Nadu, the three states have made, or are in the process of making, changes in the organizational and management oZ their extension systems to make them better suited to conditions within the state.1/ The three states are attempting to take into account the lessons and experience of project implementation and the pressures of meeting the budgetary cost of extension without external support. The Single Line of Command 2. The command structure of the extension system that was set up has been essentially similar in all the states that have participated, although there are minor variations between states. At the village level is the VEW.2/ Ideally, there is one VEW to 800 farmers (who may be located in a number of different villages), divided into eight groups with one (sometimes more) CF per group. Each group is visited on a fixed day of the week in a two-weekly cycle. During each visit, the VEW goes to the contact farmer selected from each group, passing on the messages that have been taught at the latest training session relevant at the particular time of the cropping season and involving as many other farmers as possible in the discussions and demonstrations. A group of eight VEWs are under the guidance and supervision of an Agricultural Extension Officer (AEO).3/ The area under charge of an AEO is variously known as a block or circle. 3. A number of AEO blocks, varying from 6-12, are grouped together into a sub-division, usually with three or four sub-divisions making up a district. Each sub-division is in charge of a Sub-divisional Agricultural 1/ Significant developments in Bihar and West Bengal have not occurred recently and these two states are not included in the discussion in this annex. ./ In Maharashtra the VEW is referred to as Agricultural Assistant (non- graduate) or Agricultural Supervisor (graduate); and in Tamil Nadu as Assistant Agricultural Officer (non-graduate). The name VEW is used here to cover all those serving the VEW function. 3/ In Kei the AEO is referred to as Junior Agricultural Officer (JAO). In Maharashtra and Tamil Nadu the AEO is referred to as Agricultural Officer; AEO/AOs are graduates. In Tamil Nadu a non-graduate performing AO duties is designated Deputy Agricultural Officer (DYAO). The name AEO is used to c-,ver all these. -50- Annex 1 Page 2 Officer (SDAO),4/ with the rank of Deputy Director of Agriculture (Deputy DA). Each district is in charge of a Principal Agricultural Officer (PAO) with district responsibility f.r exterzion operations throughout the district. Above him is the Joint Director of Agriculture (Joint DA). In Maharashtra three to four districts form a Division under the charge of a Divisional Joint DA, while in Kerala and Tamil Nadu there is no equivalent layer between the district and the headquarters Department of Agriculture (DOA). Within DOA the management, training and technical support of the extension service is the responsibility of the Additional DA (Extension) who himself -ts immediately responsible to the state DA. In Xerala, the Additional DA (Extension) was supported by two Joint DAs (Operation Control) with specific regional responsibilities. 4. Paralleling this administrative structure is another hierarchy of Subject Matter Specialists (SMS) to provide technical support to the VEW and to interact with the research system. These SMS are posted at sub- division, the district and the headquarters levels. At the sub-division level there are normally three SMS of Assistant DA grade, one each for agronomy of field crops, plant protection and training (in Kerala there is another SMS for tree crops). Where sub-divisions are small, two sub- divisions may share an SMS team. At the district level there is again typically a team of three SMS, this time of Deputy DA grade, usually of the same disciplines as at the sub-divisional level, although in places specializations not covered at the sub-divisional level may be represented if they are locally important (e.g. water management, horticulture). At the headquarters level the SMS team are of Joint DA grade. There may be four to eight in a team charged with strengthening the technical advice available in disciplines represented at the district and sub-division level or providing advice on subjects not represented, improving the technical quality of the extension service, developing and interpreting farm trials, promoting linkages with research and organizing training. 5. Existing alongside this extension structure there is also within each DOA an administrative structure concerned with special scheme implementation, subsidy administration, regulatory functions and input supply arrangements. Before the new extension structure was created the two functions were usually carried out by the same staff, to the detriment of extension. Hiving off extension into a separate structure did not lessen the burden of these other activities, which had negative implications for the planned transfer of staff from these programs into the extension organization, and for the work load of the AEOs at the subdivisional level who were supposed to concentrate on extension. Typically the structure consisted of one or two Additional DA at headquarters (assisted by up to ten Assistant DA), a Joint DA or Assistant Deputy DA at district level, with separate staff again at the sub-division g/ Where there are more than 10 AEOs in a sub-division the SDAO may be assisted by an Additional SDAO (ASDAO). In Tamil Nadu the SDAO is designated Divisional Agricultural Officer (DVAO), usually with the rank of Assistant Director of Agriculture. The name SDAO is used here to cover these. - 51 - Annex 1 Page 3 and block level, a Divisional Agricultural Officer (Inputs) or equivalent (e.g. in Tamil Nadu there is an Assistant DA (Seed Center) and an Agricultural Officer (Quality Control)) at the sub-divisional level and Deputy or Junior Agricultural Officer at the block level assisted by Agricultural Demonstrators. In Maharashtra there is an additional layer at the Divisional level where a Superintending Agricultural Officer is responsible for the technical guidance of the agricultural activities in the districts of the division. 6. In Maharashtra the administrative structure is farther complicated by the existence of a system of elected local government bodies. At each village is the village 'Panchayat'. Villages are grouped into blocks (usually 2 to 5 villages per block) with a "lock Development Officer (BDO) responsible for all development activities (including input supply) in the block to the Panchayat Samiti (Block Couticil). Administratively he is responsible to the Chief Executive Officer (CEO) of the Zilla Parishad at the district level, who is in charge of all development activities entrusted to the Zilla Parishad and who is the liaison officer with the State Government. The CEO is administratively responsible to the Rural Development Department. The CEO heads a number of departments at the district level. Agricultural activities are managed by an Agricultural Development Officer (ADO) who mainly deals with input supply questions and who, like other district departmental managers, operates through the BDOs. Before the introduction of a separate extension service the BDO had been responsible for extension operating through Village Level Workers (VLW). Under the project a number of VLW posts were transferred for service with the extension organization although they remained administratively within the local government sector (as were new VEW recruits). The number transferred was determined by the need to leave enough VLWs for the ADO/BDO to operate the input supply arrangements satisfactorily. In fact this dual line of command continued and still continues till the present right up to the district level with all staff, from the PAO downwards, being administratively responsible to the CEO but technically responsible to the DOA; this arrangement was in force at the time of the audit. As a result the CEO can use extension staff for non-extension work if he regards this as of higher priority. Organizational Changes in the Extension System 7. The modifications which have occurred in the three states conform to a pattern and are thus better described in terms of their intent rather than separately for each state. Th... types of modifications are related to: (i) the integration of input supply and other agricultural development activities with extension; (ii) broadening the range of advice being given by the extension system; (ii) the contact farmer as the conduit of extension advice; - 52 - Annex 1 Page 4 (iv) the role of the SMS; and (v) training. 8. Integration with Input Supply: the relationship between the input supply organization of DOA and the extension service was a frequent source of friction during project implementation. In addition, the existence of two parallel systems led to at least the appearance of a waste of resources and of unnecessarily high operational costs. In Kerala and Tamil Nadu the two systems have undergone or are undergoing some degree of integration. 9. The change began in Kerala in 1987 when it was decided to make the 'Panchayat' the base level unit for agricultural development, as it already was for other development activities. There are 1001 'Panchayats' (and 47 municipalities/corporations) in Kerala, compared with the 378 extension blocks created under the project. By amalgamating the extension blocks and the agricultural development units (i.e. the input supply structure) and abolishing some special schemes, the State Government was able to staff each 'Panchayat' with a graduate AEO equivalent and three VEWs (two in smaller Panchayats). Thus the total number of positions at the block/village level went up from 2265 at project completion to 3783, an increase of 67%. The average number of farm families per VEW thus fell from 1200-1300 to 800-900 under the project. These new units are called 'Krishi Bhavan'. 10. In Kerala, the input supply functions have been amalgamated with the extension system as have most of the other agricultural development activities of DOA. While maintaining the basic organizational structure of the extension system introduced under the project, including tb-3 single line of command, fortnightly visits and regular training, the VEW is now responsible for developing, implementing and monitoring location specific schemes such as minor irrigation projects, bio-gas plants. Furthermore, the VEW is expected to support the credit requirements of these schemes by advising on sources and utilization of credit, monitoring utilization of and assisting applications for credit; and ensuring the timely availability of input supply. 11. Each 'Krishi Bhavan' has an office in the panchayat which has become an extension center. It is permanently staffed by one of the three VEWs and frequently by the AEO. Farmers who need advice between the visits of the VEW, or who might otherwise never see the VEW, can come and find it in the extension center. One of the VEWs is responsible solely for office work (input management, development schemes, stock taking, scrutiny of credit applications, etc.) while the other two are engaged exclusively in field work. Thus, in practice, the number of farm families per active VEW has not changed much from the previous system. 12. In addition, each 'Krishi Bhavan' has an advisory committee which meets four times a year to approve location-specific schemes, follow-up on disbursement and repayment of loans, monitor the provision of inputs and review the progress of agricultural programs. The President of the local - 53 - Annex 1 Page 5 'Panchayat' is the chairman of the committee. Its members include representatives from cooperative societies and banks, farmers, agricultural laborers, field-level extension staff, the AEO and the political representative of the area. There are similar committees at district and state level. 13. This virtual elimination of the organization at the block level has been accompanied by changes at the sub-divisional level. The position of subdivisional SMS has been withdrawn (on the grounds it was duplicating the work of the district level SMS) and the staff have been re-assigned on an area basis to become the new block organization. Retaining their Assistant DA rank, the ex-SMS staff are now responsible both for their old extension-related activities and the review and implementation of location- specific schemes, financial sanctions, and a number of administrative tasks. These latter duties take up the greater proportion of the working day to the detriment of extension. 14. In Tamil Nadu, the change began in early 1989 when a new state government came into power and reintroduced the Panchayat system (see para. 7 of this annex). The DOA responded to the new Government's policy of decentralization by appointing 378 Agricultural Development Officers (ADO) at the Block level. The ADO has administrative and technical control over all the agricultural staff within the Block, extension staff as well as the staff running the input supply depots. While technically responsible to DOA, he is, however, under the admir istrative control of the Panchayat Union Council (as in Maharashtra) with some powers accorded to the Union Council Chairman and some to Panchayat Union Commissioner (formerly the BDO). The block level offices and stores have been designated as Agricultural Extension Centers. They will be developed also as places where farmers can come for advice and information as well as those inputs supplied by DOA. 15. The next stage in this reorganization is to be the amalgamation of the extension and input supply staff at the sub-division level, at the same time increasing the number of sub-divisions to coincide better with administrative boundaries, thus reducing the size of the sub-division. An additional SMS post for Information and Inputs will be created to replace the previous ADA (Seed Center). The reorganization of work at the block level will also be taken as an opportunity to redraw the boundaries of the VEWs' jurisdiction to give a more balanced work load and avoid universal application of fixed ratios. On average the number of farm families per VEW is to be reduced from 1,050 to abouL 930. 16. Although the number of farm families per VEW has gone down, the workload has increased because the VEW is now responsible for input assessment and for implementing the special state and center crop schemes operated by DOA. The VEV is required to assess the input requirements of the farm families in his jurisdiction. This is consolidated at block, sub- division and district level, up to the ADO, who then prepares a final assessment of the requirements for his agricultural extension center from cooperative and private sources. This is further consolidated up to - 54 - Annex 1 Page 6 headquarters level, indents are placed and relevant banks, cooperative societies and companies informed about expected levels of future demand. 17. Broadening the Range of Extension Advice: In the initial stages of the introduction of the new extension system the advice being offered to farmers by the VEW was almost exclusively concerned with field crops, especially major cereals and pulses (except in Kerala where tree crops predominate in some districts). This reinforced the existing parallel extension services for specific crops or activities, such as fodder production, livestock husbandry, horticulture, farm-forestry, soil conservation, etc. This process has been furthered by the establishment of separate departments for some of these subjects under the Agricultural Production Commissioner. 18. In Tamil Nadu there are six separate departments5/ and two specialized programs6/ all providing advice and/or training to farmers. Most of these, except Agricultural Engineering and Animal Husbandry, do not have enough trained staff to do more than provide a minimal service. However, the Agricultural Engineering Department is engineering-oriented and not geared to providing advice on agronomic or biological solutions to soil and moisture conservation; and the Animal Husbandry Department is mainly concerned with veterinary and breed improvement activities, not with animal husbandry, fodder production or feed management. 19. Under a proposed reorganization of the Tamil Nadu extension system it is planned to cover areas which are currently the responsibility of other departments. This would be achieved through formal notification of changes in responsibilities, in-service training of DOA's SMS on new subjects, identification of specific messages in the new areas and provision for routine training through monthly zonal workshops. The disciplines to be transferred include soil and water conservation, horticulture, oilseeds, animal husbandry (fodder crops only) and forestry. 20. In Maharashtra a similar change is being introduced for soil conservation; there will no longer be a separate soil conservation department. In addition, the monthly training sessions since 1988 have included training at the sub-divisional level on dairying, poultry, fisheries, food preservation, post-harvest technology, sericulture and farm forestry. Horticulture has long been an important topic where there is irrigation. 21. Contact Farmers: In Kerala it became evident during project implementation that under local conditions the usual contact farmer method of channelling advice to farmer groups was not working as expected. With only about 30% of farmers working full-time on their holdings, farmers were difficult to meet at a specific time and place. This was further / Agricultural Engineering, Animal Husbandry, Horticulture and Plantation Crops, Oilseeds, Forestry, Rural Development. W Farmer Service Centers, Lab to Land Program. - 55 - Annex 1 Page 7 complicated by the absence of a traditional village layout which appeared to lessen communication between individuals. In addition, the original choice of contact farmer was poor since they were selected by untrained VEWs without local knowledge. This experience led the state government to drop the CF approach with a new group approach once the project was completed. After a number of trials in 1987 a massive campaign to promote group farming in paddy was launched for the 1989 kharif crop. 22. Paddy cultivation in Kerala occurs in small concentrated patches ('padasekharams') normally lying along valley bottoms. This lends itself naturally to group farming with the groups consisting of all those who have land in the particular padasekharam. In the Kerala group farming system the individual farmer retains the ownership of his land and his produce and makes his own decisions about management practices; but farming operations such as land preparation and ploughing, communal nursery development, sowing, the purchase of inputs, water management, weed control and, above all, plant protection, are undertaken on a group basis. 23. Responsibility for the formation of a group lies with the 'Krishi Bhavan' advisory committee, with the AEO and VEWs taking the lead. Above this the AEO is responsible for organizing input supply and the SDAO for providing training and extension support. Each group elects a committee of cultivators' representatives who in turn elect a "convenor" who looks after the proper functioning of the group. In addition, the regular system of fixed visits has been modified. The VEW now holds an agroclinic once a week from 9 am till noon at each padasekharam. He is accompanied by agricultural assistants and departmental plant protection mechanics to repair equipment. In addition the district SMS, or representatives of input supply agencies and cooperatives may be present. The clinics are used for distribution of seeds and seedlings but their main use is during the growing season when the whole crop area can be closely inspectcl and pest and disease attacks dealt with promptly. The clinic is also a place where non-group farmers can bring other problems to be discussed, diseases to be identified, etc. 24. About 20% of the kharif (autumn) paddy area was covered by group farming in 1989, with a similar proportion expected for the rabi (winter) crop. The results were closely monitored and it was found that crop yields increased substantially and the total gains from reduced cost of production and increased benefits, was about 15%. But the greater a-ailability of credit, the other benefit expected to emerge from group activities is yet to be seen. 25. The Kerala Government plans to spread this group farming system to coconuts during 1990 and to other crops, particu!arly vegetables, thereafter. 26. In Tamil Nadu and Maharashtra, the experience with the contact farmer was less negative, but nevertheless experience with project implementation showed that the CF is the weak link in the transfer of technology, especially where he is a non-adopter, a non-communicator - 56 - Page 8 (except for the demonstration effect of his own farming practices), a poor communicator, or where social barriers prevent a group of farmers from receiving messages. The DOA in both states is, therefore, planning to experiment with the use of groups as a means of eventually replacing the contact farmer. Group meetings to which the whole village is invited are already being held in the evening when farmers have finished work. It is proposed to explore whether other existing groups (e.g. cooperatives, social clubs) could be suitable channels of communication, and to try to form special interest groups which the VEW could meet; examples include farmers with a joint interest on soil conservation, integrated pest management, social forestry, poultry production, or all farmers within the command area of a tank. This would also be a way to reach women farmers and agricultural laborers. 27. The Role of the Subject Matter Specialist: The abolition of the SMS at the sub-division level in Kerala has already been mentioned (para. 14). The broadening of range of subjects to be covered by the extension service in Tamil Nadu (i.e. to cover water management, soil conservation, animal husbandry, horticulture, forestry, etc.) is giving rise to the question whether there is a need to increase the number of SMS at headquarters or district level and how far existing SMS could be trained to cover these new areas. 28. Training: the necessity for fortnightly training of VEWs has been questioned in both Kerala and in Tamil Nadu. In Tamil Nadu, in fact, formal VEW training now only occurs monthly. In between these monthly training sessions a new monthly meeting of VEWs has been inserted at the ADO level in order to coordinate input supplies with extension. The meeting takes place in the Agricultural Extension Center and is chaired by the ADO. - 57 - Annex 2 Page 1 Comments on a Case Study of T & V Extension Impact 1. The audit refers to the study1/ in two contiguous agricultural districts in northern India in which only one was subject to T & V extension. The summary of the report on the study follows. Som The paper Presents results from a study designed to estimate the effect on agricultural productivity of Training and Visit (T&V) Extension. As part of the field work for the study sample surveys of farmers were conducted to 1982 and 1983 in two contiguous areas in India. One of the areas to Karnal district in the State of iaryana where the T&V system was introduced in late 1979 and the other is Zairana tehsil In the State of Utter Pradesh where the older community development system of extension still operates. The two areas are environmentally and culturally very similer except for the presence in Karnal of the more Intensive T&V extension system snd some differences In the Incidence of production problems. The farm level data show that in Karsl under the T&V system, and despite some organizational problems, the extent of farer Interaction with extension agents was greater than in Karana and that farmers In F-teal viewed the change in the extension system favorably. Moreover, an examination of the rates of knowledge diffusion for a select set of practices for Righ Yielding Varieties (N1V) of rice and wheat suggests that TV extension led, Ia the study area, to a noticeable Increase in the race of knowledge diffusion of several BRV wheat practices. The main thrust of the paper, however, is an analysis of productivity differentials between the two areas for two crops, MT wheat and rice and, the extent to which any estimated differences can be attributed to the Introduction of the T&V extension system. Using the farm level survey data from the post-project period (1982/83) and econometric estimation procedures production and supply functions we fitted which explicitly Incorporate, and hence control for, a number of variables which may cause Productivity differences. The resulting estimates suggest that in 1982/83, after three yeas of T&V extension, OV wheat yields In Carsal were about 9 percent higher than in Ratrana. This estimate however is gross of any differential that existt before the more intensive T&T extension system was introduced. The results for BTV vice were not statistically significant. Accordingly, the remainder of the analysis focusses only on SW wheat, for which the baseline differential is nest estimated. In the absence of comparable farm level survey data for the imdiate pre-project period secondary data are atiltsed to estimate the productivity differential In EV Swheat yields between the two areas in the baseline year (1979/80). These secondary data, ainly mean yield levels, while available for a number of yees are not compatible with the 1982/83 sample survey data in a number of Important respects. In particula they do not distinguish between 8W and traditional varieties or between irrigated and anirtigated conditions. Iance a number of detailed 1/ G. Feder, L.J. Lau and R.I. Slade. The Impact of Agricultural Extension. A Case Study of the T & V System in Haryana, India. World Bank Staff Working Paper No. 756, 1985. -58 - Annex 2 Page 2 adjusteents we undertaken to derive a messe of the baselie prodsctivity differential that to csistent with the evidence derived from the detailed tar @asple sarvey undertaken to 1982/83. I the order In which they we undertaken these adjustments Include, the estimation of sean yields of NV wheat from the goss yield data, the esteation of a smoth time trend (to eliaiate the effects of random distrbances) of the pre-project growth to eans yields, the removal of the effect on productivity resulting from trigation problems (which deriva from differences to the canal irrigation systes and the Incidence of tubeells In the two areas) and finally adjustments which account for systematic differences in the use of farm Iputs and other variables. This last step Is achieved with the aid of the results from the econometrically estizated production functions. These procedures establish that there was a baselie produetivity differential of between 2 and 3 percent In favor of Kernel which must be subtracted from the post-project productivity 4fferential. The final results show that, after three years of T&V extension, there was a gat In productivity for KV wheat - about 6-7 percent which can be attributed to T&V extension. Subsequently a cost-benefit analysis of the incresatal investment In T&V extension to Kernal Is undertaken taking Into aecount only the estimate of benefits for RV wheat. Under several different assemptions about the profile of productivity chages over ties, the results show, with a high degree of statistical confidence, that the Internal rate of retwo to at least 15 percent. The analysis estimated a rate of return of at least 15% for scenarios in which the gross benefit of yield differences would disappear for wheat within 6, 9 and 16 years. However, the sample in the non-T & V district (Muzaffarnagar) had less reliable irrigation with 16.45% of farmers reporting Irrigation problems compared to only 3.22 in the T & V district (Karnal). Any impact that this lower reliability (greater risk) might have on a farmer's attitudes to expenditure on production inputs was not considered, and may or may not have had a significant influence Oti tVfe result. 3. The difficulty in obtaining empirical cause and effect data on the impact of T & V is further illustrated by an examination of a longer time series (1972-1987) of yield statistics in the whole of the districts containing the above samples (graphs attached). This shows a constantly higher yield of wheat in the T & V district than in the non-T & V district, and an increasing and relatively parallel trend in both districts with no significant difference between the trends. For rice, there have been substantial yield gains after 1982/83 in the non-T & V istrict (over 40%) while the T & V district yields have not significe ly increased; the difference in trend is highly significant at the 52 level. The audit was unable to ascertain the reasons for this relative increase, but it would certainly not be safe to assume that this was due to the non-T & V extension service; factors exogenous to the extension system are almost certain to have been heavily involved.2/ 21 In a paper entitled "Strategic Issues for Pakistan's Agricultural Extension: Looking Back to Look Ahead" which was presented in an extension seminar in Islamabad on March 6, 1990, Charles H. Anholt uses 1974-1987 data from Haryana and Muzaffanagar (and other states) to query the need for and impact of intensive extension systems. He suggests that the expected steeper slopes for indexed yields in areas receiving the investment in intensive extension are not apparent. WHEAT Yild (Tns)/hc Start of 2.6 2.4 2.2 2 1.8 1.6 - 1.4- 1.2- 72/3 73/4 74/8 758/ 76/7 77/8 78/9 79/80 80/1 81/2 82/3 83/4 84/ 58/5 86/7 Years o Krnal 4 Muzefremo gr (with TV) RICE Yid (Tons)/hi Start of 2.4- 2.2- 2- 1.4- 1.2 - 0.4 -T 72/3 7f/4 74/ 75/s 75/7 77/0 75/q7q/Ms<>/1 1/2 82/3 8./4 84/ e5/S 85/7 87/» YearM 0 Kernal 4 Muzaffomo gar WHEAT Index of 'hIed3 1 40 too- statt of T&V 110 9T-- 80 5- 72/-3 7-3/4 74/5 75/6 76/7 77/8 78/9 79/800 0/1 81/2 82/3 83/4 84/5 C5/r 8/7 57/8 Yeqara 0 Karnel + Muzaffiimragr RICE Index uf eide 170- 160 - IwQ 14Q- 120- 100- 9go- 80 - Start of 7- T&V 72/a 73/4 74/3 73/ 75/7 77/0 78/979/os/1 81/2 82/ 8/4 84/5 85/5 86/7 57/8 0 Krn l Muzeffamegar - 63 - Annex 3 Page 1 MONITORING AND EVALUATION (M&E) DATA ON PROJECT IMPACT Some results from available N & E data from the extension projects in the five States are presented in this Annex. The implications are discussed in the text (paras 76 and 77). West Bensal (a) CF9 did not adopt significantly more recommended practices than other farmers - a result that is at variance with the other projects under audit; (b) C?a tended to have slightly higher yields in the seasonal surveys than other farmers; (c) extension staff vere the most important source of technical information for aman rice and irrigated aus rice, but not for jute, maize and rainfed asus rice; (d) greater adherence to fertilizer and plant protection recommendations was correlated with higher contact with extension personnel; (e) full and partial adoption was generally very high for seeding rate recommendations, moderate for fertilizer and low for plant protection and seed treatmentt (f) lack of knowledge of a practice was not a significant factor in stated reasons for non-adoption. Stated reasons for non-adoption were high cost (fertilizer, plant protection), unavailability of inputs (fertilizers) or unnecessary inputs (plant protection, seed treatment). Rihar (a) In general CFs adopted more recommended practices than other farmers; (b) CFs and others who had significant contact with extension staff tended to have higher than average yields; (c) full and partial adoption was relatively high for recommended wheat and rice varieties, nitrogen fertilizer application and sowing dates, but was low for use of pesticides, phosphate application and cultural techniques such as plant spacing and line sowing; and - 64 - Page 2 (d) lack of knowledge of a practice was not a significant factor in the stated reasons for non-adoption. Stated important reasons were Iigh cost, lack of irrigation and unavailability of inputs. Kerala (a) CFs tended to adopt more recommended practices than other farmers; (b) those farmers (includix.g contact farmers) seeking advice and receiving more visits from the extension staff tended to have higher yields for the major crops; (c) fertilizer recommendations were adopted mcderately well on rice crops, and more extensively on coconuts which had not traditionally been subject to significant chemical fertilizer applications; consumption of fertilizer per ha of cultivated land increased througbut the project period (although this may not have been wholly due to the extension service); plant protection measures on the major rice crops were moderately adopted - the level of adoption exceeded that in the other states covered by this audit; and (d) stated reasons for non-adoption generally implied that farmers had considerable faith in their traditional practice; in the case of non-adoption of fertilizer application, climatic factors and farmers' assessment of associated risks were very important factors; lack of knowledge was not a significant factor for non- adoption. Maharashtra (a) Cs adopted more recommended practices than other farmers; (b) CFs and those farmers who had frequent contact with extension staff tended to have higher yields than those who had less frequent contact (for most crops, except rainfed rice)j (c) the adoption of recommended crop varieties was very high for sorghum, peanut and cotton, and high for other crops; it was low for rainfed rice; (d) consumption of fertilizer increased over the project period, and full and partial adoption was relatively high for most of the irrigated crops and for cotton and sorghum. Seed treatment also had relatively high adoption, while the use of pesticidas was generally low, except for cotton and irrigated sorghum and grain. The use of fertilizer and pesticide increased over the project period on a state-wide basis; - 65 - An 3 Page 3 (e) common re&sons for non-adoption of pesticides were an absence of the need for protection and high cost; lack of knowledge was cited as a significant reason for non-adoption in the case of rhizobia on groundnut (251) and for recommended variety for rice (171) and groundnut (35%); in cases where lack of knowledge was a significant reason for non-adoption, there was usually a very marked difference between distric!-s, suggesting variation in the quality of messages and programming of visits; an unavailability problem was the major reason for not using rhizobia on groundnuts. Tamil Nadu (a) CFs tended to adopt more recommended practices than other fermers; (b) CFs and others who had frequent contact with extension workers tended to have higher yields than those with less frequent contact. However, this was not true for all districts; (c) adoption was high in most districts for recommended varieties; adoption was moderate for fertilizer application, although variable by district; thfre was ar Apward trend in the use of all basic fertilizer nutrients on a state-wide basis; however, seed treatment tended to have a very low adoption rate; (d) with high-value vegetable crops, adoption rates of recommended practices were high and extension staff were a very important source of the technology dissemination; and (e) major re-.sons for non-adoption were water scarcity at critical times for irrigated crops, and adverse seasonal conditions and economIfinancial considerations for rainfed crops; lack of knowledge of the technology was not a significant factor for non- adoption. - 67 - Annex 4 Page 1 Use of the Reference Group Concept 1. The concepts enunciated in this annex are not new, and are based on accepted principles of social behavior in rural situations which have been interpreted into extension techniques. The concepts of reference groups, problem census to understand farmers' perceptions of constraints, and structured group meetings to discuss problems and introduce relevant technology are being used in extension services.1/ It should be noted that the principles and concepts can be incorporated into any extension program, regardless of the title under which the methodology goes, as long as there is face-to-face contact between the farmers and the extension staff. 2. The use of established reference groups implies the need to identify them. Techniques to accomplish this, identify persons of influence, and obtain preliminary perceptions of production constraints and problems as defined by group members are available. A common method uses a quick informed-person survey technique to obtain a general knowledge of the farming systems and resource situations in each village community, followed by the design and execution of a simple, partial household survey. This requires planning skills, some knowledge of production systems, and common sense analytical ability on the part of the middle level management staff who oversee its implementation. It does not require computer hardware or software for analysis. (There are other complex systems such as FARMAP designec. by FAO to provide physical, technical, economic and social data, 1/ The concepts were introduced into the Queensland Dept. of Primary Industries extension service in Australia in the late 60s and are still being used today. In the same period, the concepts were introduced nationally into the Papua New Guinea extension service, but fell into decline along with extension in general after independence In 1975; the principles were reintroduced into oae province in 1987 (as extension is currently organized or a provincial basis). The concepts were initially adopted in Thailand in -rea/1and development projects in northern provinces which were jointly financed by the Bank and AIDAB (1980) and subsequently in AIDAB projects. In 1987, the Thailand Dept. of Agricultural Extension initiated a program to train all 11,000 extension officers in the concepts of reference groups and problem census, as well as multisectoral area analysis techniques. The program proved too ambitious, largely because of the lack of skilled and experienced trainers for such a massive program. In late 1989, the Thai Government commissioned a feasibility study to identify how the training program for adoption of the concepts could be reactivated, as their introduction remained a priority for the Department of Agricultural Extens ton. A review of the T&V extension program in Indonesia in 1988 also recommended the adoption of the concepts to improve the acceptance and effectiveness of the extension service. The concept of problem census with farmer groupa by the VEWs is now being adopted in the nine most important agricultural provinces. - 68 - Annex 4 Page 2 which do require computer hardware and software.) Although official or political representatives of a community may not be identified as influential in the adoption of agricultural technology, the surveys would identify this structure for appropriate use in the extension system. It is possible that in some circumstances the informed persons would be knowledge- able enough to reliably define the reference groups in a village community. In this case, the partial household survey could be avoided, and the VEW would ascertain the characteristics of the group in structured group meetings. However, this alternative would place heavy reliance on the skills of each VEW and would likely lead to very variable results. 3. In initially tunicating with the defined groups, the VEW would conduct a structured p.-- em census. Successful experience and methodology in conducting such meetings through division into small discussion groups of five to ten people are well documented. The review of farmers' problems and progress being made in resolving them would be repeated each year. The definition of problems as perceived by the communities is a key element in the proposed modification of the T&V system as practiced in India. 4. Many of the constraints indicated on the problem census are unlikely to be of a technical nature. They could refer to such items as unsatisfactory pricing, marketing difficulties, lack of access to credits, interruptions in power supply to irrigation pumps, or unavailability of inputs when required. These concerns would be discussed and understood by the VEW who w-uld would then convey them for attention to his supervi6ors and to interagency coordination meetings at the village and district level, as well as feeding the information into the seasonal research-extension planning and review workshops. The VEW would also maintain liaison with the interagency bodies and periodically report to the contact groups on progress in resolution of the identified non-technical problems, without this detracting from his or her primary function in dealing with the technology in farming systems. Solution of other constraints may require substantial public sector investment, e.g., regular flooding requiring flood protection and an extensive primary and secondary drainage network. However, many constraints should be able to be addressed through changes in management, technology or resource use and are responsive to technical agricultural services; these would include problems such as low soil fertility and crop yields, crop pests, variable start or finish to a rainfall season causing periodic crop losses and income risk, inadequate condition of bullocks for work at the start of the plowing season, scarcity of fuelwood supplies, a small land resource yielding insufficient food for subsistence, or static or declining real incomes. Some constraints such as marketing problems which may require village organization in the form of transport organizations (as mentioned above) can also be responsive to direct inputs from the extension service in areas of improved harvesting, processing or presentation techniques to enhance product quality or transport endurance. 5. Research and extension technicians usually believe they have or can develop technology which will improve the productivity of a particular agroecological environment or zone. In the usual T&V system, these - 69 - Annex 4 Page 3 improvements become the basis for the "messages" which are scheduled for a production season. The design of the "messages" should take into account the information concerning farmers' problems and situations as fed back from the VEWs. As pointed out in the text, however, this is commonly a weak aspect of the T&V implementation. The problem census technique ensuros the necessary focus on this critical element in technology transfer. The feedback of the perceived farmers' problems to the regular seasonal workshops can alter the orientation or prioritization of applied research and even technical recommendations, but the most prominent effect is likely to be in how the technology is formulated and presented to the contact groups. It is completely "farmer-centered" in that it is always analyzed from a return-resource cost viewpoint, the amount of risk (especially in production response and market price) involved in using the incremental resources, and how a particular technology or set of technologies can be interpreted as meeting the expressed concerns of the farming communities. The latter can only be done in a general way at the state level, but can become more applicable at the district level; it reaches its full development in the regular preparation sessions for VEWs with their supervisor and trainers prior to scheduled visits. 6. At the level of the VEW and his supporting supervisory and technical statf, a schedule would be drawn up on how the perceived technical problems should be addressed, and how available technology (previously identified as "messages") can be presented to the community as contributing to the resolution of these problems. The focus becomes "which problems should we work through with the groups in a particular month so that they appreciate the applicability of known technology?"; rather than "which technology do we recommend for particular crops at a particular time?". Presentation of the scheduled program to groups or group representatives would proceed through the sequence of agreed problem, discussion on factors involved in the problem, how the problem can be addressed, and who in the group would undertake a small demo-trial to verify a proposed solution, or to what extent should the technology be adopted and what is needed to do this. 7. Adoption of these concepts could have implications for the education level and training of extension staff, the number of staff per 1000 farmers, and the frequency of regular visits to the group/farmers. The more analytical, or "problem-solving" approach to extension would certainly benefit from a higher average education level than exists in VEWs in some States (paras. 82 to 84 in the text). It would also require, at least for a number of years, an extension methodology specialist at the AEO level who can introduce and train SMS, AEO and VEW staff in the reference group/problem census approach through case studies and in the regular training sessions. This implies having a core agency responsible for training staff and trainers. Because of the current low emphasis given in agricultural faculties to farm management economics and the social sciences, it may also be necessary to bring in outside consultants to train the specialist staff in the established techniques. 8. With less emphasis on purely information transt r, the frequency of visits for a season can be sensibly adjusted to a program agreed with the - 70 - Annex 4 Page 4 group, the VEW and his supervisor. There would be less need for a standard schedule against which VEW performance could be checked, as supervisors could more reliably assess effectiveness by obtaining responses from each group on this matter. The longer term implementation of the T&V system concerns many officials, as the concept of supplying relevant technical messages at fortnightly intervals raises questions of availability of new practices over successive seasons. The reference group approach, which incorporates an annual problem census with farmers, avoids this problem of repetitive messages by organizing the program around farmer-nominated problem. The emphasis on dealing with groups and obtaining agreed group action would also permit a reduced ratio of VEW to farmers. SUGGESTED RELATIONSIP BETWEEN ADOPTION RATE, EMTENSION TECHNIQUE, AND CHARAC=ERSTICS OF THE TICHNOLORY =N U C UITn INVOLVED TYPE OF INNOVATION * Simple, and able to be easily imposed on + CONTINUUM 4 Complex, and dependent on the adoption existing practices of a number of innovations to attain the * Relatively low incremental cost in cash input potential benefit or other resources ftSinificant incremental cost, especially * A high benefit-cost ratio is easily If associated with a spbstantiol risk demonstrated factor in obtaining the projected * Benefit is not significantly influenced by benefit uncontrollable factors so as to make it a Difficult to demonstrate (often risky Investment associated with complexity), and is a * Required inputs to adopt the technology are major departure from the practices readily available traditionally used in the community Benefits can be quickly realized * Benefits are longer-term with little benefit in the short-term OFE 1 COMMUNITY AND RESOURCES Progressive attitude to agriculture + CONTINUUM 40 A very traditionalconservatve attitude a Commercial, market orientation In production to farming practices systems 0 A sub-subsistence or subsistence farming * Higher average education level system, or one with very little marketed * Ability to organize and/or demand services production from private and public sector Institutions Lover education level " Ready access to credit * Isolated in location with difficulty In m Reliable climate or reliable irrigation obtaining agricultural services (inputs, facilities which make agriculture le risky, credit and marketing) * Poor and/or very limited land resource and capital backing * Irregularity in climatic conditions (without facilities to overcome this) which makes agricultural production risky. MOTMNION APPROACH REQUIRED FOR OUICK MDPTION Supply of up-to-date, timely information to CONTINUUM Supply of te ormation aloe is overome technical problems should result in rapid unlikely to result in rapid adoption uptake through a community of a technology which through a community; the use of socio- has a satisfactory benefit-cost*ratio gica vrinciples in working with farmer reference groups to help them appreciate how and which technology can assist them in meeting their concerns and objectives is more likely to obtain quicker adoption. 0400"totdun of technoleg enhances the effect of cI pproaches to extteion * oe euaio ee AttachIent I 73Page 1 GOVERNMENT OF INDIA MINISTRY OF AGRICULTURE kN (DEPARTMENT OF AGRICULTURE & COOPERATION) KrIabV Bhaesa, Nawa Dhi - 110001. D.C. MIS?RA Joint Secrttary- cUm-Extension Commissioner 0.0-N& r May 10, 1990 Dear how 0o&Ads , Please refer to your letter of March 29, 1990 forwarding Project Performance Audit Report covering the State Extension Projects of West Bengal, Bihar, Kerala, Maharashtra, Tamil Nadu and NARP for our comments. 2. The report which is primarily based on the project completion report of the respective State Extension Project, has since been examined and our project-wise comments enclosed. Yours sincerely, Encl : As above ( D.C.Misra ) Mr. Graham Donaldson, Chief, Agriculture, Infrastructure and Human Resource Division, Operations Evaluation Department, The World Bank, 1818 H Street, NW Washington, D.C. 20433, U.S.A. Attachment I Page 2 Government of india Ministry of Agriculture Department of Agriculture and Co-operation (Extension Division) PROJECT-WISE OBSERVATIONS/COMMENTS ON PROJECT PERFORMANCE AUDIT REPORT OF THE WORLD BANK INDIA West Bengal Agricultural and Extension Project Bihar Agricultural Extension and Research Project Kerala Agricultural Extension Project Maharashtra Agricultural Extension Project Tamil Nadu Agricultural Extension Project National Agricultural Research Project [Credits 690-IN, 761-IN, 1028-IN, 1135-IN and 1137-IN and 855-IN] West Bengal The PPAR pointed out severe implementation problems like inordinate implementation delay, non-utilisation of the World Bank credit, non-deployment of project staff which have affected the quality of the extension services in the state. Other issues like diversion of extension staff to non-extension activities and inadequate educational qualifications of the VEWs transferred from CD Departments, were the matters of permanent concern remained throughout the project period. The problems of location-specific recommendations, poor quality of SMS support, improper selection of contact farmers, inadequate use of M&E findings and ineffective support of input supply were also faced during the implementation period. The PPAR, however, points out that by and large the state had attempted to undertake the programme agreed to at appraisal and adopt the key principles of T&V extension. The problems and constraints identified by the PPAR have already been recognised by the Directorate of Extension. Besides the constraints indicated earlier, the main problem faced by the state was the weak research and extension linkage during the course of project implementation. The research set-up remained under the charge of Directorate of Agriculture while the fundamental research responsibility remained with the State Agriculture University. The vacancy position both on the research and field extension concept also remained alarming affecting programme implementation adversely. Attachment I 75 - Page 3 The project has been closed in March, 1985, however the basic tenents of the re-organised agricultural extension system are in operation in the State. The component of information support to agricul-r- extension has been strengthened by the efforts of G.. ernment of India (GOI) through a centrally sponsored SLecial Sub-Project. Also the Directorate of Extension (DOE) made a minor but important contribution by up-grading the professional competence of 20 VEWs. West Bengal's performance in utilising the GOI sponsored training SMS workshops remained at the level of 1621 only. Bihar The PPAR pointed out that the state of Bihar also had acute implementation problems like implementation delay, non-achievement of the State-wide coverage, non-utilisation of the World Bank credit, delays in deployment of extension staff, poor qualifications of the VEWs specially transferred from CD department etc. These remained matters of concern throughout the project period. As far as the technological component of the project was concerned, the PPAR pointed out that the extension messages mostly confined to the major crops during the initial period. However towards the end the project management made the nessages broad- based which however lacked the economic tocus. The other deficiencies observed during the implementation of the project covered poor SMS support, improper selection of contact farmers and inadequate use of information generated by M&E unit. The DOE endorses the observations made in the PPAR in respect of the reasons which hampered the implementation of the T&V project in the state. Inspite of the repeated emphasis at various levels, the staffing problem did not improve as also the performance of training programmes. However budgetary support and poor research and extension linkages were observed as some of the important problems still to be solved by the State. The GOI have been recently approved a Special Sub-Project to strengthen the adaptive research component in the State. In GOI sponsored training workshops for SMSs, the utilisation percentage of seats by the State of Bihar was only 40 per cent. As many as 345 VEWs were provided an opportunity to up-grade their professional competence by the financial support from the DOE. Tamil Nadu The PPAR observed that the re-organised Agricultural Extension system in Tamil Nadu was implemented successfully and the entire amount of credit was also utilised. The Attachment I - 76 - Page 4 issues of budgetary support to the project and input supply were effectively provided under the project. The PPAR further pointed out that the progress of civil work component remained rather slow initially. However it accelerated from the middle of the project. Poor qualification of the VEWs transferred from the CD department was one of the concern experienced during the project implemeatation. Some of the minor problems like improper selection of contact farmers and inadequate support of SMSs were also revealed during the project implementation. However the PPAR pointed out that the transfer of extension messages on broad-based approach has taken place, as also the focus on group approach has been witnessed during the project implementation. The DOE while going along with the observations of the PPAR, experienced that the Tamil Nadu Extension Project was one of the successfully implemented extension projects in the country. The research-extension linkage on the input supply arrangement, training support, and the information support have been duly taken care of during the project implementation. Likewise the financial support to the project and the reimbursement of the World Bank Credit also went on smoothly. The re-organised extension system in the state has taken firm roots in the state and in its second phase (under preparation), there is an innovative approach to broad-based extension. This project is likely to be pre- appraised by the end of May. The utilisation of Government of India sponsored SMSs workshops in the state of Tamil Nadu remained at the level of 46.7 per cent and the performance in up-grading VEWs' professional qualifications through GOI support remained at low level. Maharashtra The PPAR poirited out that the project was successfully implemented utilising the entire aiount of credit. The state's efforts in financial planning, expenditure and reimbursement went as per the schedule. The problems of diversion of extension staff to non-extension activities, and dual control on Lhe staff at district level were also observed. The PPAR has taken a serious note of selection and role of contact farmers. Apart from quality selection, the M&E surveys pointed out that 12 per cent of them were unaware about the selection and 54 per cent of them were not aware about their roles. However 94 per cent contact farmers knew about their VEW by the end of the project period which appeared a significant improvement. Though the project was implemented successfully, the PPAR stated that impact of the project on utilisation and statewide yield changes were not attempted. The research extension- linkage and broadening of extension services covering land- based activities broadly focusing on dryland horticultural crops, were the salient features of the implemented project. Attachment I -77 - Page 5 The PPAR further indicated that the contact farmers who had frequent contacts with extension staff tended to have higher yields. The adoption of recommended crop variety was observed to be high for sorghum, groundnut and cotton, however, it was observed to be in case of rainfed rice. The consumption of fertilisers and use of pesticides, increased over the project period on a statewide basis. The DOE goes alongwith the observations of the PPAR. However the review missions mounted by the GOI adequately emphasised on the state for proper and proportionate selection of contact farmers. The DOE further strengthened state's efforts by augmenting training facilities in the state by a Special Sub-project and also assisted in computerisation of M&E units in the state. Maharashtra's utilisation of seats in GOI sponsored 9 courses appeared to be satisfactory touching the level of o uer cent. Kerala The PPAR reports that the Kerala State '.i a successfully implemented the programme and fully disburbe. the credit. The initial implementation of the project was hampered due to vacancies at the VEW's level and the situation continued with varying degree during the project implementation. The majority of the field level staff was in position towards the end of the project period; however, the staffing position at the higher level appei-red satisfactory. The slow progress of civil works, diversion of extension staff to non-extension duties, inadequate training support, improper selection of contact farmers and inadequate use of M&E findings were some of the issues which continued as constraints on the implementation of the project. As regards the technological impact of the project, the completion report revealed that the technology transfer has been systematised. The awareness of farmers and adoption of production recommendations have consistently increased use/off-take of inputs and pressure on the research for appropriate technologies, have shown significant improvements. The re-organised agriculture extension established through T&V project, has since been modified by the State Government since September 1987 establishing 'Krishi Bhawans' at Panchayat levels which are manned by Agriculture Officers assisted by 2-3 agricultural demonstrators, catering to extension and development needs of the farming community. The DOE goes alongwith the comments/observations brought out in the PPAR. However, it may be added that the gaps identified during the project implementation have been considerably bridged in the areas like training, information Attachment I -78 - Page 6 communication, by floatirg special sub-projects. 29 VEWs from Kerala were professionally up-graded through GOI support. However the utilisation percentage of Kerala under GOI sponsored courses for SMSs remained at the level of 26.4 per cent. NARP As the backgrourd, the PPAR indicates that the National Agricultural Research Project (NARP) was expected to strengthen the capabilities of the State Agriculture Universities for conducting location-specific production- or.inted research on the basis of agro-climatic zones aiming at closer research and extension linkage. The PPAR further indicated poor disbursevent of IDA credit. However the final accomplishment appeared to be impressive covering strengthening of 22 SAUs, 21 Special Sub-projects and 86 research special-project. During the implementation of the project, it is observed that the concept of zonal research and of joint visits and joint trials took about 3 years to be jointly accepted. The positive outcome of the project appeared to be introduction of zonal research concept in the participating states. The PPAR further pointed out that the zonal research status report brought out under NARP by the participating states have not adequately addressed on the socio-economic constraints in farming system. It was also observed that the joint diagnostic field visits were less developed than other aspects of research-extension linkages and also it has been observed that there is need to improve the adoption and on-farm trial designs further. While agreeing with the observations of the PPAR on the NARP, the DOE appreciates it as a strong forum to improve the research and extension linkage through zonal research concept. This approach has begun to put pressure on research to develop technologies suitable to local agro-climatic conditions and there is enough scope for improvement to be made in the zonal research system based on the feed- back received from the extension service. ( D.C. MISRA New Delhi JOINT SECRETARY (EXTENSION)- 10 May 1990 CUM-EXTENSION COMMISSIONER Attaihment 11 .79.- Page 1 OVE-NMUT OF KERALfA 1,'4 # /P'v3 90/A t Agricul ture (Plann iln-A) Ueptk, Thiruvananthopuram,g lMtefl: 2 170 From The .ec,retary to Govefrjw.-'rit. To TheRsident MS.aion in Yndla, 140r24 oi 55 Itod. Esnte, New OelIM.13. 5ub:- Xorala Apricultuxrnl Extension Project conmmts on Pro ect Performftnec Aulit l?port.--orwardi.ing of fler:- Your etter dnted 29-3-1990. I am directed to invito a reforence to the letter cited and to forward hcrewJth the commntt etn Project Performance Audit Itport In respect of K.A.1.P (Credit rlo.1028 IN) for favour cf infnrmpt,.^n and necoesnry action. Yours faitrtuily Por .-oecretsirf toc 'loenui •I j~4 -il.* 06/ 08: 45 WOL D BAW H0W -80- Attachment 1I Page 2 COM1Mr ONM TM PWWCT 0PERFOR4UrE AUUIT ~KPOrT 1 ffiA AlRå1CUrURAL EX 9,0T PROJEC.r In the report it has high 1l6hted that the iuape- mntath* on was doe effectively and fLficiently ån the State. T tip'Wt of the PCoject in increasing the AgiJcultural Production of variour Cr4p was Dt up to the expectation# due to Ýaaious Éthgr oio-eormic jagtors Which are 1speial in nature a far as Kerala is Conerned. Trhe contaot termers sonected did not gqerve tiW purpose as envisagead by the' Project. Th# othe famezj were roluctait to bake adv1 ftom the contact formers and insisted and preferred direct delivery' of the tochnicajl m~8ages to' thlem. *The project has tremendsious InWaat on åcreasirg, the tecohnical knowledge of the tield Lur|tionaraies at all dvss esp*cially at the level of village LVel Workers and Agricultural OLLiers. In 1907 a series of changes were brought about in the activitics and style, of furationing of the Uepartmnt with reorganisation of the departmient. The inIvidual contact farmer öpproaCh was substtUtedf by group dppxoach. Fortnightly traLnirgre were changed aS monthly trainings. For the paddy crop, a Mw approOah Ugro CI.tnics* was intxoduomåd. ti Ago Clinick funation as a contaat point on tixed nMtified day& ,within the reach of the farmrs where they could interact ard consult the Agr. ofgicer ard VII abut their field problem~ and take advise *IUU for their crop ~nangemOnt. This has c'ttracted the trmers attentiÅ atad mfajority are availing the benelit by gathering at the fixed veMOs of taw oldnice. The £unotiong of the Agro..linice have modm good the Short .oiwmngS of the change in fix d schedul* visitS subsquent to thO reorganisation of th department. TOTHL P.14 Attachment 11 8 Page 5 The group farminy nyotem for Rio intxodued at a new approach inotead of khe ConteRI farmér approach. Group farming £or s has produced better roulta and the production h-# t-eafsed for Rice. Oy inplemmanttion of this maSsive paIam it has PrQVQd that the produet- Ion of Rice can be khanasd in. the Otate subetantially lreducir the Cost ofproduction. The ftatora othet than Igesearch a|d extension ayoIved in the produdtion prgoesa. couid ha geared up by. group approach åh the group farmng Foglowing the perisne gained from the group lopproach.in Iice production this ,style.A bity .Intuced other cropa like Coconut, Ppp^r, Vegetake etc. In 100nut g9oup management an eotion programm is aleo drawn £or the o;.ops health care with aCtive partiGipation of MnoUt' farmers, la$our' force ard voluntary organisations. This hag resuted in bringing up rore rs under socientillc mnanagemnent with wili-ful. part4cipation of the ermers. in the £epeCtiv$ areas, The reorganised' set up of the departmient han nade good start in impAn*tmntir useful sche* and agriciutural r0gram~4 with aQtive publio partolpationg. Xnareased rpduction 'trend is 1en in R1e, Cocnram aM Pemer wht~ teka stagnent previously. FOR DIRCWR OF AGRICUutURS. T.