Document of The World Bank FOR OFFICIAL USE ONLY RepottNo. 5197 AUDIT REPORT PROJECT PERFORMANCE TANZANIA GEITA COTTONPROJECT (CREDIT 454-TA) NATIONALMAIZE PROJECT (CREDIT 606-TA) PROJECT (CREDIT 703-TA) TABORARURALDEVELOPMENT TOBACCO 658-TA) PROCESSING PROJECT (CRED-:r TOBACCOHANDLINGPROJECT (CREDIT 802-TA) DEVELOPMENT CASHEWNUT PROJECT (LOAN 1014-TA) VOLUMEI June 29, 1984 Operations Evaluation Department This document has a resftricted distibution and may be used by recipients aoly in the perfoniame of their official duties. Its contents may not otherwise be disclosed witbout Wrld Bank anthobartion. ABBREVIATIONS ADS - Agricultural Development Staff BADEA - Arab Bank for Economic Development in Africa BRALUP- Bureau of Research, Agriculture and Land Use Planning CAN - Calcium Ammonium Nitrate CAO - Chief Agricultural Officer CATA - Cashew Authority of Tanzania CIMMYT - International Maize and Wheat Improvement Center, Mexico CNSL - Cashewaut Shell Liquid DADO - District Agricultural Development Officer DAP - Di Ammonium Phosphate DCA - Development Credit Agreement DDD - District Development Director DMPS - District Maize Project Supervisor EAPSA - Eastern Africa Projects, Southern Agriculture Division FAO - Food and Agricultural Organization GOT - Government of Tanzania IBRD - International Bank for Reconstruction and Development ICB - International Competitive Bidding IDA - International Development Association IERR - Economic Rate of Return IITA - International Institute of Tropical Agriculture, Nigeria LCB - Local Competitive Bidding MDB - Marketing Development Bureau MOA/ KILIMO - Ministry of Agriculture MPIA - Maize Production Inputs Account NADP - National Agricultural Development Program NAPB - National Agriculture Products Board NMC - National Milling Corporation NMP - National Maize Project OED - Operations Evaluation Department PCR - Project Completion Report PPMB - Project Preparation and Monitoring Bureau PSU - Project Servicing Unit RADO - Regional Agricultural Development Officer RDD - Regional Development Director RIDEP - Rural Integrated Development Project RMPC - Regional Maize Project Coordinator RMEA - Regional Mission in Eastern Africa SOA - Sulphate of Ammonia TANU - Tanganyika African National Union TAT - Tobacco Authority of Tanzania TCA - Tanzania Cotton Authority TFC - Tanzania Fertilizer Company TIB - Tanzania Investment Bank TRDB - Tanzania Rural Development '3ank TSP - Triple Super Phosphate TTPC - Tanzania Tobacco Processing Company, Ltd. ULV - Ultra Low Volume UNDP - United Nations Development Program USAID - United States Agency for International Development VEW - Village Extension Workers FOR OMCUL USE ONLY AUDIT REPORT PROJECT PERFORMANCE ( GEITA COTTON PROTECT (CREDIT 454-TA) ( NATIONAL MAIZE PROJECT (CREDIT 606-TA) TANZANIA: ( TABORA RURALDEVELOIET PROJECT (CREDIT 703-TA) ( TOBACCO PROCESSINGPROJECT (CREDIT 658-TA) ( TOBACCO HANDLINGPREOJCT (CRIT 802-TA) (CASHEWNUTDEVELOPMT PROJECT (LOAN 1014-TA) TABLE OF CONTENTS Page No. Preface.. i Basic Data Sheets...... ii - vii VOLUMEI: PROJECT PERFORMANCEAUDIT MEMORANDA I. INTRODUCTION AND OVERVIEW ............................... 1 Introduction ............................................ 1 Overview 2 ................................................. A. Macro-economic Crisis ...... .................... 2 B. Institutional Change . . ................ 3 3............. C. Parestatals ......................................... 6 D. Price Policy ........................................ 6 E. Food Supply ......................................... 8 F. Project Management ....... .................... 9 C. Sustainability ........................................ 10 H. Linkages and Inter-sectoral Balance ....... 10 I. The Size and Phasing of the Lending Program .11 J. The Balance of Future Lending .12 II. GEITA COTTON PROJECT (Credit 454-TA) . .13 A. Introduction .13 B. Project Origin, Preparation and Appraisal 13 C. Project Implementation........ 14 D. Issues .15 (i) Institutional Change . .15 (ii) Technical Package . .16 (iii) Mechanization .. 17 (iv) Management .. 18 III. NATIONAL MN E PROJECT (Credit 606-TA) ............. 18 A. Introduction .18 B. Project Origin, Preparation and Appraisal .19 C. Project Implementation .20 D. Issues . .22 (i) Failure to Grasp Implications of Institutional Change .22 This docment has a resriced distributionand may be used by recpients oy in the peformane of dutis Its contents maynot othewise be discbsed without WorldBank authorization. their offwcial TABLE OF CONTENTS(cont.) Page No. IV. TABORARURAL DEVEiLOPMENT PROJECT (Credit 703-TA) .... .... 24 A. Introduction ........................................ 24 B. Project Origin, Preparation and Appraisal .... ....... 24 C. Project Implementation .............................. 25 D. Issues ........... ................................... 29 Mi) The Need for Consensus of Opinion Regarding Project Objectives ............................. 29 (ii) Organization and Management . . 31 (iii) Lessons Learned ....... ........................... 33 Table 1 ........................ ......................... 34 V. TOBACCOPROCESSING PROJECT (Credit 658-TA) .... .......... 35 A. Introduction ................... ..................... 35 B. Project Concept, Preparation and Appraisal .... ...... 35 C. Project Implementation ............ .. ................ 36 D. Issues ...................... ........................ 37 (i) Importance of Adequate Project Preparation and Appraisal ...................................... 37 (ii) Project Supervision .............................. 38 VI. TOBACCOHANDLINGPROJECT (Credit 802-TA) .... ............ 40 A. Introduction ............................... 40 B. Project Concept, Preparation and Appraisal .... ...... 41 C. Project Implementation .............................. 42 D. Issues . ............................................. 43 Ai) Insufficient Attention to Production ...... ....... 43 (ii) Institutional Aspects ........... .. ............... 44 (iii) Lack of Focus on Requirements to be a Tobacco Farmer in Tanzania ............................. 46 VII. CASHEWNUT DEVELOPMENT PROJECT (Loan 1014-TA) .... ........ 47 A. Introduction ................... ..................... 47 B. Project Origin, Preparation and Appraisal ........... 48 C. Project Implementation .. 48 D. Issues . .50 (i) Causes of Crop Production Collapse .... ........... 50 (ii) Institutional Performance and Development .... .... 52 (iii) Bank Performance ......... ........................ 53 Attachment I: Cofinancier's Comments ............................ 56 Attachment II: Borrower's Comments ................... 57A Attachment III: Comments from TDB .57B Map * See Volume II for PCRs * AUDIT REPORT PROJECT PERFORMANCE ( GEITA COTTONPROJECT (CREDIT 454-TA) ( NATIONALMAIZE PROJECT (CREDIT 606-TA) TANZANIA: ( TABORARURAL DEVELOPMENT PROJECT (CREDIT 703-TA) ( TOBACCOPROCESSING PROJECT (CREDIT 658-TA) ( TOBACCOHANDLINGPROJECT (CREDIT 802-TA) ( CASHEWNUTDEVELOPMENT PROJECT (LOAN 1014-TA) PREFACE This is a performance audit report of six agricultural projects in Tanzania. The projectswere to support a wide range of agricultural activi- ties and geographic/agroeconomiczones; they are representativeof the Bank's overall lending programfor the agricultural sector in Tanzania. Details of the loan/creditamounts,approvaland closing dates, cancellations and final disbursements of these six projects appear on the basic data sheets which follow. Since Tanzania assumed special significancefor the Bank in the 1970s - the period when these projects were appraised and implemented - with its strong governmentcommitmenttowards alleviatingpoverty and the provi- sion of basic needs, and, in view of the fact that all of these six projects have failed to reach their objectives, a combinedaudit was undertakenin the hope that common issues of operationalsignificance could be distilledfrom project experience. This audit report, therefore, comprises two volumes: the first is the performanceaudit memorandaconsistingof an overview,which focuses upon the issues of operationalsignificance, followedby analysis of individual project experience. A second volume contains the individual Project CompletionReports (PCRs), five of which have been preparedby the Eastern Africa Regional Office, while that for the Tabora Rural Development project was preparedby the Borrower,with a brief overviewprovidedby the Eastern Africa RegionalOffice. The audit is based upon a review of the six appraisalreports, the President's reports as well as the relevant credit and loan agreements. Review of Bank files for all projects has been undertaken and staff associated with the six projects have been interviewed. An OED mission visited Tanzania in January 1984. The mission held discussions with officials of the Governmentof Tanzania, CCM, project authoritiesand the parastatals involved. Field trips to interview farmers in each of the project areas were also undertaken. A copy of the draft report was sent to the Borroweron April 19, 1984 for comments. Comments received from the Borrower and TRDB are reproducedin the report, while those received from a co-financier, CIDA, have also been incorporated. OED would like to thank the Governmentof Tanzaniaand individuals interviewe& for all the assistance of given to the mission in the preparation its report. * ~~~~~~~~~~~~- ii- PRoJECr PPFORHUACEAMDITBASIC DATASREll TANZANIACEITA COTTON PROJECr (CREDIT 454-TA) KEY PROJECTDATA Appraisal Actual or Actual an Z of ITEI Estimte Estimated Actual Appraisal Eltimte Project Coats (US$ million) 23.8 28.6 120 Credit Amot (US$ million) 17.5 17.2 99 Disbursed 17.5 17.2 99 Cancelled 0.3 Date Joard Approval - 01106/74 Date Effectiveness 04105/74 Date -Phylcal Components Completed 06/30181 12/31/82 Closlng Date 12131/82 12131/82 Proportion then completed (Z) 100Z 95Z Economic rate of return (Z) 14 negative Intitutional Performance poor DISBURSEMENTS CXUMXATIfE FY74 FY75 FY76 FY77 nY78 FY79 7Y80 FY81 7132 Appraisal estimte (USS million) .8 2.8 4.9 8.0 11.9 14.6 16.4 17.5 4 Actual (US$ million) .2 1.9 2.9 4.4 7.5 9.8 12.3 15.3 17.2fa Actual as Z of appraisal (Z) 16 68 60 55 63 67 75 87 99 Date of Final Disbursement: 07/06183 MISSION DATA Im Performance Ratlir g- Nontb/ No. of Staff days Types of ;lesion Year Persons in Field Specialization /b Status Ic Trend /d Problems le Identificatlon Preparation 07/72 3 18 Preappralsal 01/73 1 14 Appraial 02-03/74 6 210 /f Supervision I 05/74 2 10 E.A 3 T,0 Supervision 11 10/74 1 2 E 3 M,T,O SuperisLon III 04/75 2 12 E.A 3 F,H,T Supervision IT 07/75 1 5 E 3 2 P,N.P Supervision V 02/76 2 18 e,A 3 3 F,M,P Supervision VI 08/76 1 6 E - 3 1 F,K,P Technical Reviev 03/77 8 95 Z(l),A(5).G(2) Supervision VII 06/77 4 56 A(2).E(2) 3 3 F.M,P.T Supervision VIII 02/78 2 14 g,A 2 Supervision X 07/78 3 15 A,AE,FA 2 1 T,F,N,P Supervision X 01/79 1 7 A 2 1 T,P Supervision XI 06/79 2 14 A,AE 2 1 0 Supervision XII 04/80 1 10 AZ 2 2 r,o Supervision XIII 12/80 1 9 AZ 2 2 F,0.P Supervision XIV 06/81 2 22 A,AE 2 2 F,O,P Supervision XV 11/81 2 6 E,A 2 3 K,P Completion 11/82 2 28 Z,A 2 3 H.P GrE P DATA WoJECr Borrover Government of Tanzania Executing Agency Tanzonia Cotton Anthority (TCA) Fiscal Year of Borrower July 1 - June 30 Nm. of Currency (abbreviation) Tananian Shilling (TSh) Currency Ezxchange Rates: Appraisal Year Average (1973-74) US$1.00 - TSh 7.14 Intervening Years Average (1974 82) US1. 00 - TSh 8. 0 Completion Year Average (1982) US$1.00 - TSh 9.45 /a Informal extension of credit closing date granted to June 1983. 7ii A - Agriculturalist; AE - Agrcultural Economist; PA - Financial Analyst; 0 - Other. 71 1 - ProblePr-Fee or minor problems; 2 - moderate problem; and 3 - major problem. 7;i 1 - Improving; 2 - Statlonary; 3 - deterlorating. 71 F - Financial; P - Political; N - Managerlal; T - Technical; and 0 - Other. 7- Field activity only. - iii - AUDIT BASIC DATA SHECE PROJECT PERFORMANCE TANZANIA NATIONAL MZe PROJECT (CRgDIT 606-TA) KEr PROJECT DATA Appraieal Actual or Actual as Z of ITIH Estimte Estimated Actual Appralsal Estiute Project Costs (US$ million) 38.1 32.8 86% Credit Amount (US$ million) 18.0 14. 5 1Z Disbursed Cancelled 3.5 Date Board Approval 01/76 12/23/75 Date Effectiveness 05/76 05/28/76 Date Physical Components Completed 12/80 12/31/82 Closing Date 06/80 03/31/83 Proportion then completed by above date (see key Indicators - Annexes) Economic rate of return (Z) 37 probably negative or marginally positive _ Institutional Performance poor Agronomic Performance poor Number of project beneficiaries (families) 380.000 not available C1MU.ATIVE DISBURSEMNTS FY75 FY76 FY77 FY78 FY79 Fn80 FY81 FY82 FY83 Appraisal estimate (USS million) .6 4.2 8.7 13.8 18.0 18.0 - - - Actual (US$ million) .5 2.7 4.3 6.2 8.5 10.6 12.0 14.3 14.4 Actual Z of appraal (Z) 75 65 48 45 47 56 66 79 81 Date of Final Disbursement: 08/31/83 MISSION DATA ITEM Performance Rating noth/ No. of Staff days Types of Mission Year Persons in Field Specialization /c Status /d Trend /e Problema /f Identification 12/74 2 10 A,AE - - Identification 01/75 2 5 A,AE - - - Preparation /a 02/75 4 17 A,AE,E,EN - - Preparation 03/75 2 13 FA,EN - - Preparation 04/75 1 14 FA - - - Preparation 05(75 2 2 A, (2) - - Appralsal /a 06/75 4 14 A,AE,E (2) - - - Supervision I 06/76 1 5 A 2 2 F.0 Supervision II 11/76 2 9 AE,E 3 3 P,O,M Supervision III 05/77 1 16 AE 3 3 F,M,O Supervision IV /b 11/77 8 15 AE(2).A(4).E.FA - - Supervision V 7ii 02/78 1 12 A Supervision VI 7b 04/78 1 11 A - Supervision VII 10/78 1 14 A 3 1 F,M Supervision VII/b 01/79 1 7 A - - Supervision KX 08/79 2 20 A(2) 2 1 M,T Supervision X 04/80 2 11 A(2) 2 2 M,T Supervision XI 11/80 1 10 A 3 3 P,M Supervision XII 09/81 2 24 A,AE 3 1 P.M Supervilsion XIII 01/82 3 16 A,AE,E 3 3 P.M Supervlsion XIV 12(82 2 10 A,FA 3 3 P,M,T,O Supervision XV 04/83 2 5 A,FA - - rIlER PROJECT DATA Borrover Government of Tanzanla Executing Agencies Ministry of Agriculture and Regional Authorities Fiscal Year of Borrower July 1 - June 30 Name of Currency (abbreviation) Tanzanian Shilling (TSh) Currency Exchange Rstes: Appraisal Year Average (1974-75) US$1.00 - TSh 8.05 Intervening Years Average (1975-83) US$1.00 - TSh 8.33 Completion Year Average (1983) USS1-OO - TSh 11.50 Pollow-On Project: Potential IFAD project under identification. /a Preparation and Appraisal vere combined. 71 Mid-term review and follow-up ilfsions - Back-to-Office Reports prepared. 77 A - Agriculturalist; AE - Agricultural Economiet; E - Economist; FA - Financial Analyst; and EN- Engneer. 7i1 1 - problem-free or minor problem; 2 - moderate problem; and 3 - major problems. 7e 1 - Improvlag; 2 - stationary; 3 - deteriorating. 7F F - financia; P - political; M - managerial; T - technical; and 0 - other. - IV - PROJECT PERFOEWtACEAUDIT BASIC DATA SEiT TANZANIA: TABORA RURAL DEVELOPNT PROJECT (CREDIT 703-TA) UEY PROJECT DATA Appraisal Actual or Actual a 1 of 7TEM Estimate Eutiated Actual Appraisal Ectiltc Total Project Cost (USS dllion) 23.5 17.8 76 Credit Amount (US$ million) 7.2 4.2 S Disbursed 7.2 4.2 58 CanceIled - 3.0 Co-fiancing-CIDt (US$ million equlv.) 4.8 2.7 56 Cancelled - 2.11c ParaIlel-financed - ODA (US million equiv.) 5.6 6.47ii 114 Data Board Approval - 04/26777 Date Effectivenesr 09/77 11J1/77 Completion Date 09/82 09/82 Proportion (Z) Completed by Above Date 100 50/c Closing Date 06/30/83 06//l83 Economic Rate of Return (Z) 17 negattve Institutional Performance below expectations Number of project beneficiaries: - .a. CUNULATIVEDISBURSEMENTS/d FY77 FY79 FY79 FY80 FYs8 7182 le Appraisal Estimate (USS aillion) 0.17 2.27 5.64 7.46 9.86 12.00 Actual (USS amillon) 1.12 2-17 2.86 3.40 4.19 5.11 Actual as Z of appraisal (Z) 659 96 51 46 42 43 Date Final Disbursement: O/21/84 MISSION DATA - Performance Rating- Month/ No. of Staff Days Type of IT=M Year Person in Field Specialization /f StatuS Trend ehProblems /i Identification 06/74 1 9 05/74 1 17 Project lnformtion Brief 09/74 - - Preparatiou 10/74-06/75 7 151 A.B,C.EEF,J Appraisal- General 11/75 6 144 (approx.) A.BC,E.C,IK - Uater Component 02/76 1 10 0 Post-Appraisal 03,76 4 64 D,E,F.L Supervision I 11/77 1 30 A 1 2 T Superviston II 05/78 3 30 A,D,E 2 2 O,T Supervision hIIl 01/79 2 20 A.E 2 1 O,T 04/79 1 3 3 - - - Supervision lV/i 07/79 1 10 A 2 1 O,T 10/79 1 2 F SuperviaLon V/k 03/80 2 23 E,H 2 3 O,M,T Supervision VI 06/80 1 15 A 3 2 0.NMT Supervision VII/k 10/80 3 25 A,D,E 3 3 O,1NT Supervision VII171 07/81 2 18 A,E 3 1 N,O Supervision IX 01/82 2 11 A,F 2 1 P Supervislon X 10/82 3 36 A,B,E 3 2 P,l,P Supervision XI 06/83 2 28 A.E - - - OTE PROECT DATA Borrower Governmenn of Tanzania Executing Agency Tabora Regional AdministratLon FiScal Year of Borrover July I - June 30 Name of Currency (abbreviation) TInzanian ShLillng (TSh) Curreacy Exchange Rates: Appraisal Year Average (1975-76) USS1.00 - TSh 8.05 Interveanng Years Average (1977-83) USS1.00 - TSh 8.2 Completion Year Average (1983) USS1l00 - TSh 11.5 la Equivalent of TSh 20.16 million1 in 1982/83. 7ii In 1978/79 ODA increased its contribution wLth an amount T Sh 6.999 mllSon. 7;; Impossible to give a more meanLigful asseuent, due to the large number of components. aubcomponet and actlvitles within eacb of these. /d IDA + CIDA funds. Schedule of disbtremsnts are the beat figures available. 7W The exchange rate changed in February 1983 frm the previous rate of 12 shillings to the dollar to 11 shillings. /f Specializations: A - Agriculturist; B - Livestock Speclalist; C - Civil Engineer (roads); D- Civll Engineer (wster); z- Economist (agricultural); F - Forester; C - Financial Anayst; H - Disbursemet Officer; I - Training Officer; J - Rice Specialist; K - Loan Officer; L - Smal IndustrLea. If 1 - Problem free; 2 - Moderate problems; 3 - Major Problem. 1i - Improving; 2 - Stationary; 3 - Deterioratiug. 7T F - Financial; H - Manageent; T - Technical; P - Political; 0 - Other. I ODAaccompanied supervidsio mLssion. / CIDA aceompanied supervislon misLon. 73 Redesign proposals were revieved and found unsatiefactory. AUDIT BASIC DATASHEET PROJECT PERFORMANCE TANZANIATOBACCO PROCESSINGPROJECT (CREDIT 658-TA) REY PROJECTDATA Appraisal Aetual or Actual as 2 of ITEM Estimate Estiutee Actual Apprstial Estlmte Project Costs (USS million) 11.3 10.3 91Z Credit Amount (USS million) 8.0 7.5 94X Disbursed Cancelled .5 Date Board Approval 08/76 09/07/76 Date Effectiveness 10/76 02/15/77 Date Physical Components Completed 1981 Closing Date 03/81 12/31/81 Proportion then completed (Z) 100 Economic rate of return (Z) 25 negative Institutional Performance unsatis factory DISBURSEMENTS CUMULATIVE FY76 FY77 FY78 FY79 FY80 FY81 FY82 Appraisalestimate (USS million) 1.3 6.8 7.8 7.9 8.0 8.0 8.0 Actual (USS million) 1.0 1.0 6.3 6.8 7.0 7.1 7.5 Actual a. Z of appraisal (Z) 77 15 81 86 88 88 94 Date of Final Disbursement: 02/17/82 MISSION DATA rITM Performance Rating Month/ No. of Staff days Types of Mlssion Year Persons In Field Specialization /a Status /b Trend /c Problems /d Identlfication 07/75 Preparation 10/75 3 ApprAisa 10-11/75 5 EC(2),AC,C,(2) Supervision I 02/77 2 11 EN,EC 1 2 M Supervision II 11/77 3 10 EN(2),FA 2 1 M,F SupervLsion III 03/78 2 7 EN,EC 2 1 M.P Supervidson IV 01/79 2 6 EC,EG 2 1 M,F,O Supervision V 06/79 1 2 EC n.a./e n.a. n.a. Supervislon VI 10/79 2 6 EC,FA 2 1 M,F,O SupervLsion VII 04/80 2 7 EC,FA 2 2 M,F.O Supervislon VIII 12/80 2 9 FA,AC 2 2 M,F,O Supervision IX 05/81 1 12 PA 2 2 M,F,O Supervision X 02/82 2 12 AG(2) 3 3 M,F,O Supervision XI 06/82 3 10 AG,EC,FA 3 3 M,F,O Completion 05-06/83 1 0 AC - - - OTHERPROJECTDATA Borrower Govermment of Tanzania Executing Agency Tobacco Authority of Tanzania (TAT) Fiscal Year of Borrower July 1 - June 30 Name of Currency (abbreviation) Tanzanian Shilling (TSh) Currency Exchange Rates: Appraisal Year Average (1975-76) USS1.00 - TSh 8.05 Intervening Years Average (1977-80) US$1.00 - TSh 8.09 Completioa Year Average (1980) US$1.00 - TSh 8.18 Follow-On Project Name: Tobacco Handling Project Credit Number: 802-TA Credit Amount (USS million) 14.0 Credit Signiog: 06/78 /a C - Consultant; EC - Econodmst; EN - Engineer; PA - Financial Analyst; and AG- Agriculturalist. /b 1 - probler free or minor rroblems; 2 - moderate problems; and 3 - major problems. 7; 1 - Improving; 2 - Stationary; and 3 - deterlorating. Td - Management; F - Finance; 0 - Tobacco Production decline. 7V Only Back to Office report, rating status not available. - vI - POEJCT PERFOREACE AUDIT BASIC DATA SHeZr TANZANIATOBACCO HANDLINGPROJECT (CREDIT 602-TA) KEY PROJECr DATA Appraisel Actual or Actual as Z of ITM Estimate Estimated Actual Appraisal Estimate Project Costs (USS million) 20.5 6.9 332 Credit Amount (US$ million) 14.0 6.1 442 Diubursed 6.1 Cancelled 7.9 Date board Approval 05/16/78 Date Zffectiveneso 09/78 . 05/16/79 Date Physical Components Completed Not completed Closing Date 04/30/83 04(30/83 Proportion then completed (Z) 100 30 Economic rate of return (Z) 24 negative DISBURSEMENTS CUMULATIVE FY78 FT79 m8o Fre81 P182 PT83 Pn84 Appraisal estimate (USS mlllon) 0.2 4.8 8.8 12.2 12.2 14.0 Actual (USS million) - 0.2 2.1 3.3 3.3 4.4 6.1 Actual as X of appraisal (2) - 8 15 23 23 31 Date of Final Disbursement: 11/22/83 NISSION DATA Performance Rttlanrg Month/ No. of Staff days Types of ITEM Tear Person in Field Specialization /b Status /C Trend /d Problems /a Mission Identification /a NA Preparatlon 09177 Appraisal Supervision I 01/79 2 5 EC(2) 2 2 M.0 Supervision l 10/79 2 16 EC,FA 2 2 M,F,0 SupervIsion III 04/80 2 17 EC,FA 2 2 N,F,O Supervision IV 12/80 2 23 FA,AC 3 3 M,F Supervision V 05/81 1 12 rA 3 3 M.Y Supervision VI 02/82 2 24 AC(2) 3 2 F,H,0 Supervision VII 06/82 3 24 EC,AC,FA 3 3 7M3,0 Supervision VIII 12/82 1 16 EC 3 3 OTHERPROJECT DATA Borrower Government of Tanzania (COT) Executing Agencies Tobacco Authority of Tanzania (TAT), Tanzania Rural Development Bank (TRDS) Fiecal Year of Borrower and Agencies GOT - July 1 to June 30 TAT - March 1 to February 28 TIDE - March 1 to February 28 Name of Currency (abbreviation) Tanzanian Shilling (TSh) Currency Exchange Rates: Appraisal Year Average (1975-76) USS1.00 - TSh 8.3 Intervening Years Average (1977-82) US$1.00 - TSh 8.3 Completion Year Average (1982) US$1.00 - TSh 9.3 Follow-On Project: None /a Identification seems was a by-product of the appraisal of the National Agricultural Development Program (KADP). 7T AC - Agriculturaliet; K - Economist; F - Financial Analyst; end L - LivestockSpecialist. T 1 - problewrfree or minor problIms; 2 - moderate problema; and 3 - major problems. Tl Improvlng; 1 - 2 - stationary; 3 - deteriorating. 7T F - financial; P - political; N - managerial; T - technical; and 0 - other. - vii - AUDIT BASIC DATASHEET PROJECT PERPORMANCE TANZANIA CANEIUNT DEVELOPMENT PROJECT, PEASE I (LOAN 1014-TA) UT PROJECT DATA Appraisal Actual or Actual as Z of ITEM EZtiuste Eaet sted Actual Appraiesl Eut -tu Project Coat (US$ million) 30.3 31.5 1043 Loan Aount (US$ million) 21.0 20.5 9U Date Board Approval 05/74 05/21/74 Date Effectivenes 09/74 09/26/74 Date Physical Components Completed 02/79 12/04/80 Closing Date 12/31/80 12/31/81 Proportion then completed (see Table 3.1. POx) Ecoomic rate of return (Z) 39 negative Institutional Performance unstiafactory Agronomic Performnce poor Number of project bennficLsrlee: n.a. CUMULATIVEDISBUXSGE1MTS ry74 FY75 Ft76 Fn77 FY78 1179 Fr80 FY81 Appraieal estimte (US$ million) 1.0 8.3 14.0 18.0 20.6 21.0 21.0 21.0 Actual (US$ million) 4.7 8.5 10.6 13.8 16.8 19.4 19.9 20.5 Actual as X of appraisal (C) 471 101 75 77 81 93 95 98 Date of Final Dicbursement: 04/15/82 MISSION DATA Performance Ratlng -- ITEM momth/ No. of Staff days Types of Mlisson Year Fersons in Field Specialization /a Statue /b Trnd /c Problem Id Pre-Project Ident. 03/72 3 nae. EC,AG - Identification 06/72 2 2 EC,AC - - - Preparation 02/73 4 7 AG,EC,PK,AE - - - Preparation 04/73 1 2 AC - - - Preparation 0403 1 1 LO - - - Preparation 05/73 1 1 LO - - - Preparation 05/73 2 1 EC,AZ - - - Preparation osn3 1 1 AR - - - Appraisl 1O03 3 22 FA(2),PE - - - Supervision I 06/75 3 1 AG,LO,FA 1 - - Superviion II 11/75 2 4 ACE.FA 1 2 2 Supervlsion III 06176 1 3 AGE 1 2 n Supervision IV 12n76 1 2 E 1 2 N Supervlsion V 05/77 2 5 E,ACE I 2 H Supervislon VI 09/78 1 10 FA 2 3 M,0 Supervilion VII 03/79 1 8 PA 2 3 0 Supervilsion VIII li/79 2 14 PA,AG 2 3 0 Supervislon 7X 05/S0 2 14 AC(2) 3 3 O0M Supervislon X 02/81 1 10 FA 3 3 O.M Supervision XI 10/81 2 10 FA,AC 2 2 M,O Supervision XII 05/82 2 16 FA,EC 3 3 F.M,T,O OTHER PROJECT DATA Borrower Covernmnt of Tanzania Executing Agency Cashew Authority of Tanzanla (CATA) PFiscl Year of Borrower July 1 - June 30 Name of Currency (abbreviation) Tanzanian Shilling (TSh) Currency Exchange Rates: Appraisal Year Average (1972-73) US$1.00 - TSh 6.86 Intervening tears Average (1973-82) US$1.00 - TSh 8.03 Completlon Year Average (1982) US$1.00 - Tgh 9.30 Follow-On Project: Name: Second Casbewnut Development Project Credlt Number: 801-TA Credit Amount: Us$27.5 million Date of Board Approavl: 05/16/78 /I EC - Economist; AG - Agriculturalatt; Z - Enginser; LO - Loan Officer; FA - Financial Analyst; AGE - Agricultural Engineer; PE - Processing Engineer; and AZ - Agricultural Econowmet. /b 1 - problerfree or minor problem; 2 - moderate problem; and 3 - major problem. 79 1 - improving; 2 - staionary; 3 - deteriorating. 7i P - financial; P - political; H - mnagerlal; T - technical; and 0 - other. AUDIT MEMORANDA PROJECT PERFORMANCE ( GEITA COTTONPROJECT (CREDIT 454-TA) ( NATIONALMAIZE PROJECT (CREDIT 606-TA) TANZANIA: ( TABORARURALDEVELOPMENT PROJECT (CREDIT 703-TA) ( TOBACCO PROCESSING PROJECT (CREDIT 658-TA) ( TOBACCOHANDLINGPROJECT (CREDIT 802-TA) ( CASHEWNUT DEVELOPMENT PROJECT (LOAN 1014-TA) I. INTRODUCTIONAND OVERVIEW Introduction 1.01 The six projects evaluated in this audit report address the developmentof cotton in the Geita district of Mwanza region,maize produc- tion in fifteen of Tanzania'stwenty regions, rural developmentin Tabora region, tobacco processing and tobacco handling (separate projects) in Tabora, Iringa and Mbeya regionsand cashewnutprocessingand productionin the coastal areas - principal'y Mtwara region. These projects therefore cover a wide range of agriculturalactivitiesand geographical/agroeconozic zones;all have low or negativeeconomicrates of return. 1.02 Since 1974, the Bank has become closely associatedwith almost every aspect of Tanzania'sagricultural sector. Earlier programs covering export crops were augmentedby projects in support of pyrethrum,coconuts, sugar, dairy productionand fisheries. The NationalMaize Project was the first Bank effort directly to support food crops - followingthe drought of 1973-74. The Bank also promoted the concept of regionalintegrated developv- ment programs in four regionswhich were intendedto strengtheninfrastruc- tural and social services as well as provide inputs and extensionservices for increasedagricultural activities. 1.03 In view of the representative nature of these six projectsto the Bank's overall agricultural portfolio together with the fact that thls combinedaudit has highlighted a number of issueswhich are common to some or all of the projects,this report begins with a brief overview of ten issues of operationalsignificance which have been drawn from project experience. Followingthe overview,individual projectexperience is examined In greater detail in ChaptersII-VII. 1.04 At a number of points the audit and project completionreportson these projects throw additional light on the constraints on performance resulting from institutionaland policy weakness. The agriculturalsector report,l/ prepared by the region during 1982, provides a comprehensive and systematic presentation of Bank views regarding the institutional and policy 1/ Report No. 4052-TA dated August 26, 1983. -2- changes which the Government of Tanzania might implement to improve agricul- tural performance. It is the purpose of this memoranda, therefore, tc focus primarily on issues where the larger environment has important implicatious for the Bank itself in project design and implementation. Overview A. Macro-economic Crisis 1.05 Performance in relation to all the projects has been negatively affected by the extended period uf acute economic difficulties facing Tanzania. Scarcity of foreign exchange, resulting shortages of fuel, spares and other inputs, inadequate supplies of consumer goods, and stringency in the government budget have all created difficulties both in the initial implementation of projects and subsequent project performance. This was true from 1973, at the time of the first oil crisis, when most of these projects under consideration were being negotiated or designed. From 1979 when the projects should have been in full operation, the situation became much more critical with the country facing the combined effects of the break up of the East African Community, the war with Uganda and a second round of oi'l price increases; all these exogenous factors have had a critical impact on the balance of payments creating extreme foreign exchange shortages. 1.06 This is not the appropriate place to discuss the policy options open to the Government of Tanzania to resolve these difficulties. However, some lessons can be learnt for project level decision-making: (i) initial project design should take account of the likely macro- economic environment. There is no point in financing a project if the foreign exchange shortages outlined above2 /, for example, prevent its effective operation (para. 4.11). This means that appralsal at the micro-level should take account of likely aggre- gate trends, including the overall impact of all project decisions on likely supply constraints. In the early 1970's this involved, in particular, an assessment of the foreign exchange situation and, perhaps equally important, the trained manpower constraint. From 1973 onwards the international economic situation, as well as evident tightness in the Tanzanian balance of payments, should have led to a greater consideration of the foreign exchange constraint. In practice this implies effective coordination within the Bank between program and project work. (ii) project design and appraisal should be based on a realistic, rather than desirable, picture of the larger environment. Projects should be designed and appraised taking into account a feasible timetable, scarcities of foreign exchange, and subsequent difficulties with maintenance. 2/ The Region points out that in 1977/78 Tanzania had reserves of US$300 million. -3- (iii) the aid package should respond to the real resource situation. This might swing the balance in favor of program rather than proj- ect aid. It might suggest, in particular cases, a more modest scale of project which was less demanding of the use of scarce resources. It might involve incorporating into the project fndAing to ensure the supply of inputs for the operation of the project after its construction and for the maintenance of the project capital stock. This might involve the funding of projects over a much longer period than is the current practice. B. Institutional Change 1.07 Experience of all these projects illustrates the inherent risks when far-reaching institutional change is bastZly implemented by the govern- ment. 1.08 Over the years 1973-76, precisely the period when implementation of these projects began, three major changes which profoundly influenced all of the projects were carried through with great speed and energy - namely decen- tralization of government, villagization and the reorganization of agricul- tural marketing following the transfer of functions from the cooperatives to the parastatal marketing authorities. 1.09 WUhile all these moves were addressing real problems and incor- porated thinking which could have been expected to contribute to a solution, of importance is that the Bank failed to understand 3 ! three aspects of these institutional changes. First, that there was de facto as well as de jure decentralization of government to the regional level in 1972, the Bank mean- while continued to deal with the Ministry of Agriculture as if executive and administrative authority had not been divested (paras. 2.08, 3.16). Second, there was less than full understanding as to how these regional administra- tions operated in practice. Line reporting systems for agricultural activi- ties were discontinued with technical matters subordinate to the political officers at every geographical level within the region (para. 4.36 et seq). Third, failure to grasp that compulsory villagization introduced in i974 was more than movement of people into villages. It changed the decision making process at the village level with responsibility for production decisions now resting with the village government and not with individual farmers (para. 3.17). This lack of understanding has greatly undermined sound project design and implementation; in view of the importance of this issue, it is discussed in some detail. 1.10 The decentralization of government isolated the Ministry of Agri- culture from the rural areas. It separated the Ministry from any direct intervention in production; its residual role was providing advice, if called upon. Nominal responsibilities included agricultural research and control of 3/ Staff of the Agriculture Department (OPS) have some disagreement with this statement - see footnote 28, paragraph 3.15. -4- the rapid proliferation of activities being undertaken by new crop para- statals. Both were circumscribed, however, by: (a) first, work undertaken by research stations was separated from events taking place in the villages; (b) second, the Ministry had no information base, and, therefore, limited plann_ng and policy-making capacity; and (c) third, its manpower resources were severely curtailed, which resulted in the Ministry losing track of where competent agricul- turalists were working. 1.11 The institutional framework to support decentralization was vested in the political structure, headed by the Regional Commissioner, ranking as minister and appointed by the President. The Regional Colmissioner's control extends down through the hierarchy 4 ! to the district level in the office of the Area Commissioner and to the ward/village level in the Ward Secretary. The chief civil servant, the Regional Development Director (RDD) is appointed by the Prime Minister's Office, to whom he reports on all budgetary and staffing matters. In contrast with the political arm, administrative/ technical officers at each geographical level are not responsible on a line basis but report instead to their corresponding political officer. They have no direct links to their senior technical officers either in the region or in the parent ministry in Dar-es-Salaam. Thus for example, the RDD has an agri- cultural officer (RADO) as an adviser, but the RADO has no line control over the agricultural staff based in the districts, wards or villages. Without any means of information gathering, planning becomes difficult (para. 3.16) while implementation is similarly hampered when agricultural officers have no field staff, vehicles or other resources with which to carry out programs (para. 4.38). 1.12 Ujamaa and the villagization program have been two distinct policy measures for the grouping of farmers into villages. The Arusha Declaration in 1967 advocated a voluntary return to perceived traditional practices of communal cultivation and land ownership. The expected social and economic benefits were to be boosted by the provision of infrastructural services and input supplies as additional incentives to the formation of Ujamaa villages (para. 2.17). 1.13 Farmer reluctance to join these villages led the government to announce, in late 1973, the compulsory villagization program; by December 1975 over 9 million people were grouped in villages. The important feature of villagization was not the increase in the number of villages, which in fact was relatively small, but rather the substantial increase in their size. By statute, villages had to have a minimum of 250 families which has bad far reaching consequences for the traditional husbandry practices employed for both crops and livestock (paras. 4.15, 6.32, 7 .17 a). 4/ The regional administration operates at several geographical levels: region, district, ward (a group of five villages) and village. - 5- 1.14 Decisions concerning the procurement of inputs (para. 3.17), which crops to cultivate and the disposal of produce, particularly after 1976 (para. 1.17), became the responsibility of the village government. This required the village government to operate a fairly sophisticated set of accounts for each of the 250-500 families within the village, a task fre- quently beyond the capacity of the skills available at the village level. Consequently, the deduction of input costs and the distribution of crop proceeds, including that for quality production, have often been subject to arbitrary decision-making. 1.15 To promote equity, each family was allocated similar land area from the total available to the village. However, an individual's land area is not contiguous, and while the individual may have discretion on the type of crop planted on some of his land allocation, within certain areas he is ex- pected to grow the particular crop designated by village government-this applies particularly to cash crops. This approach has certain costs in that the skilled product grower (e.g., for tobacco) is limited in the area he can devote to the crop (para. 6.31), while the unskilled is obliged to devote a greater area to the crop than either his inclination or skills would other- wise warrant. 1.16 These shortages of skills at the village level have acted to constrain individual initiative, and were disincentives to production- particularly quality production. 1.17 Marketing. The role of the village government in agricultural production was reinforced in 1976 when all cooperatives, which traditionally had marketed cash/export crops and provided extension services and input sup- plies, were dissolved. These reponsibilities devolved to individual village governments with the result that the parastatal crop authorities and the credit agency were required to deal directly with a very considerable number of village governments. This increase in responsibilities to the parastatal crop authorities has also exacted a toll on their operating efficiency which is discussed below in the section on parastatals. 1.18 A conclusion that might be drawn about the current situation faced by Tanzania is that further change to improve institutional performance should be through a more considered and incremental process, rather than through a new round of abrupt transformation and that an experimental, pilot project-type approach to explore a variety of options is preferable to blanket implementation of change. 1.19 In addition to the lessons which can be learnt from these six agri- cultural projects, wider issues are raised by this experience. How should the Bank allow for/interact with institutional change in project design and implementation? Various approaches have been tried in these projects but the overall lack of success highlights the complexity of the issue which deserves more attention than can be adequately addressed here. -6- C. Parastatals 1.20 Parastatals have been used as the chosen instrument for many as- pects of Tanzanian rural development - crop purchase and storage, processing input and credit supply and, for the period covered by implementation of these projects, the provision of extension services. 1.21 Parastatal problems as observed in these projects might be viewed as being in part transitional, reflecting manpower shortages, the difficul- ties of working in conditions of economic crisis, and managerial inexperience (para. 7.17b). In part, however, they are more structural, resulting from too broad a definition of tasks which placed a heavy burden on the para- statals, and from an inadequate system of incentives and/or discipline, whether deriving from the market or from accounting and other control systems (para. 6.20 et. seq). 1.22 There have been many sources of parastatal growth, including the emphasis placed on the public sector by the Tanzanian Government, the transfer of the cooperatives' marketing and credit functions (para. 3.19), and the transfer of extension functions from the Ministry of Agriculture as a result of decentralization. Some of this growth in parastatal responsibili- ties is now being trimmed, with the restoration of the cooperatives and the transfer of extension responsibilities back to the Ministry of Agriculture. In the past year also the government has been engaged in a thorough-going investigation of parastatal performance and possible reforms. However, certainly in the period under review during which these projects were imple- mented, the parastatals were key influences on agricultural development. 1.23 Interestingly enough the Bank itself initially has been a signifi- cant supporter of parastatal growth, providing the finance for the expansion of parastatal facilities and setting the standards for staffing. As these institutions are central to Tanzanian rural development such support was necessary. However, there is a clear recognition now that the parastatals have many weaknesses, and present a considerable burden on national budget demonstrated both in the experience of these projects and in the concern cur- rently shown by the government for improving parastatal effectiveness. D. Price Policy 1.24 Price policy emerged as one significant factor affecting output performance in the cases of cashewnuts, cotton and tobacco. Appropriate incentives are a necessary condition for smallholder supply response, although in many years price effects were probably overshadowed by other factors detrimental to a good output performance. 1.25 One factor causing a deterioration in real agricultural price incentives was increasing marketing and processing costs. In the cases of both cashewauts (para. 7.17b) and tobacco (para. 5.18), when output fell short of appraisal estimates, a heavy burden of parastatal overhead expenses had to be carried by a diminishing output base, thus compounding the initial production problem. The costs incurred in this expansion of parastatal acti- vity have become one cause of poor agricultural performance. 1.26 Another general factor was that the government was slow in adjusting its agricultural price policy to the endemic inflation which it experienced after 1973, and which contrasted dramatically with the stable price levels of the 1960's. As a result agricultural price adjustments have lagged behind increases in the general price level, but during the last two years the government has begun to adjust its policy in this regard. 1.27 Even if strict conditionality were desirable in relation to price incentives in principle, it is difficult to ensure this is achieved in prac- tice. The Eighth Annual Review of project performance audit results 5 / high- lights that the determination as to what constitutes "correct" pricing policies can vary depending on whether the prices are being viewed from the perspective of the farmer, the government or the Bank. For example, farmers generally maximize returns to their scarcest production factor which may or may not coincide with the productivity goals established at appraisal. Furthermore, the government may have income distribution or food security objectives which differ from those held by the Bank. 1.28 It is also to be noted that the Bank bas not shown great prescience on the price issue. When it did raise the issue (para. 7.17b) it did so late in the day. In the mid-1970's the Bank was arguing for increases in price incentives for food production, while later it criticized the Tanzanian authorities for shifting relative prices too much in favor of food and against export production. A further difficulty is that the issue of price policy phases into questions regarding the exchange rate and the overall fis- cal situation, issues very much the preserve of the International Monetary Fund.6/ 1.29 Conclusions which can be drawn about the relation between price policy and projects are that: (i) project appraisal should include an explicit treatment of the price policies necessary for project success and the likely constraints on achievement of such policies; efforts are required to improve the weak data base at the micro-level regarding supply response; one weakness of the development effort during the past twenty years has been the failure to improve the information base for decision making so that in the short term decision makers are continually forced back to conjecture and a priori reasoning. 5/ OED Report No. 4101 dated September 9, 1982. 6/ AGR and the region state that 'the IMF's views on whether or not pre- vailing exchange rates and fiscal policies are resulting in serious distortions is current and short ternmwhile our own project and pricing policy thinking needs to be long term". (ii) the impact of the project itself on the incentivesystem (e.g., its impact on bureaucraticoverheads) should be considered; (iii) the monitoring component in projects should include, wherever possible,an attempt to assess the impact of the incentivesystem on farm performance,so as to alert the nationalauthoritiesof any problems arising in the project as a result of price policy and supply problems (para. 7.17b). This should be seen more as a matter of timely warning of potentialproblems than of condition- ality,because the broader consequences of nationalpoliciescannot be assessedin the context of one project. E. Food Supply 1.30 The increasein food prices in the seventieshas been widely idenr- tified as a reason for the shift in farmer effort away from export crops. This was so in the cases of tobacco (para. 5.02) and cotton, and may have been the case to a lesser extent for cashewnuts. 1.31 This shift in price relationships was particularly great in rela- tion to food prices in uncontrolled markets, and as such was not a direct result of official2rice policy. High food prices were a symptom of under- lying food scarcity 1L. Any reversal of that incentiveshift which resulted in a movement of resources from food to export crop production would increase Tanzanianfood import requirements. Of course, reductionsin inefficiency In uarketing,storage,regional pricing, and input supply and an across the board increase in agriculturalprice incentives could result in an increase in total agricultural supply. However, this possibilityshould not obscure the importance of the underlying food supply problem. 1.32 This problem emerged in 1973-74 and has been with Tanzania since. It is not possibleto analyze it at length here. It is a comp'lex phenomenon, an outcome of population growth, urbanization, changes in consumption patterns, deteriorationin the transport system, climatic factors, and weaknessesin the marketingand storagesystems. It was the threatposed by this problemwhich led to the NationalMaize Program, the urgencyfelt being one cause of the program'sfailure, resultingin overreaching in both size and geographical coverage. 1.33 However,it remains true that unless means are found to expand food supply, export crop projects will face difficulties, both because the cash returns to be gained from producing food for sale in local markets will divert effort from export crop productionand because of the high priority given to meeting family subsistencefood requirements. The dilemma is, as the failure of the NationalMaize Programindicates, that the food problem is 7/ This was not so in all cases. High officialprices for cassavaand some minor food crops did generateoversupplyin some years. - 9 - not to be resolved by a short-term crash program but with a persistent long- term effort backed up by steady improvement in institutional capacity, supporting services and the policy environment. 1.34 This suggests that high priority should be given to research on productivity improvement and policy related to rainfed food production in African conditions. F. Project Management 1.35 At the early stage of implementation of this group of projects, Tanzania still had a great scarcity of trained manpower. This posed problems for the management of projects. Although the situation is now very differ- ent, with a much readier availability of Tanzanians with university level and technical training, nevertheless, there are lessons to be learnt for future project design and implementation. The points to be made are fairly obvious but easily neglected if concern is concentrated on the technical rather than managerial aspects of projects. 1.36 Project design was insufficiently sensitive to the scarcity of trained manpower (para. 7.10) and failed to allow for the time factor in melding individuals recruited into a team (para. 2.20). As a result, a pre- mature buildup of disbursements was planned, with institutions expected to manage large sums before they had established the capacity to do so. There was, in implementation, some slippage but it was still the case that a slower buildup would have been a preferable path. In this regard the tendency of Bank supervision to be concerned about the achievement of physical infra- structure targets and disbursement schedules, as dated in the project docu- ments has its dangers (para. 5.16). 1.37 The experience of these projects also suggests that there is a limit to which the shortage of local managerial capacity can be overcome by expatriate talent under existing arrangements. It was not easy to recruit able people with the requisite technical and human skills to work in diffi- cult rural conditions on short term contracts. In most cases they did not stay long enough to have a major impact on project performance. As the most important function of technical assistance is to build up local managerial capacity, expatriate inputs are only likely to be effective when the need for their services is clearly recognized by project personnel and where attitudes on both sides result in a high degree of cooperation. This was often not the case. For the future the local manpower constraint is unlikely to be as tight as it was in the 1970s. The design of project organization and manage- ment should aim to utilize existing manpower resources and provide a manageable learning process. 1.38 Management performance is not only a question of availability of skills, but it also requires compatability between the tasks expected of an institution and its capacity, and the structure of incentives and disci- pline, as has been noted in the paragraphs above on parastatals. - 10- 1.39 The underlying problem is that the implementation of even small scale rural development projects demands a wide range of skills and a good deal of managerial flexibility in response to experience gained during proj- ect implementation and to fast changing external circumstances. If such capacity is not present, the record of these projects suggests that the external supervision process cannot supply it. 1.40 One conclusion which can be drawn is that the scope of a project should be limited enough that it can be managed by the institutional capacity which can be realistically expected, and that institution building should be a prime concern in project design. G. Sustainability 1.41 The sustainability of the projects over the longer term has emerged as a general issue. At the project level there is a failure to as- certain the availability of the financial and other resource requirements to maintain projects after the capital spending phase covered by the credit is completed.8/ This weakness is exacerbated by the immediate economic crisis and by the fact that further phases of projects were not pursued, as was earlier expected (e.g., Geita Cotton, Tabora Rural Development), because of poor performance. 1.42 There is no easy answer. An open-ended commitment to continue with finance even in light of poor performance would be imprudent; indeed, in some instances early project cancellation might be wise. However, rural develop- ment is a lengthy process and requires a long-term commitment to ensure that projects operate effectively in what should be their productive phase and to build on the lessons of experience. H. Linkages and Inter-sectoral Balance 1.43 One inadequacy in project design and performance relates to the linkages between project operations and the development of related economic activities. 1.44 In addition to the general dependence of these projects on the state of the national economy, in a number of cases there was a more specific dependence on conditions in closely related sectors. 1.45 The cashewnut project, which was seen by the Bank as primarily a processing project, failed because of the decline of cashewnut production (para. 7.15). In the case of the cotton project, bottlenecks emerged in cotton processing (not part of the project) which would have proved a more serious embarrassment if the project had achieved its production goals. In the case of tobacco, overoptimism on production trends led to an 8/ In some projects, moreover, financial burdens have been generated which have weakened the institutions involved (paras. 3.10, 5.07, 7.17b). - 11 - overexpansion ln processingcapacity (para. 5.01). The Bank's supervision understandably concentrates on those aspectswhich are centralto the project itself, but in so doing can easily neglect developments in related sectors which may determinethe successor failureof the project. On occasion,this was exacerbated by supervision missions tendingto comprise one or two indi- viduals each with particular technical specializations. This presents a dilemma as there are risks in redefiningprojectsmore broadly to incorporate too wide a range of activities(e.g.,agricultural production, marketingand processing)unless the basic technologyimposes such a connection (e.g., tea), as in most cases the managerial task is often already too broadly defined. However, such inter-dependency could have been treated more ex- plicitly at the design and appraisal stage, which might have led to more caution as regardsthe scale of project and also suggest the need for related projectsand policiesto ease bottlenecks affectingprojectperformance. I. The Size and Phasing of the Lending Program 1.46 Given that the projects under review all ran into difficulties related to overall constraintson foreign exchange, trained manpower and institutional capacity,it is necessaryto consider the size and phasing of the lending programin Tanzania,particularlyin light of a parallelbuild-up in bilaterallending. 1.47 There was a rapid expansion of the lending levels to Tanzania during the period. This is not to question the long-termneed for support nor the real attractionsof a stable regime and of the commitmentTanzania made to the basic needs approachwhich had, after all, widespreadcurrencyin the donor community in the late 1970's. 1.48 However, the speed of the build-up,with many projects being pre- pared and appraisedrapidly (paras. 3.06, 5.08, 7.26 a), overstretchedimple- mentationcapacityand detractedfrom the performance of all projects. This has generated what could be described as a foreign assistancecycle' in Tanzania, with donors reacting against the negative results of their own previous excess. A more carefullyphased build-up,includingdue considera- tion to the risks involved9 / particularlyof technicalrisks brought to the attentionof Bank management(paras. 3.06, 4.30, 5.15), probablywould have had a better result both in terms of short-term impact and longer-term sustainability. 1.49 While it is difficult to establish, as it involves qualitative judgement,it seems probable that enthusiasmto build up lending levels to Tanzaniain the periodwhen these projectswere appraisedresultedin a lower 9/ Most notably in respect of the inadequatepreparationand testing of technologies(paras.2.13, 4.21, 7.26b). - 12 - level of caution in decision making at both the project and the sectoral level than was desirablel°/. 1.50 In particular,there has been a misreadingof events in Tanzaniaby Bank's most senior managementll/. The annual or biannual reviews,on which the Bank's lending program was based, consistentlyindicated a tone of optimism and downplayedadverse informationemanatingfrom the field. In creating the optimistictone, these reviews clearly relied on statementsof intent as expressed by the country's senior political leaders, without adequateattentionas to how, in the past, similar statementsof intent had differed in actual implementation. Thus, the Bank's lending program was based more upon a desire to help the country in quantitativeterms (para. 3.05), rather than upon a clear strategy for the country's agricultural sector, supportedby a pipelineof well-prepared projects. 1.51 It needs to be emphasized that, with the exception of those projects supportingparastataldevelopment,notably tobacco and cashewnut, the projects were designed by the Bank with little input except at minor technicallevelsby Tanzanians. In general,however,the Bank only discussed potential projects and issues with those who did not hold the power and authorityto effect implementation,(para.3.16). 1.52 In this and in the issues previously highlighted,L 2! as the 3 agriculturalsector reportl/ concludes,in dialogue with the Tanzanians an explicitrecognition regardingpolicy improvement, of the degree to which donors must bear some of the responsibilityfor past mistakes would be appropriate. J. The Balanceof FutureLending 1.53 The negativeexperience with these projectscould lead to a danger- ous paradox in relationto future Bank lending. The experiencewith these projectsdemonstrates the dauntingtasks to be undertakento promoteagricul- turaldevelopment and, alongwith sectoralanalysis, could be read to suggest IO/ The Region attributessubsequent poor projectperformanceto inadequate projectdesignand weak technical data, both featureswhich it considers unrelatedto the size and phasingof the lendingprogram. Il/ programs have expressedthe view that to the degree that there was a misreadingof events, this was a more pervasiveproblem throughoutthe Bank and one not restingsolelywith seniormanagement. 12/ Tnter alia, the need for effective coordination between program and projectwork (para. 1.06),the importance institutional of understanding change (para. 1.09), of developingpolicies for key issues such as prices (para.1.29),food supply (para.1.31),etc. 13/ Report No. 4 052-TA,2p cit paras 1.09 and 7.62 et seq. - 13 - the need for greater levels of resourceallocationto agriculture. However, the particularfailures and the difficulties in the institutional and policy environment they reveal could lead to an understandablecaution regardingnew agricultural projects. The paradox is that having, in sectorallevel work, 1 4/ that the balance of overall investment argued in Tanzaniawas unfavorable to agricultureand biased towards other sectors, project level experience could lead in practice to the Bank shifting the balance of its own lending program away from agriculture,despite the importanceof this sector to the economy in generaland to the balance of payments in particular. II. GEITA COTTON PROJECT (CREDIT 454-TA) A. Introduction 2.01 Credit 454-TA provided US$17.5 million in support of a US$23.8 million project designed to intensifyand expand cotton productionin Geita district. At completion,the projectwas expectedto increasethe production of cotton by 42,000 bales of lint and maize by 42,000 tons. In the event, cotton output has declined from pre-projectlevels, maize output may have increasedslightly but this cannot be directly attributedto project invest- ment. The economicrate of return is judged to be negative. The Credit was closed in December 1982, with an undisbursedand cancelledamount of US$0.3 million. B. ProjectOrigin, Preparationand Appraisal 2.02 Tanzania'ssecond five-yeardevelopment plan (1969-74)projectedan ambitiousincrease in national cotton productionby 1975, from 0.4 to 0.7 million bales. Government initially expected that this would be accomplished by mechanized collective 1 5/ production, but the Bank doubted that such col- lective farmingwould be effectiveand urged a modified approach. In August 1972, government proposed a US$19.0 million project to establish large, tractor-cultivated blocks of cotton productionin the Geita area; each block would be subdividedinto smaller individualparcels for daily managementbut the conceptof collective ownershipwould be retainedwhere possible. 2.03 Bank staff continue to doubt that mechanizationcould be effec- tively managed and whether it would be economic. Problems associatedwith land identification in the large blocks and subsequentland sub-division and allocationwere envisaged;in addition,Bank staff raised technicalreserva- tions concerning the adequacy of available practices for cotton and maize production. 14/ Report No. 4052-TA, op cit. 15/ Under the auspices of Ujamaa villages (see para. 1.12). - 14 - 2.04 Nevertheless, appraisal proceeded in February/March 1973. Working with Tanzanian technical staff, the project was designed on very different lines from that initially envisaged by government. Instead it was expected that production would be attained by focussing on three groups within the farming population: (a) selected 'progressive' farmers, i.e., those with ready access to land and who were regarded as competent in the use of purchased irr- puts; (b) groups of smaller farmers who had consolidated their holdings to allow contiguous cultivation of at least 5 ha; and (c) Ujamaa villages with communal blocks of sufficient size. 2.05 Project design expected the use of tractors for land preparation 'where this would be economically and financially justified". Husbandry practices and farm input levels for crop production were to follow the blanket technical recommendations derived from the Ukiriguru cotton research station, which is in the vicinity of Geita. 2.06 To support the promotion efforts the project provided for: (a) credit for farm inputs and tractor hire; (b) tractors for some Ujamaa villages; tc) extension services together with supervision of credit handled by cooperatives; (d) storage for inputs and incremental production; (e) road construction; (f) agricultural research and land planning; (g) rural health and community education services; and (h) staff housing, offices, laboratories, etc., within the project area for the management unit established to operate the project, which was under the overall executive responsibility of the Tanzania Cotton Authority (TCA). C. Project Implementation 2.07 The PCR describes the somewhat complex story of project implementa- tion in some detail. Two months after project effectiveness, compulsory villagization was in full progress, which had an immediate impact on the project, designed to encourage progressive farmers with a range of improved services. By the second year, costs were getting out of hand and a special - 15 - review mission proposed that the roads, storage facilities,tractor and health components be scaled back by between25-50%and the educationcenters, soils laboratoryand evaluationunit be eliminatedaltogether. Thus it was not until the third year that attentionwas given to production when another fundamental problem facing the project becameapparent:the lack of viability of the technicalpackage for cotton production. This in turn underminedthe effectiveness of the extensionstaff, whilst putting the credit program in jeopardyl6/ since it was found that using the fertilizer as recommended would result in farmers incurringsubstantial financiallosses. Followinga problem project review, an eight member technicalreview missionwas mounted in March 1977 and alternative recommendations were developedand agreedwith government. At the same time, however,RMEA disagreed with these major tech- nical recommendations and six months later, the Bank suspendeddisbursements for inputs. 2.08 Throughoutthe period, the Bank was unable to focus on the implica- tions of the institutional changes at the regional and village level. The Bank had little confidencein the projectmanager (para. 2.21) and thus there was little dialogue. As a result the institutional vacuum in which the proj- ect found itself- being implemented by structures which had littleor no inr- fluence over productiondecisionsat the regional,district or village level - continued. Project activitiesbecame confined to the achievementof the infrastructuretargets and research. The audit draws on the experience describedto focus on four key issues: the effect of abrupt institutional changesintroduced by government;the inadequate technicalspecifications for the projects;mechanization and managementdeficiencies. D. Issues Change (i) Institutional 2.09 A fundamental error at the time of appraisalwas not to take fully into account the decentralization of government, which was occurringduring project preparationand appraisal. While it would have been difficult,at that time, to assess the potential of the embryo administration for the Mwanza regionl7 /, in-depthanalysis of the implications of decentralization may have indicated,however, that this administration would make all future decisionson such key matters as land allocation, and would, in any case, be heavily involved in the promotionof the principles of Ujamaa. It may even have been determinedthat, by putting responsibility for the project under the umbrellaof TCA and its parent, the Ministryof Agriculture,the project was vesting authority for implementation in institutions which soon would have little responsibility for activities at the field, district or regional level. -- 2.10 There is no doubt that anotherbasic assumptionat appraisal,con- cerning the types of farmers to be reached,was an anachronism. First, all 16/ See TRDB comments;TRDB confirmsaudit findings. 17/ Geita is a districtwithin the Mwanza region. - 16 - the evidence available at that time illustrates that government was promoting egalitarianism and not a progressive elite. Beginning in 1969, all official documents omit any reference to progressive farmers, as had been evident in earlier years. Second, the original project proposals from government clearly implied that the land of the larger farmers was intended for redis- tribution-either to Ujamaa vill -es or to large, sub-divided, blocks. Third, there was continuing emphasis towards Ujamaa and political impatience with its slow adoption on a voluntary basis, illustrated by Operation Dodoma which villagized that region in 1973. 2.11 Clearly, some form of drastic change in the area was very much a possibility and the lack of recognition of this indicates a weakness in the Bank's overall country intelligence, particularly amongst Program staff. 2.12 The audit agrees with the PCR that villagization (see para. 1.12 et. seq.) undercut the whole thrust of the project as it was designed but finds it difficult to understand why the Bank did not attempt a redesign, -8 / with new and attainable objectives being identified, which would be more in keeping with the new Institutional framework at the village and regional level. Failing that, the project should have been closed. Instead, matters were allowed to drift with ad-hoc measures being taken from time to time in an attempt to salvage the project. (ii) Technical Package 2.13 Project benefits were anticipated from improved yields an increase of 120Z from the second year. This increase was based on a recommended level of fertilizer and insecticide use. These recommendattons turned out to be inappropriate having been drawn from practices employed on the Ukiriguru cot- ton research station and not tested under field conditions. This points to two somewhat basic lessons: first, the need to check on how the research station develops its recommendations; and, second, the dangers of such stations working in isolation from farmers. Ukiriguru was founded in 1920 and had developed an excellent reputation, however, in recent years, it had stopped verifying station results on farmers' fields before publication. Over a period of time, soil conditions on the station farm had continued to be maintained or even improved, gradually becoming atypical of the sur- rounding countryside. 2.14 Evidence on files makes it clear that differing views on what constituted an appropriate package continued until project completion. Moreover, conflicting signals were given to Tanzania from time to time on the suitability or unsuitability of one particular type of fertilizer, use of which had been deemed disadvantageous in certain African soils for many years prior to project inception. 18/ The eight member review mission, mounted in 1977, made certain technical recommendations but there continued to be lack of consensus in the Bank on the viability of these recommendations. - 17 - 2.15 The inappropriateness of the technical package was in itself suf- ficient to have recognized that the project objectives could not have been achieved, and, as was concluded in respect of the institutional changes (para. 2.12), should have been recognized earlier and redesign or other remedial action undertaken. Ciii) Mechanization 2.16 The mechanizationcomponent raises two issues. First, how the prospectof additionaltractorsaffectedthe regionalauthority's view of the project;second, the technical merits of tractoruse in the projectarea. 2.17 Prior to appraisal, government expectedthat the Bank would finance a large number of tractors,which the regional authorityperceivedas being necessary for development in the region. Empirical evidence suggests that it was the prospect of receiving these tractors Which led to the authority downplaying, during negotiationl9l,its original emphasis on collective farming operations. When final projectdesign included credit for farmers to hire existing tractors, with only a limited number of tractors to be purchased by the project, the authority lost further interest in the proj- ect 2 0 /. 2.18 Bank staff had strong reservations about both the economics of tractor use and the managerialcapacity available to handle tractors. In view, however, of government'soriginal proposals the Bank felt that some tractors must be financed. The decision to proceed with project appraisal in the then prevailing political and institutional environment placed the mission in a difficult situation. Government was committed to promoting col- lective production, and wanted the tractors as incentives to encourage farmers to begin communal cultivation (para. 1.12). Appraisal,however, had to give economic considerations a high profile and while Bank staff doubted that tractors could be used economically,they could not prove this 2 1/_ Thus, a considerable number of tractors were includedbut, by imposingstrict conditions which made their procurement highly unlikely, the Bank sent coD- flicting signals to government as to what its position was on the question of tractoruse. 19/ While the regionalauthorityplayed no part in the formal negotiations, officialsbelieved that the project would result in the supply of a largenumber of tractors. 20/ Further, empirical evidence suggests, the regional authority deliber- ately chose Geita for the first forced villagization effort within the Nwanza region. 21/ This was subsequentlyverified,however,when the diseconomiesof trac- tor use, except under high standards of management, became evident during project implementation. - 18 - (iv) Management 2.19 The project was managed by a unit in the cotton authority, a para- statal with a rapidly proliferating range of responsibilities and activi- ties. Moreover, the Ministry of Agriculture had been weakened and its status diminished (para. 1.10). Thus the direction and support which could be given to the project's management team was very limited. 2.20 The pace of anticipated disbursements for the project presumed the rapid assembly of project staff and their ability to work immediately as an integrated team. This was a naive presumption, given the prevailing manage- ment constraints. First, it was known that the country was short of skilled manpower; second, although government had given assurances that manpower would be made available, there had been little planning of precisely how staff would be recuited and from where, particularly within the context of major institutional changes occuring in the country; third, when and if recruited, staff would be working in an area where housing and office accomr- modations had still to be constructed; and fourth, even under good conditions a group of individuals needs a reasonable time period before routines and team-work can be built up. 2.21 The position of project manager was filled without consultation with the Bank, contrary to the Credit Agreement. Bank staff objected to the nominee, but no effort was made to have the appointment rescinded. The result was mutual distrust between the project manager and early supervision missions which led to supervision giving expatriate staff a higher profile than was wise, thereby increasing local resistance to the recruitment of still vacant expatriate positions. 2.22 It is probable that the individual appointed as project manager had less than the full complement of skills necessary to meet the extremely onerous tasks inherent in project design. It is also probable that early supervisions had less than adequate appreciation of how difficult it was to overcome these underlying constraints. In particular, evidence suggests that the project manager was well aware of the political, administrative and operational constraints involved in handling a project within one district of a region. In contrast, supervision staff, working through the auspices of the cotton authority, inevitably had less than adequate understanding of these issues, or of the difficulty which the project faced with the regional authority not having any final responsibility. III. NATIONAL MAIZE PROJECT (CREDIT 606-TA) A. Introduction 3.01 The National Maize Project estimated to cost US$38.1 million, inr cluded an IDA credit of US$18 million and US$5 million provided by the Arab - 19 - Bank for Economic Development in Africa (ABEDIA), became effective in May 1976 and closed in March 1983. The project was expected to result in an incremental increase of 195,000 tons of mnaie per annum by 1983, but failed to reach its objectives. The economic rate of return is considered to be negative. B. Project Origin, Preparation and Appraisal 3.02 The project was a direct outcome of both a growing concern in Tanzania to maintain food self-sufficiency, following a drought and national food shortages in 1973/74, and a major recommendation of the Bank's Agricul- tural and Rural Development Sector Study 2 2 / with the adoption of the National Agricultural Development project (NADP). This strategy proposed that the Bank should finance productive investment with a minimum of support for infrastructure. In particular, it expected that investment should cor- centrate on particular commodities and in areas where technical intervention was already proven. 3.03 It was expected, therefore, that the National Maize project would support a relatively modest investment in four or five high potential maize- growing regions, concentrating on selected districts/villages within them. An underlyingassumptionwas that project area and funding could be expanded rapidly if successful. 3.04 Preparation and appraisal were condensed into a single processand carried out rapidly. Conceived in late 1974, the project was approved in December 1975 and became effective in May 1976. 3.05 The initial project concept of a carefully focused and relatively small project was affected by three developments. First, a request by government for a larger project and for simultaneous action in all twenty regions of the country;23/ second, an uncritical willingness by the Bank to increase commitments to Tanzania; and third, an unexplained decision on the part of the Bank to abandon the NADP 24 / approach shortly after adopting it as strategy. 3.06 Since project preparation had been carried out within the original terms of reference (para. 3.03), the greatly expanded focus and the rapid pace of appraisal resulted in a heavy degree of data extrapolation and less than adequate planning for those regions added to the original project plan. Concern was expressed within the appraisal team itself and specifically 22/ Report No. 541a-TA dated December 10, 1974. 23/ A compromise of 13 regions was finally agreed with government, however, during implementation a furtner 2 regions were added. 24/ The National Agricultural Development project was appraised but not implemented. - 20 - raised by the Bank's CPS members during the reviewvprocess. However, an apparent desire both to increase the lending volume and to meet Tanzania's wishes was allowed to outweigh technical reservations, notably: (a) the absence of suitable technical packages for the uninvestigated areas, (b) un- certainty as to which institutional agencies at the regional level could distribute farm inputs2 5 / and (c) concern regarding both the institutional capacity of the regions to carry out the project and the ability of the central coordinating unit in the Ministry of Agriculture (PSU), to handle such a dispersed effort. 3.07 In the decentralized administrative system of Tanzania, implementa- tion responsibility had to be given to the regional administrations through- out the country. However, a coordinating and administrative unit (PSU) was to be established and was initially expected to (a) provide limited technical support to the 4 or 5 original regions, (b) consolidate, procure and distribute the total annual input supply requirements, and (c) maintain proj- ect accounts. The work load for this unit expanded proportionately with the increase in project size, and furthermore, its final role included the onerous task of helping the regions solve the remaining technical problems, i.e. the lack of both appropriate packages and mechanisms for input supply. The powers of the PSU in carrying out its functions derived from its per- ceived status within the Ministry of Agriculture, (although it was known that regional administration staff reported to the Prime Minister's Office), and from its leverage as the dispenser of project funds, particularly its control of the revolving fund set up to finance annual input requirements. C. Project Implementation 3.08 At the start of the project, senior level staffing of the coordi- nating unit was rapidly accomplished, and government met its commitment in filling the additional posts within the regions. 2 6! Despite the lack of supporting staff and office accommodation, PSU nevertheless procured and dis- tributed the first year's input supplies. 3.09 From the second project year, the unit continued to serve as the procurement agency for equipment, but at the same time its prime function in coordinating, procuring and distributing farm input requirements disappeared, the unit thus failed to play a major role in project management as envisaged at appraisal. 25/ Appraisal had identified the existing distribution system as a major constraint. 26/ Empirical evidence suggests, however, that this may have been accom- plished by merely moving existing staff members into the designated title positions - with little or no real strengthening of regional ad- ministration. - 21 - 3.10 With the failure of the revolving fund, the government attempted to introduce a formal credit system with the Tanzania Rural Development Bank. With the assistance of the coordinating unit, TRDB extended its operations to include the handling of inputs, and in the limited field that it was able to operate, proved more efficient than the parastatals in effecting timely deliveries. However, apart from some limited successes, involvement proved a failure since, as with other agencies, the development bank was obliged to deal with villages as single entities and could not work with single commo- dity groups of farmers. Furthermore, the only collateral a village could offer was first call on the proceeds from cash crops handled by other para- statals (e.g., the crop authorities for tobacco, cotton, etc.). Thus the loan recovery problems remained much as indicated in para. 3.18. The net result of TRDB involvement was merely to transfer some of the financial 27/ strain generated by the project onto that organization. 3.11 The Bank carried out a review of the project in October/November 1977, and recommendations to restructure the project were eventually accepted by government. The project area was reduced to a few regions, much in line with original concept (para. 3.03), and funds were reallocated to strengthen the Ministry of Agriculture by providing staff in planning, monitoring and market research. 3.12 Whether maize production has increased at all as a result of the project is debatable. Presumably these inputs used would have had some positive effect, but the extent to which they were used properly, effi- ciently, or even on the maize crop, is impossible to determine. In the absence of data from the specific villages which received project inputs, the only indicators are the estimates of gross maize production by region. The reliability of these estimates is, however, doubtful. If they are accepted, production has increased overall by 17% in the 15 regions finally included in the project. Of these, however, 12 regions show a 10 decline in production, and 3 (all included in the original project plan) show a large increase of 0ocK. 3.13 Empirical evidence suggests that increased yields in these high- performing regions was not the only factor explaining high levels of output, but that the area under maize increased substantially (in some cases at the expense of cash crops). Nevertheless, in the audit's judgment, the project has had a significant impact in these three regions, thereby confirming the Bank's initial premise that concentration of investments in high potential areas would produce high returns - despite adverse institutional and macro-economic factors. 3.14 If production has in fact increased by 17% and all of this incremental value attributed solely to the project, then the project's economic rate of return just becomes positive. At the same time, however, 27/ TRDB point out that the recovery rate for maize input loans was a mere 5% (TRDE comments, page 57C). - 22 - the project has increasedthe financialproblemsof the crop authoritiesand credit agency involved,and thereforeweakened them. It has also adversely affected farmers'receptivityto productiveinputs in the majorityof regions covered by the project. D. Issues (i) Failure to Grasp Implications of Institutional Change 3.15 The single most importance explanation for project failure was the poor understanding of three institutional changes introduced by government.28 / This has already been described in some detail in the here. overview (see paras. 1.07 et. seq.), only a few aspectsare illustrated 3.16 In the first instance,there was insufficientappreciation of the fact that the Ministry of Agriculture,together with its units, had little power or influenceover activitiesat the regional level. The regionswere able to ignore both Ministry/PSU requests for information and their recommendedoperatingprocedures. When the revolvingfund for farm inputs collapsed,29 ! any residual leverage which PSU may have had, disappeared. The failure of the unit to procure vehicles in the first year 30 / further diluted confidencethe regions may have had in PSU/Ministryof Agriculture. This, combined with lack of visits by unit staff, led to PSU becoming an agency of little relevanceafter the first year of implementation.Despite this diminishedpower and respons..bility,the PSU continuedto be the major contact point for Bank supervision. 3.17 Second, there was insufficientattentiongiven to how the regional administrationoperated in practice. The lack of any technicalchain of responsibility (para. 1.11) made the collectionof systematicand reliable data on input usage and requirementsdifficult and precluded logistical planningparticularly when combinedwith the lack of any effectiveorganiza- tional/administrativecapacityat the villagelevel. Government's insistence that each villagebe treatedas a single entity, for the provisionof inputs 28/ AGR is of the opinion that the specificinstitutionalrelationships in Tanzania were well understood, but that there was an inadequate assessmentof the risks they potentiallyposed. 29/ The fund collapsedbecause (i) government decided not to reimburseits 75% subsidyelement and (ii) the regionswere unable to extract the 25% cash payments for inputs from farmers - or where they did, the funds were not repaid to the coordinating unit. 30/ PSU was unable to overcome procurement difficultiesunder ABEDIA proce- dures. - 23 - and collection of produce,31/ meant that sub-groups of farmers could not coalesce around a particular interest, in this case improved maize growing. 3.18 In the absence of demand at the village level for inputs3 2 /, some regional administrations set arbitrary targets for input requirements which were delivered to villages by fiat. This in turn led to ad-hoc distribution within the village, with little regard for either allocative or technical efficiency. Farmers became liable for debt with the repayment terms only vaguely defined. This process inevitably increased farmer resistance against using the inputs, contrary to the purpose of the project. 3.19 Third, with the reorganization of marketing arrangements, the input distribution system became even more confused. In 1976, government suddenly abolished all cooperative unions (para. 1.17), which until then had carried out this responsibility at the regional level. 33 / Government decreed that the already overstretched crop authorities assume responsibility for input supply and distribution, even for crops outside the parastatals' terms of reference. 3.20 This merely added to the financial problem of the parastatals which attempted to recoup the farmers' dues by deducting these from payments made for other commodities sold by the village. This resulted in, for example, the tobacco growers of the village paying the debts of those who had been obliged to take maize inputs. Since there were few accounting skills within the village (para. 1.14), individual liability for credit or inputs was impossible to determine. 3.21 The inequities indicated above had adverse effects beyond that of increasing farmer resistance to receiving maize input supplies, they have also further reduced incentives to grow any cash crops (tobacco, cotton, etc.), for which they now received less than full farmgate value. 31/ TRDB points out that there was no consistent policy towards villages on the question of village stores: some villages were expected to build stores on credit from TRDB, others had stores built from grants out of funds administered by the DDDs. Moreover, there was a lack of clear policy regarding ownership of these stores. (TRDB comments page 57C). 32/ Because of: (i) distrust of government-sponsored activities following the villagization program which resulted in a retreat to subsistence production; (ii) inadequate producer prices combined with an absence of consumer goods; (iii) the fact that farmers were now required to pay cash for inputs which, in earlier years, had been delivered free of charge. 331 The initial project concept of limited areas also took into account areas where Cooperative Unions (CU) operated reasonably efficiently. - 24 - IV. TABORA RURAL DEVELOPMENTPROJECT (CREDIT703-TA) A. Introduction 4.01 Credit703-TA providedUS$7.2 million over five years for the first phase of a long term rural developmentprogram in the Tabora region. The projectwas to both strengthenthe newly establishedregionalauthorityand support pilot agricultural, livestockand fuelwood programs at the village level. Total project cost was expected to be US$23.5 million with two co- financiers,the Governmentsof Canada and the United Kingdom, providinga credit of Can$5 million and a grant of US$5.9 million respectively. The remaining US$5.9 million (25X) of the project was the responsibility of the government and the villagers. The project failed to reach its objectivesand was closed in June 1983 with a total project cost of US$17.8 million. The economicrate of return is negative. B. ProjectOrigin,Preparationand Appraisal 4.02 Tabora region is the largestin the countryand covers 9 percent of the land area of mainlandTanzania. The town of Tabora, the regionalcenter, is situated some 800 km north west of Dar-es-Salaam with poor road connec- tion; the main access is by railway. Rainfall is markedly seasonal, and reducesprogressively from the west of the region to the east, where rainfall reliability is also significantly lower. 4.03 The project arose from a request by governmentfollowingprepara- tion of the Kigoma Rural Development Project (Credit No. 508-TA), and was prepared with the assistanceof the Bank's Resident Mission in East Africa (RMEA). The project was appraised in November/December 1975. A post appraisal mission visited the region in February/March 1976. 4.04 When the project was approvedon April 26, 1977, it was noted that an attempt had been made to learn the lessons of previousrural projectsin Tanzania. In particular,the need for a trial period which would allow the developmentof both technicalpackages at the village level and for strong managementat the regional level. The project was expected to be flexible since it was a pilot venture; some risks of failurewere clearly spelledout to Bank managementat appraisal. 4.05 The project supported the following components: - a program, covering about 30 villages, for developing and applying techniques designedto increaseproduction of crops; - provisionof equipmentfor the existingtobaccoresearchstation at Tumbi; - a program in approximately12 villages for developingand applying techniques designedto improve livestockproduction; - 25 - - constructionfor approximately 75-100 villages of low-cost water supply systemsand simple water supply installations. Preparation of a RegionalWater Master Plan for Tabora region; - and reconstruction maintenance of accessand network roads; - establishmentand managementof plantationsin selectedvillages to supply fuelvoodfor tobaccoprocessingand domesticuse; - preparation master plan for the of a land-useand land-capability region, and enlarging and strengthening the Regional Village PlanningUnit; - provisionof technicalassistance and infrastructureto strengthen the region'scapacityto implementthe project,monitor and evalur ate its effect and assist in the planningof furtherinvestmentsin the region. 4.06 The projectwas to be integrated into the existingregionalorgani- zation; overall responsibilityfor policy decisionsrestedwith the Regional Development Committeechairedby the RegionalDevelopment Director. C. Project Implementation 4.07 The audit commendsthe Borrower for the project completion report which providesa thoroughand frank analysisof the constraints and problems surroundingthe implementationof this project. This is the first completion report prepared by the Borrower and it provides a useful record of the experienceobtainedfrom the Bank's and the government's efforts to support rural developmentin Tanzania. 4.08 The projecthas achieved some degree of success in its efforts to upgrade regionalroads, to assist in land use planningas well as in respect of tobacco seed production, but has achievedvery little on all other compo- nents. To a large extent, successcan be attributedto the fact that these components have been implemented in isolation from the organizationand management structure supported by the project. The success of the seed tobacco productionprogram is in large part due to the efforts of a single individual who was able to identifythe requirements needed to establishthe program, to use initiativeand enterprise in obtainingthese necessaryinputs and, most importantly, to train the laboratorystaff. Only in this latter component has therebeen a substantial effortmade - and successachieved,in the trainingof local staff. 4.09 The roads component,which was the second largest in terms of fund allocationand financedby the UK grant,was to providetechnical assistance, vehicles and equipment to permit upgrading of regional roads to all-weather standard. A road maintenancepolicy was expected to prevent what had been the prime cause of deterioration of the regional roads system in the past. - 26 - The project also envisaged that the 206 villages in the region, without an access road, would be linked to the main road network by the end of the five-year project period. 4.10 Actual achievement varied significantly. Appraisal estimates regarding the speed at which the regional roads could be upgraded proved optimistic. Only 50Z of regional road targets were met; no village access roads were built. 4.11 An evaluation of the regional roads component 3 4 / by the consul- tants providing the technical assistance attributes the difference in actual accomplishments to both the severe operating conditions, particularly fuel shortages, and the lack of an appropriate infrastructure to support plant operations within the region. Another reason was the fact that there was virtually no physical progress during the first two years of the project. Bank supervision noted the difficult operating conditions as well as initial adaptability problems of internationally recruited staff3 5 /. 4.12 Another major change during implementation was in the techniques employed. At appraisal it was expected that the regional roads would be up- graded using capital-intensive equipment. With the increasing shortages of fuel and spare parts, however, casual labor was employed on a moderate scale. Initially there was some concern, particularly on the part of the MOW, regarding whether standards were being sacrificed, but given the shortages of critical inputs for capital intensive techniques the consultants showed commendable flexibility. 4.13 While there was some achievement of targets for the regional road network and a maintenance policy was put in place for the network as a whole, by project completion, no access roads had been established to any of the 206 villages, largely because the necessary self-help contributions by villagers were not forthcoming. The evaluation report 3 6/ concluded that there were at least three factors explaining general unwillingness to contribute the neces- sary self-help labor for construction of access roads. First, such roads are ouly of benefit to the village if used. During the period under review, input supplies to villages were erratic, consumer goods often unavailable, cash crops were frequently not picked up by the crop authorities and the few buses and trucks to be seen on the roads suffered from frequent breakdowns as a result of shortages of spare parts and fuel. To most villagers, therefore, there was little tangible benefit to be obtained from the presence of an 341 Tabora Rural Integrated Development Project, Roads Component, Final Report, August 17, 1983. 351 Tanzania, supervision of roads component, Tabora Rural Development Proj- ect, Credit 703-TA, Back-to-Office Report, dated October 23, 1980. 36/ op. cit. - 27 - access road. Second, an earlier construction unit for access roads in the regionhad paid village labor- as was also the case of the constructionunit in the neighboringKigoma region. Third, the village accessprogramwas seen by villagersto be a -WorldBank' scheme and, as the evaluation reportpoints out: "In village logic it was impossible to correlate the manifest wealth of the aid agencies with the con- flicting requirement for a substantial input of their own voluntarylabor" 4.14 Althoughthe roads component has only been partially successfuland more costly than expected at appraisal,the mission'sattentionwas fre- quentlydrawn to the relative improvementto the roads in the region. There is evidencealso that considerable effort is being put into road maintenance despitethe fundingproblemswith projectclosure. 4.15 The land use planning component, also funded by the UK grant, has similarly not reached appraisal expectations, but it has been implementedin an innovative manner, concentrating on how to alleviatephysical conditions and constraintsimposed by the villagization program (para. 1.13). If the methodology developed is suitablefor large scale application, the experience gained under this component could be of considerable 37 /. value However, greaterattentionneeds to be given to trainingbecausemuch of the work to date has been undertaken by an expatriate team. 4.16 The team has, in close coordination with regional,district and village authorities,developed a methodology for planning satellite vil- lages. If the methodologydevelopedis acceptableto governmentit will be importantto ensure that requiredskillsbe in place at the villagelevel and that simple proceduresare establishedso that village planningteams can be trainedfairlyrapidly. 4.17 Funds were providedunder this projectfor an expatriateagronomist to 'pursuetobacco research at Tumbi and to train the nationalstaff". The SAR is not more specifici5/and it is largely due to the calibre of the agronomistselected that some of rhe more urgent constraintsfacing the tobaccoindustryhave been identified and successfullyovercome. 4.18 One such bottleneckwas the lack of good quality commercialseed. As a result of the agronomist'sefforts,Tanzania is now self sufficient in commercialseed productionand this has created a considerable foreign ex- change saving. 37/ ODA is continuingto fund this componentbeyond the clouingdate of the project. 38/ Annex 2, Page 9. - 28 - 4.19 A training program is now in place, and handouts for extension workers have been written and printed togetherwith a tobacco handbook for extension workers. Every effort should be made to ensure that the training program initiatedand supported by the agronomistis continued. 4.20 Implementation of the remainingcomponents,notably crop and live- stock production,water supplies and reafforestation, has been poor. In order to intensifyproductionof key crops in the region - particularly cotton,paddy, sorghum,and groundnuts - thirtyvillageswere to be selected to receive subsidizedinputs and extensionsupport for crop demonstrations and trials under field conditions. And although useful broad resultswere obtained concerningrecommendedfertilizer levels for the two main soil types, other aspects of crop and area specific technicalpackages for tar- geted productionincreaseswere not developed,largely due to (a) the late arrival of the internationally recruitedagronomist;(b) lack of adequate transportation for a programwhose successwas very much dependent on mobili- ty and farmer contact;(c) lack of producerincentive prices; (d) poor input, credit and credit recoverysystems;and (e) villagization, plus the related fact that farmerswere not free to make their own decisions aboutwhat crops to grow and how to grow them, while at the same time, they were expected to make contributions and provide labor towards block farms, communal farms, villagewoodlots,etc. 4.21 By the time of the first supervision mission in December1977, one month after project effectiveness,it was concluded that the technical packageswere not firm enough to go ahead with the recommended cultivation practicesand the use of inputs. Moreover,based on experiencein ecologi- cally similar areas, notably Geita and Kigoma, it appeared that use of the packageswould involvethe farmersin substantial losses. 4.22 Crop trialswere undertaken for two seasons,but the resultswere disappointing.This can be traced to administrative, technicaland institu- tional problems.39 / The government's decision in December 1977 to recruit village managers:U/from the ranks of agriculturalextension staff meant that there were fewer than expectedextensionstaff to supervisethe trials. Furthermore, certain villages selectedwere not eligibleunder the criteria established by the appraisal report. In some villages, particularlyin Nzega district,the trialswere viewed as 'WorldBank' trials of little interest either to farmers or to agricultural 4 1 / More important, staff. however,was the inappropriateness of the packagesand the poor returns. As a result, yields of all crops improvedonly slightly. 39/ The institutional problems are describedin detail in paras. 4.36 et seq. 40/ Chief executiveand technicalofficer of the village, responsiblefor plan preparationand implementation. Not to be confused with the villagechairman. 41/ Supervision reportdated July 24, 1978. - 29 - 4.23 Farmers were also constantlyencounteringdifficultiesin input supply. Institutionalarrangements for creditwere not clear from the outset and consequentlya smaller than expected number of villages participated in the trials. 4.24 The livestockcomponentsuffered both from poor design and poor managementduring implementation. The village ranch componentassumedthat a livestockowner would subsume individualintereststo that of the national interest by destockingand adopting grazing control, despite the fact that most livestock owners in the area are from a particular tribe - the Wasukuma - to whom livestockrearing is a major, if not the major income earner. The provision of dips and veterinary services was poorly administered. There was little coordination with other componentsof the project,with the resultthat some dips were constructed in areas where there was little or no water. Dips were often poorly designedand the throughput in number of animalshas not been sufficientto guaranteeprotection. 4.25 The water componentsuffered from shortagesof material and equip- ment, as well as the lack of a componentleader for the first threeyears of project implementation. Wells were situated in areas already served with water; boreholeswere constructed which now lack pumps; piped water schemes are incomplete; villagershave been unwillingor unable to dig the necessary trenches so the PVC piping has been left lying under trees and is now badly warped. 4.26 The reafforestationcomponentwas similarlyregarded with little enthusiasmby the villagers. Planting-outof the free seedlingstended to coincidewith labor requirements 2/ for other crops.L In addition,villagers were used to obtainingsuppliesof firewoodfrom the natural forests,or sup- plied by the tobacco authorityprovidingsubsidizedtransport. Once again, therefore,this componentmay have seemed necessaryfrom the national per- spectivebut its objectives were not similarlyviewed as prioritiesfrom the smallholder point of view. 4.27 The projectwas closed in June 1983; no follow-onproject is being financed. Actual projectcostswere US$17.8million,of which the IDA Credit comprisedUS$4.2 million. The CanadiangovernmentcancelledUS$2.1 million while UK increasedits grant to US$6.4 million. D. Issues (i) The Need for Consensusof OpinionRegardingProjectObjectives 4.28 The three-year gestationperiod in the formulationof this project is a reflectionof the considerabledifferencesof opinionamong the partici- p'nts as to the project'sunderlyingconcept. At project completionit is 42/ TRDB emphasisethat the number of simultaneousself help schemes was also a constraintto other project activitieswhich assumed a substan- tial degreeof villageself help. (TRDB commentspage 57C). - 30 - evident that the project continued to be viewed from different perspectives throughout implementation; this confusion over objectives has had an adverse effect on project outcome. 4.29 When this project was first identified, the Bank was finalizing the 1974 Agricultural and Rural Development Sector Report. Project preparation reflected the NADP strategy arising from this report with its emphasis on production (see para. 3.02). 4.30 At the same time, however, the government was anxious that support be given for the newly created regional authorities - particularly in respect of their responsibilities for preparing regional development plans. With the subsequent acceleration of the villagization program, the government was also keen that donor support be given for the provision of village infrastructure and supporting services (para. 1.12). Thus, by the time of the first appraisal mission, one year later, the focus of the project had changed, the previous emphasis on productive activities was replaced by concentrated sup- port for planning activities at the regional level. This change in project objectives was the subject of considerable discussion at the issues meeting where it was pointed out that the project as designed appeared to be contrary to the Bank's policy of only investing in productive sectors. 4.31 The debate within the Bank continued but the final project version concentrated on the regional infrastructure and planning efforts, while the productive elements, principally the crop and livestock components, became pilot programs and were given a secondary role. The various stages through which the project had gone, as reflected in the components which were con- sidered and dropped, are well illustrated in the attached table prepared for the project completion report (Table 1). 4.32 It is somewhat ironic to note that despite the importance latterly attached by the Bank in this first phase to supporting the regional authori- ty's planning and implementation capacity, when the decision was taken not to go ahead with the second phase of the project, an important rationale was the poor performance of the production components.4 3/ 4.33 This confusion as to the project's concept was evident in other aspects. The completion report, prepared by the Borrower, concluded that the lack of a pre-implementation phase was a major constraint to project perfor- mance (PCR, para. 3.4 (ix)). From the issues paper 4 4 /, however, it is evident that the Bank intended that the first phase should in fact be a pre- implementation program. 43/ Letter to the Government of Tanzania dated February 23, 1983. The ab- sence of suitable technical packages was not exclusive to the Tabora region, however. Similar difficulties have been experienced in Kigoma Region and Geita District. 44/ Issues Paper dated December 12, 1975. - 31 - 4.34 Furthermore, the 1978 CPP emphasized that the rural development projects should be seen as pilot projects and not as major investments of resource transfer. The completion report, however, draws attention to the fact that the project was largely viewed as a resource procurement project, particularly amongst regional staff not directly associated with the project (PCR,para. 6.1.1). 4.35 This confusion over project objectives has, to a certain degree, strained relations between the government and the Bank since the government still finds it difficult to understand why the project bhrs been closed with no second, productive, phase. (ii) Organization and Management 4.36 A major reason for the project not achieving any of its goals was poor nrganization and management. In particular, project design should have taken greater account of first, how the regional authority operates in prac- tice; and second, since the project was an integral part of the regional structure, the consequences for project management in the event of structural bottlenecks occurring within the regional administration. 4.37 The appraisal report failed to examine how decentralization would operate in practice, particularly how the changes in reporting responsib1li- ties for technical staff would affect project implementation. Instead, decentralization was regarded by the Bank in more theoretical terms 4 5 /: 'As decentralization takes root, the Regional and District Development Directors gain experience, and the farmers begin to see their views reflected in the district plans, any outside assistance is bound to be used more effectively than would be the case if the plans were formulated and the projects designed from the 'outside-....Many of the implemen- tation problems which plague rural development proj- ects are not likely to ar'se when there is a frame- work which embraces farmers, technical staff (functional managers), Development Directors and representatives of TANU." In fact there was no mechanism in place which would allow the views of the farmers to be reflected in district plans. Neither was there a framework which would embrace farmers, technical staff and party officials alike (para. 1.11). 45/ Abraham, P. and Robinson, F. "Rural Development in Tanzania: A Review of Ujamaa.- Studies on Employment and Rural Development No. 14, September 1974. - 32 - 4.38 Decentralization of government provided for communication between the national and the regional level, as well as within the regional authority Itself. The structure, however, created no linkages between the Bwana Shamba - the village level extension worker - and his agricultural superior offi- cers. This is one of the major reasons why the crop trials program, for example, was a fallure. The component was designed by project officials working in isolation from the farmers. It covered crops which the farmers, because of the constraints on individual decision making (para. 1.15), were not interested in growing - yet there was no mechanism whereby this lack of interest or the reasons for the lack of interest could be transmitted back up the organizational structure to the officers responsible for the program. 4.39 The need to develop such linkages and technical information flows to the village level was recognized early on by project management. Since the importance of such information flows was not similarly appreciated by the regional administration this has resulted in there being two categories of 4 6 / who are staff in the regional administration, there are now 7RIDEP staff- distinct from 'nonr-RIDEP staff' as the PCR makes clear: 4 7 / -The principal and overriding constraint to imple- mentation mentioned by project management was a general lack of understanding and appreciation of the objectives and problms of the RIDEP on the part of the nonr-RIDEP staff who were involved in this project....it was felt, and project management still believe, that the only way to have a long-term impact on rural development is through the vital links with the farmer, i.e., the extension staff of the Agriculture, Livestock and Natural Resources Ministries. Thus unless the project is closely associated with those staff, there would be little chance of achieving the project's overall goals. Unfortunately, non-RIDEP staff were generally slow to realize the importance of these links- (PCR, para. 6.1.]). In effect, given the way in which the regional structure operates, the proj- ect could not accomplish both institution building at the regional level and 46/ RIDEP = Rural Integrated Development Project. 47/ Bank supervision also commented on the fact that the project was not sufficiently integrated into the regional or district programs (Super- vision Report, July 1981). - 33 - experimental productive activities at the district/village level.48I The first objective implies general acceptance of the structure as is, while the second requires either a change in its technical reporting system or direct project intervention at the district or village level.49! (iii) Lessons Learned 4.40 Precisely how the problems illustrated by the project could have been avoided, in practical terms, remains debatable, however several lessons emerge (PCR, para. 12). Amongst them: (a) A prerequisite for success is mutual understanding and agreement on project objectives between government, beneficiaries and the funding agency; (b) clearer mechanisms need to be found to deal with project redesign, in situations where either critical assump- tions change, or the original project design is faulty. 48/ The project coordinator, for example, refers to the fact that although the project's component leaders were expected to report to the region's functional managers, they tended to bypass the system and deal directly with the project coordinator. 'Tanzania: The Tabora Rural Integrated Project,- E. Mbuya, Project Coordinator, reproduced in Selected Case Studies on Monitoring and Evaluation of Rural Development Projects, IBRD 1980. 49/ This has now been recognized by the Bank in that supervision of the Mwanza/Shinyanga Rural Development Project is conceatrating much more on developing expertise at the village level. -34 - TABLE 1 TABORA RURAL DEVELOPMENT PROJ=CT (CREDIT 703-TA) Components Tentatively IdeatIfiedla ad Eventually Included Preperdaton June September November February 1974 1974 1974 1975/b Appraisal Component Activity 1. Crops Maize X X X X Ollseeds X X X X X Cotton X X X X Sorghum X X x Rice X X X X X Tobacco X X X Xld Input Credit x X x Village Stores X X X Agricultural Etension X X X Adaptive Researcb X X X Staff & Farmer Training X X 2. Land Use PlanninglSoil Conservation X X X Xle 3. Livestock Beef Cattle X X Xle Dairy r x Tsetse Control X X Cattle Buying Centers X X X Ranch Development X - X X Dips X X Veterinary Centers X X A. Been Honey and Beeswra S 5. Roado X X X X X 6. Water Supply X X X X X 7. Rural Health X S. Rural Industries /c Furniture Making X X /c Vehicle Maintenance & Training X X /c Crop Proceseing X X __c Farm Implements X X (Studies) 9. Irrigation X 10. Other Advisory Unlt X X X X Trainlng X X Project Evaluation Xle Preparation of Third Project X X 11. Reafforeatation % /a As mentioned in Back-to-Office reports and Project Implementation Briefs. lb Preparatlon Report/Loan Request. Te- A URDP Rural Industries Specialist was being consulted on activities that would justify funding. Id The appraisal mission's proposal for funding included a tobacco pro- cesslug component. It was subsequently decided to make thIs a separate project. Ie InItially, the appraisal mission vas working with a proposal which com blned project implemntation, land uze planiDng and project evaluation funtlons in a specalist team under the Regional Planning officer. - 35 - V. TOBACCOPROCESSING PROJECT (CREDIT 658-TA) A. Introduction 5.01 Credit 658-TA was the second Bank-supported project for tobacco production in Tanzania. Approved in August 1976, this credit provided US$8 million towards a total project cost of US$11.3 million to increase tobacco processing capacity in Morogoro from 18 million kg to 41 million kg of cured leaf per annum. This objective has been achieved; however, because tobacco production fell instead of increasing as had been projected at appraisal, the processing and storage capacity is now greatly underutilized. Tobacco pro- duction fell from its peak of about 16 million kg in 1973/74 to less than 10 million kg in 1982/83. As a result the project has a negative rate of return instead of the 25% estimated at appraisal. B. Project Concept, Preparation and Appraisal 5.02 For the first project (Credit 217-TA) an IDA Credit of US$9 million was provided in 1970 to increase flue-cured tobacco production in the Tabora, Rukwa and Mbeya regions by settling 15,000 new tobacco growers in villages of 100 families each. At first, farmers were reluctant to move to these new areas-possibly because of delays in providing the necessary infrastructure, particularly water, or possibly because of the requirement that farmers living in villages cultivate their land communally. With the start of the compulsory villagization program in 1974, the pace of settlement increased rapidly; the project's infrastructure facilities were an obvious attraction to farmers who were being forced to move into villages. Tobacco production continued to be much lower in the project areas than anticipated at apprais- al,5 0 1/ mostly because many of these new settlers were not initially interested in growing tobacco. 5 1/ 5.03 Despite the poor indicators which were evident in the course of this first project, planning for the processing project assumed that national tobacco production would continue to expand rapidly, maintaining trends that had been evident in the sixties. 5.04 This processing project had the following components: (a) rehabilitation of an existing, and provision of a new, processing line in the factory at Morogoro; 50/ A June 1975 supervision mission reported production at only 431 tons against the 4,000 tons estimated at appraisal. 511 Another factor which could not be foreseen by the first project appraisal was the change in farmgate prices. The tobacco/maize price ratio changed from 15:1 in 1972/73 to 9:1 in 1975/76. - 36 - (b) construction of tobacco storage space at Dar es Salaam and Morogoro; (c) infrastructure support in Morogoro; (d) construction of up-country stores; (e) technical assistance and training; and (f) a study of future tobacco processing requirements. C. Project Implementation 5.05 Project implementationstarted slowly due to delays in project effectiveness, procurementirregularities, insufficientsupervisionof some procurementunder retroactivefinancing,and cost overruns. To reduce cost overruns,some buildingswere eliminatedprior to credit approval. However, it was later found necessary to construct additional storage space at Morogoro which was not provided for in the project as approved. This was accommodated throughan amendmentto the CreditAgreementand reallocationof funds. 5.06 It was expected that upgrading of the existing processingline would be undertakenfirst, with the new expanded facilitiesinstalledthere- after. The appraisal recommendationsto implement the project in two phases-rehabilitation and expansion-werenot followed. Phase I was delayed by more than a year, and, to raduce cost overrunsarising from inflation,it was decided to carry out Phase II simultaneously with Phase I. If the phasing anticipatedin the SAR had been adhered to perhaps the declining trends in productionwould have been better emphasized.5Z/ Considering, however, that productionhad been stagnating three years prior to project appraisal,and decliningin the year between appraisaland negotiation, the importance of productionto this projectmay still not have been noted. 5.07 With the declinein total tobacco production, the projectfailed to generate foreign exchange earnings. The project also contributedto the deterioratioa of both the tobaccoauthority's and its processingsubsidiary, TTPC's financialsituation. With the low levels of capacity utilization, TTPC was able to cover only operatingcosts out of processing fees; there was no additional revenue to cover either interest or principal repayments as stipulated in the Credit Agreement. As a result, the project has become a net drain to the Tanzanian economy with government having to provide bud- getary allocationsfor the shortfall. 52/ The consultants' report, on which project preparation was based, stressedthat the expansionin processingfacilitiesas envisagedunder which comprisedPhase I-should Phase II-as against the rehabilitation be carefully evaluated in the light of actual trends in tobacco production. - 37 - D. Issues (i) Importance of Adequate Project Preparation and Appraisal 5.08 This project was prepared and appraised so rapidly that the funda- mental justification for the project-increased tobacco production-was not analyzed in sufficient detail. The audit believes that while the extent of the decline in production that did occur could not bave been foreseen at appraisal, there were sufficient warning signals-some raised within the Bank-which should have resulted in the production issue being given greater attention than it received. In retrospect, it appears that a perceived need to get the project approved quickly became the overriding concern. 5.09 Factors adversely influencing tobacco production were already evi- dent at project preparation 5 3 / but production was assumed to be increasing in a steady upward fashion. At least three widely differing forecasts were available and sensitivity analyses showed that the estimated rate of return of 25X would decline to 19% with a 10% drop in production. Nevertheless, a "conservative' annual growth rate of 7-1/2% was forecast. 54 / 5.10 The initial decision was taken by the Resident Mission in East Africa in July 1975 that more tobacco processing facilities would be needed by the time of the 1976/77 season. The proposal was strongly supported by the Tanzanian government as complementing its agro-processing strategy. RMEA considered that, even discounting the effects of villagization, crop produc- tion would have increased from 16 million kg in 73/74 season to 22 million kg by mid-1977 (TAT's forecast being 26 million kg). It was concluded that normal project preparation and appraisal would not allow for the expanded processing facilities to be in place by the 76/77 season. On the strength of a tentative suggestion that tobacco processing be incorporated into the pro- posed Tabora Rural Development project, then being prepared, RMEA hired consultants to review the existing documents in August 1975, and to prepare specific proposals in September/October, prior to appraisal in October. The project was then added to the lending program. 5.11 There was less enthusiasm in Bank headquarters about the project's prospects. If the Bank was to continue to support tobacco production this should be done through the NADP (para. 3.02). Moreover, headquarters staff viewed with concern the progress of the Flue-Cured Tobacco project being supervised from Nairobi. Headquarters did not agree with the supervision mission's proposal to upgrade the project from its previous problem project status.55/ 53/ PPAR, Flue-Cured Tobacco Project (Credit 217-TA), Report No. 2786, dated January 4, 1980. 54/ President's Report, No. P-1902-TA, dated August 20, 1976, para. 42. 551 Supervision of the Flue-cured Tobacco project: memo dated November 26, 1975. - 38 - 5.12 Similarly,in a review of the investmentproposal for the Tabora Rural Developmentproject, concern was expressedabout the proposed tobacco processing component,56/ particularly since these very poor production trends were emerging at the same time that the number of settlers in the scheme was increasingconsiderably: to 8,600 in February 1975 vs. 2,800 in February1974. Average acreage per farmer had fallen;by 1975 the ratio of maize to tobaccowas 5:1 rather than the 1:1 envisagedat appraisal. 5.13 Meanwhile,the consultantsin charge of preparationanalyzed crop production trends which had been providedby the tobaccoauthority. The pre- parationteam concludedthat these crop forecastswere optimistic and reduced the 1979/80 projectionsfrom 37.5 million kg to 30 million kg. 5 7 / Another observer of the tobacco industry stated that while the phased approachwas sensible, ... 30 million kg by 1980 representsexpansionwithout paralleland not possibleto achieve if a quality crop is the target.' 5.14 The consultant'spreparationreport was not received until a few days prior to the departureof the appraisalmission for the proposed Tabora Rural Development project, thus no issues meeting was held for tobacco processing. Having linked the appraisal of the tobacco processing to the Tabora Rural Development project for the purpose of speed, it was then separatedout for the same reason since the AppraisalReport was likely to be delayed longer than the two months it would take to complete one for a separatetobaccoproject. 5.15 Bank staff continuedto express doubt about productiontrends, on which the financialviabilityof the project rested. One review, 58 / in par- ticular,drew; attentionto the probable unreliability of productiondata for 1973/74, given that this was a year of major upheaval in Tanzania with the start of the villagizationprogram. The review pointedout that the validity of appraisalprojectionswas dependenton this 1973/74data being reliable. Aggregate productiontrendsin Tanzaniaare inevitably affectedby any change in smallholder circumstances,given the structureof agriculturaloperations, but the difficultiesfacing a smallholder growing tobaccowere not explored (para.6.29 et. seq.) (ii) Project Supervision 5.16 The consultants'preparation report drew attentionto the fact that the viabilityof the project was dependent on productionincreasingat the levels anticipated. Despite this, the two-stagephase approach was dropped and supervision missions, as would be the case in the subsequent Tobacco 56/ Memo dated September26, 1975. 57/ Report on the Proposalsfor IncreasingProcessingand Storage Capacity at the TanganyikaTobaccoProcessingCo., October1975. 58/ Tanzania Tobacco Processing Project: Yellow Cover Appraisal Report, Memo dated May 21, 1976. - 39 - Handling project,comprisedfinancial analysts and economists whose main con- cern was to ensure timely procurement of machineryand equipmentor that the buildings were constructedas planned. Equal attention should have been given to production. There were no agriculturalists on any supervision mis- sion until December 1980, when the Tobacco Handling project was also identi- fied as facing severe problemsin this regard. 5.17 Once it was finally recognized that productionwas integral to project success, there was in fact very little that the Bank could do given its existingsupervision resources. The Bank, in its supervisionmission of December 1980, comprehensively reviewed the constraintsfacing the industry outlining possible remedies which could be taken, without long delays awaiting politicaldecision making. However, productioncontinuedto stag- nate and decline. A major reason for the decline can be attributed to the institutionalweakness of the tobacco authority. As has been pointedout by a supervisionmission for the TanzaniaTea Project: 'Supervisionconcentratedlargely on TTA's institu- tionalshortcomings and possible solutions....We have been unable to effect improvementsdespite intensive action. The real cause for failure must be sought in Tanzanianeconomicmanagement, particularlyparastatal management.... Our view is that for the Bank to have done more for TTA, the Bank shouldhave put its entire lending program on the line for across the board parastatalreform. The average 14 supervisionweeks per year we have to superviseprojects could not in any circumstanceshave changed TTA's mismanagement withoutgreater Bank leverageto implementrecommenda- tionswhich have been consistently rejectedby Govern- ment.--59/ 5.18 The extent and the interrelation of the problems facing the crop authority are well illustratedin another study6 0/ and depicted in diagram form: 591 Operational Policy Review: World Bank Experience of Institutional developmentin Africa. Memo dated December8, 1983. 60/ J. Moris, "WhatWent Wrong in the 1970s",BackgroundPaper VI. - 40 - Peasants reversion < High unit charges < to subsistencecrops for crop handling Reduction in export - Reduced parastatal crop production throughput Reduced foreign - Reduced equipment - > Increased down- exchange spares time & breakdowns Where the link between the cash crop authorityand farmers is such that any action taken by the authorityfor short-termsurvival will only furtherde- press crop productiontherebyhasteningits eventualcollapse. 5.19 In conclusion,much greater care is needed during preparationand appraisalto ensure that appraisalproceduresare not short-circuited, parti- cularly when it is evident that changes of the magnitude required in this instance cannot be accomplished during supervision. There have been other projects where compressingappraisalprocedureswas thought to be justified by an urgently perceivedneed 6 1 / but at project completion,evidence conr- firmed that this has been detrimental to successfulproject outcome. VI. TOBACCOHANDLINGPROJECT (CREDIT802-TA) A. Introduction 6.01 Prior to 1978 when Credit 802-TA was approved, the Bank had pro- vided more than US$23 million in support of tobacco productionin Tanzania. Its first project, Credit 217-TA signed in 1970, concentrated on area increase of tobacco production by creating smallholdersettlement schemes similar in design to those which had been successfulin the sixties. The Bank's second project, Credit 658-TA, (ChapterV), focused on expandingthe processingfacilitiesfor flue-cured tobacco. Funds were also providedunder the Tabora Rural Development project for tobacco research and for pilot fuelwoodplantationschemes. 61/ The Drought Relief Project: A Review of Performance,OED Report No. 3173, dated October 28, 1980, and Sudan: Southern Region Agricultural Rehabilitation Project (Credit 476-SU),OED Report No. 3476, dated May 26, 1981. - 41 - 6.02 This third credit provideda furtherUS$14 million to the industry, with the emphasis on net yield increases. Research indicated that small- holders were able to produce a field crop similar to that obtainedby larger farmers, but post-harvestlosses affected their final yields by as much as 40%.62/ The project, therefore, was designed to support the introduction of new handling techniques together with infrastructural facilitiesto reduce losses experienced by smallholder tobaccoproducers. 6.03 Tobacco yields have not increased,in fact tobacco productionin 1981/82has fallen back to the level obtained in 1970, when the Bank's first tobacco project became effective. Credit 802-TA has a negative rate of return, in part due to the difficultiesexperienced in the agricultural sector during the period (Chapter I) and in part due to the tobacco authority's inability to discharge its many duties and responsibilities effectively. B. and Appraisal Project Concept,Preparation 6.04 Tobacco production in Tanzania has traditionally comprisedflue- cured tobacco cultivated by large-scaleexpatriate farmers in the Iringa region. Fire-curedtobaccowas introducedin 19b1,but undertakenon a small scale in Songea region, mainly for domestic consumption. Smallholder productionof flue cured tobacco began to expand rapidly at about this time when tobacco ceased to be regarded as a plantationcrop. Previouslysmall farmers had only been able to grow green leaf which then had to be sold before processing. With the irntroduction of settlementschemes in Tabora to process their own region where small farmerswere able, under supervision, tobacco, the economic incentiveswere such that productionexpanded rapidly with large numbers of farmers taking up the crop. By 1965, productionof flue-curedtobacco exceeded demand in East Africa and Tanzania entered the world market. 6.05 By 1970, when Credit 217-TA was signed, output had increasedfive fold to 8.8 million kg per annum with smallholders in the Tabora region pro- ducing almost 65% of the flue-curedtobacco, growers in the Iringa region accountingfor about 27% of production. Thus the first project was intended primarily to increase productionof tobacco by continuing the established trend, i.e., encouragingmore growers to take up the crop. Credit 217-TA succeededin increasingmaize production, which was not an objectiveof the project, except insofar as maize is generally regardedas a rotation crop. Incrementaltobacco productionfell far short of appraisalestimates: 1.5 63 / million kg vs. 8.9 million kg per annum. 62/ Audit has been unable to confirmthe validityof this statement,in par- ticular,when or by whom the researchwas conducted. The statementmay be no more than a perceptionby agriculturalstaff. 63/ PPAR, Flue-CuredTobacco Project (Credit217-TA),Report No. 2786, dated January 4, 1980. - 42 - 6.06 The second project, Credit 658-TA, was approved by the Bank in September 1976, concentrating entirely on the expansion of existing processingfacilities. It was not until this third Bank-supported project, Credit 802-TA, signed in 1978, that attentionwas given to increasingthe yields of smallholders. 6.07 The project provided for (a) sturdy polythene bags to enable farmers to store their cured leaves, with protectionfrom rain and damp until sufficientleaves are obtained to allow grading to take place; (b) grading and baling centerswhere tobaccocan be preparedfor the market in dry condi- tions without fear of mold; (c) marketing centers, godowns, transportfor distribution of inputs and collectionof leaf; and (d) some technicalassis- tance to the tobaccoauthority. C. ProjectImplementation 6.0B The two projectcomponentsexpectedto generatethe highestreturns were smallholder bulking bags (ERR = 320%) and village grading/baling centers (ERR - 6BZ). Appraisal envisaged TRDB providing credit for 350,000 bulking bags and 80 grading/baling centers, with the tobacco authority having day- to-day responsibilityfor the whole project. The Credit Agreementprovided, however,for TRDB to be the implementing agency for these two components. By the original project completiondate (April 1983) TRDB had not sold any bulking bags to farmers at commercialrates (a few were distributedon a trial basis). The tobacco authorityhad purchased 43,000bulking bags for distribution,but this has been slow due to farmer reluctance to accept the bags on credit. Almost 300 villages were eligible--in terms of volume of production--fora grading/balingcenter, but many were precluded from receivingfurther credit due to loan arrears. Not one of the 80 proposed grading'balingcentershas been built.64 / 6.09 Of .he 40 marketingcenters envisagedat appraisal,nine were com- pleted but at least four of thesewere constructed in areas where production was below the minimum acceptablethroughputlevel of 300,000 kg/season. In one instance tobacco productionwas no more than 34,000 kg, while the marketingcenterat Aputa, which was to cover four villageswith a total pro- duction of 652,000kg, was only partiallyconstructed. 6.10 une of the two storage facilitiesproposed at appraisalwas com- pleted, but no tobacco was stored in the 1982/83 season, the entire crop being immediatelyrailed to Dar-es-Salaam. 6.11 A:l vehicles envisagedunder the appraisal report were procured, while the technicalassistancecomponentwas implemented with minor modifica- tions. A financialsystem study was satisfactorilycompleted and implemented in an effectiveand cost-savingfashion. The project also financed the preparationof a Tobacco Rehabilitation Industry study which has been well 64/ TRDB also point out that confusionas to the ownershipof these centers and their suitablelocation,as well as shortagesof buildingmaterials were added factors. (TRDBcommentspage 57D). - 43 - receivedand is expectedto providea frameworkfor much-neededindustry-wide changes if productionis to be restoredto formerlevels. 6.12 The PCR points out that although the first (and subsequent)Bank supervisionmission expressed concern at the stagnation and decline in tobacco production, neither the Bank nor the tobacco authority managed to focus properly on the question: what policy changes would be necessaryto restore the industry to its appraisedgrowth path? Factors contributing to stagnation were correctlyidentified as inadequateavailabilityand distribu- tion of inputs, lack of good seed material and site-specific extension/ agronomicrecommendations, delays (due to transportshortagesand the Ugandan War) in crop pick-up, shortage of fuelwood, and the sharp increase in the ratio of unofficialmaize price to official tobaccoprice. 6.13 However, the lack of effective action to remedy these problems together with TAT's and TRDB's difficulty in managing the most productive project components,has meant that the project has had little impact; the crop authority'stransportfleet has been replaced,and its availablegodown capacityincreased. 6.14 The project was closed six months after the original closing date. Omly US$6.1 millionwas actuallydisbursedout of the total IDA Credit of US$14.1million, the remainderbeing cancelled. D. Issues Attentionto Production (i) Insufficient 6.15 The project originatedat a time when the Bank was giving con- siderableattentionto the need to stress agriculturalproductioncapabili- ties in Tanzania, with the National AgriculturalDevelopmentProject (see para. 3.02). 6.16 The most directly productivecomponentsof the project were the bulkingbags and the grading/baling centers. Togetherthese totalledUS$2.5 million out of the total project cost of US$20.5 million. The remainderof project funds were for physical infrastructure. Technicalassistanceunder the projectreceivedUS$1.7 million. When one tenth of total projectcost is directly devoted to productionthere is likely to be considerablepressure during implementationto concentrate on the larger,capital intensivecompo- nents, e.g. the vehicle fleet (US$3.9 million), rather than the smaller bulking bags component(US$0.3million). 6.17 Project design anticipatedthat these bulkingbags--whichhad been well received by small farmers in Malawi and Zambia--would be obtained by farmerson credit. Smallholders are quick to change their farming practices if the demonstrated benefitsare sufficiently attractive. There is no indi- cation that analysis of the qualityof the extensionservice was undertaken at projectdesign or that close supervision was earmarkedfor this component in the early stages of project implementation. A study commissioned by the - 44 - Bank 6 5 ! in 1974 concluded that yields in Urambo and Tumbi districts were strongly influenced by the quality of the extension service, a lesson which should have been noted in this project with its principal objective similarly to increase yields of smaliholders. 6 6 / The focus of institution-building efforts for the tobacco authority instead concentrated on its internal supporting services: accountants and a procurement officer together with a building officer and transport personnel. 6.18 The focus of the supervision effort has tended towards the achieve- ment of the physical infrastructural objectives of the project (see para. 5.16 et. seq.). Considerable correspondence on project files is devoted to the quantitative aspects-the number of vehicles procured for the extension service, for example-without discussion as to how effectively these facili- ties are being used by the extension staff concerned. Similarly, the project procured 80 tractors to assist in transporting inputs and fuelwood, but the monitoring unit in the Ministry of Agriculture (PPMB) reports at the end of the project that the whereabouts of only 25 is known-aIl 25 being located in the least important tobacco growing areas and with no evidence of any record- keeping in order to ensure that user charges are complied with.67/ 6.19 Of greater concern to the audit, however, is that the project was to introduce a simple technique-the bulking bags-which despite its rela- tively small size in value terms, may have a noticeable effect on smallholder yields if experience in neighboring countries is accepted. The use of these bags still remains to be tested in Tanzania, because few of the 350,000 bags to be provided under the project have been procured or distributed to farmers. It is not the purpose of this audit to determine where responsi- bility for this notable omission lies, but it is regrettable that, despite a total project cost of almost US$7 million, one cannot say whether the use of such bags will have an effect on tobacco yields or even whether they are acceptable to the farmers. (ii) Institutional Aspects 6.20 The management weaknesses in the tobacco industry essentially cover two particular aspects. First, the tobacco authority has multiple objec- tives, commercial and non-commercial, but the degree of accountability required by its principals has been vague and unclear. Second, policy and regulatory constraints within the authority's operating environment, over which it has little control or influence. 65/ Africa Rural Development Study. Background Paper: -Smallholder Tobacco Production in Tanzania: A Review of Urambo and Tumbi Schemes", M. Agarwal and D. Linsenmeyer, 1974. 661 Nor has this situation changed. Many of the recommendations of the Tobacco Industry Rehabilitation study concern the need to improve the quality of TAT's extension service. 671 "Report on visits to the TAT Tobacco Growing Areas" PPMB, Ministry of Agriculture, December 1983. - 45 - 6.21 Its range of functional activities is extensive. As the super- vision mission of January 1982 pointed out: the concept of TAT is unique in the tobaccoworld; nowhere else will you find a single organization that controls the tobacco crop from planting right through the processing,marketingand export stages. It also involves warehouses and a massive transportation system that not only transports tobacco to and from the processing plant, but also transports inputs, includingfuelvood.... 6.22 Until 1976, responsibility for credit, transport and baling services restedwith the cooperatives (para. 1.17). Farmers graded their own leaves under supervisionof extensionstaff and the cooperative deducted the cost of individualfarmers' inputs from the marketed proceeds. With the decision to abolish the cooperativesin May 1976, all these functionshad to be absorbedimmediately by TAT. 6.23 In addition to this range of functionalactivities,the authority has non-commercial objectives. TAT was establiPhed in the same year that the government was decentralized and thus obliged to competefor staff along with the simultaneous demand of twentynewly created regionaladministrations, the remaining100-200 parastatalsand the central government ministries. Since salary scales are more or less standardized, parastatalssuch as the tobacco authorityhave had to provide an array of ancillaryservices to attract and retain staff, includingdispensaries, shopb, schools,etc. 6.24 Under the provisionsof the Tobacco IndustryAct, the authorityis responsible to a number of principals,but relations in practice between governmentand the parastatalhas been vague, with the result that the dis- tinctionbetween accountabilityand autonomy has been poorly defined. TAT's parent ministry is Agriculture but control exercised has been nomina', exacerbated by the fact that the Ministry has seldom had informationwith which to measure management performance. The need for some form of perfor- mance evaluationis particularly importantin Tanzaniawhere market cignals, particularly long term profits and growth, are not availableto providebasic indicatorsof performance. To measure management performance, therefore, requires regularand reliable informationon costs, outputs, etc. thus a key requirement is accountingdata. 6.25 The Bank has been taking important first steps in this regard at both the parastataland the sectorallevel. Considerable attentionhas been given to strengthening TAT's financialsystem. At the same time, at the sectoral level, where accountability resides, the Bank has been providing support under the auspicesof the export rehabilitation credit. This unit is located in the planning division of the Ministry of Agricultureand is now startingto monitor the flow of information from this, as well as from other, parastatals. PPMB will be able to determine to what extent performanceof the authorityis a result of internalmanagementactions and to what extent the responsibility of factorsbeyond its control. - 46 - 6.26 There is no doubt that the Authority's performance has been adversely affected by policy and regulatory issues. Foremost amongst these is pricing.68! The parastatal has incurred heavy administrative costs, almost 11% of sales, but an examination of the trading results for 1980/81 shows that TAT is making a gross loss even at the trading stage, i.e., before any fixed overhead costs are added. This is due primarily to inadequate margins as well as to excessive transportation and processing costs.6 9 / 6.27 Furthermore, the tobacco authority was made responsible for credit collection on behalf of TRDB, yet it is the villages which draw up the neces- sary documentation of who has received credit and how much is to be repaid (see para. 1.14). Where the list prepared by the villages did not coincide with the list held by the credit agency, TAT was required to reconcile one with the other. 6.28 It is impossible, however, to determine how important these factors have been in TAT's poor performance overall, which underlines the need for effective management performance evaluation. Experience obtained illustrates the importance of intervention simultaneously at both the parastatal level and the sector level, i.e., PPMB, where control is to be exercised. (iii) Lack of Focus on Requirements to be a Tobacco Farmer in Tanzania 6.29 There is little evidence to indicate that there has been any systematic focus on the individual grower either at project design or imple- mentation stage. Tobacco farming requires a series of well ordered and well timed husbandry practices to produce high quality leaf. If the farmer is dependent on outside agencies or institutions for the supply of the necessary inputs then this in turn demands efficient institutional performance. 6.30 Despite the institutional changes which had occurred in Tanzania prior to the appraisal of this project, there is little evideace of any analysis of how these changes would affect a smallholder cultivating tobacco. Farmers cannot purchase any inputs for tobacco with cash. With the introduction of villagization (para. 1.12), input supply and credit admini- stratlon would almost inevitably encounter problems particularly when the cooperatives were abolished in 1976. In the early years, many villages were poorly managed as compared to the cooperatives, yet these villages were required not only to compile lists of all necessary inputs, but also to be responsible for loan recovery. In principle, the responsibility for col- lecting seasonal input loans is done at source by the crop authorities, if there is a monopoly in procurement as is the case of tobacco. In practice, 68/ Following the abolition of the cooperatives, producer prices for almost all crops are determined centrally by a cabinet committee and prevail nationally regardless of geographical location. 69/ -Performance of Agricultural Sector Parastatals- Appendix C, p.3. Background Paper, Agricultural Sector Study, 1982. - 47 - however, the tobacco authority has not always passed on the loan proceeds to TRDB with the result that the tobacco farmer can find himself with no season- al inputs either because village records have been inadequate and the credit agency has not received what it considered its due, or because the village has paid the correct amount, but the funds have been diverted by the TAT. 6.31 Furthermore, difficulties were frequently encountered in obtaining the fertilizer in time. Transportation from the Tanzania Fertilizer Company (TFC) to the regions has often been slow. Trucks are frequently unavailable to transport fertilizer from TFC godowns to rail heads or to the villages. For flue-cured tobacco, where there is only 40 days between planting of the seedlings and harvesting the first leaves, if application of fertilizer is delayed, so will be the necessary weeding, the removal of the flower heads and suckers, with subsequent effects on yield. The critical factor in tobacco growing is timeliness of operations. Concurrent with villagization came political pressure for equity within villages; thus the practice of tobacco farmers employing labor fell into disrepute. This resulted in skilled growers reducing their area considerably (from about 2 ha minimum, to about 0.5 ha maximum 7 0 /) because they had to rely on family labor only. The area given up was then taken over by new unskilled growers, operating with little extension advice. 6.32 In Tanzania the amount of fuelwood needed for the curing process is substantially higher than many other countries. A ready source of fuelwood is, therefore, an important factor in tobacco growing. With the introduction of villagization, farmers can no longer cultivate tobacco on the edge of forests, and are having to transport fuelwood over longer distances as sup, plies near the villages become depleted, Crop rotation, a necessity given the prevalence of disease, is also being affected. A final constraint is that TAT sometimes fails to pick up the tobacco or if it does, fails to pay the farmers cash. 6.33 As tobacco growing in Tanzania involves, in addition to the farmer, the village administrator, TAT, TRDB, and TFC, more attention should have been given to their Institutional capacity including how their operations directly and indirectly affect the smallholder tobacco grower. VII. CASHEWNUT DEVELOPMENT PROJECT (LOAN 1014-TA) A. Introduccion 7.01 Loan 1014-TA provided US$21.0 million in support of a US$30.3 million project aimed at increasing cashewnut processing capacity and raw nut 70/ The findings of the first project, Credit 217-TA, indicated an average of 0.4 ha per farmer rather than the 0.8 ha as expected at appraisal. - 48 - production by smallholders. Existing processing capacity of 27,000 tons of raw nuts would increase by 40,000 tons and the output of cashewnut nut shell liquid (CNSL) of about 1,200 by 2,500 tons. It was expected that the project would help increase total output of raw cashewauts in Tanzania by 19,000 tons at full project development in 1983. A small value increase was also expected due to an improvement in quality of raw nuts, as a result of exten- sion and grading activities. Decline in raw nut production, from 125,000 tons in the 73/74 season to about 41,000 tons in the 79/80 season, has caused the economic rate of return to be negative. 7.02 Despite the declining production trends throughout the course of this project, the Bank approved a second phase, follow-on project which pro- vided a further US$27.5 million for the construction of three additional processing factories, thereby expanding Tanzania's processing capacity by an additional 30,000 tons per annum. On the possible optimistic assumption that cashew production in the next 10 to 15 years is likely to stagnate at around 50,000 tons per annum, there will still be at least 50,000 tons of excess processing capacity, which makes the decision to support the Phase II project all the more surprising. B. Project Origin, Preparation and Appraisal 7.03 At the time the project was identified, cashew was one of Tanzania's principal foreign exchange earners. Production of raw nuts, which is almost entirely a smallholder crop, increased from 74,000 tons in 1965/66 to about 125,000 tons in 1972/73. About 10% to 15% of the purchased nuts were utilized by local processors, with the balance exported as raw nuts, mainly to India. The price differential between raw nuts and processed nuts was significant and the government was, therefore, interested in expanding processing capacity. It aimed to develop the cashewnut industry through a parastatal organization, the Cashewnut Authority of Tanzania, (CATA) esta- blished in July 1973. 7.04 The government was assisted by RMEA in the preparation of the project proposal, appraisal took place in October 1973, and the loan agree- ment signed in June 1974. 7.05 The project comprised the following compotents: establishment of five cashewnut processing factories and port storage facilities for cashewaut shell liquid; the establishment of a cashew extension and grading service; the improvement of cashewnut research facilities; construction of office and staff accommodation for CATA headquarters; construction of improved water supplies and education facilities in selected villages within the cashewnut producing areas; and provision of technical services, including future proj- ect preparation. The project was to be implemented by CATA under the general direction of the Ministry of Agriculture. C. Project Implementation 7.06 Overall physical execution of the project, particularly with regard to construction and equipping of factories and other civil works, has been - 49 - satisfactory. Technicalassistanceand trainingaspects of the projectwere only partially implemented, while key staff recruitment remained a problem throughout the implementation period. The research program, as proposed by the project, was not implemented. The extension service has been ineffective as far as raw nut production is concerned, although the quality of raw nuts has improved. The establishment of road maintenance units, community educa- tion centers,and improvement of rural water supplies, were dropped from the project in the early stages of Implementation as a result of project cost overruns. 7.07 The construction of buildings and installation of equipment for the five factories were completed satisfactorily by 1980, although about two years behind the estimatedcompletion date. However,the completedfactories have either had to be temporariiy closed, operatedsporadically, or operated very much below their processingcapacity. Factory problems have resulted from inadequate input supplies,decrease in demand for CNSL and, in the case of one factory,lack of water supply. 7.08 The most importantfactor affectingthe operationof the factories has been the decline in raw nut production. While productiondeclined from 125,000 tons in the 1973 season to 41,000 tons in the 1979/80 season, processing capacityincreasedfrom 27,000 tons in 1973 to 113,000tons (rated capacity)in 1980, as a result of the Phase I and Phase II projects71 / and the constructionof two additional factories with Italian financing. By December 1982, only five of a total of twelve factorieswere in operation. Of the project factories,only one was in operation. 7.09 The results of the limitedoperationsof the Phase I factorieshave been far below those of factoriesbuilt under the Phase II credit, negotiated in 1978. The superior performance of the Phase II factoriescan largely be attributed to the continuedpresenceof the Japanesemanufacturer's engineers under the agreed contractual arrangements. In retrospect, a similar arrange- ment for the manufacturer's representativeto be present for three to four years for Phase I factories,as originallyproposed under turnkey arrange- ments but dropped because of cost overruns, may have produced improved operatingresults. 7.10 The research componentwas not implemented as planned, since the government preferred to use an Italian Government grant and technical assis- tance to implemenZ a two-year breeding, pest and disease control research program which became operational only in 1979. The project appraisal envisageda trainedextensionstaffof 537, hardly a realisticproposalgiven the limitedavailability of trainedmanpower at the nationallevel. In prac- tice, the eleven senior positionswere filled with qualifiedstaff, but of the 344 extensionassistantsrecruited,only 32 had prior trainingin agri- culture. The extensionstaff have been involvedprimarilyin procurement of raw cashewnuts, (para. 7.20 et. seq). 71/ Phase I refers to this project, Phase II refers to a second project (Credit 801-TA) which provided a further three factories, thereby in- creasing processingcapacityby an additional30,000 tons p.a. - 50 - 7.11 The overall utilization of the technical services has been unsatis- factory. The provision for company representatives to supervise the opera- tion of the factories was dropped vrith consequent deleterious effects on processing. With the research program not being implemented the provision for internationally-recruited research staff was not utilized. An interna- tionally-recruited financial controller was recruited for two years from 1978 to 1980, but was unsuccessful in creating an adequate accounting system. 7.12 CATA's accounting and other records have not been adequate to reflect its operations; the last audited accounts submitted to the Bank were for the year ended September 30, 1976. 7.13 Procurement Problems resulting from the closing of the Tanzania- Kenya border, lack of foreign exchange and distance from Mtwara to Dar-es- Salaam were all problems beyond control of project staff and contractor. 7.14 However, mistakes in the initial estimates of water supply costs and difficulties with the negotiation of contracts resulted in the escalation of the cost of water supply development for the factories, from an appraisal estimate of US$150,000 to a negotiated contract of US$1.6 million awarded in July 1977. D. Issues (i) Causes of Crop Production Collapse 7.15 The decline in crop output to about one-third of the peak output levels of 1973/74 has meant that even if there had been no implementation problems, the project would still have failed to achieve its objectives. 7.16 Assessment of the causes of output decline remains inconclusive because of the extremely sketchy data available on cashewaut production in Tanzania, which was a factor evident in rhe original project preparation and at all subsequent stages of the project. In light of the dependence of proj- ect performance on crop production, it would have been desirable to build into the project, from the outset, a capacity for information gathering and analysis beyond that envisaged in the agricultural research component (which was in the event not implemented). 7.17 A review of analyses, which have attempted to explain the decline, suggests that the collapse of output is the result of a number of causes, which acted to reinforce each other. The particular impact of each, however, being difficult to isolate. These are: (a) Villagization The relocation of the rural population in the villagization program was quite thorough in the cashewaut areas (unlike in the other tree crop areas, where there was very little relocation). Labor on crop production--weeding and collecting the nuts-is not a full time occupation and is more readily undertaken when the trees - 51 - are in reasonableproximity to the home. Relocationmade it more difficult to tend the trees and collect the nuts. This is now generally accepted by government as a prime cause of output decline. The villagizationprocesswas outside the control of the proj- ect, of the Bank and of CATA. However, a more thorough under- standing of the crop productionsystem might have led the Bank to predict the consequencesof the villagization program as it was implemented,and to have urged on the government the virtue of creating villages small enough to allow for continuiag proximity to the trees. (b) Prices The decline in the real value of cashewnut prices parallel to the impact of villagization meant that a sharp decline in the return to cashewautproductioncoincidedwith an increase in the labor requirement.In the event, it was not until November 1978 that the price issue was raised by Bank supervision and then with- out much vigor. The very nature of the covenant on prices in the loan agree- ment was inadequate. The inclusionof the term 'less appropriate processing and marketingcosts" is essentiallyambiguous,and could readily be interpreted so that the farmer'sincome is the residual after prior deduction of parastatalcosts, leaving the farmer to shoulderthe costs of any parastatalinefficiency. The increasing gap between farmgate and world prices reflected, as argued in the PCR, an increasing parastatalmarketing margin, in part the result of inefficiency and high overhead costs of CATA. As those costs in turn reflected the overhead costs of the factory, includingthe debt incurredin financing the project and the staff overheadsagreed to in the project, the projectit- self must be viewed as a factor contributingto the poor farm incentives. (c) Disease There have been crop losses as a result of disease incidence, in part the result of weather conditions,in part due to factors (a) and (b) above. On the one hand, the failure to clear and weed under the trees increasessusceptibility to fungal attack, while on the other hand, the resulting decline in yield further reduces the return to collecting the nuts and maintaining the orchards. The researchprogramcurrentlyin place, funded by the Italian Government,is addressingthis question. However, if a research program had been implementedwith the same urgency as the factory - 52 - construction program, the capacity could have been created to identifythe disease problem earlier and to tackle it more aggres- sively. (d) Aging of the Trees Failure to replant (wnich may well have predated the project) means the tree population is aging and some trees are reachingthe limits of their productivelife. This is in part a reflectionof the more general disincentive to produce,but it is also believed to result from a lack of knowledgeamong farmersof the productive age of trees, and thus may relate to the absence of an effective extension effort. Performance (ii) Institutional and Development 7.18 The project completionreport concludes that the overall perfor- mance of CATA during the project period has been relativelyweak, that its effortshave almost entirelybeen directedtowardsprocurement, marketingand processing;raw nut production,extension and research aspects have been neglected. The major constraints affecting CATA's performance include shortagesof skilled and experienced manpower,high staff tuarover,and, in retrospect, an overly ambitiousrole defined for this parastatal. 7.19 The shortageof skilledmanpower can be attributedto the overali scarcityof such manpower in the country in general,and to the location of CATAheadquarters, which were moved to Mtwara near the main production areas but geographically remote. 7.20 During implementation of the project, CATAunderwent several reor- ganizations which resulted in relatively little attention being paid to researchand extension. The issue of procuringraw nuts proved to take far more attention than analysisand resolution of the issue of why raw nuts were becomingincreasingly hard to procure. 7.21 The high turnoverin CATA's top managementhas had adverse effects on efficiency. Nevertheless in the context of the overall difficulties facing the Tanzanian economy the record of the company in implementing factory constructionand handling cashewnutexportswas not unsatisfactory. Even with a much more effectivemanagement team it was unlikely that the organizationcould have establishedan effectiveoutreachto handle extension and village purchasing arrangements, given its tasks in relation to processing. 7.22 A practicalweakness has been in relation to accounting. This is relatedboth to the difficulties in recruitingeffective manpower and to the difficultyin controllingthe village level purchasingsystem. However, it should be recognizedthat the project was implementedbefore there was an adequate accounting system in place. A certain minimum level of operational performance in relation to accounting would be an appropriate prerequisite to the commitment of such large funds. - 53 - 7.23 According to the trading results, CATA has earned reasonablenet profits from 1973174up to 1978/79,except during ;975/76 when it incurreda net loss of TSh 18.8 million. However,these profitswere mainly due to the export of raw nuts as the factoriescame into operationonly during 1978/79 and thereafter. As more factoriescome on stream, an increasingproportion of raw nuts were processed in CATA's factories and with declining world market prices and high processing costs, CATA's operations became less profitable. CATA has faced seriousliquidityproblemsin the past few years of mainly due to poor recovery of cash advances to villages (for procurement nuts), financing of long-term and short-term debtors and investments in excess processing capacity. The proposals to close down some of the factoriesand lay off staff have been among drastic actions initiatedby government to improve CATA's financialposition. Other measures such as substantialreductionsin processingoverhead costs, stepped-updebt recovery coupledwith increasedraw nut productionand export sales would be essential for the restoration of CATA as a commerciallyviable organization. (iii) Bank Performance 7.24 Given the evident failure of the project, the criticalcommentsin the project completion report regarding Bank performance can be fully en- dorsed and placedin even sharperfocus in this audit report. 7.25 The failure of the project suggests four broad questions: (i) should the Bank have supportedthe project? (ii) Could better projectdesign improve performance? (iii) Could more effective Bank supervision improve per- formanceduring implementation and, relatedto that question(iv) how did the decisionmade by the Bank in May 1978 to go ahead with Phase II of the proj- ect relate to Phase I performance? (a) Shouldthe Bank Have Supportedthe Project? 7.26 Given the level of output in 1972/73and the lively growth in pre- ceding years, the initial decisionto fund cashewnutprocessing facilities in Tanzaniawas justified. At the time, the subsequent output declinecould not have been predicted. (b) ProjectDesign 1. The Role of CATA It was unreasonable to suppose that CATA, an organization whichwas only coming into existenceas the projectwas in prepara- tion, could effectivelycarry out the responsibilities of pur- chasingand marketingof cashewnut, operating a number of factories involvinga quite complex processingsystem and, at the same time, handle extension and research activitiesfor crop production.A more narrowly defined role for CATA, a more cautiously phased implementationof the project, and the use of managementservices suppliedon a contractualbasis,might have eased the problem. - 54 - 2. Crop Production Although it is unreasonable to suppose that the future fall in production could be predicted, the basis for believing that the project could increase production was, from the outset, shaky. Cashew was not a crop for which substantial research results existed or in relation to which there existed a large body of agro- noi.c specialists, either in Tanzania or internationally. Thus the belief that production could be expanded through extension was wishful thinking but laid the initial foundations for the subse- quent failure of the Bank to respond to the deterioration in the crop production situation swiftly enough when it occurred. 3. Underestimation of project costs resulted in the turnkey contract envisaged at appraisal being changed and the training of factory technicians and the provision of manufacturer's representatives to provide technical advice for the first three to four years of operation being left out. 4. Research and information. Given the evident lack of knowledge about cashewnut production the importance of research and informa- tion gathering was under-emphasized in the project formulation, and even the element included was not implemented. As a result the actual cause of output decline in the 1970s still remains a matter of speculation, with explanation insufficiently grounded in con- crete evidence. (c) Bank Supervision 7.27 The main weakness of Bank supervision was the failure to identify the fall in crop production or spell out its consequences for the project speedily. In the early years supervision was mainly concerned with the con- struction of project facilities and the staffing problems of CATA (particu- larly related to accounts). The failure to move ahead with a research program was also identified, but there was no vigorous follow-up to push the research component. It was only in the supervision report of June 1977 that the output decline was identified and the negative effect of villagization mentioned. The October 1978 supervision report details the poor output figures, and the issue was raised at that time with the Tanzanian authori- ties, but the Bank did not push for any specific action. Although the March 1979 supervision report concentrated on production problems, it was not until March 1980 that villagization was identified in unequivocal terms as a source of output decline and a strong note of alarm was struck about the output decline. The March 1980 report contains the first attempt at a comprehensive analysis of the crop production problem. 7.28 Given that crop output increase was an explicit project objective, it might even be thought that an assessment of crop performance should have been a routine aspect of the supervision. - 55 - (d) The Phase II Decision 7.29 The failure to monitor crop performance adequatelyin the supervi- sion processmust providepart of the explanation for the surprisingdecision of the Bank to go ahead with the fundingof threemore factories with a capa- city of 30,000 tons in Phase II when there was alreadya processingcapacity of about 80,000 tons in operationor in the processof construction, which proved to be in excess of crop productionof 68,000 tons. Phase 1I was appraised in May 1978, but already the Basic Economic Report of December 1977, preparedby the Bank, had indicatedthat by 1975 output had fallen to 80,300 tons from the level of 145,100tons recordedin 1973.72/ 7.30 While the Phase II project is not a subjectof this audit report of Phase I cannotbe the decisionto go ahead with Phase II before completion ignored, both as evidence of a flaw in the Bank's informationsystem as regards cashew crop productionand because of its inability to associate Phase II implementationwith Phase I performance. The installation of addi- tional capacity in Phase II contributedto the excess capacity which has resultedin four of the five factories lying idle. 72/ Report No. 1616-TA,dated December1977,Annex VI Table 3. - 56 - ATTAME I I * Agencecanadlenne d6veloppenent de Canadiarin inlemnaonal Developme Agency cOFIACIR' ternational S CONETS HSA c}LO Hu I O Canwe Cnada K1A K1A OG4 IG4 May 18, 1984 *""" " hv f lenm ova 902/00610 Mr. S.S. Kapur Director Operations Evaluation Department The World Bank 1818 H Street N.W. WASHINGTON,D.C. 20433 U.S.A. Dear Mr. Kapur, Re: Project Performance Audit Report- Tanzania Rural Development Project (IDA Credit 705-TA) Mr. Bassett has requested that I respond to your letter of April 27 on the above subject. In general we found the performance audit report on the Tabora Project to be an excellent overall analysis of the achievements, shortfalls and effectiveness of the project. Your appreciation of these issues is very much in line with CIDA's impressions gleaned from the :egular re- porting we received during the implementation of the project. Indeed, the kinds of difficulties encountered are rather similar to those that have emerged in other large rural development projects in Tanzania (a point made in the Project Completion Report). This leads to a comment on presentation of your report. It could be of value at some polnt to summarize and comment on the lessons learned by the Bark and The Tanzanian Government, perhaps using a similar format to to that in the Project Completion Report Overview. One lesson or issue that might also be included is the need to ensure projects of this scale contain the necessary mechanisms for redesign if circumstances or critical assumptions change (as they did in Tanzania) - or indeed if the original project design is found to be faulty. How this can be accomplished in practical terms given the lengthy gestation of this pro- ject is of course problematical. We look forward to receiving the final report and also wish to thank you for the material on other Bank projects in Tanzania which adds to our knowledge of the agricultural sector. Yours sincerely, Roger Dumelie Country Program Director Anglophone Africa Branch CanadEa - 57A- hL'AClT II S COMMEN BORROIWER' ZCZC DIST6507 WUI600 DIST REF : TCP FCA OEDDR WUJ600 41352 INfKUU TZ 41352 WNKUi TZ SHIV S. KAPURI OPERATIONS FVALUATION DEPARTMFNTP INTRAFRAD. RE - YOUR TELEX CONCERNING PROJEC;T PERFORMANCF AUJDIT REtPORT (.) HAVE SOME THROUGH VOLU.ME I WItH INNER FOCtS ON TABORA PROJECT (.) POINTS IN THE MEMORANDA ARE A REPI. ICA OF THOSE THlAT RAISED WERE CONSTANTLY BEING RAISED DURrNG E:VIEw miSSTIoNSC.) WE AGREE WITH ASSEStENT OF FACTORS HELD Rl.SPONSIBLE FOR FAILURf OF TflBORA PRU.IOECT WITH THE LrONS SHARE BEING TAKEN By DISRUPTIONS CAU)SED BY STATF OF THE ECONOMY. LESSONS GAINED FROM TABORA AND THE OTHER WORLD BANK AIDED PROJECTS WILL. HELP US FORMIULATE MORE REAiLISTIC PROJECrS IN FUTURE(.) REGARDS. C.N. KEENJA, PRIME MINISTER'S OFFICE, DAR ES SALAAM. 25/5/1984 41352 WUMKUt TZ 41352 WKKUU TZ...... 005113 0925 250584 0171013 7 600 =05250646 NNNN - 573 - C.-oCnts fom DB A z Um - *s Page 1 S - " f urni I7~~M Usuka y.aTanib -NW P.O. he 263JP.. 26311 SL.P. 268 inmu251I1 -' -~ l15th June, 1984. AW &2/20/AO Director, Oprations Evaluation Department, World Bank, W&-UNGTO1N DC. TZAXMKZ RU MA LP1± BA;Z C;C=--ITS 0 iM!~ AfLXIT Lf-IORA1IDL IN~ REMECT 01? GZI!fl COTTOK MW 1TJ7 (CEDI T 454 - TA) AI-D HAMIV%'VG 0AC.?o POJXCT (CRDIMr'L 802-TA) Reference is being made to y6ur telex requesting the Tanzania Rural Develoluent Bank's commerts on six projects funded by the World Bnk. It is the wish of the Tanzania Rural Develop- znent ank (TEDB) to coment only on those projects in which the lnk (i.e. mMB) was directly inzolved, and specifically on.projects compionents that utLlized credit from the Bank. 8s comnetarzy vill therefore be limited to the credit com- ponentbs of thet- i) Geita Cotton Project II) Hational Llie Project dii) Tobacco Handlin- Project. Vhe Geita Cotton Project: Fertilizers, insecticides, UIV pu.ps, and batteries were pro- vided to project villges oz credit fro. !TEB starting 1975/ 1976 season. Although there was an overall credit recovery rate of 68%, TED3 loans could not be recovered in full, Loan re-e.-yents collecte& froM both the villag_e loanees and subsidy frcn Tanzania Cotton Authority (TCAL) could not fully liquldate the outstandi.ng loan arrears. To date T.Shs. 7mil. is stiln outstanding again3t the project village3. The inability of project beneficiaries to meet their loan obli- gations is attributable to poor cotton :roduction. This seems to sup_iort tae observation that the in_zut recammendations were inappropriate. If the anticipated inproved yields (an incre- ase of 12% from the eecond year) were attained the loans might have been recovered fu13y. In this respect, therefore, TD3 Shares the view of,the AUDIT that the project lacked technically vinable input packages. - 57C - ATA T III Page 2 Me National aJize ProJect: The Credit component administered by R perfmed poorly: Saize Inputs -Lgaus: 1Mherecovezy rate of maize izput loan was a mere 5%. This is attributable to:- i) Sale of maize in the lucrative infoxmal market where prices offered are higher than those of the National Milling Corporation. Loans could not be recoveredat source from. these markets, ii) lack of overall anticipatedmaize surplus: P}roduction in excess of subsistencewas hard to come Xy in some instances due to poor technoloar (use of the primitive hnd hoe and aninal pomer) and poor crop husbandxy practices. IDA may wish to consider adopting a more liberal policy in sup- porting tractorIzation in areas where poor techno- logy is a proven draw-back. Vi,la e Stores: Under the National Maize Project a total of 914 village stores were to be built out of IDA/(SAID credit funds. The following problems hindered a successfulimplementationof the programmez- a) Some Of the villages were expected to build stores on credit from MTM while others had their stores built out of grants from Rural Develolpment Funds administered by District DevelopientDirectors and 'onmmxmity Development Trust Pund (CDIE). This arra_ nf-ement was a diB-incentive to the would be borro- wers for village stores. Therefore only 39 stores were built out ofZ Gredit from MRDB. b) During prosramme implementation there were too mamy self help schemes on-going in the villaZ-es. These included construction of schools, dis-pensaries, roads and the like. Villa6es found the provision of la- bour for constructing stores stitl on self help basis over taxing, c) Leck of clear policy on ownership.of stores: Vill3avs are Lf?aketing Agents for .he lational Ydlling Corporation (131LC. They carzy out on 12C's behalf the crop _urchasing functions and store pur- chased produce before it is transported to District/ Regional/National storage centres. Since ROC did not pay for the storaze costs for produce stored in the villar-e stores, villa-=es rithout stores were not very much attracted to owning stores and expected MUCto build stores in villages in need off storage facilities (i.e. those not hiaving benefitted frcm MR and CDIF funds .or building stores). .---- /3- 57D - hATTACH III Page 3 d) An acute shortage of building materials contribu- ted to the delays in construction of the approved. st0res. Tobacoo Handling Project Credit 802 - Us i) Balking Bags Component: 350,000 tobacco bulking bags were to be given to farmers under this project component, Due to the fact that the would be borrowers had not been gi- ven adequate extension advice and training on the use and benefits of the bags, they were reluctant to apply for loans from the Tanzania Rural Develo- pment Bank. ii) Baling and Grading Centres: The following problems hindered implementation of this project components: Dispute of Ownership: a) Initially there was a dispute as to whether the centres should be ovmed by TAT or villa-' ges. Sorting this out took sometime and this led to delays in inplementation. b) Failure of villages to fulfill credit conditions: Villages could qualify for bazling and grad,ing centre loans only if they bad attained a mini- mm productionlevel of 100,000 k)-s. of tobacco. However villages that attained the minimum levels were found to have performed poorly in the repay- ments o' TREB loan. These did not qualify for credit due to heavy loan arrears. c) Shortage of building materials: At the time when the aforesaid issues were sor- ted out i.e. towards the end of the project imp- lementation, the country was experiencingseri.- ous shortages of building materials. d) Location of hleCentres: Faxmers wanted such centres located within or very near their fazns rather than at places far removed from these -narms as envisaE:ed under the project. 'le hone you will find this commentary helpful. TAi3LlAU?aU- DWV1IOZ%MOT BAJIK (G.A. a) ~~~~~~~~~~~~~~~~~~.. .4.. / - 57E- ATTACT III Page 4 c.c. P'incipal Secretaxr, Ministry of Finance, P.O. Box 9111, DR ES SALL=. Paincipal Secretazry, iDiistry of Agriculture, P.O. Box 9192, DER ES SALLAM. 6?I, 0 j ~~~~~~~~~~ ~~~~~~~~~~ ..'~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... ~~~~~~~~~~~~~~~~~~~~~~~~~.r. A~~~~~~~~~w~~~~~~~~~ L"' . ... .0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0 ,a--, -, .. ~ ~~~~'- - .>~~~~~~~~~~~~~~~~~~~~~~~...... .... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.~~~~~~~~~~~~~C - IBRD 18072 (PPAJ APRIL 1984 TANZANIA AGRICULTURAL PROJECTS . :,> r,qCREDITS454-TA, 606-TA, 658-TA, 703-TA,802,-TA,LOAN 1014-TA Arusho ' Moshi [ZZ M3Geita CottonProjecr,454-TA Notional Maize Project,606-TA : U R S H A { A ^ 4 <</\ / TobaccoPrucessing Project. 658-TA lilfi l Baoi>g boGroving Areos Tbcl Factory _*. oMomhaaf g' *r 0/X//0 oTabora Rural Development,703-T4 TobaccoHandling Project, 802-TA 0~i\\ Warehouses CoshewnutProject, 1014-TA n ls / _Tago :~:-:: - CashewGrowingArea Seger. 7 \ t.i T^N FactoryCenters D (0 D A\M^ - . / Mcin Roads ,Xd orn 3~ -| |Railways -. e J t -- - ( \ fl- Rivers Regional Boundaries InternationalBoundaries- THIS MAP hIIRO 1728-PPS WAS SPECIFICALLY PREPARED g-CO r.,p A S:t : -/ FORTHISPROJECTPERFORMANCEAUDrrREPORT : - 11 f11,| X r ~ - : - / INVIAN OCEAN - KIL0M^RS .- O, :- i_o 150 .s/H ,''-,R,-;x,.............. >/, MILES 6 S W O T/MRG OO O R ;RRv .* ... .... R:R b 160iini / . >{y-~~~~~~. .-R-.................. l-N---I- ) 5 f -- - R-~~~~~~~. -7 .- ::R:R:::::R: :. :: ............. ::,;:........... J R.. -... ...- R. .......................... ~ ~ ~ ~ ~ .*,I.,. dTh U V~ .. .- .*R - -R I.'- r-J." De W- U--s I iwt \ -:::-:::::: ~~~~~~~~~~... :-::::'St:- ~~~........ .::.:.::...W..ordt# .............. .... : > ..................... , . .N 2 MOZAMBIQUE 2'o_ 36 0ad d~e
Группа Всемирного банка · Project Performance Assessment Report
Tanzania - Agricultural Products and Rural Development Projects (Vol. 1 of 2)
Открыть оригинал документа
Полный текст размещён на сайте публикующей организации. lawenc.com индексирует метаданные и ведёт на официальный источник.
Вернуться к постатейному просмотруПолный текст
Основные сведения
Организация
Группа Всемирного банка
Тип документа
Project Performance Assessment Report
Страна
Танзания
Источник
Всемирный банк