Tanzania - Agricultural Products Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 9253 PROJECT PERFORMANCE AUDIT REPORT TANZANIA SECOND CASHEWNUT DEVELOPMENT PROJECT (CREDIT 801-TA) PYRETHRUM PROJECT (CREDIT 1007-TA) SMALLHOLDER TEA CONSOLIDATION PROJECT (CREDIT 1037-TA) COCONUT PILOT PROJECT (CREDIT 1070-TA) DECEMBER 28, 1990 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS US$1.00-TSh 8.30 (1978) 8.30 (1979) 8.20 (1980) 8.30 (1981) 9.28 (1982) 11.14 (1983) 15.29 (1984) 17.47 (1985) 32.70 (1986) 64.26 (1987) 112.00 (1988) 130.10 (1989) GOVERNMENT OF TANZANIA FISCAL YEAR July 1 - June 30 ABBREVIATIONS CATA - Cashewnut Authority of Tanzania CPIPP - Cashewnut Production Improvement Pilot Project CPS - Central Projects Staff FAO/CP - Food and Agriculture Organization/Cooperative Program FRG - Federal Republic of Germany GOT - Government of Tanzania GTZ - Gesellschaft fUr Technische Zusammenarbeit IDA - International Development Association Kilimo - Ministry of Agriculture NCDP - National Coconut Development Program NORAD - Norwegian Agency for International Development OED - Operations Evaluation Department PCR - Project Completion Report PPAR - Project Performance Audit Report RIMU - Road Improvement and Maintenance Unit RMEA - Resident Mission in East Africa SAR - Staff Appraisal Report TCMB - Tanzanian Cashewnt Marketing Board TPB - Tanganyika Pyrethrum Board TTA - Tanzania Tea Authority UAC - Uyole Agricultural Center FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office to Directorenrcohal Operations Evaliatia December 28, 1990 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on Tanzania Second Cashewnut Development Project (Credit 801-TA) Pyrethrum Project (Credit 1007-TA) Smallholder Tea Consolidation Project (Credit 1037-TA) Coconut Pilot Proiect (Credit 1070-TA) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on Tanzania - Second Cashewnut Project (Credit 801-TA), Pyrethrum Project (Credit 1007-TA), Smallholder Tea Consolidation Project (Credit 1037-TA) and Coconut Pilot Project (Credit 1070-TA)" prepared by the Operations Evaluation Departmen:. Attachment This document has a restricted distribution and may be used by recipients only in the performance or their official duties. Its contents may not otherwise be disclosed without World Bank authorization. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT TANZANIA SECOND CASHEWNUT DEVELOPMENT PROJECT (CREDIT 801-TA) PYRETHRUM PROJECT (CREDIT 1007-TA) SMALLHOLDER TEA CONSOLIDATION PROJECT (CREDIT 1037-TA) COCONUT PILOT PROJECT (CREDIT 1070-TA) TABLE OF CONTENTS Paae No. PREFACE ........................................................ i BASIC DATA SHEETS .............................................. iii EVALUATION SUMMARY ............................................ xi I. BACKGROUND................ . ......... 1 II. THE PROJECTS...................................... 3 A. Second Cashewnut Development Project........ 3 B. Pyrethrum Project........................... 6 C. Smallholder Tea Consolidation Project....... 8 D. Coconut Pilot Project............. 10 III. PROJECT OUTCOMES............... ............... s. 12 IV. FINDINGS AND ISSUES............. ...... .... 13 A. Preparation and Appraisal Inadequacies...... 14 B. Timing of Follow-on Projects................ 17 C. The Importance of Applied Research to Development .... ............ 18 Tables 1. Raw Cashevnut Purchases........................... 21 2. Pyrethrum Production Statistics................... 22 IBRD 13096R - Cashevnut Growing Areas and Factories IBRD 14083R - Smallholder Tea Areas IBRD 14814R2 - Coconut Pilot Area This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. PROJECT PERFORMANCE AUDIT REPORT TANZANIA SECOND CASHEWNUT DEVELOPMENT PROJECT (CREDIT 801-TA) PYRETHRUM PROJECT (CREDIT 1007-TA) SMALLHOLDER TEA CONSOLIDATION PROJECT (CREDIT 1037-TA) COCONUT PILOT PROJECT (CREDIT 1070-TA) PREFACE 1. This is a Project Performance Audit Report (PPAR) on four IDA- assisted agricultural sector projects in Tanzania. The credits,11 to a total value of US$58.3 million (M) were approved by the Board in May 1978, April 1980, June 1980, and October 1980, respectively. They were closed in June 1989, December 1985, June 1989, and March 1989, respectively. A total of US$8.3 M in undisbursed balances was cancelled.2/ 2. The PPAR is based upon a review of the Staff Appraisal Reports (SARs),3/ the corresponding President's Reports,4/ the Development Credit Agreements and Project Agreements,5/ the Project Completion Reports (PCRe),6/ and the transcripts of the Executive Directors' meetings at which the projects were considered. Bank files and other relevant documents were also studied, and discussions were held with Bank staff associated with the projects. 3. An OED mission visited Tanzania in June 1990. The mission held discussions with officials in the ministries and other agencies responsible for the projects. Their kind cooperation and valuable assistance in the preparation of this report is gratefully acknowledged. Field trips to a 1/ Credit 801 - US$27.5 M; Credit 1007 - US$10.0 M; Credit 1037 - US$14.0 M; and Credit 1070 - US$6.8 M. 2/ Credit 801 - US$1.4 M; Credit 1007 - US$5.1 M; Credit 1037 - US$1.0 M; and Credit 1070 - US$0.8 M. 3/ Report numbers: 1868-TA dated April 26, 1978; 2749b-TA dated March 20, 1979; 2814-TA dated May 9, 1980; and 2906-TA dated August 14, 1980. 4/ Report numbers P-2314-TA dated May 1, 1978; P-2754-TA dated March 31, 1980; P-2814-TA dated May 14, 1980; and P-2878-TA dated September 18, 1980. 5/ Dated June 14, 1978; October 24, 1980; August 21, 1980; and October 24, 1980. 6/ Report numbers: 8625 dated May 8, 1990; 7044 dated December 16, 1987; 8510 dated April 13, 1990; and 8626 dated May 8, 1990. - ii - representative selection of project sites were also undertaken for the purpose of familiarization and to interview staff responsible for implementation and project beneficiaries. 4. Following standard OED procedures, copies of the draft PPAR were sent to the Government of Tanzania (GOT), the parastatal Borrowers and the cofinancing agencies for comments but none were received. - 111 - PROJECT PERFORMANCE AUDIT REPORT TANZANIA SECOND CASHEWNUT DEVELOPMENT PROJECT (CREDIT 801-TA) BASIC DATA SHEET YEY PROJECT DATA Appraisal Actual or Actual as Z of it-m estimate Estimate Actual appraisal estimate Total project cost (US$ million) 36.3 n.a. Credit Amount (US$ million) 27.5 24.5 89 Economic rate of return X 16 negative Institutional performance poor CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS FY79 FY80 DLS FY82 FY83 FY84 FY65 FY86 FY87 FY88 FY89 Appraisal estimate (US$ M) 4.6 15.0 22.6 24.6 26.0 27.2 27.5 24.5 24.5 24.5 24.5 Actual (US$ M) n.a. 16.0 17.3 19.3 19.6 19.6 19.8 19.9 20.4 21.3 24.1 Actual as % of appraisal n.a. 106 77 78 75 72 72 81 83 87 94 Date of Final Disbursement: December 29, 1989 PROJECT DATES Original Plan Revisions Actual Appraisal 07177 Appraisal Follow-up 02178 Negotiations 04117/78 Board Approval 05110178 Credit Agreement Date 06/14/78 Effectiveness 10102178 Date physical components completed 12!83 06189 Proportion completed by that date (2) 100 100 Closing Date 12184 12131187 06130189 STAFF INPUTS (staff weeks) FY76 FY77 FY78 FY79 FY8 FYI FY82 FY83 FY84 Y8 FY86 FY87 FY88 Total Preappraisal - 12.4 .2 - - - - - - - - - - 12.6 Appraisal - 1.6 47.1 - - - - - - - - - - 48.8 Negotiation - - 9.2 - - - - - - - - - - 9.2 Supervision - - .6 5.7 9.5 4.0 16.9 5.2 9.5 26.7 17.5 9.8 7.7 113.0 Other - - 2.1 - - - - .3 - - - - - 2.5 TOTAL - 14.0 59.4 5.7 9.5 4.0 16.9 5.4 9.5 26.7 17.5 9.8 7.7 186.0 - iv - MISSION DATES Date No. of Specializations Performance Type of (anlyr) persons Represented Rating Trend Problems LA Lb La Ld Identification/Preparation. 1976 Appraisal 06-07/77 Supervision 1 09178 1 FA 2 2 0 Supervision 2 04/79 1 FA n.a. n.a. n.a. Supervision 3 12/79 2 FA,A 2 3 0 Supervision 4 06/80 2 FA,A 3 3 0M Supervision 5 02/81 1 FA 2 2 M,0 Supervision 6 10/81 2 FA,A 2 2 1,0 Supervision 7 02/82 le Supervisiot% 8 06/82 2 FA,Eng. 3 3 F,M,T,O Supervision 9 12/82 2 A,E 3 3 F,M,T,O Supervision 10 06/83 1 E 3 3 M,P,T,F Supervision 11 10/83 1 A 3 3 M,P,T,F Supervision 12 10/84 2 A,A 3 2 M,P,T,F Supervision 13 02/85 2 A,EL_ n.a. n.a. n.a. LA Supervision 14 07/85 2 A,E 1 1 Supervision 15 02/86 2 A,E 2 2 Supervision 16 06/86 1 A 2 2 Supervision 17 10/86 1 A 2 1 Supervision 18 01/87 2 A,A 2 1 Supervision 19 05/87 1 A 2 1 Supervision 20 11/87 2 A,A 2 1 Supervision 21 03/88 1 A 2 1 Completion 06/89 2 A,E OTHER PROJECT DATA Borrower: Government of Tanzania Executing Agencies: The Cashewnut Authority of Tanzania (CATA) (later changed to Tanzania Cashew Marketing Board) Follow-on Project: Name: Cashew and Coconut Tree Crops Project Credit Number: Credit 2050-TA Amount (US$ million) 25.1 Approval Date June 27, 1989 /a FA-Financial Analyst, A=Agriculturisti Eng-Engineer: E-Economist. /b 1-Problem-free/or Minor Problems; 2-Moderate Problems, 3-Major Problems. Lc 1-Improving; 2-Stationary; 3=Deteriorating. Id F-Financial; M=Managerial; T-Technicalt P-Political; 0-Other. le Visit to evaluate and discuss with GOT their proposals for a Cashewnut Rehabilitation and Expansion Project. If This mission discussed the extension o' the closing date and the redesign of the project from processing facilities construction and operation inro a pilot cashevnut production project. L& Status and Trend no longer included in supurvision mission reports. From 1985 onward status and trend parameters indicate "overall status" and "development impact." PROJECT PERFORMANCE AUDIT REPORT TANZANIA PYRETRUM PROJECT (CREDIT 1007-TA) BASIC DATA SHRET KEY PROJECT DATA Appraisal Actual or Actual as I of Item estimate Estimate Actual appraisal estimate Total project cost (US$ million) 13.2 9.1 g 69 Credit Amount (US$ million) 10.0 10.0 100 Economic rate of return X 17 negative Institutional performance - marginal CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS IM E82 FY83 FY84 M I Y86 FY87 Appraisal estimate (US$ million) 0.3 2.6 5.0 7.4 9.3 10.0 10.0 Actual (US$ million) - 1.4 3.2 3.7 4.3 4.8 4.9 Actual as Z of appraisal (2) - 54 64 50 46 48 49 Date of Final Disbursement: September 9, 1986 PROJECT DATES Original Plan Actual Appraisal 04/79 04-05179 Negotiations 02/80 02116/80 Board Approval 04/80 04122180 Credit Agreement Date 06/80 10/24/80 Effectiveness 01/23/81 04120/81 Date ph7sical components completed 06/86 06/86 Proportion completed by that date (%) 100 55 Closing Date 12131185 12131185 STAFF INPUTS Lb (staff veeks) FY79 FY80 PY81 FY82 PY83 FY84 FY85 FY86 Total Identification/Preparation 20.9 - - - - - - - 20.9 Appraisal 51.4 33.3 - - - - - - 84.7 Negotiation - 6.0 - - - - - - 6.0 Supervision - 5.3 16.7 29.5 14.9 14.4 20.2 7.9 109.0 Other 1.6 0.8 - - - - - - 2.4 Total 73.9 45.5 16.7 29.5 14.9 14.4 20.2 7.9 223.0 - Vi - MISSION DATES Dat No. of Specialisations Performance Type of (molyr) personA Raoresented Rating Tren4 Probema La LA La Ll Identification/Preparation Appraisal 04-05179 6 a,b,c,d,e - Supervision 1 06/80 3 a,b,c 2 2 M Supervision 2 10/80 2 b,d 2 2 M Supervision 3 06/81 1 b 3 3 M,F Supervision 4 01182 3 b,d 3 - - Supervision 5 03/82 1 e - - - Supervision 6 06/82 2 b.c 3 3 F,M,O Supervision 7 02/83 1 o 3LA - - Supervision 8 02/83 1 a 3 2 T,r.M Supervision 9 10/83 1 c 3 2 T.F,O Supervision 10 05/84 2 c 3 2 T,F,O Supervision 11 11/84 1 a -LA - - Supervision 12 01/85 1 c -Lh - - Supervision 13 05/85 2 c,d 3 2 T,F,O Supervision 14 02/86 6 c 4L. - OTHER PROJECT DATA Borrower: Government of Tanzania Executing Agencies: Tanganyika Pyrethrum Board (TPB) Uyale Agricultural Center (UAC) Ministry of Works (MOW) Follow-on Project: None /a Estimated on basis of incomplete data (see para. 3.60 of PCR and Annex Table 7). lb Input as staff weeks. Source: World Bank Planning and Budgeting Dept. 1e a-economistg b-agricultural economists c-financial analyst, d-agriculturalistiagronomists e-highway engineer. 14 1-Problem-free/or Minor Problems; 2-Moderate Problems, 3*Major Problems. a4 1-Improvingi 2-Stationaryt 3-Deteriorating. If F-Financials M-Manaeriali T-Technicals O0-Other. LA Mission reviewed only project's roads component, and issued a Back-to-Office report. Lh Mission reviewed only preparation of action plan for the pyrethrum industry, and issued a Back-to-Office report. IL 1-None or minor problems, 2-Moderate Problems, 3-Major problems which are not being addressed adequatelyl 4-Major problems which are not being addressed adequately and which require further action on the part of the borrower/implementing agency. -Vii - pROJECT PERFORMANCE AUDIT REPORT TAN.ZAN.IA SMALLHOLDER TTA CONSOLIDATION PROJECT IrREDIT 1037-TA) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Actual as X of LM AstiLmato Sstimate Actual ODorasL& estimate Total project cost (US$ million) 20.0 14.2 71 Credit Amount (US$ million) 14.0 13.04 93 Economic rate of return I 21 28 133 Institutional performance Satisfactory CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS 81 E82 FY81 FI§4 FI EY86 E10W EM Appraisal estimate (US$ million) 4.5 9.0 11.2 12.9 14.0 14.0 14.0 14.0 14.0 Actual (US$ million) 0.25 1.25 2.3 3.57 4.3 5.72 6.47 8.33 13.04 Actual as X of appraisal (Z) 6 14 21 28 31 41 46 60 93 Date of ..nal Disbursement: December 12, 1989 PROJECT DATES Original Plan Revisions Actual Appraisal 07178 Post-appraisal 10179 Negotiations 04/80 Board Approval PY79 06103/80 Credit Agreement Date 08/80 Effectiveness 11/80 01-02/81 04/81 Date physical components completed 12/85 12/88 Closing Date 06/86 06189 STAFF INPUTS (staff weeks) Pre- FY81 FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 FY89 PXY9 Total Preappraisal 6.4 6.4 Appraisal 181.6 181.6 Negotiation 1.5 1.5 Supervision 7.3 17.8 11.1 19.3 18.4 16.4 18.0 8.4 8.8 125.4 Other 5.7 1.0 .1 .1 .6 .3 2.4 6.5 2.7 19.5 Total 195.1 8.3 17.8 11.2 19.3 18.4 17.1 18.4 10.8 15.3 2.7 334.3 - Viii - MISSION DATES gasg No. of Specialisations Performance Type of (molyr) prse Reiresented Rat1ns andI ZERUbAAM La LA La LA Identification/Preparation 1976 Appraisal 07178 Post-Appraisal 10/79 Supervision 1 10180 2 n.a. 2 2 F,N,T Supervision 2 09/81 2 F.A 3 3 F,N,T Supervision 3 04/82 1 F 2 3 F,M,T Supervision 4 03183 2 F,A 3 3 Y,M,T Supervision 5 06/83 2 F,A 3 3 F,M,T Supervision 6 09/83 2 n.a. 3 3 M,P,T Supervision 7 03/84 1 F 3 3 F,M,T Supervision 8 07/84 1 F 3 2 F.M,T Supervision 9 10/84 2 F.A 3 1 N'F'T Supervision 10 02/85 2 F,A 3 3 MN,F,T Supervision 11 10/85 1 F 3 n.a. Supervision 12 02/86 1 A 3 3 n.a. Supervision 13 05/86 2 FIE 3 3 n.a. Supervision 14 12/86 1 F 3 3 n.a. Supervision 15 03/87 1 A 3 3 n.a. Supervision 16 11/87 2 P,A 2 2 n.a. Supervision 17 02/89 2 Ec,E 2 1 n.a. OTHER PROJECT DATA Borrower: Government of Tanzania Executing Agencies: Ministry of Agriculture (MOA) Tanzania Tea Authority (TTA) Follow-on Projecti None ag A-Agronomist or Agriculturalistj * - Engineer; Ec-Economists F-Financial analyst. lb 1-Problem-freelor Minor Problemsi 2*Moderate Problems, 3-Major Problems. Le 1-Improving; 2-Stationarys 3-Deteriorating. Id F-Financiala M-Managerial, T-Technical - ix - PROJECT PERPORMANCE AUDIT REPORT TANZANIA COCONUT PILOT PROJECT (CREDIT 1070-TA) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Actual as X of Item estimate Estimate Actual appraisal estimate Total project cost (US$ million) 8.5 8.1 95 Credit Amount (US$ million) 6.8 6.8 100 Economic rate of return % - - - Institutional performance - satisfactory - CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS CY81 CY82 CY83 CY84 CYS5 CY86 CY87 CY88 CY89 Appraisal estimate (US$ million) 1.700 2.900 4.100 5.500 6.500 6.800 - - - Actual (US$ million) .326 .723 1.069 1.649 2.402 3.775 5.140 5.953 5.971 Actual as X of appraisal (%) 19 25 26 30 37 55 Date of Final Disbursement: September 19, 1989 PROJECT DATES Original Plan Revisions Actual Appraisal 10/79 Negotiations 07180 Board Approval 10/07/80 Credit Agreement Date 10/24180 Effectiveness 01/23/81 04/20/81 Date physical components completed 09/30/85 12/87 12i88 Proportion completed by that date (%) - - Closing Date 03/86 03/87 03188 03/89 STAFF INPUTS (staff veeks) PY78 FY79 FY80 F8l FY82 FY83 FY84 FY85 FY86 FY87 ?Y88 Total Preappraisal 11.2 47.4 2.5 - - - - - - - - 61.1 Appraisal - - 30.3 - - - - - - - - 30.3 Negotiation - - .0 6.8 - - - - - - - 6.9 Supervision - - - 3.5 12.8 4.3 8.2 20.8 21.1 10.6 5.1 86.5 Other - .1 .2 - - - - - - - - .3 Total 11.2 47.5 33.0 10.3 12.8 4.3 8.2 20.8 21.1 10.6 5.1 185.0 MISSION DATES Date No. of Specialisations Performance Type of (molyr) persons Represented Ratin Trend Problems La Lb Ic Preparation 05/79 Appraisal 06/80 4 n.a. - - - Supervision lid 02182 2 n.a. n.a. - * Supervision 2 08/82 2 a,a 1 2 - Supervision 3 01/84 1 e 1 2 - Supervision 4 08184 1 e 1 2 - Supervision 5 11/84 1 e - - - Supervision 6 03/85 2 e,a 1 1 - Supervision 7 07/85 2 e,a 1 2 - Supervision 8 12/85 2 a,e 1 - - Supervision 9 06/86 2 a,e 1 - - Supervision 10 01187 2 a,e 1 - - Supervision 11 12/87 1 a 1 - - Supervision 12 04/88 1 a 1 - - Supervision 13/e 10/88 4 a,e,f 1 - - OTHER PROJECT DATA Borrower: Government of Tanzania Executing Agencies: National Coconut Development Program (NCDP) Ministry of Agriculture & Livestock Development (MALD) Follow-on Project: Name: Cashew and Coconut Treecrops Project Credit Number 2050-TA Amount (US$ million) 25.1 Approval Date June 27, 1989 /a a-agronomist; or agriculturalisti e = Economist or Agricultural Economistg f-Financial analyst. Lb 1-Problem-free/or Minor Problemsi 2-Moderate Problems, 3-Major Problems. a 1-Improvings 2-Stationary 3-Deteriorating. /d Supervision Report not available on file. Le Coinciding with appraisal of Cashew and Coconut Treecrops Project. -xi - PROJECT PERFORMANCE AUDIT REPORT TANZANIA SECOND CASHEWNUT DEVELOPMENT PROJECT (CREDIT 801-TA) PYRETHRUM PROJECT (CREDIT 1007-TA) SMALLHOLDER TEA CONSOLIDATION PROJECT (CREDIT 1037-TA) COCONUT PILOT PROJEC7 (CREDIT 1070-TA) EVALUATION SUMMARY Introduction 1. The projects covered by this PPAR were the seventeenth, nineteenth, twentieth and twenty-first agricultural sector projects in Tanzania to be supported by the Bank/IDA. The four projects have been combined in one report because their different experiences and outcomes afford interesting and useful lessons of experience with important implications for future lending to Tanzania. Obiectives 2. All the projects aimed to improve the production and enhance the earnings from their respective crop sectors. The cashewnut and tea projects aimed to increase the processing capacity and quality of output, while the pyrethrum and coconut projects were directed toward increasing crop yields and production. The cashewnut, pyrethrum and tea projects were expected to increase the country's foreign exchange earnings, while the coconut project was intended to increase the supply of coconuts on the domestic market. Implementation Experience 3. The projects had very different implementation experiences. The cashewnut project got off to an excellent start ±Ln physical terms, thanks to bid evaluation for processing equipment and contract award for factory buildings in advance of credit negotiation, but ran into serious problems of a financial, managerial, political and technical nature in the course of implementation. The quality of implementation improved considerably in the seventh year, after a complete redesign of the project. The pyrethrum project encountered similar problems from the start and these problems persisted through to completion. The tea project experienced a slow start- up and was also beset by problems of a financial, managerial and technical nature but with the passage of time, including a three year extension, was able to overcome its difficulties. The coconut project rapidly got underway following an initial short delay in meeting a condition of - xii - effectiveness, and was efficiently implemented after due allowance is made for logistical problems and local funding constraints. Results 4. The cashewnut project, except for its redesigned research element, and pyrethrum project had most unsatisfactory outcomes with negative re- estimated rates of return. The tea project eventually overcame most of its problems and had a satisfactory outcome with a re-estimated economic rate of return of 28%. The coconut project satisfactorily attained its primary objective of establishing a strong, multi-disciplinary research program which represents a sound investment for the future. Sustainability 5. The cashewnut project as originally designed and the pyrethrum project are clearly unsustainable. The tea project is lik-Ay to be sustainable provided Government of Tanzania maintains the current improved pricing, foreign exchange and investment policies favorable to the Tanzania Tea Authority and the smallholder industry. The coconut project is likely to be sustainable provided the Government of Tanzania continues to encourage the further development of the small cadre of research scientists built up through the project. Findings and Lessons Learned 6. The cashewnut and pyrethrum projects constitute, through their overall lack of success, unfortunate additions to a past series of unsatisfactory Bank-assisted projects, which floundered due to a combination of inadequate preparation, over-optimistic appraisal expectations and weak implementation capacities and capabilities on the part of the responsible project agencies (paras. 52, 54-64). 7. The importance of timing of follow-on projects to enable experiences from earlier projects to be assimilated in the design of successor projects is illastrated by a comparison of the experiences of the cashewnut and tea projects (paras. 53, 65-67). 8. The importance of applied research in ensuring the attainment of project supported crop production objectives in a situation of poor agricultural standards, weak institutions, and an absence of improved locally proven technologies, is emphasized by the varied experiences of the cashewnut, pyrethrum and coconut projects (paras. 68-73). PROJECT PERFORMANCE AUDIT REPORT TANZANIA SECOND CASHEWNUT DEVELOPMENT PROJECT (CREDIT 801-TA) PYRETHRUM PROJECT (CREDIT 1007-TA) SMALLHOLDER TEA CONSOLIDATION PROJECT (CREDIT 1037-TA) COCONUT PILOT PROJECT (CREDIT 1070-TA) I. BACKGROUND Context 1. The Bank has been a strong supporter of Tanzania since its first credit in 1962, and particularly since the Arusha Declaration in 1967 when an accelerated program in pursuit of social equality, self-reliance, and the eradication of poverty and social transformation was introduced with an emphasis on rural development and public ownership of the principal means of production. In restructuring the political, economic and social life of the country the leadership introduced a series of far-reaching institutional reforms which centered on the nationalization of most large- scale units in manufacturing, finance, and wholesale trade. 2. Three of the four projects covered by this audit, all of which were prepared and appraised in the mid to late seventies, were intended to assist the parastatal organizations responsible for purchasing, processing and marketing of cashewnut, pyrethrum and tea, while the fourth aimed to strengthen the research capability and technical support to smallholders in the government agency responsible for coconut development. Obiectives 3. The cashewnu roject had the objective of increasing the country's raw nut procesing throughput by 30,000 ton/annum through the construction of three 10,000 ton factories. The pyrethrum project aimed to increase the output of dried pyrethrum flowers through provision of improved planting material, research, extension, marketing, roads and road transport services. The tea project had the primary objective of improving the quality of made tea and providinF additional processing facilities to cope with an expected increase in .aen leaf production. The primary objective of the coconut project was to rovide a sound basis for the development of the coconut industry, with rarticular emphasis on coconut production research. All the projects were expected to benefit smallholder producers through enhanced earnings. The cashewnut, pyrethrum and tea projects were expected to increase the country's loreign exchange earnings, - 2 - while the coconut project was intended to increase the supply of fresh coconuts on the domestic market. Des imn 4. The cashewnut project was proposed by the Government of Tanzania (GOT) as a 55,000 ton follow-on to an earlier raw nut processing and ancillary cashewnut research, extension and grading services project.1 Preparation was undertaken by GOT staff with World Bank technical assistance. At appraisal the additional processing capacity was reduced to 30,000 ton on the understanding that other sources of funds would be sought for the remaining 25,000 ton. The project was to be implemented, as was the first, by the Cashewnut Authority of Tanzania (CATA), a parastatal agency established in 1973 to conduct all activities connected with the cashewnut industry. 5. The pyrethrum project was prepared by GOT staff with assistance from Bank staff. The responsibility for the smallholder production component in the project as appraised rested with the Tanganyika Pyrethrum Board (TPB) which was to be responsible for producing and distributing improved planting material, extension advice, and operation of pyrethrum marketing centers. The Uyole Agricultural Center (UAC) in Mbeya Region was charged with strengthening its research program, particularly with a view to developing higher yielding plant varieties, while the roads component was entrusted to a new Road Improvement and Maintenance Unit (RIMU) under the control of the Ministry of Works. 6. The tea project as prepared by GOT staff was to be a further expansion of an earlier project designed to make good a shortfall in new estate plantings under the first project, and to further increase the green leaf processing capacity of the Tanzania Tea Authority (TTA). The project was reformulated by Bank staff in the course of an appraisal and a post- appraisal whereby the emphasis was given to consolidation of achievements under the first project, with accent on increasing production from existing areas, and improving the quality of made tea. Institutional strengthening was addressed as also were sector policy reforms as they affected smallholder green leaf prices, the domestic made tea market price, and ways and means for strengthening the financial viability of TTA. Road construction and maintenance was to be undertaken by the respective district roads authorities. 7. The coconut project, as designed with the assistance of FAO/CP staff, although presented in the Staff Appraisal Report (SAR) as a pilot development operation intended to complement an on-going Federal Republic of Germany (FRG) Coconut Development Program, was essentially an applied research operation with emphasis on agronomic trials, hybrid seed garden establishment, surveys and studies, which operated as an ancillary to the FRG operation. Both projects were designed to be jointly executed by 1/ Cashewnut Development Project (Loan 1014-TA), Project Performance Audit Report Number 5179 dated June 29, 1984. - 3 - Gesellechaft fUr Technische Zusammenarbeit (GTZ), the principal FRG agency responsible for technical cooperation, under the overall control of the Director, Crop Production, in the Ministry of Agriculture (Kilimo). Finance Plan 8. Project costs were as follows: Proiect Appraisal Actual IDA IDA Credit GOT NORAD Total Credit GOT NORAD Total ----------------------------$M-------------------------- Cashew 27.5 6.6 - 34.1 24.1 5.1 - 29.2 Pyrethrum 10.0 2.7 - 12.7 4.9 5.0 - 9.9 Tea 14.0 4.4 1.6/a 20.0 13.0 3.0 0.0 16.0 Coconut/b 6.8 1.7 - 8.5 6.1 2.0 - 8.1 Total 58.3 15.4 1.6 75.5 48.1 15.1 0.0 63.2 /a The NORAD financed component did not eventuate. /b In addition, approximately DM 20.0 M in parallel grant funds had been contributed to the GOT's National Coconut Development Program (NCDP) by the end of FRG's fourth phase project in mid-1990. 9. The lower actual total expenditures in US dollar terms as compared with appraisal expectations were due to progressive devaluation of the Tanzanian currency during implementation, and to reductions in project scope which affected the cashewnut, pyrethrum and tea projects. II. THE PROJECTS A. SECOND CASHEWNUT DEVELOPMENT PROJECT 10. Identification and Preparation. The cashewnut project was proposed in mid-1976 by Kilimo as an addition to the 40,000 ton processing capacity of the first Cashewnut Development Project, which had been approved by the Board in 1974. At the time construction work under the first project was going well but none of the five factories had been commissioned and no experience was therefore available regarding operating efficiency. Washington based Bank staff expressed reservations about proceeding with a second project before completion of the first, and before experience was available concerning the operating efficiency of at least one of the new factories. They also raised some reservations regarding the need for additional capacity at that time in view of a severe fall in raw nut production in 1974/75 and 1975/76 of around 30% from an all-time high of 145,000 ton in 1973/74. 11. Staff in the Resident Mission in East Africa (RMEA) in Nairobi, and in the Resident Mission in Dar Es Salaam on the other hand argued that preparation and appraisal should proceed swiftly because the factory processing techniques were clearly proven and because the adequacy of management at the five new factories under the first project would be assured considering that the equipment suppliers would be retained to operate the facilities for a specified period. The two year decline in raw nut production was explained as being a transient phenomenon due to bad weather and to temporary disruption caused by the GOT's villagization program.2 A further argument put forward by RMEA and the Resident Mission for accelerated processing (which suggests a strong element of lending pressure, real or assumed, and belies the Bank's role as a lender of last resort) was that any delay would disappoint GOT and raise the strong possibility that they would seek an alternate source of funding. With some reluctance, Bank Headquarters agreed to bring appraisal forward to March 1977. Preparation was assisted by Bank-seconded staff attached to Kilimo through the later half of 1976 and finalized in time for appraisal which eventually took place in June 1977. 12. Appraisal. The appraisal mission concluded that a 30,000 ton/annum additional processing capacity at three 10,000 ton/annum factories was appropriate, based on its finding that the 1974/75 and 1975/76 declines in CATA raw nut purchases to 118,000 and 106,000 ton respectively had been a temporary, weather related phenomenon which was already undergoing reversal. Increased production was assumed to be further assured by a massive five year new planting program at the rate of 30,000 ha/annum which the Cashewnut Authority of Tanzania (CATA) had initiated in 1977. Accordingly, the mission projected that the improvement from the 1975/76 low to 110,000 ton in 1976/77 would continue with a steady increase to 220,000 ton by 1992, and with 100,000 ton of this increase accounted for by the new planting program. Furthermore, the appraisal concluded that the additional 30,000 ton processing capacity was justified even without the new planting program, as the additional capacity, which would raise the national total to 113,000 ton/annum, would still be below the 120,000 ton/annum raw cashewnut production potential from the existing plantings. 13. The mission was more concerned with CATA's proposal to arrange a negotiated purchase of the processing equipment for the three factories from the supplier which had been awarded the equipment contract under the first project. The mission was successful in persuading CATA to procure the equipment through international competitive bidding from prequalified bidders and to evaluate the bids prior to credit negotiation in order to / A grouping of dispersed farm families into villages which was introduced as a key element of GOT strategy for rural areas in 1970. The program was considerably accelerated in 1974. - 5 - ensure reasonable accuracy in project cost estimates (factory costs in the first project had proved to be 3.6 times the SAR estimate). 14. Although the Bank officially endorsed the appraisal mission's proposals at the Decision Meeting, the support was far from unanimous. Doubts as to the economic viability of mechanized cashewnut processing in Tanzania worried several staff, one of whom suggested in a memorandum that the ERR was likely to be so low as to be unacceptable and wrote that "...it is just possible the project could be a disaster," while another called for a rigorous re-examination of the project proposal on account of its doubtful economics. 15. An appraisal follow-up mission was undertaken in February 1978 to resolve several outstanding issues ranging from production research through physical concerns about staff housing needs and factory sites, recruitment, training programs and management information systems, to financial projections and the weak state of CATA finances. Surprisingly, in view of the earlier concern over the decline in cashewnut production (paras. 10 and 12), the post-appraisal appears not to have pursued further the matter of cashewnut supply, which in fact was then continuing its decline to 69,000 ton for 1977/78 rather than increasing to 120,000 ton as forecast by the earlier appraisal. The green cover appraisal report was finalized shortly after the return of this mission, whereupon Central Projects Staff (CPS) raised strong objections to proceeding with the project at that time on account of what they viewed as the considerable technical, managerial, financial, and economic risks. They considered the project to be premature and recommended that further processing should be postponed, if necessary for a few years, pending full knowledge of factory performance under the first cashewnut development project. Despite these protestations project negotiations took place in April and the credit was approved by the Board in May, after two Board members had expressed some concern about absorptive capacities of Tanzanian parastatal institutions in general, and enquired as to the Bank's role in improving the performance of parastatals. 16. Start-up. Bank insistence on bid evaluation prior to credit negotiations for the cashewnut processing machinery tender, and GOT initiative in awarding civil works construction contracts for the factory buildings several months before the project became effective, facilitated an early start-up to the cashewnut project. 17. Sequence and Design Chanpes. All three cashewnut processing factories were satisfactorily completed within schedule before the end of 1980, bringing the total installed capacity up to 113,000 ton. However, as noted by the first supervision mission in September 1978, a month before effectiveness, the increase in processing capacity was offset by a further marked decline in production of raw cashewnute. In consequence, CATA's installed processing capacity exceeded actual raw cashewnut production levels even before the project factories were completed, and all three factories have operated only intermittently and for short periods since completion at extremely low levels of utilization. One of the factories - 6 - only operated for 116 days and closed down in April 1982. The economics of production at these levels is such that they operate at a loss, and CATA finds it more profitable to export raw rather than processed nuts. 18. The cause of the decline in raw cashewnut production was attributed in the early years of implementation to a range of factors including adverse weather, dislocation and abandonment of plots due to villagization, pests, labor shortages, unattractive farmgate prices, and delayed payment procedures by CATA to growers. By 1984 powdery mildew, a fungal disease which had been recognized as affecting cashewnuts since the 1960s, was seen as the principal cause of the decline, and remains so to the present time. 19. From mid 1982 Bank missions were rating the project as having serious problems of a financial, managerial, political and technical nature, and by mid-1983 the Bank was considering suspending disbursements because of failure on the part of CATA to submit audited accounts for 1980/81 and 1981/82, and because of the negative opinions of the auditors regarding several previous years' accounts. CATA senior staff appeared to lack the will or capacity to rectify the financial situation or to effect efficient management. Finally, in 1985, recognizing the seriousness of the decline in production the Bank proposed, and GOT accepted, a radical redesign of the project in the shape of a US$3.1 H Cashewnut Production Improvement Pilot Project (CPIPP) to be implemented by the Crop Production and Extension Service of the Ministry of Agriculture and Livestock Development and financed from savings in the unallocated and technical assistance disbursement categories of the original project credit agreement. At the same time CATA was transformed into the Tanzanian Cashewnut Marketing Board (TCMB). The closing date was extended by 3 years, and subsequently further extended by an additional 18 months, to June 30, 1989, to permit the implementation of this revised project. 20. The CPIPP, which was essentially an applied research, extension and training program with design features similar to the coconut project, initiated a baseline survey of cashewnut growers and a marketing study, both undertaken by consultants. It also undertook much useful investigative work into the powdery mildew problem and the promotion of sulphur dusting techniques for combating mildew damage on affected trees, and other simple agronomic improvement measures. B. PYRETHRUM PROJECT 21. Identification and Preparation. A pyrethrum improvement program was originally proposed in mid-1977 as a component for an EEC-financed rural development project for the Iringa Region, and again for a proposed Bank-financed national rural development project, but did not eventuate. Preparation of a freestanding pyrethrum project by a GOT/RMEA team began in earnest in April 1978 and was completed in March the following year, a month before appraisal. - 7 - 22. A declining international pyrethrum price in the face of severe competition from synthetic pyrethroide; TPB institutional weaknesses and processing inefficiencies; and a low GOT-controlled producer price for dried pyrethrum flowers, which made cultivation unattractive to farmers in comparison with maize in the growing areas, all made for a difficult preparation. This led one astute *'ank staff member to observe that the project as prepared, which aimed to reverse the decline in production from a low 1977/78 level of 2,600 ton back to the mid-sixties level of around 6,000 ton through the provision of better roads and improved TPB agricultaral services to growers (improved planting material, spraying, buying centers and research), could achieve the same result at less cost by the alternate expedient of simply increasing the producer price to a level comparable with that prevailing in neighboring Kenya. 23. Appraisal. The appraisal mission in April 1979 affirmed the project as prepared, and concluded that the synthetic pyrethroid threat should not be overrated. Nevertheless, the mission was concerned at the apparently weak and obscure financial state of TPB, and recommended the inclusion of a financial controller to strengthen the accounting procedures in a technical assistance team, which was also to include two agronomists and a research officer. 24. A follow-up mission in September to appraise the roads component led to the inclusion of three additional technical assistance posts for a roads engineer, a construction supervisor and a mechanical superintendent. An outstanding issue as to which agency would be responsible for this component was also resolved at that time, when the Ministry of Works agreed to be responsible. 25. At the SAR review stage details were sought by CPS about the new high-yielding pyrethrum clone, which was an essential element for the success of the project. On being advised by a member of the appraisal team that the clone (59/347) was fully tested and capable of giving yields of 570 kg/ha of dried flowers against a 1974 average yield of 270 kg/ha CPS gave its support to the project. Negotiations took place in February 1980 and the Board approved the project under special procedures a:.d without discussion in April. 26. Start-up. A six month gap between Board presentation, in April 1980, and credit signing, followed by a further six month gap between credit signing and effectiveness in April 1981 prevented the project from getting off to a good start. Several troubling events at about the same time, including the suspension of the entire 15 person Board of TPB in January 1981, at which time TPB was experiencing difficulty in paying farmers for their dried pyrethrum flowers, or its suppliers and contractors for goods supplied and work done on a new pyrethrum extraction plant at Mafinga which was to replace an old plant in Arusha, added to the start-up problems. Receipt of the 1977/78 audited accounts of TPB in April 1981, which showed a net deficit for that year of TSh 5.2 H as against an SAR estimate of TSh 1.2 M, and the knowledge that a sharp drop in the pyrethrum price had occurred in 1981, and that the financial situation of the Board had almost certainly worsened since 1977/78 was of great concern to the Bank. Bank supervision reports reveal virtually no progress in the first year of implementation and an acceptance of the fact that the project was in jeopardy. 27. Sequence and Design Changes. Implementation of the project was continuously plagued by major problems of a technical, financial, or managerial nature. By early 1982 a new extraction plant (which was not part of the project) had been commissioned, but it was operating well below its rated capacity and its production costs were far higher than the old plant, which had been decommissioned. TPD with its associated Tanganyika Extract Company was running at a combined annual loss of about TShl5.0 M, while its cash flow situation was poor and overdraft was getting steadily worse. Pyrethrum crude extract was costing around US$70/kg to produce but selling for about US$38/kg. The roads component, under the control of the Ministry of Transport, was seriously lagging, while Uyole Agricultural Center was making very slow progress with the research component. 28. By the time of the eighth supervision in May, 1984 production, at 1,500 ton, was at its lowest for 20 years and 60% below appraisal estimates; TPB's financial position was rapidly deteriorating with current losses estimated at TSh40 M/year, and was operating with negative working capital. Annual economic costs were estimated to exceed benefits by almost 50%. Producer prices were not attractive to growers in comparison with alternative crops such as maize and potato, and it became clear that at the then prevailing international market price for pyrethrum products its production, processing and marketing was not viable in Tanzania. The implication of these findings was that it would be inappropriate to continue the project as appraised. 29. Towards the end of 1984 the Bank, concerned at the unviable state of the industry and poor progress of the project, took up the matter of the project's future with GOT. A separate analysis of the situation conducted by the GOT's Marketing Development Bureau in early 1985 came to a similar conclusion as the Bank's earlier analysis (para. 28). After several months consideration of alternate strategies for possible recovery of the industry, GOT decided, essentially for political reasons, to continue supporting TPB in its purchasing, processing and marketing role while limiting future annual dried flower production to 1,600 ton. The Bank declined to provide further support for the project under those circumstances and, with Ministry of Finance agreement, closed the credit at the end of 19!5 with only US$4.9 M disbursed of the US$10.0 M total. C. SMALLHOLDER TEA CONSOLIDATION PROJECT 30. Identification and Preparation. The tea project as originally proposed and prepared by GOT was, like its predecessor3 in the early 3/ Smallholder Tea Development Project (Credit 287-TA), Project Performance Audit Report Number 2787, dated December 27, 1979. - 9 - seventies, designed to expand the area of tea and increase the processing capacity with new and enlarged factories. However, the Bank did not accept this expansionary proposal but persuaded GOT to revise its proposal in favor of consolidation with emphasis on increasing yields of green leaf from existing areas and improving the quality of made tea by concentrating attention on factory efficiency and substantial institutional strengthening of the TTA. 31. Appraisal. The appraisal of the tea project in July 1978 was only partly successful in reformulating the project from an expansionary mode, as proposed by GOT, to a consolidation mode as desired by the Bank. At first, GOT was unhappy with the appraisal mission's reduction in the new tea planting area from 2,629 ha to only 200 ha new planting and 630 ha of infilling in existing areas, and with the increased emphasis given by the Bank to efficiency improvements in the existing factories rather than new factory construction. The Bank, on the other hand, felt that insufficient progress had been made during appraisal in reforming the policy environment on the basis of which the heavily subsidized and overstaffed TTA could be strengthened financially and institutionally. A post-appraisal was therefore mounted in October 1979 with the express purpose of assisting GOT in developing policy options for salvaging the Tanzanian tea industry. 32. The second mission concluded that, in addition to several internal reforms designed to improve the efficiency of TTA, its financial viability required a downward adjustment in the producer price paid for green leaf, and an upward adjustment in the domestic sale price of made tea to reflect parity with world prices. 33. CPS made several constructive comments at the appraisal report review stage concerning management efficiency, TTA's financial viability, consultant inputs, etc., most of which were incorporated into the SAR text. Negotiations took place in April 1980 and Board approval, under special procedures and without discussion, followed in June. 34. Start-up. The tea project experienced a slow start-up, due to a postponement of 5 months in effectiveness caused by delay in concluding a subsidiary loan agreement with TTA. The first Bank supervision mission in November, 1980 reported that the project was already 9 months behind appraisal schedule, but that good groundwork was being laid under a new general manager. 35. Sequence and Design Changes. The tea project was beset by problems of a financial, managerial and technical nature which adversely affected progress through its early years of implementation, but, after a three year extension, the project was eventually executed in a generally satisfactory manner, although with several modifications from appraisal intentions. Much of the credit for overcoming the problems experienced during implementation is due to the quality of TTA senior management, technical assist-rce personnel, and Bank supervision staff, whose joint efforts were successful in attaining most of the project's physical - 10 - objectives in spite of excessive GOT controls and bureaucratic procedures. Many of the difficulties faced by TTA in the early years were of a financial nature which were beyond its power to control, but the successful introduction of covenants concerning adjustment of producer green leaf prices in relation to international made tea prices, in the first instance througb the expedient of holding the green leaf price constant ikn nominal terms; raising the domestic price of made tea to equal the export parity price; and transferring the costs of agricultural extension away from TTA to Kilimo; together with increased tea factory throughputs, upgrading in quality of output and general efficiency improvements within TTA, contributed to its turn-around from a loss making to a profitable enterprise by the closing year. 36. Much of the credit for this turn-around is attributable to the close attention given by RMEA based supervision missions to the financial health of TTA, commencing in the third year of implementation, which had a markedly beneficial effect on the eventual project outcome. The sixth supervisfon mission in September 1983 highlighted TTA's extreme financial difficulties, in spite of improvements at the operating level, as a result not only of an inflated exchange rate but particularly because of a lack of equity capital stemming from past losses, and deficient financial support by GOT. Subsequent supervision missions referred repeatedly to the need for an equity capital contribution by GOT. Progressive injections of new capital by GOT in the subsequent four years, following strong representations by the Bank, combined with further improvements in operational efficiency, eventually led to the writing off of TTA's accumulated losses in the 1987-88 accounts and the dawn of a new era of profitable activity by the time of credit closing in June, 1989. D. COCONUT PILOT PROJECT 37. Identification and Preparation. A Bank consultant conducting a study in early 1978 for the revival of the Tanzanian coconut industry identified an expansionary project designed to introduce high yielding West African developed exotic hybrid coconut varieties and, after a short period of seed garden establishment and adaptability trials, to gradually develop 60,000 ha of nucleus estate and smallholder plantings, all under the direction of an autonomous coconut development agency modelled on SODEPALI, a State-owned plantation crops company in the Ivory Coast. GOT reacted favorably to this proposal and urged the Bank to proceed speedily with preparation with a view to commencing a first stage IDA-assisted development project by mid 1979. However, the Bank considered the consultant's proposal too ambitious and risky under the circumstances and mounted a joint Bank-FAO/CP identification mission to reassess the situation. In its report the mission expressed alarm at the presence of an undiagnosed disease problem affecting coconuts in many parts of Tanzania and concern at the moisture stress to which West African hybrid coconuts would be exposed under the more severe dry season conditions of Tanzania. The mission recommended a cautious approach to the introduction of hybrid planting material, the performance of which was unproven under Tanzanian - 11 - conditions. It urged the Bank to support an existing Federal Republic of Germany (FRG) bilateral assistance applied research, extension and training project managed by GTZ which emphasized hybrid planting material development using local as well as exotic parent varieties, agronomic trials, and disease and pest control investigations before investing in any substantial commercial planting operations. 38. FAO/CP subsequently prepared a project, which was described as a pilot project to distinguish it from the earlier, large scale commercial production proposal. 39. Appraisal. Disappointment on the part of GOT at the scaled-down concept for the coconut pilot project as prepared by FAO/CP prompted the Bank to mount a pre-appraisal mission in June 1979. In addition to refining the investigatory themes for what in reality now constituted an applied research, extension and training project, the mission was able to reformulate the project to supplement the ongoing FRG assistance project with its emphasis on disease and pest control, extension and technical services, with Bank-supported components for agronomy, breeding and seed multiplication, and plantation management. 40. Appraisal followed in November 1S79 at which time agreement in principle was reached between GOT, f'G and the Bank regarding their respective roles in building a strcng agro-technical base for rehabilitation of the Tanzanian coconut industry under the aegis of GOT's National Coconut Development Program (CIP). It was also agreed that Gesellschaft fUr Technische Zusammenarbeit (GTZ), the FRG's te hnical cooperation agency, should manage tha t-o projects, in order to ensure their integration at the operational leve). 41. CPS consideted the project to bS well conceived and prepared. Negotiation3 took place in Dar Es Salaam In July 1980 and the project was approved by the Board, under special proce-rs and without discussion, in October. 42. Start-up. Effectiveness of the coconut project was delayed for 3 months due to difficulty in meeting a concition of effectiveness which called for the selection of the seed garden and trial sites. Once this was done, in April, 1981 start-up activities rapidly got underway, thanks in large part to the early appointment of the credit-financed coconut agronomist and the support of the GTZ persc...el financed under the BMZ grant who were already in plEce. 43. Sequence and Design Changes. The project was efficiently implemented under GTZ management, after due allowance is made for logistical problems and local funding constraints. The project, which was an integral part of the NCDP, was given high priority by GOT because of concern for the reha"ilitation of the coconut industry after a period of neglect t'irough the previous decade. Accordingly, the project was favored with proportionally high allocations of scarce local budget funds and - 12 - qualified counterpart staff. The technical assistance funded through the credit was especially effective in establishing and maintaining a wide range of varietal and agronomic trials as well as developing seed gardens as a source of improved, high yielding planting material. 44. However, the commercial development aspects of the project received a serious setback in the early years, when the imported West African Tall X Malaysian Dwarf hybrid material proved susceptible to the destructive lethal disease which is widespread over much of the Tanzanian coconv.t area. While not entirely unexpected, this finding underlined the importance for the future of the industry of the more time-consuming task of selecting, testing and multiplying improved local (East African tall) varieties and developing improved hybrids through crossing the local, more disease tolerant varieties with exotic dwarf materials. 45. The project henceforth concentrated on the applied research, extension and training objectives, coupled with a cautious release of locally produced improved seed nuts to farmers, which, although below appraisal expectations, nevertheless reached 500,000 in 1988, the last full implementation year, following a three year extension of the credit closing date. III. PROJECT OUTCOMES 46. The outcomes were unsatisfactory in the cashewnut and pyrethrum while satisfactory in the tea and coconut projects. 47. The cashewnut project as approved by the Board was an abysmal failure on account of the fact that the three factories at completion provided an unneeded additional processing capacity and can serve no other useful purpose. The re-estimated economic rate of return is undoubtedly negative. However, the redesigned CPIPP introduced after the failure of the original project and financed by US$3.1 M of residual credit funds, was successful in initiating useful research into disease tolerance and vegetative propagation of improved cashewnut varieties, vhich has laid the groundwork for a follow-on cashewnut (and coconut) research and development project designed to assist the rehabilitation of the Tanzanian cashewnut industry over the long term through further research, extension and a pilot credit scheme. 48. The pyrethrum project failed miserably in that it was completely unsuccessful in its objective of increasing the yield or production of pyrethrum; on the contrary, both yield and production appear at best to have remained stable, or have declined, through the '80s. The re-estimated economic rate of return can be assumed to be negative. 49. Despite serious implementation problems, which at one stage were so severe that the Bank threatened to suspend disbursements pending equity - 13 - refinancing of TTA by GOT, and even briefly considered cancellation on account of inadequate borrower performance, the tea project underwent a heartening improvement in its closing years, and made a profit in the final (1988-89) implementation year. The eventual outcome is undoubtedly satisfactory, and the re-estimated economic rate of return of 28% presented in the PCR appears realistic. 50. The coconut project satisfactorily attained its primary objective of assisting in the establishment of a strong, multi-disciplinary, coconut production research program staffed with a critical mass of well motivated, trained and experienced scientists (local and foreign) and technical support staff. Although the high appraisal expectations of a ready adaptation and rapid commercial planting of the exotic hybrid materials were not met, the project contributed substantially to the NCDP through its financing of coconut seed gardens, hybrid seed production facilities, field trials, and its support of consultant services for conduct of the research program and the training of local staff. The success of the jointly funded efforts of GOT, FRG and the Bank are evident from the achievements of the hybrid coconut seed production unit and the quality of the applied research work carried out to date, particularly into lethal disease, coreid bug control, and coconut rehabilitation and maintenance techniques. A six year follow-on project4 also jointly assisted by FRG and the Bank should further consolidate the achievements of this first project through continuing support for research and extension, and larger scale multiplication and distribution of improved planting materials. IV. FINDINGS AND ISSUES Overview 51. The contrasting fortunes of these four audited projects afford useful lessons of experience given their markedly different implementation circumstances and outcomes. The two unsatisfactory projects were inadequately prepared and appraised, especially in respect to their agricultural production parameters and to judgments regarding the capabilities and capacities of the parastatal agencies responsible for implementation, which were overoptimistic. On the other hand, the two satisfactory projects display evidence of more careful preparation and appraisal, which reflects well on the Bank staff and consultants responsible for those aspects of the cycle. Related Experience 52. OED has now evaluated 22 Bank/IDA financed agriculture and rural development operations in Tanzania, of which 17 have been judged 4/ Cashewnut and Coconut Treecrops Project (Credit 2050-TA), SAR No. 7681- TA, May 30, 1989. - 14 - unsatisfactory. Thirteen, including the cashewnut and pyrethrum projects, were found to have negative rates of return. These two projects therefore perpetuate the poor record of experience established through past lending. A recent country study by OED concluded that Tanzania could have been better off with less Bank lending.5 Pressure to lend on the part of the Bank, and to borrow by GOT often led to poorly prepared, complex or overambitious projects which overtaxed the absorptive capacity of the agencies responsible for their implementation, resulting in highly unsatisfactory outcomes, as was the case with the cashewnut and pyrethrum projects. 53. On the other hand, the tea project, which was a follow-on project which benefited from the lessons of experience of an earlier project, and whose presentation to the Board was delayed to permit thorough preparation and appraisal represents a considerable improvement on the level of previous Bank involvement with Tanzania. Similarly, the coconut project (and the redesigned cashewnut pilot component which was modelled on the coconut project), with its emphasis on strengthening the applied research capability as a precursor for large scale crop rehabilitation or expansionary planting of locally improved materials was entirely appropriate as a contribution to the long term development of the Tanzanian coconut sector. Issues Arising A. Preparation and Appraisal Inadequacies 54. The inability of the Bank and GOT to appreciate the seriousness, and significance to the project, of the decline in cashewnut production in the years immediately preceding credit negotiation is regrettable, and points toward imperfections in the preparation and appraisal processes as they operated in the case of the cashewnut project. The circumstances deserve close analysis for the possible lessons which might be drawn from this experience. 55. As previously noted (paras. 10 and 12) CATA raw nut purchases (which are assumed to closely mirror actual production as CATA was the sole purchasing agent, and hand processed local kernel production is minimal) for the years preceding appraisal as accepted by the appraisal and post- appraisal missions, showed a decline from a 1973/74 high of 145,000 ton to 118,000 ton in 1974/75 and 106,000 ton in 1975/76, and an improvement to 110,000 ton in 1976/77 (SAR Table 1), from which levels they were projected to increase gradually to 220,000 ton by 1992 (SAR para. 2.13). However, the above figures for 1975/76 and 1976/77 presented in the SAR were exaggerated and the estimates for subsequent years unrealistically high. Actual purchases were below 84,000 ton in 1975/76 and 98,000 ton in 1976/77 and had slumped further to around 69,000 ton by 1977/78 (Table 1). Records 5/ Tanzania: World Bank/Tanzania Relations, 1961-1987, Report No. 8329, January, 1990. - 15 - of CATA purchases, which were regularly updated through the year on the basis of returns from the local purchasing offices, should have been made available to the missions, which mistakenly appear to have accepted highly inflated figures for those crucial years. 56. The figures presented in the SAR are contradicted by other statistics available from GOT sources at about that time. For example, a Tanzania Investment Bank analysis6 of a "Proposal for a Second Phase Cashewnut Development Project" dated February, 1977 (four months before appraisal) contains a statement that Kilimo was reporting a 1975/76 cashewnut output of only 85,000 ton; while a Bank economic report published in December, 19777 (two months before the follow-up appraisal) prepared following an August 1976 economic mission to Tanzania quotes a figure of 80,300 ton for marketed production of cashewnuts in 1975/76, based on Ministry of Finance and Planning sources and data provided by the Bureau of Statistics. 57. Obviously, more accurate cashewnut purchase/production figures than those presented in the SAR for the years immediately preceding appraisal were available from several sources, but were either overlooked or rejected by the missions in favor of figures provided by CATA. Enquiries during a visit to CATA headquarters in Mtwara in June, 1990 leads the audit to conclude that the figures accepted by the appraisal missions were derived from budget estimates prepared in the previous years rather than actual purchase figures compiled during or at the end of the years in question. 58. Given the crucial importance of accurate production figures for the immediate past years, the appraisal missions can be criticized for lack of diligence in checking the validity of their production statistics and comparing them with other credible data sources. 59. Again, in projecting future production after 1977/78 the appraisal missions appear to have accepted uncritically CATA's plans for planting an additional area of 150,000 ha of cashewnut in the five year period 1977/78- 1981/82, on the basis of which Tanzanian raw nut production was expected to rise from an assumed 120,000 ton annual average to 220,000 ton, and thereby further ensure that production would comfortably exceeded total installed processing capacity after the project factories were commissioned. Furthermore, the missions appear to have failed to assess the quality of the 8,200 ha planting program reported to have been carried out in the 1976/77 season. A review of CATA's past and current financial records up to 1977/78 and budget proposals for 1978/79 should have revealed the improbability of CATA being able to finance a planting program of the size 6/ Bank Archives, Document No. 101,589. 7/ Tanzania Basic Economic Report, No. 1616-TA, December, 1977, Annex VI, Table 3. - 16 - envisaged in the time contemplated. In the event, the new planting program was inadequately funded and proved to be completely ineffective. 60. Similarly, in th case of the pyrethrum project the appraisal mission concluded that an improved clone, 57/347, described as fully tested, (SAR para. 8.01) and capable of giving dried flower yields of 570 kg/ha on well maintained farms could be relied upon as the basis for an increase in average yields from around 200 kg/ha before the project to 300 kg/ha by 1989/90. In fact, as the audit learned from discussions with TPB and UAC staff in June, 1990, clone 57/347, which was recognized as a promising candidate for the highest elevations as early as 1957, was either lost through mislabeling, or had otherwise deteriorated to a stage where its high yield characteristics had been suppressed around the time the research program was transferred from Igeri, its originating experimental station, to lyole in the early to mid-1970s when the former station was closed. This move occurred several years prior to appraisal. Neither clone 57/347 nor any other selected materials distributed to farmers through the implementation period, or since, has been successful in improving average annual yields; which remained stationary at or below 200 kg/ha through the 1980s according to official statistics from the Ministry of Agriculture (Table 2). 61. In this instance failure to confirm the true status of the reputed high yielding material led to false expectations as to likely project impact, and contributed to the project's unsatisfactory outcome, notwithstanding the weak international market for pyrethrum through the implementation years. A better understanding by both the preparation and appraisal missions of the limits to yield improvement afforded by clonal material available at the time could have resulted in a different, more realistically designed project with a better chance of success, or possibly no project at all. 62. Bank files show that many Bank staff expressed reservations about the cashewnut and pyrethrum projects during the prenegotiation phases. Most importantly, the cashewnut project was criticized by CPS as premature and better postponed, if necessary for a few years, until CATA could be shown to be capable of efficiently operating the five factories being financed under the first cashewnut development project (para. 15). Had the correct purchase/production figures for 1975/76, 1976/77 and the first eight months of 1977/78 been revealed to Bank staff reviewing the SAR draft prior to credit negotiation it is highly likely that the argument for postponement would have won the day and led to a very different project, probably on the lines of the CPIPP. 63. The experiences with these two projects underscores the extreme importance for the Bank of ensuring that Bank staff engaged in preparation and appraisal of agricultural projects undertake thorough critical reviews, and cross-checks whenever possible, of all relevant crop production data. Reservations expressed by CPS and others were too easily brushed aside by Bank management in both these instances, most likely out of concern for - 17 - project processing schedules. Neglect on the part of appraisal mission staff to perform adequately their tasks detracted from the integrity of the appraisal process, with unfortunate consequences in respect of project outcomes. 64. Borrowers could also learn from these experiences, for the inaccurate information provided to Bank staff came from staff of the GOT institutions which were to benefit from the projects. Central authorities on the borrower side responsible for handling donor-assisted projects share with Bank management the final responsibility for ensuring that projects are prepared on the basis of the most current and reliable information. B. Timing of Follow-on Projects 65. The importance of timing for successor projects, to enable experiences from earlier projects in a series to be fully assimilated in the design of the follow-on projects, has been referred to frequently in OED reports.8 The contrasting experiences of the cashewnut and tea projects illustrate respectively the dnngers of premature processing and the advantages of prudent delay until such time as the lessons from an earlier project can be applied to its successor. 66. GOT and Bank anxiety to proceed with appraisal of the cashewnut project even before the factories financed under the earlier project had been commissioned is well documented (paras. 10-11). In fact, the two projects overlapped by 41 months, between appraisal of the second in July, 1977 and physical completion of the first in December, 1980. If the exhortations of CPS and other Bank staff to delay project appraisal, if necessary for a few years, had been heeded, not only might the project have been better designed to take account of inefficiencies experienced in the operating of the factories financed under the first project, which were enormous, but with the passage of time the seriousness of the decline in raw nut production would have become so obvious as to have demolished any argument favoring an increase in processing capacity. 67. On the other hand, the Bank proved much more circumspect in its approach to the tea project. GOT's initial proposal, in mid 1976, some three and a half years before the credit closed on the first project, was basically for an expansionary project intended to make good on the considerable shortfalls in terms of planted area and factory processing capacity expected under the first (para. 6). In this case, however, the Bank withheld appraisal for almost two years (until July, 1978) pending reformulation of the GOT proposal in the direction of institutional strengthening and tea sector policy changes designed to maximize the g/ See, for example, Korea: Second Integrated Dairy Development Project (Loan 1193-KO), PPAR Number 4270, dated December 30, 1982; Niger: First Forestry Project (Credit 800-NIR), PPAR Number 4569, dated June 21, 1983; and Indonesia: Nucleus Estate and Smallholders I, II and III Projects (Loans 1499-IND, 1604-IND and 1751-IND), PPAR Number 7955, dated June 30, 1989. - 18 - returns on investments to TTA made under the first project. Then, when the appraisal proved unable to define a project which would ensure the long term financial viability of TTA without subsidization from GOT, the Bank carried out a post-appraisal to achieve that purpose. The post-appraisal, in October 1979, took place six months after the PCR for the first project had been finalized, thus enabling its lessons of experience to be fully assimilated. C. The Importance of Applied Research to Development 68. The cashewnut, pyrethrum and coconut projects, each in its different way, illustrate the importance of applied research to the attainment of project supported crop production objectives, and the effectiveness or otherwise with which the Bank and GOT dealt with this matter. 69. The coconut project affords a particularly striking example of how locally researched improved production technology can impact favorably on a traditional low input - low output agricultural subsector. The Bank wisely rejected a Bank financed consultant's proposal for an ambitious new planting project using West African developed hybrids, which was at first strongly favored by GOT, and instead accepted the FAO/CP prepared project in support of an ongoing FRG assisted applied research, extension and training project, described as a pilot phase for the GOT's NCDP (paras. 37- 41). The decision to support a long-term program involving surveys, fact finding, identification of constraints, and development of advanced technologies to overcome local production problems, as a precursor to an eventual large scale rehabilitation and new areas development program, was an appropriate investment for an assistance agency concerned for the future of the Tanzanian coconut industry, given the neglected condition of the subsector and the limited institutional strengths and locally proven technological base existing at the time of appraisal. Although no immediate impact on production was foreseen, the benefits from this approach are now being felt in terms of higher husbandry standards and increased yields on the part of those farmers who have adopted the new technology developed through the project, to their considerable financial advantage. These benefits are likely to spread rapidly in the near future, particularly as the work initiated under this project is being continued under the follow-on Cashewnut and Coconut Treecrops Project (para. 50). 70. The initial research experience of the cashewnut project contrasted unhappily with the coconut project. The Bank readily acquiesced (too readily as it turned out) to a GOT request for an early follow-on project (para. 11) which, like its predecessor, aimed primarily to expand the raw nut processing capacity, but nevertheless provided a substantial sum of about US$1.0 M for cashewnut production research intended to support 16 staff years of technical assistance in the third through sixth years of project implementation (i.e. from 1980), the details of which were not defined in the SAR, in succession to an Italian technical assistance grant aided program. GOT was at first reluctant to use credit funds for engaging - 19 - research consultants and arranged for continuation of the Italian assistance through to 1985, during which time the widespread and damaging impact of the powdery mildew disease on production was recognized, the causative agent isolated, sulphur dusting control techniques demonstrated, and several disease tolerant varieties identified. However, by 1984, about the time that the Italian program was winding down, the need for more substantial research effort to tackle the problems associated with the steep decline in raw nut production of the preceding years became more widely recognized. Accordingly, the unsatisfactory cashewnut project as appraised was closed in 1985 and changes were made to the credit agreement to permit the implementation of the CPIPP (paras. 19-20). 71. The CPIPP was modelled on the Coconut project (para. 69) although more modest in content and adjusted to the organizational structure of the Ministry of Agriculture. It proved to be an appropriate intervention. It consolidated much of the earlier work of the Italian team, developed successful 1-idding and grafting techniques which considerably enlarges the scope for propagating disease tolerant varieties, and introduced, through extension, simple improvement techniques including motorized sulphur dusting for rehabilitation of abandoned cashewnut holdings. The follow-on Cashewnut and Coconut Treecrops Project is further expanding on this work which should be of great benefit to responsive farmers over the medium term. 72. Finally, the experience of the research component in the pyrethrum project provides a worst case scenario. The objective was to support pyrethrum improvement research through the construction and equipping of a tissue culture unit, provision of an internationally recruited researcher, and study tours and training for local staff. This objective would have beer entirely appropriate had the world market outlook for pyrethrum remained as buoyant as was assumed at appraisal, and had the returns to growers increased as projected through the rapid propagation and widespread distribution of the (supposedly) fully tested higher yielding clone 59/347 (para. 60) which would have improved the comparative advantage of pyrethrum against competing crops. However, this was not to be, while bureaucratic delays and inefficiencies common to many parastatals in Tanzania, much of which can be attributed to the general deterioration in Tanzania's economic environment during the implementation years, held back progress to the extent that little in the way of research had been achieved by the end of the fifth year, when the Bank declined to extend credit beyond the original closing date on the grounds of overall unsatisfactory implementation progress. 73. The audit sees a direct relationship between satisfactory and unsatisfactory outcomes and the different emphasis given to applied production research in the case of these three agricultural projects. In situations where agricultural standards are poor, institutions weak and locally proven improved technologies are absent, investments in large scale crop rehabilitation or expansionary programs can be extremely risky. The risk can be considerably reduced only if the investments are preceded or at - 20 - least accompanied by investments in high quality applied research capable of generating the requisite technologies. The coconut project has been particularly satisfactory in this regard; the cashewnut project, as redesigned on the pattern of the coconut project, has established, belatedly, the basis for a similar development; while the pyrethrum project was unable to make any favorable contribution in that direction. - 21 - Table 1 RAW CASHEWNUT PURCHASES BY CATA/TCMB La Year b Metric Ton 1965/66 73,000 1966/67 84,000 1967/68 73,000 1968/69 118,000 1969/70 110,000 1970/71 112,000 1971/72 126,000 1972/73 126,000 1973/74 145,080 1974/75 118,947 1975/76 83,755 1976/77 97,645 1977/78 69,288 1978/79 57,028 1979/80 41,376 1980/81 56,558 1981/82 44,419 1982/83 32,966 1983/84 48,307 1984/85 32,073 1985/86 18,802 1986/87 16,548 1987/88 24,187 1988/89 18,934 1989/90 16,473 /c /a TCMB figures as at June, 1990. At July 1 to June 30. /c up to June 9, 1990. - 22 - Table 2 PYRETHRUM AREA, PRODUCTION AND YIELD /a YEAR AREA PRODUCTION YIELD (July-June) ('000 ha) (Dried flowers '000 mt) (Crude Extract, KeIha) 1979/79 8.05 1.60 198 1979/80 n.a. 1.62 n.a. 1980/81 n.a. 2.0 n.a. 1981/82 10.00 1.9 190 1982/83 9.00 1.6 178 1983/84 8.00 1.44 180 1984/85 8.00 1.53 191 1985/86 7.50 1.35 180 1986/87 7.73 1.23 159 1987/88 7.78 1.41 181 1988/89 8.40 1.31 156 1989/90 /a Source: Basic Data, Agriculture and Livestock Sector, Table 5.17, Planning and Marketing Division, Kilimo. UGANDA - -- KENYA TANZANIA 32- i B/- 1i l CC AcA SECOND CASHEW NUT DEVELOPMENT PROJECT Bukobo uaa Trm MTWARA Processe cnters R~~f WANDAcOries I~rOssing c a itisi tons.), RWWD A102.z500) Inoeration UJnder corstructionl or plonned m W Bónk or IDA oaancm 'Ngoras m - MN Cosew growing brems <E0z3/5 0 Anerose onnoe o ro. nts P,,rha.sed ro nar hr CATA Bi romel dunn. thep od 1970/71-1976/77, by di.trict Gek Mriy dugurd, ++ + AN-2AMriwy Dsctbound-res I U Shinyongo S " AVIA 0 N Nugo KimbalR KIENYA /l.Ndole K gomaTQbora Singida L UKondbo a e TongPEMBA Klottrosni D~dm ZANZ1ßAR Mganda \mý,w m N DGul BA Rungwo Kilosa aagr R e5 SALAAM e a o12,500) ZAIRE -8* Sebowong -ing- i nci ýd Chunye CIGHi Mohenge 1 "ll- Kl o s~k ukull Ndelaong Lupernbe Tuntd.m)mb zA A - 3 -. t&ngi .frf I D i MTWA NAC NtWEA *5 .*,0,000 2ýom'.5 ) '(5"On00 m Mboffbo Boy A LAWI -- 0 z A m e 1o uE APRILt 1990 徊口口口口口口 _｀一’ IBRI? 146]4R2 JGANDA E r, Y A K N Y A uRuND)1 TANZANIA D. I Z A l R E mos l zAmBIA vo. mozAmelauE TANZANIA MOMBASA COCONUT PILOT PROJECT Moln Romods ARjSI-lå Secondory Roads Riw$ o TA + Radroads " NGA %). 108'.alion.l Airpoth P-ga.- Scheduled SeNice Airports C, A Region Boundaries Inlernalionot Bovndarjes Z-b. r7 - Chornbc.i lh.hue Zg-p' Kundmhi + Da, os Sola.. ROGORO DAR ES SÄ! AMVI KILOMETERS ý0 1?0 160 io b NYLES K,ka KNiýje 11..ale + 4 i nd i Mtwam + Tuoduru ýV A 76,...ph-b- -d i, ký 0. &d -d d. ~~ F- c~ MOZAMBIQUE F- 4o- APRIL 1990 MAP SECTION