Turkey - Third Livestock Development Project

EW=uMMa of The World Bank FOR OFFICIAL USE ONLY Report No. 6559 PROJECT PERFORMANCE AUDIT REPORT TITRKFY THIRD LTV!TOCK DEVELOPMENT PR0J'FT (LOAN 1265-T11) December 31, 1986 OrwrainicY va1-tion Don-r-.nt This document has a restricted distrbution and may be used by recipients only in the perfoirmance of their official duties. Its contents may not otherwise be disclosed without World Fiank euthoriation. FO OK IML ONLY THE WORLD BANK Washngton. D.C. 20433 U.S.A. Otk of Dwcto-Geweal Opweaium tvaluation December 31, 1986 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on Turkey: Third Livestock Development Project (Loan 1265-TU) Attached, for information, is a copy of the report entitled "Project Performance Audit Report - Turkey Third Livestock Development Project (Loan 1265-TU)" prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their ofcial duties. Its contents may not otherwise be disclosed without World Bank authorization. VOM OnVIwu ta ONLY ABBREVIATIONS DVS - Directorate of Veterinary Services EBK - State Meat Organization EDLP - Encouragement and Development Loans Department, TCZB ERR - Economic Rate of Return GDP - Gross Domestic Product IDA - International Development Association IDPD - Intensive Dairy Production Division of LDP LDP - General Directorate for Livestock Development Projects, MAFRA MAFRA - Ministry of Agriculture, Forestry and Rural Affairs OED - Operations Evaluation Department PCR - Project Completion Report PPAM - Project Performance Audit Meorandum TCZB - Agricultural Bank of Turkey TSEK - Milk Corporation 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 TURKEY THIRD LIVESTOCK DEVELOPMENT PROJECT (LOAN 1265-TU) TABLE OF CONTENTS Page No. Preface.............................. i Basic Data Sheet .......................... it Evaluation Summary ....................................... iii PROJECT PERFORMANCE AUDIT MEMORANDUM I* SUMMARY ................................................. 1 Background ....................................... . 1 Project Design .................... ... 2 Prcjeclz Revision ........... ...... 3 Project Implementation .............................. 4 Project Dutcome ..................... 7 II. MAIN ISSUES ........... ................ . 8 A. Impact of Structural Adjustment Process on the Project... 8 B. Rate of Return .......................................... 10 C. Alternative Strategies which both Increase Livestock Productivity and reach Small Farmers .... .. 12 Appendix: Comments received from Government .............. 15 PROJECT COMPLETION REPORT 1. Introduction ....................................... 23 II. Project Formulation and Processing ...................... 24 III. Project Implementation .................................. 26 IV. Institutional Performanceo............................. 36 V. Operating Performance and Project Impact................. 43 VI. Financial and Economic Resulte........................... 48 VII. Bank's Performance . . . . . . .......... 49 VIII. Conclusions and Recommendations ......................... 50 Annexes Map: IBRD 12037R1 PROJECT PERFORMANCE AUDIT REPORT TURKEY THIRD LIVESTOCK DEVELOPMENT PROJECT (LOAN 1265-TU) PREFACE * This is a performance audit of the Third Livestock Development Project in Turkey, for which Loan 1265-TU was approved on May 25, 1976 in the amount of US$21.5 million and closed on March 31, 1984. Final disbursement took place on November 19, 1984. The audit report consists of an audit memorandum (PPAM) prepared by the Operations Evaluation Department (OED) and a project completion report (PCR) dated November 19, 1985. The PCR was prepared by the Europe, Middle East and North Africa Regional Office following a mission to Turkey in May 1985 and is based upon a study prepared by the General Directorate of Project and Implementation (GDPI), Ministry of Agriculture, Forestry and Rural Affairs. The PPAM is based upon a review of the Staff Appraisal Report (No. 1027-TU), dated May 10, 1976, the President's Report (No. P-1847-TU) dated May 12, 1976, the Loan and Project Agreements dated May 26, 1976, and the PCR dated November 6, 1985. Correspondence with the Borrower and internal Bank memoranda on project issues as contained in relevant Bank files were consulted and Bank staff associated with the pruject were interviewed. An OED mission visited Turkey in November 1985. Discussions were held with officials of GDPI, TCZB in Ankara and in the field, of provincial offices and of the project. A field trip to visit the project area was undertaken, a series of visits was made to producers and information was directly obtained to test the validity of the conclusions of the PCR. The audit finds that the PCR covers the project's salient features and it supports the PCR's mair. findings. In addition to summarizing the project's objectives and results, the PPAM expands upon the effects which the Bank-supported structural adjustment process had had on this project; questions the assumptions -underlying the rate of return as recalculated by the PCR; and reviews briefly alternative approaches to increasing livestock productivity which both address constraints in the subsector and contribute to raising incomes of small farmers and the landless poor. The draft of the report was sent to government on April 30, 1986 for comments; these are attached in the Appendix and have been reflected in the final report. The valuable assistance rendered by government of Turkey and the project staff met during the preparation of this report is gratefully acknow- ledged. 7 it - PROJECT PERPONANCE AUDIT REPORT TURKEY THIRD LIVESTOCK DEVELOPMENT PROJECT (LOAN 1265-rU) BASIC DATA SKEST KEY PROJECT DATA Appraisal Actual or Actual as I of Estimate Estimated Actual Appraisal Estimate Total Project Cost (USS million) 34.7 34.0 98.0 Total Project Cost (TL million) 520.3 4,978.2 857.0 Loan Amount (US$ million) 21.1 21.5 '00.0 Date of Board Approval 05/04/76 05/25/76 Date of Loan Agreement 05/26/76 05/26/76 Date of Loan Effectiveness 09/27/76 02/25/77 Closing Data 03/31/82 03/31/84 Economic Rat of Retirn (2) 22.0 Likely to be tl low 10%* Institutional Performance Less than adequate Number of Beneficiaries (jobe created) 1,200 not estimated STAPP INPUT FY76 FY77 FY78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 Identification/Preparation 6.1 - - - - - - - - Nego'mtion 10.5 - - - - - - - - - Appaiss) 73.9 - - - - - - - - - Supervision 2.7 13.1 12.8 10.2 16.1 28.0 34.2 14,3 8.3 10.0 TOTAL 93.2 13.1 12.8 10.2 16.1 28.0 14.2 14.3 8.3 10.0 CUNUIATIVE DISBURSEMENTS FY77 FY78 FY19 FY80 FY81 ?Y82 VY83 FY84 Appraisal Estimate (USS millt-n) 3.1 7.1 11.4 16.7 21.5 - - - Actual (US$ million) - 0.4 3.8 7.7 11.0 13.9 16.2 21.5 Actual as % of Estimate (%) 0 6 33 46 5t 65 75 100 Date of Final Disbursement November 19, 1984 MISSION DATA Date No. of Specializations Periormance Types of Mission (Mo./Yr.) Persona Represented/& Rating/b Trend/c Problems/d Identification 08/72 1 c - - - Identification 10/73 1 d - - Preparation 05/75 1 d - - - Preparation (FAO/CP) 08/75 2 d,c - - Appraisal 1W/75 4 a,b,c,d - - - Supervision I 07/76 2 bd I 1 - Supervision II 04/77 3 b,d 1 M Supervision III 11/77 3 b,d 2 2 M,T Supervision IV 07/78 I d 2 1 M,T Supervision V 11/28 1 d I I M,T Supervision VI 08/79 3 b,d 2 1 M,T Supervisi)n VII 07/80 2 - 3 3 F,T,M Supervision VIII 11/80 1 d full supervision report not prepared Supervision IX 05/81 2 d 3 3 M,T,0 Supervision X 11/81 5 b,d 2 1 M,T,0 Supervision XI 12/82 3 a,b,d 2 2 PO Supervision XII 08/83 1 d 2 1 F,O Supervision XIII 04/84 2 a,d 2 1 F,0 Supervision XIV 10/84 1 a 2 P F,M OTHER PROJECT DATA borrower Government of the Republic of Turkey Executing Agency General Directorate for Livestock Development (LDP) On-Lending Institution Agricultural Bank of Turkey (TCZB) Follow-zn vroject: None Tiscal Year March 01 - February 28 Name of Currency (abbreviation) Turkish Lira Currency Exchange Rate: Appraisal Year Average 1975 US$I.00 * LT 15.0 Intervening Years Average 1986-83 US$1.00 * LT 100.0 Completion Year Average 1984 US$1.00 * LT 444.0 /a a - agriculturalist; b - agricultural economist; c - financial analyst; d * technicel specialist b 1 - Problem Free or Minor Problems; 2 * Moderate Problem; 3 - Major Problems. 1 * Improving; 2 * Stationary; 3 - Deteriorating. 7_ F - Financial; N * Managerial; * Technical; P * Political; 0 - Other. Recalculated for the dairy component only, which comprised 60% of total project costs. PROJECT PERFORMANCE AUDIT REPORT TURKEY THIRD LIVESTOCK DEVELOPMENT PROJECT (LOAN 1265-TU) EVALUATION SUMMARY Introduction The project was part of Government's efforts to modernize the livestock sector through dairy intensification in twenty-five provinces surrounding major urban milk consuming centres. The project was the Bank's eighth loan to Turkey for agriculture and first repeater project in the livestock sector. Objectives The principal project objective was to increase the supply of commercial milk production through the importation of higher yielding animals distributed under a supervised credit program to encourage the adoption of modern techniques of animal husbandry by participating farmers. Secondary project objectives included efforts to reach small village farmers situated in the poorer, eastern region of Turkey. The project was to be implemented by the General Directorate of Livestock Development Projects (LDP) which was to benefit from institution building under the project. Credit was to be extended by TCZB, on the basis of farm plans prepared by LDP field staff working in collaboration with farmer sub-borrowers. Implementation Experience Importation of over 10,000 heads of genetically superior in-calf dairy heifers was successfully completed with lower unit costs in dollar terms than expected at appraisal. The Government's adoption of structural adjustment policies in the course of project implementation (including a realistic exchange rate and improved export incentives) adjusted the returns to farmers on various activities to reflect more closely Turkey's comparative advantage. As a result meat production became relatively more attractive than milk production and it was more profitable to purchase domestically-bred offspring of importel dairy cattle than to import new cattle. The loan agreement was revised twice to provide for these changes and the project focus shifted to include fattening as well as dairying. The basic concept of a supervised credit programme was retained, however, though credit was no longer the major constraint. Somewhat surprisingly, no other measures to deal with other important constraints were introduced at the revisions. The second project revision did, however, have the effect of completing disbursements. - iv - Institution building was only partially successful. LDP suf- feredl/ from lack of government support in respect of both shortages of counterpart funds and delays in appointing consultants and a lack of focus, tending to concentrate on preparation of individual farm plans to the exclu- sion of almost all else. Technical supervision of subloans was weak and veterinary services were inadequate as a result. Training was not completed as planned and technical assistance fell short of appraisal expectations. The project closed two years behind schedule in March 1984. Total project costs at completion were USS34.0 million compared with US$34.7 million proj- ected at appraisal. Results The importation of such relatively large numbers of genetically superior animals contributed to raising the genetic potential of local stock but full benefits as expected at appraisal were not obtained because: (a) the financial attractiveness of dairying activities declined relative to other livestock/crop activities; (b) the technical package for animal husbandry was only partially adopted by farmers thus affecting yields and output; (c) animal health services proved inadequate, which further affected produc- tivity. Weak monitoring of production data by LDP also meant that the actual benefits are difficult to quantify. At project completion, the ERR for the dairy component, which comprised 60% of actual project costs, was recalculated by the PCR at 19%. No economic rate of return was calculated for the fattening component, which constituted the remaining 40% of project costs. Although the region does not agree, the audit concludes that the 19% ERR recalc!ulated by the PCR is overoptimistic, given (a) that from the farmers' perspective dairying became relatively less attractive to fattening over the project period, (b) the use of farm models not representative of actual subloan size, (c) no account taken either of the animals slaughtered or of the possible impact of poor technical support on the production coefficients used. In view of these factors, the ERR is likely to be below the acceptable 10% cut off point. Sensitivity analysis undertaken by the PCP. Annex 2, page 1 reducing project benefits by 20% beginning with the project completion year (1984) causes the rate of return to fall to 10%. If project benefits are similarly reduced by 20% over the life of the project, the rate of return falls to -2%. Sustainability Sustainability of dairy operations will depend on restoration2/ of financial profitability to farmers which, given that producer prices are 1/ Government states that these problems have since been resolved. 2/ The Region states that "dairying in the developed areas based upon Turkish bred offspring of Imported cattle remains profitable." -v - unlikely to improve, will require steps being taken to address the con- straints to more efficient production noted Above, LDP was disbanded at project completion and the staff reassigned under a new extension system, with responsibilities largely decentralized to the province level. The reorganization of extension setvices in Turkey towards an area specific orienta-ion based on the training and visit system may assist in this regard. Findings and Lessons ?voject performance was substantially affected by changing economic policies arising out of the structural adjustment process implemented concur- rently with this project which dampened demand for credit by farmers for imported dairy cattle and increased demand for improved locally-bred cattle and for meat production activities (PPAM paras 26-30). The project was accordingly revised twice to adapt to these changes. Some US$14.1 million (41%) of total project funds was disbursed for fattening activities which were not originally envisaged in the original project design. The advisability of using such a high proportion of project funds for such purposes, which probably could mostly have been financed from other sources, is questioned by the audit which feels that aiequate funds were already available for this purposes from TCZB sources (PPAM paras. 19-20). The rate of return calculation which is based on the dairy program is deemed to be overoptimistic (PPAM paras 29-33). Parallel and on-going activities in the livestock subsector indicate, however, that if a more inte- grated management package is employed - with training and extension, supply of inputs and marketing combined with farmer participation, livestock produc- -ivity can be improved amongst even the small farmers and landless poor (PPAM paras 34-3,). PROJECT PERFORMANCE AUDIT MEMORANDUM TURKEY THIRD LIVESTOCK DEVELOPMENT PROJECT (LOAN 1265-TU) I. SUMMARY Background 1. During the mid seventies, agriculture constituted a preeminent if declining position in the economy: contributing 26% of GE", 57% of export earnings, employing about 65% of total labour force and producing almost all of Turkey's food requirements. 2. Within the agriculture sector, livestock produced about 30% of GDP. The livestock population is amongst the highest in Europe and the Middle East, over three quarters of the herd raised according to traditional animal husbandry methods with the extensive grasslands of eastern Turkey providing the core of the animal production syst3m. Growth in the livestock subsector during the First (1963 -67) and Second (1968-72) Five Year Development Plans was below expectations, due to the small quantitative increase as well as to low productivity. In its Third Five Year Development Plan (1973-77) govern- ment focussed attention on increasing livestock production primarily through better use of fallow land and genetic improvement of various species. Within the livestock subsector, poultry and dairy production were to receive pri- ority. Government strategy for dairy development in the Third Plan called for (a) increasing producer incentives; (b) improving milk collection and processing facilities; and (c) modernization of the sector through provision of technology and the necessary inputs. 3. When this project was appraised in May 1976, the Bank group had extended seven loans and seven credits totalling US$350 million to the agri- culture sector in Turkey; principally for irrigation and supporting services. Previous support for livestock comprised a first phase projectl/ for inten- sive dairy production around four main milk consuming centers. Commercial farmers in 23 provinces would be provided with credit for tht purchase of high quality imported animals and machinery/inputs for the cultivation of fodder crops, supervised by a newly established General Directorate for Live- stock Projects (LDP). This institution would be responsible for technical support and would coordinate with the credit institution (TCZB) and animal health department (DVS). A second livestock project,2/ also a supervised 1/ Intensive Dairy Production Ptoject (Credit 236-TU). 2/ Second Livestock Development Project (Credit 330-TU). -2- credit program was directed towards increasing productivity in the remote rural areas of northeastern Turkey through improved animal husbandry and development of small scale fattening operations on the Anatolian plateau. Both projects were considered at completion to be generally successful in economic and institutional terms.3/ Project Design 4. The Bank attached importance to this third livestock development project, it being the first repeater project in the agricultural sector in Turkey. The project would cover almost the same area as that of the first with two additional provinces in the eastern, poorer part of Turkey, and would continue the approach adopted by the first project emphasizing genetic improvement and fodder development. Principal objectives were to (a) increase the stock of genetically superior dairy animals for upgrading of native herds; (b) continue to develop modern techniques of animal husbandry begu, under the first project; (c) develop a pilot program for lending to small village dairy producers and (d) expand milk production in five large urban areas and in two provinces in the east with high potential. 5. As approved by the Board in May 1976, the project consisted of the following productive components: - importation of 10,500 Holstein and Brown Swiss in-nalf heifers and bulls to be distributed on credit to 750 dairy farms as follows: 300 subloans to small farmers to enable them to increase their herd from about three native cows to five genetically superior cows over a three-year period, 200 subloans to farmers similarly to increase their herds from six to twelve and 250 subloans to large farmers similarly to increase their herds from ten to forty; - construction of barns and provision of tractors and equipment necessary for forage production; - continued strengthening of LDP to ensure proper supervision of the loans through formulation of farm plans and advice on modern production practices including on-farm production of feedstuffs; - training of local technicians and farmers, support for studies and surveys for further development of dairy production and milk marketing. 6. The loan agreement required that three technical advisers would be recruited for the duration of the project Zo provide technical advice to LDP managers, assist in organizing training for LDP staff and in the design and 3/ PPARs Turkey: Intensive Dairy Production Project (Credit 236-TU) OED Report No. 2542 dated June 18, 1979 and Second Livestock Development Project (Credit 330-TU) OED Report No. 3517 dated June 25, 1981. -3- supervi6ion of the technical studies. The project would be implemented by LDP. TCZB, a long established, government-owned bank which provides over 95% of institutional agricultural credit in Turkey, being responsible for credit administration. Coordination between the two institutions would be governed by a protocol established under the first project. 8. Total project costs were estimated at US$34.7 million financed by an IBRD loan of US$21.5 million, a TCZB contribution of US$6.9 million and a sub-borrower input of US$6.3 million. The economic rate of return was esti- mated at appraisal at 22%. Project Revision 9. The loan and project agreements were revised twice in an effort to overcome weak demand by farmers for credit for dairy operations. The first came In October 1977, eight months after effectiveness, when the project agreement was revised in order to address several perceived constraints to TCZB's lending policies. Thereafter, farmers were able to obtain credit for both investment and operating costs. Under the revision, farm plans could include more than 25% of the cost of barns, and incremental costs of labor, forage cropping and feedstuffs for the first year. Demand for loans increased in 1979 but this proved temporary and by the time the project should have closed in March 1982 only US$13.4 million of the US$21.5 million l.oan had been disbursed. 10. The loan and project agreements were further revised in July 1983 to permit loan proceeds to finance the indirect foreign exchange content of local costs, chiefly the purchase of local progeny of imported stock for both cattle and sheep breeding operations. (Under the terms of the original agreement, Bank financing was to be solely for the foreign exchange costs of imported dairy animals.) Under the revised agreement, project funds were allocated for 45% of incremental lending by TCZB for short term fattening and 45% for medium term loans. 10. Despite these revisions, by the amended Closing Date of March 1984, US$5 million still remained to be disbursed. Agreement was reached that would allow IBRD to participate in TCZB's ongoing, livestock credit port- folio; the Bank financing its subloans which would qualify under the revised Loan Agreement and retroactive to July 1983 - when the Loan Agreement was amended. The Loan Agreement was satisfied by having LDP draw up farm plans for the loans already agreed by TCZB and to incorporate those loans found eligible into LDP's supervised credit problem. Fvnds were rapidly disbursed under this facility, with final disbursements taking place in November 1984. 11. As a result of the two project revisions, at completion approxi- mately 58% of total project costs went for dairying activities compared to 98% at appraisal with approximately 41% assisting livestock fattening. -4- Project Implementation 12. Cattle were procured as planned at considerably lower unit costs than expected at appraisal. At project completion TCZB had disbursed only TL 184 million (in constant prices) for the dairy component compared with TL 469 million expected at appraisal, i.e., 39% of total expected for dairying subloans. Actual numbers of subloans processed were greater than expected (1,024 vs 750 at appraisal) but many farmers withdrew after TCZB approval,4/ although the overall number is not known. The total number of subloans actually in operation is unclear, the project completion report citinb 1,024, the government 853; neither of these figures includes farmers who sold their cattle after receiving subloans. 13. Reasons for poor demand for imported cattle can be traced to the changing financial incentives to farmers resulting from the effects of the structural adjustment process on the project as well as from weak technical support.5/ 14. Changes in the exchange rate and improved export incentives made meat production more attractive than previously. Whereas under the first livestock project, incentives for dairying had generally been favorable, ref- lecting policies at the national level promoting food self sufficiency, including high incentive prices for milk production. Soon after effective- ness of the second project, farmers were faced with a deteriorating relation- ship between input costs and producer prices for milk in part due to the official producer price of milk shifting to a protective floor price. This was comgounded in early 1961 by a sudden drop in milk prices in actual terms by TSEK / which resulted in many farmers losing confidence in dairy activi- ties. Although producer prices were raised, and thereafter until project completion increased more or less in line with inflation, any increase in real terms was not considered by either the Bank or government to be a possible option. This then emphasized the importance of lowering production and processing/marketing costs to restore financial viability, which in turn highlightod the need for effective institutional performance by the supporting services. 4/ In August 198U the supervision mission reported that, as of March 31, 1980, 225 farmers had withdrawn from the project of a total of 887 who had had loans approved by TCZB; records show that a further 146 withdrew in a six month period during 1981. 5/ Government attributes lack of demand solely to factors arising out of the structural adjustment process. 6/ The state owned enterprise which purchased milk from farmers handling about 20% of the market for milk, cheese and yoghurt. 5- 15. Technical support to farmers fell considerably below appraisal expectations. Institution building experienced difficulties during implemen- tation. In early years, this was deemed a consequence of Turkey's debt cri- sis when LDP failed to receive sufficient recurrent funding with the result that the Special Operating Fund (SOF)7/ was used to supplement LDP's operating budget. The Fund was insufficient, however, to make good the shortfall to LDP's budget and also resulted in its own original objectives not being fulfilled. As a result, LDP experienced rapid turnover of staff, due to more attractive incentives from other government agencies and insufficient budget for its operating funds, which affected mobility in the field as well as training. Constraints to project implementation were also noted in respect of other supporting institutions such as health risks to the high quality genetic pool of animals due to the shortages of vaccines and drugs from DVS and erratic supply of forage seeds from state seed farms thereby affecting this important component of the project. 16. At the end of 1980, following the threat of suspension of disburse- ments by the bank, broad, corrective action was taken by government in respect of these institutional problems. Performance of LDP and, in turn, of the project did not improve, however. Six months later, in May 1981, LDP was suffering from overstaffing and lack of training, while TCZB's decision to capitalise incerest payments - on the assumption that the floating interest rates were the reason why farmers were having difficulty in repaying their loans - affected the inflow of funds into the SOF. Moreover, supervision missions noted--poor quality of field work including a lack of awareness on the part of LDP staff of the financial implications of the subloans to farmers--TCZ8's reluctance to accept LDP's farm plans resulting in continued slow approval of loans (which previously had been thought to be a function of LDP's restricted mobility in the field)--monitoring of production data of the imported stock not being undertaken on a national level thereby losing both the opportunity to measure project impact and the benefits for upgrading indigenous livestock--the three technical specialists posts funded under the project vacant since 1979/80. 17. When the project was reviewed in depth in November 1981, although project performance had improved somewhat, LOP was continuing to suffer major problems: the technical package of forage production was not being adopted by most farmers which, in addition to the loss of benefits to production re- sulted in increased input costs; farm plans were technically unrealistic; training was not being undertaken; and the consultant positions were still vacant. However, the mission noted that since most farm plans had been drawn up, the extension of project closing by a further two years should allow LDP to concentrate exclusively on extension activities. 7/ The SOF had been expressly established under the first project to provide additional finance to LDP for payment of miscellaneous costs, for example, additional incentives for recruitment of staff, hardship allowances etc., and derived from 0.5% of the interest received by TCZB from farmers. -6- 18. The project revision towards fattening was based upon the results of revised economic incentives more in accord with Turkey's comparative advantage and the desire to improve disbursements by financing subloans in accordance with farmer demand. The other difficulties facing the livestock sector including poor technical services being extended to farmers were unfortunately not addressed8/ and continue to be a constraint today. A strategy paper prepared in part for the mid-term review of Livestock V, com- mented: "The deficiencies in the provision of technical services by LDP and in coordinating with the veterinary service have all along been a major constraint to satisfactory on-farm performance under the Bank-assisted live- stock development projects." Much of LDP's performance limitations can be attributed to the fact that the technical assistance and training components were not implemented as planned principally due to lack of government commit- ment (PCR para 4.10-4.18). Similarly, studies were not undertaken, again in part due to insufficient attention being given to this component by LDP and the Bank. 19. Total project costs at completion were US$34.0 million compared to US$34.7 million. Costs of individual components differed cubtan:ially from that expected at appraisal as can be seen in the following table: Actual Project Costs (US$ m) Including Contingencies Component Appraisal Actual Cattle 16.3 11.1 Buildings 7.3 2.3 Equipment 9.9 5.3 Technical Assistance 1.2 u.6 Crop Production 0.6 Subtotal 34.7 19.9 Fattening - 14.1 Total 34.7 34.0 b/ UPS comments that: "It is not correct to inter that the Bank did not try to address the difficult problem of improving technical services; trie Bank's refusal to extend the closing date until arrangements to employ expatriate consultants were made it evidence of the Bank's pri- ority in this regard. Furthermore, technical components were included in the ongoing Fifth project (e.g. animal breeding improvement, artifi- cial insemination, mastitis control, animal disease control and vaccine production)." -7- 20. The most significant deviation concerns the fattening component, which was introduced following revision of the project in July 1983, with an estimated US$14.1 million or over 40% of project costs funding this compo- nent. The project revision towards fattening was based upon finding a solu- tion to the problem of disbursing the remaining proceeds of the loan (PCR, para. 3.05), rather than addressing the iesue of constraints facing the live- stock sector, including the poor technical support to farmers. Fattening has long been a profitable industry in Turkey, well supported by institutions such as TCZB which has a large livestock portfolio of its own; Government also considers that providing loans in the project area from both TCZB and LDP resulted in a low demand for loans from LDP. This project revision had the effect of speeding disbursements; the project closed two years late in March 31, 1984 - with final disbursement taking place nine months later in November 1984. Project Outcome 21. The importation of 10,500 genetically superior in-calf heifers improved milk production in participating herds and assisted in raising the standard of living of the iivolved rural population. Dairying was considered particularly effective in this respect since it provided farmers with a daily income. Project records indicate that considerably more subloans were extended for five cow units than expected at appraisal although, all such subloans were not necessarily as the PCR states, to small farmers. The effects of the change in macroeconomic policies to reflect more closely Turkey's comparative advantage (see paras 27-28) as well as the weak tech- nical support extended to farmers leads the audit to consider that the 19% rate of return is overoptimisfic (paras. 29-33). 22. Future sustainability of the dairying activities depends in large part on the long-run impact of the policies emanating from the structural adjustment process. Government feels that credit is not now the binding constraint in the livestock sector, a conclusion reinforced in a general review of the livestock sector undertaken as part of the supervision of Livestock IV. The latter review proposed that "future development strategy for the subsector should tackle a number of problems: marketing, animal health care, breeding, animal production (particularly improved herd manage- ment and feeding practices through domestic production of forages), research and extension, in a more comprehensive manner than has been attempted so far."9/ Parallel experience in the livestock sector in Turkey would also reinfjrce the merits of such an approach (see paras 34-39). 23. Sustainability of eff>rts in respect of institution building have also been affected by events at the national level. Under the auspices of the Agricultural Extension and Applied Research project (Loan 2405-T1) and 9/ Turkey Livestock IV (Loan 1586-TU) supervision mission dated Decem- ber 13, 1985. -8- the follow on Agriculture Sector Adjustment Loan,10/ both supported by the Bank, extension activities are now organized according to the training and visit system rather than the commodity specific approach as formerly employed. While LDP as an institution has been disbanded, most of the former LDP staff have been assigned to the Projects and Statistics Sections of the recently decentralized provincial extension services. In these position LDP staff can use their experience in preparing farm development plans for pro- posed investments and, by providing advice and assistance in livestock matters to the village extension workers (who now provide advice in both crop and livestock production), can be instrumental in strengthening technical services to farmers and contributing to the sustainability of project con- cepts and Government's new institution building initiatives. II. MAIN ISSUES A. Impact of Structural Adjustment Process on the Project 24. During the seventies, Turkey did not make the necessary adjustments to the effects of the steep rise in oil prices and the recession and high inflation in the OECD countries with the consequent deterioration to its terms of trade. Until 1977, however, growth rates in Turkey remained high largely because of sustained public investment financed to a great extent by remittances and short term borrowings. Policies during this period resulted in an overvalued exchange rate and active public participation in the economy through state economic enterprises whose practices addressed national social and economic needs rather than commercial profitability. In the agricultural sector, food self sufficiency was an important goal to be achieved through import substitution by means of subsidized inputs, high producer prices and shelter behind domestic tariffs.11/ 25. By 1977, however, the rapid growth in GNP came to a halt, and the exter-ial debt burden became such that restructuring of the economy became essential. A program to initiate the required adjustments was adopted by the new government which came to power in 1980. This resulted in a considerable change from previous policies with the introduction of an outward oriented development strategy to liberalize the economy by relying on market forces for allocation decisions. This export-oriented approach shifted the emphasis from food self-sufficiency to increasing agriculture's net contribution to the balance of trade with imports of certain foodstuffs, for which Turkey 10/ Both these lending operations are in fact the mechanism for implementing objectives laid down in SALs II & III: namely to streamline and improve the delivery of agricultural extension and related supporting services. 11/ Turkey: PPAR, First Structural Adjustment Loan (No. 1818-TU) and Supplement (No. 1915-TU), OED Report No. 4015 dated July 6, 1982. -9- does not have a comparative advantage, being allowed.12/ The implications of this adjustment process for the agricultural sector largely concerned: reducing the overall level of subsidization which had contributed to the strain placed on the budget and included, inter alia, reduction of subsidies on inputs ranging from fertilizers to agricultural credit; reduction in the number of production price supports as well as their gradual conversion from high incentive prices to protective floor prices; w.th the introduction of a realistic exchange rate policy, a shift in the terms of trade in favor of exports; and reform -i state owned enterprises.13! 26. The implications for the Third Livestock Development Project were (i) the favorable TL price to farmerr of imported cattle (the largest component in terms of project costs) relative to that of locally improved animals at appraisal was reversed, with the price of the latter falling to 50-60% of the price of imported animals; (ii) the relative price of milk/meat which, prior to structural adjustment, had favored milk was reversed with fattening operations becoming financially more attractive to farmers; (iii) the actual value of subloans constantly being revised upwards from that approved in the initial farm plans to take account of devaluation of the Turkish lira (1700% over the project period) with farmers having to bear the foreign exchan&e risk between preparation of the farm plan and actual import of the cattle. 4/ The repayment terms of the subloans similarly increased with the introduction of variable interest rates on agricultural credit; and (iv) increasing marketing problems due to government efforts to try to restore some form of financial viability to TSEK, the state-owned enterprise involved in collection and marketing of milk. 27. The effects of these policy changes can be seen in a decreasing demand by farmers for subloansl5/ for the purchase of imported dairy animals, reflecting that the structutil adjustment process, undertaken between 1980-84 in effect changed the basic assumptions on which the project was based. While the changed environment encouraged production decisions more in line with Turkey's comparative advantage, cerL,in agricultural undertakings including dairying definitely became less attractive to farmers as time passed. Other activities, such as fattening, in turn b;:.acme more attractive. 12/ Turkey: PPAR, Second and Third Structural Adjustment Loans (Loans 1987-TU and 2158-TU), OED Report No. 5763 dated June 28, 1985. 13/ Turkey - The Fifth Five Year Plan in the context of Structural Adjustment, Report No. 5418-TU dated July 20, 1985. 14/ Which could result in "great differences between the loans as projected in the farm plans and effective loans ..... exact figures about the loans were often not known to the farmer or to LDP field staff." Aide M6moire dated December 24, 1980. IS! As Government pointed out, the fear of not being able to repay the loan prevented some farmers from requesting loans. - 10 - 28. The project was prepared and appraised four years before these adjustments were introduced, and it is doubtful whether any project could be designed with such sufficient flexibility to withstand shocks of this magni- tude. However, given the fact that the Bank was intimately involved in the structural adjustment processl6/ from the beginning of the dialogue under- taken by government with external donors, the question has to be raised as to why the performance of this project was not analyzed more within the context of the changes being introduced at the national level and why it took the Bank from 1980 to 1983 to recognize and adjust project design. Instead, supervision at the time was preoccupied with the effects of the financial crisis on LDP's operating budget and although attention of government was drawn to the effects of the drop in milk price following TSEK's efforts to restore its operations to financial viability, the project's position within the sectoral/national context was not considered in depth. If this had been done, it is possible that the need for project revision could have been identified sooner. B. Rate of Return 29. The project was appraised as a dairy project with the economic rate of return calculated at 22%. At project completion, an economic rate of return was recalculated for the dairy "component" of the project at 19%. The audit questions some of the assumptions underlying this reestimated rate of return. 30. From the farmer's perspective, it is evident that the effects of the structural adjustment process resulted in first, incentives in the form of an attractive producer price for milk deteriorating vis a vis other live- stock/crop activities with the result that many farmers found it more profi- table to sell the offspring and/or enter fattening operations.17/ In the seven out of 25 provinces visited by the mission, both farmers and officials repeatedly drew attention to the low price of milk and stressed the profitability of fattening or crop activities. Second, farmers in certain areas encountered marketing problems, described by Government as one of the most important deficiencies of the project. Farmers relying on the processing plants of the TSEK encountered difficulties both with plants closing on weekends and sudden fall in milk prices fixed by TSEK; subsequently certain TSEK plants were also closed in an effort to stop financial losses. Third, individual farm plans, once approved by LDP and TCZB, changed in value terms reflecting the difference in TL t2rms of dairy animals from the time of subloan approval to the time when the animals 16/ Since March 1980 the Bank has supported in Turkey five structural adjustment loans totalling US$1,556.3 million and an Agricultural sector loan of USS 300 million. 17/ Government states that one of the most important reasons why animal fattening was preferred over dairying was that 'milk could not find a market for what it was worth.' - 11 - arrived at the farm. Moreover, with the freeing of interest rates following the structural adjustment program, farmers were equally concerned about the variations likely to occur in the servicing of their subloans. In addition there were continuing constraints to on-farm performance of improved animals, notably poor management, feeding (forage production was seldom introduced; Government attributes this to farmers insisting on retaining their tradi- tional farming methods in some regions and the profitability of alternative crops) and disease control.18/ These combined to result in farmers losing interest in dairying activities. The extent of farmer withdrawal has not been documented, but in 1980, supervision reported that 225 farmers had with- drawn out of 880 who had subloans approved; in 1981 a further 146 farmers withdrew.19/ 31. The 19% rate of return for dairying is based upon the number of animals actually imported (10,300) taking no account of animals which have since been slaughtered. No data is available on this, but the issue was considered sufficiently serious to warrant supervision mission attention.20! No actual production data exist and thus the coefficients of dairy production contained in the PCR are estimates only. As Table 5 (PCR, para. 5.02) shows, these estimates are almost identical to those estimates used at appraisal. This similarity in production estimates calculated at appraisal and at completion is hard to reconcile given the likely impact on production coeffi- cients of the substantial weaknesses both within the individual supporting services - lack of attention to forage production, adequate disease control, regular breeding - and in coordination between these services. Such weak- nesses were regarded by supervision missions as contributing to standards of farm management and animal husbandry "often far below" what was deemed neces- sary to obtain efficient production. And, as the project completion report pointed out "after subloans were approved, extension assistance to the subborrower was virtually non-existent." While farms located close to urban centers were a5le to bypass the weaknesses noted in the public extension and research services with their access to markets and good management, in general upstream and downstream production and marketing industries played only a small part in transferring technology thereby fostering increased and more efficient on-farm production, 18/ "The Livestock Subsector in Turkey, Current Constraints and Possible Medium term Strategy" page 5, Supervision Mission Livestock IV dated November 1985. 19/ The Region attributeb this to a loss of interest in imported cows. 20/ The Region comments: "No evidence is given of dairy cattle being slaughtered (beyond normal culling rates) ... staff involved do not believe that this has been a problem." However, the first importation of cattle took place in December 1977. Since only in calf heifers were procured and since culling on grounds of age begins around ten years, it seems likely that "normal culling rates" throughout the life of the project would have been extremely low. - 12 - 32. The rate of return is based upon farm models which are not alto- gether representative of actual subloan sizes. Three farm budgets were prepared at appraisal, for a five, ten and forty cow herd. At completion, only a forty cow model has been recalculated. However, the data available on the actual subloans by size shows that in fact few subloans were extended for a forty cow unit. LDP data indicates that only 3% of subloans were for this category - 62% being for five cow units, while the project completion report indicates that 20% of subloans were for 20 cow units and over, with 55% being for five cow units. The mission was also told that owners of five cow units were able to market the milk themselves thereby obtaining a higher return; Government has added that as a result of the economic conditions in Turkey, enterprises were small scale. 33. Although the region does not agree, the aulit considers that the 19% rate of return recalculated by the PCR to bz overly optimistic and dravs attention to the sensitivity analysis undertaken by the PCR (Annex II, page 1) wherein if project benefits are reduced by 20% from project completion in 1984 to full development the rate of return falls to 10.0%. If the project benefits are reduced by 20% from the start of the project, the rate of return falls to minus 2%.21/ In view, therefore, of the far reaching effects of the structural adjustment process -- unforeseen at appraisal, the limited adoption of technology together with the weak technical supporting services, the ERR for the dairy component is likely to be below the acceptable 10% level. The audit notes that the rate of return was recalculated for the dairy component only. The PCR felt able to use farm models and best estimates for the dairy component but not in respect of the fattening component, although models were built in 1983 when the project was restructured. The PCR states that "too little information" was available on the benefits to undertake an economic analysis, although this component constituted almost 40% of total project costs. Since fattening has long heen a profitable industry in Turkey, if a rate of return had been calculated for this component, it may well have had a favorable effect on the rate of return overall. C. Alternative Strategies Which Both Increase Livestock Productivity and Reach Small Farmers 34. This project was noc designed as a poverty oriented project but it contained a pilot programme for lending to small village dairy producers in two provinces in Central/Eastern Anatolia. This region has considerably less production potential, with a relative paucity of resources and harsh climatic conditions when compared to the Western/Mediterranean reaches of Turkey. Livestock production fits well with these agroclimatic conditions, flocks migrating to higher pastures for the spring and summer months then either fattened for western markets or returned to villages for the winter months; it is also an important means whereby landless people are aole, with their few animals, to obtain milk, meat and fibre for immediate family needs. 2/ The Region comments that the presentation of the sensitivity analysjs is uhalanced and that there should be no inference thit a drop of 20Z In enef ts is expected. -13.. 35. While this project would have had some impact in increasing stan- dards of living of villagers under the auspices of the pilot programme, par- ticularly with the introduction of pure bred animals, several constraints remain. Most commonly cited concern the marketing of milk; credit recovery in certain villages; the difficulties for an established credit bank to onlend to individual, small farmers; and lack of extension support in part due to the animals having been distributed to farmers over wide distances. Meanwhile, the po:ential contribution which livestock activities can make to rural incomes is growing due to the fact that arable land is becoming increasingly fragmented as small plots are divided up among the large fami- lies characteristically found in Eastern Anatolia. 36. Given both the identified constraints and the importance of, and benefits to be obtained from, increasing livestock productivity, it is of interest to analyse a parallel strategy in Turkey to livestock development. This has been introduced and refined over the past decade by a non- governmental organization (NGO), and appears to be rather successful both in overcoming the constraints noted above as well as reaching small farmers or landless people. 37. The basis for the NGO's strategy for dairy/sheep improvement, which is based on an earlier model developed specifically for poultry operations, is an integrated package approach which goes beyond providing supervised credit and also relies heavily on local participation. As a first step, fodder crops for improved feeding of cows and sheep are introduced by pro- viding seeds at half price for the first planting, subsequently farmer groups are responsible for the procurement of seeds; bulk purchase of other inputs for example, building materials for shelter, are undertaken to reduce costs; veterinary services are supplied, much of the work now being undertaken by farmers themselves; and, a key element, milk collection centres and mobile units established which not only provide the livestock owners wit! guaranteed a daily income but are an important means whereby the repayment of cvedit is secured. The livestock improvement program also undertakes crossbreeding activities as well as removing indigenous bulls from local herds since AI services are rarely used. The emphasis is towards small farmers: the 1517 families reached by the programme in 1985 have on average less than two improved head of cattle per family. 38. The strategy, therefore, is based upon an integrated approach, pro- viding training and extension, supply of inputs and marketing; and ensuring farmer participation in planning and implementing the activities and commu- nity organization. With the benefit of hindsight, it is evident that the design of the Third Livestock Project, with LDP essentially having responsi- bility only for animal husbandry and with little ability to influence the activities of other supporting services - particularly health and marketing/ processing has adversely affected both LDP's efficiency and public sector efforts to increase livestock productivity. 39. The question then emerges as to whether there could have been a role for such an organization in this particular project. This would have 14 - been difficult at appraisal because the organization was too small or diffuse to be able to benefit from Bank financing.2/ By the time that this project was restructured in 1983, however it may well have been possible to channel the undisbursed funds towards such an organization. While its experience in respect of improved sheep and dairy production was still in the early stages at that time, the formula employed is similar to that which had been devel- oped in the poultry subsector, which has been successfully supported by the Bank under the Fifth Livestock Project, which had been prepared and appraised five years previously, between 1978 and 1979. In addition, this formula also complements the more strategic and comprehensive approach now being outlined by the Bank for a possible future lending program to the livestock sector in Turkey. With the benefit of hindsight, it seems evident that an opportunity was missed23/ when the project was restructured in not supporting this organization's dairy/sheep activities for landless/poor people particularly when TCZB has such an active portfolio for livestock fattening. 22/ It is possible, however, that this particular problem can be overcome in future if for example the experience obtained in some Bank-financed rural development projects in Brazii proves to be successful. There, the Bank is providing seed money whereby between 10-20% of total project funds are non-specified to be used to support community projects identified by the farmers themselves. Such funds could be used to promote institutional mechanisms such as this in Turkey. 23/ The Region states: "TKV was specifically asked at that time to demonstrate its livestock activities, but was not able to present any evidence of successful experience or technical expertise in the livestock field (other than poultry)." COMMENTS RECEIVED FROM - 15 - GOVERMNT EMBASSY OF THE REPUBLIC OF TURKEY OFFICE OF THE CHIEF COUNSELOR FOR ECONOMIC AND COMMERCIAL AFFAIRS 2523 Massachusetts Avenue. N.W., Washington, D.C. 20008 TELEPHONE (202) 4836366 Ref: 31.3.45/1524 (2o2 4033 904143 TUFN-UR July 24, 1986 Mr. J. Roy Acting Division Chief, CP II Emena/IBRD Washington, D.C. Re: Third Livestock Project Loan, No. 1265-TU Mr. 0. Maiss', Acting Director of OED, telex dated July 11, 1986 - Dear Mr. Roy, I am forwarding to you the 7urkish versions of the comments prepared by MAFRA on Project Evaluation and Completion Reports of the Third Livestock Project. The English versions of the above-mentioned comments will be sent to you in due course. I hope the delay in sending the comments will not cause any inconvenience. Thank you for your cooperation. Yours sincerely, Hikmet Ulugbay Chief Counselor Enc.  - 17- TRANSLATION OF RECEIVED IN IBRD ON GOVERNMENT COMENTS PAGE 1 OF 5 NOTE CONCERNING THE PROJECT EVALUATION REPORT Implementation Experience In spite of the fact that the General Directorate of Livestock Development Projects (HAYGEL)(LDP)* and the provincial organization suffered partially from insufficient personnel and budget during the years of its establishment, the personnel Page iv problems of this General Directorate has been eliminated by making appointments for both agricultural ergineer and veterinary positions in the following years. Moreover, the budget of the General Directorate has been supported by the allocations needed for allowinces and the control of the enterprises. The training program has been realized as mentioned on page 11 and 114 ' the project completion report. Results The technical and health package proposed to the producers by (HAYGEL) has been adopted by the farmers as much as possible. For instance, each farmer who wanted to benefit from the (HAYGEL) projects by obtaining loan from the project was asked to cultivaLe feed plants (clover, trefoil, common vetch, etc.). Included However, there have been difticulties in adopting the new methods in para 30 in the area of feeding due to the insistence of the farmers in implementing the traditional farming methods in some regior and the fact that the vegetables and industrial plants are more profitable. In spite of this, the knowledge of the farmer has been improved by the training activities and demonstrations made in recent years. For instance, they started grazing and not to feed the animals with hay. In cases where milk inarketing, which is one of the most important deficiencies in the project, is well managed, it was observed that there was not any probl*em left in the field of feeding, that is, the farmers followed the instructions. The animal health services which are directly related to the productivity have been taken care by the establishment of a (HAYCEL) provincial directora e in every city and the support with sufficient number of veterinarian and animal health officer.5 in recent years (before 1985). Every enterprise that has been established by the project was controlled periodically, two times a month; the required health advise was given, and; the required medicine and vaccine were provided on time and in sufficient amounts upon a protocol made with the General Directorate of Veterinary Affairs. (*) (LDP) - Livestock Development Projects - 18 - TRANSLATION OF GOVERNMENT COMMENTS PAGE 2 OF 5 Although, the pedigree and productivity records were kept at each provincial directorate, and these were recorded systematically by the Health and Pedigree Department of the General Directorate, a recordation at the national level was not possible for the proj-cts due to the lack of a master plan prepared for livestock during the commencement of the project. Studies, like national pedigree systems, etc. could not be made, since the "Institution Building Concept" was not given place in the III Livestock Project. This deficiency was felt during the project implementation, and the livestock strategy studies were started with the World Bank. During the course of the project, demand for stock farming for milk production was higher at the beginning, however, animal fattening received preference later on. One of the most important Footnote 17 reasons of this was the fact that milk could not tind market for para 30 what it worth. In addition, the farmers were able to quickly convert into cash the money they invested in because of the shorter stock fattening period. The fact that the farmer who wanted to obtain a loan for stock fattening was not required to separate a land for rough feedstuff production encouraged stock fattening for persons who did not have a land. In addition to this, the extensive pasture grazing in our country reduces the cost of stock fattening significantly. Project Implementation Those who sold their animal after obtaining the loan are not included in the 853 enterprises mentioned in the PCR*. The &44f53 enterprises at that date are the enterprises in active Para 12 implementation, amended - The reason for low demand for imported cattle is the fact that the prices of the breeding animal purchased domestically were lower and that the project became active by obtaining the Footnote animal in a very short time after the initiation of the project. added to It is not related to the weakness in the technical service. para 13 - The special fund was used for the purposes of meeting various training expenses made for the trainers, as allowances during the training in the country, the stationary expenses related to the project, for their printing, etc. There is a need for a special fund account in these kind of projects in order to increase the project operation. * (PCR) - Project Completion Report -2- TRANSLATION OF - 19 - GOVERNMENT COMMENTS PAGE 3 OF 5 - During the recent years, job-leaving by the Livestock Development Organization personnel has been prevented, and, on the contrary, the transfer of the personnel of the other organizations to HAYGEL has been encouraged. - Providing loans for livestock activities in the project implementation areas by both, TCZB (Agricultural Bank of the Republic of Turkey) and HAYGEL has created a low demand for Para 20 HAYGEL projects. However, there has been an increase, again, in amended the demand for loans, after it was decided, in 1984, that only HAYGEL loans would be utilizel in the provinces which are under the coverage of the project as required by the agreements made with the TCZB. (Loans have been provided from a single source). - 20 - TRANSLATION OF GOVERNMENT C(0ENTS PAGE 4 OF 5 NOTE CONCERNING THE PROJECT COMPLETION REPORT 3.03. There was no slowness in the preparation of farm plan and in extending credit. However, increase in the investment costs, low level of milk prices and the marketing problems affected the loan amounts adversely. 3.08. The issue of insufficiency of the farm plans as mentioned in the examination report by the World Bank staff in Added tc August 1980 was originated from the economic conditions which paras 27 & 32 Turkey was under during that period. The high level of tnterest rates, the high cost of investments and the market problems have reduced the demand. As a result, the enterprises were of small- scale, and, the fear of not being able to repay the loan has prevented some farmers from requesting loans. 3.13. As a result of the fact that the imported animals could not provide sufficient productivity and that the delivery of the imported animals was made before the enterprise came to operational stage, animal imports were stopped and the breeding animal was started to be supplied to the enterprises from the State Farms in Turkey. 3.15. The fact that the economic conditions of the preparation and implementation stages were different caused the Turkish lira counterpart of the loans, calculated during the evaluation, to be exceeded significantly. 3.16. The realization, as 63%, of the 62% World Bank contribution to the total project costs as calculated in the evaluation report indicates that a good planning has been made. 3.18. The personnel requirement of the projects is met in accordance with the employment policy of the government. For this reason, the number of the personnel may look low in certain cases. However, well-training of the present personnel fills this gap. 3.22. Programs, abroad, that are oriented towards increasing the knowledge and the experience of the farmers covered by the project have been prepared. However, this kind of training programs have been kept limited, by taking into consideration that the training of the farmers would create some administrative and technical problems, if it is made before the completion of this type of training of the technical personnel assigned at the project site. 4.03. No adverse effect has oc urred on the quality and the preparation period of the farm plans prepared by the technical personnel at the project site. The technical personnel who are on duty in this area have been continuously subject to training and control. - 21 - TRAMSLATION OF GOVERNMENT COMENTS PAGE 5 OF 5 4.06. Control of the milk enterprises that are financed under the coverage of the III Livestock Development Project and the provision of new services have been fulfilled to the extent possible by the present personnel at the project site and other capabilities. However, the majority of the animal-fattening enterprises established under the same coverage could not be controlled since their activities in this respect did not continue. 4.08. Since the technical information to the established enterprises is also provided by nther ministerial organizations at the project site during the project implementation, the Page iv evaluation of the LDP/IDPD performance alone is wrong. For this reason, increase of the service efficicncy by a new project was planned. 4.16. Training of the personnel abroad has always been a problem due to the language incompetency. However, it is expected that this problem will be overcome by the effective operation of the language laboratories established at our Ministry and by the intensive language courses to be opened in the cultural centers in Ankara within the coverage of external credits. 4.23. A new M+E unit has been established in our Ministry. However, both the small number of personnel at the present moment and the large number of projects to be studied resulted in the limitation of the documents in this respect. 5.05. The figure indicating the average milk productivity per animal does not represent the impacts of the project. The milk productivity is low in Turkey, however, there is large number of locally well adapted domestic animal, for milk production. For that reason, a comparison with Iran and Greece is not appropriate. 6.01. Since record-keeping operation is not extensive in the enterprises in Turkey, the surveys made at the enterprises established by loans do not provide positive data in every circumstances. However, surveys have been tried to be made systematically as much as the capabilities permitted. 8.01. The realization of the targets with respect to the organization was not possible due to the fact that a new organization is to be made in our Ministry. The effective Para 23 extension service foreseen by the project in the new structure is being provided also with the support of the veterinary services. 8.02. The wage benefits provided by our Ministry for the personnel at the site in the de%eloping areas are at the levels which can not even be compared with the private sector in certain cases. ～ご2一 メり－;2ひ分2承ルぐ 23 TURKEY THIRD LIVESTOCK DEVELOPMENT PROJECT Loan 1265.-TU PROJECT COMPLETION REPORT November 6, 1985 Regional Projects Department Europe, Middle East and North Africa Region - 24 - TURKEY THIRD LIVESTOCK DEVELOPMENT PROJECT (LOAN 1265-TU) PROJECT COMPLETION REPORT I. INTRODUCTION 1.01 Milk and milk products are traditional in the Turkish diet. At the time of appraisal (1975), there were about 13.8 million cattle, 41 million sheep, 18.8 million goats and one million buffaloes producing 4.8 million tons of milk. Of this, 60% or 2.9 million tons were cow's milk with the remaining 40% coming from sheep, goats, and buffaloes. Annual per capita consumption of milk and dairy products was estimated at 100 liters of milk equivalent, compared to 218 liters in Ireland and 56 liters in Greece. The increase in milk production in the decade before appraisal was less than one percent annually, well below the population growth rate. As a consequence, access of the rapidly growing urban populations to milk and dairy products had been gradually decreasing. 1.02 By 1975, the bulk of Turkey's milk was produced by small and widely scattered small farmers, the only modern dairy farms being a few government farms and those established under the Bank's First Livestock Development (Intensive Dairy Production) Project (Credit 236-TU). Average annual milk yields of indigenous cattle were below 600 liters, or only one-fifth of those of imported cows. To increase milk availability, both genetic potential and feeding and management practices needed to be improved. 1.03 A second problem in the dairy subsector was the lack of a well integrated marketing infrastructure. Ninety percent of milk and dairy products were marketed by small-scale, generally uncontrolled peddlers wich peak supplies in spring and summer and low supplies in winter. Government milk plants were handling less than 2% of total production, and 5% of marketed milk but a national program was underway to provide modern processing and storage facilities for about 20% of total marketed milk and dairy products. At the time of appraisal, a recent 50% increase of floor prices offered by government processing plants had greatly increased farmer interest in dairying. 1.04 The Intensive Dairy Production Project (Credit 236-TU, for US$4.5 million) was the first attempt by Government to provide investment credit facilities and technical expertise to dairy producers around the country's four largest milk consumption centers (Adana, Ankara, Istanbul, and Izmir). The project became effective in December 1971 and provided for imported dairy cattle, credit to farmers, technical services and training for staff and farmers. It closed on March 14, 1979 and the Project - 25 - Performance Audit Report (No. 2542) sho%ed that project performance slightly exceeded appraisal estimates in terms of bio-economic efficiency parameters and financial/ economic returns. Overall, the project was successful in establishing a nucleus of over 3,500 high-quality dairy cows and in promoting dairy modernization through provision of supervised credit and improved technical services to dairy farmers. The third livestock project was to build on this experience.1' II. PROJECT FORMULATION AND PROCESSING A. Origin and Processing 2.01 A Bank assisted Third Livestock Development Project was first mentioned in August 1972 by the General Director of Turkey's Meat and Fish Organization who was interested in slaughtering facilities and meat packing plants. In October 1973, however, a Bank supervision mission was told that the Third Livestock Development Project would likely be an extension of the First Livestock Development Project (Credit 236-TU), and that preparation wasi being done by the General Directorate of Livestock Development Projects (LDP) of the Ministry of Food and Agriculture (MFA). A two man FAO/IBRD Cooperative Program mission also contributed to preparation. 2.02 A proposal was submitted to the Bank in mid-1975 which included: (a) extension of the First Livestock Project (which had covered 23 provinces) to an additional 21 provinces, for a total of 44 provinces; (b) minimum herd size of 10 cows, with 260 farms to have an average of 40 cows and 140 farms an average of 12 cows; (c) technical assistance (two staff) plus one-half the cost of LDP's technical service; (d) technical studies, which were not described in detail. The total cost of the proposed project was TL 324.6 million (US$23.2 million). The project design incorporated some of the lessons learned from the First Livestock Project in which the minimum herd size was 20 cows, and a Lequirement for farmers receiving credit was possession of irrigated land. Both factors were thought to limit the number of participants and so the proposed project lowered the minimum herd size and did not require possession of irrigated land for subborrowers. 1/ A Second Livestock Development Project (Credit 330-TU, for US$16.0 million) supported village livestock development in Turkey's northeastern highlands and cattle and sheep fattening in the Central Anatolian Plateau. Re-estimated financial and economic returns in the PPAR No. 3517 exceeded appraisal estimates. - 26 - 2.03 In October/November 1975, the Bank appraised the project and recommended continuation in the 23 provinces of the First Livestock Project and addition of 2 provinces; a total of 600 subloans to be made for farms of 5, 12 and 40 cows, with about 25% of the subloans to be made to the smaller, 5-cow farms; technical assistance, and training. The total cost was estimated at US$38 million. Thus, the emphasis was on livestock development in a smaller area on a more intensive basis (50% more subloans were included at appraisal compared to the number proposed) and for smaller farms. 2.04 The issues identified in the issues/decision process were the shortage of LDP technical staff (particularly for the Second Livestock Project), the lack of salary incentives, the interest rate of 11% in a period of inflation for the previous three years of 15%, and a spread to the Agricultural Bank of Turkey (TCZB) of 2-1/2%, which was thought to be too high. On the staffing situation, the Bank would ask Government to make recommendations on staffing and incentives prior to negotiations. It was agreed that the interest rate and spread were acceptable for the first year of the project, but that during negotiations agreement would be sought that the Bank and the Government would jointly review the 11% rate one year after loan effectiveness, and the spread would be examined after further analysis of TCZB and LDP administrative costs. 2.05 The decision meeting also agreed on the smaller geographic area of 25 provinces rather than the 44 proposed, and on the distribution of subloans to smaller farmers. Having LDP rather than TCZB as the executing agency was thought to be a practical arrangement. The meeting concluded that a marketing study should be included in the project to examine the milk collection system in the project area. 2.06 Before negotiations, a Turkish delegation negotiating the Agricultural Credit and Agroindustries Project (Loan 1248-TU) agreed to a review of subloan intE.rprt rates at such time as the Bank, the Borrower (TCZB), or the Government would request; such a review under the livestock loan was thus no longer necessary. In addition, Government gave the Bank assurances that progress had been made both in number of staff under the Second Livestock Project and on incentives paid to staff. 2.07 Negotiations were held from April 14 - 22, 1976. Small changes emerged in project area and numbers of project subloans and, although the Bank was satisfied with progress to that point on LDP staffing and remuneration issues, several covenants were included in the loan agreement to ensure continued progress (para 3.18). 2.08 The loan was approved by the Executive Directors on May 25, 1976 and signed on May 26, 1976. The original loan effectiveness date was September 27, 1976; the loan became effective on February 25, 1977 after TCZB and LDP had agreed to and signed a protocol for the credit aspect of the project, which was the condition of effectiveness. - 27 - B. Project Description and Objectives 2.09 The project was to further the development of the dairy subsector and to expand and intensify the efforts of the First Livestock Development Project, to upgrade the dairy infrastructure, including the improvement of the extension service and expansion of a nucteus of genetically superior cattle. The project included "a pilot program of lending to small village dairy producers" who would, over the long term, provide the basis for any larger effort to modernize the dairy subsector in Turkey. 2.10 The project was to finance investments to be undertaken during a 5-year period, including (a) farm development costs of 750 dairy farms; (b) internal and external training of technicians in modern dairying; (c) training of farmers through study tours and extension activities; (d) provision of technical assistance; (e) studies in dairy production and milk marketing; and (f) preparation of a future intensive dairy development project in additional Provinces. About 200 of the 750 dairy farms were expected to increase their herds from about 6 indigenous milking cows to 12 genetically superior cows over a 4--year period; and about 250 farms were to increase their herds from about 10 indigenous cows to 40 genetically superior cows also over a 4-year period. These 450 farms were expected to receive the bulk of about 10,000 purebred in-calf heifers and all of the 450 purebred breeding bulls planned to be imported under the project during the first three project years. In addition, about 300 subloans would be made to small farmers who would increase their herds from about 3 indigenous cows to a genetically superior herd of 5 cows over a 3-year period. Project implementation was expected to be completed in early March 1982, and the project was to reach full production in 1986. 2.11 Total project costs were estimated at US$34.7 million (TL 520.3 million) with an estimated foreign exchange cost of US$21.5 million (TL 322.1 million), or 62% of total estimated project costs. The proposed Bank Loan of US$21.5 million was expected to cover the total foreign exchange costs of the project, and was for a term of 17 years including 5 years of grace, and an 8.5% interest rate per annum. 2.12 LDP, through its Int-nsive Dairy Production Division (IDPD), would have overall implementation responsibility for the project. IDPU would carry out all the required technical functions, and the Encouragemc.nt and Development Loans Department (EDLD) of TCZB would handle the project's credit aspects. IDPD would be supported by 3 internationally recruited technical specialists, as well as by veterinarians of the Directorate of Veterinary Services (DVS) of MAFRA. III. PROJECT IMPLEMENTATION A. Start-up 3.01 The project had a slow start. In the first year of implementation, from February 27, 1977 through March 1978, only 78 subloans had been prepared by LDP, compared to the 115 subloans expected to be - 28 - approved by TCZB at appraisal. No disbursements had been made either by TCZB or by the Bank compared to an estimated Bank disbursement of US$6.0 million. 3.02 According to the supervision report of May 1977, written several months after effectiveness, the staffing situation was satisfactory. Two of the three internationally recruited technicians were in post and the third was expected shortly, a staff training program had been drawn up, and a farmer training program was soon to follow. 3.03 It is not clear why subloan preparation and processing and disbursements lagged so far behind appraisal estimates, given that this was a repeater project and that staffing levels were satisfactory. Several reasons given by supervision missions included the relative inexperience of the staff in preparing farm development plans; low morale because of low levels of remuneration and consequent delays in preparing the plans; lack of interest by farmers because of rising investment costs and a fear of stagnant milk prices because of marketing problems, and the limitation of TCZB's disbursements to 25% of building costs and exclusion of forage crop establishment costs (this was thought to affect the smaller 5-cow farmers in particular). Nevertheless, it appears in retrospect that appraisal estimates of start-up time for this larger project, which included smaller farms and thus additional constraints compared to the First Livestock Project, were overly optimistic. Subsequent implementation eventually caught up, in physical terms, to appraisal estimates (para 3.07 and Table 1). B. Revisions 3.04 Within the first year of the project, when farmers seemed reluctant to borrow, the Bank and LDP thought that several constraints on TCZB's lending policy were responsible. As a result, the Bank agreed in October 1977 to change the Project Agreement to allow for the possibility of more than 25% of the cost of the buildings being financed by TCZB, and for the possibility of financing incremental costs of labor, forage cropping and feedstuffs (for feedstuffs, only the first year's costs would be eligible). The number of farm plans did increase after that, but it is likely due to a combination of factors (speedier loan processing, for example) rather than these changes alone. 3.05 By March 1982, when the project should have closed, only US$13.4 million out of US$21.5 million had been disbursed. At that point the number of subloans and the amounts disbursed by TCZB in local currency exceeded appraisal estimates (Table 1). In order to disburse the remaining dollar amounts from the loan, LDP and Bank supervision missions suggested disbursing against livestock fattening (cattle and sheep) and dairy projects with local purebred cattle and sheep, as opposed to disbursing only against imported cattle or dairy projects as the Loan Agreement was then written. Supervision missions pointed out the financial attractiveness and viability of investments in local purebred or crossbred cattle and sheep, and in July 1983 the Bank agreed to amend the legal documents to allow disbursements against these investments. - 29 - 3.06 By March 1984 funds not disbursed amounted to about US$5 million. But the demand for livestock loans was not yet satisfied. Also TCZB had under its large lending volume for livestock development numerous subloan applications which qualified for Bank financing. Therefore the Borrower and the Bank agreed that MAFRA would participate in TCZB's large lending volume for livestock development which amounted to US$200 million equivalent for CY 1984. In order to satisfy legal stipulations (Schedule to PA) requiring IDPD to prepare farm plans the agreement included that MAFRA would approve subloans prepared by TCZB before they were sent to the Bank for disbursement. These arrangements greatly facilitated the disbursement of project funds. The project was fully disbursed by November 1984, eight months after the revised closing date of March 31, 1984 which had been extended from March 31, 1982 to allow for the implementation of the additional types of credits described above as well as completion of importation of cows and equipment as foreseen at appraisal. C. Physical Implementation 3.07 The physical progress is shown in the Implementation Chart and Table 1. The pace of subloaa approvals and TCZB disbursements picked up in 1978 and the project proceeded smoothly, with minor problems, as far as these aspects were concerned. 3.08 In 1980, however, the project was rated as a problem project (from August 1980 through June 1981), although in terms of subloans approved and TCZB disbursement it was doing well compared to appraisal estimates: 75%-80% of the appraisal target had been met on the number of subloans and number of cattle imported. (Over 100% of estimated TCZB disbursements had been made, although the financial progress on TCZB disbursements was due to domestic inflation). Bank disbursements were only about 50% of appraisal targets during this period. The August 1980 supervision report noted that the quality of subprojects was not high and that a large number of farmers (225) had withdrawn from the project after cattle and machincry had been purchased by the project for further distribution. The report stressed the inadequate supervision and extension work by LDP, inadequate support from the veterinary service (DVS) for disease control, inadequate training of LDP staff, and inadequate budgetary support to LDP, and noted that the pool of imported high quality cattle was not being used properly as an asset because of inadequate breeding and selection programs. 3.09 This situation continued into 1981 with increasing numbers of farmers withdrawing from the project (146 in a six-month period) and financial difficulties by farmers reported as a result of poor feeding practices, increased costs of farm operations relative to farmgate milk prices, and problems of milk collection and marketing. - 30 - Table 1: Physical Implementation Nov. July Aug. Nov. Oct. Nov. Aug. Sept. 1977 1978 1979 1980 1981 1982 1983 1984 1. Subloans under implementation /1 (TL million) Appraisal 163 213 315 419 469 469 469 469 Actual /2 - 13 322 696 1,246 1,551 1,753 2,950 Actual in 1977 prices 11 166 141 177 173 162 184 2. No. Subloans processed /1 Appraisal 230 310 475 660 750 750 750 750 Actual /3 78 195 320 531 760 840 861 1,024 3. Cattle Imported /1 Appraisal 2,090 4,180 6,270 8,360 10,450 10,450 10,450 Actual 352 1,838 5,071 7,125 8,315 9,876 10,385 /1 Cumulative, TCZB disbursements. /2 TCZB disbursements, which .hould represent 80-85% of total cost of subloans. /3 Operational; more subloans were approved than became operational because some farmers withdrew after TCZB approval 3.10 By the end of 1981, however, project progress was reported as satisfactory again, with the number of farmers withdrawing from the project decreasing sharply and the financial profitability of the farms improved with increases in milk prices. The same problems of LDF staff and extension service were noted but, with the increased profitability of the farms, the Bank was optimistic that farm productivity would be maintained. 3.11 Objectively, based on the pace of subloans approved by TCZB and of Bank disbursements, the progress of the project improved continuously after 1978 in terms of physical and financial implementation. The success of the project, however, was periodically judged on more subjective but equally important criteria which included the impact of the project on institution building and on developing a nucleus of high quality cattle that could be effectively exploited to upgrade the livestock sector (see paras 4.01 and 4.08 for discussion on institutional performance). 3.12 By the end of the project in November 1984, 1,024 subloans had been approved by TCZB for imported cattle compared to 750 at appraisal; 10,400 cattle had been importei and distributed to farmecs compared to 10,450 at appraisal, and 160 fattening subprojects had been approved for -31- 9,050 head of cattle and 6,700 head of sheep, which had not been included at appraisal (see Table 1 for figures on imported cattle for dairy subprojects). The distribution of dairy farms was as follows: Table 2: Dairy Farms by_Size: Actual vs Appraisal Subloans Approved to Actual Appraisal Small farms /1 563 300 Medium farms /1 256 200 Large farms /1 205 250 Total 1,024 750 /1 Small: 5-10 cows; medium: 11-20 cows; lar e: over 20 cows. In spite of initial caution on lending to small-scale farmers and calling this portion of the project a pilot scheme at appraisal, a higher number and proportion of 5-10 cow units were approved and operational than expected. D. Procurement 3.13 Ptocurement under the project was handled efficiently. Altogether 10,317 breeding heifers and 68 breeding bulls, costing the equivalent of US$11.1 million, were imported from West Germany, Israel, Austria, Holland, and Denmark. Purchases were based on three quotations from at least two countries, as stipulated in the Loan Agreement (Schedule 4, Section A, para 1). On several occasions, however, Bank supervision missions reported that imports had an unduly high proportion of poor quality cattle. The project also financed the importation of various equipment (tractors, forage harvesters, mowers, milking machines), and bidding interest for these items was satisfactory. Supervision reports noted that changes in make of machinery between bulk purchases created some problems in machinery utilization and maintenance. In addition, failure to synchronize bulk purchases of farm machinery with importation of cattle resulted in out-of-phase on-farm development where the cattle arrived before the farms were prepared to receive them. The appointment of the three international consultants was based on direct negotiations satisfactory to the Bank (para 4.10). E. Pro ect Cost 3.14 The borrower was unable to provide information on total project costs for the local cost components (for dairy farms this includes the buildings and farm establishment and the entire costs of the livestock fattening units). The mission estimated these costs based on livestock development plans, which in turn were projected costs rather than actuals, and on periodic supervision reports that contained farm budgets and costs. The following paragraphs on project costs and financing are therefore the mission's best estimates made on the basis of available information. - 32 - 3.15 The total project cost was almost ten-fold greater in Turkish lira terus than estimated at appraisal, but lower in dollar terms by 2% (see Table 3 below). The reason for the cost overrun in Turkish lira was that domestic inflation was much higher than projected at appraisal.!' In addition, project changes towards the end of the project included disbursements for livestock fattening subprojects and the lira value of these additions represents over 65% of the project cost in lira terms. In dollar equivalent, the value of the livestock fattening component was roughly 40% of the project cost.-' At the same time, the lira devalued by over 1,700%, from US$1.0 = TL 19 in 1978 to US$1.0 = TL 310 in the first quarter of 1984, when the project was completed (see Table 3, FN 5). In fact, the unit costs in dollar terms of the imported cattle and much of the imported equipment were below the base unit costs, before contingencies, estimated at appraisal. 1/ 1977 1978 1979 1980 1981 1982 1983 %---------------- p.a.----------------- Projected inflation 12 12 12 10 NA NA NA Actual domestic inflation (GDP deflator) 24 44 70 102 42 28 28 2/ Very little information was available from the Borrower on this component and these figures have been derived from a combination of information from TCZB and from the Bank's disbursement records. - 33 - Table 3: Total Project Costs: Appraisal vs Actual Component Appraisal /1 Actual % of % of US$ Total US$ Total % Cost Million Cost Million Cost Overrun Dairy Farms Imported cattle 16.5 48 11.1 33 -33 Equipment 9.9 29 5.3 16 -46 Buildings 7.4 21 2.3 7 -69 Farm establishment /2 - - 0.6 2 Livestock Fattening /2 - 14.1 41 Technical Services /3 0.5 1 0.4 1 -20 Training/Studies /4 0.4 1 0.2 1 -50 TOTAL PROJECT COST 34.7 100 34.0 100 -2 TL Million Equivalent /5 520.3 4,978 857 /1 All contingencies included. /2 Not included at appraisal. /3 -Foreign technical specialists only; excludes local staff. /4 Only cost of overseas training included; excludes costs of local training. /5 Appraisal exchange rate: US$1.0 = 15.0 TL. Actual exchange rates: 1977: 18.0 TL; 1978: 24.3 TL; 1979: 31.1 TL; 19S0: 77.0 TL; 1981: 111.6 TL; 1982: 162.0 TL; 1983: 232.1 TL; 1984 (March): 310.0 TL F. Financing 3.16 The financing plan shown in Table 4 is based on mission estimates (para 3.14). The original financing plan in the appraisal report showed that for the farm development costs the subborrowers would contribute 18% of the costs of the dairy farm development as included in the appraisal estimate, while the subloan would be 82%. The Project Agreement stipulated in the Terms and Conditions of the subloans (para 4.1(iv) of the Schedule on Lending and Operating Policies) that an individual subloan, which would consist of TCZB's disbursement and World Bank's portion of the financing, would not exceed 80% of the total cost of the farm development plans. The difference in the figures comes from the incremental costs of labor and forage crops, included as part of the development plans but not shown in the appraisal report's costings of the project. From the information submitted by the Borrower, the subborrowers' contribution is estimated at about 15% of the total, excluding the labor and forage crops, and is thus close to appraisal estimates. For total project costs, including the costs of technical services and training, the World Bank portion was about 63%, as against 62% at appraisal. -34- Table 4: Financing Plan Appraisal Actual Sub- Sub- borrowers TCZB IBRD Total borrowers TCZB IBRD Total ------------------------US$ million--------------------------- Dairy Farms Livestock 0.5 1.1 14.9 16.5 - 0.5 10.6 11.1 Equipment 0.2 4.0 5.7 9.9 - 0.1 5.2 5.3 Buildings 5.6 1.8 - 7.4 1.7 0.6 - 2.3 Farm establishment - - - 0.6 - - 0.6 Livestock Fattening - - - - 2.8 6.2 5.1 14.1 Subtotal 6.3 6.9 20.6 33.8 5.1 7.4 20.9 33.4 (% of subtotal) (18) (20) (33.8) (15) (22) (63) Technical services - - 0.5 0.5 - - 0.4 0.4 Training/ Studies - - 0.4 0.4 - - 0.2 0.2 Total Project Cost 6.3 6.9 21.5 34.7 5.1 7.4 21.5 34.0 (% of Total) (18) (20) (62) (100) (15) (22) (63) (100) --------------------------TL million------------------------- Total Project Cost /1 94.5 103.3 322.5 520.3 752.0 1,067.0 3,159.1 4,978.2 /1 Contains rounding errors. G. Disbursements 3.17 Throughout the project, disbursements lagged behind appraisal estimates. One year after loan effectiveness (from February 1977 through March 1978) no funds were disbursed, as against disbursements of US$6.0 million estimated at appraisal. Except for one quarter, disbursements were never more than 50% of appraisal estimates and by the original loan closing date of March 31, 1982 only US$13.4 million out of a total of US$21.5 million had been disbursed. The loan closing date was extended to March 31, 1984 and the final disbursement was on November 19, 1984. At that date the loan was fully disbursed. The estimated and actual disbursements schedule is at Annex 1. 35 H. Covenants 3.18 The Borrower's performance in complying with covenants was mixed. For example, on staffing, the Borrower was required to employ three internationally recruited technical specialists who were to assume their duties no later than January 1, 1977 (Loan Agreement, Section 3.01(b)). Two of the three specialists were in post by this date and the third was in the process of being hired. When their contracts expired in 1979-80, however, the Borrower was slow to replace them. Two of the three were replaced by other consultants towards the end of November 1982, but the replacements were not issued work permits until March/April 1983. On local staffing, the Borrower was to employ within IDPD 18 incremental technicians in its fiscal year 1976 and another 16 technicians in 1977 (Loan Agreement, Section 3.01(c)). According to supervision reports, the 18 staff were employed in 1977 and the additional 16 were employed only in the Borrower's fiscal year 1982. It is unlikely, however, that this delay in hiring additional local staff would have made a significant impact on project progress, as even when the staffing levels were satisfactory, supervision reports noted that lack of experience, low morale and inadequate incentives were major problems throughout the life of the project. 3.19 On reporting, the quarterly reports were submitted in a timely manner but the audit reports for LDP (Loan Agreement, Section 4.03(b) and Supplemental Letter No. 3) and for TCZB (Project Agreement, Section 4.01(c) and Supplemental Letter No. 3) have been sent with one to two year delays. The files contain an LDP audit report only for 1982. The TCZB audit report for 1984 was not received as of October 1985.- Although LDP kept a separate account of Project funds, this involved only funds received from the World Bank used for training and technical assistance under the Project, which amounted to about 1% of project costs. Separate accounts were also maintained by TCZB for World Bank funds onlent under the Project, as stipulated in the Project Agreement (Section 4.01(b)). The Borrower, however, did not keep records of the total cost of the Project as stipulated in the Loan Agreement (Section 3.06(b)), and thus it was difficult to estimate total project costs and financing percentages, including farmers' and TCZB's contributions.z/ 1/ The 1983 TCZB report sent to the Bank was in fact translated excerpts from the full audit report. 2/ The Project Agreement (Section 4.01(b)(iii) stipulated that TCZB should register in its accounts the amounts provided by the Borrower or TCZB from their own resources to finance the subloans. These amounts were not clearly identified in the translated excerpts we received of the accounts. 36 - 3.20 The Special Operational Fund (SOF) to be established by the Borrower for use by LDP (Loan Agreement, Section 3.04) and to be financed by TCZB out of credit repayments (Project Agreement, Section 4.02) was established, but TCZB payments into the account did not correspond to credit repayments from the loan proceeds and LDP did not make use of the SOF as was intended.- 3.21 TCZB did not inform their subborrowers of the amounts charged to their accounts and TCZB/IDPD staff did not supervise regularly the use of the subloans, as stipulated in the Project Agreement (Sections 5.1 and 6.2 of the Schedule). Supervision and extension activities of both TCZB and LDP were inadequate throughout the project (para 4.06) 3.22 IDPD did prepare a training program for staff as stipulated in Section 3.07 (b) of the Loan Agreement and it was submitted within the time required, although farmers were not included as was intended. No mention was made in supervision reports of IDPD's submission to the Bank of plans for technical studies as required under Section 3.07 of the Loan Agreement, so it must be concluded that the annual plans were not drawn up and the studies, with the exception of one study on dairy marketing, were not carried out (paras 4.19 and 4.20). This result corresponds to the 16% of all studies that experienced major problems as a recent review in EMENA countries revealed which also states that "Turkey has by far the greatest number of projects experiencing major problems with studies." 1/ If payments to the SOF were to be made from repayments of subloans under this project, as is clearly stated in the Project Agreement (Section 4.02), then they would not have begun until some time in 1982-four years after the first disbursements--and would continue as long as subloan repayments are made to TCZB. Correspondence indicates that the Bank wished to change this arrangement to cover repayments on all outstanding livestock subloans made by TCZB but there is no indication in the files that TCZB explicitly agreed to this. In any case, the information from LDP and TCZB shows that TCZB credited funds to SOF starting in 1978 and LDP has reported that the inflow stopped when the Loan closed in 1984. LD? did not make use of the fund until 1980 and then it was either for certain per diem expenditures or for unexpected or extraordinary expenses that could not be met from the reguiar LDP budget; no evidence exists either from LDP or from supervision reports, that the funds were ever used on a regular basis to supplement LDP or IDPD staff salaries as a form of incentives, as was the intention. 2/ Review of current status of project-related Technical Assistance in EMENA Countries, Memorandum of October 3, 1985. - 37 - IV. INSTITUTIONAL PERFORMANCE A. Institutional Design 4.01 LDP, through IDPD, was the executing agency of the project. Regional control was carried out through offices in Ankara, Istanbul,. Izair, Adana, Konya, Kayseri, and Malatya, each headed by a Manager, and staffed with a Deputy Manager, at least one Agronomist, one Veterinarian, two or three Field Assistants and support staff. IDPD was responsible for promoting the project, preparation of dairy development plans, supervision of project implementation, provision of extension services to project farmers, and maintaining records on project performance. 4.02 Three internationally recruited technical specialists, responsible to the LDP Director General, were to provide technical advice and training of field staff during the duration of the project and to assist in design and supervision of technical studies. One each was covering the areas of Istanbul/Izmir, Ankara/Adana/Konya and Kayseri/Malatya. Credit aspects of the project were handled through TCZB's Encouragement and Development Loans Department (EDLD). Subloan applications were submitted to IDPD and TCZB. After determination of a farmer's creditworthiness and credit limit by TCZB, a dairy development plan was prepared by IDPD and submitted to the TCZB branch office. 4.03 This arrangement for credit approval is, in retrospect, questionable. TCZB relied on LDP to develop farm plans and indications from supervision reports and the project completion mission are that the quality of these plans was not good and the preparation time for the plans was often quite long (para 4.05). 4.04 Because of this situation, LDP/IDPD's role as the preparer of farm plans has to be questioned: was it essential for the project or should TCZB have been allowed to do its own appraisal? An important question is whether the quality of subloans or loan performance made by TCZB under this project is significantly better than other subloans for livestock made by TCZB without LDP input. Information was not available to answer this question, but the mission Lelieves that the basic design of the project was overly cumbersome and demanding of LDP. Since the project closed, LDP and IDPD have been essentially disbanded, with their functions divided at the Provincial level between Project and Statistics (for farm plans), Extension and Animal Health. Less emphasis will probably be put in the future on developing farm plans to submit to TCZB and more on the role of livestock technical staff as technical "advisors" to TCZB and as extension service to subborrowers. B. LDP/IDPD's Performance 4.05 The main functions of IDPD staff were to prepare livestock development plans and supervise implementation and provide general extension assistance; other functions included coordination with TCZB in making subloans and in recording and reporting on the details of the output from the farms. LDP was also to carry out a series of studies. In the - 38 - preparation of the livestock development plans, IDPD became more efficient, in terms of the amount of preparation time required, as the project progressed. Initial delays in IDPD of up to 4 months and 4.7 staff months per plan were reduced to 2.2 staff months per plan, but supervision missions reported this was still too high. The quality of the plans, as examined by the completion mission, is mixed, however, and many of the forecasts of yields and returns are overestimated. Supervision missions made several recommendations on how to improve the quality of the farm plans and reduce the preparation time, but these recommendations were made in 1982, only after four or five years of project implementation. 4.06 Coordination with TCZB on appraisal/credit approval has been satisfactory, in the sense that staff from the two agencies had regular contact, at the field level as well as at the center in Ankara. But extension assistance to subborrowers by IDPD and/or TCZB staff was virtually non-existent, due to lack of motivated staff, of transport and funds by IDPD and lack of staff by TCZB. Veterinary services were to be provided by staff from the Directorate of Veterinary Services (DVS), but supervision missions reported that these services were also inadequate. This in turn affected disease outbreak and overall animal health. The Bank made the point repeatedly throughout the project that LDP needed a larger budget to ensure adequate funds to cover supervision, extension, veterinary and reporting functions and although several Bank reports and information supplied to the mission by LDP indicate substantial increases in LDP's budget, the budgets were still insufficient to provide adequate incentives to maintain staff and allow for regular field visits. The establishment of the Special Operational Fund had no impact on this situation because it was not used to provide regular incentives to the staff (para 3.20 and FN). 4.07 Bank supervision missions, while commenting that the number of staff was adequace, noted the low staff morale throughout the project area due to inadequate incentives and supporting operating funds, and resulting in high staff turnover. As a consequence, several functions were not properly carried out. On recording dairy production data, a sophisticated system was developed in 1980, but LDP staff were not aware of how these data could be used to select genetically superior animals for breeding. As a result, the recording was reported by supervision missions to be inaccurate and sometimes incomplete. The failure to report accurate production data was also a problem for monitoring the project performance and for evaluating the results (paras 4.21-4.24). On the studies that were to be carried out under the project only one study on the dairy industry was carried out by consultants. The list of other production studies included at appraisal and referred to in Section 3.07(a) of the Loan Agreement was never done (paras 3.22 and 4.20). 4.08 In sum, LDP/IDPD's performance was not as good as expected at appraisal. Appraisal expectations were probably based on LDP's satisfactory performance under the Livestock II Project'" until 1978/79, after which performance began to deteriorate be-ause of budgetary and staff constraints. In retrospect, the Bank may have been overambitious in its 1/ PPAR No. 3517 of June 23, 1981. - 39 - expectations of staff who were inadequately trained, poorly paid and with little incentives to carry out the many tasks involved in introducing on a fairly large scale new breeds of cattle and a new technology to traditional dairy farmers. C. Credit Extension and Supervision 4.09 Supervision reports noted long delays in credit processing from the time the farm plans were sent from LDP to TCZB to credit approval, up to 64 days on average per farm plan. TCZB claims, however, that delays are not due to its own procedures but to LDP's processing times. In reviewing the entire procedure, from the farmer application to approval, the mission found that the chain of steps was quite long and involved both LDP and TCZB, so it is difficult to trace exactly where the delays occurred. Within TCZB, branch offices can approve farm plans for credit only below a certain amount, an amount which has been raised from time to time and also varies depending on the branch and the sector. (For livestock loans, the limit is now roughly TL 2 million (US$2,860). Above this limit and up to TL 10 million (US$14,300) the application is sent to the head office in Ankara and approved by TCZB staff, and above TL 10 million the application must go to TCZB's Credit Committee.) During this project, as many as 50% of the applications were probably processed at the branch office. Whether or not delays occur at the branch and head offices of TCZB, the staffing situation of the bank is inadequate in terms of numbers of technically qualified staff to appraise loans. In spite of a policy to supervise loans twice a month, consistent supervision of loans is not possible and TCZB relies almost entirely on LDP staff for this function. There is little evidence, however, that LDP's supervision is adequate considering the experience under this Project (paras 4.08 and 4.18). Technical input needs to be increased in appraisal and supervision of subloans in future projects. D. Technical Specialists 4.10 The Loan Agreement (Section 3.01(b)) stipulated that three internationally recruited technical specialists would be employed and in post by January 1, 1977. The two technical specialists employed under the Livestock I Project agreed to a two-year contract extension under the Livestock III Project, while the third technical specialist arrived in Turkey in September 1977, six months after Loan Effectiveness. The supervision reports did not specifically mention the consultants' performance except to rate their performance on the supervision form as being essentially problem free. In addition, when their contracts expired (in 1979 and 1980) and they were not replaced, the Bank made the case that project momentum and quality control decreased after their departure and that they should be replaced as soon as possible. 4.11 The Bank finally made it a condition of extending the Loan Closing Date for two years that LDP would hire six technical specialists for the Third and Fourth Livestock Projects combined by July 1982.1/ By then LDP 1/ A second condition was the submission by July 31, 1982 of a training program for technicians and farmers, acceptable to the Bank. This was also done. - 40 - had agreed with an internationally recruited consulting firm to supply four expatriates and two local technicians acceptable to MAFRA and the Bank. The Bank extended the loan Closing Date until November 30, 1982 and a further extension was contingent on contract signature and employment of the six technicians, at least four of whom should have started to work. As of the end of November 1982, two of the four expatriate consultants were in post and LDP had accepted the two other expatriate candidates who were due to start in January 1983, and two local specialists had been identified. The Bank agreed to extend the loan Closing Date to March 31, 1984. 4.12 The two consultants were not issued residence or work permits for about four or five months, although they were already in Turkey. They were not allowed to travel outside Ankara or to discharge their duties until March or April 1983, after the Bank had intervened. The other two expatriates who had been hired were not issued visas to go to Turkey until sometime in 1983 (between March and June), again after Bank intervention, and by then one of the two consultants was no longer interested. The contracts of two of the three consultants expired in March 1984 when the project ended, although the loan remained open until November 1984. The two local specialists were never hired (no mention is made of the two vacant positions in the files or in Bank reports after February 1983). 4.13 It is difficult to evaluate how effective the technical specialists were. LDP's field staff reported that the specialists were highly effective and supervision missions felt that they provided momentum and quality control to many of the activities (livestock development plans, records keeping, veterinary services, training programs) and that these activities suffered in their absence. 4.14 In spite of this, however, the Ministry appeared reluctant to replace the two technicians and, even after replacing them with two others, LDP did not provide a welcoming atmosphere in which they might be most effective. Overall this experience corresponds well to EMPTA's findings.' according to which "Turkey has by far the greatest number of projects experiencing major problems with consultants." In the future, it will be important to design expatriate participation in a more effective way, perhaps by part-time participation and/or frequent visits, as presently tried under the Erzurum Rural Development Project (Loan 2094-TU), so that the technical assistance can make a proper contribution. E. Training 4.15 According to Supplemental Letter No. 7, training was to be given priority in the project and would comprise: (a) In-service training of LDP staff at all levels in animal nutrition, animal breeding, crop production, agricultural machinery, general farm management, farm planning, financial analysis, accounting and recording. 1/ Review of current status of project-related Technical Assistance in EMENA Countries, Memorandum of October 3, 1985. - 41 - (b) Farmer training, particularly field days, oafarm demonstrations and visits to State farms. (c) Group tours for LDP staff and prominent fare leaders, to be coordinated if possible with trips for cattle procurement (not less than five such study tours in the course of the project); and (d) Study leave for individual LDP or Government staff for specialized overseas training in advanced dairy production and technology--such leave to be practically-oriented studies of direct application to the developing dairy industry in Turkey. The Project Technical Specialists were to play a major role in organizing training programs which were to be effective by the beginning of the second year of the project. Thirty man-years of local professional staff time had been allowed in the staffing estimates to provide for absences of staff in training and for staff time needed for planning and executing training programs for farmers and middle-level Project staff. 4.16 Although LDP submitted an ambitious overseas and local training program to the Bank within six months of effectiveness, the training component was only partially implemented. LDP did not want to send its staff abroad for training, mainly because of the language barrier. LDP felt that, in the past, professional staff with limited knowledge of English sent to the United States for a Forage Crop Course were handicapped by insufficient English. In light of this experience, the design of future projects should explore the possibility of uverseas training for better educated foreign language-speaking Turkish trainees, who could then provide appropriate training in the country in Turkish. LDP also felt the knowledge on livestock farming available within Turkey was more than enough to educate the most traditional farmers in the country. As a consequence, LDP placed more emphasis on extension and demonstration activities in the country than on external training (para 4.17). The latter was confined to cattle buying missions abroad at which several regional or headquarters staff participated. This exposure probably had only little impact on the ability of relevant staff to stimulate improved dairy practices in Turkey. The Bank periodically encouraged LDP to send its staff on overseas courses and occasionally sent information on specific courses being held. This bad no apparent impact on LDP's reluctance to send staff abroad. 4.17 Despite LDP's relatively greater attention to internal training, achievements in this area were also below expectations both in quantitative and qualitative terms. Although LDP finally established its own Training Division in early 1980, three years after the start of the project, and planned to provide close to 500 man-weeks of staff training in 1981, it had actually provided only 30 man-weeks by the end of 1981. Training courses held and numbers trained were reported in quarterly and semi-annual reports from LDP but Bank supervision missions paid sporadic attention to the content and frequency of training and the composition of trainees (staff and farmers). - 42 - 4.18 In retrospect, given that one of the main objectives of the project was to build up a modern dairy infrastructure including an effective extension service, the training component should have been a critical element in achieving this. It was however only one element out of many which included, inter alia, preparation of farm plans, staff morale and salary structure, lack of operating funds, TCZB credit policies, and internationally recruited expatriates, on which LDP and Bank staff spent time and energy. As a result, training did not have the central focus it needed to be effective. This result differs little from the one under the First Livestock Project-" which had a shortfall in the formal training of project staff and from the recent analysis of EMPTA-' according to which projects in Turkey show the least output of training as well as the least achievements of training objectives. F. Technical Studies 4.19 Supplemental Letter No. 7 states that two types of studies would be financed under the loan. (a) Production studies, to include surveys and practical research and demonstration, particularly on forage production and utilization and improved livestock husbandry. Priority would be attached to (i) forage species and cultivar adaptability; (ii) fertilizer responses; (iii) forage utilization; (iv) animal health; (v) feed requirements; (vi) animal breeding; (vii) early weaning; and (viii) costs of production. These studies were expected to be done mainly on project farms. (b) Milk marketing studies, which would include surveys and studies on the coordination of production, cooling/collection, transportation, processing, pricing and marketing. In addition, a review of milk production potential and its relation to processing facilities and market outlets would be given priority, and ways to accelerate dairy cooperative formations in Turkey would be analyzed. 4.20 Although one supervision mission noted in May 1981 that the studies had been virtually completed, no evidence exists in the files that either the annual plans or the production studies under (a) in para 4.19 above were done. A milk marketing study was done, however, by a Turkish consulting firm, but the study was later incorporated--considerably revised and expanded--in a larger milk marketing study financed under the Fourth Livestock Project (Loan 1586-TU). 1/ PPAR No. 2542 of June 12, 1979. 2/ Review of current status of project-related Technical Assistance in EMENA Countries, Memo of October 3, 1985. -43 - G. Monitoring and Evaluation (M&E) 4.21 The Borrower was required (refer LA Section 3.06) to monitor and evaluate the progress of the project by preparing quarterly progress and evaluation reports. Neither contents nor details had been defined for these reports, probably because the milk recording system introduced under the First Livestock Project!" (Cr. 236-TU) worked well in monitoring onfarm milk production. Throughout project implementation the Bank received - and never asked for more than - quarterly information that reported on progress of preparation and approval of subloans. The penultimate supervision mission's (March/April 1984) proposal to the Borrower to include in their project completion report on-farm production data was toc late Lo prepare adequate infotniation for the Bank's PCR. Therefore the completion mission had to estimate project results with the help of farm plans that were updated by MAFRA and seemed to best represent project results. 4.22 Despite the findings of the PCR of the First Livestock Project (Cr. 236-TU) which concluded that a reasonable balance had been struck in allocating staffing between monitoring and field development work, the M&E performance of this project was inadequate, and corresponds to OED's general findings regarding agricultural projects.-' Specifically, the M&E requirements were not clearly defined (see para 4.21), Bank staff paid little attention to M&E during supervision, and there was insufficient borrower support. These shortcomings point to the need to build M&E into the project design at preparation stage, as was eventually done in the Erzurum Rural Development Project (Loan 2094-TU), although not in the repeater Livestock IV and V projects (para 4.23). At a minimum, ongoing and future projects should require regular reporting of project output performance as part of regular progress reporting, with content and details well defined and agreed with the Borrower. 4.23 In the two ongoing projects Livestock IV (1586-TU) and Livestock V (1862-TU) requirements for M&E are basically the same as under Livestock III (1265-TU) with the difference that (i) detailed records are to be maintained on the implementation of not less than 20% of the farm development plans; and (ii) semesterly progress and evaluation reports due to be prepared and sent to the Bank within two months after the end of each semester. Except for the poultry component of the Fifth Livestock Project (1862-TU) for which detailed information has been received, including information on production and income of poultry farms, the Bank has .eceived only general progress reports on farm plan preparation and approval, procurement, training and staffing. Detailed records on the implementation of farm development plans were neither available nor received for either of the two projects (1586-TU and 1862-TU). 1/ PPAR No. 2542 of June 12, 1979. 2/ Report No. 5781 of June 28, 1985: Built-in Project Monitoring and Evaluation: An Overview. -44 - V. OPERATING PERFORMANCE AND PROJECT IMPACT A. Technolqoical Change 5.01 The general standard of onfarm development and management appears to be satisfactory, although on several farms too much reliance is still being placed on the feeding of concentrates rather than forage crops. Farmers in the Istanbul, Izmir and Adana regions have more easily adopted the modern dairy farm technology advocated by the Project, than those in the other regions. Silage production has been introduced into the Istanbul, Izmir and Adana regions. Few farmers have however taken up silage making in the Central Anatolian regions, where a high proportion of project farms have little or no experience in growing forage crops and LDP technicians frequently lack relevant experience. Because the number of farmers taking subloans is about 37% greater than anticipated at appraisal (para 3.12 and Table 1), the project's impact on introducing technological change should be greater than expected at appraisal. B. Bio-economic Production Efficiency Parameters 5.02 The estimates in Table 5 below on production efficiency parameters are taken from livestock development plans that LDP sent to the Bank in 1984. Table 5: Production Parameters Appraisal PCR Assumption /1 Estimate Calving rate (%) 95 80 Mortality rate calves (%) ns 8 Mortality rate adults (%) 7 4 Culling rate cows (%) 30 20 Milk production per cow per year (kg) 3,000 - 3,800 3,000 - 4,000 Milk sold per cow per year (kg) 2,600 - 3,300 2,500 - 3,400 Cull cow liveweight (kg) ns 400 Concentrate feed per cow per year (t) ns 1.5 ns = not specified /1 At full development. C. Incremental Production 5.03 The estimates of incremental production shown in Table 6 derive from the completion mission's calculations based on the parameters shown in Table 5 above and on the data in Table 1 showing the number of farms established per year during the project. * 45 - Table 6: a 1A 1m 1I=Q 1Wi 12A 19A 19& 1Wm i 1i im8 Total milk sold (tons) 7.509 8.189 12.810 17,568 18.800 21,727 25.205 25,033 27.359 27.646 Total milk home consumed (tons) 899 966 1.50S 2.067 2.207 2,550 2,961 3.029 3,206 3.239 Total incremental milk production (tons)41 8,408 9.155 14.315 19.635 21,007 24,277 28,166 28.862 30,565 30,885 Total cull cows L 952 -.182 -196 436 -26 438 780 424 682 816 Total cull heifers LZ 238 -105 7 109 -32 13 69 41 - - Total in-calf heifers - - 252 1.008 644 802 1,120 1,050 1,171 Total young breeding bulls - 714 637 910 1.384 1.306 1,487 1,797 1,737 1,819 Total young fattening bulls -238 224 238 269 601 641 676 911 937 991 Total calves 357 21 21 174 39 39 123 - 1 Excludes milk to calves, which is estimated at roughly half the amount of home-consumed milk. LZ Since the local cattle are sold in the first year to be replaced by the pure bred, and culled pure bred cows are not sold until year 4 of a subroject, negative numbers of incremental culled cows are the result of ne culled cows minus pre-project culled local cows. The amount uf incremental milk from project development is less than 1% of the total milk supply in the country and the numbers of culled animals for slaughter are likewise unimportant compared to the national totals. The real impact on the project in terms of output, however, can be considered to be the annual sales of in-calf heifers (about 1,200) and breeding bulls (about 1,800) which should add considerably to the national nucleus of pure bred, high yielding dairy cows, the size of which is difficult to estimate but may now consist of roughly 15,000-20,000 animals. D. Marketing and Prices Milk 5.04 Throughout the project's life, there have been reports on marketing difficulties, reflecting the general difficulties of the dairy industry. 5.05 Supply. Most of the total national supply of 5-5.6 million tons of raw milk produced annually since 1975 come from a large number of small herds. The average milk yield of cows from the counry as a whole was 581 kg/year in 1981, compared to 780 kg/year ;n Iran, 1,860 kg/year in Greece and 2,690 kg/year in Bulgaria. The generally small-scale and scattered nature of suppliers affects the structure and composition of the dairy industry. 5.06 Demand. Over half of the milk produced for human consumption is consumed in rural areas, on farms and within villages, and never enters the market network. Average per capita consumption of milk and milk products was estimated at 95.4 kg in 1982, and in urban areas the annual per capita - 46 - consumption was estimated at 131 kg for 1982. The most popular dairy product in Turkey is yogurt and a diluted form of yogurt (60% yogurt, 40. water) called ayran. Drinking milk and yogurt account for between 40% and 60% of the final uses of raw milk.!" 5.07 Milk collection and processing.. Marketed milk has traditionally been purchased at the farm gate by small-scale peddlers and traders for sale as raw milk in nearby urban areas or for processing into cheese, butter, yogurt and .ayran. Many small scale processors, called mandiras, are found throughout the country, close to the sources of milk supply, and are either privately owned or private sector cooperatives. These two groups, the traders and mandiras, together handle about 80% of the raw milk marketed in Turkey. In addition, a modern dairy sector has developed in the last 20 years, comprised of public and private sector factories. The latter are generally established in or near large urban areas, while the public sector (mainly the milk corporation TSEK)' factories are spread across the entire country. 5.08 The capacity utilization of the modern dairy sector has been low, particularly for TSEK, which averaged about 30% over the period 1980-83.-!/ In spite of this, TSEK capacity has steadily expanded as investments in new processing plants have been made. Established in 1963 to serve the needs of dairy farmers selling surplus milk and of consumers to supply good quality product at reasonable prices, TSEK has consistently lost money and been in a serious financial situation for the last 12 years. The losses reached 30-54% of th- sales revenue (for every TL million of revenue, the company lost TL 500,000) in 1976-80. Since then, the losses were reduced to 10-15% of revenues because selling prices were allowed to be increased. 5.09 The reason for the losses is partly due to low capacity utilization. The farmgate purchase price established by Government is often insufficient to prevent traders and peddlers from outbidding TSEK. The selling price, also established by Government, is below market price. TSEK is required to supply institutions such as the military, hoapitals and schools, and in fact has little to sell on the market. Their impact on either the purchase price from farmers or on the general level of prices for dairy products is thus minimal, but their mandate on fixed prices and establishing an extensive network results in low capacity utilization and constant financial losses. 5.10 The mandiras and private traders, on the other hand, serve the rural sectors with a relatively efficient marketing network. They are unregulated, in terms of prices and thus are able to pay farmers higher prices than TSEK. They are also unregulated in terms of hygienic practices 1/ The estimates >n the distribution of final uses of raw milk vary widely, depen ng on the source of information. 2/ The project : d not provide support for TSEK. 3/ For the pr' s , exact figures are not available. - 47 - and quality of milk and milk products delivered, however, and this aspect of their operations may need further attention from government, they do nevertheless provide an effective marketing network for dairy farmers. 5.11 The problems of the dairy industry have been the subject of a study carried out by consultants, financed under the Fourth Livestock Development Project (Loan 1586-TU), which has made recommendations for both the traditional (peddlers, mandiras) and modern (particulary TSEK) sectors emphasizing investment in training rather than equipment. The study, still in draft, is under discussion with the State Planning Office and MAFRA. 5.12 Prices. The prices paid to farmers for raw milk vary widely across the country and by season, as well as, for TSEK purchases, by fat content. It is difficult therefore to use an average price of milk to calculate average returns to a typical farmer. The trend in milk prices, however, across regions and seasons, kept pace with inflation. Since appraisal in 1976, the farmgate price of milk has increased from 3.5 TL/liter to 70 TL/liter in 1984, or twenty-fold, which is about the same as the general price increases during this period. Meat and Live Animals 5.13 During the project period there was no report of difficulties in selling animals for meat. In addition, however, the market for full bred heifers and bulls was particularly strong during the project period and as the prices offered for the pure bred cattle increased some project participants sold their herds to make large nominal capital gains rather than continue to run their farms. The prices in Table 8.show that between 1982 and 1984 the prices for pure bred in-calf heifers and pure bred bulls doubled or tripled, again generally in line with domestic inflation and devaluation. 5.14 Supply. As with milk supply, most of the meat supply comes from a large number of small producers, concentrated in the eastern part of the country. Carcass weights are low, but have been increasing; between 1975 and 1982, average carcass weight for cattle increased from about 100 kg to about 128 kg. Red meat production has shown a steadily increasing trend, from about 475,000 tons in 1973 to 816,000 tons in 1982. In 1982 about 130,000 tons of red meat was exported to the Middle East and North Africa. 5.15 Structure of Meat Processing Industry. Together with municipal abattoirs, the State Meat Organization (EBK)-' had a monopoly on slaughtering meat from 1952, when EBK was established, until 1980, when private slaughterhouses were allowed to operate. By 1982, private sector slaughterhouses accounted for about 7% of total "controlled" slaughter of cattle, which inciudes slaughter in EBK, municipal, private and religious slaughterhouses and live exports. The breakdown in 1982 for cattle, sheep and goats was as follows: 1/ The project did not provide support for EBK. - 48 - Table 7: Slaughter of Livestock, 1982 1982 Actual 1982 Capacity Cattle Sheep & Goat Cattle -------('000 head)------ ('000 head) (% utilization) EBK 369 2,119 1,336 28 Municipal 2,706 7,949 4,089 66 Private 231 428 704 33 Religious 136 2,647 - Live exports 142 2,454 - Subtotal 3,584 15,597 6,129 Uncontrolled 398 - Total 3,982 21,077 Source: Union International Consultants: Livestock Products Development Study, 1985. The total capacity of the slaughterhouses was, however, much higher: for large animals total capacity was estimated as 6,130,000 heads, and EBK, the municipal and private slaughterhouses together were operating at about 30-40% capacity utilization. Some of this excess capacity is to be expected because of the recent entry of private sector companies; thirty-one have been established since 1980. Table 8: mula_LPrict.f_rnlk.JAaaLAnimals UAppraisa -282L 1983 ____28 1976 (Nov) (Aug) (Apr) (Nov) Milk (TL/kg) 3.5 35 41.5 so 70 Liveweight sale Cattle (TL/kg) na 240 222 320 420 Cull cow (TL/head) 6,500 70-90,COO Cull heifer 6.000 40-60,000 In-calf heifer pure bred 9,500 80-185,000 125-150,000 Z1 220-250,000 Z_ 270-300,000 L Breeding bull pure bred 10,000 130,000 123-136.000 /1 200-225,000 Ll 225--250,000 /1 Cull calf 3.000 25-30,000 Ll The higher price in for Holstein; the lower price for Brown-Swiss. - 49 - VI. FINANCIAL AND ECONOMIC RESULTS A. Financial Results 6.01 Because no systematic surveys were carried out in the project, the physical outputs had to be taken from farm models. These farm models are based on information from LDP field staff, which is, in turn, their best estimate of the results. The rates of return based on these data are thus not founded on good survey data but on the only estimates available to the mission./ Financial and economic analyses are summarized in Annex 2. Financial rates of return were calculated for a 40-cow model based on parameters from the farm model referred to above and 1982 prices, and the result was a 20% rate of return compared to 21% estimated at appraisal. Sensitivity analysis shows that if all benefits were 20% lower, the rate of return would drop to 11% and, if operating costs were 20% higher, it would go down to 13%. B. Credit Recovery 6.02 On credit recovery, according to TCZB, as of December 31, 1984 out of a total amount of subloans outstanding from this loan of TL 2,054.4 million, TL 184.6 million or about 9% of the outstanding amount were under legal action, which means that they were more than one year overdue and another TL 146.3 million or 7% were overdue. Looked at another way, however, repayments in 1984 were only TL 164.3 million: thus the total amount due in 1984 was the repayment plus the overdue amount (TL 164.3 million + TL 146.3 million = TL 310.6 million) and the overdue represents 47% (TL 146.3 million divided by TL 310.6 million) of the amount payable. Almost half of the repayments due in 1984 from the subloans were not paid. 6.03 Interest rates followed the pattern of similar TCZB lending in the agricultural sector. Until March 1980 the interest rate remained at the 11% established at the beginning of the project, but thereafter variable rates were applied to all new and outstanding subloans as follows: TCZB Rate of Interest on Agricultural Credit February 1977 to March 1980 1i% March 1180 to September 1980 18% September 1980 to January 1982 24% January 1982 to January 1984 22% January 1984 to Janua-y 1985 28% January 1985 to August 1985 30% Since August 1985 30% for loans under TL 2.5 million 32% for loans TL 2.5 million and over 1/ It can be noted that the revisions under this project, which expanded its scope, as well as two subsequent livestock projects have been justified on similar information. - 50 - Nevertheless, these interest rates were consistently negative in real terms. In consequence, agricultural lending, including the loans under the project, has required direct and indirect subsidization. The Government had stated its intention in connection with SAL V to move progressively to achieve positive real interest rates for agriculture by the end of 1986. C. Economic Results 6.04 The main economic benefit of the project, in addition to the incremental milk supply and some animals for slaughtering, is the long term provision to other farmers of surplus improved dairy stock for breeding. The economic rate of return to the entire project, estimated in constant 1977 prices, is 19% compared to 22% at appraisal.L' Sensitivity tests show that with a decrease of 20% in prices of milk and breeding cattle or an increase of 20% in operating costs, the rate of return falls to 13% or 16%, respectively. VII. BANK'S PERFORMANCE A. Project Formulation and Objectives 7.01 As a repeater project of the First Livestock Development Project, the concept was appropriate in terms of size (number of subloans) and scope (number of Provinces covered). The appropriateness of the institutional design, however, must be questioned. Preparation of individual farm plans by LDP staff was deemed necessary because the Bank thought that TCZB was technically weak. This meant long delays in farm plan approvals and a preoccupation by LDP with training staff to prepare the plans. More emphasis should have been put on strengthening TCZB's ability to prepare and appraise farm plans, and on training LDP staff more relevantly in extension matters and providing them with the means to supervise the farms in operation. Secondly, apparent lack of commitment on the part of the Government to carry out the studies financed under the project, and inadequate supervision by the Bank led to studies which were never done or which, when done, were eventually incorporated into another study. Finally, the appraisal should have provided more explicit guidelines for establishing a systematic data collection and analysis, and the uses to which such data could be put. B. Supervision 7.02 Project processing was efficient and Bank supervisions started early (the first one was in July 1976, before effectiveness) and continued at regular six month intervals. Missions had appropriate staff and staff continuity, but the focus of the missions and the judgment on project progress shifted throughout the project period. In the first two years, the missions were positive and reported satisfactory progress on subloan processing, staffing and farm operations. For a period during 1980/81, the focus changed to criticism of LDP's effectiveness, quality of farm plans and farm operations. Thereafter, without any factual evidence that 1/ The economic analysis includes only the dairy component as too little information was available on the livestock fattening component. - 51 - progress had improved (except that the number of farmers withdrawing from the project decreased), supervision missions again rated the project's progress as basically satisfactory. Several missions were thorough in their reviews and reports but only one, in February 1983, recommended any basic changes in project design. 7.03 This 1983 mission finally recognized that conditions had changed and that local purebred and crossbred cows were as financially attractive as imported ones. They recommended that the Loan Agreement be changed to allow financing of locally purchased dairy cattle, and cattle and sheep fattening subprojects. This argues for allowing for more flexibility at the beginning of a project so as to maximize the chances of achieving the objectives. In this case, the decision to finance livestock fattening subprojects from the loan can be questioned as to how this enhanced the dairy subsector; but the basic lesson is that to achieve an upgrading of the livestock subsector, a project should consider a range of means to accomplish the same end. VIII. CONCLUSIONS AND RECOMMENDATIONS A. Overview 8.01 The project's physical goals were generally attained. They included the establishment of a nucleus herd of purebred dairy cattle and the introduction to small scale farmers of modern dairying practices, including forage production and use of feedstuffs. The institutional goals were attained to a less satisfactory degree. The objective of building up an effective extension service with back-up veterinary services was not met. LDP was hampered by inexperienced staff and staff turnover, and by serious budgetary constraints. This in turn affected the quality of extension service and the quality of the dairy farms. It has also limited the extent to which the nucleus herd is fully exploited for its genetic potential. As in the case of other TCZB loans, negative real interest rates were subsidized by the Borrower and, while profitable to TCBZ, made the project expensive for the Government. There were no incentives for TCZB to use Bank funds in preference to its own funds, for which it did not have to meet Bank requirements for farm plans and reports. 8.02 While individual subprojects are likely to continue, the sustainability of the overall project as it was designed is questionable. The Borrower's commitment and institutional capacity would need to be greater, and the project design more balanced and more effective in its technical and institutional dimensions. The late and inadequate provision of recurrent resources including personnel, funds and equipment showed the extent of Borrower's commitment. MAFRA's institutional capacity remained weak, partly as a result of loss of salary competitiveness vis-a-vis the private sector. The recent reorganization of MAFRA, decentralizing responsibility to the provinces and merging various extension staffs, should help to create an extension service better equipped to deal with farmers who normally combine crop and livestock production. However, it will make livestock-only projects more difficult to coordinate through Ankara. The project design in retrospect, would have been better if it had used MAFRA staff primarily for training and extension programs and - 52 - subproject monitoring, while building up technical capacity for subloan appraisal by the credit institutiuon, rather than using MAFRA staff for developing farm production plans. Bank supervision needed to be oriented more towards course correction and getting the Borrower to act appropriately on supervision missions' recommendations. The Bank should consider re-evaluating the financial and economic results of the project once the required production data that are being collected are available. An Impact Evaluation would be useful especially since two more livestock projects, the Fourth (1586-TU) and the Fifth (1862-TU) are under implementation and should close in December 1985 and June 1987 respectively. B. Lessons Learned 8.03 One important lesson learned from the shortcomings of this project design is that subloans should be appraised by the credit institution, which would enable the te'chnical institution (MAFRA) to concentrate on subloan supervision and extension service. If the credit institution is technically weak, then the project should be designed to build up the capability of that institution, through training and technical assistance if appropriate. While this strengthening is taking place, it may be necessary to involve the technical institution in the appraisal process as a temporary expedient. Interest rates on subloans need to be adequate to at least cover the credit institution's costs. 8.04 If the objective of future livestock projects is to improve extension and veterinary services, then Government must show support of this objective by providing adequate recurrent resources to pay salaries, travel allowances and transport (vehicles, fuel, repair and maintenance costs). The Bank and the Government agreed to establish a Special Operational Fund to provide special incentives to IDPD staff. This arrangement was designed to correct inadequate salary remuneration and was not successful. One important lesson from this and other- experiences is that, without adequate government support in the form of regular, sufficient budgetary allocations, special funds and exceptional arrangements are not satisfactory substitutes. In addition, international specialists need to be recruited as stipulated and be provided with required work facilities. 8.05 Furthermore, the project should be designed to encourage staff training and the Bank should be concerned with the specific content of in-service training components and, if appropriate, overseas courses. This livestock project did include these components and supervision missions frequently commented on the desirability of overseas courses, but the Bank did not get involved in the content of the local training courses, concentrating more on the number of staff weeks being provided by LDP. In addition, LDP was reluctant to send staff overseas, except on cattle-buying missions. Thus, the effectiveness of the training component is questionable. In the future, if one objective is to provide institution building, then it would be important for the Bank to be involved in every aspect of a program designed to do this, and for the Government to be fully committed as well. 1/ Refer PPAR No. 3517 of June 23, 1981 on Livestock II (Credit 330-TU). - 53 - 8.06 Both the Bank and the Borrower should be prepared to be flexible in implementing credit projects designed to provide subloans involving new technology. This loan was designed to finance only imported dairy cattle and only direct foreign exchange costs at a time when dairying was profitable and had been successfully improved on a small scale in a previous project and at a time when foreign exchange and access to imported cattle were serious constraints. Starting in 1980 Government's policy brought about dramatic changes internally in relative prices, and in the relative attractiveness of meat production. It was only after several years, however, and near the end of project implementation, that project scope and flexibility were expanded to allow domestic animals and equipment and livestock fattening activities to be included. Other ongoing pfujects incorporated this lesson and allowed a wider range of sources (local, imported) as well as a broader range of livestock in general (dairy cattle, cattle for fattening, as well as smaller livestock). The advantage of this approach is that as returns to farms change in different subprojects during the course of the project, credit would continue to be extended on the more attractive subprojects. Under this project, when dairy subprojects became unattractive for a period of time, the project slowed down until changes were incorporated. 8.07 Monitoring requirements should be defined more precisely. Quarterly progress reports could meet the requirements if they were detailed enough and included production and income data of representative farm samples. These reports could also facilitate the evaluation of the project's impact. Supervision missions should follow-up more closely on such requirements. 8.08 Finally, studies should be included in projects when the implementing agency itself is keen to carry out the study, which will result in clear practical recommendations for actions or policy measures. Otherwise, the inclusion of studies results either in the frustration of the Bank, unspent loan amounts and unnecessary commitment fees because they are not done, or in a waste of funds because they are done and not used. This should be avoided in general (and not just in livestock projects). Once studies are included in projects their implementation needs to be adequately followed through by Bank supervisions, with respect to reviewing carefully the terms of reference for the studies, the selection of consultants, their progress, the final outcome of the studies and the uses to which they are put. -Ann4 1 TURKEY THIRD LIVESTOCK DEVELOPMENT PROJECT (LOAN 1265-TU) PROJECT COMPLETION REPORT Actual and Appraisal Cumulative Disbursements IBRD Fiscal Year Actual Appraisal and Quarter Disbursements Estimate ---------- US$'000-------- 1977: July-Sept. 0.7 Oct.-Dec. 1.5 Jan.-Mar. 2.3 Apr.-June 3.1 1978: July-Sept. 4.1 Oct.-Dec. 5.0 Jan.-Mar. 6.0 Apr.-June 0.4 7.1 1979: July-Sept. 0.5 8.1 Oct.-Dec. 3.1 9.2 Jan.-Mar. 3.6 10.3 Apr.-June 3.8 11.4 1980: July-Sept. 5.0 12.6 Oct.-Dec. 5.3 13.8 Jan.-Mar. 5.4 15.2 Apr.-June 7.7 16.7 1981: July-Sept. 9.5 18.2 Oct.-Dec. 9.9 19.8 Jan.-Mar. 10.3 20.7 Apr.-June 11.0 21.5 1982: July-Sept. 12.9 Oct.-Dec. 13.0 Jan.-Mar. 13.4 Apr.-June 13.9 1983: July-Sept. 14.5 Oct.-Dec. 14.6 Jan.-Mar. 15.4 Apr.-June 15.4 1984: July-Sept. 15.6 Oct.-Dec. 15.7 Jan.-Mar. 15.9 Apr.-June 16.2 1985: July-Sept. 20.2 Oct.-Dec. 21.5 Note: Fully disbursed on November 19, 1984. - 5 - Annex 2 w Page 1 Tg LIES'QK DVELQPMET PROJECT (LOAN 1265-T) PAQJECT CO"LTION .PORT ECONOIC RATE OF RETURN TL IILLION 1977 1978 1979 1900 1981 1982 1983 1984 1985 1986 1987 1988-1997 ECONOMIC COSTS Investeent Costs 7.1 70.3 110.7 73.5 43.4 47.7 31.6 0.4 - - - - Operating Costs 11.8 17.9 40.3 52.8 66.9 78.4 102.4 99.1 109.3 113.1 113.1 113.1 --- - -- ----- ----- ----- -- - - - - - - - - - - - ----- ----- TOTAL COSTS 18.9 88.2 151.0 126.3 110.3 126.1 134.0 99.5 109.3 113.1 113.1 113.1 ECONOMIC BENEFITS Milk - 61.7 54.9 51.5 75.1 91.3 114.6 108.3 111.0 117.6 118.7 118.7 In-Calf Heifers - 0.0 0.0 0.0 4.6 19.8 14.9 12.5 17.4 16.3 18.2 18.2 Dreeding Bulls - 0.0 16.7 10.4 11.4 26.4 23.2 20.4 24.7 26.6 25.0 25.4 Cull Cows - -2.5 -2.3 3.5 -0.2 3.7 5.6 2.0 3.3 3.9 4.1 4.9 Cull Heifer - 28.5 -1.0 .0 0.6 -0.2 0.1 0.2 -0.1 0.0 0.0 0.0 Benefit Fattening Bulls - -5.3 3.9 23,3 1.6 1.7 3.5 2.3 4.1 5.6 5.7 6.0 Culled Calves - 2.1 0.1 0.1 0.5 0.1 0.1 0.2 0.0 0.0 0.0 0.0 TOTAL BENEFITS 84.5 72.4 88.9 93.6 142.8 162.1 146.0 160.4 170.0 171.8 173.3 NET BENEFITS -18.9 -3.7 -78.6 -37.4 -16.7 16.7 28.1 46.5 51.1 56.9 58.7 60.2 ECONOMIC RATE OF RETURN 19.4% SENSITIVITY ANALYSIS Operating costs +207. Istarting 1986) 15.5% Benefits -20Z e 12.8% o.- o cr. i.-- . o ic ot 0 . .r' i . .: * i2i- '06 --, ,- . E: -:2 .:-: .3 . ' . - - . -ia. - x - 2 - a E . a -.. C=. g - : - 00 i ø' e :: 0000 g I : s 5 : = " : 2 - "t m i:" ,- O oO: O o= - 0 0: - : 0x Ot- O: Ce . 0 0 . :- i 0000x g:-g.: ), = g : : --g'. : : 1 x a : 3 - = . s*a e -et- - - a : e:--- e: :- - - - : -i TURKEY THIRD LIVESTOCK DEVELOPMENT PROJECT (Loan 1265-TU) Implementation Schedule CaLendoar Yeo 1976 1977 1978 1979 1980 1981 1982 1983 1984 Quoreri l 1 ij i N 1 iv Iv i e W i i tåi iv " 1.3 [ a.51 "uo 861 1.024 en en en>rb, 1 ~4 3 JUIJJimu aneaia en an et §linne ,se-,end da a 35it3 c/t 1s 1 98/S 1 0385 l i AAd.ul Uu,r~oAppfowlso ALTudIau . 1 taucunon4 ut Stud1d 1. AppfaiSQl umsessesses Actual ~w,I( tko l 726 、〕引