Document of The World Bank FOR OFFIAL USE ONLY Report No. 5157 PROJECT PEdFtRMANCE AUDIT REPORT INDONESIA - TRANSMIGRATION I PROJEIET (LOAN 1318-IND) June 25, 1984 lDerations Evaluation Department This document has a restricted distribution ad may be used by recipients only in theperformance of their official duties. Its contents may not otherwise be disclosed without World Balk aulhoriation. ABBREVIATIONS BIMAS - Credit System for Agricultural Inputs BRI - Bank Rakyat Indonesia CRIA - Central Research Institute for Agriculture DGT - Directorate General of Transmigration FAO/CP - FAO/World Bank Cooperative Program GOI - Government of Indonesia KUD - Village Cooperative PCR - Project Completion Report PHU - Project Management Unit PTP X - Para Statal Enterprise for Rubber Plantation REPELITA - Five-Year Development Plan RSI - Bank Resident Staff in Indonesia SAR - Staff Appraisal Report FOR OFFiCIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT INDONESIA - TRANSMIGRAIIONI PROJECT (LOAN 1318-IND) TABLE OF CONTENTS Page No. Preface ............................................................ Basic Data Sheet .................................................... Highlights .............. ........................................... PROJECT PERFORMANCE AUDIT MEMORANDUM I. PROJECT SUMMARY .......................................... 1 II. AUDIT FINDINGS ........................................... 3 General .................................... 3 Problems and Constraints ................................ . 4 Main Issue ........................ 10 Annex I - Comments Received from the Minister of Transmigration .... 15 Annex II - Comments Received from the Director General of Livestock Services .................................... 17 PROJECT COMPLETION REPORT I. Summary and Conclusions .. 21 II. Background ............................................... 22 III. Project Formulation ...................................... 24 IV. The Project Scope. ......................... 24 V. Implementation ........................................... 25 VI. Economic Evaluation ...................................... 34 VII. Procurement, Disbursements and Project Costs .... .........35 VIII. Lessons Learned .......................................... 37 Tables 1-7 Map - IBRD No. 11316R1 (PPA) This document has a restricted distribution and may be used by recipients only in the performance of their officialduties. Its contents may not otherwise be disclosed without World Bank authorization. PROJECT PERFORMANCE AUDIT REPORT INDONESIA - TRANSMIGRATION I PROJECT (LOAN 1318-IND) PREFACE This is a performance audit of the Transmigration I Project in Indonesia, for which Loan 1318-IND in the amount of US$30.0 million was approved in July 1976. The loan was closed, 17 months behind schedule, and an undisbursed balance of US$23,039 cancelled on April 13, 1983. The audit report consists of an audit memorandum prepared by the Operations Evaluation Department (OED) and a project completion report (PCR) dated January 4, 1984. The PCR was prepared by the Bank's East Asia and Pacific Regional Office. An OED mission visited Indonesia in November/December 1983. The mission held discussions with officials of the Department of Transmigration and with evaluation staff of the Bogor Agricultural Institute (IPB). The project area was visited and project officials and farmer settlers inter- viewed. The information obtained during the mission was used to test the validity of the analysis and conclusions of the PCR. The audit memorandum is based on these discussions, on interviews with Bank staff associated with the project, and on a review of the PCR, the Staff Appraisal Report (No. 1119b-IND) dated June 10, 1976, the President's Report (No. P-1885-IND) dated June 25, 1976, the Loan Agreement of July 21, 1976, correspondence with the Borrower, and internal Bank memoranda on proj- ect issues as contained in relevant Bank files. A copy of the draft report was sent to the Borrower on March 5, 1984. Comments received from the Minister of Transmigration and from the Director General of Livestock Services are attached as Annexes I and II and have been taken into account in the preparation of the final report. The audit finds the PCR comprehensive and accurate with respect to the project's principal achievements and shortcomings and has no reason to question its conclusions. The audit memorandum deals with problems and con- straints related to project design, contractor performance, monitoring and evaluation, agricultural development, settler welfare and the post-settlement transition, and raises the issue of the suitability of the transmigration settlement model. The valuable assistance provided by the Government of Indonesia and by local officials and the farmers visited is gratefully acknowledged. sCr inowANC agDT ato Snauzsn - I PSZUC TnaugRAnXo. (CLO 1315-CD) sIc DM SWm Appraisal Actual or Actul - Z of Item stimate Katimtad Actual Apraisal Estimate Totea Project Coats CsS milon) 56.8 57.3 101 Loan Amount CUSSmillo) 30.0 30.0 LOO Diuburad 30.0 30.0 /a Cancelled - is Date Doard Approval - 77E5/76 Loan Agraement Date - 07/21/76 Data Effectivaeam 10/19/76 02/2B/77 233 lb Date Physical Compo_mta Coplated 0W/ll 10/B2 125 7T; Proportin Thb Completed zoo 100 Closig Date 11/30/81 04/13/13 ,127 /b Economic Rata of Ratzr (C) 12-19 le 17-22 /e 129 Institutioal Perfonce - Pair Agroncmc Perfor_mae - ain - Number of Direct BemeficiarieS (1983) 14,500 14.500 1W0 CIWLAfvZ DIS_U5SUDlTS F776 n77 1778 179 mrBo Ml 1r82 FT83 Appraisal estimte COSS milion) 1.0 6.5 16_0 23.0 28.5 30.0 - Actul (55 million) - .1.6 6.4 10.6 16.2 22.5 27.4 30.0 ActuaXl s C of estimte 0 25 40 46 57 75 91 100 Date of flna diboem t April 13. 1983 MISSION DATA Date No. of Handays Specializations Performnce Types of uiasSio (mo/Ir) Persons in Field ReDreamted Id Rating le Trend /f Problem /i identification 06173 2 Preparation 01/74 - Appria 11174 8 Follow-up 03)75 1 Appralaal Follow-up 09175 5 Appral-1 Follow 05/76 5 Supervision 1 09176 4 20 A.-,D 1 1 N.T SupervlsIon 2 08/77 2 4 A.3 2 1 n SuperviSLon 3 02J78 4 24 A,3,D,F 3 2 1 Supervision 4 06/78 3 9 A,11D 3 1 1 SuperviSion 5 (9/78 5 25 A,ZRD,C 3 2 X.F Supervision 6 0/79 5 40 A,C.D 2 2 X.F Supervision 7 01180 4 16 A.C,D 2 2 H.P Supervison 3 8 0/80 5 25 A,C,D,Z 2 1 M.F Supervision 9 05181 4 24 A,C,D 2 1 H.P Supervision 10 02/82 6 30 A,C,D 2 1 n.r Total Superwicion 217 oru PROJCT DATA Borrower Republic of Indonesia Ixcutiug Agency Directorate General of Transmigretion Fiscal Tear April 1 to Harch 31 Ease of Currency (abbreviation) Rupiab CRp-) Currency Exchange Rate: Appradal Ter Aerage US 1.00 - Rtp 415 latervening Tears Average US$1.00 - Up. 627 Completion Year Average 551.00 - Rp. 650 Follow-oe Project: ame Transmigration II Tranmigrartoon IIn TranSamgration IV Loan/Credit haber ta. 1707/Cr. 919-11D Ln. 2248-1CD Ln. 2ZUB-IRD Loan/Credit Amount (05$ m) 90.0/67.0 101.0 63.5 Date Board Aprowval 05/29/79 03/22/83 05/24/83 /a Actual disbursemet mas 5329,976,961; the balance of US$23.039 sa cancelled. 7i Caleclated in terms of oath from the date of hoard approval. 7; For Iaturaj and Way fblug ubareas, respectively. 71 A - Agronomlat; 3 - Agricultural Economist; C - Financial Aalyat; D - Clvil Engineer; K - ealtb Specialist; F - Procurement Specialist. /e 1 - Problerfrae or minor problems; 2 - Hoderate problems; and 3 - Hajor problems. 7-f 1 - Improving; 2 - Stationary; ard 3 -Deteriorating. 7 F - Financial; h - hansgeril2; nd T -Techancal. - iii - AUDIT REPORT PROJECT PERFORMANCE INDONESIA - TRANSMIGRATION I PROJECT (LOAN 1318-IND) HIGHLIGHTS The project was the first Bank support for the transmigration pro- gram in Indonesia and comprised subprojects in South Sumatra and Lampung. One subproject assisted the settlement of 4,500 transmigrant families at Baturaja, while the second aimed at the rehabilitation of an existing settle- ment involving 10,000 families at Way Abung. The project's main objective was to test strategies for the agricultural, social and economic development of upland transmigration settlements based on an integrated rural development approach. Project start up was delayed by seven months and the closing date was extended by one year because of weak management and procurement problems during the first three years. Thereafter the project gained momentum and met most of its targets at no additional cost. Targets for road construction, water supply, and agricultural production were exceeded. However, a minor shortfall was noted in the distribution of cattle, and difficulties were encountered with the monitoring and evaluation program. The project can generally be considered as having been successful in spite of some initial doubts. The Baturaja settlement showed that good agricultural results can be achieved on red/yellow podsolic soils provided farmers apply an adequate package of inputs and soil management techniques; some 27Z additional spontaneous transmigrant families were attracted to the area. Marketable food surpluses were produced. The resulting incomes, to- gether with future revenues to be generated by settlers' rubber plots, are expected to greatly exceed transmigrants' incomes at their places of origin. The development of cassava industries at Way Abung is a promising sign of the effect on secondary economic activities. Among the contributing factors to the favorable outcome were the project's rural development approach, its implementation through a Project Management Unit, the semi nucleated village design, the high quality of rubber plantations, and the health, educational and marketing infrastructure being established in the project area. The economic rates of return of 17Z and 22% for Baturaja and Way Abung, respec- tively, exceeded appraisal estimates. Other points of interest are: - a project specific management unit (PMU) has proven effective for the coordination of project activities. However this concept requires groups of settlements totalling 15,000 to 20,000 families to be cost effective (PPAM, para. 10; PCR, paras. 5.02-5.04, 7.02- 7.03, and 8.01); - iv - the provision,free of charge, of agriculturalinputs during the initial three years, as well as applied research and the agricul- tural extension services proved to be important factors during the start-up phase of the new settlement (PPAM, paras. 13, 19 and 22; PCR, paras. 5.18-5.23, and 8.01); the provision of one head of cattle per family has helped to alle- viate labor constraintsand increase crop production,although difficulties were encountered in the cattle distribution program (PPAM, para. 21; PCR, paras. 5.26-5.29); a single mixed farming model per settlement scheme has been insufficient to take account of the variable transmigrant attitudes and land conditions (PPAM, para. 19; PCR, para. 8.01); the results have shown that monitoring and evaluation should not be over ambitious to be implemented and used efficiently (PPAM, paras. 17-18; PCR, paras. 5.15-5.17); the construction of small earth das- surrounded by shallow wells has yielded an adequate vater supply for settlements with inade- quate groundwater resources (PPAM, para. 14; PCR, paras. 5.09- 5.11); poor contractor performance contributed to cost increases and implementation delays (PPAM, para. 16; PCR, paras. 5.07 and 5.10- 5.11); the developmentof village cooperatives has been disappointing and requires more extensive technical and financial assistance than has been provided (PPAM,para. 22; PCR, paras. 5.35 and 8.01); and the Baturaja settlement,due to its pilot nature, is not repre- sentativeof Government's overall settlementprogram. Its above- average agricultural and economicalresults, especiallyregarding cultivation of red/yellow podzolic soils, should be considered as potential levels achievable under intensive investment and tech- nical assistance conditions (PPAM, paras. 20 and 30; PCR, paras. 5.22 and 8.01). AUDIT MEMORANDUM PROJECT PERFORMANCE INDONESIA - TRANSMIGRATION I PROJECT (LOAN 1318-IND) I. 1/ PROJECT SUMMARY 1. Although transmigration has been taking place for decades in Indo- nesia, it has received priority Government support only since the early 19 70s. Growing inter-island disparities, i.e., overpopulation in Java and Bali and lagging economic development in the Outer Islands, have increased the urgency of transmigration. Unfavorable conditions for food production on the Outer Islands have been a major impediment to rapid settlement. 2. FAO was invited to assist in overcoming technical production con- straints in transmigration areas, and later, during 1972-74, the FAO/IBRD Cooperative Program undertook the preparation of this project' The project followed the rural development approach and was kept small in scale as it was intended to develop and demonstrate a viable pattern to be replicated else- where.2 / It was estimated to cost US$56.8 million. Appraisal took place during 1974-75, and a Loan (1318-IND) in the amount of US$30.0 million was approved in July 1976. The loan became effective in February 1977, four months behind schedule, and was closed in April 1983, about 17 months later than expected. A small undisbursed balance of US$23,039 was cancelled. 3. The objective of the project was to devise and implement two test schemes, one for the establishment of a new settlement at Baturaja, South Sumatra, and the other for upgrading an existing settlement at Way Abung, Lampung. The main features of the proposed Baturaja scheme were the conr- struction of 10 village centers with about 4,500 settler houses and adequate health and school facilities; the construction of roads and tracks; block- planting of 4,500 ha of rubber and maintenance for six years; distribution of about 4,500 head of cattle and of farm inputs; the establishment of agricul- tural credit, extension and cooperative services; and the provision for one year of food and other essential commodities. The proposed Way Abung scheme comprised the construction of roads and tracks; block-planting of 2,500 ha of rubber and maintenance for six years, benefitting about 5,000 settlers; dis- tribution of 5,000 head of cattle, benefitting another 5,000 settlers; construction of health and educational facilities; and expansion of agricul- tural support services. Technical assistance was to be provided for both schemes, and various studies were to be undertaken. General program support was also included. l/ Adapted from the PCR. 2/ The original and official title of this project was -Transmigrationand Rural Development Project'. In view of the follow-on projects, this project became subsequently known as -TransmigrationI-. -2- 4. Implementation of the project was in large measure successful, although, as in most projectsof this nature, therewere considerable diffi- culties,and more time was requiredthan expected. Project start up was very slow due to delaysin establishing projectauthorities, concluding agreements with cooperatingGovernment agencies, and recruitinga consultingfirm. After further slow progress during a learningperiod for project staff, the pace of implementation became satisfactory.In Baturaja,4,500 families were settledin 11 villageunits. In both schemes together, 840 km of roads were constructed or improved, exceeding the originaltargetby 50% and the revised target (1981)by 10%. The provisionof domestic water supply at Baturaja was successful only after the originalproposal,i.e., collecting rain water,was found inadequate and well constructionproblems related to unfavorable groundwater conditions and inexperiencedcontractorshad been resolved. Eventually, a combination of small earth dams and shallowwells yieldedample water, though of low quality for householdpurposes. At Way Abung, the con- structionof 2,170 shallowwells exceeded the target by 10%. The combined total of 51 established health facilities and schoolswas exactly on target. The planting of rubber trees proceededas planned at Baturaja,with 4,500 settlersreceiving1 ha each. At Way Abung, difficulties were encountered with land availability, and instead of 5,000 farmers receiving0.5 ha each, about 2,700 farmers were allocatednearly 1 ha each. Although the 9,300 cattledistributed represented a shortfall of only 2X compared with appraisal estimates, this program had suffered from inadequateholding grounds and excessivelosses of animals in transit. Despite the delays in implementa- tion, the total project costs of US$57.3 millionwere practicallyidentical to appraisal estimates. 5. The achievements to date are encouraging. Adaptive researchwork produced cropping systems which incorporate adequate soil conservation measuresand are financially rewarding. However,the high levels of cropping and labor intensityhave in some cases exceededsettlers'managerialcapa- bility and familylabor supply. The development of the residualarea within the 5 ha allocatedto each settler for crop productionhas thereforebeen slow, althoughthe total of over 5,800 ha broughtunder food crop cultivation exceeded original estimates. The experiencewith cattle has been mixed, mostly becauseof settlers' unfamiliarity with cattle. However,availability of animal draft power is reportedto be essentialfor the cultivation of the whole of farmers' tillable area as well as for the maintenanceof soil fertility. Rubber development, under the country'sparastatalenterprise for rubber plantation,has shown excellent results, and the expected yields exceed original estimates by about 20Z. The health and educational facili- ties established under the project are operational. The achievements to date with cooperative development have been less than satisfactory,and agricul- tural supportservices have been inadequate. Marketing,creditand agricul- tural extension all requireimprovement potential if the full agricultural of the project area is to be realized. Another weakness concerns road main- tenance, which has been highly inadequate. 6. Institutionally,the project had some innovativefeatures. The Project ManagementUnit (PHU), althoughinitiallyplagued by staffingdefi- ciencies,eventually performedwell and has been creditedwith successfully - 3 - overcoming the often intractable coordination problems inherent in Indonesia's public sector. The requiredtransferof responsibility from the transmigration authoritiesto regular,mostly local government, agencieshad only started,but there are indicationsthat this change-overwill not be accomplished without difficulties. Financialand staffing resourcesavail- able in the futureare likely to be inadequate,and the futureuse of general projectfacilities, e.g. officesand staff housing,is as yet undetermined. 7. The project'sbeneflts include increasedincomes for 10,000 farm familiesin Way Abung and the 4,500 settlersin Baturaja. In addition, about 1,200 spontaneoustransmigrant familieshave settledat Baturajaand are conr tributingto the local economy. As a result of the project, additional insight has been gained Into the complexities associatedwith the agricul- tural utilizationof the unstable,low-fertility soils of the project area. Project cost, as expected,was relativelyhigh, amountingto an average of about US$8,000per family at Baturaja,and US$2,200 at Way Abung. However, thesehigh costs were more than offsetby project-generatedbenefits,leading to reestimated economic rates of return of 17Z at Baturaja and 22% at Way Abung, exceedingcombinedappraisalestimates by 29%. II. AUDIT FINDINGS General 8. This project marks the outcome of Bank's first involvementin transmigration, which began with an attitude somewhere between excessive hesitancy and prudent caution. This attitudewas justified, however,in view of the multiple objectives of transmigrationprograms- economic development is only one, and perhaps not even the predominant,goal - and the absenceof proven and viable technical,economic and sociologicalmodels. Although transmigration had been ongoing for a long time in Indonesia,the determi- nants of success and failure had not been sufficientlyunderstood. This project made a contributiontowardsan improvedunderstanding of settlement problems. However, that this contribution was modest is reflectedin the fact that the three follow-ontransmigration projects supportedby the Bank (Credit 919-IND/Loan 1707-IND, Loan 2248-IND and Loan 2288-IND) are all designeddifferentlyin an effort to addressdifferentproblemsor searchfor and have made very uneven progress. new solutions, 9. While the scope for generalizing from this experienceis limited, within its own set of circumstances the project successfully capitalized on its innovative character. As problemsarose,anticipated or otherwise, these were eventuallydealt with, and the project graduallybecame known for its interimachievements. The key ingredients for this successwere the flexi- bility and coordinativetalent exhibited by project authorities. Whether it concerned such obstacles as unfavorable topography for locating village centers, lack of accessiblegroundwater and of experiencedcontractorsfor providingwater for the villages,or an underestimation of the difficulties of acquiring suitableblocks of land for rubber plantings,as was the case in -4- Way Abung, ilmaginative solutions were found and good resultsobtained. With- out effectivecoordinationof the many agencies Involved In this project, however, such resultsvould not have been possible. 10. A successfulfeatureof the projectwas its implementation through a ProjectManagement Unit (PMU). The PEU was establishedespecially for this project and actually encountered serious problems at the beginningwhich required substantialsupervisioninput. For about two years, staffing as well as operatingprocedureswere inadequate, and the projectprogressed dis- appointingly (paras.5.02-5.04and 7.02-7.03). But with effectivepersonnel in place and procedures streamlined,performancebegan to improve after 1978. The PHU vas particularlysuitable for coordinating the activitiesof participatingagencies. However, despite these encouragingresults, the Ministry of Transmigrationhas indicatedthat it is inclinedto abandon the PMU concept in future transmigration projects,as not being in line with Governmentpolicy. As such a move would eliminatea proven success factor, the PCR suggestion(para. 8.01 (a)) to retain the PMU concept at least for large schemes- the featuresof the presentPMU may be difficultto replicate widely - seems worthy of consideration. 11. The project's chance for sustainable successwas greatly enhanced by the very positive experiencewith its rubber component. Operationally, the involvementin the establishment and maintenancefor six years of the block rubber plantings of the experiencedestate company (PTP X) produced rubber tree stands with very good yield prospects. Conceptually, the exis- tence of the rubber trees is likely to be extremelybeneficialin the medium term for the settlersfinancially. Income from rubber trees should begin to flow at the critical time when off-farmemploymentopportunities created by the project itself are ceasing, and farm productionand permanentoff-farm work are far from fully developed. While there are still existingproblems requiringsolutions(para. 23) and equity considerations give rise to argur ments (para. 37), individualsettlers and the project as a whole stand to gain from the rubbercomponentas it was conceived and implemented. Problemsand Constraints 12. ProjectDesign. Severaldesign featuresof the project proved in- adequateor inappropriate.Among the most important of these was the size of holdings establishedunder the project which was 5 ha. In most cases this area exceeds present requirementsper family (para. 8.01 (g)). Effective manpower availabilitygenerally does not allow adequate cultivationof the whole area, either because the family work torce is too small, or partial off-farm engagementof the work force leaves an insufficientresidual to operate the farm. There are also indicationsthat intensivecultivation of 5 ha in Baturaja, i.e., with 2 ha under rubber and nearly 3 ha under food crops, will generate a potential family income far in excess of national standards. One advantageof providingmore land to the settlersis that it can easily be dividedup among settlers'childrenin the future. In fact, the potentialcarrying capacity of this land has already attractedinmigra- tion of additionalsettlers ('inrfixmigrants - mostly relativesof project settlers), who are sharingthe land. In view of the growingpressureon land - 5- use, however, an extension to thls project, financed under Loan 2248-IND, provides only 3.5 ha per settler. This allocation is still generous compared with traditional settler lots of about 2.5 ha, and the maximum of 1 ha per- mitted for spontaneous settlers. 13. The original project concept provided for the distribution of fer- tilizers, seeds and implements to participating farmers in the first year of settlement. This provision proved inadequate in the course of implementa- tion, and farmers were actually given support for three years. Officials associated with the project consider this an effective way of demonstrating to settlers that it pays to apply modern Inputs and to get them accustomed to their use. However, there are open questions In this connection. First, it is possible that this subsidy could have been reduced had the farmers' sup- port services been wore adequate (para. 22). Second, it Is open to conjec- ture to what extent the favorable agronomic results and optimistic conclusions in the PCR (para. 8.01 (f)) are the result of the three-year input grant. But regardless of what the answers to these questions are, the project has demonstrated the three-year grant to be effective in promoting food crop production under the prevailing circumstances. 14. Incomplete information on topography and other environmental con- ditions in the project area had several design implications. It was thus not possible to locate ten large enough sites which would have accommodated about 450 settler families each. Instead, the project had to establish 11 village centers averaging a smaller number of families. For this reason, and also because of general underestimation at appraisal, the proposed road network had to be expanded by over 50Z during implementation. Another - one of the biggest - design-related problems concerned the water supply component for Baturaja (paras. 5.09-5.12). While a groundwater resource constraint was recognized at appraisal, the suggested alternative, i.e., roof water collec- tion, proved inadequate. Avoidance of a potentially disastrous situation required nothing less from project management than to come up with a tech- nically ground-breaking innovation. This innovation, a combination of small earth dams and shallow wells, deserves careful scrutiny for possible replica- tion where similar hydrological and climatic conditions exist. 3/ 15. Another project proposal that ran into difficulties was the rubber component at Way Abung. It was assumed at appraisal that sufficient land would be available within settlers' aggregate entitlements for block-planting of 2,500 ha of rubber trees (para. 5.31). For various reasons, however, including a partial misreading of the situation at appraisal, implementation delays, and a lack of agency coordination and forward planning, It was diffi- cult to acquire blocks of land in suitable locations. Consequently, the allocation of rubber plantings had to be changed. Instead of 5,000 farmers receiving 0.5 ha each, only about 2,700 farmers were given on average nearly 1.0 ha of rubber trees; the target area of 2,500 ha was thus met. Project authorities attempted to compensate farmers not benefitting from the rubber scheme through preferential allocation of cattle and farm inputs. 3/ This scheme is already being adopted in Nucleus Estates and Smallholder (NES) projects. -6- 16. Contractor Performance. As has been the case in other projects4/ poor contractor performance has contributed to cost increases and implementa- tion delays. Specifically, inexperienced contractors, by not using casings, caused 140 shallov wells to collapse. Twenty earth dams were washed away as a result of faulty construction (paras. 5.10-5.11). Contractors were also used for road maintenance, and the experience was unfavorable (para. 5.07). Although inadequate supplies of construction materials and financial diffi- culties adversely affected the contractors' performance, inexperience and ineffective supervision by project staff were the prime causes. Aspects of contractor involvement have been shown by this project to require increased attention in future projects. 17. Monitoring and Evaluation. Careful provision was made under the project, including the employment of consultants, for monitoring and evalua- tion (part II.A of Schedule 2, and Section 3.03 (b) of the Loan Agreement). The objective was to follow the socio-economic development of the settlement areas and to identify project alternatives conducive to enhancing the partic- ipation of the settlers in the establishment of viable farming coumunities. Unfortunately, implementation of this component was not satisfactory, and the PCR correctly portrays it as the weakest (paras. 5.15-5.17). Project manage- ment was not able to arrange for the required studies to be carried out in a timely fashion. Initially, this component was delayed, together with the rest of the project, as a result of the PMU start-up problems (para. 10). Subsequently, there vas a shortage of funds for this purpose, and it took a long time to approve the consultants' contract. Complete agreement on the nature of the monitoring and evaluation exercise actually was never reached with the consultants - a team from the Institut Pertanian Bogor (IPB). Con- sequently, the consultants' work was of limited value as it covered only one year (1980/81), publication of the results was delayed, and the focus did not meet expectations with regard to, for example, identifying alternatives for reducing costs in future replications of upland settlements. Bank supervi- sions, being preoccupied with the implementation of the settlement component, were surprisingly ineffective in resolving conceptual matters for these studies and in expediting contractual arrangements, although the problem was reported as early as February 1978. Indications are that the supervision task was not made any easier by Government's reluctance to undertake the studies. In the audit's view, this is an area where the Bank's input, de- cisively delivered, could have chaniged the situation. However, despite these shortcomings, progress of the project is relatively well documented in the records of the PHU and some participating agencies. 41 See PPAR for the Agricultural Research and Extension Project (Loan 1179-IND) and the Natirzal Food Crops Extension Project (Loan 1267-IND), OED (forthcoming). -7- 18. The IPB team produced a series of quarterly reports, and a final report was issued in 1983.5/ The latter offers a picture of the project which, in some respects, is in contrast with other available information. It basically underlines the heterogeneity of the settlers' economic achievements and argues for more comprehensive and effective support to be accorded the settlers by the project. A major contribution of the project was to provide an opportunity for both the Indonesian transmigration authorities and the participating IPB staff to gain on-hand experience with outside evaluation of transmigration activities. While it is necessary to improve the focus and timely execution, the Bank should continue to encourage involvement of local agencies in monitoring and evaluation activities of projects in Indonesia. 19. Farming Aspects. The project generated impressive insights into the agricultural complexities facing newcomers to the area. The Central Research Institute for Agriculture (CRIA) conducted an adaptive research pro- gram over a six-year period and issued practical recommendations which are currently being field-tested for long-term adoption by farmers (paras. 5.18- 5.25). While an encouraging beginning has been made, the audit agrees with the PCR's argument that a single, predominant farm model is inadequate for meeting settlers' diverse requirements (para. 8.01 (d)). Farming systems will have to continue to be tested for their suitability to overcome farm families' labor constraints, to meet market requirements, and to be consis- tent with the need to preserve soil fertility. 20. The audit is advising caution, however, on the conclusion reached in the PCR that under certain conditions sustained food crop production on red/yellow podsolic soils has been shown to be feasible under this project (para. 8.01 (f)). First, a six-year test period is relatively short for assessing the long-term effects of continuous, high-intensity cropping. The results are still test results and not actual achievement levels of unsuper- vised farmers. The adequacy of cropping patterns, soil management systems and input levels - not too clear to begin with - is liable to change given the unstable soils in the project area, and economic limits may eventually be exceeded if, for example, rising input levels are required to maintain con- stant yields and soil fertility. Second, even if the above PCR conclusion is correct in principle, there is insufficient evidence that settlers wifl actually adopt and continue to apply adequate husbandry methods. As Govern- ment policies (i.e., price interventions, input subsidies) and market condi- tions change, recommended practices may not remain attractive to farmers and suboptimal practices may be substituted. Once podsolic soils are eroded it takes many years for the top soils to form again. The valuable lesson conveyed by this project, however, is that even if the long-term farming systems were to be based on perennials, transmigrants can become self- sufficient in food crops during the establishment period of the settlement and continue raising food crops at least as a subsidiary activity. 5/ IPB, Study on Evaluation of Long-term Monitoring of Baturaja and Way Abung Transmigration and Rural Development Project, Final Report 1980- 1982, July 1983. -8- 21. Experience with the project also shows that ownership of draft animals is beneficial in a farming system based on food crops. However, the results of the project's cattle distribution component were not entirely satisfactory, partly because of insufficient cooperation between the PHU and the Provincial Livestock Services Office. 6 / Initial losses were high due to transport deficiencies and lack of holding grounds (paras. 5.26-5.29). Farmers were largely inexperienced in the care and use of draft cattle; shortcomings in extension were especially felt in this area. The calf return program has been seriously behind schedule, partly due to poor breeding results. The utilization pattern of draft cattle raises the question of timing of cattle distribution. Utilization during the initial seasons was low due to the as yet small land area available for cultivation, lack of training, and farmers' off-farm employment. Moreover, a significant portion of the original recipients no longer own any cattle on account of misfortune or mismanagement. This experience suggests that the viability of settlements like Baturaja could possibly be improved by phasing the allocation of draft cattle in relation to an agreed expansion in cultivated land, making it con- ditional on a minimum cultivated area, or requiring several families to share an animal for some time. In contrast to cattle, small animals could be more easily cared for by the settlers and generated ready cash income. Increased ownership of small animals should therefore have been promoted more strongly. Aspects of trade-off or complementarity between draft cattle and small animals should be thoroughly reviewed. Such a review should also aim at determining the suitability, on a selective basis, of dual-purpose cattle instead of the draft breeds presently being distributed to new settlers. 22. Support Services. The development of support services was disap- pointing under this project. Agricultural extension, especially in animal husbandry, was deficient, greatly slowing the process of acquainting settlers with unfamiliar practices or improved methods. CRIA encountered problems in conveying its recommendations to the farmers. A major reason for these shortcomings lies in extension organization, which militates against an ade- quate concentration of extension coverage and effective coordination of participating agencies in the newly settled areas. Furthermore, the transfer of responsibility upon project completion from the transmigration to local authorities is prone to disrupt, rather than improve, extension activities. The development and operation of cooperatives has been most disappointing, and their contribution to date tb settlers' economic activities has been minimal. Under the circumstances, the future role of cooperatives in trans- migration projects needs to be reviewed. 23. The marketing function was assumed almost entirely by the private sector, and performance, while imperfect, has been encouraging. Farmgate as well as central market sales of farm produce have been commonplace, and mar- keting activities have been conducted regularly and frequently. The supply of goods to farmers, especially farm inputs, has been more limited, however. There is obviously much room for increasing marketing efficiency, and the 6/ Comments received from the Director General of Livestock Services are in Annex II. -9- best chance to accomplish this is through improved competition among traders and provision of better infrastructure. As long as transportation is diffi- cult and costly and institutional farm credit remains underutilized, large trade margins and seasonal price fluctuations will adversely affect farm incomes. A particular weakness exists with respect to the marketing of rubber. Although tapping is about to start, effective control of tapping standards is lacking, and arrangements for the collection and processing of latex have yet to be put into effect. 24. The promotion of agricultural credit has not been very successful. One explanation for this is that the high level of subsidies (i.e., inputs, food, capital grants in the form of settler and farming infrastructure) and some nonrcredit project features (i.e., calf return program), as well as the low initial level,of commercial farm production and continued use of tradi- tional trade credit adversely affected demand for institutional credit. Active promotion of institutional credit at an early stage most likely would have generated cost savings for the transmigration program and eased the transition from the sheltered construction phase of settlement to the normal operations following thereafter. 25. Settler Welfare. Overall, farmers in both Baturaja and Way Abung are relatively prosperous. Signs of material well-being are abundant, and the state of health and education is reported to be satisfactory. However, available data picture a diverse income situation. The PCR (para. 5.25) estimates that family income from food crops alone will eventually reach about US$1,500 per year. If two hectares of rubber were cultivated this would add another US$2,600 annually. The future income prospects for a Baturaja settler thus appear bright. However, incomes to date have been far more modest. According to the PCR, an average cropped area of 1.30 ha in 1983, under favorable conditions, produced a net family farm income of US$614, or US$472/ha. The IPB survey (132 farmers) showed an average total family income in 1980/81 of US$524, of which 52X was from farm production and 48% from off-farm sources. The farm income component, with an average area of 0.87 ha cultivated, was thus US$270, or US$310/ha. With adjustment for inflation, the IPB results (US$375) for on-farm incomes were thus about US$100 lower on a per-hectare basis than the PCR estimates, partially reflecting the latter's assumption of enhanced productivity through CRIA- assistance. 26. While it is rather certain that average incomes of project farmers currently exceed subsistence levels, there exists considerable variation among farmers. This variation is in part due td differential adoption by farmers of improved practices on account of weak extension services. Total family income group averages in the IPB study ranged from US$398 to US$720. A sizeable number of families must thus be assumed to be at or below subsist- ence income levels. This result is especially significant as the average daily calorie intake (1980/81) was found to be 1882, i.e., about 8% below estimated (standard) requirements; protein intake was 6% above recommended levels. There is a strong likelihood, therefore, that a portion of the set- tler population was inadequately fed at the time of project completion, while another portion was very well off. - 10 - - 27. It is interesting that the IPB sample farmers reported almost one- half of their family income to have been derived from off-farm sources. Moreover, returns to labor were higher for off-farm employment than for farm work. There was thus a strong tendency to seek off-farm employment, with the result that land clearing and agricultural development were relatively slow. This outcome reflects a rational short-term reaction on the part of farmers, but with the heavy public investments in an agror-based settlementany aspect that adversely affects the utilization of agricultural resources, especially land, must be carefully watched. Whether there actually was an imbalance between off-farm employment and or-farm development is difficult to say, how- ever, without further in-depth studies. 28. Aspects of Transition. It is Government policy that, as settlement units are established and the mandate of the transmigration authorities expires, public sector responsibilities are transferred to local authori- ties. This process has started for project entities, although the planned expansion financed under the Transmigration III Project requires continued use of some project infrastructure at Baturaja. The main questions arising, aside from the continuity of services to be provided to the settlement communities, concern the future use of project-specific facilities and the maintenance of roads. There are office buildings, guest houses, storage facilities, generators, vehicles, and a seed farm and tree nursery which will be partially or completely vacated with the departure&of project staff. The future use. particularly of buildings, is likely to create problems for design and locational reasons, and because the upkeep is costly. Unless a suitable use can be found - and that had not been the case with the facili- ties at Way Abung at the time of the audit mission - valuable assets created under the project will be wasted. 29. Potentially more serious, however, is the maintenance of roads cor- structed or rehabilitated with Droject funds. For major roads, provincial authorities' maintenance responsibility started at the time of completion of work on those roads. However, the only maintenance undertaken to date was by the project, and that was highly inadequate. Consequently, rapid deteriora- tion is making some stretches impassable. The economic cost, as a result of worsened transport conditions and the need to eventually rebuild rather than merely maintain the roads, is offsetting a growing portion of project bene- fits. It is reasonable to conclude from this experience that transmigration project roads should be designed to standards which most closely conform to the conditions prevailing in the Province where the project is located, and, with respect to buildings, the facilities provided should be useful beyond the implementation phase of the project. Main Issue 30. The Transmigration Settlement Model. The project was a rest case with organizational features and levels of investment which differed from the country's overall transmigration program (para. 8.01 (k)). As such, at issue is whether this model has been successful and can be recommended for replication elsewhere in Indonesia. There is no question that implementation has been on the whole satisfactory and the majority of settlers are reaping - 11 - sizeable benefits. However, the cost of the project, at nearly US$8,000 per family in Baturaja, has been high for dryland farming development, and the crucial question is whether similar benefits could be achieved at a lower cost to the Government. 31. It has been estimated that 1,200 spontaneous transmigrant families settled in the project area. These families are either relatives of spon- sored settlers with the same socio-economic background, or specially moti- vated independeinttransmigrants. The latter tend to be part-time farmers who possess some nonrfarm skills, have more financial resources of their own, and are economically especially active. These characteristics are desirable in a settlement community and should be promoted. If a project could be designed in such a way as to increase the number of spontaneous settlers in relation to sponsored transmigrants, then the cost to the Goverament of settling a given area could be reduced, even if some of the project services were extended to cover spontaneous settlers. One possibility would be to plan phased development around core communities or facilities, similar to the regional growth pole concept promoted by the Ministry of Public Works. 32. Relatively little analysis is available on the implication of settler composition with respect to place of origin, kinship, motivation to migrate, previous occupation, educational background and own financial resources. However, there are clear indications that these factors are crucial determinants of individual settlers' ability to adjust and progress, and of the dynamism of the settlers as a group. The greater the heterogene- ity the more difficult it is to match project services with individual requirements and the greater is the disparity in the levels of achievement. The better educated and financed transmigrants tend to become dominant groups. While the intrinsic advantage of the existence of such groups for the settlement community, i.e., entrepreneurial leadership and effective com- munication with authorities, needs to be preserved, exploitative tendencies have to be discouraged in order for a balanced community to evolve. 33. The degree of settler involvement in site and land development also has an effect on settlers' commitment and the rate of progress. While settlers initially should not be overburdened and left prematurely to their own devices, their early and extensive involvement is desirable. At Baturaja, settlers generally were not' involved in the selection of their site, the clearing of the homestead lot or the construction of their house. This is generally the case in sponsored settlements because involvement in these activities, though desirable, creates extra expense in temporary camps where health problems also could develop. Nevertheless, a review of proced- ures applicable to this early construction stage with a view to advancing the involvement of future settlers could eventually result in an improvement of the overall settlement quality. 34. Experience with this project has reaffirmed the importance of proper site selection. A multitude of variables need to be taken into account in the selection of village sites, and errors made in this selection are likely to be costly and have long-lasting effects. The Baturaja area - 12 - turned out to be more difficult than expected on account of adverse topo- graphic features and unfavorable groundwater conditions. Land acquisition became a problem in Way Abung (para. 15). A priori determination of site suitability must thus be standard procedure of any transmigration program. The support provided for this purpose under the Transmigration III Project, which is expected to be continued under the proposed Transmigration V Proj- ect, is therefore most appropriate. 35. Concerning the pace and pattern of development of a settlement, experience with this project implies that the chance of success tends to increase with increasing levels of investment. However, it is not possible to draw firm conclusions as to what the optimum level of public investment might be. Equally important to know would be what the minimum critical investment intensity is for a satisfactory take-off of a settlement's socio- economic activities. This is an area where the monitoring/evaluationstudies produced very disappointing results. It is difficult to say which project activities could have been curtailed and which others could usefully have been expanded or added. Preliminary indications are, as mentioned earlier (para. 21) that cattle distribution could have been more selective and phased over a longer period of time, while the keeping of small animals should have been promoted earlier. Given the level of maintenance, hard-surfacing of some of the roads was unnecessary. In view of the temporary nature of proj- ect management activities, the cost of project management infrastructuremust be considered high. Furthermore, as a general notion, in a settlement consisting prirarily of families with small children, high school facilities could be provided at a later date. As for underspending, the development of n-on-farmingactivities at Baturaja has been slow, and additional expenditures to stimulate such activities would have been appropriate. 36. Post-establishment interventions, or the overall phasing of public transmigration expenditures, are also of key importance. While project support for Baturaja was an expensive first-time effort, the activities in Way Abung represented second-phase investments. Public follow-on investments in previously under-financed transmigration schemes are currently gaining popularity in Indonesia. The question thus arises as to whether initial investments should be such that a settlement becomes immediately viable and sustainable, or whether additional phases should follow a relatively low- level initial investment phase. A comparison of the two project schemes shows that a second phase should not primarily serve to catch up with invest- ments that should have been made initially, but it also demonstrates that some investments can be advantageously deferred. For example, farm input support, improved services, and better roads and health facilities should have been provided much earlier at Way Abung. The planting of rubber trees and the distribution of cattle conceivably can be accomplished at a slower pace than other investments. Phasing of investments has the advantage of flexibility in responding to proven weaknesses and in avoiding public expenr- ditures that are not absolutely needed. The 'full-package Baturaja model promises a more rapid take-off, especially in an area where substantial location-specificproblems exist and the settlement is not a replication of a successful operation already being implemented elsewhere under similar condi- tions. But even in the Baturaja case it is not immediately known whether the - 13 - settlementwill be self-sustainingwithout further help, or whether follow-on investments are required. Accumulated experience to date is favoring a multi-phasedapproach,but with higher initialinvestments than were provided in traditionaltransmigrationschemes,and with more extensiveprivate sector involvement. 37. A difficultelementwith socio-economic and politicaldimensionsis cost recovery in a transmigration project. The attractiveness and replica- bility of high-cost schemes like the ones financed under this project could be increased through improved cost recovery. The levels of recovery presently achieved through the collectionof a land tax are minimal. In- creased use of the credit mechanismis, of course, an attractivemethod for reducing public expendituresin the first place. Current inputs are the prime candidatefor credit financing,but an essentialprerequisitefor such financing is the-presence of an efficient credit apparatus, of effective extension services,and of adequate input supply and marketing facilities. Capital investmentsalso could be subject, at least in part, to credit fi- nancing. The cattle distributionscheme providesfor recovery under a quasi revolvingfund principle: calves returned by the beneficiaries are immedi- ately allocatedto other farmers. The theoretical possibilityof eventually recoveringpublic expendituresthrough the sale of returned calves exists, althoughactual prospectsgiven past performance are not promising. 38. Cost recovery would normally be expected for capital investments in cash crops like rubber. To establishand maintainfor six years one hectare of rubber trees cost US$2,800-3,000under this project,i.e., more than one- third of total project cost per family in Baturaja. These investmentsare expected to generate about US$1,500 in incomes annually, according to the PCR. Such incomes, combinedwith substantialexpected earnings from food crops production, clearly create a capacity for repayment of investment capital to the public sector. Such repayments would be in addition.topublic revenuesgenerated,if any, at the processingand final marketingstages for rubber. While the cost of agricultural support services, social infrastructure and roads is not expected to be recovereddirectly from bene- ficiaries,investmentsto create private (e.g. housing) and productive(e.g. land, inputs) assets could be subjected to cost recovery provisions for equity and public financingreasons. The Baturajamodel has demonstrated the favorableimpact in a settlementscheme of ample grant financing. It has, however, not succeededin promotingcost-sharing by beneficiariesthrough an effective credit mechanism. Furthermore,it has not even addressed the question of cost recovery through levies or similar means, despite the fact that preliminaryresults point to potential income levels on the basis of which direct cost recoverywould be justified. 39. It is the conclusion of the audit that this project offers valuable lessons, although inadequatemonitoring has hampered thorough evaluation. Transmigration project investmentscan reasonablybe expected to generate attractive benefits. However, opportunities present themselves in future projects for improvingsettlers'economic incentives, making the investment package more effective,and reducing the overall cost of and the share of the public sector in Indonesia's "transmigration program. S ?IA t c% -7 g 'k,4 - /7/ - - 15- Ansna x Coments received from the Hinister of Traumizration ZCZCtJIST4185 UU1045 DrST _ REF : TCP FCA OEDDRZ UU1045 45123 TRANS JKT 45123 TRANS JKT HENTERI TRANSHISRASI R.I. N E n O R A N D U M NOMOR . 160 f H / V / 1984 .- TO. : MR. SHIR.S. KAPUR DIRECTOR v OPERATIONS EVALUATIQN DEPARTMENT THE WORLD BANK v WASHINSTON - DC. FROM : n A R T O NO MINISTER OF TRANSHIGRATION THE REPUBLIC OF INDONESrA RE. : PROJECK PERFORMANCE AUDIT REPORT TRANSHIBRATION I DATE : MAY 16 , 1984 RE YOUR LETTER OF 5 MARCH 1984 I FULLY CONCUR WITH unAFT PROJECK PERFORMANCE AUDIT REPORT INDONESIA TRANSMIGRATION I PROJECK C LOAN - 13138 - IND ) AND LOOK FORWARD FOR RECEIVING COPY OF FINAL REPORT. SIGNATURE N A R T O N O MINISTER OF TRANSMIGRATION * 601 CC. 1. HR. RADIUS PRAHIRDi MINISTER-OF FINANCE. 2. MR. SOENAS. 3. F I L E SEND BY : DJOKO DJATMIKO .9. 45123 TRANS JKT 158553 01710171 045 =05220530 NNNN I~~~~~~~~~~~~~~~~~~~ I ,.,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ i "a 'V1~ - 17 - Annex II DEPARTEMENPERTANIAN Page 1 DIREKTORATJENDERAL PETERNAKAN JAKARTA Telex 48125 JI. SalembaRaya 16 (Tromol pos; 402 TeleponNo 883935 Comments received from the Director General of Livestock Services Hamor s 229A1 Vum/= Jakarta, 29 Ma r c h 1984. Lampiran : Peribal : Comnent on Zrwsnsigra Mr. Shiv. S. Eapur tion I Project Report. Director, Operation Evaluation Depar- tement The World Bark. Dear Sir, 1. Thank you for your letter dated March. 5 1984 of the Project Performance Audit Reports, Indenesia 2ransigration I Project ( Loan - 1318 - IND ). 2. Regarding the slomess of cattle development in Tranflnigration I Project, we would like to forward to you the comments on the project report, hoping it will be useful for cattle distribrtzion and development inder Trans 13m Project. a. Since the implementation of the Transmigration I,Project the Provin- cial Livestock Services Office has not completely been involved by the RIU. Tb avoid the same situation may happen again in Trans III. ProJect, the P4U should worked close together with the Prov-cial Livestock Services Offioe. In sch co-operation, the office will reponsible for the operation df all technical activities of cattle distribation and development within the Transnigration III Project area. For this purpose, one qualified staff from Provincial Livestock Services Office should be a member of the P!RTU's. b. The suggestion on increasing the bull to cow ratio to 10 % is highly appreciated. The bull Units in small size will be set-up in the place relatively close to the covws of the surrounding area. One bull will serve about 10 cows. c. DCLS agrees to the prog-amof cattle repayment. The progran will operate satifactorily under Mransmigration III Project only by an accurate plan and good field preparation. -18- Anmex II Page 2 d. Tte proble rising from day-to-dq operation &Puld be mlmed by both the RIU and the Provincial Livestock Services Office; Kind regards. RIERER OF'LI SCKES:- c C. r. Waldsn MmeIwld B1k, Jakarta. - 19 - CONFIDENTIAL THE WORLDBANK INDONESIA PROJECT COMPLETIONREPORT TRANSMIGRATIONI PROJECT (Loan 1318-IND) January 4, 1984 Projects Department East Asia and Pacific Regional Office V 0 0 NZ v I % 0 %It, I \ I Si d i QN 4 ll- .4 14 k - 21 - INDONESIA TRANShIGRATION I (LOAN 1318-IND) PROJECT COMPLETIONREPORT 1. AND CONCLUSIONS SUMMARY 1.01 Transmigration has become, since 1970, an important part of Indonesia's five year development plans: Under Repelita III (79/80-83/84) the target was to settle 500,000 families in the Outer Islands. 1.02 The Bank loan 1318-IND of US$30.0 million was the first to support the transmigration program. It became effective on February 2, 1977 and was closed on April 13, 1983 with 16 months delay. After a slow start, due mostly to managerial problems, the Transmigration I project achieved its major objectives which were to test new approaches to upland settlement and to explore ways and means to upgrade ongoing settlement schemes including: (a) the settlement of 4.500 families in Baturaja with requisite agricultural, physical and social infrastructure and services; Cb) the upgrading of infrastructure and services in an existing settlement of 12,000 families in Way Abung; and (c) the monitoring of the settlement development. 1.03 The analysis of key indicators shows that most project components were completed as foreseen in the SAR at no additional cost. The targets were exceeded with regard to road construction, water supply and on-farm benefits; however, a shortfall was noted in the distribution of cattle to settlers (4%) and in the monitoring and evaluation of the settlement program. 1.04 The Transmigration I project, and particularly its Baturaja settlement; which is regarded as a "show-piece" by GOI, can generally be considered a successful project inspite of the initial doubts. The encouraging results stem from various observations: (a) this project has demonstrated the possibility of maintaining agricultural production on red/yellow podsolic soils if fa-rmers apply an adequate package of inputs and soil management tech- niques. When this package is combined with the intensive cropping pattern developed by CRIA during project implementa- tion and the use of animal draft power the risks are reduced and above average yields can be achieved thus providing attractive benefits to transmigrants; -22- (b) after 6 years of project, the transuigrants of both sites have an average family income about 15% above subsistence level. Incomes will increase further in the 7th year (1983/84)when the one hectare rubber plantationgranted by the project enters into production. At full developmenteach farmer at Baturaja is expected to have an annual family income of about US$4,060 with 2 ha food crops and 2 ha rubber under production. Thus the Baturaja transmigrants, chosen from among the poorest inhabitants of Java and Bali, will have achieved, in less than 20 years, an income about 7.7 times above their initial level. (c) the inflow of about 1,200 spontaneous transmigrant families into Baturaja,representingabout 27% of the official transmigrants, and the development of cassava industries in Uay Abung demon- strates the attractiveness of the project and will further contribute to the area's economic development; (d) other factorshave also contributedto the success of this project among which: the integratedrural developmentapproach, the PMU management,the semi-nucleated village design, the high quality of rubber plantations,the agriculturalextension services, the primary and secondary schools, the health centers and the development of a large permanent market by private traders. 1.05 At the end of February 1983, the first four villages of the 11 coustructed in Baturaja were officially transferred to the Provincial authorities. However, to consolidate the benefits of this project, several Issues remain: the maintenance of roads and basic infrastructure; the organization of the rubber tapping and marketing; the allocation of land and the assistance to spontaneous transmigrants; the adjustment of the BIMAS extensionand credit program; the availabilityof medium and long-term credit for rubber planting,cattle and farm implements; and the timely supply of agricultural inputs, mainly fertilizers, chemicals and seeds. Some of these issues will be addressed by the PKU and provincial authorities under the Transmigration III project (Loan 2248-IND) which includes the extensionof the Baturajasettlementby another 2,000 families. II. BACKGROUND 2.01 Transmigration has been implemented in Indonesia for over 75 years. However, it was not until 1970 that it was accorded priority in the country's five year development plans with 40,000 families settled in Repelita I (69/70-73/74), 83,000 in Repelita II (74/75-78/79), and a target of 500,000 - 23 - families in Repelita III (79180-83/84)/1 and possibly 0.75 million in Repelita IV (84/85-88/89). During the same time, the agency responsible for transmigration was strengthened and upgraded from the status of a Directorate General, to a Junior Ministry (1978) and lately (March 1983) to a Ministry. 2.02 The earlier settlement schemes were based on irrigated agriculture and estate plantations. In those days, it was considered that rice and rubber production were the best means of providing a decent living for transmigrants. But the high investments and the scarcity of government resources hampered the trausmigration programs. It is roughly estimated that, between 1905 and 1970, about 500,000 families, or 1.5 million people, were settled through government sponsored transmigration programs. 2.03 To meet the targets fixed by the Repelitas since 1970, low cost settlements based on rainfed agriculture in upland red/yellow podsolic soils, known for their low fertility and acidity, had to be developed. In the initial upland settlements, the Government allocated small areas of land to the transmigrants (2 ha) and a minimum package of technical, social and economic assistance. As a result, in a majority of sites, after a few years of production yields declined, some settlers left agricultural production for off-farm employment and other transmigrants reverted to their initial poverty level. As such, in the early 1970s, the achievements of transmigration were not encouraging. 2.04 In view of these limited achievements and of the necessity to help solve the problems of overpopulation and unemployment in the Inner Islands together with the underutilization of land in the Outer Islands, GOI sought technical and financial assistance from various development agencies, including the Bank. GOI's main concern was to elaborate and test, in upland settlements, new development strategies that would ensure a minimum target income to transmigrants. The Bank became involved in the transmigration program in 1976 with the financing of a US$30.0 million loan (1318-IND) for the Transmigration and Rural Development Project in the Baturaja and Way Abung sites (S. Sumatera), the Transmigration I project. /1 In March 1983 the Junior Minister for Transmigration announced in a public speech that 415,568 families had been settled in the first three years of Repelita III. - 24 - III. PROJECT FORMLATION 3.01 Due to FAO's involvement in the PematangPauggang transmigration project, the identification of the Transmigration I projectwas carried out with the assistanceof the FAO/CP, from January 1972 to February 1974. The strategyfor upland settlements, proposed by the FAD/CP identification mission, consistedof a rural development projectin which the major agronomic,social and economicactivitieswere addressed. The projectwas based on the principlesof replicability, viabilityand low cost. Taking into account these aspects, the project was initially designed to develop four schemes in South Sumatra comprising the settlement of 13,000 families and the upgradingof a typicalongoing schemeinvolving12,000 families. 3.02 The projectpreparation and appraisal were carried out jointly by DGT, FAO/CP and the Bank froa February 1974 to October 1975. These two steps were time consuming because of delays in the implementation of aerial photography, soil surveys,physicalplanningand also because of the pioneeringnature of this project. Many featureswere innovativeand needed to be thoroughly analyzedand discussedbeforebeing acceptedby GOI and the Bank. Among these were the minimum income level of transumigrants, the optimal size of holdings, the ratio between food crops and cash crops, the level of food aid and inputs for farm establishment, the level of cost recovery,the projectmanagement,the layoutof villageinfrastructure, the health servicesand health control, the intensity and nature of agricultural support services,the animal husbandryand the number of draft animals. 3.04 During appraisal,the projectwas scaled down to .twotest areas comprisingthe settlement of 4,500 familiesin Baturajaand the upgradingof 12,000 familiesin Way Abung. 3.05 Project negotiations took place in June 1976 and 26 agreements, assurances and conditions of effectiveness bad to be negotiatedbetween GOI and the Bank. The projectwas not only innovative in many fields but also difficult to implement because of the involvement of eight government agencies. 3.06 Board presentationtook place on July 15, 1976 and a loan of $30.0 million, to be disbursedin five years,was approvedout of a total project cost of $56.8 million. The development periodwas estimatedat 10 years and the economicrates of return were estimatedat 11.5% for Baturaja and 19Z for Way Abung. IV. THE PROJECTSCOPE 4.01 In Baturaja,each of the 4,500 transmigrants would be given a 5 ha holding, a 35 sq m wooden house, inputs and food aid for farm development during the first year of settlement, one head of cattle and one hectare of - 25 - block planted rubber which would be maintained by the project during the first three years. The farm model was designed to allow a net annual income of $1,000 (1978 value) per family at full development under the assumption that each settler would crop intensively a house garden of 0.2 ha, produce food crops less intensively on 0.5 ha in rotation with 1.5 ha of a legumin- ous fodder crop and start tapping 1 ha of rubber 6 years after planting. Farmers would have access to short term credit through the BIMASIBRI program for annual crops from year 2 onwards and to long term credit for planting rubber or perennials on 2.8 ha of undeveloped land. Village cooperatives (KUD) were to be formed before year 6 for supply of inputs and rubber processing. 4.02 In Baturaja, the settlement infrastructure comprised the construc- tion of access-and village roads, 10 village centers with public facilities and the construction of settlers' houses with wells for water supply. The services provided in each village included a primary school, a health post, a BRI and KUD office, a village hall, a market, godowns, a place of worship, a project office and staff housing. The villages were designed for 400 to 500 families subdivided in clusters of 25 to 30 houses called "dukuh" which would form the basic social units. 4.03 In Way Abung, about 5,000 families would be granted 0.5 ha of rubber plantation with 3 years maintenance financed by the project and about 5,000 other families were to receive one head of cattle. The BIMAS/BRI credit system was to be strengthened for the supply of agricultural inputs and the village and road infrastructure was to be rehabilitated for better access to markets. 4.04 Extension services at the rate of one village worker (PPL) per village were to be provided to both sites and adaptive research on crops and cropping patterns was to be carried out by CRIA. Two holding grounds and two cattle farms were to be established. 4.05 The project was to be managed by a Project Management Unit (PMU) attached to the DGT. The PMU would comprise local personnel assisted by a multidisciplinary team of consultants. A senior accountant, placed directly under the project manager, would keep the accounts. V. IMPLEMENTATION A. General 5.01 The loan became effective on February 28, 1977, seven months behind schedule because of delays in meeting conditions of effectiveness. Thereafter the project was implemented in 6 years instead of 5 and disbursements were terminated on April 13, 1983 (loan closing). Altogether, the project was - 26 - and settlement about 2 years behind schedule,but almost all infrastructure targets have been met or exceeded. The remainingundisbursedbalance of $23,039.12was cancelled. 5.02 The project started slowly due mostly to inadequatestaffingof the PMU. In June 1978, this problemwas solved and project implementation improved steadily. However, these improvements were slow as all staff, includingthe projectmanager,were not familiarwith the multidisciplinary approach of this project. Despite severalmanagerialdraw backs and weaknessesparticularly in the cattle, credit, and monitoringcomponentsthe project achievedbetter results than expectedin the field of roads, water supply, agricultural productionand settler income. 5.03 The consultants, SCET, also requireda period of adaptation. Thereafter, they performed satisfactorily particularly in village designing and in solving the water supply problem in Baturaja. The supervising missions requestedthem, on several occasions,to provide additionalinput to assist the PHU in accounting, monitoringand other subjects. 5.04 A specialmention should be made of the project'saccountsand dis- bursementswhich improvedafter the recruitment in mid-1979 of a senior accountant. The presenceof this senior accountant within the PMU facilitatedsignificantlythe work of the Bank's missions. B. Project Components (a) Villages 5.05 In Baturaja11 villageswere constructed instead of 10 planned at appraisaldue to undulatingtopographyand lack of arable land immediately around village centers. The average number of settlersper villagewas thereforereducedfrom 450 to about 400. The villages were laid out accord- ing to two basic designs:semi nucleated(or semi linear) and nucleated. The seminucleated villageswere subdividedinto clustersof 25 to 30 houses and each house had direct access to 1 ha of arable land. The nucleated villages were built in one group of contiguous houses and each house had direct access to only 0.20 ha of'arable land for home gardens. In both cases the remaining arable land was distributed in 2 or 3 plots located within walking distancefrom the village centers. 5.06 The nucleateddesign led to cheaper infrastructure (25 to 30%). It also stimulated the social and economic life and facilitated adult and children education. On the other hand, the semi-nucleated design allowed farmers to tend more closely their first food crop lots particularly during the initialyears of settlement when pests and diseases threatenthe house- hold income. Though the nucleateddesign is cheaper, the semi-nucleated layout is better suited to agriculturallife in newly developedareas and is - 27 - preferred by farmers. Consequently the semi-nucleated design was recom- mended in the GOI transmigration program and used in subsequent Bank assisted projects. (b) Roads 5.07 In Baturaja, the lengths of roads were underestimated in the SAR: 15 km instead of 42 km for main access roads, 107 km instead of 140 km for village roads and 286 km instead of 478 km for farm roads. All roads suffered from the lack of maintenance during the disbursement period. The PMU had neither the know how nor the means to carry out adequate mainte- nance. This problem was partially resolved by hiring contractors for the maintenance of main roads and purchasing road equipment for the maintenance of smaller roads. However, the main roads were still in bad condition at the time of the mission's field visit. This was due to the bad performance of contractors, the inadequacy of road design (asphalt or macadam topping too narrow) and also to the difficulty of maintaining surfaced roads by a non specialized agency. Conversely the smaller village roads were found to be in better condition possibly because of the easier maintenance and less traffic. 5.08 If all roads had been constructed with an all weather compacted aggregated subgrade (AWCAS standard), their cost would have been reduced and their maintenance would have been a lesser problem for both PMU and District authority. The same observations and recommendations apply to Way Abung, where the SAR targets have been met after somn minor modifications. (c) Water Supply 5.09 In Baturaja, the water supply rapidly became a major problem after project inception for lack of ground water. The SAR had acknowledged this problem and had foreseen that the main water supply would be from the collection of roof water and the digging of 900 shallow wells in selected locations. It had also considered the drilling of 10 trial deep tube wells and the construction of some small earth dams to secure an additional source of water. 5.10 The roof water collection failed because of long drought periods and the first 140 shallow wells constructed under the project collapsed because of the lack of casings. Once the problem of casings was remedied the shallow wells became dry during the drought periods. Similarly the deep tube well program ran into problems after striking coal in one well and natural gas in another one./1 /1 The natural gas was still burning at the time of the mission's field visit despite several attempts to close the well. It was reported that the amount of gas was not sufficient for industrial applications. - 28 - 5.11 In view of its broken and rolling topography Baturaja was ideal for the construction of small earth dams. However the first 20 dams were washed away due to the lack of experience of contractors in this field./l In spite of these initial drawbacks the small earth dam program turned out ot be a success: 209 were constructed and have provided ample water supply during the severe 1982 drought. 5.12 Village water supply is done by shallow wells, equipped with hand pumps, and located around the dams. This supply system is easy to operate and hygienic. However, during project implementation, the Government decided to equip villages with water treatment plants and in 1981/82, 11 units were constructed at great expense (about $30,000/unit) out of Government resources. During the mission's visit, it was observed that some treatment plants were operating without chlorine thus defeation their purpose. The PMU also explained that the operation and maintenance of these plants was expensive and that some were stopped because of the lack of fuel. 5.13 In Way Abung, the construction of 2,170 shallow wells was imple- mented without difficulty in view of the flat topography and a shllow ground water table. The major problem stemmed from the lack of spare parts and mechanics to maintain the hand pumps. (d) Land Clearing 5.14 The land clearing operation was carried out without any major problem. The project developed a method that combined machinery and manual labor by: (a) clearing mechanically the 0.02 ha house lots so that trans- migrants could be settled in the project area as early as possilbe and (b) clearing the food corp lots with a combination of machinery and transmigrant labor force. This method allowed the early settlement of transmigrant provide off-farm income and generated minimal soil distrubance. However, it is applicable only to light secondary forest where small logs and branches can easily be removed by hand such as in Baturaja. (e) Monitoring and Evaluation 5.15 Inspite of the emphasis laid on Monitoring and Evaluation in the SAR, this component has been the weakest in project implementation. The starting of monitoring and evaluation was delayed several years due to problems of bugeting, contracting and mobilising. It was finally implemented by the Bogor Agricultural Institue (IPB) during the last three project years. Data collection took place during one agricultural year (1980/81) only and the final report was not completed at the time of the PCR mission. IPB's /1 The dikes were not properly compacted and spillways wer too small. - 29 - data analysiswas largelydescriptiveand the figuresmentionedin the provisionalreportswere lacking in accuracy: crop yields and farm income data were erratic and unreliable. 5.16 In general,IPB could not cope with large data collectingand with a far too ambitious processingand was overwhelmed monitoringand evaluationprogram. 5.17 The PMU and the CentralResearch Institutefor Agriculture(CRIA), through their routinecollectionof data, were able to compensatepartially for IPB's deficiency, particularlyin the field of agriculturalproduction. Researchand Farm Production (f) Agricultural 5.18 Adaptive agricultural research was carriedout by the Central Research Institutefor Agriculture(CRIA) and startedon the first year of project implementation.This Instituteworked successfully during six years in test farms and with about 1,000 project farmersin Baturaja and Way Abung. The adaptiveresearchprogram was introducedto maximize food crop productionunder rainfedconditionswhile maintainingsoil fertility, controllingerosionand limiting labor requirements. Based on these criteria,five croppingpatternswere tested togetherwith differentlevels of fertilizerapplications and improvedsoil managementtechniques. 5.19 The outcome of the adaptive researchwork was a croppingsystem based on three croppingseasons: - upland paddy, intercropped 1st season (October-February), with maize and cassava; - groundnutor soyabeansintercropped 2nd season (March-May), with cassava; - cowpeasor mungbeansstill intercropped 3rd season (June-August), with cassava. 5.20 In additionCRIA recommended: (a) to terrace the fields with more than 8% slope; (b) to plant maize and cassavaat fixed intervalsalong con- tours to control erosion;(c) to mulch with paddy straw and dibble without tillage the groundnutand cowpea crops for soil and water conservation and also to alleviatelabor constraints; and (d) to plow the paddy straw mulch into the soil, before the main paddy crop, to increasesoil organicmatter content. 5.21 The best yield/cost ratio for the full CRIA croppingpattern,was obtained with-200kg of urea and 200 kg of TSP per hectare and per year. Recent trials suggestadditionalbenefits from the applicationof 2 tons/ha of lime once every five years but no data was availableto quantify the benefits. - 30 - 5.22 The results achieved by CRIA show that crop yields increased during the first two years and stabilized from the third year onwards. This would demonstrate that, with adequate plant, soil and fertilizer management, the red/yellow podzolic soils are able to sustain rainfed food crops. The yields achieved by CRIA farmers in several locations during the last four project years are compared with the estimated SAR yields at full development and those achieved by average farmers after six years of production in the following table. WITH SAR Ct/ha) AND CRIA FARMERSCOMPARED AVERAGEYIELDS OF AVERAGE Average farmers Baturaja Way Abung (both sites) CRIA SAR CRIA SAR Upland paddy (C22/Serendah) 1.40 - 1.60 2.2 2.1 2.5 2.0 Maize (UG/Arjuna) 1.00 - 1.10 1.5 - 1.7 _ Cassava (Goding/local) 10.00 -11.00 14.4 12.0 15.0 15.0 Groundnuts (Gajah/Kidang) 0.50 - 0.55 0.7 1.0 0.8 1.2 Cowpea or Ricebean (NO.191/local) 0.35 - 0.35 0.5. - 0.4/a - /a There was a problem of seed germination. The above CRIA yields were obtained without potassium fertilizer or lime. The introduction of new varieties of blast resistant paddy and improved maize and groundnut together with adequate fertilizers, soil management and pest control are likely to further increase these yields. However these yields represent a potential achieved by advanced farmers rather than an average applicable to the entire project area. 5.23 The main drawback to the CRIA cropping pattern is its high cropping intensity (260%) which generates large labor requirements through- out the year (500-600 mandays/ha) and peak labor requirements between each cropping season. Its full application is therefore limited without resort- ing to additional labor force or animal draft power. However, the PMU estimated that about 70% of the farmers followed CRIA's recommendations in full for the first cropping season and partly.for the second. Generally the areas cropped in the third cropping season were extremely limited for lack of time and labor. 5.24 At Baturaja and Way Abung, after six years of project lise, the physical area cropped per transmigrant was about 1.30 ha (Table 5). These transmigrants were therefore able to crop larger areas than in most other GOI settlements because of the cattle allocated by the project (ref. 5.28). The cropping intensity, at both sites, was about 150% and the average yields were about 65% of those obtained by the CRIA assisted - 31 - transmigrants. The average income was computed at $614 i.e. 15% above the poverty line estimated at $530 for a family of five in rural Java and Bali. 5.25 The PCR mission has estimated that at full development the Baturaja farmers will obtain yields equal to those of the CRIA assisted farmers in year six, will increase the cropping intensity to 180% and will crop 2.0 ha on average. At that stage the family income due to food crops and the home garden will be about S1,496 or 2.8 times above the poverty level. Presently, the major problems facing the transmigrants for meeting these targets are labor constraints, lack of adequate seeds particularly for the leguminous crops and a reduction of fertilizer use by the farmers indebted with the BIMAS credit system. The breeding of two cows per holding will overcome the labor constraints and the rubber production, starting in the seventh year, will progressively reduce the indebtedness. As for the availability of adequate seeds further selection work is necessary and should be carried out under the Transmigration III project (Baturaja extension). (g) Livestock 5.26 The project purchased about 15% more cattle than the SAR estimates i.e. 4,661 heads for Baturaja and 4,327 for Way Abung to compensate for delays in the program and losses during the distribution. In one shipment 243 heads died due to the insufficient preparation of the holding grounds. The cattle supply program was suspended for one year and the creation of communal pastures was then decided in Baturaja by the supervision missions. Altogether, about 4,300 farmers in Baturaja and 5,000 in Way Abung received one head of cattle. 5.27 The repayment of the cattle by the transmigrants was to be made through the heifer return program at the rate of two heifers per head of cattle. The heifer return program did not operate satisfactorily and at the time of the PCR mission about 200 farmers in Baturaja still had not received their head of cattle. The heifers were to be inseminated in bull units managed by the PMU, but because of the distance to these bull units, the low number of bulls purchased, and the refuctance of some farmers to return one calf, the calving rate remained below expectations, at about 60%. One of the solutions proposed by the PMU to increase the calving rate, was to increase the bull to cow ratio to 10%, and distribute the bulls to the best farmers. 5.28 The project introduced livestock to improve soil fertility through the application of manure, provide animal draft power and progressively increase nutritional standards of the transmigrants' aiet. These improve- ments, necessary for sustained food crop production and increased farm income, were starting to provide additional benefits at the time of the PCR mission. On that occasion, several farmers reported using manure instead of - 32 - inorganicfertilizersin their home gardens. They also indicatedthat the quantity of manure produced by one head of cattlewas insufficient to fertilizethe food crop lots. A rapid census showed, on the other hand, that about 60% of the Baturaja farmers shared their cattle with neighborsto plow their fields and that farmersusing animal draft power were able to increase the area croppedby about 0.5 ha./I 5.29 During project implementationthe lack of animal draft power hampered the extensionof cropped areas and the full use of the CRIA cropping pattern. A trial to introducesmall scale mechanization did not prove successful. The distribution of two heads of cattle would have allowed a quicker and larger developmentof croppedareas in Baturaja and Way Abung. (h) Rubber Development 5.30 The rubber developmentprogram met the appraisaltargetsof 4,600 ha /2 in Baturajaand 2,500 ha in Way Abung. Land clearing,planting and maintenancewere contractedto PTP X and providedmost of the off-farm employmentin Baturaja. The quality of the rubber plantations, in both sites, is of high standardand peak yield projections were revisedupwards by 20%, from 1.2 t/ha to 1.45 t/ha. In view of the high quality of the rubber plantations and taking into account the relativeinexperience of transmigrants in rubber plantation,the maintenanceperiod, by PTP X, was increasedfrom 3 to 6 years to preserve the potentialyields until the first year of tapping. 5.31 In Baturaja,each of the 4,500 transmigrant received,as planned, 1 ha of rubber but, in Way Abung, the availability of contiguousland for block planting encountered serious problems. At appraisalfarmershad received only one hectareof their two hectare holdings and few difficulties were foreseen in findingcontiguousland for block planting2,500 hectare of rubber. But, before plantingstarted, all land was allocatedto farmers and titles processed. To implementthe block plantingwould have involved repossession, splittingand reallocationof land. Finding contiguousblocks was also further complicated by the progress of the Way Rarem irrigation scheme and its competingdemand on land. Consequently of the 5,000 farmers planned to receive 0.5 ha of rubber about 2,700 were allocatednearly 1 ha. The number of beneficiaries was thereforereduced while the area planted remained as planned. /1 1.1 ha on average croppedwithout animal draft against 1.6 ha with animal draft. /2 Including100 ha for trials and demonstrations. - 33 - 5.32 At the time of the PCR mission, the trainingof farmers for rubber tapping had begun in the oldest villages,but, the two crumb rubber fac- tories, one In Baturaja and one in Way Abung, were still not constructed. As a result the financingof these factorieshas been included in the Trans- migrationIII project under the Baturaja extensioncomponentand PTP X was to construct,own and operate these factories. (i) Other components 5.33 The health componentwas carried out beyond the SAR estimateswith the constructionof two health centers and 12 health posts In Baturaja,and three health centers and 19 health posts in Way Abung. These facilities were staffed by the Governmentwith doctors,midwives, nurses and health extensionists, and sufficient medicine and supplieswere also provided. In addition,a malaria control program was carried out by the project. In general, the health conditionof settlers is satisfactoryin both sites. 5.34 The educationcomponentwas also carried out beyond the SAR estimateswith the constructiona senior high school, 2 junior high schools and 12 primary schools in Baturaja,and 4 junior high schools in Way Abung. These facilities were adequatelystaffed by the PMU with teacherswho were granted the same benefits as the transmigrants. The PMU reported that 100X of the children attended school and a demand for college level educationhad been expressedby the transmigrants. 5.35 The cooperativecomponentwas implementedwith an uneven success tbough the project had constructedthe village level facilitiesas planned. Village cooperatives(KUD) were formed in each village but, in Baturaja,the number of participating farmersvaried from 0 to 86% per KUD with an overall average of 41%. Most KUDs were not very active because of the lack of funds to acquire processingand transportfacilitiesand also because of the lack of managementskill. The marketingof rubber, with the tochnicalassistance of PTP X, may reactivatesome of the KUDs. 5.36 Farm inputs do not seem to have encounteredspecial problemswith regard to the distribution of fertilizers and insecticidesin both areas, but the need for timely availabilityof seeds, especiallypeanuts and soybean was not met. The seed farms have been mostly used for field trials by CRIA. 5.37 The BIMAS/BRIagriculturalcredit program started slowly and never operated on a large scale in both areas. It was reported that 39% of the transmigrants in Way Abung and 50Z in Baturaja were using presentlythe program. Three reasons can be attributedto this rather low performance. First, the BIMAS/BRIloans are made on a per crop basis which does not correspondto the needs of the mixed croping pattern advocatedby CRIA; second, the grant of fertilizers during the initial three years of settle- ment did not encouragethe Baturaja transmigrants to apply for BIMAS/BRI loans; and third, the low repaymentrecord, particularlyin Way Abunv, progressively reduced the number of loans. However, it should be noted that the use of fertilizershas been well understoodby transmigrants, especially in Baturaja,where they are now availablein local markets. -34 - VI. ECONOMIC EVALUAIION A. Yields and Production 6.01 The principal economic benefit of the project would be increased rubber production. The appraisal target plantations of 4,500 ha in Baturaja and 2,500 ha in Way Abung have been well established, and the peak yield is now reestinated at 1,450 kg/ha compared to 1,200 kg/ha at appraisal. The other quantified economic benefits are increased food and garden crops and livestock production. A survey carried out by the PCR mission indicated that there was a surplus production of about 850 t of paddy in Baturaja and about 20,700 t of cassava in Way Abung where four cassava processing plants have been constructed in recent years. Areas under food crops production are about as estimated at appraisal i.e. 5,850 ha against 5,625 ha in Baturaja but for the economic evaluation SAR yield projections have been revised downwards, for rice to 0.9 t/ha in Baturaja /1 and 1.2 t/ha in Way Abung and for groundnuts to 0.7 t/ha in Baturaja and 0.9 t/ha in Way Abung. 6.02 Other modifications have been made on the farm models. For Baturaja, it has now been assumed that during the 30-year project life only two instead of three hectares of rubber would be established by each farmer./2 In Way Abung, one instead of one-half hectare of rubber has already been planted and this has decreased the number of farmers partlci- pating in the project's rubber program to 2,700 instead of 5,000. It has further been assumed that all increases in food crops production in Way Abung would occur on the farms benefitting from cattle under the project's distribution program valued at the net output of one-half hectare. In Baturaja, the food crop production of 1,200 spontaneous transmigrants has been included-in the project's output on the basis of one-half hectare each. B. Pricing 6.03 Costs and benefits are expressed in constant mid-1982 prices. Actual costs from 1976-1982 have been adjusted to mid-1982 level on the basis of the GDP deflator. The appraisal shadow-priced foreign exchange 25Z above the official exchange rate (Rp 415 = US$1). Following two devalua- tions of the rupiah (November 1978 to Rp 625, and March 1983 to Rp 970), however, a standard conversion factor of 1.0 has been assumed for the reevaluation of the rate of return. 6.04 The appraisal shadow-priced farm labor at the minimum wage rate (Rp 210/day), estimated to provide 50% of the minimum subsistance require- ments. For the economic reevaluation farm labor has been shadow-priced at Rp 500 per working day (half the existing market wage rate in Baturaja) con- sidering the over-abundant labor supplies and low productivity employment in Java and Bali. /1 The appraisal projected a gradual increase of rice yields from 1.2 t/ha to 2.1 t/ha, and of groundnut yields from 0.6 t/ha to 1.2 t/ha. /2 Transmigrants who have started tapping, have opened savings accounts with BRI for reinvesting in the second hectare of rubber. - 35 - 6.05 In addition to including all costs along the appraisal's method- ology in the economic reevaluation, the overhead costs of the Project Management Unit and the Project Surveys and Studies (mainly CRIA and IPB) have been apportioned to Baturaja and Way Abung at a ratio of 2:1. C. Economic Rate of Return 6.06 Using the foregoing assumptions and discounting cjsts and benefits over 30 years, the economic rate of return for Baturaja is 16.6Z /1 (appraisal: 11.5%) and 22% (appraisal: 19%) for Way Abung. The higher rate of return of Way Abung reflects the level of sunk investment in settlement. In both project areas, cost overruns are expected to be offset by higher rubber yields then estimated at appraisal and higher rupiah values of the world market price forecasts. D. Other Benefits 6.07 The major indirect benefit of the project was the settlement of about 1,200 spontaneous transmigrant families in Baturaja (27% of the project settlers) who contributed to the project economy by the supply of incremental production. Another indiree-t benefit refers to the road rehabilitation program in Way Abung which reduced the cost of transportation for the subproject area, contributed to the development of cassava factories and generated off-farm labor employment. Lack of data, unfortunately, prevented the valuation of this latter benefit. It has also not been possible to quantify the benefits from improved health and nutritional status of the settlers. VII. PROCUREMENT, DISBURSEMENTS AND PROJECT COSTS A. Procurement 7.01 The weak management and the inexperienced staff during the first three years of the Project caused many procurement problems. The PMU was not yet effectively established and recognized as the central management and coordination unit of the Project. The Provincial Transmigration Director and the site managers often let contracts without reference to the PMU. Purchases of necessary materials and equipment were inadequately planned; specifications of civil work contracts were lacking or of poor quality. Contract approval procedures were complicated and time consuming because they involved numerous agencies (BAPPENAS,Ministry of Finance, Bank of Indonesia); contractors for local competitive bidding were not prequalified, and qubsequently work standards were often poor and quality of materials often unsatisfactory. For example, road construction was of poor standard and soor required upgrading; 140 wells collapsed and were unusable. /1 11.8% in the unlikely event that farmers would not plant the second hectare of rubber assumed in para. 6.02. - 36 - 7.02 The Bank's supervision missions played an important role in strengthening the PMU./1 They took the position that the Bank would no longer disburse unless contracts were cleared and work completion certified- by the PMD staff, including its consultant advisers. As from late lS79 onwards until project completion, the effectiveness of the Project Director and his Financial Advisor speeded up and improved procurement to a satis- factory level. However, the contracting for rubber development still remained a problem until 1981. Because of disagreement on PTP X's (formally PNP X) contract prices, the signing of the contract with DGT was long delayed. As a consequence, as little as 255 ha rubber were planted in Baturaja ia 1980/81. 7.03 The decisive factor in improving procurement was the improved staffing of the PHMU. Besides channelling all procurement through the PMU - which was certainly important enough by itself, no procedure was changed. Many of the Bank's new projects in Indonesia still face in their early years the same procurement problems as Transmigration I did because of the unfamiliarity of staff with local procurement procedures, agencies, and personnel involved. B. Disbursement 7.04 The delayed implementation of the Project during the first three years, mainly due to weak management, slowed down disbuisement substan- tially. For about two years, total disbursements remained at 30% or less of appraisal estimate. Only since late 1979, in the fourth project year, did disbursements gain momentum, helped by the decision to disburse 65% instead of the original 30% of total costs of rubber development. The first reallo- cation of funds in 1981, in fact, was necessary to meet the higher financing requirements of rubber. 7.05 The flow of disbursements was also hampered by the slow processing of disbursement applications - again due to the weak management/coordination factor - and the lack of supporting data/documents after submission of applications. Again, with the improved staffing of the PHU the problems were overcome. 7.06 The Closing Date of the Loan was extended once, from December 31, 1981, to December 31, 1982 and on an exceptional basis withdrawal applica- tions were accepted up to April 13, 1983, the date of the last payment. As per Hay 31, 1983, the Loan had an undisbursed balance of $23,039.12. /1 The total number of staff weeks provided over six years amounted to 200 or an average of about 33 staff weeks annually. - 37 - 7.07 The Bank's funds share of total project cost was as estimated at appraisal (53%). The increase in total rupiah project cost by 38.6% was off-set by the dollar appreciation against the rupiah since late 1978. As compared to the original allocation by expenditure category, Category IV - rticularly large increase (para. 7.04) in disburse- Rubber Development had a p.. ments, while Category VII (Village Development) used less funds than esti- mated at appraisal. For continued rubber maintenance until six years after planting, the Bank has included funds in the Transmigration III Project. C. Project Costs 7.08 Total project cost up to December 31, 1983 amounted to Rp 32,700 million ($57 million) against an appraisal estimated of Rp 23,580 million ($57 million). The increase in local currency costs was mainly due to the delay in implementation and higher than anticipated inflation rates. To cover part of the increase in local cost, the Bank increased the disbursement percentage for rubber development (para. 7.04). Goverment provided the budget funds to finance other cost increases. 7.09 The cost per family was US$7,923 (Mid 1982 value) against $6,787 at appraisal in Baturaja and $2,160 against $1,990 in Way Abung based on 4,500 and 10,000 beneficiaries respectively. However, the cost per family would be reduced to about $6,263 in Raturaja if the 1,200 spontaneous transmigrants were considered as project beneficiaries. 7.10 Shortages of funds were occasionally caused by budgeting proce- dures - as they occur still nowadays - and resulted sometimes in implemen- tation delays. If items were budgeted in previous DIPs but current costs exceeded the budgeted amount, lengthy revisions were required, involving several agencies. VIII. LESSONS LEARNED 8.01 The lessons learned from the Transmigration I project are manifold and some have already been applied to subsequent GOI and Bank assisted transmigration projects. (a) The PMU proved to be an efficient management tool when it is given the authority to control and coordina*teall project activities including those not falling directly under the jurisdiction of the Ministry of Transmigration. However to be cost effective this concept should be applied to groups of settlements totalling 15,000 to 20,000 families. (b) The PHU manager's position requires a broad knowledge of technical and administrative matters. Experienced managers are scarce in Indonesia and due consideration should be given in future projects to the training of adequate high level staff before entrusting them with the responsibilities of PMU managers. - 38 - (c) The successof a transmigration settlement project depends not only on good managementduring the first five project years before it is transferred from the Ministry of Transmigration to the Provincial Authorities but also on the levels of assistance and trainingprovidedto the transmigrants during that period. One year of inputs granted to new settlersproved to be insufficient. The extensionto three years of granted inputs appearedadequateto allow time for the transmigrants to pro- duce marketable surpluses,for the BIMAS/BRI credit structures to become operational and for traders to establish new links and developmarketsin the project area. Applied researchand intensiveagricultural extensionwork during the initial five years are also two prerequisites for sustainedagricultural development. (d) The mixed farmingsystem based on garden crops, fruit trees, food crops,livestockand rubber used in the Transmigration I Project, thoughprovidingthe best insurance against risk, has put excessive strain on transmigrants.Not only each transmigrant had to learn in a short time a multiplicity of agricultural tasks but also had to cope with problems of soil erosion, animal and insect attacks, low soil fertility and erratic rainfallwhich are intensified in newly opened settlements. The mixed farming systemled to lower initial food crop yields and longer rubbermaintenanceperiod by the PMU than estimatedin the SAR. Rather than to grant the same amount of land and the same agricultural package to all transmigrants it is the PCR mission's belief that higher returns per family and per project could be achieved in future projects by considering several farming systems according to individualpreferences of transmigrantsand land suitability. Food crop, tree crop and livestockfarm models could be considered in each upland settlement. (e) The provision of cattle in Baturajaenabledtransmigrants to crop, on average,largee areas (1.3 ha) than other GOI sponsoredsettlements without cattle. The provisionof one or more heads of cattle through credit operationsalleviateslabor constraints, intensifies crop productionand reduces risk. (f) Sustained food crop productionis feasibleon red/yellow podsolicsoilswhen adequate croppingpatterns,soil management systemsand input levels are applied. Under these conditions and under the provisionthat enough aniifaldraft power (2 cows) and land (3 to 4 ha) is granted to transmigrants,a food crop farm model yieldinga family net income over $1,500/year appears feasiblewithout additional off-farmor tree crop cash income. - 39 - (g) The initial allocationof 5 ha holdings to each transmigrant in Baturajawas excessive. Areas of 3-4 ha dependingon the farm model (food crops, tree crops or livestock)would be more in line with the labor capacityusually expected from standard transmigrantfamilies. However, larger holdings of 3 ha or above induce more spontaneoustrausmigration than smallerholdings of 2 ha or below. (h) The monitoringand evaluationfailed in the Transmigration I project. It was too comprehensive and complex to be carried out in remote transmigrationsites. Future projects should limit the scope and number of indicatorsto increase the efficiencyof monitoringand evaluation. (i) In rolling topography where groundwateris unevenly distributed, the construction of small earth dams surrounded by shallow wells proved to be an excellent water supply system. (j) Technicaland financialassistanceto the lKUDsis necessaryfor their proper operation. The technicalassistancecould be in form of training coursesfor the managers and accountants. The financialassistancewould be for the purchase of processing and transportfacilitiesand also for the creation of revolving funds. (k) The BaturaJa settlementachieved in general better agricultural and economicalresults than most other Governmentsettlements developedduring RepelitaIII. These results were made possible because of higher levels of investmentsand technicalassistance per transmigrantthan in the rest of the program. The Baturaja settlementis thereforenot representative of the Repelita III transmigration program but should rather be consideredas a pilot settlementestablishedto test mixed farming on red/yelllow podzolicsoils and to determinepotentiallevels of production under given conditions. When compared to Baturaja,most other transmigrationsettlements can be consideredunderdeveloped.As the success of settlements depends largely on the provisionof a sound economicbase, it would thereforebe recommended to: (i) increase the levels of investmentand technicalassistance in the future settlements to be developedunder Repelita IV; and (ii) implementa second stage settlementdevelopmentprogram to improve the agriculturaloutput and living conditionsof the settlementsdeveloped under Repelita III. -e -460 b TRAUIclaClain I (LOAN 1318-The0 PROJErCTUSLUT10W URPOIT Revised Percent of Apprasal target Achieve- Revised Unit total 7!81182 sent SAR target Settle_mt at Baturais Number settled No. 4,500 4,500 4,500 100 100 Houses constructed No. 4,500 4,500 4,500 100 100 Village centers No. 10 11 11 110 100 constructed load Constrnition Naturoas main access roads km 15 15 42 280 280 Secondary roads hk 107 118 140 130 120 Frai roads k 286 449 478 170 106 Main access roads hk SO 50 50 100 100 Secondary roads km 50 30 30 50 100 Road improvements km S0 100 100 200 100 Tater Supply Saturaja Shlow wells 1No 900 1,665/a 1,525 170 100 Earth dams No. - 22971b 209 - 100 water treatuent uidts No. - li 11 - 100 Shallow wells No. 2,000 2,000 2,170 110 110 Daturala Health and Schools Health centers No. 1 1 2 200 100 Health poets No. 12 12 12 100 100 Schools No. 10 11 11 110 100 Health centers No- 2 3 3 150 100 Health posts NO. 22 19 19 85 100 Schools No. 4 4 4 100 100 lubber Plantin g Baturaja ha 4.600 4,600 4600 100 100 Way Abung ha 2.500 2,500 2.500 100 100 Cattle Distribution Progr | Baturaja No. 4,5W 4.500/c 4,300 96 96 Way Abung No. 5,000 5,000 5.000 100 100 Seed Garden Development Baturaja ha 200 200 200 100 100 Area uder Tood-Crops Batcraja ha N/& 5.625 5.850 - 104 Distributilon of inputs Baturaja No. 4,500 4,500 4,500 100 100 Studies Progri support ha 200,000 0 0 /d 0 - Way Abung Irrigation No. 1 0 0 7 0 - MonLtoring No. 1 1 0 7T 0 0 Applied Agri. Res. NO. 1 1 1 100 100 Health control No. 1 1 1 100 100 /a 140 collapsed duriag constructlon. 7ii 20 collapsed during construction. 71 243 did after the arrival of one sblpmt. Id was carried out under the Tranmigration I1 Project CLoan 1707-IND). 7T was carriled out under Japanese Technld Aistnce T7 the final report was not reAdy at the time of the PF mission. - 41 - Table 2 INDONESIA TRANSMIGRATION I (LOAN 1318-IND) PROJECT COMPLETIONREPORT Project Cost Summary /a Appraisal Actual (up to estimate Dec. 31, 1982) Rp M US$M Rp M US$1 A. Investment Cost Baturaja 6,313 15.20 15,256 26.50 Way Abung 3,863 9.30 8,585 14.90 Project Management Unit 58 0.10 1,467 2.50 Other 1,814 4.50 2,406 5.00 Subtotal 12,048 29.10 27,714 48.90 B. Operating Costs Baturaja 1,389 3.30 2,739 4.70 Way Abung 249 0.60 1,393 2.20 Project Management Unit 643 1.50 789 1.50 Studies (program support) 2,477 6.00 48 0.01 Cattle distribution - central facilities 208 0.50 - - Subtotal 4,966 11.90 4,969 8.40 Contingencies 6,565 15.80 - - Total Project Cost 23,579 56.80 32,683 57.30 C. Investment cost per family including contingencies Baturaja 5,393 6,722 D. Operating cost per family including contingencies Baturaja 1,324 1,211 E. Total cost per family including contingencies 6,787 7,923 /a 1976-1978 US$ 1.00 = Rp 415, 1979-1982 US$ 1.00 = Rp 625. - 142 - Table 3 INDONESIA TRANSMIGRATION I (LOAN 1318-IND) PROJECT COMPLETIONREPORT Allocation of Loan Proceeds (USs Million) (As of February 28, 1982 /a Original Category allocation Actual use I. Civil works 7.5 25.0 8.0 27.1 II. Vehicles & equipment 2.8 9.3 2.0 6.8 III. Agrochemicals 1.2 4.0 1.9 6.4 IV. Rubber development 2.7 9.0 5.3 17.9 V. Cattle 1.4 4.7 2.2 7.4 VI. Consultants' services & surveys 6.4 21.3 6.7 22.7 VII. Village completion 4.5 15.0 3.4 11.5 VIII. UnalLocated 3.5 11.7 - - 30.0 100.0 29.5 100.0 /a Excluding five outstanding withdrawal applications totalling US$568,529. Table 4 - 43 - INDONESIA I (LOAN 1318-IND) TRANSMIGRATION PROJECT COMPLETIONREPORT Disbursement Schedule (US$ Million) Appraisal estimate Actual Calendar X of Z of year Annual Cumulative total loan Annual Cumulative total loan 1976 1.0 1.0 3 - - - 1977 5.5 6.5 22 1.64 1.64 5.5 1978 9.5 16.0 53 4.74 6.38 21.2 1979 7.0 23.0 77 4.24 10.62 35.4 1980 5.5 28.5 95 5.54 16.16 53.9 1981 1.5 30.0 100 6.31 22.47 74.9 1982 - - - 4.90 27.37 91.2 1983 - - - 2.63 29.98 /a 100.0 /a There is an undisbursed balance of US$23,039.12. Table 5 - 44 - INDONESIA TRANSNIGRATIONI (LOAN 1318-IND) PROJECT COMPLETIONREPORT Cropping Pattern and Farm Budget at Baturaja Full Crops Unit Year 6 development/c Yields (Pure Crops) Paddy ton/ha 1,400 2,200 Maize ton/ha 1,noo 1,600 Groundnuts ton/ha 500 800 Cassava ton/ha 10,000 15,000 Cowpeas ton/ha 350 500 Harvested Area Physical area cropped ha 1.30 2.00/a First Season Paddy ha 1.00 1.30 Maize ha 0.25 0.40 Second Season Groundnuts ha 0.55 1.20 Third Season Cassava ha 0.05 n.30 Cowpeas ha 0.10 0.40 Cropping intensity 150 180 Crop Production Paddy kg 1,400 2,860 Maize kg 250 640 Groundnuts kg 275 960 Cassava kg 500 4,500 Cowpeas kg 35 200 Net Farm Income (Excluding Family Labor) Food crops Rp 280,000 847,000 Home gardens Rp 150,000 200,000 Rubber First hectare Rp - 1,052,000 Second hectare /b Rp - 744,000 Total Rp 430,000 2,843,000 US$ 614 4,061 /a With two cows. /b Second hectare rubber financed by transmigrants through long-term credit. /c The Baturaja yields and income will be substantially higher than those of the Government settlements developed during Repelita III.- - 45 - Table 6 INDONESIA TRANSMIGRATION I (LOAN 1318-IND) PROJECT COMPLETION REPORT Economic Rate of Return Calculation - Baturaja Constant Mid-1982 (Rupiah Million) Costs Benefits Farm Foodcrops production including Year Investment O&M costs Total Rubber garden Livestock Total 1976/77 793 713 41 1,547 - 23 - 23 1977/78 1,281 826 68 2,175 - 75 8 83 1978/79 2,313 747 135 3,195 - 150 26 176 1979/80 3,515 738 286 4,539 - 364 55 419 1980/81 3,093 772 1,190 5,055 - 838 86 924 1981/82 3,486 836 1,310 5,632 189 1,478 88 1,755 1982/83 2,388 917 1,414 4,719 494 1,591 254 2,339 1983/84 1,623 281 1,681 3,585 876 2,217 254 3,347 1984/85 1,315 281 1,839 3,435 1,626 2,260 260 4,146 1985/86 2,307/a 281 1,956 4,544 2,297 2,284 260 4,841 1986/87 3,432/a 281 2,129 5,842 3,307 2,288 280 5,875 1987/88 2,363 281 2,285 4,929 4,309 2,313 320 6,942 1988/89 1,965 281 2,315 4,561 5,305 2,313 *363 7,981 1989/90 1,799 281 2,315 4,395 5,347 2,313 410 8,070 1990/91 1,622 281 2,544 4,447 9,291 2,790 414 12,495 *1991/92 788 281 2,846 3,915 11,993 2,790 432 15,215 1992/93 - 281 2,944 3,225 13,663 2,790 455 16,908 1993/94 - 281 2,944 3,225 14,327 2,790 480 17,597 1994/95 - 281 2,944 3,225 15,210 2,790 499 18,499 1995/96 - 281 2,944 3,225 16,094 2,790 529 19,413 1996/97 - 281 2,944 3,225 16,978 2,790 529 20,297 1997/98 - 281 2,944 3,225 17,862 2,790 529 21,181 2005/06 - 281 2,944 3,225 17,862 2,790 529 21,181 ECONOMIC RATE OF RETURN = 16.6% la Planting of second hectare of rubber on farms: 2.250 ha each 1985/86 and 1986/87. - 46 - Table 7 INDONESIA TRANSMIGRATION I (LOAN 1318-IND) PROJECT COMPLETIONREPORT Economic Rate of Return Calculation - Way Abung Constant Mid-1982 (Rupiah Million) Costs Benefits Farm Foodcrops production including Year Investment O&M costs Total Rubber garden Livestock Total 1976/77 345 54 - 399 _ _ _ 1977/78 905 169 - 1,074 - - - 1978/79 1,603 289 133 2,025 - 218 - 218 1979/80 1,286 446 163 1,895 - 293 - 293 1980/81 1,738 362 237 2,337 - 621 - 621 1981/82 1,744 142 373 2,259 89 1,120 184 1,393 1982/83 515 212 416 1,143 190 1,665 184 2,039 1983/84 499 50 432 981 408 1,684 184 2,276 1984/85 418 50 503 971 921 1,698 184 2,803 1985/86 264 50 640 954 1,718 1,710 206 2,634 1986/87 - 50 917 967 2,210 1,754 234 4,198 1987/88 - 50 894 944 2,502 1,828 256 4,586 1988/89 - 50 918 968 2,802 1,924 289 5,015 1989/90 - 50 949 999 3,133 2,054 300 5,487 1990/91 - 50 980 1,030 4,546 2,649 316 7,511 1991/92 - 50 1,012 1,062 4,745 2,811 338 7,894 1992/93 - 50 1,046 1,096 4,745 2,981 358 8,084 1993/94 - 50 1,083 1,133 4,745 3,166 382 8,293 1994/95 - 50 1,122 1,172 4,745 3,359 399 8,503 1995/96 - 50 1,163 1,213 4,745 3,562 426 8,733 ECONOMICRATE OF RETURN= 22% ;,~~ -. - -- D': 7 soling Pr-. -h -- Tobt~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C _~~~~~~~~ A Tl E R A p w4ouon~ MnogDsa 4-d *1- _ _ A TL o j,"' ,, . cso / \ \ t )a- -v ..- ;5- ere!A:/ S -- ~ t-. : ~ -a-io5dS-e8e* .;~~~~~~~~~~~~~~~~~~~L& .1~~~~~~~~~~~1 ;55~~~~ -C V,405d S@'c A, N ~~ 4.77 ;:C C:;-0u' ''' LAM'' U'"N' -- c.l.v : ~ ~ C-S.I S/&44 g_S { '-5,_~~~~~~~~~~~=/4 a,.ND1ONESco _____ TRN RTO AN RURA DEVEOMN PROJECT ARA SETEET .N N AN .A .UATR SEA N PRVICE j~~~~~~~~~ ._! t C. - -t .w -w - - / - A - ~~~~~~~4~~~~~L ; < I T.. _ - ~_ rA - a-'ie I '''' - - - ;''1' -r pK_- ~ _Aw~~~~ c .. . dOC?.i1- - \ I - _ 'IVL e - )uII3_r df- , '' ' S- k 16 4 _ d3 _fl _ .v __ _ __._ _ __._ CVdd S -~II - -- _ __._ _ __-_ _ __._ _ _ _ _