Document of The World Bank FOR OFFICIAL USE ONLY INDONESIA. VILLAGE INFRASTRUCTURE PROJECT (Loan 3888-IND) March25, 1999

Transcription

1 Public Disclosure Authorized Document of The World Bank FOR OFFICIAL USE ONLY Report No Public Disclosure Authorized Public Disclosure Authorized IMPLEMENTATION COMPLETION REPORT INDONESIA VILLAGE INFRASTRUCTURE PROJECT (Loan 3888-IND) March25, 1999 Public Disclosure Authorized Transport Sector Unit Indonesia Country Management Unit East Asia and Pacific Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.

2 CURRENCY EQUIVALENTS Currency Unit = Indonesian Rupiah (Rp) US$1 = 2,200 Rp (January 1995) = 2,350 Rp (January 1996) = 2,400 Rp (January 1997) = 8,000 Rp (January 1998) = 13,000 Rp (June 1998) = 8,000 Rp (January 1999) GOVERNMENT OF INDONESIA FISCAL YEAR April 1- March 31 WEIGHTS AND MEASUREMENTS Metric Utits ABBREVIATIONS AND ACRONYMS BAPPENAS - National Development Planning Agency BRI - Bank Rakyat Indonesia (state-owned commnercial bank) BPKP _ National Audit Agency Bupati - Head of Kabupaten Camat - Head of Kecamatan DPUK - Department of Public Works at Kabupaten INPRES - Presidential Instruction, including a class of grants IDT - Inpres Desa Tertinggal (Program for villages left behind) ICR - Implementation Completion Report Kabupaten - District Kecamatan - Subdistrict KDP Kecarnatan Development Project Kepala desa - Head of village and of LKMD LKMD - Village commnunity group MoF Ministry of Finance Pimpro - Project manager from DPUK PMU - Project Management Unit Repelita - Five-year Development Plan SOE - Statement of expenditure SUSENAS - Household expenditure survey Vice President: Country Director: Sector Manager: Task Manager: Jean-Michel Severino Dennis de Tray Jitendra Bajpai Frida Johansen

3 Village Infrastructure Project -i- FOR OFFICIAL USE ONLY INDONESIA VILLAGE INFRASTRUCTURE PROJECT (Loan No INI)) IMPLEMENTATION COMPLETION REPORT Table of Contents PREFACE... i EVALUATION SUMMARY... iii PART I: IMPLEMENTATION ASSESSMENT... I A. STATEMENT/EVALUATION OF OBJECTIVES... 1 B. ACHIEVEMENT OF PROJECT OBJECTIVES... 2 C. MAJOR FACTORS AFFECTING THE PROJECT... 6 D. PROJECT SUSTAINABILITY... 6 E. BANK PERFORMANCEE... 7 F. THE BORROWER PERFORMANCE... 9 G. ASSESSMENT OF OUTCOME H. FUTURE OPERATIONS I. KEY LESSONS LEARNED PART II: STATISTICAL ANNEXES TABLE 1: SUMMARY OF ASSESSMENT TABLE 2: RELATED BANK LOANS TABLE 3: PROJECT TIMETABLE TABLE 4A: LOAN DISBURSEMENTS: CUMULATIV ESTIMATED AND ACTUAL TABLE 4B: CONSULTANT CONTRACTS TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION (A, B & C) TABLE 6: KEY INDICATORS FOR PROJECT OPERATION TABLE 7: STUDIES INCLUDED IN PROJECT TABLE 8A: PROJECT COSTS TABLE 8B: PROJECT FINANCING TABLE 9: STATUS OF LEGAL COVENANTS TABLE 10: COMPLIANCE WITH OPERATIONAL MANUAL STATEMENTS TABLE 11: BANK RESOURCES: STAFF INPUTS TABLE 12: BANK RESOURCES: STAFF MISSIONS ATTACHMENT 1: SAMPLE ECONOMIC RETURNS TO VILLAGE INFRASTRUCTURE ATTACHMENT 2: AVERAGE NO. OF BENEFICIARIES FOR 96/ PART III: BORROWER CONTRIBUTION TO THE ICR MAP IBRD 26797R This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.

4 Village Infrastructure Project -ii- IMPLEMENTATION COMPLETION REPORT INDONESIA VILLAGE INFRASTRUCTURE PROJECT (Loan No IND) Preface This is the Implementation Completion Report (ICR) for the Village Infrastructure Project (VIP) for which a loan of US$72.5 million equivalent to the Republic of Indonesia was approved on May 23, 1995 and made effective on August 25, The Loan was closed on September 30, 1998, the original closing date. The final disbursement took place on February 22, 1999 at which time the loan balance of US$ 6,850, was cancelled. The ICR was prepared by Frida Johansen, Principal Economist and Task Manager. It was reviewed by Messrs. Dennis de Tray, Country Director, Indonesia and Jitendra Bajpai, Manager by the time of project completion of the Transport Sector Unit, East Asia and Pacific Region. Preparation of some ICR tables began in February 1998, by Ephrem Asebe, Consultant (EASTR). The ICR is based on materials in the project file including data provided by the Government of Indonesia. BAPPENAS commented on a draft of the ICR and provided the Govemment's evaluation (Part EI), including their assessment of the project's execution, effectiveness, and impact.

5 -iii- Implementation Completion Report INDONESIA VILLAGE INFRASTRUCTURE PROJECT (Loan No IND) IMPLEMENTATION COMPLETION REPORT Evaluation Summary Introduction The Government requested the Bank assist its new poverty program in This project was the result of this request, and it started implementation in Project Objectives The objective was to reduce poverty in poor rural villages in new ways, by promoting village participation, transparency and decentralization. Each selected village was to build infrastructure of its choice, to receive a) a one tirne grant, then equivalent to some $54,000, via the local commercial bank branch and b) technical assistance, directly, via a consultant field engineer assigned to a cluster of 5 villages. The local public work agency assisted with the public administration. The central government organized and managed the system, including procurement, training of the engineers, audits and monitoring. The project also was to fund a study of non-infrastructure constraints to trade and transport in rural areas, with the objective of addressing the main constraints identified. Implementation Experience and Results The above objectives were met, except addressing the non-infrastructure constraints. The project was implemented in 2 years. A total of 1230 villages compared to 1200 foreseen built a total of 3,680 km of rural roads, 7,790 meters of bridges, 2,427 water systems, 1,230 communal sanitation units and 2 piers totaling 61 meters. Quality and price were overall good. Benefits from the infrastructure are significant, in addition to the construction having provided employment for cash to villagers. Transparency was considerably better than under other projects. The field engineers gained experience that is valuable for subsequent activities. The main concem remains future maintenance, which is the responsibility of the villages but villagers may not appreciate the importance of preventive maintenance, and don't like to do it. The study on constraints was well done. With almost all costs in local currency, the rupiah depreciation reduced the loan dollars needed for the program. Thus the loan was able to also finance some preparation of the VIP2 (this project followed seamlessly starting implementation in FY97/98) and preparation of a kecamatan development project during VIP's third allowed year. The pilot project proved more difficult than envisaged including difficulties with the consultant retained and definition of an approach; eventually the funds spent in infrastructure and economic activities by the selected villages seem to have a good economic return but it may not accrue to the poorer villagers as emphasis was placed on repayment potential (the Kecamatan Development Project started implementation late in FY98). The loan also financed a study on rural road construction methods, that did not provide new insights. Loan funds were left over despite these added activities as their cost in dollars turned out to be about a third of the original estimate due to the 300% depreciation of therupiah in FY98. The project was subject to intensive supervision and scrutiry, but the supervision cost was no more than for other projects because of its short implementation period. Three independent reviews were also made, by QAG, the Water Unit of the Bank/UNDP, and an NGO hired under the Bank aegis, that found the project to be positive.

6 Village Infrastructure Project - iv - Summary of Findings, Future Operations, and Lessons Learned The project is considered to be highly satisfactory. The project demonstrated that a large scale project can be used to address village level needs at considerable speed (as replicating franchises) and that villages can be the "subjects of development rather than the object of development". Also, that task managers should not simply look up "best practice", but analyze local needs and find local solutions with the local people; and that Bank managers taking what they perceive to be risks is needed to allow innovation. That the strong government commitment to the project played a major role in its success is no surprise. Very good consultants at the center were given latitude and were instrumental in achieving the objectives. The project was designed to be a short term very simple program, and this indeed facilitated achieving the desired results and to move on, including to less simple programs. Concems remain about maintenance, but this, and increasing transparency and participation are gradual processes that have been given a push. The study on constraints was distributed widely and subsequently served as an input to policy reforms. The review of poverty lines used for Indonesia, indicates the need to review the appropriateness of the methodology used in their estimation. The VIP was succeeded by the VIP2 (Ln.4100-IND, $140.1 million FY97) and by the Kecamatan Development Project (Ln.4330-IND, $225 million FY98). VIP2 extended the approach to an additional 4,500 villages on Java and on Sumatra, and has completed its second year; having 2 projects rather than a larger initial one seems to have been the right approach. The KDP expands the VIP approach to assist villages by subdistrict (kecamatan), with both infrastructure and economic activities. It began in 500 kecamatan and will expand over 3 years to 1500 districts throughout Indonesia. Other projects adopted some of the modalities the VIP pioneered, both in Indonesia and in some other countries. The project team has become wary of the usefulness of pilots (from the KDP) and of independent monitoring. Pilots take as much effort as a full fledged project. Independent monitoring does not provide deep and early insights to enable corrections to the project design on the go if needed (if a country has good monitoring capacity, which is unlikely, monitoring could be more useful). In-depth supervision by staff familiar with the project and who know what and where to look for, get more reliable information, can translate it into corrective action faster while keeping an "institutional memory". However, some independent reviews are needed.

7 - 1 - Implementation Completion Report INDONESIA VILLAGE INFRASTRUCTURE PROJECT PART I: IMPLEMENTATION ASSESSMENT A. STATEMENT/EVALUATION OF OBJECTIVES Background 1. Poverty reduction was one of the main objectives of the Government's Five Year Plan Another objective was decentralization. While earlier programs had been effective in reducing poverty, by 1994 there were still some 25 million "absolute" poor and many more "near" poor, in a population of 190 million, some two-thirds of them in rural areas. The instruments foreseen to achieve the poverty reduction objective were a labor intensive outward oriented growth, human resource development and a new program targeted to poor villages (IDT). The IDT program provided 20,000 poor villages with Rp2O million each (then almost $10,000 equivalent), 3 times for the larger villages, 2 times for the medium sized and one time (year) for the smallest ones. The villages were the bottom third in each district, identified through a census based on a number of indicators, and mainly rural. The funds were to provide capital to the poorer households in the poorer villages; beneficiaries had to repay the funds so that other poor could benefit in turn. This program was fully funded by the Government. 2. In parallel to the IDT program, the Government sought to continue to provide infrastructure to poor villages, but in a more effective manner than under most earlier and ongoing programs. Some programs had provided experience on labor intensive methods and showed they could be effective in Indonesia. The Government requested foreign agencies to assist with the village infrastructure provision, in a decentralized manner. In response, OECF financed a 1994 study that identified the needs and the cost of provision of minimum village infrastructure, by area. The Bank agreed to assist with a project as poverty reduction and decentralization were in line with the Bank's priorities. The Bank also obtained a PHRD grant for independent monitoring of the IDT program, towards gaining lessons from it for future poverty projects. A completion report dated is being issued separately for the grant. Project Objectives and Components 3. Objectives. Thus far, the Bank had assisted road, water, agriculture and other infrastructure projects in Indonesia, mainly defined and implemented by national ministries and contractors, and with little beneficiary involvement. Objectives of the project were to innovate, more specifically to a) promote village participation, empower villagers to decide priority uses of a grant and to implement the works; b) provide public infrastructure needed in poor villages; c) create jobs paid in cash for unemployed/underemployed villagers especially in the seasons of low agricultural activity; d) increase transparency; e) mobilize villager contributions; f) support the Government decentralization agenda; and g) reduce non-infrastructure impediments to trade and transport in rural areas. 4. The planning aimed at project completion in two years. Bank management requested that three years be allowed for implementation and that the project be considered a "pilot project". 5. Components. A total of 1200 villages on Java were to be assisted. Each selected village was to receive Rp120 million (then equivalent to about $54,000) towards public infrastructure of their choice. The loan categories were kept to two, very simple: "goods, works and services under grants", in order to allow for villager choices, and "consultant services" for management, implementation and monitoring of the village works, and for a study of constraints to trade and transport (the Govemment funding the administration cost). Despite the programmatic nature of the approach but because of its newness, Bank management asked that an amount be left 'inallocated".

8 Village Infrastructure Project As the project objectives were met and the components completed in 2 years as planned, and at less than the foreseen cost in dollars, it was agreed that the loan also finance: a) a study to review the desirability of expanding the VIP approach to Sumatra, in preparation of a VIP2; b) a study on the effectiveness of altemative approaches to building rural roads, as it was found that access roads were the top infrastructure choice of villagers; c) data collection for one audit, as the project design required the documentation to be kept inthe villages while for the first year, the national audit agency requested the documentation for the sample villages be available in Jakarta; d) a kecamatan (subdistrict) development pilot and completion of a Local Institutions Study in the third year (FY97/98) allowed for project implementation. 7. Covenants. The covenants related only to project implementation. Bappenas indicated that it could not accept dated covenants relating to the removal of non-infrastructure constraints to trade and transport, as it could not guarantee implementation and did not want to default on any covenant, but that it would try to act on recommendations that arose from the study, and the draft action plan included in the appraisal report (Report No IND, April 26, 1995). Overall Assessment B. ACHIEVEMENT OF PROJECT OBJECTIVES 8. The project has been highly satisfactory. More villages than planned were assisted (1230vs 1200). A total of 415 and 815 villages were assisted the first and second years (FY96 and FY97), compared to 400 and 800 planned. The Bank chose to assist on Java, while OECF assisted villages off-java where contractors had more of a role (due to lower population densities and longer access roads needed, and thus insufficient labor for labor intensive techniques). The first year villages were selected by Bappenas in order to facilitate an early start; for the second year the provinces and districts made the selection from a long list of eligible (IDT) villages prepared in Bappenas. The village selection was appropriate. All the villages were poor, needed infrastructure; and were eager about the project. Assessment by Objective 9. The project fully met its objectives. It has served as a demonstration project, and its decentralized modalities have been adopted in several more projects (especially the new disbursement mechanism, management by the villages, technical assistance directly to villages instead of to government agencies). It also helped prepare two follow on projects. More specifically, by objective a) Villagers participated and were empowered to decide priority uses of a grant. Village meetings were held to select the public infrastructure to be built and to set the level of the incentive payment to workers. The Project Manual provided examples of 5 types of public infrastructure (roads, bridges, drinking water, communal sanitation units, piers); unfortunately most people understood this list to be restrictive. Nevertheless, villagers chose based on their priorities; sometimes the village head had undue influence but his choice tended to reflect the consensus, because the infrastructure needs had been felt for a long time and decisions on priorities were generally quite clear. Mostly as was desired, the district or subdistrict officials did not influence choices. b) Public infrastructure was provided in poor villages. In only 2 years, compared to the average Bank project implementation period of 6-7 years, the project assisted the building of 3,680 km of rural roads, of 7,790 m of bridges, 2,427 water systems, 1,230 communal sanitation units and piers totaling 61 m. Villagers had to agree to undertake maintenance of the infrastructure with their own means (the Government cannot finance maintenance in 60,000 villages), and field engineers returned to the villages in subsequent years to assist with maintenance. Maintenance however, needs further encouragement. Roads cost only $10-15,000 per kilometer. Villagers were able to build bridges even 25 m long, for 4 wheel vehicles (and 100 m long suspension bridges) at some $800-$1,000 per linear meter; a contractor-built bridge would have cost 2-3 times more.

9 - 3 - Implementation Completion Report c) The project created jobs paid in cash. The project gave equal opportunity to women, men and youth; local cultures resulted in women workers generally being a minority, though growing, and still sharing in the benefits even if they did not work. The number of village workers totaled 314,924 men, 41,770 women, and 117,552 youth, who worked for a total million days (unskilled), 1.1 million days (skilled) and 0.54 million days (leaders). Workdays averaged 10,000-15,000 per village. More people sought work than could be hired so generally a rotation system was used, often supplemented by a target system to facilitate correct payment per output. The objective of doing most of the work off the agriculture season (mostly the dry season) was only partially met partly because of the fiscal year cycle of budget approvals, but this did not dissuade participation. The incentive pay was around Rp3,000-4,000 per work-day of 6 hours (equivalent to $1-1.50). Some of the work-days were not "additional", as some villagers decided to stay in the village to do the works rather than to go for short-term jobs in town, that while they paid more also entailed expenses and left neither a higher net pay nor the local infrastructure benefits. The project also provided work for 83 field engineers the first year and 163 the second year, and for the senior engineers supervising and managing the project. (details in Tables 5) d) Transparency was achieved (and improved upon in the VIP2). Large sign boards were placed in front of the village office, indicating inter alia a) the grant amount (a standard amount across villages rather than variable amounts reflecting the cost of subprojects was intended to facilitate transparency, and it did); b) the daily incentive pay, that villagers knew and was the same for men and women and tended to be lower than the prevailing minimum wage; c) the cost of the subprojects; d) the payments allowed for administration in the village. Villagers had to sign against their payments. The village kept books and documents to substantiate the expenditures. e) The project mobilized villager contributions. Contributions were nrminly in the form of village land for roads, but also in the form of work at less than the minimum wage, even communal work at no pay, in order to make the grant pay for more materials and build more infrastructure. Contributions were not recorded systematically (as they were not mandated), leading some outside reviewers to question their existence. Reviews indicate that contributions were no less or even more than under community projects with mandated contributions (mandated contributions generate the wish to avoid them, and it is easy to "manipulate" the numbers...). f) The project supported the Government's decentralization agenda. The grants were disbursed directly to the villagers from the local branch of the commercial bank selected (BRI), bypassing some lengthier and less transparent bureaucratic processes. The project demonstrated that villagers could be entrusted with project definition and implementation. Several similar projects have followed (see H. "Future Operations"). g) The project did not achieve reduction of non-infrastructure impediments to trade and transport in rural areas during the project period. The related study was well done and identified many problems in rural areas (the study was the first such study in Indonesia). While no follow up was made initially, the study has been widely read and used, and has served as a basis for a number subsequent initiatives. In 1998 the central Government asked local governments to eliminate a number of the identified locally imposed restrictions; compliance was being verified starting in December The project achieved more than its objectives. The consultant field engineers gained valuable experience on the job and had at least 2 training sessions per year, that enhanced their skills. Capacity was built up at Bappenas. For the VIP2 and its extension into Sumatra, a field study/rapid appraisal of the feasibility of the VIP approach in Sumatra was carried out; the study proved less satisfactory than desired but was still useful. A pilot was carried out in 6 kecamatans (2 in each of the provinces of West Sumatra, Central Java, and East Timo&) for a Kecamatan Development Project; the pilot expanded the VIP infrastructure menu to include economic activities and credit, provided funds to targeted kecamatan rather than targeted villages, and added "facilitators" while reducing field engineer assistance (reflecting the preferences of the anthropologist who became task manager of the project). The pilot was fraught with difficulties, but served the purpose. The study of alternative methods for building access roads did not provide new insights. The consultant inventoried works and their cost, condition and construction method but did not analyze the data systematically. 1 West Sumatra, kabupaten Solok, kecamatans Lembah Bumanti and Lembang Jaya; Central Java, kabupaten Sukoharjo, kecamatans Weru and Bulu; NTT, kabupaten Belu, kecamatans Malaka Timor and Lamaknen.

10 Village Infrastructure Project Economic Rate of Return 11. The economic return from the infrastructure is satisfactory, of at least 20% overall, though variable across villages. Rates of return were estimated from a sample of villages. The return on the infrastructure, mostly from quantifiable transport cost savings as road building absorbed almost 80% of the grants, would be a minimum 20%, even if the infrastructures were to last only 3 years while they are expected to last considerably longer (bridges and water systems should last 20 years; the life of roads will depend on the maintenance effort). Average transport cost reduction was 30% in the second year program villages upon completion of their subprojects (by March 1997). Water subprojects had high returns; sanitation subprojects were the least beneficial. In general, villager preference is first, to have access, meaning they will choose to build a road and bridge if one is needed. After access is secured, the choice is drinking water. Only when this is also assured, do villagers choose to build sanitation systems. Unquantified benefits accrued from the experiences of bookkeeping in the village, and of the "transparency" under the project. 12. A review in early 1998 of 172 project villages, revealed that in 1997, one or 2 years after subproject completion, increased economic activity generated an average Rp30 million per village and savings in existing activities amounted to some Rp2O million per village in 155 villages. In addition, the skills acquired under the project enabled new activities of an average Rp20 million cost in 73 villages. Other benefits were increased school attendance, the employment benefits, and electrification (made feasible after an access road was built). Only 7 villages did not report any of these benefits. These impacts are detailed in Part III. The 2 basic benefits alone imply a one year return of 38% on the village grant and the technical assistance cost, and a rate of return much higher, though local contributions reduce returns on total costs. 13. Benefits also accrued outside the villages. A majority of the field engineers are still working under VIP2 and will be able to continue this kind of activity on their own later on. Some Bappenas staff involved in VIP subsequently became managers of other similar projects. A good data base was established in Bappenas, identifying all villages by their national statistical code, that will allow review of impacts later (one is planned at the end of VIP2) and coordination of assistance programs by location. And the demonstration effect was high; it has resulted in several other projects adopting at least some VIP features. 14. Infrastructure and economic activities were also achieved under the pilotkecamatan project. Funds were allocated in similar proportions for infrastructure grants and for subloans. Economic activities generally have much higher returns than infrastructure; however, they have fewer beneficiaries, at least initially. Revolving of the subloans for economic activities is continuing. Private sector participation 15. Participation has been discussed above, as it was one of the project objectives. The partnership (public grants, local management and work) proved positive; non project villages were inquiring when their turn was. Environmental impact 16. The environmental impact is generally neutral. The new roads allow motorized and increased traffic, and this has some impact. Road widening was the more common and only involved cutting the border vegetation. The larger impacts came from opening up new roads. Poor villages are often in mountainous areas; opening new 3-5 km access roads often involved moving large amounts of earth from deep cuts, and fill, and still roads are steep. Care was taken to follow reasonable alignments and reduce the potential for landslides; occasionally, landslides occurred and were repaired. Attention was also paid to including appropriate culverts for sustainable drainage and avoid erosion of soft soil covers. Project implementation 17. Project implementation proceeded fast, as planned. It started with governmentprefinancing, ahead of loan effectiveness (and also prefinancing of the advance to villages the second year). Villages were clustered in groups of 5 to enable their assigned field engineers to spend as little time as possible on the road. Every village completed its

11 - 5 - Implementation Completion Report program within a year. Every province had a senior engineer. Only a thin consultant staff (3 senior professionals) was located in the Jakarta Project Secretariat for general management. The Secretariat handled all the consultant procurement, managed the performance indicators (MIS) that were updated monthly, and managed the independent monitoring. The National Bureau of Statistics conducted a household expenditure survey in 120 project villages, with the standard national questionnaire, that allows later comparisons. A non-profit organization did an independent monitoring the second project year on a first year village sample; the effort proved difficult and yielded little extra information, and was thus not repeated for the second year villages. The Borrower performance section expands on the implementation topic. The kecamatan pilot was implemented the allowed third project year. There were problems with contracting the NGO, the NGO experienced cash flow problems and delivered only partially. The Project Manual for the pilot was revised even as the pilot was underway, and this created confusion and delays in the kecamatans. Project costs and contracts 18. The original project cost was estimated at $83.8 million equivalent, and the actual was $75.2 million. The cost was as estimated in rupiah, as all costs were in rupiah except for two contracts that included a small foreign exchange part. The lower dollar costs arose from the devaluation of the rupiah (from 2,200 per dollar at appraisal to 5,000 in 1997) and late claims by MoF for reimbursement of the village grant advances or for payments to consultants. The grants to villages, at Rpl20 million each for 1,230 villages, totaled Rpl47.6 billion or about $61 million. The 4 engineering contracts (#2, 3, 4, 6), covering the 2 years, cost the equivalent of $8.1 million, or some 14% of the cost of works being managed and supervised. Monitoring cost the equivalent of $365,276 (contracts #5, 7, 11). The study on constraints in rural areas cost $650,228 equivalent (contract #1). For cost details, see Table 4, Consultant contracts, and Tables 8, Project Costs. 19. The added components cost about $1 million equivalent. The road study cost $60,670 equivalent (contract #13). Preparation of VIP2 on Sumatra cost $58,767. The pilot kecamatan project cost about $860,000 equivalent, including the kecamatan allocations. The allocations to the 6 pilot kecamatan, at Rp400 million each, totaled Rp2,400 million or $570,000. A consultant firm was hired to manage and carry out the pilots, and prepare the follow on project; the contract with the firm was equivalent to about $640,940 (total payments were $195,524 equivalent; the consultant had claims outstanding; contract #15). Completion of a "Local Institutions Study" reviewing rural villages was also financed (at an original cost equivalent to $157,762 and final cost $67,578 equivalent; contract #14). And, the advisor originally hired for VIP was retained to advise start up of the KDP, at $26,672 equivalent (contract#16). This was supplemented by a Japanese Grant of $250,000, that was fully disbursed. 20. A total of 14 contracts were undertaken under the project (see Table 4). Most went smoothly and were completed within the original cost. But, a) as the study to review the effectiveness of alternative road construction approaches proved disappointing, the final payment was withheld. The firm proposed to undertake further work to justify the final payment but this was not accepted by Bappenas and the Bank, that considered that the consultant would not be able to deliver; and b) as the consultant firm undertaking the kecamatan pilot experienced staff and funding difficulties, the contract was eventually suspended and some staff who had worked and not been paid by the firm were paid as individuals under the Bank's consultant budget. 21. The loan was $72.5 million, to finance 90% of village grants and 100% of consultant services. The consultant payments totaled $9,637,440 equivalent (under the $11 million appraisal estimate). The contracts had much smaller unused balances in their original currencies. Disbursements in 1998 were made at about 1/4 of the appraisal exchange rate (the rupiah devalued to 8,000 per dollar). The final disbursement was February 22, 1999 for Rp3,933,356,237, or about $449,526, equivalent; $3,611,203 were recovered from the Special Account in February 22, $43,947 were lost due to exchange rate fluctuations on disbursements related to the Special Account. Bappenas did not have easy access to information on real loan balances. A total of $6,850,505 was cancelled upon closing the Loan Account in March 1999.

12 Village Infrastructure Project C. MAJOR FACTORS AFFECTING THE PROJECT 22. Factors generally subject to government control. The simplicity of the project and the commitment to it by Bappenas, MoF and MoHA, and by the Bank, greatly facilitated its implementation which went smoothly. Because of the time standard processes take, the Government prefinanced the start of project implementation the first year (to enable works during the dry season) rather than delay it for another year awaiting loan effectiveness. 23. Factors not generally subject to government control. The devaluation of the rupiah, though small in the basic project period, resulted in the loan allocations exceeding those needed to achieve the original objectives. D. PROJECT SUSTAINABILITY 24. The project modalities have been demonstrated to be sustainable. However, targeting villages was not intended to continue, as with increased coverage targeting becomes less meaningful; it was the intention to assist poor villages once, and then switch to "competition" for funds: this was started under the KDP in Indonesian consultants have acquired skills and are able to continue similar activities. The pilot revolving funds continue to operate, and they continue to be audited. 25. The "sustainability" of the infrastructure built, and thus benefits derived from it, foremost depend on maintenance. The technical quality of construction was generally adequate, but design standards were kept low in order to keep costs low. Every year since the village built infrastructure under the project, a field engineer has returned to assist with maintenance. During 1999, with funds from the VIP2 loan, repairs needed will be inventoried and funds for materials and technical assistance will be provided to those villages where the problem is due to faulty construction (that could be said depended at least partly on the field engineer) and not due to lack of maintenance (that depends on the villagers); the villages will have to provide the labor to do the repairs. 26. Roads (that took up some 80% of the village grants) were to be built to gravel surface only, while allowing steep slopes to be surfaced with asphalt or concrete. Most road construction followed the rules, and roads can be expected to last 5 to 10 years; their quality is as good as that achieved at higher cost under contracts by the department of public works. But, often with support from the public works district office, many villages proceeded to asphalthe roads at considerable cost; this makes the roads last longer prior to maintenance being first needed but also makes it more difficulto do repairs once they are needed. Some steep road sections on the other hand were left with gravel surface, and heavy rains take a heavy toll on unprotected slopes. Also, trucks heavier than the new roads were built for, started to use them with damaging effect; not all villages installed portals to prevent heavy trucks using the roads, particularly when they are wet. Villagers understand the importance of maintenance and prevention. 27. Road bridges have been built to withstand truck weights, and should last for many years. In cases, the wooden deck may need replacement of some planks in a few years. Another area of concern are foundations, how well they were designed and are maintained to withstand erosion-often the bridges built under the project replace an older bridge that had collapse due to poorly protected piles and side walls. In fact, bridges built by the department of public works should be inspected more regularly to avoid the recurrence of costly collapses, that could be prevented at very little cost. Some bridges built under the project are suspension bridges for pedestrians or motorcycle traffic only, and they should last for considerable time, also. 28. Water systems normally have proven to meet real demand. They were provided at competitive cost. The modalities for maintenance vary, but given user satisfaction it is likely that systems will be maintained. In some coastal areas wells did not provide drinking water (due to saline intrusion) but nevertheless provided water for other uses and were appreciated; a problem there is that the pumps tend to rust. 29. Communal sanitation systems were not in much demand. Some of those built could have had a more appropriate design had the main users been consulted about their preferences. Lack of familiarity with toilets and little cleaning, meant that some toilets were not used for long. Public bathrooms including toilet, bathing and washing places and being used by as many as twenty families, were built mostly were housing density is high and the need was felt.

13 - 7 - Implementation Completion Report 30. Only a few piers were built; most of the poor villages selected were in uphill areas. And very few "other infrastructure" were built, mostly probably because the understanding of the field engineers, that could not be straightened even in the first year of VIP2, that only the five types of infrastructure given in the example could be approved. 31. There is a trade-off between standards and quantity affordable: the project chose relatively low design standards in order to facilitate labor intensive works, to make the funds reach further and thus to benefit more villages, and to obtain the best possible economic return. Building public village infrastructure is a continuous process; there is a recurrent upgrading as time passes, the old infrastructure deteriorates by use and even simply by weather, and demand grows, that justify somewhat higher standards the next time around rather than only maintenance. Identification E. BANK PERFORMANCE 32. When the Government launched its IDT program, the Bank was invited to participate. After various internal discussions, the Bank felt that this program was too risky to support. A Bank identification mission addressed Bappenas request and proposed to proceed with a complementary village infrastructure project. Many studies were available on rural needs, among them the OECF study mentioned above, that were perused to identify what a project could do and at what cost. It was also identified that the largest number of poor and most easily reachable at lowest fiscal cost was on Java, and that some earlier programs had been successful in addressing village needs there (such as a Yogyakarta Uplands project). A team consisting of a Bappenas staff and a Bank staff proceeded to visit rural villages on Java to test the viability of leaving the decision making and implementation to villagers themselves, and to review the banking support available in rural areas to channel funds directly to the villages. The reviews (that can be called rapid appraisals) were positive, and the Bank management eventually approved the approach. 33. The Bank obtained a PHRD grant (TF029182, signed July 31, 1995), foremost to assist the Government monitor the IDT program. Bappenas executed this grant (which is the object of a separate completion report). Preparation 34. The Bank financed a consultant for 6 months, mainly to assist Bappenas with the identification of villages to be assisted in a first year program, though he spent much of this time on more general matters. Another consultant, with much experience in labor intensive works on Java and project management, was engaged shortly thereafter to start preparing the Project Manual and detailed implementation modalities of the project. In parallel, the required Bank documentation was written and the various steps accomplished in a short time. The preparation cost was minimal as existing knowledge, both in Indonesia and in the Bank for other projects, was tapped. Simplified procurement forms and systems, and compensation guidelines, were developed that could meet Bank standards while being understandable to villagers. Throughout preparation there was a constant exchange of faxes between the Bank and the Bappenas team working in Jakarta; responses were fast as was progress. 35. The greatest debates on the project design in the Bank, were a) A number of staff thought that giving money to villagers instead of channeling it through ministerial agencies would result in loss of "control", and waste. But as the agriculture projects were implemented slowly and reached only small numbers of people, management supported proceeding with the project innovations that provided a potentially faster way to reach more people at lower unit costs and getting the beneficiaries involved-this was a main reason to call the approach a "pilot". b) Between the task manager, and some staff who had worked with the household expenditure survey and considered misguided the proposed focus on Java on account of their estimated greater poverty incidences in the outer islands. Sheer number of poor, population densities, economies of scale, lower prices, supervision possibilities and probability of success being greater on Java prevailed in keeping the focus on Java.

14 Village Infrastructure Project c) Between the task manager and the lawyers who wished to have only "standard involuntary resettlement and compensation guidelines". Eventually, simple guidelines including the possibility of voluntary contributions (without compensation), were agreed--and used in several other projects. Appraisal 36. The appraisal took place 9 months after the first identification mission. It consisted of reviewing the completed draft manual, agreeing with the selected bank on the mechanism to channel funds to villages, agreeing on the list of villages, and seeing that the central and local agencies that needed to be involved were ready. Not only were terms of reference ready but advance procurement was followed to ensure that the management team and the engineers needed for the project could be in place quickly and trained even before loan effectiveness as the government wished to proceed fast. Several ways of monitoring were also agreed on, given the "pilot" nature of the project and the management request for quick feedback to enable taking corrective action quickly if one was needed. Supervision 37. On average 3 missions per year were undertaken, and during each of them subteams formed to visit each of the four provinces of Java. The teams visited project villages at random to discuss with villagers, see the works and the appropriateness of the village selection, check the local bank accounts, discuss with local project managers, field engineers and senior engineers, etc. To facilitate evaluation of progress across the various teams, standard supervision questionnaires and reporting formats were developed and followed. There is thus a good supervision data base on the project progress. Bank supervision missions proved effective for understanding what went right and what needed adjustment. Essentially based on the supervision results, and knowing the Bank's lead times, the same project team started preparing at the end of the first project year a Second Village Infrastructure Project in order to continue with a 3rd and 4th VIP years on Java, and extending the approach to Sumatra. 38. Independent monitoring of villages proved more difficult. Several monitoring exercises were done: a) an NGO was chosen following competitive bidding to review a sample of villages of year 1. Eventually, an individual consultant was hired to assist the NGO improve its analysis and the presentation of results so they could be more useful for others; but representing qualitative data as generic results was not easy; and b) ad-hoc SUSENAS surveys were undertaken in project villages, that would facilitate comparison of trends in them and in non-project villages. The SUSENAS survey did not demonstrate significant differences across project villages and non-project villages, perhaps because it was too soon to show. Supervising the independent monitoring took considerable time. It was decided that it was not worth replicating the monitoring the second year (and currently, following the economic downturn, it would be difficult to see significant project impacts from SUSENAS results). 39. The loan agreement was amended twice. First the loan, denominated in a basket of currencies, was amended the second project year for the balance to become a single currency loan (dollars). Second, the loan was amended to permit the pilot kecamatan project, mainly to operate outside Java; other additions to the project could be accommodated within the generic "consultant services". Disbursements were monitored throughout, some were also audited by the Bank team. There is a problem with the Bank's disbursement data system: for contracts not denominated in dollars, the system provides a loan disbursement dollar value based on the exchange rate of the day of the enquiry-thus the real loan disbursement and balance in dollars cannot be known when the exchange rates fluctuate. This system deficiency is serious with large exchange rate fluctuations, as was the case in Indonesia: it was part of the reason for the large loan cancellation. The Loan Department has been urged to tackle this problem. 40. The pilot KDP was almost cancelled, but eventually was continued. It required intense supervision.

15 - 9 - Implementation Completion Report Overall 41. The total Bank direct cost for the project is below average ($414,000: $111,800 for preparation, that focused on project start up activities; $47,100 for appraisal to effectiveness; $240,000 for supervision, intense supervision compensated by a short implementation period (essentially only a 2 year implementation period, compared to 6-7 years for most projects) followed by a third year mostly for preparation work for follow on projects. The completion report bank direct cost is about $15,000. No grant funds were used. 42. From identification to completion, that took about 3.5 years, the project had the same task manager/economist, the same engineer and continuity in other team members, in particular the anthropologist, who assisted with field reviews (all the Bank managers working on Indonesia at the time of the project, are gone). Preparation and Implementation F. THE BORROWER PERFORMANCE 44. The Borrower showed great commitment to the project since its beginning. A full time, dedicated Bappenas staff was assigned to the project, and quickly developed the Secretariat under which the project was implemented. This person and the Bureau Chief stayed with the project from beginning to end and gave it all the attention and support needed; they were prompt in communicating with the local level agencies involved, gave latitude to their consultants in their tasks, and followed up on problems. Unfortunately, the Bappenas Deputy who started the dialogue passed away early in the process, but his successor showed equal support. One high level official in Public Works objected to Bappenas being the implementing agency, and the bypassing of the district level public work agencies except for administrative approvals (indeed Bappenas, the planning agency, normally does not implement projects, but in this case the involvement of multiple ministries, the new approach, and the desired speed, to some extent justified Bappenas being the implementing agency). Pimpros, though initially weary of a new modality, ultimately endorsed it; their operating budget was often late orunderfunded; this shortcoming has been addressed under VIP The Secretariat provided a forum for representatives of other ministries to be involved and assist with implementation. Representatives of the ministries involved (Bappenas, Public Works, Home Affairs, Finance) assisted with the definition and improvement of the Project Manual and other materials, participated in field trips to review progress, and eventually not only endorsed the approach but adopted it for other programs. The Ministry of Finance prefinanced village grant advances before the loan became effective, to allow an earlier start (see also para.49 below). Procurement was handled expeditiously. The Secretariat coordinated area coverage with other projects (e.g. OECF's) and was prompt in requesting the follow on project. 46. The implementing consultants performed at top rate. The advisor hired originally by the Bank to help Bappenas with village identification helped achieve fast project preparation, continuous dialogue by fax with the Bank task manager; he was retained by the Secretariat to continue assisting implementation, of the VIP mainly with monitoring/procurement activities, and of the Japanese Grant obtained in parallel for poverty reduction activities, with efforts to monitor the IDT program (and continued working for the KDP). The senior engineer, hired during project preparation to draft the Project Manual and set the various processes in motion before appraisal, formed a good team with the advisor, and remained in charge of the consultant engineering/management team overseeing project implementation, the field engineers and the information system. All project data is recorded in a computerized data base against the Central Statistical Bureau codes for the project villages, and will allow future impact reviews; qualitative surveys were also made by field engineers. The senior engineer/team leader also organized training for the engineers and field engineers, about 2 times per year, a week each time; this training is a valuable project output in itself. The team leader performed at the highest level till the project end (and is managing the VIP2). 47. Independent audits proved difficult and did not add much. For the first year results, as the national audit agency may not have recognized the difference in the project's procedures compared to standard ones for projects implemented by contractors, it requested documentation in Jakarta that had then to be assembled from the

16 Village Infrastructure Project districts/villages. A consultant was hired by Bappenas to assemble the documentation, and paid under the loansubsequently the same audit agency indicated this was not needed. No consultant was hired for the second year audit; however, this imposed a considerable burden on the Secretariat itself, as again, BPKP requested more documents than the project design. Eventually, all the audits were "non-qualified" (and a different approach followed for audits of VIP2). The regional audit agencies also audited the project, and some were reported to have become a sort of "harassment" of villages. 48. A disappointing aspect was the lack of action on removing the constraints to trade and transport in rural areas, that could have had high pay off. An interministerial working level committee was established to review the findings of the study and suggest ways of implementing them. The knowledge spread (or was confirmed). However, as no clear action plan was forthcoming and meetings did not show progress, the matter was not pursued. Subsequently, deregulation initiatives seem to be proceeding. 49. The Government financed the advance (Rp2O million) to the villages both project years, even though this amount was more than its 10% share. Subsequent disbursements for the villages were made at the ratio 90-10%, on the assumption that the Government would seek reimbursement of the advance. However, it could practically only do so when the full Rpl20 million had been spent and disbursements proven. MoF getting the SOEs for the last payments to each village, when the villages no longer had to present papers to get further funds, proved quite difficult and delayed MoF's reimbursement requests. The difficulty was taken into account for VIP2. There were occasional delays in paying consultants. 50. The Government requested the loan agreement be amended from a multiple currency to a single currency denomination in dollars when it considered that this provided more certainty of funding availability. And also acted promptly to request an amendment to permit the kecamatan pilot. G. ASSESSMENT OF OUTCOME 51. The project outcome is highly satisfactory. Much infrastructure has been built, underemployed villagers have used their time productively and earned cash, the engineers gained experience. It is acknowleded that some infrastructure has turned out substandard, but even a 5% failure rate can be considered highly satisfactory. The field engineers have been coached and trained and will be able to perform better after the project; they have asked for certificates stating their participation in the project, which is widely respected in rural areas. The approach has been replicated, both in Indonesia and in some other countries. The Ministry of Public Works has accepted its effectiveness despite early doubts; other sectors have also adopted some of its features. Several countries (Guinea, Vietnam, Cambodia) sent teams to see first hand the project and assess the potential for village implementation in their countries; the project manual and other documents have been translated into French. The QAG team assessing the quality at entry of VIP2 largely based their views on the visible results of this VIP, and found the approach to be bold and several aspects to deserve "best practice" classification 2. The project was also the subject of other reviews, namely a) an international rural water study undertaken by the Bank, that found that not explicitly requesting villagers' contributions did not result in less contributions than when these are demanded, and that cost tended to be lower under VIP 3 and b) an independent NGO study, that suggested more 2 The QAG report found strengths: a) ongoing learning by doing, b) village participation, c) well conceived project design, d) strong government commitment, good leadership by individuals involved in designing and implementing the project, excellent supervision and engineering support; and weaknesses in: a) targeting (it thought the poorest may not have been reached), b) participation (in decision-making, as most subprojects were roads), c) replicability (it thought the project enjoyed a level of support other projects would not have), d) occasional inadequate design and insufficient reporting and follow up, e) concerns about sustainability and maintenance and about monitoring and evaluation. (Uma Lele et al., QAG Review of East Asia Rural Poverty Reduction Projects, Annex 4, 1998, internal document). 3 Gillian Brown, consultant, Responding to Demand: Two Approaches from Indonesia, 1997, internal document, that served as an input to the International Workshop on Rural Infrastructure, Washington, May 1997.

17 Implementation Completion Report NGO involvement, primarily, though it found the results to be satisfactory. More than the original activities were undertaken. 52. Some criticisms voiced have been a) that not enough village grass root participation took place in planning, and by women: may be; but participation was higher than in other projects and as time passed, both increased. The project approach was to provide "equal opportunity", not affirnative action that tends to preserve stereotypes. Women's preferences had to be considered in designs of subprojects they would use most, such as washing places/water outlets. b) the lack of required contributions; in fact many were made though not well recorded and this has been remedied in VIP2. c) that incentive payments were not set at a level below the official minimum wage; villagers were free to choose the level, and it is unfair to ask the very poor to work for less than is considered necessary for daily needs. In fact the relatively better off agreed more to lower pay to make more funds available for materials. d) that the standard grant amount did not match the cost of the village needs; this is true, but the grant level was chosen such that meaningful works could be made with it, and a standard amount greatly enhanced transparency. e) that other projects are unlikely to have the high level of commitment this project had, are therefore less likely to succeed and put sustainability in doubt; the question would then be, why proceed with projects that do not have a high commitment level? 53. Independent monitoring of the activities was not very satisfactory. Part of the problem was the culture, an environment in which few people reported bad news; the few bad news reported were seldom corrected but the reporter could face official pressures. Also, monitoring capacity in Indonesia is scarce (though improving). Thus independent monitorers of VIP had a difficult time. R. FUTURE OPERATIONS 54. The operation and maintenance of the infrastructure built is the responsibility of the villages, as the government could not, and should not, finance maintenance in 60,000 villages. VIP2 is assisting in 1999 to do construction repairs as may be needed, however. 55. Operations based on the VIP are mentioned below; the three additional Bank projects involved the same Bank staff: a) the VIP2, that expanded the coverage to Sumatra; the VIP2 was prepared during the second year of VIP, in order to avoid a hiatus in assisting villages at the completion of VIP, and it is starting a third program year, with a much larger coverage than VIP (Loan 4100, US$140.1 million, effective in 1997). b) the Kecamatan Development Project, that expands/adapts the VIP principles to selectkecamatans instead of villages; to comprise economic activities in addition to infrastructure (the EDT grants having run their 3 years and the rural credit markets still not working properly), and to provide "social" facilitation and less engineering assistance. By government request the coverage was greatly expanded compared to plans (Loan 4330, US$225 million, effective mid-1998). c) several Bank projects in other sectors (agriculture and education) have adopted the VIP modalities, in particular the simple disbursement mechanism; OECF has since adopted a mix of the VIP and its own contractor-based modalities; and the government has used it to some extent for its labor intensive programs in both rural and urban areas to palliate the economic crisis. d) an Urban Poverty Project has been negotiated that benefited from the VIP. The national audit agency participated in its appraisal (the first time in Indonesia) to ensure their concerns were addressed, and that they understood the project design. 4 Bina Swadaya, A World Bank Project Monitoring Participation, 1998, included in the Bank-NGO International Conference on Upscaling and Mainstrearning Participation of Primary Stakeholders, Washington, November 1998.

18 Village Infrastructure Project I. KEY LESSONS LEARNED 56. It was demonstrated that large scale projects can be used to address village level needs at considerable speed, and that villagers are able to be "the subjects of development instead of the object of development" as one Indonesian put it. Still, developing a maintenance culture is a long term proposition; the project has provided a push to it. Continuity of the project team permits good communications and fast actions. A short implementation period is better; it allows in depth supervision within standard costs. Commitment by government agencies to the main thrust was essential. Frankness of the Government indicating that it would/could not act on a problem for which a study or actions were recommended was appreciated, as it saved time, money and frustrating dialogue. Bank management hands-off was most beneficial. The executing agency should have easier information on actual loan balances, especially at later stages and when exchange rates fluctuate widely and impact on the dollar equivalency of disbursements. The project team intends to write a publication for a wider audience upon completion of the VIP The team has become wary of the usefulness of pilots. The KDP pilot implementation did not go as well as desired; it took as much effort as a full project (it needed a manual, decrees, hiring of consultants, training, loan amendment, etc.). The conclusion of the team is, that such pilots are hardly worthwhile. 58. Independent monitoring cannot be relied upon to detect problems quickly for corrections on the way, at least when local monitoring capacity is weak and foreign experts are costly, as in Indonesia. Lack of continuity and perhaps understanding of local circumstances may mean that what foreign experts learn is mostly lost, at least to the project. Constant supervision by project staff, government and Bank, who know what to look for and are not afraid of admitting problems in order to address them, is more effective and provides valuable insights for follow on activities. However, some independent monitoring (or supervision) is of course needed. 59. Involving the audit agency during project preparation is useful, to avoid the agency later requiring documents that are not foreseen in innovative project designs. 60. Recommendations on the infrastructure per se include: a) bridges built by the department of public works should be inspected more regularly at very little cost to avoid the recurrence of costly collapses, that could be prevented; b) sanitation system construction needs some parallel education to be fully effective, but such education is costly. Thus the justification of sanitation systems for rural areas should be looked at carefully; and c) faucets available for water systems are of poor quality; it may be better to do without them and use local approaches. 61. It is worth reviewing the methodology for using the SUSENAS data in the estimation of poverty incidences in Indonesia.

19 Implementation Completion Report INDONESIA VILLAGE INFRASTRUCTURE PROJECT (Loan 3888-IND) IMPLEMENTATION COMPLETION REPORT PART II: SUPPORTING TABLES AND ANNEX Table 1: Summary of Assessment A. Achievement of objectives Substantial Partial Negligible Not applicable Macroeconomic policies El l E m Sector policies F1 El m E Financial objectives El E El Institutional development ml El El Physical objectives [F El El Poverty reduction IN C E l Gender concerns El E El Other social objectives l El Environmental objectives al E E l Public sector management ml E El Private sector development El 0 0 El Economic benefits EN l 0 El B. Project Sustainability Likely Unlikely Uncertain 1A 0 1-E C. Bank Performance Highly Satisfactory Satisfactory Deficient Identification El El Preparation assistance El El Appraisal El El Supervision El El D. Borrower Performance Highly Satisfactory Satisfactory Deficient Preparation El El Implementation El Ela Covenant compliance El El Operation (if applicable) El El E E. Assessment of outcome Highly ighly Satisfactory Satisfactory Unsatisfactory Unsatisfactory mne E El

20 Village Infrastructure Project Table 2: Related Bank Loans Loan Title Purpose Year of Status l ~~~~~~~ ~ ~ ~~~~~~~~approva Preceding Operations: None Following Operations: Village Infrastructure This is a targeted poverty reduction intervention project Ongoing, Project II The main objective is to build small infrastructure in Satisfactory Loan 4100-IND, $140.lm poor villages on Java and Sumatra. Other objectives are to increase decentralization and community participation, as under VIP. Kecamatan Development This is a targeted poverty reduction project and capacity 1998 Ongoing, Project building project. It expands the VIP into economic Satisfactory Loan 4330-E1D, $225.0m activities and into allowing villages collectively in selected subdistricts to choose what to do, with village implementation. Municipal Innovations This is a leaming an innovation loan. It focuses on 1999 Starting Project capacity building, govemance, service delivery. Some Loan 4440-IND, $5.Om of the "municipalities" are substantially rural. Urban Poverty Project This project is also a targeted poverty reduction project, Not yet Board Credit.. -IND, $100.Om but in urban areas. It follows the community decision on equivalent expected participation approach and capacity building. A loan was negotiated in October 1998, conversion into a IDA pending credit, for which Indonesia is newly eligible, has delayed processing. Note: Several agricultural and education projects have also adopted VIP modalities, in particular the disbursement and community implementation, some through amendments and some new ones. Table 3: Project Timetable Steps in project cycle Date planned Date actual Identification N/A 02/15/94 Preparation 10//94 10/01/94 Appraisal 01/05/95 02/14/95 Negotiations 04/20/95 04/21/95 Board Presentation 05/09/95 05/23/95 Signing 06/95 06/29/95 Effectiveness 06/95 08/25/95 First Loan Amendment 5 08/25/95 Second Loan amendment 6 12/2/97 Project Completion 03/31/98 03/31/98 Loan Closing 9/30/98 9/30/98 5 Loan Agreement was amended effective August 25, 1995, with US$25.4 mnillion disbursed under a currency pool, and the remainder US$47.1 million converted into a in single currency, USD, loan. 6 To allow for the kecamatan development pilot.

21 Implementation Completion Report Table 4A: Loan Disbursements: Cumulative Estimated and Actual (US$ million) FY96 FY97 FY98 FY99 Appraisal estimate Revised Actual " Actual as % of estimnate 85% 106% 98% 91% Date of final disbursement 2/99 1 Including disbursements into the Special Account, the balance of which ($3.61 million) was refunded by the closing date. Table 4B: Consultant Contracts Contract Curr. Initial Value (w/o Not Invoiced Original US$ Total Paid Total Paid, taxes) equivalent Original Currency in US$ 1 Study of non-infrastructure constraints to trade GBP 321,675 11, , , ,228 and transport Rp 545,210,000 67,043, ,166,465 (at Rp2350) 2 Engineering services for West Java (Years 1 & 2) Rp 4,624,810, ,968,000 4,624,809,470 1,898,558 3 Engineering Services for Central Java and Yogya Rp 6,520,745,450 34,599,130 2,937,273 6,486,146,320 2,673,240 4 Engineering services for East Java (Years 1 & 2) Rp 5,414,690, ,439,050 5,414,690,690 2,219,018 5 Independent monitorng of year 1 results Rp 576,287, , ,287, ,843 6 National management support US$ 322, , , ,347 1,333,824 Rp 3,090,830,000 1,315,200 3,165,933,600 7 SUSENAS-VIP survey, of year 1 Rp 304,326, , ,326, ,877 8,9 # were not used (?). 10 Review of VIP approach applicability on Sumatra Rp 137,574, , ,574,000 58, Individual TA to assist independent monitoring US$ 27,500 8,250 38,068 19,250 28,556 Rp 22,409,040 22,409, Data collection for audit Rp 227,730,000 96, ,730,000 94, Review of rural road construction methods US$ 54,815 16,445 87,479 38,370 60,671 Rp 76,694,400 23,008,320 53,686, Local Institution Study (completion) Rp 385,730, , ,730,000 67, Services Kecamatan Development Project pilot US$ 109,680 36, ,943 73, ,524 Rp 1,298,938, ,938,350 1,045,999, Individual TA for KDP preparation US$ 26,672 26,672 20,000 20,000 Totals US$ 540, , ,296 9,637,440 _ GBP 321,675 11, ,456 Rp 23,688,400, ,598,222 22,923,733,9111 Total in US$ 384,417 11,209, ,637,440 Grants for Works Grants to villages (1,230 x Rp.120 m) Rp 147,600,000,000 Grants to kecamatan (added) (6 x Rp.400 m) Rp 2,400,000,000 Total Rp 150,000,000,000 = US$ 68.2 m at Rp2,200: US$62.2 million actual Disbursements at 90% $55,968,108

22 Village Infrastructure Project Table 5A: Key Performance Indicators for Project Implementation" Key Implementation Estimated Actual Indicators in SAR IHighly Satisfactory I. Villagers empowered to decide what to 1200 poor 1230 poor villages. build villages. II. infrastructure built by the above villages n.a. Details by year (Table 5B) and by kabupaten and province (Table 5C) a) roads 3,680 km b) bridges 7,791 m c) water systems 2,427 units d) communal Sanitation units 1,230 units e) piers 2 totaling 61 meters ILA. Jobs created n.a. (# of workers) a) men 314,924 b) women 41,770 c) youth 117,552 d) field engineers 83 (first year); 163 (second year) IIIB. Jobs created n.a. (# person-days) a) unskilled million b) skilled 1.10 million c) leaders 0.54 million d) field engineers 246 field engineer years IV. Mobilized private contributions n.a. -land and workdays (but totals are not available) V. Supported decentralization -transferred resources directly to villagers; -strengthened village institutions; demonstrated the viability of village communities being in charge for their own development; -prompted VIP2 and parallel government funded program; and -the pilot (enabled with fund savings) for Kecaratan Development Fund concept to follow the VIPs on a regular basis and no longer targeting the villages. Table 5B: Summary of Infrastructure Results 1995/ /971 Year Measures Roads Bridges Water supply Sanitary units Piers 1995/96 #villages Results 1349 km 3295 m 1,023 units 546 units 15 m Average 3305 m 16.5 m 7.2 units 7.8 units 15 m 1996/97 #villages Results 2337 km 4496m 1404 units 684 units 46m l AveTage 2913 m m This excluded the added kecamatan pilot: some 60 villages in 6 kecamatans were assisted.

23 VILLAGE INFRASTRUCTURE PROJECT FINAL REPORT 1995/1996 No. yjaiaomishments Mandays Targt and RealizationWokFrePoes Kabupaten of Ras Bridges Jetties Water Sanitary - Mandays Taro at I Mandas Realized JPson,, oo Vill. (molar) (Motor) (reter) (unf!0 (unit) Ordinary Skilled Leaders Ordinary Skilled Leaders Men Women Youth P Phys.1 Fin. Mdys. PANDEGAG ~ , LEBAK _100,0 98,1 SUKABUMI a 8 _ i114.0 r. CIAWUR a , GARtf _ , IND~RAMAYU _ _ O95 SERANG ~~~~ ,0O 108:3 BAWARNE~GARA 20 82, M.8W , KEBUMEN ,0 100, PURWJORE,JO , 0 WONOSOBO _ I _ ,0. GROBOGAN 15~ _ , 0.0 r' REME3ANG _ 31 0o i _ _ 383, ,0 100,0 0r PAT] , _ , BATANG =00 GUNUNOKIDUL , j,09,0 106,7 0' PROBOUJNGGO _ , PASURUAN _ i ,61 NGiANJUJK 10 38i a ,5 BOJONEGORO a ,0 98, 957 BANGKALAN 30 94, S ,0 100,0 100,0 SAMPANG , , a 8 9 SUMENEP l 01 I_ i , IWO 100:0 101:3 TOTAL i (b

24 VILLAGE INFRASTRUCTURE PROJECT FINAL RePORT FY 1996/1997 No. Infrastructure Constructed Mandays Target and Realization Work Force Progress Kabupaten of Roads B3ridges Jetties Water Santy Mandas Target Man asrealized (Por VIII. (meter) (meotor (meter (unit) (urit) Ordinary ISkilled Leaders Ordinr Skilled ILeader Man Women IYouth- Poor Phys. Fin. Md s... Pandegiang ~ go ,01 100, Lebak J00,0 1L ,0 - Bo12E~ , ,0 - Sukabuml , aaniw ~~~~20 -il , o. o. o. Bandun~~~~ 5 ~ , , Garut Tasikmaay Ciamis 5, ,01 Kunngn , , ,0 Cirebon , Malaingka , Sumedang ~T2 _ ,0,- oo Indramnayu Ss , Suba!g , ,0 Puxwakarta SW Is _ Karawang ,0 100, Serang Ill Cilaca , Ba umas ~ ,0 100, PwbaIiigg~, , Ban*anegara I t i 6, i ,0 99,7 Kobumen _ PLirwaro ~~ AW ,5 Wonosobo _ ,0 Magga tlj,, Bwoyil _ , Klaten , ,0 Sukohario ,4 Wonogrlr _ , ,0 Karanganyr GO a ,9 Sragen (~~obogan , Bloare_ ,0 Rembang _ 10, ,0 Put _ Jepara ~~~ r ,0 Oemak I , 100,010. 9, Semarang 5 11.~~809I

25 VILLAGE INFRASTRUCTURE PROJECT FINAL REPORT FY 1996/1997 No. lnkastucturo Constucted Makndas Targe and Realization Wotk Force Progress Kabupaten of Roads Bridges Jetties Water Sanitary Manday& Tar let I MAndayRgaIIzed (Pros - VIII. (mtf (motor) (unit)~5 I (Jh OrdinvI Skilled Leaders Ordlniv Skilled Leader Men Women Youth Poor Phys. Fin.I Mdys. 1emngung Kendal a0 % ,0 Batang 91,4 I ,0 Pekaongan ,0 PomWang 96, _ ,0 Tagal 104,9 ~ Brebes , KulonProgo W BantuL , Guinungkidul ,0 100, Sleman 100, Pacitan ,0 Ponoroao 08.6 _ i Rtrn%qatek 100, ll , ,, , Tulungagung 100, ,0 Blitar , , ,0100,0 Kedid 111, , _ , ,0 1-I0 29i i , , _ 0 3, ,0 100,0 100, _ I ,0 10-0,0 105.~7 Jember , ,146 1, ,0 100,0 100.n ~Uwaft~l 5~ al a a i Sondowoso 10o J i , ,0 100,0f92.6 Sltubondo J , , _ 0 33Y ,0 100.J Procllng is F ,783 10, Pasuruan _ , , ,0 00, 0 10l,ri Mqj~~~~to ~~~~~~~ ,0.Jomba _ a ,0-100,0 100,0o Nanjuk ,58 1, ,0100 Mad5un 100, Maget~i ,0 100, , 8, ' f l ,0 100,0 88,2' INgawj 10 27,21016s ,0 100,0 99,8 ' qoionagoro r , 0. 0, Tuban ~~~~ Lamongan ~~ !00,0 ( _ Gkeslk l so Bangkatan ,0 1000~ I Sampang a Pamulkasan o , oo er %TAL , Lm90. TOTAL d ~ I & ,6201 1Ai

26 Village Infrastructure Project Table 6: Key Indicators for Project Operation Summary of Surveys of Local Government Support Aspect Rated Rating Number of Respondents Percent Supporter Support of VIP Neutral l Does not Support Assumption about success VIP will succeed Neutral VIP will not succeed Participate Leadership Style Neutral Decides himself Source: A survey undertaken in Table 7: Studies Included in Project Study Purpose as defined at Status Impact of Study appraisal/redefined 1. Review of non- Identify potential cost Completed Good, after project infrastructure constrains to reduction measures implementation priod trade and transport in rural areas 2. Monitoring of the VIP Evaluate VIP impact Completed Will allow longitudinal analyses Data Processing and later Analysis 3. Project Manual General implementation Initial draft guidelines completed within six weeks 4. Review of Altemative Approaches to Building Completed Negligible Road Construction Low Traffic Road Study Methods 5. VIP on Sumatra To prepare VIP2 Completed Used for VIP2 6. Kecamatan Pilot subproject in 6 Completed Basis for a follow on project Development Pilot Kecamatans to serve as an started in FY98 input for proposed KDP 7. Local Institutional Input for KDP Id Id Study

27 Implementation Completion Report Table 8A: Project Costs (US$M).l Appraisal Estimate Actual Local Foreign Total Total Goods, Works and Services TA Implementation Support Increased Administration Subtotal Contingencies (exchange rate loss) Total Project Costs Table 8B: Project Costs (Rp billion) Appraisal Estimate Actual Local Foreign 0 Total J Total Goods, Works and Services TA hnplementation Support hicreased Administration Subtotal Contingencies (exchange rate loss) Total Project Costs Table 8C: Project Financing (US$M) Appraisal Estimate Actual Local! Foreign Total Total GOI Bank k Total Project Costs $6.85 million were cancelled.

28 Village Infrastructure Project Table 9: Status of Legal Covenants Loan/Section! Covenant Present Original Revised {type Status fulfillment fulfilime date nt Description of covenant Comments l ~~~~~~~~~~~Datel Sch. 5, Part A, 10 C 12/31/95 Borrower to select Project villages not Complied Para. 1(a) later than 12/31/95 for the 2nd and 3rd with. Project years. Bupatis and Camats to l provide public information for. j subprojects to villages. Sch 5, Part A, 10 C Selection and information of Complied Para. 2(a) subprojects to be carried out in with. accordance with criteria listed in l subparas. (i) - (iv). Sch. 5, Part A, 10 C Field engineers shall complete Complied Para. 2(b) technical reviews of subprojercts prior with. to approval or replacement by an altemative subproject, in accordance with guidelines (i) - (viii). Sch. 5, Part A, 9 Quarterly progress reports and an Complied Para. 4 annual report by March 31, of each with. Project year Sch. 5, Part A, 9 C 01/01/96 Prepare and furnish to the Bank a Complied Para. 5(b) report integrating the results of its with. l monitoring and evaluation activities. Sch. 5, part A, 9, 10 C 02/01/96 Review monitoring and evaluation Complied para. 5 (c) report with the Bank and take with. measures to ensure efficient completion of the project. Sch. 5, Part A, 1, 9 C 10/30/96 02/18/97 Carry out annual technical audits and Complied Para. 6 furnish them to the Bank. with late due to work overload. } Sch. 5, Part B 5. 6 I C Project agencies and institutions Complied responsible for Project implementation with. adequately staffed and Project Manual made available and updated as necessary in consultation with the Bank. Sch. 5, Part C 7 C Compensation for affected persons to Complied be avoided or minimized, and where with. not possible, ensure that they are compensated in accordance with the compensation Guidelines agreed -between GOI and the Bank. Compliance Covenantype C = CovenaTnt complied with I = Account/Audits 8 = Indigenous people CD = Complied with after delay 2 = Financial performance/revenue 9 = Monitoring, review, and reporting CP = Complied with partially generation from beneficiaries 10 = Project implementation not covered by categories 1-9 NC = Not complied with 3 = Flow and utilization of project 11 = Sectoral or cross-sectoral budgetary or other resource funds allocation 4 = Counterpart funding 12 = Sectoral or cross-sectoral policy/regulatory/institutional 5 = Management aspects of project or action executing agency 13 = other 6 = Environmental covenants 7 = Involuntary resettlement

29 Implementation Completion Report Table 10: Compliance with Operational Manual Statements Statement number and title Description and comment on lack of compliance l No significant lack of compliance with applicable Operation Manual Statements OD or OP/BP Stage of project cycle Table 11: Bank Resources: Staff Inputs Actual Weeks US$ '000l Pre-appraisal Appraisal Appraisal-Effectiveness Supervision Žl Completion Ž Total ' Including travel expenses. ' FY '000$ to 2/

30 Village Infrastructure Project Table 12: Bank Resources: Missions Stage of Month/ No. SW Specialty Performance Types of problems project Year 7 of In ratings cycle persons field IS DO Preparation 03/ NA E _ 1 mission 10/14/94 2 NA E l Appraisal 01/20/95 3 NA E, HE, A through board approval Board through 06/20/ E effectiveness Supervision I 09/29/ E, A, HE HS HS Strengthening of engineering team through hiring another EMC senior engineer per province & defining responsibilities for dealing with problems. Addressing the desire for road asphalting by villagers; updating the draft manual by end of Nov. 1995; and increased transparency to allow villagers to know the flows of expenditures. Supervision II- 02/09/ E, A, HE HS HS Main concems: quality of civil works, mid-term in particularoads; Admin. LKMD withdrew standard monthly installments from BRI; in one village, increased benefits could have been achieved with a subprojectype not specified in the Manual (a flood gate); VIP related operational budgets were made available late, and not clearly allocated to the agencies involved. Supervision 07/20/ E, A, HE HS HS In few kabupatens Pimpro's formal III appointment had not been completed; delays in release of the advance to some villages; some concern about the quality of the civil works. Supervision IV 11/15/ E, A, HS HS The project is ahead of schedule, ENV except for a slightly delayed audit by BPKP. Issues: a) quality of road works a problem in some villages, improvement is not easy as often soil conditions are the cause; b) the foreseen balance of $3 million may be used for TA start-up under VIP II if agreed. E = Principal Economist/Task ENV = Senior Environmental Specialist EG = Engineer Manager HE = Senior Highway Engineer A = Senior Anthropologist IS = Institutions expert SW- Staff weeks DO - Development objectives IS - Implementation progress 7 The mission end date.

31 Implementation Completion Report Stage of Month/ No. SW Specialty Performance T7ypes of problems Project years of In ratings Cycle persons field IS DO Supervision V 03/20/ E, A, EG, HS HS None. All works were completed, 7 IS within original cost estimates and ahead of schedule. A balance of over $3 million of "unallocated" funds is to be used for pteparation of follow-on pilot Kecamatan project and a related study. Supervision VI 1' 10/ E.A. S* S* KDP pilot: problems of definition, with consultants; good progress with local institutions study. Supervision VII 01/20/ E, A, EG HS HS None. The original project was 8 completed ahead of schedule. The added pilot and study are under implementation. Supervision VIII 3/ E HS HS MoF overdue in withdrawing from 1/ SA. Project villages to be revisited; KDP experiencing implementation difficulties; local inst. study draft interim reports available (VIP2 proceeding) Supervision IX 8/ E S* S* Funds were invested, but the poorer were bypassed. Supervision X 9/ HS HS Start of ICR Supervision XI 2/ E HS HS Final discussion of ICR with GOI 1/ Some 5 field visits for the pilot and the local institutions study were undertaken by team members and independent parties residing in Jakarta (sucb as GTZ, anthropologist). * Rating applies to pilot only. 8 The mission end date.

32 Village Infrastructure Project ATTACHMENT 1 Sample Economic Returns to Village Infrastructure Roads and bridges. About 76% of the village grant amounts was used for road upgrading or construction, and 11% for new bridges. The road lengths improved varied generally from some 1.5 km to 5 km, but occasionally were longer. With 87% allocated to road upgrading, this annex peruses the transportation benefits mentioned in a sample of village completion/benefit reports. A majority of the road upgrading projects allowed motorized vehicles to replace porters. Transport cost savings per ton range from Rp.35,000 to Rp.150,000, between 20% and 75% ofthepreproject/porter cost. Some projects allowed the use of larger vehicles, for instance trucks instead of cars, and others improved the road surfaces; these projects had some 20% cost savings--less than Rp.5,000 per ton, i.e. much lower in absolute unit terms compared to the reductions noted above, but probably such roads had larger traffic volumes (see overleaf). Still, some others and especially bridge projects allowed access to new markets, changing routes altogether, with higher benefits. Benefits from personal transport have not been reported, but these are significant as well, as minibus service substitutes motorcycle "taxis" or even overnight stays outside the village, at a fraction of the cost, for access to schools, markets, health centers etc. The larger the population in the village, and thus the village production and personal transport, the larger the benefits of the road improvement. The products commonly exported are rice, coconuts, coffee, fruits, fish; the transport cost reduction increases farmn gate prices considerably. However, all transport volumes are not known to us. A simulation, assuming a $1 per day income, that just IO% is spent on transport without the project and that the above noted saving rates apply with the project, and assuming that the road will last some 6 years and including in the initial cost the engineering overheads, provides rates of return up to 200%, with a sample median of about 50% ERR (see overleaf). For road works to have a minimum 10% rate of return, the threshold traffic is only some 700 trips per year of a 7-ton truck or equivalent (some 5000 tons, or 10,000 tons both ways) during 5 years, assuming an average saving of Rp.7,000 per ton. Alternatively, the road cost recovery may be achieved in less than 3 years; just one 7-ton truck replacing porters with an average savings of Rp.100,000 per ton, and making 250 trips per year, would in only one year ensure cost recovery; or, if just five 7-ton trucks or equivalent could use an improved road with an average saving of Rp.5,000 per ton, and also made 250 trips per year each, cost recovery would take 3 years. Future road maintenance will be essential for the roads to last, and villagers have agreed to undertake the maintenance; but even if roads were to last less, the investment would still be justified. Unemployed porters are reported to find alternative jobs without much difficulty. To be conservative, only a 20% ERR is reported in the main text, but there is no doubt that the upgrading of village roads has favorable returns compared to higher standard and higher traffic roads. Drinking water and sanitation. Drinking water and sanitation were allocated only 13% of the village grant amounts. From VIP supervision missions it is known that water supply improvements also have rates of return exceeding 20%, even when water is purchased in bulk from a water enterprise. When needed, water user groups are formed to ensure payments or monitor water use; field engineers have to ensure a group is formed for subsequent maintenance of the facilities. Women are the main beneficiaries from water supply installations. Benefits from sanitation facilities are less clear, partly because villagers may require further awareness campaigns and training in maintenance of the installations, but this will be addressed under the project.

33 Implementation Completion Report Average No. of Beneficiaries for VIP 96/97 Total per Kabupaten in the Project Provinces ATTACHMENT 2 Activities West Java Central Java Yogyakarta East Java Total No. of Villages Reported Total No of villages No. of Roads ,177 Average No. of 2,366 1,641 1,224 1,761 1,823 Beneficiaries/Kabupaten No. of Bridges Average No. of 1,718 1, , BeneficiarieslKabupaten I No. of Jetties.- I 1 Average No. of - - 2,980 2,980 Beneficiaries/Kabupaten No. of Water Supply Average No. of Beneficiaries/Kabupaten No. of Sanitation 53 III Average No. of Beneficiaries/Kabupaten Location of Beneficiaries Outside 149, ,437 1, , ,842 In Village 458, ,440 42, ,175 1,690,194 Total 607, ,877 44, ,842 2,162,036 Source: Based on a survey of second year villages. Reduction in Transport Prices to5 U ma ~ ~ ~ ~ ~ ~~~PretRduto nfi

34 Village Infrastructure Project PART HI BORROWER CONTRIBUTION TO THE ICR Introduction Completion Report Village Infrastructure Project on Java Rudimentary infrastructure at the village level, roads, bridges, water supply, jetties, and sanitary facilities is often lacking on Java, home to over 60% of the population of Indonesia and among the most densely populated places in the world. The Central Bureau of Statistics in 1993 identified over 6,000 villages on Java that had been "left behind" by development, using a definition that was based in part on deficiencies of basic infrastructure because such deficiencies are a clear cause of poverty. With poor access to markets, for example, crops might rot in the fields because they could not be transported cheaply or quickly. Poor roads often limited not only the possible profit from crops but the choice of crops to grow. The lack of roads and bridges also limited access to health facilities found usually at the subdistrict and district levels, which might take hours to reach even when they are only a few kilometers away. The level of health in many villages is also low due to lack of reliable water supplies and unsanitary disposal of waste, causing economic loss as well as suffering. It is often the task of already overworked women to fetch water for family needs, sometimes making several long and tiring trips each day. This waste of human resources could be prevented by providing water nearer to homes. Most of the inhabitants of Java live in rural areas, as farmers or farm laborers working small plots of land in some 18,000 villages, growing rice, corn, cassava, and a large variety of other crops in conditions that are usually quite favorable to agriculture. Many "left behind" villages, though, are in areas that are much less productive, especially in, the uplands and some coastal areas. Infrastructure in the villages had been addressed by a scattering of government programs but even more commonly by self-help activities of the village communities, who were quite familiar with organizing themselves to meet limited village needs, such as road cleaning or building mosques. Constructing high quality infrastructure was always considered to be beyond their means, as it would mean sacrificing their time to spend months laboring on construction as well as purchasing construction materials that cannot be found in the village. The villages had very limited funds to spend, and most farmers could not afford to abandon their daily search for income. In the past, the government's approach to the identification and construction of infrastructure was very top-down through either sectoral programs or block grants of the Inpres program, implemented through provincial or district levels of govemment. The works were carried out normally via contractors, selected through the normal bureaucratic bidding procedures. In several ways this situation was unsatisfactory: The provision of infrastructure in the villages was often ignored or given a low priority, as typical methods of project selection favor higher level roads in the network or large water supply systems. The system did not target the poor areas that required infrastructure, because the selection criteria favor more rapidly developing areas. The use of contractors tended to lead to low quality works since many contractors were more concerned with profit than quality, and top-down decision making did not give the communities a sense of ownership, which in turn caused problems in maintenance. This system did not lead to improvements in the communities' capability of developing itself, since the village tended to wait for further top-down solutions to its remaining needs.

35 Implementation Completion Report The gap between developed areas and left-behind areas continued to expand, compounding existing problems of migration to the cities, a decrease in the quality of life in the villages, and social unrest from the growing disparity in quality of life between urban and rural areas. The Inpres Desa Tertinggal program was created to improve the conditions in the poorest villages in Indonesia. Under this program, several community groups in each poor villages were given approximately $30,000 per village over three years as capital for group economic activities. The groups of poor persons determined for themselves the activities they will invest the money in, and the funds in turn were revolved to other groups. Lack of infrastructure was seen as one major constraint to developing these poor villages. Indeed, in some of the villages the investment of money in economic activities cannot succeed because of the restraints caused by poor infrastructure. Therefore, the government decided that additional grants should be allocated to these villages to provide infrastructure. The Village Infrastructure Project was a result of this decision. It was designed in late 1994 and implementation began in April Project Design and Implementation The Village Infrastructure Project took an innovative approach to solving the problem of inadequate infrastructure in the villages. Many of the standard regulations for project implementation were waived in order to make the activities attainable by villagers, following principles of simplification and practicality. Both the World Bank and the Project Secretariat actively shielded the project from pressures to conform to established practices that would interfere with VIP implementation. The key aspects of the Village Infrastructure Project included: 1. Village selection was made by district governments from long lists based on Central Bureau of Statistics surveys using criteria developed jointly. 2. Selection of infrastructure was the responsibility of the villages. 3. Villages were assisted in the design and calculations by consultant engineers recruited from the private sector, one field engineer for every five villages. 4. The basis for construction was a very simple grant agreement signed by the village and the representative from district Public Works. 5. A fixed amount of funding was channeled directly to each village (Rp 120 million, or about $50,000) using a simple and responsive funding process. 6. The villages alone were responsible for funds, assisted by the field engineer. 7. Senior local Engineering Management Consultants provided advice and guidance to the more junior Field Engineers, and as they were from another firm they also provided a degree of objective monitoring. 8. Construction was done by villagers themselves for an incentive wage, using local materials wherever possible so that more money stayed in the village. 9. Information boards were located in strategic places so that anyone in the village had access to key data, with community-wide meetings to decide salient issues.

36 Village Infrastructure Project Project Results and Impact The government gained valuable experience in bottom-up processes and in simplified delivery of infrastructure. This experience led to changes in the design of existing and proposed village development activities. The World Bank provided a second loan to expand the VIP to the island of Sumatra in 1997/1998 and 1998/1999, with a target of 2600 villages in two years. In addition, in 1997/1998 the government funded 2000 additional villages from its own funds using the same methodology. Database of Subproject Information A database of subproject information was collected during the two years of implementation, based on subproject designs and budgets. Data for the 1996/1997 subprojects were collected from 731 of the 815 villages. In each village there was an average of 2.8 subprojects reported. Some interesting conclusions include: * 26.4% of the reported budget was used for labor payments, with a higher portion of the budget going to materials (68.3%). The portion of the budget going to materials often included the procurement of materials labor intensively by the villagers, for which workers were paid based on the daily wage. Roads were the most labor intensive subproject type. Database of Prior Conditions Simple data were recorded by the field engineers early in the construction season about conditions in 770 of the 815 villages (94.5%) active in the second year. The most interesting conclusions were as follows: * Nearly all VIP roads were all-weather roads. Of almost 1100 roads reported, only 10 did not function immediately. Most roads were built over existing earth roads, but a significant proportion (32.4%) were built over inadequate existing hard-surfaced roads. The largest number of roads connected to an asphalt-surfaced road (42.1%) or to a good rock-surfaced road (16.3%). Nearly all bridges were steel girder, concrete, or wood-beamed bridges. The largest number of bridges connected hamlets (dusun) within the village (53.9%), but a significant number connected two villages (34.4%). Most of the bridges replaced existing bridges (74.1%), but these were most often damaged or constructed simply of bamboo or wood. Nearly all water supply projects were used as a source for drinking, washing, and bathing. In fewer than 10% of the water supply sites had the water been checked at a lab by the time of filling out the questionnaire. The most common source of water before the project was from local wells that were limited in supply. The largest number of water supply projects drew their water from springs (55.9%), followed by ground water (31.5%). Only 5.5% used surface water sources. Some 79.6% of sanitary facilities used septic tanks with leaching fields. Most water for the latrines came from wells (55.0%), and most of the remainder came from springs (38.9%). The majority of people had used the river or fields (57.9%) prior to building latrines. The remainder used inadequate or insufficient latrines. Less than 20% of the Operations and Maintenance groups for water supply or sanitation projects had been formed at the time of filling out the questionnaires.

37 Implementation Completion Report Community Participation To demonstrate the level of participation in decision making, consultant field engineers reported about Participation of the LKMD in Preparations Percentage of Villages Where LKMD was Highly Involved ' 09,~'' '5 e Preparatory Activities the level of participation in project preparations. The resulting graph is shown above.there was usually a sharing of tasks between the field engineers and the villages, but for a number of activities the engineer did nearly all the work: calculating budgets, technical drawings, verification. For other tasks, the community performed the tasks without almost no assistance from the engineer: labor recruitmnent, proposing infrastructure, identifying sources of materials, solving land acquisition issues. Participation during construction can be seen in the following table: Category of Worker 95/96 96/97 Persons % Persons % Men 102, , Women 10, , Youth 34, , Poor 73, , All Workers 112, , Workers/Viliage Person Days 5,543,888 9,452,164 Person Days/Worker

38 Village Infrastructure Project There was a higher proportion of women, youth, and the poor implementing the project in 1996/1997 compared to 1995/1996. At the same time, there was a reduction in the average number of days worked by each of the workers, only part of which was explained by the increased participation of women, who typically worked for shorter periods. The key change was in number of person-days of labor per village, which fell from 13,360 to 11,600. Changes in Commodity Transport Prices Changing transport prices was one of the earliest occurring indicators of benefits from an improved road. The prices paid by farners for transporting their commodities to market was monitored by the field engineers. At the start of construction a point was selected near the end of the proposed road, and the engineer asked farmers in the vicinity about the types of commodities sold, their destination, the transport method, and the unit price for transporting them by available means. This investigation was repeated immediately at the end of construction. The hypothesis was that transport prices would be lower after construction, because of increased access by vehicles and lower operating costs. Data were received from 752 of the 815 villages (92%). In 64 of the villages a comparison could not be made because of incorrect data collection, or because the before and after data could not be compared (different and unconvertible units of measure or different destinations recorded for products), or because there was no road constructed (13 villages). The graph below shows the results for the 688 villages from which valid data were received: Reduction in Commodity Transport Prices VIP 1996/1997 Number of Villages l I 80 li I 40 i _l NI 0 cz o50 w s I- Ci %O lp 9 tp' 4. q4 0 4)\O l eo No~ go o oq 43E Percent Reduction The majority of villages reported a reduction in transport prices of at least 20%. The average percentage reduction was 30,3%. Only three villages reported an increase in transport prices after construction. About 13% reported no change in transport prices. Impact Data Impact data were collected in February 1998, targeting all clusters constructed during the village infrastructure project, but only 172 clusters were visited. This represented 69.9% of the 246 clusters in the Village Infrastructure Project (59% of the first year villages and 75% of the second year villages).

39 Implementation Completion Report Impact data were collected by the field engineer in four categories: Increases in usage or economic activities due to the construction of the subproject. This included increased transportation, more production because of easier selling, new products, and the like. Savings due to the subproject, such as reduced outlays for water, reduced transport costs, or reduced cost of inputs. These were quantified using coarse estimates that reflected the true benefit according to the enumerator. Instances of applying skills learned in the VIP to other activities, such as undertaking additional infrastructure construction or applying skills on another project or private undertaking. Other impacts that could be attributed to the project, most of which could not be easily quantified. The results of the data collection and preliminary analysis were as follows: * 155 of the 172 villages in the sample reported increased economic activity (90.1%), and on the average they generated a reported 30,928,306 rupiah each. This average was distorted by some very large numbers: more than 10 villages reported more than 100,000,000 rupiah in increased activities. The median increase was slightly less than 10,000,000rupiah. The table on the next page shows the distribution of the increases. Also in 155 of the 172 villages were there reported savings. Savings on the average were slightly less (Rp 20,195,923) than the increase in activities, because there were fewer extremely large values in the list. The median, however, was higher at Rp 13,005,000. Only 7 of the 172 villages did not report either of the two quantifiable impact measures above. A summary of these two measures can also be seen in the table on the next page. 73 of the 172 villages (42.4%) reported applying their new skills on another activity, with 81 projects cited by the enumerators. The average value of all the projects cited was slightly more than Rp 20 million. The list shows both public and private activities undertaken. Perhaps the most interesting data were found in the miscellaneous types of benefits that were reported on the data collection form. On the second page following is a summary of the kinds of benefits attributed to the project. The most common benefit was in the ability of a portion of the community to send their children to a specific school or level of schooling, which previously had been beyond their means, probably because of inadequate transport. In 101 villages there was a quantified increase in numbers of students, showing on the average more than a 109% increase in attendance, with 66 of the villages (65.3%) showing more than a 100% increase. This does not include another 16 villages where school going was reported as easier after the project, but without an increase reported in attendance. The second most common benefit was electrification, as 41.9% of the villages reported receiving electricity since the project was completed. Access to the villages normally is a prerequisite for public electrification, as power lines and poles have to be transported into the village. Health benefits were cited in many villages, and four main kinds of health benefits were cited: O E El Improved water supply Improved sanitation New health facilities or services

40 Village Infrastructure Project O Better access to health care because of improved transportation Improved transport was often cited as a benefit, as was increased motorcycle ownership. There were changes mentioned in the way that village products were marketed. Some farmers preferred to bring their own produce to market (now easier to do) while others preferred to be able to have buyers come into the village. Another item cited often was the improved availability of building materials in the village, which led to improved housing. Among the items cited by just a single enumerator were two cases of negative influences of the project: bad elements coming into the village and an increase in disturbances. Follow-up The villages have primary responsibility for maintenance of the completed infrastructure. They received training in how to determine what needs to be done and in how to do it. They were required to devise a maintenance plan, which should include aspects of raising funds that are required to purchase materials for repairs. It is the normal practice in many Indonesian villages to undertake repairs throughgotong royong (selfhelp) sessions, wherein the community as a whole joins to clean out ditches, repairs damaged surfaces, and the like. An impact analysis will be conducted to determine the quality of the infrastructure after several years have elapsed, to what extent the subprojects are beneficial, and to what level they are maintained. It will be attempted to compare the development of villages that received assistance with similar villages that did not receive assistance with infrastructure, even though it will be difficult to isolate the effects of this one project from all the other programs that have occurred over the intervening years. Evaluation of the Performance of the Borrower At the beginning of the project, many observers both within the government and outside of it doubted the ability of the government to implement this kind of project, one that intentionally limits the government's role and opportunities for intervention. There were also doubts about the ability of villagers to implement the project and the ability of the government to recruit so many dedicated engineers from the private sector. The cooperation among the central government agencies was outstanding, as they presented a unified and consistent position that this project and its approach must be implemented and it must succeed. Temporarily the agencies could put aside their sectoral egos for a common goal. The central government also approved a series of simplifications to standard project rules that greatly assisted the VIP in working in the village, such as paying the grants directly to the village, freeing the project from difficult procurement and contracting rules and local regulations, freeing the project from tax considerations, and freeing the project from revenue stamps. The central government and particularly the project secretariat were also very responsive in dealing with problems that occurred and were brought to its attention. This can be noted by observing the differences between implementation in the first year and the second year. Written project guidelines also changed a great deal. The success at the local level was not uniform. While the majority of kabupaten offices gave their support to the project and to its unusual mechanisms, there were a few other kabupatens that resisted the new approach and attempted to intervene for other purposes. In these districts progress was slower and less in line with VIP principles, although they too had to bend to the project regulations. Provincial offices varied as well, although rarely by actively opposing the project principles - more by being somewhat unexcited and slow to react when help was requested.

41 Implementation Completion Report Lessons Learned While the Village Infrastructure Project was not a pilot project per se, it shared with pilot projects the goal of trying innovative approaches that might be applicable to other projects. In this regard the project was highly successful and taught many interesting and applicable lessons. Some of the most applicable are given below: 1. Villages and their organizations were capable of undertaking the construction of infrastructure in their villages, managing the works and administration. Works were performed labor intensively with equipment usage normally limited to compaction of the road surface by heavy roller. All 1230 villages completed their subprojects. Thousands of kilometers of roads, hundreds of bridges, and thousands of small water supply and sanitation projects were constructed, at an average cost much lower than if contracted in the usual way. 2. Written guidelines were very important in ensuring the adherence to project procedures and standards. Technical standards were very flexible, more for guidance than for standardizing. 3. Technical assistance was necessary to balance the relationship between the village and the project manager from the bureaucracy. It was possible to employ a large number of private sector engineering consultants to support the VIP (163 field engineers), and their role was the most important on the project. The field engineers worked hard in difficult conditions, sometimes with less support than they would have liked, and the large majority exceeded all but the most optimistic expectations. The transfer of technical and management skills to the village was very important. 4. Except in those few areas where water supply had been a perennial problem, villages responded to the opportunity to build infrastructure of their choosing by choosing foremost to build roads and bridges to improve access. Although the project used an open menu for economic infrastructure, the majority of villages, officials, and consultants considered the menu limited to roads, bridges, water supply, sanitation, and jetties. 5. Transparency was the key to village implementation. Large information boards and communitywide meetings were the main ways to promote it. Lack of transparency contributed to several cases of malfeasance, top-down planning, difficulties in attracting labor, and asset acquisition problems. 6. More attention to the quality of works should have been made from the very beginning, since long term benefits are absent if the infrastructure lasts only a year or two. The project developed a thirteen point quality program after the first year and the quality level improved, but even then quality of works was frequently an issue. 7. The VIP with its innovative funding mechanism suffered very little leakage of funds, with most of that leakage the product of village chief domination or unscrupulous local officials. Very few consultant engineers were implicated in malfeasance cases, and these were dealt with quickly by the Secretariat by enforcing a strict Code of Ethics. 8. The use of a standard grant amount (Rpl20 million) was very helpful, particularly in the beginning, as it promoted transparency and simplified administration, particularly funding flow from the government to the village. While it was obvious that villages needs were not uniform and that this amount could not solve most villages' needs for improved infrastructure, the grant was sufficiently large to enable the construction of useful infrastructure within the given time frame of one dry season. 9. The project emphasized its ability to learn by doing, and to achieve this the engineers were given considerable time to discuss problems and recommendations, including an orientation workshop

42 Village Infrastructure Project and an evaluation workshop each year, as well as structured meetings once or twice a month. These opportunities, plus the support given to communications in the budgets and the willingness of the Secretariat to alter its procedures and guidelines, led to the feeling of joint ownership of the project between the government and the private sector engineers. It also led to the continual improvement in guidelines and instructions. 10. Maintenance was a weak point of the project. While training was given to villages in maintaining their infrastructure, inertia and the lack of outside funding often led to inadequate maintenance. Over half of the first year's roads exhibited maintenance deficiencies after one year. Consultants were sent back to previous year's sites to promote better maintenance, but lack of funding for materials and equipment took its toll. 11. It proved difficult to get local government or consultant engineers to report on problems. Government officials were reluctant to reflect badly on their districts, but the engineers reasons were more complex. In part they felt that problems should be solvable without involving higher levels, and local problems need not be reported. They, too, were reluctant to report problems that might reflect badly on themselves or their company. But in many instances the problem was that the engineer did not recognize the problem as a problem, necessitating frequent visits by more senior engineers. 12. The project engaged an NGO to conduct independent monitoring of the VIP during the two years of implementation. Only one recommendation of the NGO was implemented, to have the engineers monitor community participation more closely. All other recommendations and conclusions were either obvious or inappropriate. The project could have benefited from outside monitoring, but it did not. 13. It has been concluded by some that the project methods are not sustainable because they required intensive or even heroic support from the government, the donor, and the consultants. A condemnation of others cannot be considered a weakness in the VIP. Evaluation of the Performance of the World Bank The World Bank supplied consistent support for the project, urging the VIP Secretariat and consultants to increase their efforts to ensure that the project was successful and innovative. The World Bank officials worked closely with the members of the Secretariat and the consultants to improve project mechanisms. On many things the World Bank task manager and the Secretariat were able to achieve results much more quickly than had been the norm with other projects. The World Bank was also extraordinarily quick in issuing approvals required for quick implementation. Supervision missions, although sometimes painful, provided critical feedback. They were painful because they pointed out where things were not going according to plan. One such piece of feedback was directed at the quality of implementation. It resulted in a plan to improve subproject quality in the field, but it was still painful to hear. One area where there might be some improvement would be with the relationship to the Resident Staff in Indonesia. As the task manager was based in Washington, the project on occasion had no permanent and well informed senior local official involved. Consequently there was some occasional confusion, wherein contradictory statements would be made. It should be noted as well that more visits to the field were made by senior Washington staff than by senior RSI staff on their own. Comments on the World Bank Implementation Completion Report The World Bank Implementation Completion Report is quite clear and contains nothing that is surprising as these are issues that have been frequently discussed with the project secretariat. We would, however, make a few clarifications, as follows:

43 Implementation Completion Report 1. [Paragraph 3] The mobilization of village contributions is something that was never emphasized in the project. Asset acquisition on Java was rarely an issue, as the communities are familiar with community participation in the form of labor or contributed land. It therefore was never stressed, although contributions occurred normally. In the Second VIP it was attempted to quantify contributions to be able to answer criticisms that the communities are getting something for nothing and therefore will not value it. 2. [Paragraph 6] We found most of the studies connected with the VIP to have little benefit to the project. The study of conditions in Sumatra was of very limited usefulness; much more was learned by sending people familiar with the VIP for just a few days to potential sites. The study of alternative approaches to building roads did not yield a useful product. In part this was the fault of the Secretariat, which at that time could not afford to give much direct assistance to the expert, but the final product was not as envisioned in the terms of reference. 3. [Paragraph 25] While maintenance is important, sustainability is not entirely reflected in maintenance. The project has in its training sessions and orientations stressed that empowerment is the number one goal of the project, which should be measured by how much improvement there was in village capabilities and in how they approach future development activities. 4. [Paragraph 52] With the advantage of hindsight, the preparation study for Sumatra and VIP2 was inadequate. 5. [Paragraph 47] One major handicap about involving auditing agencies in the project was that the agencies themselves were mostly not interested in changing their methods and not interested in learning about the VIP approach. Some continued to be difficult even after an approach was made to explain the differences. 6. [Paragraphs 8 to 7] More should have been said about village selection, as this had such an influence on project results. Because of the speed with which the project was started, the villages in the first year were selected very much top-down. Dissatisfactions were aired, and in the second year the villages were selected in a more bottom up fashion, utilizing consultants, but this too was considered unsatisfactory. No matter how the villages were selected it would be easy to find some more appropriate villages that were not selected and more inappropriate ones that were selected, and finding such sites led people to be suspicious of the process. Examples of poor selection efficiency were probably nothing more than the inefficiency of any large complicated system. Any way to select villages would be subject to similar claims of unfairness.

44 Village Infrastructure Project SUMMARY OF OTHER TYPES OF BENEFITS CITED BY ENUMERATORS Jumlah desa % Desa Jenis Manfaat No. of Villages % Villages Type of Benefit Peningkatan jumlah anak sekolah / Increase in number of school children Listrik masuk desa / Electricity entered the village Peningkatan sanitasi melalui MCK / Improved sanitation Transpor lebih lancar / Improved transportation Sekolah dipermudah / Easier to go to school Bahan bangunan masuk / Building materials enter the village Peningkatan fasilitas kesehatan / Improved health facilities or access to them Kesehatan meningkat karena sumber air / Improved health from healthy water source Pengaspalan / Asphalting of road Perubahan dalam cara menjual hasil / Change in method of selling products Pembukaan isolasi desa / End of village isolation Pemilikan motor meningkat / Increase in motorcycle ownership Konstruksi rumah baru / Construction of new houses Kesempatan kerja ojek I Job opportunities for motorcycle taxis Mobil tangki bisa masuk / Tanker trucks can enter the village Dana dapat dikumpulkan untuk PDAM / Funds collected for city water Pengaruh negatif orang masuk / Negative effects of strangfers Warung tambah / Increase in shops Penyebrangan sungai berbahaya / Remove danger from river crossing Pengajian lebih lancar / Ease of religious meetings Terdapat gangguan keamanan IDisturbances Total sample 172 villages Increase in Numbers of School Children: Before 3154 After 6597 An increase of 109% in numbers of students cited (for particular schools or levels only) 66 of the 101 villages reported more than 100% increase.

45 Implementation Completion Report Summary of Reported Economic Benefits Sample Size 1995/ /1997 West Java Central Java Yogyakarta EastJava Total Summary of Quantifiable Economic Benefits Number of Villages Amount of Reported Benefits Increase Economic Combined Activities Savings Benefits I None Reported Economic Benifits Summary VIP 1995/1996 & 1996/ = 1;35-0, co Z Range of Benefits 2Economic Savings Ei Increased Activities

46 IBRD 26797R Sum era~~~~b.sua 100 1Cr U 114? 114?~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Lebak 0 urk,o0 AW ra Searnan. GotO - Siimedong t,4oiolengko ' Pemnolng S.k.b..---i i, S,eb.., Tegal 0:/ w,loa 00 i Kun.~g00 0 Baaegr 0Srnrn 00Bingr 0uoby dsaloaal \~~~~~~~ 0 00nob 0 Skat ja 00, ~~Y~~vgya~K SuialwProbJombggn ~~~~~Pagmnyarngle filr MIng Lni A>~ ~-0--~ 0 v ' 100'~ ~~ ~~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~dri0pb VILLAGIE IN FRASTUCTfUREPO'C OPEINRPR JC) - N,u 110~~~~~~~- Flores Sa 14 c 2~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~9 Indi on a12o ~,oestimr ei foosngar ntrnaioalboudaie NiJSA TiMU~~~~~~~~~ TENGOAkA /0 Thnor Poria Mooro -~~~~~100 Sumbo l.-~~~~~~~~~~~~~~~~~~~ T~~~~~~p~~~~~s~~~~A Sowohiomlo ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ARL 9