Paper No. 39

March 2002

Indonesia ’s Kecamatan Development Project

Is It Replicable?

Design Considerations in

Community Driven Development

Judith Edstrom

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Social Development PapersEnvironmentally and Socially Sustainable Development Network

Paper No. 39March 2002

Indonesia’s Kecamatan Development ProjectIs It Replicable?

Design Considerations in Community Driven Development

Judith Edstrom

This publication was developed and produced by theWorld Bank’s Social Development staff, comprisingWorld Bank staff who work on social issues. TheEnvironment, Rural Development, and SocialDevelopment staff are part of the Environmentally andSocially Sustainable Development (ESSD) Network.

Papers in the Social Development series are not formalpublications of the World Bank. They are publishedinformally and circulated to encourage discussion andcomment within the development community. Thefindings, interpretations, judgments, and conclusionsexpressed in this paper are those of the author andshould not be attributed to the World Bank, to its affiliated organizations, or to members of the Board ofExecutive Directors of the governments they represent.

Editing, design, layout, and production by Dorst MediaWorks.

Copies of this paper are available from:Social DevelopmentThe World Bank1818 H Street, N.W.Washington, D.C. 20433 U.S.A.

www.worldbank.org/socialdevelopmentEmail: socialdevelopment@worldbank.org

i

Contents

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .i

Glossary and Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ii

Project Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

Bypassing Government to Get Good Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2

Commitment to a Cross-Sectoral Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3

Credit for Economic Activity as a Project Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4

Project Scope: Scale from the Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5

Existence and Influence of NGOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5

Regimes of Land Tenure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6

Human Resources and Remuneration Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7

Locus of Project Responsibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8

Demands on the Bank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9

Annex: KDP Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15

Acknowledgments

This paper is the result of a mission toIndonesia in July 2001 to study the Bank–financed Kecamatan Development Project.

I would like to thank the members of theKDP teams—World Bank and KDP staff whomade the KDP visit such a rich learning experi-ence. This includes Cyprian Fisiy who extendedthe invitation, Scott Guggenheim, Victor Bottiniand other staff of the Jakarta Country Office who made the arrangements, Bhuvan Bhatnagar,Anto Juntak, Susan Wong, as well as LeniDharmawan and Sono Sadoko, who participat-ed in the field visits and shared their valuableinsight.

Many thanks to Scott Guggenheim, VictorBottini, Leni Dharmawan, Susan Wong, DanOwen, and Gillian Brown who gave me valu-able comments on the paper.

I ’d like to especially thank Ralf J. Leiteritzwho joined me on the trip to Indonesia andprovided useful comments and significantresearch assistance, with special reference toland tenure and education data.

Finally, my thanks to Carmen Martinel, whoprovided excellent administrative assistance.

ii

Glossary and Acronyms

Adat Traditional system of land rightsin Indonesia

Bappeda Regional Development PlanningAgency at the province or district level

Bappenas National Development PlanningAgency

BAL Basic Agrarian Law

BPKP Finance and DevelopmentSupervision Agency

BPM Ministry of Home Affairs

CBO Community-Based Organization

CDD Community DrivenDevelopment

GNI Gross National Income

KDP Kecamatan DevelopmentProgram

Kecamatan Sub-district containing 20 villages on average

LKMD Village council

MFI Microenterprise Finance Institutions

Musbangdes Village development meeting

Musbangdes III Meeting to report results ofUDKPII meeting, to organizecommunity implementation team to oversee project imple-mentation, and to decide on the continuation of the villagetechnical assistance

NGO Nongovernmental Organization

NMC National ManagementConsultants, located at the central level

PIU Project Implementation Unit

PjOK Project manager for KDP at thekecamatan level; section head PMD at the kecamatan level

PMD Community DevelopmentAgency at the BPM

Rps Rupiahs (Indonesia currency)

Simpan pinjam Revolving microcredit schemes

TA Technical Assistance

UPKP Sub-district Council of villageheads, includes additional village representatives as theKDP planning and decision-making body

UPKPII Kecamatan-level meeting todecide which village proposalswill be accepted and to electmembers of UPK

UPK Financial management unit usually composed of three peo-ple who handle the kecamatanand village finances for KDP

VIP Village Infrastructure Project

iii

1

Indonesia’s Kecamatan Development Project Is it Replicable?Design Considerations in Community Driven Development

The Kecamatan Development Project(KDP), launched in Indonesia in 1998,has been referred to as an outstanding

model of participatory development, demon-strating community driven development in itsmost “evolved” state.1 It is designed to operateon multiple levels in many sectors, involvingboth national and local institutions to allow villagers to propose virtually any investmentthey would like—from infrastructure provisionto small-scale economic activities.

KDP takes into account both financial flowsand flows of authority, and establishes a robustsystem of checks and balances to encouragetransparency and social inclusion, with particu-lar attention to gender issues. Moreover, it hasachieved a national scope, and now coversapproximately 30 percent of the rural sub-districts (kecamatans) in the country.2 Of 67,500 villages in Indonesia, more than 20,000 haveparticipated in KDP thus far (18 percent), witha total beneficiary population of almost 10 mil-lion people. This amounts to approximately 5percent of the total population and 8 percent ofthe rural population, which has been especiallytargeted because of the greater incidence ofpoverty there. Under the second project (KDPII), it is estimated that the total beneficiary pop-ulation will be approximately 20 to 30 millionpeople with 200 kecamatans added per year.

A project design with these characteristicsnaturally invites inquiry as to its replicability in

other contexts and countries. Are certain pre-existing conditions necessary for its success?What components are essential? Can the KDPapproach be applied in a single sector context,for example, a community-based water project?

There may be no clear-cut answers to thesequestions. However, in assessing its relevancefor other country situations, some key featuresof the KDP project environment need to be considered. Some factors relate to fundamentalissues, such as project objectives, power rela-tionships, and the scale and scope of the pro-gram. Others, such as the human-resource baseand salary levels, may seem more mundane,but are also as important to consider. Becauseevery country context is unique, lessonslearned from the KDP experience would need to be adapted.

The following study seeks to identify thosefeatures of KDP and the Indonesian sociopoliti-cal landscape that contribute to its successfulimplementation, and thereby to contribute to the general knowledge base for those development practitioners who are contem-plating replicating or adapting the KDPdesign elsewhere.

Project Objectives

KDP is a “programmatic project”–better charac-terized as a program than a project. The statedobjectives of the second KDP project are to: “(i)

1 A description of the program is provided in the annex. A loan to finance KDP was approved by the Bank’s Board in 1998, and a second project was approved in 2001.

2 Poverty in Indonesia is largely a rural phenomenon. Hence, KDP is targeted at this constituency. Approximately 60 percent ofIndonesia’s total population live in rural areas, and about 85 percent of the population in KDP areas live in rural areas. Withrespect to the allocation among the provinces, a few small provinces have an especially high density of KDP investments, forexample, Aceh, NTT, Maluku, and South-East Sulawesi. In the larger, and typically wealthier, provinces of Java only 12–20 percent of the populations live in sub-districts where KDP is in place. As for poverty levels, in 1999 the poverty rates wereapproximately 22 percent of the population for non-KDP areas versus 30 percent for KDP areas. See World Bank 2001a.

Indonesia’s Kecamatan Development Project: Is it Replicable?

2

support participatory planning and develop-ment management in villages; (ii) support abroad program of social and economic infra-structure construction in poor villages; and (iii) strengthen local formal and informalinstitutions by making them more inclusive,accountable, and effective at meeting villagers’self-identified development needs.”3 The objec-tive to “raise rural incomes,” included in thefirst KDP project, has been dropped, both dueto the difficulty of measurement and to the primacy of the overarching objective—creatingparticipatory institutions and processes. Theessence of the program, therefore, is establish-ing vital participatory processes, rather thandelivering tangible physical evidence of theinvestment, although the latter remains impor-tant for assessing project success. For example,the extent of participation and transparency is measured in terms of the proportion of thepopulation served, their ability and willingnessto maintain investment, repayment rates onloans, leakage of funds, and availability of information.

The project ’s objective is fundamentally radical: it aims to institute transparency anddemocracy from the bottom up in a countrywhere serious abuse of office and top-downplanning have been endemic. It calls on vil-lagers to demand accountability from both thegovernment and their neighbors, and to takeresponsibility for the investments they deemimportant. This is a highly ambitious, indeeddownright subversive, objective and one that is very risky in terms of program sustainability.The program will only achieve and sustain itsobjectives if people are able to insist on exercis-ing and retaining this voice in decisions thataffect them. This sustainability, therefore, is notbased on maintaining the current level of KDP

financial flows to the village level. In fact, theseflows will most likely decrease once externalfinancing ceases, even if the proportion of government spending targeted at the villagelevel rises.

Consequently, the fundamental challenge for those inclined to adopt the KDP design is whether these objectives of transparency and participation are to be given sufficient priority, and whether they can be achieved and sustained. (This a challenge still con-fronting KDP as well.) The chief problem with partial achievement of these objectives—for example a modest level of empowerment—is that if people are unable to insist on demo-cratic and accountable government at the levelwhere they can see and feel it, both govern-ment and poor people may revert to formerpatterns of behavior.

Bypassing Government To Get Good Government

It is not surprising that with such bold objec-tives, a primary hurdle in launching a KDP-type program is ensuring government accept-ance. Indonesia’s sociopolitical context offerssome instructive lessons. The KDP concept waslaunched4 at a time when the government rec-ognized that traditional means of gaining thepatronage of rural people—pumping moneydownward through successive levels of govern-ment—had become counterproductive.Corruption was so rife5 that few funds actuallyreached the villages, and the credibility of alllevels of government had long been destroyedin the process. As a result, the target popula-tion’s perception of government and, possiblymore important, the government’s own percep-tion of its credibility with this population, were

3 Project Appraisal Document KDP-II, p.2.4 Several design features of KDP were tested earlier in a Bank-financed Village Infrastructure Project (VIP). The objective of VIP

was also “to reduce poverty in poor rural villages in new ways, by promoting village participation, transparency and decentral-ization.” Each selected village was to build infrastructure of its choice (limited to 5 types of public infrastructure: roads, bridges,drinking water, communal sanitation units, piers), to receive (i) a one time grant, then equivalent to some $54,000, disburseddirectly to the villagers from the local branch of the commercial bank selected (so bypassing some lengthier and less transparentbureaucratic processes); and (ii) technical assistance directly via the consultant field engineer assigned to a cluster of 5 villages.The central government organized and managed the system, including procurement, training of the engineers, audits and moni-toring, and the local public works agency assisted with the public administration. See World Bank 1999b.

5 According to the Corruption Perceptions Index 2001 published by the NGO Transparency International, Indonesia ranks number 88out of 91 countries surveyed. The country has been at the bottom of the index ever since it was first established in 1995.Indonesia usually scores approximately 2 on a scale from 10 (highly clean) to 0 (highly corrupt). See under http://www.globalcorruptionreport.org.

Design Considerations in Community Driven Development

3

so negative that the government was willing toexperiment with a radical, new approach. KDPprovided for funds to flow directly from a central project account to a joint village accountat a local sub-district bank, processed by thebranch office of the national Treasury. At nostage do funds pass through a governmentministry. In these circumstances, governmentofficials have supported KDP in the hopes ofbenefiting from the associated political capital,even though they have no access to the fundsthemselves.

KDP remains a calculated political gambleby the project team to improve government bytemporarily bypassing it—through direct trans-fer of funds and the creation of parallel institu-tions to oversee program implementation. Itruns counter to a Bank-wide trend during the1990’s to move away from heavy reliance onproject implementation units (PIUs) by “put-ting government in the driver’s seat,” strength-ening government institutions, and insistingthat projects and programs be implemented bygovernment agencies. In fact, the KDP NationalManagement Consultants (NMC) office inJakarta is in reality a PIU, or at best an in-houseNGO. It also runs counter to some decentraliza-tion initiatives that originate with federal government offices, and subsequently cascadedown through development plans prepared atvarious levels.

“Bypassing” government, however, may be too categorical a characterization of KDP.Indeed, KDP is not without precedent amongtraditional fiscal-decentralization measures,such as block grants to municipalities. AndKDP funds are registered on local governmentbooks. In fact, the government itself channelsfunds through KDP. Technical advice and project management support are also providedby the kecamatan government staff, many ofwhom have become strong supporters of theprogram. Nonetheless, financial flows and criti-cal areas of management and audit are outsidenormal government channels.

Full replication in other countries of this fun-damental design feature requires an assessmentof the extent to which existing governmentfinancial and technical advisory systems can beused. Introducing well-financed parallel struc-tures can have the undesirable consequence ofweakening government. Therefore, it should

only be contemplated if the risks of undermin-ing government financial processes do notexceed the prospects for reforming existing systems. Furthermore, the feasibility of using acombination of government and direct-fundingchannels in countries where some governmentagencies work more effectively than othersdepends on the likelihood that funding will notonly reach municipalities, but rural areas aswell. In Indonesia, the provincial governmentconfronts the greatest potential for loss ofpower of any government level. This is attribut-able to the central government’s strategy (unre-lated to KDP) to allocate funds directly to all ofIndonesia’s 4,028 kecamatans. This also pres-ents challenges in other countries, in terms ofappraising the appropriate role of the province.

Commitment to a Cross-SectoralApproach

As noted above, KDP provides a block grantdirectly to participating kecamatans, each con-taining about 20 villages. Villages participatingin KDP can submit proposals for a wide rangeof activities, from village infrastructure projectsto seed capital for small business creation orexpansion—what is termed “open menu.”Infrastructure projects include construction or repair of communal infrastructure (roads,bridges, sanitation) or the repair or expansionof buildings used for service delivery (health or education facilities). In the latter case, thevillage is required to obtain confirmation by the service provider that the facility will bestaffed. KDP thus differs from traditional village infrastructure activities, such as thosefinanced under the Village InfrastructureProject (VIP) that provide a menu of five typesof infrastructure projects, all of which comeunder the jurisdiction of the public worksauthority (see footnote 2). The wide range ofchoice is a key design feature, because it rein-forces the participatory processes necessary tosustain it. The process itself, therefore, culti-vates community capacity and confidence toprioritize and select project proposals—a keyprogram objective.

Community driven development (CDD)projects that do not embrace such an openmenu cannot be qualified as “KDP-type” proj-ects. (Indeed, a case can be made that a projectwithout investment choice cannot be consid-

Indonesia’s Kecamatan Development Project: Is it Replicable?

4

ered a community “driven” project.) This is notto say, however, that KDP succeeds in stimulat-ing a preference for wildly distinctive projects.The majority of projects fall into the category ofvillage infrastructure because that is what peo-ple are most familiar with, particularly follow-ing their experience with the VIP. There are fewproposals for rehabilitation of social infrastruc-ture, such as schools or health posts. Thisstands in contrast to the experience in socialfunds where education and health rehabilita-tion projects are often selected, demonstratingthe important influence of Bank or governmentstaff in shaping community preferences for spe-cific infrastructure projects. KDP has also expe-rienced difficulties in capturing the interest andtechnical support of sectoral ministries in par-ticipating kecamatans, the backing of which isimportant in advising on project options.Nevertheless, the carte blanche offered to vil-lagers is an important signal of their ownershipof the process and outcomes, and is essential toachieving KDP objectives. Any country consid-ering replicating KDP would therefore beadvised to include the cross-sectoral projectdesign option.

Credit for Economic Activity as a Project Option

A feature of KDP’s open menu that merits spe-cific attention is the inclusion of both infra-structure and credit for economic activities asproject options for villages. In KDP, financing isprovided on a grant basis for village infrastruc-ture projects; the funds are lent at market ratesof interest for economic activities. Loan repay-ments are made to the kecamatan or retained inthe village to finance new economic activitiesor public goods.

Credit schemes have for years been a contro-versial component of social fund projects.6

Economic activities have often been excludedbecause of the difficulty of measuring returns,the risk of elite capture for private gain, lowrecovery rates, and the difficulty of providingsuitable technical advisory support on a sus-tainable basis for widely diverse micro-schemes.7

KDP’s credit program has not escaped theserisks and weaknesses, and there was consider-able debate within Indonesia whether to retainthis component in KDP II. Repayment rates for economic activities vary greatly, but areimproving. Nationwide they were onlybetween 50 and 60 percent in KDP. On Java,where more than one-third of the KDP keca-matans are located, the repayment rates are 90percent and higher. Criteria for allocating fundshave been nearly impossible to adhere to, andtechnical and legal advice for small ventureshas been close to nonexistent.8 At the sametime, the credit scheme has had some signifi-cant gains. Returning borrowers are proving to be good credit risks, and some of the smallrevolving credit schemes (simpan pinjam) areshowing good internal credit discipline andproviding community social safety nets. Inaddition, production schemes appear to be generating small-scale employment and tradingnetworks. Anecdotal evidence suggests thatKDP credit provides first-time credit access for poor people, and that the KDP’s economicdiscussions are a key method for bringing ingroups, such as women, who often do not par-ticipate in community decisionmaking. Thecredit program has been extremely popular inthose Indonesian villages that already havebasic infrastructure, particularly road access, as it has been used to increase agricultural,food processing, and trading activities.

6 See, for example, CGAP 2001.7 According to a recent OED study about the Bank’s experience with microenterprise finance institutions (MFI) between FY87 and

FY97, only slightly more than half of the total number of 81 World Bank–supported projects are likely to have satisfactory out-comes and only 25 percent of the 40 projects providing funding at below-market rates were doing so efficiently. The overallresult is that only 45 percent of Bank-supported projects are likely to be sustainable because, for sustainability, the MFI must beeither fully self-financing or have clear, accessible resources to deal with any revenue shortfall. See World Bank 1999a.

8 These include requirements that borrowers have to (i) demonstrate ineligibility to access bank loans; (ii) pay a 20 percent rate ofinterest, and (iii) repay the loan within 18 months. In addition, only groups that have existed for more than one year can make aproposal, a specific cost proposal must be made, and further credit is denied to groups or villages that do not repay. KDP IIimproves the economic option in some of the following ways: all proposals must be verified by a consultant hired from a localbank; communities that are eligible for the economic option must show year-to-year repayment improvements of 20 percent ormore (or already be greater than 80 percent); funds revolve at the village rather than kecamatan level, to increase social pressurefor repayment; and communities adopting the economic option must use a portion of their technical assistance budget to con-tract a consultant who will train village members in basic bookkeeping.

Design Considerations in Community Driven Development

5

A consultant review of the KDP microcreditscheme concludes that “the very same designthat brought about project success doomedKDP’s microcredit component to failure…(T)heparticipatory planning and decisionmakingprocess by which funds are competed forbetween villages…is good for allocating fundsearmarked for public goods…(but is) incompat-ible with the sound management of credit andrevolving funds.”9 In short, on grounds of finan-cial sustainability, the evidence points to a negative return for the KDP microcredit scheme.KDP II is working to strengthen the credit com-ponent by linking it to pre-existing microfinanceinstitutions or banks to the extent possible.

The inclusion of a credit finance componentsuch as KDP’s can be a considerable drain incases in which resources are severely limited oradministrative arrangements need to be kept assimple as possible. Whereas the kind of infra-structure supplied under project funding isgenerally perceived to be part of government’sfundamental role, the provision of credit hasusually been viewed by the international devel-opment community as the role of the market.Acknowledging that commercial credit marketsoften do not successfully reach the poor, gov-ernment schemes have not proven sustainablein providing credit either. Adoption of creditas part of a community development program

should therefore be approached with caution.

Project Scope: Scale From the Start

Launched in July 1998, KDP had its antecedentsin a pilot program within the VIP. This phasewas short (from June 1997 to June 1998), andthe KDP design called for a very rapid nationalroll out, rather than gradual expansion of apilot operation. The project reached 501 keca-matans the first year of implementation, andadded 271 kecamatan the second year and 257the third year, which now total 984 kecamatan.Whereas this entailed a “ready-fire-aim”10 learn-ing approach, it avoided the pitfalls inherent in“boutique” operations, with their heavy, oftenexpatriate, technical assistance, their unrealistic

cost structure, and their rarefied implementa-tion environment. Not only did the programexpand rapidly, but also the design called for—and succeeded in achieving—a short periodbetween initial sensitization of a village and thecompletion of the village-level project. Key tothe project’s success was the four to six monthlong facilitation process leading up to projectselection, at which point funds are transferredto the village account.

The average implementation period of vil-lage infrastructure projects was three to fourmonths in the first year of KDP. In general,rapid roll out is important to test a project’sreplicability within a country, to retain simplici-ty, and to gain credibility and enthusiasm necessary to garner villagers’ confidence. Thesimplicity and “direct-deposit” nature of thefunding scheme, from central project account to deposit in a kecamatan account, has con-tributed to the speed of the KDP roll out.Substantial pre-implementation work andagreement with national partners is thereforeessential, and should be factored into consider-ation to adopt a KDP approach.

Existence and Influence of NGOs

In contrast to many countries where NGOs areimplementing agencies for community-basedprojects, KDP relies on village-level processesand institutions that interface with the civiladministration. In some cases these are thelocal assemblies which, under recent decentral-ization directives, are now elected by villagers.Project briefings and selection of proposals areheld in village meetings where all villagers areinvited; these are supplemented by meetings atthe hamlet (sub-village) level and with variousgroups of villagers. In many respects, KDPembodies direct democracy, where decisionsare taken by all and in the presence of all.

Implementation is also broadly participatory,particularly for infrastructure. Because of thepoor track record of contractors in terms ofmeeting commitments and eschewing corrup-tion, work is organized by local implementa-

9 Holloh (2001).10 This refers to a scaling-up approach that starts with a large number of trials early on to learn from, through as large a sample as

possible, and to follow with refinement of approach for subsequent roll out.

Indonesia’s Kecamatan Development Project: Is it Replicable?

6

tion teams with the assistance of a local govern-ment project manager, using village labor,which may be paid or voluntary.

Indonesia does not have a strong tradition of freely operating NGOs. Many existing NGOshave been found to collude with government,and have therefore been discredited. Moreover,since KDP aims at empowering individuals toparticipate as community members who insiston government accountability, intervention byany intermediary whatsoever, even well-inten-tioned NGOs, can be counterproductive. To theextent that NGOs often represent particular sec-tors, their preferences could be particularlydetrimental for village prioritization of projects.

Over the years, many NGOs have proven tobe significant advocates for the poor in manycountries. Although NGOs are not implement-ing agencies in KDP, monitoring by independ-ent advocacy NGOs and the media has beenpivotal for the inclusion of civil society organi-zations. In addition, a legal advocacy pilot in two provinces pairs KDP project villageswith national and district advocacy NGOs and university-based law faculties. However,the challenge remains how to best involve them constructively in KDP-type projects. That decision will likely depend on particularcountry contexts.

Regimes of Land Tenure

Indonesia’s Basic Agrarian Law (BAL) of 1960protected traditional land (adat) rights andwove principles of adat into the formal law,thereby supporting continuity and stability.11

This in turn has permitted the development ofa flexible land tenure system with many levelsof security—a system that allows many low-income households to acquire land rights withsome security where a more formal system (ofregistration and hence ownership) would not.As a consequence, a central feature in rural

Indonesia, in contrast to many countries, is thesmall proportion of large landholdings and theassociated number of tenant farmers: the aver-age size of a farm is one hectare, and only 0.2percent of all farms are 10 hectares or larger.More than 70 percent of farms in Indonesia aresmaller than one hectare, compared with only38 percent in the Philippines.12

Thanks to the relatively secure access ofhouseholds to land for production, ruralIndonesians can anticipate direct benefits fromboth infrastructure and economic investments.Their commitment to the KDP processes ofboth planning and implementation are there-fore more assured. Building a bridge increasesaccess to their fields or to markets for their pro-duce. Microcredit obtained through the simpanpinjam can be used to clear a small plot andpurchase some pepper seedlings; repayment ofthe loan prior to maturation of the pepperplants can be made from sale of another crop.In Indonesian villages, more powerful familiesmay capture a disproportionate share of thebenefits through their larger land holdings orability to manipulate decisionmaking processes.However, their dominance does not normallytotally shut out other villagers from benefits.Moreover, anecdotal evidence from KDP sug-gests that the relatively small size of the grants,coupled with the time commitment required formeetings, are too much bother to attract thewealthier members of the community.

In feudal land regimes, such as those insome Latin American countries where themajority of land at the village or district level isowned by one or a few families and a largeshare of the population works as laborers, it isunlikely that a KDP design would be able tocounter the economic and political dominationof the elite. Even in feudal systems, therewould still be demand for public goods infra-structure, particularly that aimed at improvingthe lives of the poor, such as village water sup-

11 Two types of land systems existed in Indonesia before a unified system was introduced under the Basic Agrarian Law (BAL) of1960. One type was that of the traditional adat rights, which still exists in many parts of Indonesia, where land is owned andhanded down according to customary law without formal registration or title, and with ownership and holdings based on com-munity acceptance of boundaries and claims. The second type was the more Westernized system of written land titles and landregistration. The BAL established a system of land rights with varying degrees of tenure linked to citizenship requirements. Inprinciple, the BAL also established its precedence over the adat system of laws, although rights acquired under adat prior topromulgation of the BAL were protected. Based on: MacAndrews 1986 and Behuria 1994.

12 The data are for 1993; Rosegrant and Hazell 2000, p.99.

Design Considerations in Community Driven Development

ply and social service infrastructure. However,the stimulus that KDP infrastructure and creditattempt to provide to stimulate agriculture-based livelihoods would be muted. And theimpact of KDP on accountability and respon-siveness of local officials towards the popula-tion at large—the underlying process objectiveof KDP—would also be weak in the face of theconcentration of power by a small elite.

Human Resources and RemunerationLevels

A key feature of the KDP design is the largequantity of high-caliber staff. During the sec-ond year, the program employed a total of20,500 contract staff serving a population ofalmost 10 million. By far the largest percentagewas at the village level with 15,300 village facil-itators, or one facilitator for every 650 people.In addition, some 1,200 consultants from 16 pri-vate companies at the national (15), provincial(21), district (164), and kecamatan (959) levelsprovided technical assistance and guidance tothe project. The UPK (Unit PengelolaanKeuangan), the financial management arm ofKDP, employs about 3,000 people country-wide, most at the kecamatan level.13

The large staff is due to the need for a rela-tively high degree of facilitation and financialoversight. The project identification and selec-tion process at the village level is both time-and labor-intensive in order to ensure the par-ticipation of women, to mobilize interest at thesub-village level, and to promote transparency.While government staff from sector depart-ments or public works are expected to providespecialized technical assistance, the projectmust also rely on its own technical staff to vetproposals and advise village implementationteams or groups undertaking microenterpriseventures. The UPK employs at least three staffat the kecamatan level for bookkeeping, expen-

diture reporting, overseeing any large procure-ments, and completion documentation. Thesejob profiles require a well-educated cadre ofstaff: even at the village level, facilitators musthave strong literacy, accountancy, and facilita-tion skills. KDP has been able to find staff withrelatively high levels of education. Kecamatanfacilitators are usually university graduates,and many village facilitators have a high schooldiploma or even some university education.With the economic downturn in Indonesia,these are often young people who have beenunable to continue their studies or find jobs in cities, and have returned to their villages.14

While education levels are relatively high inKDP, salaries remain low. A village facilitatorearns US$15 (October 2001 exchange rate) amonth (on a five-month contract in the past, to increase to a full twelve months in KDP II15

with probably the same pay). A kecamatanfacilitator earns a monthly base pay of $135 to$170, depending on experience.16 The projectcan therefore afford to employ a large numberof well-qualified staff for both communitycapacity building and program oversight.These education and salary figures are consis-tent with Indonesia’s overall education andincome statistics relative to other countries ofEast Asia. While its average education levelapproaches the overall level for East Asia,17 itsGNI per capita is only 54 percent of the EastAsia and the Pacific Region average for 2000.Average government salaries, which have some bearing on KDP salary levels, are low, at 414,000 Rps per month ($53 at 1999 rates).18

During 1996-2000, the ratio of governmentwages to per capita GDP was 1:1 in Indonesia,compared with an average of 2.9:1 for the EastAsia and the Pacific Region.19

In many countries, education and salary levels, and terms of employment would makethe KDP model unaffordable. These include

13 Most staff are employed on temporary contracts of 5 to 12 months. 14 Even with this level of skills available, KDP leadership reports that they struggle with a dearth of trained staff at all levels to

assist communities in local level planning and decision making. The importance of technically competent staff to provide assis-tance to communities should not be underestimated.

15 All dollar amounts are U.S. dollars.16 In addition to base pay, the kecamatan facilitator gets about $70-80 for motorbike and house rental, a travel allowance, and

funds for operating the office.17 Primary gross enrollment ratios are greater than 100 percent on average in the East Asia and Pacific Region. The secondary

gross enrollment ratio was 56 percent in Indonesia and 69 percent in the East Asia and Pacific Region in 1997. See World Bank2001b.

18 Filmer and Lindauer 2001.19 See http://www1.worldbank.org/public sector/civilservice/development.htm.

7

Indonesia’s Kecamatan Development Project: Is it Replicable?

8

20 Whereas the average conversion rate between the Rupiah and the U.S. dollar in 1990 was 1,842, it stood at 7,855 in 1999 and at8,500 in 2000.

countries where average education levels arelow and a secondary or university educationcommands a premium, where public sectorsalary levels are generally high, or whereemployment terms and expectations are lessflexible. Indeed, the affordability of the KDPwage bill is due in part to the dramatic declinein value of the Indonesian Rupiah since the1997 financial crisis.20 Adapting KDP’s person-nel profile and density to higher-cost countries,as well as ensuring its sustained implementa-tion within Indonesia, is likely to require inno-vations to reduce overall numbers and improveeffectiveness of a more limited number of per-sonnel. This will nonetheless be a challenge tocountries such as Indonesia where intensivefacilitation is required to overcome traditions of corruption, a lack of transparency, and top-down planning.

Locus of Project Responsibility

It is traditional wisdom in project design toidentify as strong and capable a governmentagency as possible to champion a project, onewhose staff embrace the vision of the programand can spearhead and sustain the necessaryreforms and oversee implementation. A chal-lenge to the success of community develop-ment programs has been the relatively weakstaff and political clout of ministries of commu-nity development or social welfare. Bank staffhave often steered clear of these agencies infavor of community development arms of economic or other sectoral ministries.

KDP has taken a contrarian approach. It hasas its implementing agency the CommunityDevelopment Department (PembangunanMasyarakat Desa-PMD) of the Ministry of HomeAffairs, an agency that is less powerful thanother ministries in the Government. It nonethe-less has a number of features in its favor:

• KDP is a bigger fish in the PMD pond thanit would be in the portfolio of other min-istries—it represents their opportunity forgreater influence and therefore commandstheir growing attention and efforts;

• PMD has less vested interest in pushingone sector or another, and therefore buys

into the open menu, community choiceobjectives of KDP;

• PMD is reluctantly willing to cede some ofits power to the project management con-sultants in the interest of speedy and effec-tive implementation; contracting out suchproject implementation details might bemore difficult if the implementing agencywere more powerful and able to intervene,potentially to the detriment of projectobjectives and transparency; and

• As part of Home Affairs, PMD—and par-ticularly KDP—benefits from the adminis-trative, logistical, and political support thatextends from central to provincial and sub-provincial levels; this includes provision ofoffice space and infrastructure as well asaccess to authorities at these levels.

The downside risk of implementation by aweaker ministry is the greater likelihood ofunsustainability of funding and strong gover-nance once project funding and oversight areno longer shepherding the program. Theseopportunities and risks among different imple-menting agencies are highly specific to eachcountry, and need to be carefully weighedwhen considering the launch of a communitydriven program such as KDP.

Demands on the Bank

Design and implementation support for a KDPrequires a large upfront investment by theBank. The upstream work entails significantanalytical work on community-level institu-tions and their interface with formal publicinstitutions–the organizational structures,stakeholders, and the formal legal, regulatory,and political regimes. It requires fastidiousattention to the details of procurement andaccountancy practices and the flow of fundsand goods. An understanding of genderdynamics is particularly important to deter-mine how cultural practices can be respectedwhile responding to the development prioritiesof both men and women. Tackling corruptionrequires an understanding of the incentives thatdrive it, creativity in design of counter-incen-tives, and fortitude to go public in cases

Design Considerations in Community Driven Development

9

of fraud. And finally, the Bank needs to ensurethat its own advice and lending portfolio areinternally consistent with and supportive of theobjectives of KDP-type projects, particularlywith regard to community choice and insis-tence on good governance.

It is unlikely that the Indonesia KDP wouldhave been feasible in the absence of a highlycapable international Bank staff member basedin-country who was able to devote a largeshare of his/her work program and energies to KDP. Part of the KDP’s success is attributableto a rapid response to project complaints,which has led to course corrections and practi-cal solutions the villagers can see. It has alsorequired the vigilance of Bank staff to minimizeelite capture at local and national levels. Suchwork can simply not be carried out during periodic missions.

At the same time, the intensive preparatoryand supervision work paid off in terms of rapiddisbursement and project implementation. Theproject disbursed 25 percent faster than antici-pated at appraisal, and scaled up by 50 percentmore than was expected during each year ofimplementation. This occurred despite a periodof political crisis in the country, and the factthat disbursements in the entire Indonesia portfolio averaged only 40 percent of the levelanticipated at appraisal. Finally, KDP has beenthe training ground for a considerable numberof Bank staff working on CDD issues in othercountries.

Conclusion

To its supporters, the Indonesia KDP appears tobe almost venerated as the ultimate communitydriven development. To its detractors, it is aparallel delivery system that cannot be sus-tained. In reality, it carries the benefits and risksof both, in its attempt to undertake a bypassoperation in terms of development financewhile stimulating communities to take greaterresponsibility for driving their developmentpriorities and holding their governmentaccountable.

However, the design of KDP is particular tothe Indonesian context. The summary belowshows which factors have shaped its featuresand contributed to date in its successful imple-mentation, distinguishing between project char-

acteristics that were influenced by the Banktask team and those external conditions on the ground.

Exogenous Factors

• The level of disempowerment of poor ruralpeople was such that its bold process andoutcome objectives—participatory processes,strengthened accountability of institutionsand village infrastructure—were deemedworthy by a broad range of stakeholders;

• The high level of corruption in governmentand contract services led the government toagree to bypass its own financial processesand allow funding to go directly from a cen-tral project account to the local level;

• Broad-based land ownership or communaluse of land permitted a wider range of villagemembers to benefit from the program; and

• Relatively high levels of education and lowsalaries allowed KDP to retain a large num-ber of well-qualified staff for implementationand capacity building.

Design Features

• The commitment to a multi-sector, open-menu approach provides the basis for vil-lagers to genuinely choose the projects andactivities that are most important to them;

• The inclusion of credit for economic activity,still a questionable method in terms of viablerepayment rates, increases options for choiceat the village level, and may provide a formof social risk management;

• The rapid roll out, designed as a scale opera-tion from the start, provides credibility to vil-lagers that the program is capable of provid-ing benefits to them;

• Villagers are encouraged to speak out forthemselves, rather than through intermediaryNGOs or CBOs;

• Vesting implementation responsibility in arelatively weaker agency, but with structuresat regional and local levels, has allowed theBank to shape the program; and

• The Bank has been willing to commit a rela-tively high level of staff resources, based in-country, to see through the design and imple-mentation of the program.

The matrix above lays out which features aredeemed indispensable, adjustable, and dispen-sable by the framers of KDP.

While no program can be directly replicated,those seeking to adapt a KDP approach shouldweigh each of the above characteristics todetermine to what extent they might be adapted for successful adoption elsewhere.

Indonesia’s Kecamatan Development Project: Is it Replicable?

10

Design Feature Adopt Adjust Abandon Comments

Direct X Important to be rapid, responsive, disbursement reliable, and free from tampering.flows Independent audit

Open menu X Choice and decision process essential

Microcredit X Benefits mixed; if open menu for infra-structure in place, credit not essentialoffering

Project scope X Scale needs to be figured into design fromthe start; rapid roll out required to build and maintain confidence

Role of NGOs X Stronger NGOs bring advantages in roll out and disadvantages of NGO overpowering villagers’ decision process

Land tenure ? Concentrated ownership likely to stifle buy-in by poor but returns possible forsocial infrastructure

Human resources X Higher salary scales likely to requirereduced numbers, contracting out, otherfacilitation modes

Locus of X Governed by country contextgovernment champion

Access to X Important to inform choice, ensureInformation transparency of financial flows,

underpin rapid roll out

11

Annex

A Summary of Indonesia’sKecamatan Development Project1

Introduction

The Kecamatan Development Program (KDP)is a Government of Indonesia effort to alleviatepoverty in rural communities and improvelocal governance. The program targets thepoorest kecamatans (sub-districts) in Indonesia.KDP aims to foster more democratic and partic-ipatory forms of local governance by strength-ening kecamatan and village capacities andimproving community participation in devel-opment projects. KDP provides block grants ofapproximately $43,000 to $125,000 a year direct-ly to kecamatans and villages for small-scaleinfrastructure, and social and economic activi-ties. The program works with village councilsand kecamatan development forums. KDP issupported by facilitators and consultants fromboth the village and national level, who pro-vide technical support and training.

Most importantly, KDP emphasizes the prin-ciples of community participation—especiallyfor women and poor villagers, transparency,competition for funds, and sustainability. AllKDP activities aim at allowing villagers tomake their own choices about the kinds of projects that they need and want.

How Does The Program Work?

The KDP project cycle has several stages: infor-mation dissemination, planning, proposalpreparation and verification, funding decisions,implementation, and follow-up. All stages aimto have a high degree of community participa-tion and transparency throughout the process.

Information Dissemination. Information dis-semination about KDP occurs in several ways.Workshops are held at the provincial, district,and kecamatan levels to disseminate informa-tion and popularize the program. The work-shops involve community leaders, local govern-ment officials, local press, universities, andNGOs. Dissemination of information at the village level occurs through large village meet-ings as well as through group- and hamlet-levelmeetings to share information and encouragepeople to propose ideas for KDP support.

Planning. Planning meetings occur at the sub-village and village levels. Village and keca-matan facilitators disseminate informationabout KDP procedures and encourage villagersto submit ideas for KDP funding. To encouragethe participation of women in the planningprocess, a women’s-only meeting is held to dis-cuss their proposal. At the second village meet-ing, villagers’ discuss their ideas and the forumdecides which proposals the village will putforward to the inter-village forum (UDKP).

Proposal Preparation. Each village administra-tive body (LKMD) can submit up to two pro-posals to the UDKP with a minimum of one ofthe proposals originating from women’sgroups. Proposals from one village can total upto a maximum of Rps 300 million ($33,300).

Project menus are open to all productive invest-ments except those on a negative short list. Theopen menu policy is an important element ofKDP. It allows villagers themselves to decidewhat they want. Proposals can include a mix-ture of various economic, social, and infrastruc-ture activities if the villagers so choose.

1 Excerpt from internal memo, KDP, World Bank Indonesia office.

Indonesia’s Kecamatan Development Project: Is it Replicable?

12

Verification. Verification of the proposals’ technical elements occurs during the proposalreview stage prior to project selection. A keca-matan verification team usually includes com-munity leaders, the kecamatan facilitator, andappropriate technical staff recommended by thedistrict engineer. The district engineer also doesa final check before the results of the verification are presented and considered in the second kecamatan level meeting (UDKPII).

Project Selection. Village proposals are dis-cussed and decided on at the UDKPII. TheUDKPII forum reviews the findings of the verification team and discusses the merits andbudgets of each proposal. Final decisions fromthe UDKPII meeting are posted on KDP infor-mation boards and shared with the villagersthrough the third village-level meeting(Musbangdes III) and at smaller sub-villageand group meetings.

Project Implementation. Once it has beendecided which projects should be funded, village technical assistance is mobilized andcontracted. Contracts are signed with theLKMD and acknowledged by the district engineer. Technical designs are finalized andthe mobilization of workers takes place. Acommunity implementation team of five per-sons in each village is elected during theMusbangdes III meetings to oversee projectimplementation. The community contributionsare also collected. The kecamatan facilitatorsand the kecamatan project manager (PjOK)who is appointed by the government, supportsgroups and LKMDs on administrative andfinancial matters. Communities, governmentofficials, and KDP consultants supervise theimplementation of activities. Village meetingsreport on the status of project implementation.

Figure 1: Indonesia Second Kecamatan Development ProjectProject Costs and Financing Arrangements

$ ’000

Implementation Period

Year 1 Year 2 Year 3 Year 4 Total

Project Costs

Kecamatan Grants & Pilots 80.5 103.5 103.5 23.0 310.6

Community Capacity Building 20.4 20.4 19.6 5.0 65.5

Implementation Support 10.2 11.1 11.1 8.0 40.4

Monitoring & Evaluation and Studies 0.7 0.7 0.7 0.7 2.9

Front End Fee 2.1 2.1

Total 114.0 135.7 134.9 36.8 421.5

Financing Arrangements

IBRD/IDA 97.7 103.9 103.1 15.5 320.2

Government 5.8 18.3 18.3 18.3 60.8

Central 5.8 5.8 5.8 5.8 23.3

District 0.0 12.5 12.5 12.5 37.5

Beneficiaries 10.5 13.5 13.5 3.0 40.5

Total 114.0 135.7 134.9 36.8 421.5

13

Maintenance. Once the infrastructure is built,maintenance teams and user fees are estab-lished. Villagers form operation and mainte-nance committees that are responsible for over-seeing the proper maintenance of facilities andfee collection. For economic loans, repaymentsare collected by the kecamatan financial man-agement unit (UPK) for a maximum period of 18 months.

Management Structure of KDP

The management of KDP is the responsibilityof the Ministry of Home Affairs (BPM). At thenational level, BPM deals with the day-to-dayoperations of the project. Government coordi-nation teams, representing various ministries,also assist with KDP. At the national level, thecoordination team is headed by the NationalDevelopment Planning Agency (Bappenas). Atthe provincial and district levels, coordinationteams are headed by the Regional DevelopmentPlanning Agency (Bappeda). At the kecamatanlevel, the role of the BPM section head (PjOK) isessential because he/she serves as KDP’s localproject manager. The PjOK is assisted byadministrative staff from Home Affairs.

KDP is assisted technically by consultantteams. Some 3,800 consultants from 20 privatecompanies at the national, provincial, districtand kecamatan levels provide technical assis-tance and guidance to the project.KDP has alsodeveloped a comprehensive monitoring andevaluation system. In addition to semi-annualWorld Bank supervision missions and regularmonitoring by field consultants and govern-ment officials, local journalists and 45 NGOsconduct independent, external monitoring ofKDP in each province. KDP also established a Complaints Handling Unit to channel com-plaints and questions from the community and to follow-up on individual cases. TheGovernment Audit Agency (BPKP), and theKDP Financial Supervision and Training Unitaudit and provide supervision and training onlocal finances. Several thematic studies andimpact evaluations have also been conducted to assess the program’s benefits and impact.

Annex

14

Figure 2: Indonesia: Second Kecamatan Development ProjectSequence of Activities

Phase Activity• Provincial and district workshops to publicize KDP among local government

officials, press, universities and NGOs• Kecamatan Facilitator training• Kecamatan forum meetings (UDKPI) to publicize KDP and to

evaluate project staff (just in old kecamatans) and select Year 2 villages (justin new kecamatans)

• 1st village meetings (Musbangdes I) to publicize KDP in the village andselect Village Facilitators

• Train Village Facilitators• Information boards placed in all villages• Inventory of all village groups and locations of poorer villagers• Sub-village and group meetings, cross-visits• Updating information on village boards• Further dissemination of project information• Special meetings with women’s groups• Special meetings with women’s groups to discuss women’s proposal ideas• 2nd village meetings (Musbangdes II) to discuss proposal ideas • Village TA for proposal preparation selected• Formation of team to prepare village proposals• Preparation of proposals, with designs and budgets• Updating information on village boards• Cross village visits to learn about process• Formation of verification teams• Field visits and inspections• Team meetings to compile results and recommendations• Feedback to villages and groups• 2nd Kecamatans Forum Meeting (UDKPII) to select village projects for fund-

ing and to select UPK (Financial Management Units), DPRD joint teamsformed.

• The team from one village visits another village to learn about process• Evaluation of project staff by UKDP II forum in old and new kecamatans.• 3rd Village meetings (Musbangdes III) to discuss results of UDKP II meeting,

elect community implementation Team, decided on village technical assis-tance and elect teams to control Implementation in village.

• Updating information on village boards• Training for Implementation Teams, UPKs, Village TA• Administration• Release of funds:

- from the state treasury in 3 stages (40%, 40% & 20%)- from collective village accounts as needed (not pre-set %)

• Mobilization of village laborers• Procurement of materials and equipment

- less than Rp 15 million, information from 3 sources (not written)- Rp 15 million or more, 3 quotations (written) evaluated by village teams

• Implementation of village activities, loans disbursed• Supervision, monitoring, reporting and monthly meetings of facilitators• Village meetings to account for funds (when about 50% of funds expended)• Formation of maintenance teams• Training for village councils and maintenance teams• Repayment of loans and relending of repaid funds, on UDKP instructions• Maintenance of infrastructure• Evaluation of project and staff by villagers• Village meetings for hand-over of completed projects, to account for

funds used

SocializationDissemination of

Information

Selection of Activitiesfor Village Proposals

Socialization inHamlets and Activity Suggestions Collected

Feasibility /Verification Stage

Funding Decisions(Selection of Proposals

to be Funded)

Post Implementation

Stage

➧➧

➧➧

➧➧

➧

Preparations forImplementation

Implementation

15

References

Behuria, Sutanu. 1994. “Some Aspects of LandAdministration in Indonesia: Implications forBank Operations.” Asian Development BankOccasional Paper No.10, Manila.

CGAP (Consultative Group to Assist the Poor).2001. “Microfinance, Grants, and Non-Financial

Responses to Poverty Reduction: Where DoesMicrocredit Fit?” CGAP Focus Note 20.

Filmer, Deon and David L. Lindauer. 2001.“Does Indonesia Have a ‘Low Pay’ CivilService?” World Bank Research Paper.

Holloh, Detlev. 2001. Review of the KDPMicrocredit Approach, prepared for WorldBank, Jakarta.

MacAndrews, Colin. 1986. “Land Policy inModern Indonesia. A Study of Land Issues inthe New Order Period.” Oelgeschlager, Gunn &Hain, Boston.

National Management Consultants 2000.Kecamatan Development Program. SecondAnnual Report 1999/2000, prepared for theKDP National Secretariat.

Rosegrant, Mark W. and Peter B.R. Hazell.2000. “Transforming the Rural Asian Economy:The Unfinished Revolution.” Oxford UniversityPress.

Woodhouse, Andrea. 2001. “FightingCorruption in KDP.” Draft, World Bank JakartaOffice, July.

World Bank. 1998. Kecamatan DevelopmentProject. Project Appraisal Document, ReportNo: 17397-IND.

World Bank. 1999a. “The World Bank andMicroenterprise Finance: From Concept toPractice.” Operations Evaluation Department.

World Bank. 1999b. “Village InfrastructureProject.” Implementation Completion Report,Report No: 19099.

World Bank. 2001a. “Second KecamatanDevelopment Project.” Project AppraisalDocument, Report No: 22279-IND.

World Bank. 2001b. World DevelopmentIndicators.