SERVICES FOR THE PEOPLE, BY THE PEOPLE: INDONESIAÕS PROGRAM FOR COMMUNITY EMPOWERMENT, 1998 Ð 2006 SYNOPSIS When financial crisis and weather-related natural disasters ravaged IndonesiaÕs economy in 1997, national leaders searched for ways to cushion the impact on poor rural households. A team of public servants within Bappenas, the countryÕs powerful national development planning agency, suggested an aggressive, nationwide expansion of an experiment in community-driven development. The Kecamatan Development Program (KDP), which worked at the kecamatan, or subdistrict, level, furnished block grants directly to poor communities and empowered villagers to determine how they wanted to use the fundsÑwhether for building roads, bridges, schools, or health clinics. Communities chose, planned, implemented, and maintained projects on their own, supervised by village volunteers, subdistrict committees and verification teams, and specially trained facilitators. Planners at Bappenas worked with the World Bank to modify and scale up the original KDP experiment. The Ministry of Home Affairs, which also participated in the early phases, took over the program two years later. During an eight-year period, the new KDP provided direct benefits for more than half of IndonesiaÕs 70,000 villages, helping communities move out of poverty in greater proportions than their counterparts in non-KDP areas did. Rushda Majeed drafted this case study based on interviews conducted in Jakarta, Indonesia, in October 2013. The research benefited from additional interviews conducted by Jonathan Friedman in May and June 2013. This case study is the first in a two-part series; see ÒExpanding and Diversifying IndonesiaÕs Program for Community Empowerment, 2007 Ð 2012.Ó Case published February 2014. INTRODUCTION The Asian financial crisis that struck Indonesia in mid-1997 reversed years of economic progress and plunged millions of rural poor below the poverty line. During the previous two decades, the country had made significant gains in improving standards of living. The number of people earning less than US$1 a day had fallen from more than 50% in 1970 to less than 20% by 1997.1 But almost overnight, many of those gains evaporated. Adding to the countryÕs misery, a period of intense ocean warming called El Ni–o began to distort weather patterns. On top of fiscal chaos, Indonesia faced massive crop failure as a result of drought and forest fires. The government had to act quickly to provide relief. Leaders thought a community- ISS is a joint program of the Woodrow Wilson School of Public and International Affairs and the Bobst Center for Peace and Justice: www.princeton.edu/successfulsocieties. ISS invites readers to share feedback and information on how these cases are being used: iss@princeton.edu. driven antipoverty program could hold the answer. A few years before the crisis, in the early 1990s, the deputy minister of the Ministry of National Development PlanningÑcalled BappenasÑhad assembled a team to help develop new ways to address rural poverty. Gunawan Sumodiningrat, one of the agencyÕs deputy ministers, helped launch a series of experiments with the help of Herman Haeruman, deputy director of regional development; Tatag Wiranto, director of rural development; and Sujana Royat, director of urban land and special programs, among others. The initiatives aimed to improve livelihoods and reduce regional inequality, which had become an increasing source of worry in several parts of the country, an archipelago of more than 17,000 islands.2 Some of the experiments in poverty reduction had transferred money from the central government to village heads, who then took charge of helping poor residents buy livestock or other assets. Other projectsÑ organized by a small team under the direction of Danish-Argentine economist Frida Johansen of the World BankÑused labor-intensive methods for village infrastructure construction. But in all cases, some of the resources intended failed to reach the intended beneficiaries and ended up, instead, in the pockets of local leaders, politicians, or civil servants. In early 1997, just before the financial crisis hit, the Bappenas team worked with the World Bank to launch pilot community-driven development projects in 12 of IndonesiaÕs roughly 4,000 kecamatan (subdistricts, or local- level administrative units under a district, each of which comprised 20 or so desa, or villages). Breaking with past procedures, the new program transferred funds directly to villagers and then helped the villagers organize to hold each other accountable for the use of the resources.3 When the financial crisis and El Ni–o struck, later in 1997, scaling up the pilot projects acquired new urgency. As the value of the countryÕs currency, the rupiah, collapsed and business slowed, the plight of the poor intensified. Scott Guggenheim, who had headed the World BankÕs social development team in Indonesia and had advised the government since 1994, recalled: ÒThe government was scared out of its wits. Actually, so was the bank.Ó THE CHALLENGE Both IndonesiaÕs leaders and foreign donors were desperate for programs that would provide quick relief for poor communities far from the capital city of Jakarta. The Ministry of Home Affairs, which oversaw local government, sent Ayip Muflich, Prabawa Eko Soesanta, Eko Sri Haryanto, and Bito Wikantosa to help Bappenas design the new program. In addition to Guggenheim, the World Bank sent financial and operations advisers Unggul Suprayitno and Yogana Prasta, economists, and other technical staff to provide advice and assistance for planning and implementation. The question was whether the Bappenas team could throw a lifeline to rural Indonesians by expanding its pilot projects in the 12 kecamatan to tens of thousands of villages on thousands of islands spread across Southeast Asia and Oceania. But the programÕs designers had to address complex and thorny problems that could easily undermine the proposed scale- up. The first challenge was to ensure that project funds would reach the poor villages that needed the aid. The enemy was corruption, spread throughout a multitiered government system that flowed from the national level through provinces, regencies, cities, districts, subdistricts, and villages. In 1997, Transparency InternationalÕs Corruption Perceptions Index, which measures public perception of corruption, ranked Indonesia 46 out of the 52 countries evaluatedÑjust above Russia, Colombia, and Nigeria. Royat said, ÒWhen government officials sent money downward, they demanded their cut. And there were no sanctions.Ó Contractors also skimmed what they could. Further, rigid government rules discouraged innovation. Guggenheim commented, ÒThe regulations wouldnÕt let some parts of government give money [to communities] even if they wanted to.Ó Devolving more responsibilities to local governments, as Bappenas policy makers contemplated, would at least temporarily add to the confusion. There were potential risks at the village level, too. Indonesia had already tried two forms of community-led development, but with limited results. In 1993, the government had launched the Inpres Desa Tertinggal (IDT), called the Òprogram for left-behind villages.Ó The IDT program gave villages block grants of 20 million rupiah (US$10,000) to invest in development or income-generating activities, with minimal restrictions on spending. Johansen, the World Bank economist, had worked with the government to evaluate the program and adjust the design. Two years later, with World Bank support, the government created the Village Infrastructure Program, which provided much larger block grants of 150 million rupiah (US$68,000) to help poor villages build roads, bridges, and other infrastructure. In both instances, it became clear that only a handful of people typically benefited within the villages, and that the poor and other marginalized groups were shut out. Royat said: ÒOur experience with the IDT was good, but the money was still in the hands of the village elite. Common people did not have access to it.Ó A statistical evaluation of the IDT found that the program had no net impact on poverty reduction.4 An additional challenge involved lack of capacity at the local level. Some villages had a tradition of involving their residents in decision making, but others had little or no experience in the kind of group deliberations first needed to identify the projects that would help them most and then to manage such initiatives. For the rural poor in many parts of the country to benefit, the new program would have to empower people and ensure that they knew what to do with that power. Further, because of IndonesiaÕs extreme cultural and geographic diversities, the model would have to be simple enough and sufficiently flexible to work under a wide variety of circumstances. Finally, the design would have to provide a voice for women and other marginalized community members. To further complicate matters, a political crisis was beginning to unfold because the level of public frustration with corruption during a period of economic instability was already high. Removal of electricity subsidies that would help balance the budget proved to be the last straw. In May 1998, public demonstrations and riots shook the country and triggered the downfall of Suharto, who stepped down after more than three decades in power. Suharto transferred power to his vice president, J. B. Habibie, who promised electionsÑas well as government reformÑwithin a year. The risks were high. No one knew who the next president would be, what direction policy might take, or whether any government would hold together for long. For the reform team at Bappenas, however, the fast-changing situation created an opportunity to break with the past and move ahead on an unprecedented scale. FRAMING A RESPONSE In early 1998, as Indonesians marched in the streets and demanded change, the reform team huddled in conference rooms to improvise a way to realize twin goals: first, improving rural livelihoods, and second, giving villagers a stronger voice in the strengthening of institutions by holding government and their own traditional leaders responsible for using public resources well. The team included anthropologists and other people who had spent time on the ground in communities. The team members studied community-level planning and management of water and sanitation facilities in India and education experiments in Brazil that transferred resources directly to communities; they thought hard about their own experiences; and they brought an ethnographic perspective to the task at hand. One conclusion was clear: Scaling up the KDP pilot program would require a carefully constructed design that could accommodate diversity and promote accountability. The groupÕs final proposal applied several tacticsÑ social capital, competition, transparency, checks and balances, limited discretion, and procedural simplicityÑto help ensure that the countryÕs scarce resources would have a significant impact on the lives of the rural poor, as follows. Mobilize community social capital The Bappenas team remained committed to the community-driven development strategy with which they had experimented earlier, encouraged by research that had found that projects selected and managed by villages performed better than did government and nongovernmental-organization (NGO) projects.5 The model they proposed built on IndonesiaÕs tradition of reciprocity and sharing, or social capital. The principles of gotong royong (joint bearing of burdens) and musyawarah (consensus decision making) were foundations of communal life in many parts of the country. At that level there was an ethic of social accountability to neighbors that did not always exist in government or other organizations. Introduce competition However, earlier experiments with community-based programs had shown that that sense of obligation and shared responsibility often failed to carry through in the absence of incentives. Leaders and their friends frequently kept the benefits of government programs for themselves. The Bappenas team members reasoned that they might be able to reduce that kind of Òelite captureÓ by distributing project support to villages on a competitive basis. Unless there were high levels of collusion among leaders of separate villages, the requirement that local representatives vote to select the best projects and leave others unfunded would encourage villages to favor projects that would affect the lives of many people rather than just a few. If the villages in a kecamatan submitted 40 proposals per competition, as anticipated, a typical intervillage forum would be able to fund 10 to 15 depending on cost and the size of the block grant available. Projects not selected could be resubmitted for reconsideration the following year. There were other important reasons for choosing the kecamatan as the main administrative unit (see Textbox 1). Create checks and balances To make sure that the rural poor would truly benefit from government grants, the Bappenas team designed step-by-step procedures that would set up a system of checks and balances and boost transparency. In the first step, the government would announce a grant competition at a kecamatan intervillage forum. Each kecamatan would elect a subdistrict committee to manage the competitive process. The committee would include five representativesÑthree women and two menÑ from each village. The representatives would be nominated at a public meeting in a village and chosen by a show of hands or by paper ballot. Textbox 1: The Kecamatan Each of IndonesiaÕs 34 provinces was divided into districts, which were further divided into roughly 4,000 subdistricts or kecamatan, with about 20 villages in each. Historically, a kecamatan was a cohesive social unit. Haeruman said: ÒIt was a central place where you had markets, and it was a social unit as well. Kingdoms were at the kecamatan level, so people were used to that.Ó Even though district-level administrators did not accord kecamatan much importance, villagers still saw them as their main line of communication with the government. Each kecamatan also had a council composed of all village heads. Although the councils met only once or twice a year, the KDP could build on that existing structure. Further, kecamatan had no budget or contracting powers of their own and were therefore less likely to be influenced by vested interests. Organizing the KDP at the kecamatan level had one other important advantage: it made it possible to manage projects like roads that had spillover effects or that required coordination. Samsul Widodo, a member of the team at the directorate of rural infrastructure at Bappenas, said, ÒWe had built district roads; now we wanted get the network or connection going between villages.Ó The committee would create two teams: one to oversee financial transactionsÑcalled the Financial Management UnitÑand another to verify results. (The teams would receive further training in simple accounting, price comparison, costing, and engineering evaluations.) Next, each village would appoint two volunteersÑa man and a womanÑcalled village facilitators. The jobs of those volunteers were to organize project-proposal meetings in the hamlets, or neighborhoods, which made up their villages and to help convene an open meeting to discuss community needs and ideas. Communities were free to propose almost any kind of projectÑfor example, public works, educational investments, or income-generating activitiesÑalthough the Bappenas team decided to ban projects that served military or religious purposes, or involved illegal drugs manufacture or trade, or required land acquisition. The third step would involve a second village meeting, at which residents would select two priority projects, at least one of which had to come from womenÕs groups. (Later, villages would be allowed to add a third proposal, which had to come from womenÕs revolving loan fund associations.) The villagers also selected representatives to write down the proposals by using simple templates that included each projectÕs location, the expected number of beneficiaries from each project, and a rough cost estimate for each. The proposal writers received modest compensation for their efforts. The kecamatan verification team would visit and help the village complete the proposals. Then the intervillage forum would reconvene to discuss the merits of village proposals, aided by reports from the verification team. The forum would rank village proposals based on feasibility, impact, how well villages had maintained any previous projects, contributions from villagers, and other factors; and the forum would select winning proposals, seeking consensus. In the fifth and final step, special facilitatorsÑor private consultantsÑwould help communities whose proposals received high marks prepare detailed design plans and cost estimates. Streamline and add financial checks The design had several features that protected project funds from unauthorized diversion. After review by a district engineer from the local government, winning villages had to prepare grant agreements for the proposals selected. Each agreement included details of the project design, a budget, a map, and the amount of the community contribution, and the information had to be posted in public spaces in the community. The kecamatan financial-management unit then submitted to its local government a request for payment. The government forwarded the request to the national treasury office. The treasury transferred money to a special bank account that each community had set up at a local commercial bank. To ensure accountability, the program rules required cooperating banks to collect the signatures of a kecamatan-level facilitator, a village representative, and the head of an elected kecamatan committee before opening an account or releasing payments. Representatives could not withdraw money for private use. Grant sizes ranged from 350 million rupiah to 1 billion rupiah (US$40,000 to US$114,000), depending on population density, poverty incidence, and remoteness. Villages with populations of 5,000 or fewer received block grants of 500 million rupiah (US$55,000) or less. For populations of 5, 001 to 20,000, grants ranged from 500 million rupiah to 1 billion rupiah (US$57,000 to US$114,000). Villages with 20,001 people or more received the maximum amount. (Later, the amounts of block grants rose to 1.5 billion rupiah [US$171,000].) Most grants were roughly equivalent to the cost of building an average farm-to-market road. The plan called for disbursement of the funds in three installments: two of 40% each and one of 20%. The program released the first installment based on project plans; the second installment was contingent on submission of an expenditure and progress report; the final installment was released on receipt of a completion certificate from the district engineer. Build transparency to aid delivery Once planning was complete and the first installment of the grant was in the communityÕs bank account, residents elected a three-person implementation team to contract with suppliers, hire poor villagers for some of the work, purchase supplies, and chart progress. Procurement of supplies exceeding 15 million rupiah (US$1,500) followed simplified competitive- bidding procedures. The teams reported on project progress at village accountability meetings held at the halfway mark and at the end of the project period. District engineers had to certify completion of the projects before the Treasury released the final 20% installment. Following completion of the projects, village implementation teams presented results to residents, who then would have to select operation and maintenance teams to sustain the projects. Target the neediest areas KDP planners wanted the nationwide program rollout to target the poorest kecamatan as well as the most marginalized citizens, especially women. They created an initial list of 1,500 kecamatan on the basis of village census data, surveys of spending and infrastructure, geographic location, population size, and number of people below the official poverty line. And they contacted district local governments to confirm that the choices they made corresponded to the poorest subdistricts or those in need of basic infrastructure.6 Ensure inclusive participation Because the targeted villages and groups were those least likely to have the skills to navigate the grant process, the Bappenas team included coaches and quotas in their design. Throughout the project cycle, a network of Indonesian consultants, known as facilitators, aided the villages. The facilitatorsÕ job was to encourage participation, help form representative committees, guide the discussion of community needs, train people in project and budget management, and ensure projectsÕ technical quality. The team also set rules designed to guarantee inclusion of marginalized groups in addition to the requirement that at least one of every two proposals had to come from womenÕs groups (later raised to two of every three). For instance, because traditional gender roles sometimes discouraged participation by women at community meetings, the team decided to require that women be allowed to meet separately and devise their own proposals beforehand. If a husband would not allow his wife to attend a meeting in another area, facilitators would hold the meeting near the coupleÕs house. One out of every three village representatives at each level had to be female, as did 30% of the participants in village meetingsÑand facilitators could adjust the timing of meetings so that larger numbers of women could attend. Royat said: ÒOur role was to convince people and community groups that it is your money. You have to watch this and have the right to ask where even one rupiah goes and to whom.Ó The program adopted the slogan ÒPlanned by communities, executed by communities, controlled by communities.Ó GET TING DOWN TO WORK With the design parameters established, members of the KDP team moved to implement their ideas. Specifically, they had to determine how the program would fit into the national government, how to organize financial support, how to train facilitators, and how to ensure active involvement by the local citizens, who would have bottom-line responsibility for projects in their communities. Creating an administrative home Initially, Bappenas took responsibility for rolling out the program. The agency had hosted the KDPÕs original pilot programs and had played the lead role in designing the expanded effort. But because BappenasÕs mandate involved planningÑwhereas other ministries oversaw local governanceÑthe agency eventually had to transfer leadership of the KDP. After two yearsÑin 2000Ñthe Community Development Agency (Pemberdayaan Masyarakat Desa, or PMD) at the Ministry of Home Affairs took over. Erman Rahman, who was at Bappenas at the time, said: ÒAs a planning agency, we were strongly criticized for directly implementing projects. It was clear to us that we couldnÕt implement the program ourselves. The most appropriate agency was the village-and-community-development directorate general [again, the PMD] within the Ministry of Home Affairs. Although the PMD had only a limited amount of experience in implementing large-scale and donor-funded projects, it was the department legally responsible for village development.Ó Bappenas retained a planning and supervisory role, however. Capacity constraints led both Bappenas and the Ministry of Home Affairs to vest daily supervision in a group of about 50 Indonesian consultants called the National Management Consultant team. Haeruman, the agencyÕs deputy director of regional development, said: ÒIn government, you always hear good things. I said I donÕt need that kind of reporting . . . I need to know the problems. So this group was hired for that [purpose].Ó The group consisted of engineers, auditors, microfinance experts, specialists who could manage information systems, and people who provided human resources support and monitored complaints. Samsul Widodo at the directorate of rural infrastructure at Bappenas said, ÒTheir role was to assist the government, provide oversight, monitor implementation in the field, and train facilitators.Ó The consulting teamÕs provincial and regional coordinators spent half their time in the field, where they monitored and recorded activities; and they submitted monthly reports to Bappenas (and, later, the Ministry of Home Affairs). The consultantsÕ monthly reports formed a critical accountability measure for the team, and accuracy and timeliness were crucial. The World Bank continued to provide technical support, focused primarily on learning and external oversight. Planning was BappenasÕs responsibility; the Ministry of Home Affairs (the PMD) executed; and the Ministry of Finance organized the resources, but there was no initial provision for external oversight. The Indonesian consultants were not in a strong position to take on corruption, even though their tasks included reporting potential incidents. The World Bank team played that role. Financing During the first phase, IndonesiaÕs government had funded 90% of the program through World Bank credits and loans.7 Starting in 2002, the government increased its contribution to 22% in the forms of budget allocations and local government support. Community contributions and grants as well as trust funds from other donors added to KDP resources later on. The implementation would cost US$700 million for its first two phases (1998Ð2006).8 Decision makers relied on the World BankÕs continued involvement for another reason. Royat said that having to account to the World Bank for resource use created a Òcheck on the politicians,Ó adding that ÒWhenever we had conflict, we referred to bank regulations. This was also a protection for us.Ó The World BankÐfunded loan went into a special Bank of Indonesia account earmarked for the KDP. The Ministry of Finance oversaw the account. Making it work: Local issues The KDP team anticipated the communitiesÕ need for facilitation support, especially in the programÕs early phases. The plan called for gradually building villagesÕ capacities to implement projects and monitor progress. The National Management Consultant team worked with private Indonesian consulting companies to recruit and supervise kecamatan facilitators. Allocated up to 3% of total project funds, the companies were responsible for paying facilitatorsÕ salaries and transport costs and for providing operations and technical support. There were two types of kecamatan facilitatorsÑtechnical and socialÑand together these men and women played an important role in making sure the system worked at the local level. The facilitators built awareness of the program, recruited two village counterparts, mobilized and organized the communities they served, helped communities plan and implement projects, provided technical assistance, identified and fixed problems in the field, and reported progress to central management. When some of the early projects fell short of quality standards, the ministry took additional steps to train and oversee engineers at the kecamatan level, beginning in 2000. Before working in communities, facilitators underwent about a month of instruction. Initially, the program hired CAREÑan international NGOÑto handle the training. Later, as experience grew, senior facilitators started training new recruits. The training focused on KDP principles, ways to encourage discussion and handle complaints or problems, and specific tasks. Facilitators could also consult a simple operations manual if they needed help regarding policies and procedures. A big part of the job was to elicit active participation in a country where such behavior had often been discouraged. Sri Kuntari, one of the KDPÕs early facilitators based in West Java, said: ÒIt sounds very simple, but it wasnÕt that easy at the time. Because of the authoritarian government, people werenÕt used to speaking, and all structures were co-opted by the elite.Ó The management team knew that elite capture would be difficult to overcome, largely because village heads had sweeping authority. Kuntari said: ÒThe village head is a very powerful person. He controls everything. He dominates the discussion in village meetings and has the power to stop a meeting and tell everyone to go home. . . . And he can instruct people not to come to meetings.Ó Sentot Satria, a former consultant team leader for North Sumatra, said: ÒFacilitators, especially the young men and women who were leading meetings for the first time, trembled and didnÕt know how to deal with a seasoned village head. It is not easy, even until now.Ó KDP leaders recognized that they would be unable to remove such roadblocks easily, so they tried various workarounds that produced varying degrees of success. For instance, trainers coached new recruits about how to do their jobs without offending their village leaders. In remote areas where social structures varied significantly, facilitators found it difficult to gather together people from different social classes. In some instances, facilitators approached community leaders in mosques and churches or in traditional meetings, using religious uniformity to access social diversity. Facilitators also had to persuade villagers that their involvement in the program was a good investment of their time and effort. Called upon to devote long hours to learning about the program and then even far more time implementing and monitoring projects, some community members sometimes complained that the program was too much of a burden. Enurlaela Hasanah, a facilitator in charge of organizing and mobilizing rural communities and later part of the national management team, described the problem: ÒSometimes they said, ÔWhy do we have to be busy like this? This is a government project, and they should do it. Why are we doing these meetings at the village and hamlet levels? That was one of the constraints in the beginning.Ó Similarly, some village volunteers complained that they should be paid for their work. Hasanah said: ÒWe say we will reimburse only the cost, which is the operational cost. ItÕs not a real salary. For instance, village facilitators get only about US$15 to cover transportation.Ó The constraints continued to be a problem. Because of limited financial incentives, some village facilitators and volunteers either did not do their jobs well or stopped contributing. But in most cases, communities cooperated because of the promise of new or improved services in their villages. KDP empowered them to take charge instead of sending requests and waiting for responses through official government channels. In addition to winning participation, building trust, and helping guide the proposal development process and project management, the facilitators acted as monitors. Facilitators would keep track of project funding at each step, along with community representatives of the Financial Management Unit. Tracking progress Provincial facilitators received monthly reports from districts and subdistricts and submitted them to Bappenas (and, later, the Ministry of Home Affairs). They relied on information passed upward from hamlets to villages and then to kecamatan. Kecamatan facilitators logged data on the number of people who participated in meetings, the ratio of male to female attendees, the groups that had proposed the projects, project progress, problems, and actions taken. The reports often included pictures documenting both progress and the time it took to complete projects, so that the management team in Jakarta could track projects nationally. The central management team of consultants collected the data and provided the ministry with a monthly review of KDP progress in all provinces along with details on how much money had been disbursed. In 1999, KDP moved to bolster its tracking and accountability mechanisms by launching a system for soliciting and handling public grievances, tips, and complaints. The ministry and the management team provided field staff with guidelines for citizens on how to file complaints, whom to contact, and the criteria to use. People could lodge complaints with their facilitators either in person or in writing. Later, they sent complaints via cell phone text messages to a hotline. The management team worked closely with the ministry to keep track of complaints and their resolution. Hasanah said: ÒTo strengthen complaints handling, there was a complaints monitor at the provincial level and also in Jakarta. So, we would get reports from provincial-based monitors, such as NGOs and reporters. I would put these issues in reports in a matrix and follow up.Ó Hasanah divided all the complaints from the field into four common types: violation of regulation; corruption or misuse of funds; interference by local officials; and Òcatastrophe not under the control of the community,Ó such as an earthquake. Another category tracked conflicts between consultants. Hasanah compiled the information, made recommendations, and sent them to the regional coordinator. The coordinator would visit the village to discuss and monitor the problems. The team followed up with the regional coordinator every two weeks. Although the alliances did not enjoy long- term success, KDP recruited NGOs and journalists to monitor the KDP. Journalists and NGO representatives visited villages to gather information about community participation and then reported on progress in newspapers and via reports. Managing corruption risk KDP planners built in sanctions to discourage misuse of funds, including suspension of funding in cases of fraud or corruption. Guggenheim said: ÒIf a village committed an infraction, we could suspend the entire kecamatan. If it was the kecamatan, we could suspend the kabupaten [district]. And for two infractions in a kabupaten, we could suspend the province. That puts a lot of social pressure on the [administrative] unit by all their neighbors to say that you better fix this.Ó Because a fair number of corruption cases involved officials, warnings were usually issued to the provinces, which could then warn the units below them. However, the program did allow for government officials and communities to make amends for breaking program rules. Guggenheim continued: ÒOn a somewhat positive side, we could allow them informally . . . to pay restitution. So, I donÕt need the money back as long as you finish the bridge we officially gave money for.Ó The KDP publicized corruption cases and the enforcement measures taken against the offenders. When government officials turned out to be the source of the problem, communities could bring legal cases with the help of NGOs. During 2000Ð2001, three years into the programÑ5% of the KDPÕs technical staff was replaced for corrupt practices, complicity, or failure to report corruption.9 By 2002, at least 12 government officials had incurred fines or received prison terms for misuse of KDP funds.10 Provinces that failed to act on corruption cases knew they could lose program funding if they did not comply. Program leaders learned from experience and adapted procedures, as the program developed, in order to reduce opportunities for corruption. Royat said: ÒThere were many brokers in local government. They knew the amount of the budget. They could go directly to the village and say, ÔBecause of me, you will get the money, so my fee is 20%; this is what you owe me.Õ So we changed the strategy. We would notify first the community [that it had received a grant] and then, a day later, send the letter to the bhupati [district head].Ó Some KDP projects were especially vulnerable to corruption or capture. For example, when rural agricultural economies were reeling under the twin impacts of the financial crisis and El Ni–o, the early phases of the KDP included a microfinance component that provided low-cost loans for the poorest villagers and poorest communities. Guggenheim explained: ÒIt was because of the fear of starvation. From the beginning, no one thought it would be sustainable . . . but there was no [other] way to get money into villages, with production collapsing all around because of the weather.Ó Although the loans succeeded in getting money to poor communities, repayment rates were very low; on average, 40% of loans were repaid.11 Vested and powerful groups captured some of the microfinance projects for themselves. Although the KDP leaders could not win support for elimination of microfinance when the problems arose, the program adapted: in later years, it funded only womenÕs groups, which had a higher likelihood of repayment. Soesanta noted, ÒWe were surprised the repayment increased to more than 98%.Ó OVERCOMING OBSTACLES The program faced three challenges that required it to adapt its design further. The first challenge involved the relationship with local government. The second had to do with convincing facilitators to pass along bad newsÑ failures as well as successes. The third challenge was the sheer variety of situations that a country as diverse as Indonesia presented. Building the relationship with local government One of the plannersÕ long-term aims was to transfer the KDP to district governments and to empower villagers to hold those institutions accountable for performance. Agung Widodo, part of the Bappenas team responsible for selecting the kecamatan, said, ÒWe thought that by the end of three years, the local government would provide for it and run it.Ó But discord sometimes marred the relationship. For instance, civil servants, unable to take a cut, would tell villages to hand over the money to them so that they could complete the projects. Others resented that the program had bypassed them. Agung Widodo said: ÒIt is a trade-off. If you give money directly to local governments, then maybe the money that reaches villages would be less because of corruption. But if you donÕt involve them, then itÕs a problem as well because they will take over the program in the future. We need to foster a sense of belonging to the program for the local governments.Ó In some instances, villagers reached out to local officials on their own. Hasanah recalled: ÒOne village was upset that it hadnÕt gotten the money at all. They complained to the subdistrict head. That is why it was important to involve local officials.Ó After the first phase, KDP leaders began to involve local government in new ways. Facilitators requested subdistrict local government representatives to chair meetings with the heads of villages. Starting in 2002, the KDP also allowed districts to add kecamatan for the program, provided they put up 80% of the block grant from the district budget. During the KDPÕs second phase (2002Ð 006), about 40% of districts provided supplemental grants for communities. By the KDPÕs third phase (2006Ð08), all districts provided contributions from their own budgets. Community contributions averaged 17%, with wide provincial variations.12 By 2004, all participating districts were providing matching grants on a scale calibrated to their poverty ranks, especially for subdistricts they wished to keep in the program for more than three years. But there would still be much to do to implement the planned transfer of management to local governments even as the KDP would get folded into a larger program in 2007. Embracing bad news KDP planners asked facilitators to be open about problems and report them. Hasanah said: ÒPeople thought the government would want to hear only good things. So they didnÕt tell about bad things. We made sure to tell the facilitators it is no problem to have problems. ÔWe will evaluate your performanceÑbut only in handling the problem, not in reporting it.ÕÓ Facilitators eventually became the single biggest channel of information about corruption. (From time to time, though, facilitators were implicated in corruption cases. From 1998 to 2002, 13% of reported KDP-alleged corruption cases were against facilitators.13 ) The ministry staff in Jakarta encouraged field staff to report delicate situations. Former consultant team leader Satria said: ÒIf they have problems with village heads or if someone threatens them, the instruction is simple. They have to get away from the situation and report directly to the local government or the directorate [general] of village and community empowerment at the Ministry of Home Affairs. And the project management would tell the kecamatan that, ÔYou wonÕt get money if people donÕt follow rules.ÕÓ The relationships between facilitators and Jakarta-based staff were not seamless, however. Hasanah said: ÒSome people didnÕt understand that this is a community-driven development project. . . . It is an open menu, and people can pick what they want.Ó But people in Jakarta would say, ÒIt is not the right project.Ó One model doesnÕt fit all The planners made small changes so as to account for certain kinds of regional differences. Population densities and distances varied and sometimes made the intervillage forums hard to travel to. Guggenheim said: ÒIn the isolated areas, a kecamatan is too big. It could take two weeks to walk from the village to the kecamatan [center] in Papua.Ó In those areas, facilitators could arrange to hold meetings in villages instead of holding one big meeting at the subdistrict level. Representatives from those meetings could then attend the subdistrict meeting. Density and distance affected cost projections as well. Transportation and travel costs were higher in such regions as East Indonesia. KDP planners provided for higher facilitator salaries and higher expenses in those areas. Planners also acknowledged the need for design adjustment in conflict-prone areas, such as Aceh, East Timor (which became independent in 2002), and Papua. Many areas also had their own local customs and traditions. Citing an example, Satria said: ÒNias and Aceh have basically the same social conditions and traditional structures. Those provinces and villages are older than this country itself. They have their customary laws and mechanisms to solve problems and deal with conflict. But the Suharto government made them administrative villages. So you still have a customary way of living, along with a uniform administrative setup.Ó Under the KDP, villages had to compete for funding at the subdistrict level. Agung Widodo said: ÒWe worried that such competition could lead to further conflicts. So we tried to soften competition in certain areas. We tried to see that all participating villages got their proposals funded, even though proposals were developed in a competitive manner.Ó And to minimize the exclusion of minorities, facilitators also made sure all communities were represented. For conflict-ridden areas, the KDP added extra facilitators and gave them more time to work with communities and put projects in place. The program progressed more slowly in those areas and usually took more than one fiscal year to complete. Sometimes the KDP was the only government program that could operate amid low-intensity conflict. In Aceh, Satria said, ÒWhen the rebels found out that the planning and implementation [of projects] were by the communities themselves,Ó they let the villagers continue with the relationship. In 2004, when a devastating tsunami and earthquake hit the area, the KDP was the only donor-funded infrastructure project in the region. With the help of facilitators, the government quickly expanded the program to provide relief. Satria said: ÒThe rehabilitation and reconstruction used the structure of facilitators and villagers who are part of the program. Whether victims or not, they could be part of the recovery. The nice thing about their involvement in the recovery process was that they were part of the healing process. And, following the 2005 Helsinki peace agreement [between rebels and the government], the government also used KDPÕs existing platform to provide additional block grants to AcehÕs post-conflict areas.Ó Even with extra facilitators in the region, however, there was too much for each to do. As time went by, the Aceh facilitators began to skip steps or omit responsibilities in order to cope with the demands on their time. The program began to diverge from the KDP model, and some of the problems that had afflicted earlier versions of community-driven development reappeared. Adapting to conflict areas and to some of the varied cultural terrain would remain a problem despite some signal successes at the time of the tsunami. ASSESSING RESULTS If geographic breadth is a measure, the KDP achieved a good deal. The rollout took place in several phases. The original 1997 pilot consisted of 25 villages in 20 subdistricts. The first phase of the scale-up, from 1998 to 2002, covered 15,481 villages in 986 subdistricts, distributed across 130 districts and 22 provinces. By 2005, the program had expanded to 25,651 villages in 1,520 subdistricts, 239 districts, and 30 provinces. By 2006, it was reaching 34,233 villagesÑor nearly half of all villages in Indonesia. A third phase continued to 2008 and brought the total project cost to US$1.3 billion. The original idea was to rotate out of an area after three years, during phase two, which began in 2002, but the government decided to continue operating in the first set of subdistricts while also adding new ones. From 1998 to 2006, KDPs helped villages build or rehabilitate more than 65,500 kilometers of roads, 9,000 bridges, 11,000 irrigation systems, 28,300 drinking-water systems, 17,500 sanitation facilities, 6,950 schools, 120,000 scholarships for poor students, and 5,700 health posts. By 2007, more than 6.1 million villagers had earned direct short-term employment; nearly half of them (45%) were women or rural poor. On balance, a variety of metrics suggested that the program had achieved significant success. Households in KDP areas moved out of poverty in greater proportions than did households in non-KDP areas.14 Studies found that most infrastructure projects were high quality, carried high economic rates of return, and cost less than if the work had been handled through traditional government channels.15 The roads that connected villages to markets and the irrigation systems that opened up new areas to cultivation produced some of the largest economic impacts. The program helped engage some of IndonesiaÕs poorest people. On average, about 50% of all participants in meetings represented the poorest groups, as did 70% of workers employed in KDP projects.16 Citizen participation levels were higher in KDP project meetings than in other village decisions, which were often limited to the core elite. On average, 60 to 100 people joined KDP meetings, and 200 to 300 villagers gathered at subdistrict meetings.17 Women accounted for 26 to 45% of participants. However, evaluations showed that certain marginalized groups were largely excluded. For example, widows did not participate in village-level planning meetings. Most villages opted for infrastructure projects, often a major need in communities. Kuntari said: ÒThey thought of produce and access to markets and schools and water. Later on in the program, they started asking for schools, bridges, and clinics. But in the beginning, it was roads and water.Ó Designers noticed that most of the communities ended up choosing infrastructure projects because such projects were the easiest to propose and implement and because such projects imposed lower long-term costs than health or education programs did. Villagers often suspected that ministries would not put teachers or nurses into the schools or clinics their communities built. Over time, the planners encouraged diversification of projects, but the public works projects often had important social impacts that standard metrics did not take into account. Guggenheim pointed out, for instance, that roads Òmake a huge difference to people when it comes to maternal mortality and to education.Ó The extent of corruption was difficult to measure, and even those closely involved with the programÕs design suggested that wrongdoing was underreported. Still, data indicated that the KDP had done a decent job of managing the problem. Audits by independent agencies found that irregularities in handling money amounted to less than 1% of the total budgetÑa percentage some questioned, including the World Bank project officerÑand the KDP was able to recover 40% of the missing funds. By 2006, corruption complaints were amounting to US$650,000Ña rate of 0.4% of US$170 million disbursed.18 About 45% of the complaints were resolved.19 A 2004 community survey reported that 95.2% of respondents said they felt there had been no malfeasance during KDP implementation.20 Not all evaluations found community- driven efforts the most effective in reducing corruption, however. A donor-funded randomized, controlled trial conducted in 2003Ð04 found that government audits reduced corruption more effectively than did efforts to expand participation in community accountability meetings.21 The trial found that mandatory audits caused corruption to drop by eight percentage points to 19.2% and that special invitations to accountability meetings and distribution of comment forms had much less effect (See Textbox 2). The KDP suffered from certain significant drawbacks. Central-government management of the program was not always consistent. For instance, there were sometimes long delays in disbursing funds, and the program had problems in tackling technically difficult activities or activities that involved recurrent costs, such as Textbox 2: Accountability Meetings, Audits, and Corruption in Road Projects A study conducted by Benjamin Olken and researchers at the Massachusetts Institute of TechnologyÕs Poverty Action Lab asked whether expanding participation in accountability meetings or mandatory government audits had greater ability to reduce corruption. The study focused on road projects and measured corruption by comparing the estimated cost of the materials usedÑbased on core samples and other measurementsÑwith the reported cost. The study randomly assigned 608 villages with road projects to one of six groups. The researchers told one set of villages that the government would audit their projects and that it could bring criminal charges against people who diverted funds. In a second set of villages, residents received invitations to the accountability meetings at which the road project managers reported on their work. In these villages, the risk of audit remained at 4%Ñthe normal level. In a third set of villages, an anonymous consent form accompanied the invitations, and the risk of audit was 4%. The fourth and fifth sets combined the participatory elements with the mandatory audit. The final set of villages received no special treatment and served as a control group. The study found that mandatory government audit was most successful in reducing corruption, but the increased community monitoring had little average impact. Community accountability meetings were most effective in villages without histories of elite capture. And there was some evidence that when residents both used the comment form and received invitations to the meetings, the residentsÕ village was more likely to take serious action to prevent corruption. paying teachersÕ salaries or providing healthcare supplies for clinics. Other countries learned from the KDP model. Timor-LesteÕs Community Empowerment Project copied the KDP, and the PhilippinesÕ Kalahi program drew from the KDP experience. In Afghanistan, the National Solidarity Programme took a similar approach, directly transferring money to communities to implement projects that citizens chose. REFLECTIONS The KDPÕs effectiveness in improving the lives of some of IndonesiaÕs poorest rural residents stemmed partly from seven elements the program shared with many other successfully implemented reforms. First, public servants had a clear vision of what they wanted to do and seized an opportunityÑthe reform momentÑto introduce a new way of doing things. The Asian financial crisis and the El Ni–o onslaught imperiled millions of rural poor and created political cover for a program that deliberately sidestepped the vested interests that had usually and historically exacted significant payments for their participation; the economic crisis thereby helped the program ease the grip of patronage politics. Second, the team that designed the program and oversaw its implementation enjoyed the support of senior levels of government as well as donor alignment. Although he was not generally considered a proponent of the people, President Suharto actively encouraged community-driven programs, as did his successors. And the KDP corresponded closely (1) with new ideas at the World Bank and (2) with the bankÕs conviction that dealing with IndonesiaÕs twin crises required a social safety net for the countryÕs poorest people. Third, KDP planners designed and adapted the program to work within both formal government structures and the informalÑbut no less significantÑcustoms, norms, and traditions that dictated lines of authority at the local level. Instead of attempting to impose a new system for deliberations, for instance, the program exploited existing social mores and principles that embodied accountability, volunteerism, and reciprocity. Fourth, the programÕs clever and simple design won praise from both insiders and observers. Meaningful incentives, effective checks and balances, transparency, and competition played important roles in implementing procedural standards, limiting corruption, and building acceptance and participation. Indeed, the initial time and effort spent on the programÕs design paid dividends by reducing the need for rigorous enforcement of complex rules. Fifth, the programÕs field presenceÑin the form of its extensive network of facilitatorsÑ contributed importantly to success. Operations adviser Yogana Prasta said: ÒThe success of the program depends very much on the facilitators. They are the KDP troops that help communities in planning and empower them.Ó Relying on facilitators rather than officials subject to the countryÕs civil service rules also lent greater flexibility in implementation of the program. Flexibility was a sixth important element of the reform. The team was willing to adapt the system to villagersÑand to local conditionsÑ rather than forcing villagers to adapt to the system. For example, the team adjusted the length of the project cycle, the roles of representatives, and the number and job descriptions of facilitators in order to accommodate the lower population densities in some areas. Finally, from the beginning, senior managers set a high premium on monitoring and learning. With the World BankÕs help, they commissioned research to assess the impact of changes in procedures. They developed an information system that showed the status of the program throughout the country at any given time, including the completion rates of projects in different subdistricts, types of projects, and numbers of women who participated in community meetings. The information system helped decision makers spot what worked and what did not and helped them make changes, assess results, and adapt further. Starting in 2006, Pres. Susilo Bambang YudhoyonoÕs new government built upon the KDPÕs progress and expanded the program to the rest of the country. The new version became the National Program for Community Empowerment (Program Nasional Pemberdayaan Masyarakat, or PNPM). In December 2013, Indonesia wrote the KDP/PNPM system into law. Timeline August 1997 Asian f inancial crisis 1997 KDP pilot in 25 villages May 1998 fall of SuhartoÕs New Order regime 1998 Ð 2002 KDP I 2002 Ð 2006 KDP II 2006 Ð 2008 KDP II 2006 President Susilo Bambang Yudhoyono announces plans to expand the Kecamatan Development Program and combine it with other projects under the umbrella of PNPM Mandiri. 2007 KDP is folded into the PNPM. President launches the PNPM in April 2007. References 1 Scott Guggenheim, ÒCrises and Contradictions: Understanding the Origins of a Community Development Project in Indonesia,Ó Paper presented to the Yale University Agrarian Studies Program, no date, 8. Accessed at http: //www. yale. edu/agrarianstudies/colloqpapers/22guggenheim. pdf. 2 http: //www. seadiproject. com/0_repository/SEADI- 13R140%20Evolution%20of%20Inequality%20in%20Indonesia. pdf. 3 ÒPak Suharto agreed to bring everyone together under poverty alleviation. So, we set targetsÑlike poverty should be reduced by X%. We [regional boards] worked with sectoral ministries to implement [projects]. But it was elitist still, and we wanted development to happen at the community level.Ó Sujana Royat, director of urban land and special programs, made this comment. 4 John Molyneaux and Paul Gertler, ÒApplying a Simple Method to Measure Program Impact: An Evaluation of a Large-Scale Micro-credit/Poverty Alleviation Program in Indonesia,Ó 1999. 5 In 1996, the World Bank undertook local-level-institutions studies, which comprised a series of case studies on villages in three countries, including Indonesia. 6 Scott Guggenheim, Tatag Wiranto, Yogana Prasta, and Susan Wong, ÒIndonesiaÕs Kecamatan Development Program: A Large-Scale Use of Community Development to Reduce Poverty,Ó 2004, 6. 7 Ibid., 25. 8 Susan Wong, ÒIndonesia Kecamatan Development Program: Building a Monitoring and Evaluation System for a Large-Scale Community-Driven Development Program,Ó 2003, 1. 9 IndonesiaÕs Kecamatan Development Program: An Effective Project-Level Anticorruption Strategy,Ó 2010, 5. 10 Scott Guggenheim, ÒCrises and Contradictions: Understanding the Origins of a Community Development Project in Indonesia,Ó 7. 11 Susan Wong, ÒIndonesia Kecamatan Development Program,Ó 11. 12 Scott Guggenheim, Tatag Wiranto, Yogana Prasta, and Susan Wong, ÒIndonesiaÕs Kecamatan Development Program, 11. 13 Andrea Woodhouse, ÒVillage Corruption in Indonesia: Fighting Corruption in the World BankÕs Kecamatan Development Program,Ó 2002, 17. 14Anthony Torrens, ÒEconomic Impact Analysis of Kecamatan Development Program Infrastructure Projects,Ó January 2005, 13. 15 Ibid., 2. 16 Scott Guggenheim, Tatag Wiranto, Yogana Prasta, and Susan Wong, ÒIndonesiaÕs Kecamatan Development Program, 11. 17 Ibid., 32. 18 Scott Guggenheim et al., ÒIndonesiaÕs Kecamatan Development Program: An Effective Project-Level Anticorruption Strategy,Ó 2010, 4. 19 Ibid. 20 Ibid., 5. 21 Benjamin Olken, ÒMonitoring Corruption.Ó 15 Innovations for Successful Societies makes its case studies and other publications available to all at no cost, under the guidelines of the Terms of Use listed below. The ISS Web repository is intended to serve as an idea bank, enabling practitioners and scholars to evaluate the pros and cons of different reform strategies and weigh the effects of context. ISS welcomes readersÕ feedback, including suggestions of additional topics and questions to be considered, corrections, and how case studies are being used: iss @princeton. edu. Terms of Use Before using any materials downloaded from the Innovations for Successful Societies website, users must read and accept the terms on which we make these items available. The terms constitute a legal agreement between any person who seeks to use information available at www.princeton. edu/successfulsocieties and from Princeton University. In downloading or otherwise employing this information, users indicate that a. They understand that the materials downloaded from the website are protected under United States Copyright Law (Title 17, United States Code). b. They will use the material only for educational, scholarly, and other noncommercial purposes. c. They will not sell, transfer, assign, license, lease, or otherwise convey any portion of this information to any third party. (Republication or display on a third partyÕs website requires the express written permission of the Princeton University Innovations for Successful Societies program or the Princeton University Library.) d. In all publications, presentations or other communications that incorporate or otherwise rely on information from this archive, they will acknowledge that such information was obtained through the Innovations for Successful Societies website. A suggested citation format is as follows: [Document author if listed], [Document title], Innovations for Successful Societies, Princeton University, accessed at http://www.princeton.edu/successfulsocieties on [date accessed on Web] e. They understand that the quotes used in the case study reflect the intervieweesÕ personal points of view. Although all efforts have been made to ensure the accuracy of the information collected, Princeton University does not warrant the accuracy, completeness, timeliness, or other characteristics of any material available online. f. They acknowledge that the content and/or format of the archive and the site may be revised, updated or otherwise modified from time to time. g. They accept that access to and use of the archive are at their own risk. They shall not hold Princeton University liable for any loss or damages resulting from the use of information in the archive. Princeton University assumes no liability for any errors or omissions with respect to the functioning of the archive. Innovations for Successful Societies (ISS) is a joint program of Princeton UniversityÕs Woodrow Wilson School of Public & International Affairs and the Bobst Center for Peace & Justice. The Woodrow Wilson School prepares students for careers in public service and supports scholarly research on policy and governance. The mission of the Bobst Center for Peace & Justice is to advance the cause of peace and justice through mutual understanding and respect for all ethnic traditions and religious faiths, both within countries and across national borders. ISS is a joint program of the Woodrow Wilson School of Public and International Affairs and the Bobst Center for Peace and Justice: www.princeton.edu/successfulsocieties