GOVERNANCE and ANTICORRUPTION

Transcription

1 GOVERNANCE and ANTICORRUPTION A S S E S S I N G R E S U L T S I N P R O J E C T S This study is one in a series of papers that describe project-level approaches to governance and October corruption 2010issues, Case Study 1 KDP, a community-driven development project implemented in Indonesia since 1998, was the first World Bank-financed project ever to introduce an anticorruption action plan to address the corruption related risks that were identified by specific diagnostic studies. Several innovative mitigation measures (such as multilayered monitoring, complaint-handling mechanisms, and audits) were included, leading to tangible results that were documented by robust evaluations and research studies in the field: villagers perception of a low level of corruption, reduced percentage of missing funds, removal of abusive village and subdistrict heads, recovery of missing funds, and repair of poor-quality infrastructure. Indonesia s Kecamatan Development Program: An Effective Project-level Anticorruption Strategy The Project The Kecamatan Development Program (KDP) was launched in 1998 to (a) alleviate poverty by raising rural incomes, (b) strengthen local government and community institutions, and (c) improve governance. Since then it has covered 30 provinces, 260 districts, 1,983 subdistricts, and 34,200 villages, and the total loan, grant, trust fund, and Government financing has amounted to US$1.3 billion. The program empowers communities by placing the planning and decision-making process directly in the hands of villagers, who come together to plan local projects and submit proposals for consideration at an intervillage forum. The selected projects are then funded through block grant transfers, and villagers themselves implement subprojects and manage the funds. KDP is now being followed up by the National Community Empowerment Program, which builds on the success of its predecessor. Diagnosis of the Problem Corruption risks are very common in development projects in Indonesia. At the village level, corruption can occur in several ways: for example, there can be collusion between implementation teams and suppliers; suppliers can inflate prices or quantities; wage payments can be manipulated; and corrupt officials can bill the project for labor that is actually unpaid. The World Bank task team undertook several assessments indepth ethnographic interviews, field experience reviews, and analysis of the incentive structure throughout the project cycle to determine actor interest, motivations, and constraints related to corruption. They found that corruption was mainly a problem of incentives and could be fought by changing the costs and benefits attached to corrupt behavior. This study is one in a series of papers that describe project-level approaches to governance and corruption issues, identify their results, and draw lessons that other task teams may be able to use.

2 G O V E R N A N C E a n d A N T I C O R R U P T I O N TABLE 1 Mitigation and Results at a Glance Conditions at entry Corruption-related risks GAC mitigation measures Results Strong history of community engagement Government commitment Rural areas Fiduciary risks Specifi c risks during project cycle: proposal preparation (formation of false borrower groups for small loans); release of funds (collusion among bank account signatories to embezzle funds); and implementation (collusion and corruption in procurement) Highly dispersed communities Corruption diagnostics Anticorruption strategy Multilayered monitoring with social oversight Complaint-handling mechanism with sanctions Block grants Transparency measures Low level of corruption Few funds irregularities (1%) 40% of missing funds recovered 976 reported corruption cases as of 2006 KDP village infrastructure built costs 56% less than other government programs Infrastructure quality satisfactory or very satisfactory (94%) Low level of corruption perceived by villagers (92%) 2 GAC Mitigation Measures KDP was the fi rst World Bank-fi nanced project anywhere to introduce an anticorruption action plan as part of its strategic approach to reducing corruption risks. The plan, which refl ected the fi ndings of diagnostics, had three main areas. Eliminating complexity and reducing transactions. Designs that pass funds to communities through various governmental levels are subject to delays and leakages. To avoid this risk, KDP block grants go straight from the national level to the village accounts without stopping at each desk or level along the way. Focusing on financial transactions and making them public. All fi nancial information is made public in the villages. For example, local GAC shopping IN PROJECTS price quotations HOW-TO for NOTES materials are not valid until they are read aloud in public meetings; signboards posted around the villages state material and labor unit costs; and all bookkeeping is managed by an elected implementation committee. Moreover, KDP involves a broad range of entirely independent groups provincial nongovernmental organizations (NGOs) and independent journalists whose job is to inspect KDP sites for signs of irregularity and who have full access to KDP documents. To ensure national-level transparency, KDP sends its audit summaries to civil society oversight groups; and PMD, a division of the implementing ministry, publishes in commonly read provincial newspapers a list of problems encountered so that independent organizations can check for themselves whether they have been reported and fi xed. Monitoring the whole process (and responding quickly to complaints). KDP has a multilayered and robust monitoring system with an external and internal side. The internal monitoring system includes the following: A computerized Management Information System (MIS) that records progress on planning and works, levels and types of participation, complaints and their resolution, and end result evaluations. A grievance and complaints resolution mechanism under which communities can direct their questions or complaints to KDP facilitators, government staff, or

3 CASE STUDY 1 NGOs, or send inquiries directly to a P.O. box or by text message (SMS). National and provincial units record and follow up on inquiries and complaints (see Box 1). For fi nancial management oversight, the national KDP team includes a sevenperson fi nancial unit that provides on-site supervision and training to improve the quality of local recordkeeping. Supervision by fi nancial management (FM) consultants contracted in each province. The external monitoring includes the following features: BOX 1 Complaints-Handling Mechanism Under KDP s unique and innovative mechanism to track and resolve communities complaints, the project maintains a national- and regionallevel complaints database, updated weekly, that is routinely shared with civil society watchdog groups, including the press. At the provincial level, reports by fi eld staff, villagers, and the monitors are logged in, reported to local government and the national team, and then pursued every two weeks until they are resolved. In several districts and provinces, increasingly active project coordination teams also pursue corruption problems. Community participatory monitoring: the benefi ciary communities elect a monitoring committee to oversee project implementation and fi nances. Committee members check on prices, quotations, supply of goods (quality and quantity), benefi ts to the community, fi nancial bookkeeping, and the progress of infrastructure implementation. The implementation team also reports to the communities twice in accountability meetings regarding project progress and fi nances. KDP requires project information to be posted on village information boards. Independent audits: KDP is routinely audited by BPKP and by Bawasda (there have also been audits by Price Waterhouse, Moores Rowland, and others). Independent civil society monitoring through contracts to 28 independent provincial NGOs and 32 independent journalists; inclusion of university students is being considered. Provincial NGOs have monthly meetings with the consultant and Government management teams to list the problems they encounter and review proposed corrective actions; by contract, they have full discretion to share all fi ndings. The journalists contracts provide them with operational expenses so they can visit Results and report on KDP project sites in regional and national newspapers; their articles are not subject to prior review by the project before publication. The journalists provide a route by which the independent NGO monitors and other civil society groups can report on KDP activities. Evidence from all the evaluations, audit reports, and research studies conducted on KDP 1 reveals lower levels of corruption than standard projects. Audits by Price Waterhouse, Moore Rowlands, and the Internal Financial Audit Team reported that fund irregularities were less than one percent; and KDP recovered 40 percent of those missing funds. BPKP (the offi cial government auditor of KDP) found very little leakage and gave KDP a good score of 81 out of a possible 100 points. An independent study 2 conducted a randomized, controlled analysis of corruption 1 World Bank audits and other papers, Olken study, Government audits, Internal Financial Audit Team, etc. 2 Ben Olken, Monitoring Corruption: Evidence from a Field Experiment in Indonesia. 3

4 G O V E R N A N C E a n d A N T I C O R R U P T I O N FIGURE 1 Results Chain Diagram KDP Implementation approach KDP method costs (unit costs for construction and labor) Cost comparison Village infrastructure built through KDP methods costs 56% less than that built through government programs; quality is good. Block grants Anecdotal evidence Fewer delays and leakages. Anticorruption strategy Anticorruption components Access to information and transparency measures Surveys Indirect measures of corruption Community awareness raised, and low level of corruption perceived. Direct measures of corruption Low level of corruption in KDP Multilayered monitoring Complaint-handling mechanism (when sanctions applied) Official audits CS oversight Community oversight Field experiment Data from CHM unit Percentage of missing funds reduced by about 8 percentage points when increased probability of an audit. Missing expenditures in labor reduced by percentage points when increasing community participation and monitoring. Removal of abusive village heads and subdistrict heads, recovery of missing funds, and repair of poor quality infrastructure. 4 and found that announcing an increased probability of a government audit and reporting the audit results directly to a public village forum reduced the percentage of missing funds (funds that could not be accounted for by an engineering inspection of the completed infrastructure project) from approximately 29 percent in control villages to 21 percent in villages receiving the audits. The study also found that increased community monitoring participation was very effective in reducing missing expenditures in labor by 14 to 18 percentage points. By 2006 the number of corruption complaints GAC had IN PROJECTS reached a total HOW-TO of 976 NOTES for a total value of $650,000 a corruption rate of only 0.4 percent of the more than $170 million disbursed in that period. However, it should be noted that only 45 percent of the complaints were resolved, compared with the target resolution rate of 75 percent 3 partly because the target may have been overly ambitious, and partly because it was much more diffi cult to resolve cases related to the misuse of funds and corruption than those involving procedural violations or intervention by outside parties. Grievances or inquiries originated primarily from consultants reports (83% of reported cases during 1998 to June 2004), followed by other sources, audits, and community letters. Civil society groups usually played an important role in following up and monitoring cases. For example, once grievances such as misuse of funds were reported to KDP, journalists followed the stories and reported on them as the cases unfolded. 3 On a positive note, that number of resolved corruption cases was 45% higher than the number resolved in the fi rst two years of KDP.

5 CASE STUDY 1 Anticorruption victories were widely publicized, sending a strong signal to the public that corruption in KDP would be punished. Moreover, there were many examples of sanctions and remedies as a follow-up to complaints. In one year at least 15 legal cases were successfully brought against village and subdistrict offi cials by communities working with NGOs and justice offi cers. Those cases resulted in the removal of abusive offi cials (some even went to jail), the BOX 2 Handling Complaints Successfully in South Sumatra In September 1999, the KDP Facilitator (FK) of Tulung Selapan Kecamatan in South Sumatra discovered that a village head had taken approximately Rps 8 million ($1,000) of KDP funds for personal use. The FK reported the case directly to the KDP district coordinator and the government district coordination team, who reported the details to the NMC, which in turn informed the national-level PMD about the case. At a village meeting the village head was asked to return all the funds immediately: he returned about half the funds, Rps. 4.5 million, promising to repay the balance as soon as he could. The outcome of the meeting was reported to the NMC Complaints Unit, which tracked the case for the next several weeks and in October suggested that another village audit meeting be held to review the case. At the meeting the village head was interrogated again, and the community leaders concluded that he should be fi red. He was relieved of his duties, but the local leaders, KDP consultants, and NMC continued to press for him to return the stolen funds. Every two weeks, the NMC Complaints Unit phoned the district coordinator to check on progress and urge local offi cials to resolve the case. Finally, in December 1999, the village board and some of the village elders met formally with the ex-village head and demanded that he return the debt by the following month or the case would be sent to the police authorities for further action. He paid back the remaining amount in January Source: Susan Wong, Indonesia Kecamatan Development Program, May recovery of missing funds, or, more commonly, the repair of poor-quality infrastructure. In the fi rst three years of implementation, approximately 5 percent of the technical assistance staff was replaced for corruption or for letting corruption happen without reporting it. Moreover, 3 of the project s 19 consulting fi rms were not extended in the project s second year because of concerns over delays or cuts in payments to fi eld staff. Some provinces were suspended because they failed to act on corruption reports; in one the suspension was lifted when the offending parties were removed and the funds returned, but in another, the main district where the problems were happening was dropped from the project entirely. Several studies have shown that KDP is much more cost-effective than other government programs. An independent evaluation carried out in 2004 of 113 villages in four provinces reconstructed the unit costs for construction materials and labor, and found that village infrastructure built through KDP methods costs an average of 56 percent less than equivalent works built through the Ministry of Public Works or through local government contracts (43% for irrigation systems, 59% for roads and bridges, and 66% for water supply), and that 94 percent of the projects were of good or very good technical quality. In 2005 a separate evaluation of 288 villages and 355 infrastructure works had similar fi ndings on cost and concluded that percent of infrastructure was of at least suffi cient quality. This would not be possible if corruption rates were high. Finally, indirect measures of corruption point to relatively low levels of corruption in KDP: villagers own perception is that much less goes missing in KDP projects than in other development projects. In a community survey conducted as part of the 2004 independent evaluation, 95.2 percent of respondents said that there was no malfeasance during implementation of KDP. 5

6 G O V E R N A N C E a n d A N T I C O R R U P T I O N 6 Lessons Learned Several major lessons have been learned from the KDP experience and can be applied in similar environments. Project Design and Diagnostics Corruption diagnostics. Tackling corruption requires an understanding of the incentives that underpin it. Therefore, it is important to conduct corruption diagnostics and corruption mappings when designing projects in high-risk environments, and to link them with the Bank s Country Assistance Strategy. Throughout the program mapping exercises must be updated and new mitigation measures instituted as appropriate. Changing incentives related to corruption. In KDP the approach was to decrease the benefi ts of corrupt behavior, raise the costs of corrupt behavior, close the loopholes for corruption, and increase the possibility of being caught and held accountable. Effective anticorruption plan with robust and multilayered monitoring. While it is extremely diffi cult to eliminate corruption entirely, KDP reduced it signifi cantly through a combination of internal and external controls, use of block grants, and transparency measures. The KDP strategy included important concrete measures: contracting provincial NGOs and journalists to independently monitor KDP, increasing technical oversight, ensuring follow-up and sanctions for culprits, increasing auditing, sharing information with the public, and conducting studies to learn which methods are effective or not. Some GAC components IN PROJECTS did not HOW-TO work as NOTES planned (i.e., NGO monitoring), but overall the multilayered structure was of crucial importance (although complaints from villages that were audited many times point to the need for coordination between monitoring units). The government has promised to identify 11 concrete steps to improve governance and reduce fraud and corruption. Community participation. KDP shows that community participation throughout the project cycle is crucial when tackling corruption. Communities can help task teams understand where the corruption risks are; and building public understanding of the project makes it harder for project actors to manipulate project information. A sense of community ownership can help the project succeed and divert benefi ciaries from corrupt actions, and benefi ciaries can exert effective power during selection of project goods and services. When benefi ciaries participate in the construction of infrastructure, costs decline dramatically while quality remains high. Community empowerment. Under KDP, community participation and capacity building translated into community empowerment, so that villagers now have avenues to make their voice heard. KDP also incorporated gender considerations: for instance, village women s groups were set up and trained to monitor materials procured by the village implementation teams. Government willingness. Not surprisingly, government commitment is crucial when tackling corruption issues. Indonesia s Government showed willingness to experiment and adopt such unique and progressive design features as independent monitoring by media and NGOs and information transparency mechanisms. Bank staff deserve credit for proactively negotiating with the Government and getting it involved and committed. Project Implementation Complaint-handling mechanism. This mechanism is very effective at shining light on corruption cases. Part of KDP s success is attributable to a rapid response to project complaints, which has led to course corrections and practical solutions the villagers can see. The following are some lessons on making the mechanism effective:

7 CASE STUDY 1 Internal capacity (e.g., in collecting and analyzing data) is essential, and training may be needed. All complainants must be assured of confi dentiality. Complaints against staff should be handled by an independent entity. The mechanism should have a simple structure and clear procedures clarity as to how to fi le a complaint, whom to contact, what criteria to use, about how long it takes to resolve complaints and incurred costs, and who is responsible for taking corrective action without confl ict of interest. The mechanism should be updated regularly and specifi c monitoring exercises should be conducted regularly (effectiveness reviews of solved cases, unsolved cases, or monthly reporting of complaints and action taken). An online complaints monitoring system may be useful. It is important to develop communication material with greater use of visual aids, community education material, and radio and local television. Applying sanctions. Offi cial complaint channels are of little use if evidence of corruption is not followed by sanctions. Reparation. KDP reinterpreted the Bank s rules on zero tolerance so that cases of corruption require reparation but not repayment. Once the Ministry of Finance knew that it did not have to repay the funds for village-level corruption but only had to replace the stolen money, the Government itself became better at pursuing corruption. For villagers, a disincentive to complain has been that by complaining they would lose all their funds (while they would lose only the stolen amount if they did not complain). The Government has recently mandated that funds recovered do not go to the state treasury but back to those most disadvantaged by the loss or crime. Computerized system. Computerizing information can play a fundamental role, but the system has to be well designed and appropriately funded throughout the project. KDP s MIS was archaic and did not work as planned, causing delays. Capacity building. World Bank staff were proactive in training Government offi cials in several areas. For instance, although it took some time, the Bank trained BPKP appropriately; now BPKP itself is training Bawasda in auditing. Block grants. Direct transfers to communities from the national level can be effective in reaching target benefi ciaries, reducing leakage, and accelerating disbursement. KDP funds pass through fewer intermediaries than funds from other projects, and KDP infrastructure is built by villagers and not contractors, giving fewer opportunities than in other projects to skim off funds. However, as several studies have emphasized, this approach needs to include creating and strengthening a participatory planning and decision-making mechanism at the local level to enable communities to decide on their priorities. Moreover, it should be mentioned that block grants have been criticized for undermining the role of local governments. Facilitators. Facilitators play a crucial role in limiting corruption in KDP and villages in general. They act as conveners for the project, getting villagers to come to project meetings. They can also give villagers information about projects and KDP procedures, and they can help check facts if accusations are made. Moreover, they usually do not fear redress if they get involved in pursuing corruption cases. Monitoring and Evaluation Grassroots monitoring. Grassroots monitoring is essential in community development projects, but research results suggest that it may be more effective in certain contexts for example, where villagers have 7

8 G O V E R N A N C E a n d A N T I C O R R U P T I O N good information about potential corruption and a strong personal stake in minimizing theft of funds, as in programs providing subsidies for food, health care, education, or microcredit projects. Community training in collecting and analyzing data is crucial. For projects where information gathering is diffi cult and benefi ts from reducing corruption are more diffuse, such as infrastructure projects, grassroots monitoring alone is unlikely to be suffi cient. Intensive World Bank supervision. Intensive World Bank supervision was useful because the auditors and controllers knew that they were being watched and there was some authority of last recourse. Supervision missions were very well funded and planned (i.e., mapping areas and matching fi ndings). Moreover, technicians participated in the supervision missions and adopted new techniques for example, engineers took pictures of building to show communities where the problems were. In some cases, Government offi cials asked the World Bank to do on-site inspections and initiate disciplinary actions. Robust evaluations. Methodologically sound evaluations and studies showed program results and identifi ed problems to be addressed. References Bappenas. Finding of Post Construction Economic Impact Analysis Study for CDD Programs,, October Brief on KDP s Anti-Corruption Work, World Bank Indonesia KDP Task Team: EXTERNAL/COUNTRIES/EASTASIAPACIFICEXT/EXTEA PREGTOPSOCDEV/0,,contentMDK: ~pagePK : ~piPK: ~theSitePK:502940,00.html Guggenheim, Scott, Tatag Wiranto, Yogana Prasta, and Susan Wong. Indonesia s Kecamatan Development Program: A Large-Scale Use of Community Development to Reduce Poverty. Presentation at Shanghai Conference on Scaling Up Poverty Reduction, May 25 27, worldbank.org/etools/docs/reducingpoverty/case/88/ fullcase/indonesia%20kdp%20full%20case.pdf National Management Consultants and Kecamatan Development Program Secretariat, Final Report Evaluation of Infrastructure Quality KDP Cycle IV, September 25, Olken, Ben. Monitoring Corruption: Evidence from a Field Experiment in Indonesia. Journal of Political Economy, 2007, vol. 115, no. 2. Campos, J. E., and S. Pradhan. The Many Faces of Corruption: Tracking Vulnerabilities at the Sector Level. World Bank, Wong, Susan. Indonesia Kecamatan Development Program: Building a Monitoring and Evaluation System for a Large- Scale Community-Driven Development Program. East Asia and Pacifi c Region Discussion Paper, World Bank, May Woodhouse, Andrea. Village Corruption in Indonesia: Fighting Corruption in the World Bank s Kecamatan Development Program, World Bank June GAC IN PROJECTS HOW-TO NOTES World Bank. Project Appraisal Document: The Kecamatan Development Project, April 21, worldbank.org/servlet/wdscontentserver/iw3p/ib/199 9/06/03/ _ /Rendered/PDF/ multi_page.pdf World Bank. Implementation Completion Report: Kecamatan Development Project, June 30, worldbank.org/servlet/wdscontentserver/iw3p/ib/2 003/07/09/ _ /Rendered/ PDF/261630ID0Kecamatan0ICR.pdf World Bank. Project Appraisal Document: Second Kecamatan Development Project, May 23, worldbank.org/servlet/wdscontentserver/iw3p/ib/200 1/06/29/ _ /Rendered/PDF/ multi0page.pdf World Bank. Implementation Completion Report: Second Kecamatan Development Project, August 11, imagebank.worldbank.org/servlet/wdscontentserver/ IW3P/IB/2008/09/12/ _ /Rendered/PDF/ICR7670P Box334053B01 PUBLIC1.pdf World Bank. Project Appraisal Document: Third Kecamatan Development Project, June 2, worldbank.org/servlet/wdscontentserver/iw3p/ib/2003 /06/10/ _ /Rendered/PDF/ ID13rd0Kecamatan0Dev1r pdf World Bank. Project Appraisal Document: Community Recovery through the Kecamatan Development Project, August 22, WDSContentServer/IW3P/IB/2005/09/27/ _ /Rendered/PDF/332030rev.pdf World Bank. Project Performance Assessment Report: Indonesia Second Village Infrastructure Project and Kecamatan Development Project, IEG, March 21, This study was prepared by Tommaso Balbo di Vinadio under the supervision of Ivor Beazley and with the assistance of several World Bank staff: Bolormaa Amgaabazar, Victor Bottini, Steven Burgess, Scott Guggenheim, Sentot Surya Satria, and Susan Wong. 8