93700 GOVERNANCE REVIEW OF PNPM RURAL Community Level Analysis Final report Andrea Woodhouse 22 January 2012 PART ONE: INTRODUCTION 3 WHY CONDUCT A GOVERNANCE REVIEW OF PNPM RURAL? 4 A GOVERNANCE REVIEW IN THREE PARTS 4 OBJECTIVES 5 METHODOLOGY 6 1. P RELIMINARY ANALYSIS . 6 2. STAKEHOLDER REVIEW, ACTION RESEARCH AND CROSS-LEARNING 6 3. FINAL REPORT 7 PART TWO: FINDINGS & OBSERVATIONS FROM FIELD TRIPS 8 WHAT IS THE STATE OF LOCAL GOVERNANCE IN PNPM? 9 WHY DO THESE CHALLENGES EXIST? 12 CHANGES IN THE WIDER CONTEXT 12 CHANGES WITH THE PROGRAM DESIGN 13 PROBLEMS WITH MANAGEMENT AND IMPLEMENTATION 13 PART THREE: RECOMMENDATIONS 22 RECOMMENDATIONS FROM WORKSHOP 23 PART ONE: INTRODUCTION WHY CONDUCT A GOVERNANCE REVIEW OF PNPM RURAL? PNPM is the biggest community driven development program in Indonesia, and one of the biggest in the world. It has grown significantly since its inception as KDP in 1998. In 2006, the government announced that it would expand PNPM to cover the entire country. In the space of a few short years, between 2006 and 2009, PNPM thus grew almost fourfold to cover every sub-district in Indonesia. This growth took place extremely quickly and, as might be expected from any rapid scale-up, many elements of the program—including management and oversight mechanisms and staffing— struggled to keep pace. Since 2009, supervision missions for PNPM have identified a growing number of governance problems in the program. These include a rise in serious fraud and corruption cases1 and evidence of problems at community level with the quality of facilitation and participation and other key elements of governance in the program. This Governance Review is an attempt to assist the government in better understanding the extent and causes of such problems and to generate recommendations for improvement. A GOVERNANCE REVIEW IN THREE PARTS This community governance review (this report) is one part of a broader three-part governance review of the program:  Part One consists of independent validation and monitoring of benchmarks for governance improvement committed to by the Government of Indonesia in the preparation of PNPM IV. Part Three: Community  Part Two consists of a central level analysis Part One: Governance Benchmarks Review (this of governance strengths and weaknesses focusing on (ongoing) report) internal management at central level and management of entities below the central agency. Part Two:  Part Three is this community governance Management review review. It consists of an analysis of governance, (forthcoming) accountability and corruption strengths and weaknesses at the local level (hamlet, village and sub- district). The community governance review draws on and is closely related to other ongoing work on governance in PNPM. These include the PSF fiduciary team’s quarterly analyses of corruption risks 1 Reported rates of corruption are overall still low, though are thought to under-report the true extent of the problem. within the program and the work conducted by the PSF Local Level Institutions (LLI) team on participation. The fiduciary team’s analyses are released in the form of the six-monthly anti- corruption updates. These contain a detailed assessment of corruption risks within the program and analyze the practices, processes, institutions and interests at community level that affect corruption in PNPM. The LLI participation work analyzes community participation and the motivation of community members to participate in PNPM. This report thus does not focus on these issues: instead, it should be read in conjunction with those reports. OBJECTIVES The overall objective of the community governance review is to identify practical recommendations to support the POKJA Pengendali and PMD in its decision-making to strengthen governance in the program. This involves gaining a better understanding of: 1. What is the nature and extent of the community governance challenges facing PNPM? This includes an analysis of what kinds of things are working well and what kinds of things are working poorly, whether this varies from place to place, and whether the types of governance and social accountability challenges facing PNPM can be broken down: for example, into:  Participation  Transparency and information  Corruption  Marginalization and elite capture  Accountability 2. Why do existing governance challenges exist? This includes an analysis of the extent to which governance challenges in PNPM are a result of (i) changes in the wider environment in Indonesia since the program’s inception; (ii) risks, strengths and weaknesses inherent in the program design, or the way that changes in the program design have affected current conditions; and (iii) problems in implementing and managing the program. It analyzes the ways in which five practical elements of the program affect local governance:  Social accountability mechanisms  Financial management  Facilitators  Complaints-handling mechanisms  Management information systems & reporting 3. How can governance in PNPM be improved? This includes an analysis of what kinds of changes to the existing program design or management and implementation arrangements would be most effective at improving community governance, as well as the risks, costs and likely consequences of introducing these. METHODOLOGY The review is a practical exercise intended to generate discussion, exchange ideas and identify recommendations for improvement. The methodology was thus based on a program of ‘action research’ and cross learning in which a wide range of people with different areas of expertise visited field locations, shared their observations and generated ideas and discussion through a series of workshops. The process was three-fold: 1. P RELIMINARY ANALYSIS . The aim of the preliminary analysis was to develop a set of observations with which to guide the wider program of action research. It consisted of both a desk- and field-based review. The desk- based review took place in October 2011. It consisted of a literature review and interviews and meetings in Jakarta with staff from PMD, the World Bank, NMC and donors. The field visits took place in October and November 2011. They included missions to West Java and West Sumatra, both independently and as part of a PNPM supervision mission. Locations were chosen to ensure the maximum diversity in the kinds of factors that might affect local governance and to see how PNPM functions in a range of contexts. They included (i) ‘old’ areas that had participated in PNPM since the early days of KDP, and ‘new’ areas that were fresh to the program and had joined during the scale-up; (ii) areas where local governance in the program was working poorly and well; (iii) rural and urbanizing areas and (iv) areas that were still in the program, and ones that had been phased out of PNPM Rural and migrated to PNPM Urban. Locations also included those that were more or less remote, though time and logistical constraints prevented any truly remote locations from being included. They also prevented visiting any ‘pemekaran’ areas where new sub-districts had been created due to administrative ‘splitting’ of districts. The field visits included interviews with:  Local government, including PNPM project coordinators  Provincial coordinators and specialists, including in training and complaints-handling  District facilitators and specialists, including in financial management and complaints-handling  Sub-district facilitators (empowerment and technical)  Staff of the sub-district level ‘activities management’ units, the UPK, which manages funds  The ‘BP-UPK’ – Badan Pengawas UPK, an elected body intended to oversee the UPK  The ‘BKAD’, Badan Kerjasama Antar Desa, an elected body intended to handle inter-village affairs, including holding the UPK and other entities to account  Village head  Village cadres  Women borrowers  Community members 2. STAKEHOLDER REVIEW, ACTION RESEARCH AND CROSS-LEARNING The aim of the stakeholder review, action research and cross-learning was for a much larger group to reflect on and assess the preliminary analysis, visit field locations to observe for themselves the dynamics of local governance in PNPM, share their experiences and observations, and identify recommendations for improvement. The work took place in December 2011. It consisted of a preliminary workshop, field visits to four provinces, a wrap up workshop in Jakarta to share observations and identify recommendations, and a briefing and discussion session with PMD. Participants were selected to ensure a range of expertise in the group. They included staff from NMC, the World Bank, PMD, AusAID and local NGOs. They also included PNPM facilitators and community members who had hands-on experience of dealing with governance challenges in PNPM (the community members were ones who had been elected to the ‘BKAD’ and ‘BP-UPK’ bodies within the program, which are intended to increase accountability), as well as financial management and social accountability specialists from India and Bangladesh, who shared their experience from elsewhere in the world. As with the preliminary analysis, the field teams visited a diverse set of locations. The provinces included West Java, Central Java, East Java, and South Sulawesi, and included one extremely remote location only reachable by boat. As with the preliminary analysis, the field teams interviewed a range of stakeholders at the village, sub-district, district and provincial levels. 3. FINAL REPORT The aim of this report is to present the findings of the preliminary analysis and the stakeholder review and consolidate the recommendations generated during the workshops. It focuses on the key recommendations discussed with government at the briefing in December. The final report is intended to complement and feed into the work being conducted by IRAI on Part Two of the Governance Review, the Management Review, and to assist with the benchmarks process. PART TWO: FINDINGS & OBSERVATIONS FROM FIELD TRIPS WHAT IS THE STATE OF LOCAL GOVERNANCE IN PNPM? Summary. The state of local governance in PNPM is mixed. It is still remarkably strong, but faces problems. Participation rates are still mostly high, but the quality of participation has in some places declined. There are weaknesses in transparency and information sharing. The performance of accountability mechanisms in the program—such as the village accountability meetings and the creation of the BKAD and BP-UPK—varies, and the incidence of serious corruption is up. Despite this, the ‘foundations’ of PNPM are still strong. These foundations, though, are being eroded by serious pressure from ‘higher-up’ problems related to the broader governance environment, changes in the program design, and problems with implementation and management. Although the foundations of PNPM are mostly still strong enough to withstand these challenges, in some places, they have combined to cause multiple accountability and local governance failures. PNPM must address these program design, implementation and management issues to avoid undoing its years of good work at community level. 1. The state of local governance in PNPM is mixed: it is still strong but faces clear problems. Overall, local governance in PNPM at the community level was still remarkably strong. In most locations, participation rates were high and community members had good knowledge of PNPM and its processes. This was especially true in locations that had long participated in PNPM, where community members reported that participation rates had either stayed high or increased because villagers saw that the program delivered tangible results. PNPM was popular with community members, who compared it favorably with other programs and overwhelmingly wished it to continue. However, there were some clearly identifiable problems. Most of these are encapsulated in the commonly held view in Jakarta that PNPM has begun to ‘lose its soul’2: that its accountability mechanisms are functioning, but in a mechanical and procedural way that bypasses the true purpose of the program. There is a good deal of truth to this, though it is partly a by-product of the program’s success in ‘normalizing’ and therefore ‘professionalizing’ the participatory decision- making model. The ways in which it is true are as follows: (a) Participation is mostly still high, but its quality is mixed Although participation rates were still high, the quality of that participation was variable, and in some cases appeared to be declining. Community members tended to understand project rules and procedures, often in impressive detail. They often also understood the empowerment aims of the program. They reported being motivated to participated for private but also public reasons: they attended meetings partly in the hope that their households would receive savings and loan funds but, especially in smaller rural communities, also because they wanted to participate in community decision-making about things that would benefit their villages as a whole. However, attending meetings did not necessarily mean that they spoke, debated, or participated in a meaningful way. This was captured by the trend in some locations towards more ‘representative’ decision-making, in which there was much community discussion of priorities at the hamlet level, but where decision-making processes at the village level and above were dominated by elites. Although the 2 This view was often also expressed by the province, district and sometimes sub-district PNPM field staff who were interviewed. It was less commonly expressed, though, at the village level, where community members tended to feel the ‘spirit’ of the program rem ained. ‘representative’ nature of such decision-making does not by itself lead to poor governance, the lack of direct participation increases opportunities for village elites to misuse these processes to further their own agendas. (b) There are weaknesses in information sharing and transparency There were also weaknesses in information sharing and transparency. Although community members had good knowledge of project procedures, they tended to know little about project finances or where and how to complain about misuse. In many of the locations visited, community notice boards were poorly maintained. A common problem appears to be that the wider purpose of the notice boards is poorly understood. The organizing principle should be that they be used to share regularly the kind of information that would best enable ordinary community members to monitor the project themselves and hold it accountable (which would include clear, simple information about project finances, PNPM’s principles, the rights and responsibilities of community members, and where and how to take action). Yet where they are updated, they are often used in an ad-hoc way to share information such as photographs of infrastructure and community meetings or detailed, yet hard-to-understand, financial information: information that may be interesting but is insufficient to enable ordinary community members to act as monitors. (c) The performance of accountability mechanisms such as the BKAD and BP-UPK has varied The field teams observed the performance of the BP-UPK and BKAD accountability mechanisms, as well as the use of community monitoring teams, cross-audits, women’s procurement teams, and village accountability meetings. Such mechanisms varied in their effectiveness from place to place. Village accountability meetings tended to be held and were relatively well attended, but —as with the notice boards—information can be shared in a way that is hard to understand: for example, instead of presenting simple numbers and analysis to enable meeting participants to understand how funds have been used, long lists of complex numbers are read out in a way that makes them difficult to understand. The efficacy of the BKAD and BP-UPK mechanisms varied. In some places they were effectively non-functional or had members who were either too close to local elites or— in the case of BP-UPK members—did not necessarily have the financial literacy or clout to effectively monitor how funds were used. In other places, however, they functioned very well, and were an essential accountability mechanism. Those involved in PNPM at the sub-district levels tended to feel that the BKAD and BP-UPK mechanisms had huge potential. (d) The incidence of serious corruption is up, indicating broader vulnerabilities The research did not unearth any new information about corruption rates in PNPM, and the rates of reported corruption remain low (though many believe that, due to an unwillingness of community members to complain, the true numbers are underreported). Nevertheless, the incidence of serious corruption cases, as flagged by the supervision missions, is up. Most of these are related to SPP repayments. Corruption can only become serious in PNPM if there is widespread collusion and the multiple accountability mechanisms that are built into the program fail: that the incidence of serious cases is up suggests a broader accountability problem. It is also mainly through such missions that such cases are identified or followed up, suggesting that corruption is under- reported. 2. The quality of local governance varies by context. The quality of governance in the program varied from place to place. The research did not seek to map the quality of governance directly onto whether an area was old or new, rural or urban, remote or non-remote, phase-out or non-phase out3, or pemekaran or non-pemekaran. Nevertheless, there are features of each that affect the quality of governance. Whether an area is new or not to the program appears to matter, but not in the way that one might expect. In old areas, contrary to expectations that people might get bored of the program, participation tended to remain high. Community members reported that this was because they had seen PNPM deliver results in previous iterations of the program, and that they knew much about PNPM’s processes through repeated engagement. Whether an area is rural or urban does not appear to matter in a direct manner, but what does tend to matter is how tight-knit and socially cohesive a community is, which is easier in smaller rural communities whose participants know—and are often related to—each other. In such communities, villagers appeared to be more motivated to participate by the public good of contributing towards community affairs. They also appeared to be less likely to wish to use project means—such as the SMS complaints number—to resolve problems that arose, preferring to resolve problems informally and directly. The phase-out area visited during the research was performing well, mainly due to its committed and competent staff. There is a risk, however, of increased corruption of UPK funds in phase-out areas, because the standard project controls no longer apply: this creates a potential ‘black box’ where there is little scrutiny 3. Despite the strength of local governance, the project is facing strong pressure from ‘higher up’ problems in implementation and management. Despite the overall strength of local governance at community level, though, the project is facing strong pressure from problems that have arisen at ‘higher-up’ levels beyond the community. These are in three main categories. First, the variance in the wider governance environment at local level means that, in areas with high wider corruption and poor local governance, there can be political interference on PNPM, pressure to collude, weak government oversight, and political capture of the kinds of institutions, such as the prosecutors and courts, that provide a means of redress. Second, the proliferation of PNPM offshoot programs, combined with the massive growth of ‘old’ revolving funds managed by UPKs, has significantly increased the workload of facilitators and UPK staff, decreasing the quality of supervision. Third, a range of problems with implementation and management—related to the scale-up—have created local level pressure. Pressure to disburse, heavy workloads, difficulties with deploying and recruiting facilitators, a decline in training quality, weaknesses in complaints handling, and unrealistic reporting requirements have led to a decline in the quality of supervision and facilitation, creating an enabling environment for corruption and other governance problems to grow. These problems will be discussed later in the section. 3 These are areas that are no longer in the program. 4. These problems can overwhelm, and disrupt the success of PNPM In many communities, the local social controls are strong enough to withstand this kind of pressure. But where they are not, as in the increasing number of large-scale corruption cases, they have failed badly: in these cases there have been multiple accountability failures. This has serious consequences for PNPM: it erodes people’s faith in the program. WHY DO THESE CHALLENGES EXIST? Summary. The causes of the local governance problems facing PNPM are threefold. First, the variation in the quality of wider local governance, though not determinative of poor governance, can create an enabling environment for corruption and put pressure on the main actors in the program. Second, factors related to the program design—primarily the lack of adequate automatic staffing provisions for new PNPM offshoot programs and the lack of adequate financial management provisions for the increase in the ‘old’ revolving funds of KDP—have created the conditions for poor governance. Finally, and most importantly, problems with implementation and management have created an environment for poor local governance. First, the strength of social accountability mechanisms varies. Second, there are problems with financial management, including the pressure to disburse, late payments, manual bookkeeping, the high number of windows and reports, the lack of clarity with accountability mechanisms, and the unnecessarily high number of cash transactions. Third, there are problems with facilitation. Facilitators are increasingly in short supply, poorly trained, under pressure or lacking support, or overworked. Fourth, the complaints handling system needs improvement. Community members tend to be reluctant to complain. This is compounded by a lack of capacity in following up complaints and problems with the complaints ‘tools’, including the SMS system. Finally, reporting requirements are unnecessarily heavy, and detract from downwards accountability. The causes of the local governance problems of PNPM can be organized into three categories: 1. Changes in the wider context 2. Changes with the program design 3. Problems with management and implementation:  Social accountability  Financial management  Facilitators  Complaints handling  MIS & reporting CHANGES IN THE WIDER CONTEXT There is now more variance than before in the quality of local governance in Indonesia The first (albeit weakest) is related to changes in the wider context. Indonesia has changed since the inception of KDP in 1998, and especially since its ‘big-bang’ decentralization. Corruption remains widespread. Many argue it has worsened. There is now much more variance than before in the quality of governance at local level. Weak local governance is not ‘determinative’ of poor governance in the program The variation in governance in PNPM is related much to the ‘nuts and bolts’ of program administration than wider political and decentralization-related changes. Poor wider local governance does not automatically mean poor governance within the program: areas with high degrees of corruption and collusion among the local government can nevertheless be ones where governance within the program at community level is strong, particularly if there are good facilitators and program staff and communities have long participated in the program. But it does change the incentive structures of the main actors Nevertheless, in districts with weak governance, there can be weak government oversight, political pressure to collude, and ‘capture’ of the kinds of institutions, such as the prosecutors and courts, which would otherwise provide an alternative means of redress for complaints. These kinds of dynamics have been evident in the cases of serious corruption that have emerged. In one case that emerged recently, large amounts of money were misused through systematic collusion and corruption spearheaded by the project coordinator at district level, a government employee. Because of his status as a local strongman, he was able not only to ensure that key positions within the program (such as facilitators and local accountability mechanisms such as the BKAD) were ‘captured’ by his friends and family, but also, by virtue of his friendship with and ‘capture’ of the prosecutors and local courts, to make it difficult for villagers to find local legal redress for complaints that would reasonably be followed up on. CHANGES WITH THE PROGRAM DESIGN The second is related to the program design and problems with implementation and management. This report does not analyze the program cycle for corruption risks and benefits: such analysis is contained in the Anti-Corruption Updates. Nevertheless, it was clear that two elements of the current program design are problematic not simply for corruption but also for local governance in PNPM more broadly defined. These are (a) the lack of automatic staffing provisions accompanying the proliferation of PNPM offshoot programs, and (b) the fact that the current program design lacks adequate financial management provisions for the increase in ‘old’ revolving funds managed by UPKs. They will be in the sections on ‘facilitators’ and ‘financial management’ below. PROBLEMS WITH MANAGEMENT AND IMPLEMENTATION The third (and strongest) is related to implementation and management. Several factors have combined to decrease the quality of and time available for community facilitation. First, weaknesses in the way social accountability mechanisms are supported create an enabling environment where poor governance can flourish. Second, problems in financial management, including the over-use of cash and the growth of UPK funds without adequate staffing, create corruption risks. So too does the requirement to disburse. Because the program is a government program, there is a legal requirement to disburse by year-end. This has decreased the time available for facilitation and created a culture where disbursement targets outweigh the need to engage in long-term community planning and problem solving. Third, difficulties in recruiting and deploying facilitators create vacancies and further increase the workload of existing facilitators; the problem is made more acute by high facilitator turnover, a decrease in the competitiveness of facilitator salaries, increased competition in the ‘facilitator market’ due to the ‘professionalization’ of PNPM and the success of the participatory project model, and the burden placed on facilitators by the proliferation of PNPM offshoot programs. The quality and length of training has decreased, which leads to a decrease in the quality of facilitation: it concentrates now on function, not the ‘why’ of empowerment. Fourth, weaknesses in complaints handling—stemming partly from work overload—can, especially in areas with high political or other interference, contribute towards an environment of impunity. Finally, an increase in reporting requirements, combined with unrealistic deadlines, has increased the workload of PNPM field staff members at sub-district, district and province level and decreased the quality of oversight and community engagement. These issues are discussed in more detail below. 1. S OCIAL A CCOUNTABILITY Summary. Social accountability is at the heart of PNPM. For PNPM to succeed, ordinary community members must have a meaningful say in the kinds of decisions that affect their lives. They also must be able to have healthy debate and hold the program to account if it falls short. Several elements have to be in place for this. They include (i) strong information sharing and transparency, (ii) good quality participation, including women’s participation, and (iii) well-functioning mechanisms for ordinary people to be able to hold the program to account, including the social accountability mechanisms, complaints handling and the musyawarah pertanggungjawaban. As outlined earlier in the report, the state of these is variable. This is because of problems with management and implementation, particularly with facilitation and complaints handling, which are analyzed further below. Well-functioning social accountability mechanisms are at the heart of PNPM Social accountability is at the heart of PNPM. Although the precise boundaries of how social accountability is defined may vary, the heart of it—empowerment and civic engagement—remains the same. For PNPM to have achieved its social accountability aims, ordinary community members, including the poorest and most marginalized, must have voice and decision-making power in the program and an ability to hold it to account. Without this, the success of other aspects of the program means little: infrastructure may be built, financial management systems may be run efficiently, and community members may improve their livelihoods through savings and loans programs, but the aims of PNPM will not be met if community members have little voice in or ownership over the program and cannot hold it to account. Indeed, there may often be trade-offs between the aims of social accountability and other program aims, such as speed and efficiency. Social accountability rests on two key principles: first, that community members have a full understanding of their rights and responsibilities in relation to PNPM as active, engaged participants, and second, that the mechanisms they can use to hold the program to account— including complaints handling and the BKAD and BP-UPK—work well. For these principles to be upheld several elements have to be in place: facilitators must provide excellent training at the ‘socialization’ stage of the project during which project principles are explained; community members have to participate meaningfully in decision-making about the program; community members must receive good and regularly updated information about the program; and the social accountability mechanisms of the program must work well. Yet their strength and effectiveness vary The strength of this, however, varies in PNPM. As outlined earlier, review participants found that, although local governance at community level was fairly strong, there were problems with information and transparency and with the quality of participation. They found too that the BKAD and BP-UPK mechanisms had a lot of potential: the mere fact of having village representatives with ‘hiring and firing’ power over those who administer the funds (UPK) widens the locus of accountability in the program, making it less dependent on how good the facilitators and complaints-handling mechanisms are. However, the BKAD and BP-UPK mechanisms had been implemented unevenly, and did not yet have the strength or capacity to act effectively as oversight mechanisms. Furthermore, it is important that they are set up well, because there is the risk of them being ‘captured’ by elites. It is important also that the members of BP-UPKs have adequate financial literacy to be able to perform the audit function effectively. Much of this rests on the quality of management and implementation. Much of this rests on the ‘nuts and bolts’ of program management and implementation, in particular the strength of facilitation and training and the strength of complaints handling mechanisms. These are explored in more detail below. 2. F INANCIAL M ANAGEMENT Summary. PNPM is experiencing problems with financial management that decrease the quality of local governance in the program. The legal requirement to disburse has led to immense time pressure, created bad incentives and created a ‘culture of disbursement’ where the needs of disbursement trump empowerment goals. This is exacerbated by time-consuming manual bookkeeping, by the high number of funding windows, boxes, and reports, and by the high number of cash transactions. The lack of transparency caused by manual bookkeeping is exacerbated in areas that have participated in the program for a long time, where the size of revolving funds has grown significantly but without adequate staffing and financial management provisions. There are several financial management challenges facing PNPM that affect the quality of local governance. These include the following: The legal requirement to disburse has led to immense time pressure, created bad incentives and created a working culture where disbursement trumps empowerment. There is a legal requirement in PNPM to disburse funds by the end of each financial year. This has had both direct and indirect effects. First, it has created huge pressure on PNPM field staff to get things done quickly. If the aim of PNPM were simply to build infrastructure or issue small loans, this would not in itself be an issue. The aim of PNPM, though, is not simply to raise rural incomes, but also to promote community governance and promote participation in development. Those processes—ensuring community members are well informed, mobilizing wide participation, encouraging wide-ranging debate and discussion, following up on allegations of misuse and ensuring that accountability mechanisms function well—take time. There are often clear trade-offs between speed and participation, and the current pressure to disburse skews the way these are valued. It means disbursement almost always trumps empowerment, and changes the incentive structures facing field staff: it means that allegations of misuse may be overlooked in the need for speed, and accountability steps rushed. Second, it has a knock-on effect on the working culture of the program. The ‘culture’ of the program, perpetuated either through the informal norms and working culture of project staff or formally through what kinds of people are hired, how they are trained and how they are evaluated, has a ripple effect on governance. It affects what aspects of the program design are emphasized in practice, and has an impact throughout the program chain —for example, in how project procedures are explained to villagers and how members of the UPK, BKAD, BP-UPK and so forth understand their role. Bookkeeping is conducted manually and takes too much time. This reduces transparency. Field teams found that books and local accounts in general tended to be kept well. Keeping the books and accounts, though, proved to be extremely time-consuming. Books are kept manually, and reports are written separately. There is an absence of simple accounting software that automatically generates reports and which might reduce opportunities for misusing funds. This is especially acute in “old” KDP locations where the size of revolving funds has grown significantly, and where UPK staff may lack the technical skills or staff capacity to cope. In many ‘old’ locations that have participated in PNPM since KDP days, the size of the revolving funds managed by UPKs has grown significantly. There are now reportedly over 370 UPKs with over IDR 3 billion in revolving loan fund capital. The following chart shows the amounts of money being handled by a sample UPK in West Sumatra, at the beginning of its participation in KDP and now. UPKs use computers to record and process financial million rupiah transactions. But the data is usually entered manually into Excel: computer operators do not have adequate or 400 standardized software to automate transactions. This 300 means that, with the growth of both the size of the funds and the numbers of borrower groups, the amount of work 200 million involved in processing the funds and maintaining the rupiah books grows too. 100 Yet PNPM lacks adequate staffing provisions for UPKs 0 with large amounts of revolving fund growth. Before After The number of funding windows and reports is high. This increases complexity and reduces transparency. Field teams observed that there were too many funding windows and reports. This causes complexity, reduces transparency, and increases the workload of the UPKs and local facilitators. Oversight and accountability functions are unclear. This makes them less effective. Field teams reported that the financial management oversight and accountability functions were unclear and weak. There was no simple overall accountability framework communicated to all actors, outlining who approves payments, who makes payments, who oversees transactions and accounts, who receives reports, who makes management decisions and who audits. As a result, community members lack clarity on how things are supposed to work, and the main project actors lack clarity on who is responsible for what. This lack of clarity means that there is no clear division of tasks, which can leads to conflicts of interest, and which reduces transparency. The number of cash transactions is unnecessarily high. This increases the risk of corruption. The financial management and social accountability specialists on the field teams observed that in PNPM, there are a high number of cash-based transactions, including transactions that could be conducted by bank transfer. Cash-based transactions are less transparent and carry a higher risk of corruption than automated transactions. 3. F ACILITATORS Summary. Good facilitators are critical to community governance. Yet facilitators are increasingly absent, poorly trained, lacking in the right skills or motivation, or facing political pressure and attempts at collusion by local government. They also are increasingly greatly overworked. The problem of overwork faces not only sub-district facilitators, but also district and provincial facilitators and specialists, who argue that the pressures of the scale-up, the rise in the number of areas they have to cover, and the demands of reporting leave them little time to do their jobs effectively. Local governance depends on good facilitation The strength of project facilitation within communities—through the role of the sub-district facilitator and the attendant effect on village cadres—is one of the most critical elements of good governance within PNPM. Facilitators are the locus of most of the accountability measures in the program. A good facilitator, with the skills, time and support to do her job well, can help ensure that community meetings are well attended and facilitated, that villagers understand the principles of PNPM, that village cadres are well trained, that notice boards are well maintained, and that complaints are handled well. Yet facilitators are increasingly in short supply. They are underpaid… Facilitators, though, are increasingly in short supply. The number of vacant positions is high. Salaries are no longer competitive, particularly for remote regions. According to PNPM field staff, this is partly due to the ‘professionalization’ of the PNPM model: they argue that because the participatory decision-making model (which is dependent on facilitation) has become popular and adopted by other programs, demand for good facilitators has risen sharply. This means that the bulk of facilitators see their work as a standard job, not as part of a social movement, and accordingly can treat their role as one of executing a project, not as doing the long, hands-on work of helping to empower communities. “Before they were called by their hearts,” one provincial consultant said of facilitators. “Now they are called by money.” The below-market pay also has a knock-on effect: according to one workshop participant, it reportedly leads to facilitators feeling that their work is not respected or recognized as much as it should be. Facilitator shortages add to the workload of existing facilitators, who report being chronically overworked, with little time to spend among the communities they serve. …poorly trained, The standard and length of training has declined. Facilitators are supposed to receive two full weeks of routine refresher training per year, but tend to receive much less than this. Materials are reportedly undifferentiated and not engaging. The key issue appears to be that training focuses too much on the ‘what’ of project procedures and rules, rather than on the ‘why’ of empowerment and the ‘how’ of problem solving. Instead of learning the principles behind PNPM and being given the peer-to-peer learning opportunities necessary to learn how to solve problems that arise in different contexts, facilitators are learning project rules in a formulaic way. This constrains their ability to deal creatively with problems that arise. The problem in training applies not only to facilitators but also to UPK, BKAD and BP-UPK staff, village cadres, villagers and others. Facilitators reported they received little assistance from Jakarta in developing differentiated training materials for communities. This in turn causes program quality to decline, adding to the sense in which the project in followed in a ‘mechanical’ way in communities. It also detracts from the sustainability of the model: if facilitators are not supported to develop locally appropriate materials, they are forever dependent on Jakarta to survive. …under pressure or lacking support, In areas with weak wider local governance, facilitators (at the province, district and sub-district levels) can often come under pressure from political actors to collude in misuse or face other forms of pressure and political intervention. It can be difficult for them to withstand this, as they are technically hired by the provincial Satkers and so can feel beholden to them. In KDP, facilitators were usually from a different area to the one in which they worked and were backed up by the national level if they faced local intimidation. That now is mostly not true. In the cases where facilitators have stood up to pressure and intimidation, they have sometimes reported feeling that the PNPM national level management do not back them up. When this happens, they have little redress. In one location, facilitators had reported alleged misuse by local political actors up to the national level for months before they got a response. The response was to demote most of the facilitators, who had not been involved. The choice facing the facilitators was either to protest the demotion, and ‘appeal’ or stay quiet. The problem is that the consequences of losing such an appeal are, in PNPM, grounds for immediate firing. It is thus almost never in the interest of any facilitator to protest wrongful treatment by NMC. This in turn decreases the incentives for reporting misuse. …and overworked All PNPM field staff, from the provincial down to the sub-district level reported feeling massively overworked. The causes of this are several. First, there are too few facilitators for the number of villages and sub-districts covered. Review participants reported that since the scale-up, there has been an imbalance between the number of facilitators and the number of villages covered: facilitators cover a much larger number of areas than before, with an attendant rise in workload. Although there are provisions in the program design for positions to be added as the number of locations increases, in reality the actual implementation of this has been patchy. This has been exacerbated by the high number of vacancies in facilitator positions: when this happens, existing facilitators get more work. Second, the proliferation of PNPM ‘offshoot’ programs4, especially ‘PNPM Integrasi’ has come without adequate staffing provisions. Although some of these offshoot programs (such as PNPM Generasi) provide extra facilitators, they either do not provide all of the right kind of facilitators (for example PNPM Green provides environmental but not extra technical facilitators) or do not provide sufficient numbers of facilitators. PNPM field staff consistently reported that their workloads had greatly increased as a result of such additional programs, saying that this limited the time they had to do the kind of hands-on facilitation and supervision work at community level that is essential for strong governance. PNPM field staff reported that their workloads had increased most as a result of PNPM Integrasi. Finally, reporting requirements are too heavy and unrealistic. Facilitators are spending a much greater proportion of their time than before on writing and filling out reports for the PNPM management. They report that this leaves them much less time than before to conduct supervision and facilitation work in communities. These problems of overwork are faced not only by sub-district facilitators but also by district and provincial facilitators. This overly heavy workload is faced not only by sub-district facilitators, but also by district and province level staff. All staff interviewed during the preliminary analysis reported that their workloads had increased significantly since the days of KDP. 4. C OMPLAINTS H ANDLING Summary. There are three kinds of problems with PNPM’s complaints handling system. First, there is an incentives problem with complaints at community level. Various factors, including the threat of being dropped from PNPM and the fear of social disharmony, combine to weaken the incentives for individual villagers to complain. This is exacerbated by the low expectation that complaints will be acted upon and a lack of information about what will happen when people complain. Second, the program’s capacity for resolving complaints is overloaded. Third, the SMS gateway and CHS application are not working. These problems with handling and processing complaints combine even further to weaken incentives for complaining. The review found room for improvement with PNPM’s complaints handling system . Review participants identified three main problems:1. Community members are reluctant to make complaints The field teams reported that community members tended to be reluctant to complain about problems in the program. There were three main reasons for this. First, community members tended to value social harmony, and feared that making complaints would threaten this. Second, 4 As part of its expansion and growth, PNPM instituted a series of pilot programs in some areas. These included PNPM Generasi Sehat dan Cerdas (PNPM Healthy and Bright Generation), launched in July 2007, to contribute to achieving the three Millennium Goals of universal basic education, a reduction in child mortality and improvement in maternal health; PNPM Green, designed to further integrate environmental issues into CDD planning and implementation; and PNPM Integrasi, designed to integrate the proposals made during the community planning platform into local government development plans. there was a mismatch of sanctions: the risk of being dropped from the program tended to outweigh the benefits of complaining. Community members realized that complaints—especially if about large-scale corruption—might cause their locations to be suspended from the program, which would cause their entire communities to suffer and negated the potential benefit from complaining. Third, people tended not to know if their complaints would be addressed or followed up properly, and felt that there was little protection for those who complained. The fear of intimidation or local social pressure acted as a disincentive for complaining. 2. The program’s capacity for resolving complaints is overloaded and the SMS gateway and CHS application are not working. The review also found that even if people wished to complain, the capacity of the program to resolve complaints was inadequate. This played out in two main ways. The first stemmed from the increased workload of facilitators. Usually, the role of facilitators in resolving complaints has been critical. Community members tend to prefer to resolve problems locally through existing community mechanisms. Where these fail, though, and where community members want to resolve the problem through the program itself, the facilitator has been the person through whom complaints are lodged. Yet facilitators have been so busy focusing on day-to-day program management that they have had little time to handle complaints: in many locations a backlog has been built up. They also have had little time to help build capacity among community members to resolve problems. Second, the program structures above facilitators have been slow to respond to complaints, during which time problems can escalate. This is partly because of the sheer scale of the program, partly because of problems in managing the complaints handling system, and partly because there is a work culture in the program in which facilitators and district and provincial staff feel they can only report complaints up through official channels (the standard reports) but have no other way to call special attention to big problems that demand follow-up—to do so would be seen as straying outside official procedure. The impact of this was seen in one large-scale corruption case, in which the sub-district, district and provincial facilitators had reported a large-scale corruption case up to the national level every month for several months with no response. Third, there are technical problems with the existing complaints handling systems. The SMS gateway and CHS application are not working. Community members in rural villages tended not to use or know about the SMS system, preferring instead to resolve problems personally or through using social structures. Where those local systems fail, though, it is important for the program to have a safe, anonymous means for community members to report complaints—and the knowledge that those complaints will be followed up on. These are currently lacking. SMS numbers often didn’t work: and where they do, community members do not know when their problems will be acted upon; furthermore, at the time of writing the report the CHS application was not working properly. 5. MIS & R EPORTING Summary. Problems with the MIS & reporting system affect governance at the local level. Reports are too long and time-consuming and come with unrealistic deadlines. This decreases the quality of information, facilitation and complaints handling. The lack of standardization of software applications also causes problems: it makes it difficult to extract analysis or to compare information among regions. The review identified several problems with MIS and reporting. They include the following. First, routine reports tend to be extremely long and are time-consuming to produce. The monthly reports encountered in district and provincial offices were often hundreds of pages long, containing information whose sheer volume and density made it difficult to understand—and meant they were rarely read. Much of the length of reports is because attachments are submitted in hard copy, not soft copy, but the effect is that the reports look too daunting to read. The reports are time- consuming to produce, which takes time away from community facilitation. One district office, for example, reported that they spent over eight working days a month on reporting up to the province-level office: about four for each monthly report, one each for weekly report, and another day or two for incidental reports. This was on top of time required to meet and otherwise coordinate with the government coordination teams at province level. This left them little time to spend with community members or supervising the work of sub-district facilitators. Second, the deadlines for reports are unrealistic and leave little time to verify information. According to one provincial consultant, UPKs were required to close their books on the last day of the month and sub-district facilitators had to submit their reports to the district level one day later. The district then had to report up to the province by the 5th of every month, and the province up to the national level by the 10th. This meant that verification of information in the reports was almost impossible. The software applications associated with the reports are not standardized and come with bugs. The software applications associated with the reporting system are not standardized. Different areas use different applications for their databases. These are not integrated, are untested, and often infested with viruses. This significantly decreases their utility. It also means that it is difficult to compare information among regions: the variables in the applications used in one region do not match those of another. It is thus very difficult to extract any analysis from the information in the databases or generate any automatic comparative reporting. PART THREE: RECOMMENDATIONS RECOMMENDATIONS FROM WORKSHOP Recommendations for improvement were generated at a wrap-up workshop in Jakarta in December 2011 attended by PMD, NMC, BKAD & BP-UPK members, community members, facilitators, donors, civil society actors, World Bank staff members, and specialists in financial management and social accountability from elsewhere in the world. The overall message emerging from the action research and workshop process was that improving local governance in PNPM requires PMD to solidify their vision of what PNPM should achieve and then take practical steps to translate that into reality, including making the necessary changes to the program design or the system of management and implementation. Participants of the workshops had a number of practical suggestions for improvement in the five areas identified. Recommendation Details SOCIAL ACCOUNTABILITY 1. Commit to strengthening the BKADs Workshop participants stressed the need to strengthen and BP-UPKs and undertake the BKAD and BP-UPK accountability mechanisms, which diagnostic work to explore how best have tended to be effective when put in place properly. to do this. PNPM should thus commit to finding ways to strengthen these mechanisms and undertake further diagnostic work to identify how to strengthen them, including a SWOT analysis of different options. Some preliminary options identified in the workshop included (a) taking steps to ensure the BKAD and BP-UPK mechanisms had a clear legal foundation; and (b) forming a BKAD/BP-UPK association. 2. Provide better district-level oversight Workshop participants stressed the need for better over PNPM, for example by involving independent district-level oversight of PNPM. PMD NGOs and paralegals in monitoring. should explore ways to provide this, for example by involving NGOs and paralegals in monitoring and oversight, building on the experience in KDP of providing grants to independent journalists. 3. Institute social audits at village level Workshop participants highlighted both the success of social audits in other contexts as a means of strengthening community monitoring and the success of cross-audits of villages and UPKs when they take place. PMD should review the experience in other countries and projects of doing this. A study tour to India or Bangladesh to review the experience there would be a good starting point. 4. Provide assistance in prosecuting Workshop participants suggested PMD should be provided assistance in helping to prosecute corruption cases through the legal system cases and that legal assistance should be provide to help villagers with taking their information to the police and helping to follow up cases through the system. One suggestion was to create an MOU between the executing agency and a legal organization at national level supported by TNP2K. FINANCIAL MANAGEMENT 5. Reduce cash transactions: pay Workshop participants observed that too many suppliers by bank transfer transactions were in cash and suggested exploring ways to reduce cash transactions (and therefore reduce corruption opportunities) by paying more suppliers by bank transfer. 6. Work with banks to create ‘deliver The financial management and social accountability banking to the people’ specialists from India and Bangladesh recommended that services/barefoot bankers PNPM work with the private banks to deliver mobile banking services to communities that lacked them. This has been done in India/Bangladesh. Discussion of this could form part of a proposed study tour. 7. PMD should explore the possibility of Some workshop participants recommended that funding TPKs having a village-level account (to should flow directly to the village level, which would which funds would flow directly). reduce opportunities for fraud and corruption and further put power in the hands of communities. Others, however, disagreed, saying such a system would be impossible in practice – that it would simply be far too difficult logistically and administratively given the vast number of villages in the program. PMD should investigate the feasibility of such a system. 8. The executing agency should explore how to improve the training of BP- UPKs. 9. The number of cross-audits should be increased. 10. PNPM/PMD should review options for Most corruption cases that exist are related to SPP, not enhancing community monitoring infrastructure. Workshop participants recommended that across the board, but particularly for PNPM should review options for enhancing community SPP groups. monitoring across the board, but particularly for SPP. 11. Fix the financial applications used by PNPM should have simple, custom-designed software for UPKs to reduce their workload. UPK management. Such software should be applied nationally, which would improve MIS and reporting in addition to reducing workload. FACILITATORS 12. Structure facilitation so that it is done The recommendation here is not to reduce the number of by teams, not individuals (i.e. FKs facilitators, but rather to deploy and task them differently covering more than one kecamatan in teams of facilitators, so that they may work together, but in teams). learn from one another and share their workloads. 13. Look into improving the quality and Workshop participants recommended that PNPM identify length of refresher training and create practical ways to improve the quality and length of a ‘learning house’ for facilitators so refresher training and increase peer learning they have better cross support and opportunities for facilitators, the overall aim being to peer learning opportunities ensure regular review, supportive exchange and cross- learning at decentralized levels. One suggestion was to do so through creating some kind of ‘learning house’ for facilitators and/or to link this to the facilitator certification process. The World Bank/PSF could consider financing this. 14. Ensure that the association of The facilitators’ association should take steps to ensure facilitators actually sanctions poor that poor facilitators are actually sanctioned. facilitators 15. Use the mechanisms of ‘social Workshop participants proposed that the mechanisms of controls’ for facilitator performance social controls could be used as part of the performance review for facilitators so that there was an incentive to conduct good facilitation. This has been implemented successfully in other contexts and could form part of an India study tour. COMPLAINTS HANDLING 16. PNPM should explore how to conduct better ‘socialization’ through different media, religious institutions, and village cadres, and should link to Creative Communities to do this. 17. PNPM should explore how to improve the anonymity of the SMS system. 18. Fix the complaints handling system (SMS gateway etc). MIS & reporting 19. Simplify existing reporting formats Existing reporting formats should be simplified so that they are less burdensome for project staff. 20. Amend the reporting deadlines so PMD should amend the current reporting deadlines to they are more realistic make them more realistic – the current deadlines are too short. 21. Separate routine from ‘special’/’urgent’ reports, especially for corruption cases 22. Ensure better quality control of reports. Involve NGOs, universities 23. Institute intensive work on improving One proposal was that the project could automatically community notice boards. fire facilitators if no community notice board is found on supervision missions.