79453 AUTHOR ACCEPTED MANUSCRIPT FINAL PUBLICATION INFORMATION Can Participation Be Induced? Some Evidence from Developing Countries The definitive version of the text was subsequently published in Critical Review of International Social and Political Philosophy, 16(2), 2013-04-08 Published by Taylor and Francis THE FINAL PUBLISHED VERSION OF THIS ARTICLE IS AVAILABLE ON THE PUBLISHER’S PLATFORM This Author Accepted Manuscript is copyrighted by the World Bank and published by Taylor and Francis. It is posted here by agreement between them. Changes resulting from the publishing process—such as editing, corrections, structural formatting, and other quality control mechanisms—may not be reflected in this version of the text. You may download, copy, and distribute this Author Accepted Manuscript for noncommercial purposes. Your license is limited by the following restrictions: (1) You may use this Author Accepted Manuscript for noncommercial purposes only under a CC BY-NC-ND 3.0 Unported license http://creativecommons.org/licenses/by-nc-nd/3.0/. (2) The integrity of the work and identification of the author, copyright owner, and publisher must be preserved in any copy. (3) You must attribute this Author Accepted Manuscript in the following format: This is an Author Accepted Manuscript of an Article by Mansuri, Ghazala; Rao, Vijayendra Can Participation Be Induced? Some Evidence from Developing Countries © World Bank, published in the Critical Review of International Social and Political Philosophy16(2) 2013-04-08 http://creativecommons.org/licenses/by-nc-nd/3.0/ © 2013 The World Bank Can Participation be Induced? Some evidence from Developing Countries i Ghazala Mansuri 1 and Vijayendra Rao 2 Abstract: Influenced by Amartya Sen, over the last decade the World Bank has allocated nearly $80 billion to local participatory development projects targeting poverty, improved public service delivery, and strengthened social cohesion and government accountability. But the success of these programs is hindered both by endogenous local factors and by flawed program design and implementation. Two especially important local obstacles are (1) entrenched interests of political agents, civil bureaucrats, and NGOs with either incentives to resist or capabilities to appropriate program resources, and (2) poverty and illiteracy, as the poor and illiterate participate less and benefit less from participatory projects than do the wealthier, more educated, and more connected. After reviewing hundreds of participatory projects, three lessons are clear for program planning: First, contextual factors like inequality, history, geography, and political systems (among others) are important. Second, communities do not necessarily have a ready stock of ‘social capital’ to mobilize. Third, induced participatory interventions work best when supported by a responsive state—donors cannot substitute for a nonfunctional state, and successful programs combine enlightened state action from above with social mobilization from below. Future participatory development projects would benefit substantially from revised planning and considerably more attention paid to evaluation and monitoring. Project managers have historically paid little attention to context, monitoring, or evaluation, in part because the World Bank’s operational policies did not provide incentives to do so. Donor agencies should also exercise greater patience and allow for flexible, long-term engagement to facilitate contextual and programmatic learning, including learning from failure. An important principle that runs through much of Amartya Sen’s work is the importance of public reasoning. He has argued that political voice and public reasoning are deeply interrelated with economic performance and social opportunity, and thus central to the development process. Moreover he has also posited, in the Argumentative Indian (2005) and the Idea of Justice (2009), that public deliberation has deep nonwestern roots, saying that the “attraction of participatory governance...has surfaced and resurfaced with some consistency in different parts of the world (Sen 2009: 323). 1 Lead Economist, Poverty Reduction and Equity Department, The World Bank. Email: Gmansuri@worldbank.org. 2 Lead Economist, Development Research Group, The World Bank. Email: vrao@worldbank.org. 1 Sen’s views on participatory development, particularly with the publication of Development as Freedom (1999), have been very influential - particularly at the World Bank. Influenced by Sen, the World Bank has allocated close to $80 billion to participatory development projects at the local level over the last decade. Other development agencies - bilateral donors and regional development banks - have, in all probability, spent at least as much, as have the governments of most developing countries. This is not the first time that the development community and governments have embraced the idea of participation. It has waxed and waned before. The current wave of interest in participation began as a reaction to the highly centralized development strategies of the 1970s and 1980s, which created the widespread perception among activists and NGOs that ”top-down” development was deeply disconnected from the needs of the poor. Underlying this shift was the belief that giving the poor a greater say in decisions that affected their lives, by involving them in at least some aspects of project design and implementation, would result in a closer connection between development aid and its intended beneficiaries. Indeed local participation has, over the past decade, acquired a life of its own and is now proposed as a means to achieve a variety of goals – better poverty targeting, improved public service delivery, better maintained infrastructure, greater voice and social cohesion and strengthening accountability in government. It is by no means clear, however, that civic participation can be successfully induced through government and donor-funded participatory programs. Achieving participatory governance, and building civic capacity, has historically been an organic rather than a state-led process. By “organic,” we mean spurred by civic groups acting independently of government, and often in opposition to it. Organic participation is usually driven by social movements aimed at confronting and reforming the structure of power, both in 2 the private sector and within government. Such processes are often effective because they arise endogenously within a country’s trajectory of change, and are often directed by highly motivated, charismatic leaders who mobilize citizens to give voice to their interests—their grievances, their rights, and their concerns—and can create or exploit political opportunities to achieve their goals. Organic participation is therefore a broad term. At one level, it includes social movements that fight for greater democratic expression—such as the American Civil Rights movement and the anti-Apartheid movements—or labor movements which aim to unionize workers. The goal of such movements is ultimately to influence or transform the larger political process. At another level, it includes civic watch dog groups, and trade associations formed to represent the interests of a particular industry. It also includes cooperatives, such as the Self-Employed Women’s Association in India; membership-based organizations which aim to improve livelihoods and living standards, such as the Grameen Bank in Bangladesh; or NGOs which seek to build citizenship and develop local communities through self help, such as the Orangi Pilot Project in Pakistan. They engage in a process of creative destruction. First, they imagine a world in which social and political relationships are more equitably arranged, or at least restructured in a manner congruent with the interests of the movement. Then they articulate their vision of this world to expand their influence, and mobilize citizens who believe in this vision to fight for the cause, often at considerable personal cost. Induced participation, by contrast, refers to participation promoted through policy actions of the state and implemented by bureaucracies. Note that the “state” here can include external governments working through bilateral and multilateral agencies, which usually operate with the consent of the sovereign state. 3 There is often some overlap between organic and induced participation. Governments may decentralize because of the efforts of social movements, and the designs of induced participatory programs are often built on organic models. A government may decide to scale up the efforts of small-scale organic initiatives and thus turn them into induced ones. One important question is whether efforts initiated by organic participation can be scaled up by policy interventions in the form of projects. Rather than wait for the slow process of the endogenous development of civic capacity, can policy interventions speed up the process and harness the capacity of citizens to help themselves, and improve the quality of government and the functioning of markets? There are two major modalities for inducing local participation: community development and the decentralization of resources and authority to local governments. • Community development supports efforts to bring villages, urban neighborhoods, or other household groupings into the process of managing development resources, without relying on formally constituted local governments. Community development projects are variously labeled as Community Driven Development (CDD), Community Based Development (CBD), Community Livelihood Projects, Social Funds, and include efforts to expand community engagement in service delivery. Designs for this type of aid can range from community-based targeting, in which only the selection of beneficiaries is decentralized, to those where communities are also involved to varying degrees in project design and management as well as the management of resources. • Decentralization refers to efforts to strengthen village and municipal governments on both the demand and supply sides. On the demand side, decentralization strengthens citizens’ participation in local government by, for example, instituting regular elections, improving access to information, and fostering mechanisms for deliberative decisionmaking. On the supply side, it 4 enhances the ability of local governments to provide services by increasing their financial resources, strengthening the capacity of local officials, and streamlining and rationalizing their administrative functions. The organic development of civic capacity is a complex process—deeply imbedded in a country’s history, its internal conflicts, its conception of nationalism, its levels of education and literacy, the distribution of education and wealth, the nature of the state, the nature of economic and political markets, and a variety of other conditions. Importantly, organic participation is driven by self-motivated leaders who work tirelessly, with little compensation, and often at high personal cost—constantly innovating, networking, and organizing to get the movement to succeed. When this complex process of organic change, driven by intrinsically motivated people, is turned into policy—into projects and interventions to induce participation—then it has to be transformed into manageable, bureaucratically defined entities with budgets, targets, and extrinsically motivated, salaried staff as agents of change. This transformation is common to all large-scale, state-led policy initiatives, and has been famously characterized by Scott (1999) as “seeing like a state.” But, participatory interventions are different from other types of policy interventions in an important sense—encouraging greater participation can be quite threatening for those who enjoy privilege and power – including political agents, civil bureaucrats and NGOs at the local level. Moreover, by devolving power to the local level, higher levels of government are giving away power, authority, and finances to communities over which they may have little control. Yet, those who stand to lose from a shift in the distribution of power are typically also charged with the implementation of participatory development projects and the organization of consultative 5 bodies, like participatory councils, at both the central and the local level. This suggests that project implementers need to be prepared to act against what may well be their own self-interest, in promoting institutions whose purpose it is to upset the prevailing equilibrium. This presents an interesting conundrum that does not face organic movements. Reviewing the Evidence In a recently completed book, we synthesize the available evidence on the effectiveness of induced participatory projects. ii In this section, we take up some of the more salient points to illustrate our main argument regarding the challenges of induced participation We begin with the questions that are at the heart of efforts to go local. Do participatory programs enhance the involvement of the poor and the marginalized in community-level decisionmaking bodies, and does this increase ‘voice’? Does allowing communities a greater say over decisions that affect their lives result in choices that are more aligned with their preferences? Does it produce more resilient and inclusive local institutions? Does it reduce capture and corruption? On balance, we find that participants in civic activities tend be wealthier, more educated, of higher social status (by caste and ethnicity), male, and more politically connected than non- participants. This may reflect, to a degree, the higher opportunity cost of participation for the poor. There is some evidence that the poor spend both less time and less money—as one might expect—on collective activities. There is also some evidence that the poor attend community meetings in larger numbers when they stand to directly benefit by doing so—usually because beneficiaries for targeted programs are selected in such meetings. 6 However, it also appears that the poor, regardless of incentives to participate, tend to benefit less from participatory processes. For one thing, resource allocation decisions typically reflect the preferences of elite groups. Whether or not this distorts the pro-poor intent of programs then depends quite substantially on community characteristics. Studies from a variety of countries show that high-inequality communities do much worse, especially when there is a concentration of political, economic and social power in the hands of a few. Capture also tends to be greater in communities which are remote from centers of power, have low literacy, are poor or have significant caste, race and gender disparities. Policy design can also have unintended consequences—there is some evidence that a large injection of resources can induce greater participation by the wealthy and increase exclusion. In sum, in the absence of explicit and enforceable rules of inclusion, a few wealthy, and often politically connected, men—who are not necessarily more educated than other participants—tend to make decisions in community meetings. Some studies find, moreover, that civic organizations tend to self-sort into those led and attended by the wealthy and those attended by the poor. This may serve to create organizations of the poor that remain weak and vulnerable and undermine the goal of social cohesion. The potential for resource capture by political elites is often no less worrying than the rent-seeking activities of traditional elites. There is plenty of evidence that local politicians use public budgets to reward political allies and loyal constituents, to improve their own future electoral outcomes and to increase their private fortunes. The ultimate question of interest, though, is whether the overall scope for capture is dampened under democratic decentralization. The answer appears to be a cautiously optimistic one. Available studies suggest that under democratic decentralization spending decisions are better aligned with local needs. There is also 7 a shift of resources in favor of the less advantaged. However, much depends on the nature of electoral incentives and the capacity of higher levels of government to provide oversight and to ensure downward accountability. Ultimately, the impetus for most participatory projects, as well as for decentralization, is the hope that greater engagement by citizens will lead to a faster and more inclusive development process. We now assess the extent to which participation improves the targeting of anti-poverty projects, enhances the delivery of public services, increases incomes and leads to better managed natural resources and better maintained infrastructure. We also try to identify, as far as the literature permits, the channels through which community engagement improves outcomes and the factors which appear to limit its effectiveness. To do this we turn to a very large and diverse set of studies which have looked at the following issues: • local versus central targeting of public goods and private benefits; • community engagement in the provision of infrastructure; • decentralized and community based natural resource management; • the decentralization of education and health to local governments or directly to communities through school based management and community based health initiatives of various types; • the impact of community-based projects on the livelihood opportunities of the poor and on their incomes. 8 Studies that have looked at the relative gain from decentralized targeting find that local agents are often able to target beneficiary households better than the center can. However, it appears that the gains tend to be small. There is also considerable evidence that communities and individuals who get left out tend to be poorer and less well connected to relevant political networks. They also tend to be the least literate. Areas that are remote, poorly served by media or disadvantaged in infrastructure quality also tend to have worse targeting outcomes. Community management of common pool resources is often seen as the most natural expression of local collective action. In some sense, it is the model for all other types of community engagement. The creation and maintenance of small scale community infrastructure in CDD projects certainly proceeds from the assumption that communities are well placed to manage such shared assets. There are two central ideas that drive this literature: that community management increases the sustainability of the resource or infrastructure scheme and that it promises greater hope for equity in the distribution of benefits. On balance, the evidence does suggest that greater community involvement tends to improve resource sustainability and infrastructure quality. However, four general concerns permeate the literature. The first is that inequality tends to worsen both efficiency and equity, but greater resource sustainability need not imply greater equity. Second, and perhaps more surprisingly, the transfer of management responsibilities for a resource or an infrastructure scheme is not usually a matter of handing over control to an organic entity with the requisite capacity. More often it requires the creation of local management capacity. In the absence of deliberate efforts to create such capacity—and to plan for the requisite technical and financial resources required for maintenance and management—investments in infrastructure are largely wasted and natural resources are poorly managed. The third is that clear mechanisms for 9 downward accountability are central. The literature on common pool resource management is rife with cases where decentralization is more a mechanism for tightening central control and creating larger incentives for upward accountability than it is a mechanism for greater local discretion—often because it is a holdover from colonial rules designed to extract resources from local communities rather than to benefit them. Finally, communities need to benefit from the resources they manage. For natural resources which create substantial externalities, this benefit should be commensurate with the size of the externality created by the resource and should at least compensate communities for the alternative uses to which they could put the resource for immediate gain. These concerns suggest the need for sustained engagement by higher tier governments or implementing agencies, on the one hand, and substantial local discretion, on the other—the former for setting the broad parameters under which management is devolved, for building requisite local capacity, and for adequately monitoring outcomes, and the latter for strengthening downward accountability. The few studies which have compared community-managed infrastructure projects with similar types of projects created by governmental line departments using a more ‘top-down’ approach serve as a good example. These studies find that community engagement does seem to improve both the construction quality and maintenance of local infrastructure. However, these are cases where the implementing agency provides significant oversight during construction. Maintenance and recurrent costs are also explicitly budgeted for, or the implementing agency is available to provide training and support for maintenance. The case study literature supports these findings. It provides numerous cases of failure when these conditions are not met, particularly in communities with low capacity and for projects that require technical know-how. The good news from these studies is that carefully designed projects can limit the adverse 10 influence of ‘bad’ community characteristics to a substantial degree. While inequality, community heterogeneity and low literacy dampen project quality in these studies, as in others, there is some evidence that they do not overwhelm the positive effects of project design. We turn next to the role of participation in improving service delivery. Efforts to engage communities in improving basic health services or primary schools are now pervasive and the evidence for community engagement is also, on the whole, more heartening—though there are some caveats. First, efforts to engage communities in improving basic health services or primary schools are usually multi-faceted interventions which also involve a substantial injection of funds. This makes it difficult to isolate the impact of community engagement and most studies don’t even try to do so. However, a handful of randomized evaluations of community based health programs are an important exception. These evaluations find that while greater community engagement alone is not sufficient for improving any health outcome, community participation can be quite beneficial where projects also provide trained health personnel at the community level, or make investments in upgrading health facilities. Projects achieve the greatest improvements under these conditions—delivering substantial reductions in maternal and infant mortality, greater use of health facilities and significant modification of health behaviors. These results suggest that community participation may well account for at least some of the positive health effects observed in less well identified studies. Interestingly, the most successful health programs have been implemented by local governments. A number of successful pilot programs have also been implemented through regular government health delivery systems. This is encouraging news, since scaling up health initiatives usually requires working through government delivery systems. In contrast, devolving 11 programs to NGOs appears to have worked less well, though there are too few studies to allow for any clear assessment. Finally, health interventions that provide information to households and communities about the quality of services in their community, as well as government standards of service, tend to improve both household level health outcomes and the quality of service provision, even when no additional resources are expended. This suggests that there might be tremendous room for expanding service quality through interventions that focus on improving community capacity to monitor service providers. Community engagement in education has had similar impacts in some respects, though the evidence is more heterogeneous and the effects are more muted. Overall, studies report an increase in school access as well as improvements in retention rates, student attendance and grade repetition. However, the impact on student learning tends to be weak. This may reflect the short evaluation periods of most studies. The typical time periods associated with improvements in learning range around eight years, while the bulk of studies have data over two to three years, which is a concern. In fact, some studies do show a worsening in student quality at school entry as children from less privileged backgrounds enter schools for the first time. This may account, to an extent, for the negligible improvements in learning, even as there is an improvement in attendance and school retention. As with other interventions, though, poorer and more remote areas are less able to realize gains from decentralized service delivery. Localities also do worse when they are less well administered and more embedded within a nexus of corruption, with poor democratic practices and a politicized administration. Low literacy levels impose a further constraint. Several studies 12 find that low literacy levels make it harder for communities to participate in decentralized programs, and to make informed decisions when they do. Funding matters as well. Increasing the fiscal burden of poor communities can worsen public service delivery. Similarly, as we see repeatedly across various contexts, when maintenance and recurrent costs are not provided, poor outcomes follow: crumbling schools with no teachers, and clinics without medicines or doctors. Community based development projects are often promoted as a tool to eradicate poverty. This is an area where the evidence is most mixed and is largely negative-though the diversity of interventions and the lack of rigorous studies make it difficult to draw clear conclusions. That said, most carefully done studies find that participatory anti-poverty projects have had little impact on income poverty. The studies which find some positive impact on income, savings or assets, also tend to find that gains are observed only for the wealthy or only during the life cycle of the project. When sustainable gains are realized, they are confined to specific sub-groups – such as the more educated among the poor—who are not the largest beneficiaries of the program being evaluated. Some projects also find improvements in nutrition and diet, but only when considering sub-groups who self-select to participate in the project being evaluated. This tends to be problematic since it is unclear whether those who choose to participate are also different from eligible non-participants in other dimensions that the researcher cannot observe, but which could influence their success. The one small exception is perhaps from a couple of recent studies of community based projects which provide micro credit, cash grants and business or livelihood related training to individual community members organized into groups. The results indicate positive and 13 economically significant gains within a relatively short time period. However, none of the studies assesses the sustainability of income gains over the longer run or the impact of the more traditional community livelihood fund. Programs which provide microcredit also do not tend to focus on the poorest. Given the rapid expansion of community based livelihood projects, this is clearly an area where much more work is needed. Next, we turn to the question that is at the core of the participatory approach to development: In what ways do participatory efforts seek to confront and repair civic failures, and how successful have these efforts been? We look at efforts to build social cohesion—in conditions of peace, as well as in the aftermath of conflict; efforts to redress entrenched social inequalities of caste, ethnicity and gender through explicit inclusion mandates; and the creation of participatory forums where collective decisions can be made in a deliberative manner or where, at the very least, the exercise of ‘voice’ can be practiced and refined. There is little evidence that induced participation builds long-lasting cohesion, even at the community level. Group formation tends to be both parochial and unequal. What seems to happen instead is that during the course of the project, people are induced to participate and build networks. But they do so in order to benefit from the cash and other material payoffs provided by the project, an effect that tends to melt away when the incentives are withdrawn. One indication of this is that the observed impacts of participation tend to be in self-reported attitudes about outcomes promoted by projects, such as the “role of the community in development” and “the community can achieve things on its own…”. Impacts are not, however, seen in broader attitudes towards democracy or in objective measures of collective action. Only when projects explicitly link community-based organizations with markets, or provide skills training, do they tend to improve group cohesiveness and collective action beyond the life of the project. 14 With these caveats in mind, there is some heartening evidence that participation has intrinsic value. Communities tend to express greater satisfaction with decisions in which they participate even when participation does not change the outcome or where outcomes are not consistent with their expressed preferences. On the whole, however, the ballot box, though far from perfect, provides a clearer mechanism for sanctioning unpopular policy choices or excessive rent-seeking by traditional or political elites, as compared to more informal forums for deliberation. In decentralized settings, electoral incentives—credible and open elections—tend to help by aligning the decisions of politicians with the demands of their constituencies. Moreover, in such settings we find that participatory and deliberative councils, when they exist, can foster a significant degree of civic engagement. It is less clear how citizens can collectively sanction negligent or corrupt officials or local leaders where such venues for the exercise of voice are not available. Repairing civic failures also requires that social inequalities be addressed. One way of trying to do so is to mandate the inclusion of disadvantaged groups in the participatory process. There is virtually no evidence for whether this works from evaluations of CDD projects. However, a growing body of evidence from village democracies in India indicates broadly positive impacts. Quotas in village councils and presidencies for disadvantaged groups and women tend to change political incentives in favor of the group that is privileged by the quota. Mandated inclusion also appears to provide an incubator for new political leadership. Evidence indicates that women and other excluded groups are more likely to stand for office for non-mandated seats once they have had some experience on a mandated seat. Quotas can also weaken prevailing stereotypes that assign low ability and poor performance to traditionally 15 excluded groups. However, lasting change requires that the inclusion mandates remain in place for long enough to change perceptions and social norms. It is less clear that community-based projects, which are typically ad hoc in duration, can induce similar shifts in attitudes, opportunities and political dynamics. Decentralized programs usually have a constitutional mandate or other legal sanction from the center and are relatively permanent. They may thus be better able to affect change by shifting social and political dynamics over the long term. More evidence is needed on this important question. Participatory interventions are also seen as a valuable tool for getting funds on the ground in post-conflict settings—where the state is viewed as dysfunctional. The limited evidence of the effectiveness of community-based projects in such settings mirrors what we know about inducing participation in situations without conflict. Context matters a great deal, as does the quality of the intervention. Overall, projects tend to have very limited impact in building social cohesion or in rebuilding the state. They tend to exclude the poor and are dominated by elites. However, evidence from Africa seems to suggest that people emerging from civic conflict have a strong desire to participate and that a well-designed and implemented project could effectively draw on this need. To sum up, three lessons appear to be abundantly clear: First, context, both local and national, is extremely important. Outcomes from interventions are highly variable across communities. Inequality, history, geography, the nature of social interactions, networks, and political systems all have a strong influence. Thus a successful project designed for one context may fail miserably in another. This suggests the need for strong in-built systems of learning and monitoring, great sensitivity to context and the willingness and ability to adapt. As some of the evidence shows, carefully designed projects, whether they are implemented by governments or 16 by donor funded implementing agencies are able to limit the negative impact of ‘bad’ community characteristics, at least to a degree. Second, the idea that all communities have a ready stock of ‘social capital’ that can simply be harnessed is naïve in the extreme. Instead, all of the evidence suggests that building citizenship, which includes engaging communities in monitoring service providers and governments and supporting community based management of natural resources or infrastructure, requires a serious and sustained engagement in building local capacity. Finally, induced participatory interventions work best when they are supported by a responsive state. While local actors may have an informational and locational advantage, they appear to use it to the benefit of the disadvantaged only where institutions and mechanisms to ensure local accountability are robust. In fact, local oversight is most effective when other, higher level, institutions of accountability function well and communities have the capacity to effectively monitor service providers and others in charge of public resources. This appears to increase, rather than diminish, the need for functional and strong institutions at the center. It also implies that project implementing agencies for donor funded projects need to have the capacity to exercise adequate oversight. However, there is little evidence that donors alone can substitute for a nonfunctional state as a higher-level accountability agent. When funds are parachuted down into communities without any monitoring by a supportive state, decisionmaking is more susceptible to capture by elites who control the local cooperative infrastructure, leading to a high risk of corruption. Instead, reforms that enhance judicial oversight, allow for independent audit agencies and protect and promote the right to information and a free media appear to be necessary for effective local participation. 17 In this, our findings are consistent with the large body of case study evidence of what Fox (1993) described, many years ago, as a “sandwich movement” of enlightened state action from above interacting with social mobilization from below. The state does not necessarily have to be democratic—though that helps a great deal—but in the sphere where the intervention is being conducted, at the level of the community or the neighborhood, the state has to be responsive to community demands. For example, schools that incorporate parents into decisionmaking will be more responsive to the demands of those parents if the parents have a measure of control over school budgets. And village governments will become more responsive to the needs of citizens when they both function within an electoral democracy supplemented by deliberative interactions. This is not to say that participatory engagement cannot make a difference in the absence of a supportive state under all circumstances. It can, but usually when that engagement is organic and thus outside of, and often in resistance to, the state (Gaventa and Barnett 2010). The Importance of Context It is clear that participatory processes are deeply conditioned by culture, politics and social structure, and vary from place to place. A policy that works in one country, or even one municipality, may fail miserably in another. And, ironically, effective collective action is usually conditioned by a “cooperative infrastructure” that is provided by a strong state. It is not at all clear that strong governments are created by the presence of a strong civil society; it is, rather, a chicken and egg problem that does not lend itself to easy answers. Furthermore, while empowering civic groups may often lead to good outcomes, it is not at all clear that inducing 18 civic empowerment is always superior to a pure market-based strategy, or a strategy that strengthens the role of central bureaucrats. Context matters at both the national and the local level, and comprehensively categorizing it is difficult, but we sketch a few important aspects (see Goodin and Tilley, 2009). At the national level, nationalist ideologies—the manner in which the state (colonial and post- colonial) has created and propagated identity—can create symbolic public goods that facilitate collective action by building a participatory ethic. History matters. The way policies and institutions have evolved—land reforms, education systems, the judiciary, the media, and efforts at social inclusion—can influence the responsiveness of governments to civic mobilization, and consequently affect the incentives for collective action. Similarly, a history of organic participation matters greatly for several reasons. Some countries have a long history of civic participation developed in the process of struggles for independence from colonial rule, or against the rule of entrenched elites. Such significant social movements help give legitimacy to civic activists and create a culture that facilitates civic mobilization. A history of organic participation creates a community of peer-educators who can train others in how to reach a consensus, how to engage in participatory planning and how to hold governments accountable for their actions. In time, organic participation can make it easier to institute a cadre of trained facilitators who can spearhead scaled-up community-based interventions. A history of organic participation also creates an enabling environment for social entrepreneurs to spark participatory innovations, the most effective of which can have important lessons for scaled-up induced interventions. Both induced and organic participation are likely to be influenced by elements of context like the nature and extent of social and economic inequality, and the composition and diversity of 19 groups. The cultures and norms of cooperation that have evolved within the community are strongly influenced by inequality and heterogeneity. These have a bearing not only on the nature of collective action but also on the role of local leaders. Do they act in ways that are benevolent or malevolent to the larger interests of the community they claim to represent? Do they enforce compliant behavior and maximize rents or do they lead with the collective welfare of the community in mind? Geography matters—in particular, remoteness from more developed areas, difficult terrain or weather conditions can create greater vulnerability to isolation with poorer outcomes. Both social heterogeneity and geography have a bearing on the local cooperative infrastructure—the community’s capacity for collective action. If the village has a long history of successfully managing common property resources, then that capacity could potentially translate into a collaboration to manage a school, for example. Migrant urban communities, depending on the nature of their settlement, could either consist of people from the same region and therefore carry over common rural norms and customs to the city, or they could come from different places, which could make cooperative behavior more challenging. Politics matters: the nature of the local state, and the relationship of local communities to the state, deeply affects the extent to which the nexus of accommodation hampers development. In contexts with compound market, government, and civil society failures, local and national political leaders, bureaucrats, and strongmen are often embedded within an extractive equilibrium where the interests of citizens are given the lowest priority. Breaking this nexus, i.e. changing the equilibrium in a manner that makes the state more responsive to the needs of citizens, is at the heart of effective participatory development. For all these reasons, effective participatory interventions necessitate a deep understanding of context: the nature of the 20 cooperative infrastructure, politics, social structures, history, geography, and inequality, at both the national and community levels. Donors, Governments and Trajectories of Change A major problem with donor-induced participation is that it works within what might be described as an ‘infrastructure template.’ Their institutional structure and incentives are optimally suited to projects with short timelines and linear trajectories of change with clear, unambiguous projected outcomes. One could argue that this is due to the path-dependent history of donor organizations as agencies of post-war “reconstruction” rather than “development”. When a bridge is built the outcome is easily verified, the trajectory of change is predictable, and the impact is almost immediate. This is not the case with participatory interventions which engage in the much more complex task of shifting political and social equilibria and have very different trajectories. Unfortunately, however, most participatory projects that emerge from donor agencies are designed within the same assumed trajectory as infrastructure projects. They are also forced to work within three or five year project cycles, and are approved on the belief that at the end of those three years various civic objectives will be reached such as better social capital, community empowerment, improved accountability, etc. Almost all community driven projects go further in projecting gains in outcomes such as a reduction in poverty, improved school enrollment, improvements in sanitation and health, etc. The assumption is that within the period of the project cycle, the intervention will activate civic capacity to the extent that it will repair political and market failures enough to have an observable impact on “hard” outcomes. There are three assumptions inherent in this – first that civic engagement will be activated in the initial 21 period of the project, second that civic capacity will be deepened enough to repair government and market failures, and third that this improvement in the quality of governments and markets will result in a measurable change in outcomes. Figure 1: Possible Trajectories of Local Participation Household Welfare, Projected Public Goods, Quality of Development Path Realized Public Services for Welfare Development Outcomes Path for Welfare Outcomes Time Realized Development Civil Society and Path for Civil Society Governance Outcomes and Governance Outcomes Figure 1 illustrates the problems with these assumptions. The project-based assumption is illustrated by the dotted lines, where civil society and governance outcomes will improve in a predictable linear path that is congruent with changes in measurable outcomes. The problem with this reasoning is that civic change is a highly unpredictable process. Consider the kinds of things that have to take place to make it happen. Individuals have to believe that collective mobilization is worth the effort and be willing to participate; civic groups have to solve the collective action problem and exploit political opportunities to effect change; the nexus of 22 accommodation in government has to be disrupted by the rising cost of ignoring citizens’ interests, so that politicians and bureaucrats change their actions; and their new actions have to result in changes in outcomes. In other words, a change in outcomes has to be preceded by an improvement in civic capacity, and this improvement will then, possibly, unleash a series of changes that will change outcomes (Woolcock 2009). The reality is depicted in Figure 1 by the solid lines. Predicting when change will meaningfully occur in each of these nodes is extremely difficult, because a number of factors come into play. The nature of the cooperative infrastructure; the history of civic engagement and politics; the level of development; the extent to which the state has committed to the process of change and is therefore effectively incentivizing, enforcing, and monitoring the actions of its agents; the level of literacy; information flows—in other words, everything that affects bonding and bridging civic failures. Social equilibrium is hard to change because it has evolved after years of repeated interactions within particular economic, political, and social environments. Therefore, whether at the micro or the macro level, civic engagement often tends to be absorbed, in its early stages, within a nexus of accommodation, with the leaders co-opted by elites. Furthermore, until citizens are convinced that the high cost of engaging, of fighting for their interests and resisting elite domination, is worth the effort, it is unlikely to happen in an effective manner. Widespread participation occurs when a tipping point is reached; when enough people are convinced of the value of participation, when they sense a fundamental change in the nature of politics and power, and when enough people have convinced enough others, resulting in a participatory cascade. Sociologists, borrowing from evolutionary biologists, describe this as a “punctuated equilibrium” (Koopmans 2007)—where long periods of stability 23 are punctuated by brief periods of extremely rapid change. This unpredictability in the timing of change is reinforced, at the local level, by the huge diversity in the nature of communities. Each community is likely to have a different change trajectory. Thus, induced participation—particularly when it is packaged within a project—is almost set up for failure due to unrealistic predictions that emerge from bureaucratic imperatives. The challenge of policy interventions is to figure out where each community is within this complex trajectory of change, and to create an enabling environment for that change to occur in a manner that improves development objectives. For induced participatory projects to have a chance of meeting their objectives, they have to attempt to bring in the spirit of experimentation, learning, and persistent engagement that characterize organic participatory change. Unfortunately, donors are bound by strict timelines, imperatives to disperse money both quickly and effectively, and internal incentives that make honest and effective monitoring and evaluation a low priority at the project level, even with all the rhetoric around the issue. We turn next to this issue. Monitoring and Attention to Context in World Bank Projects The variability of local context highlights the importance of developing effective monitoring systems in large participatory projects. Such projects require constant adjustment, learning in the field, and experimentation in order to be effective. A notable example of an effectively monitored induced community development project was the $1.3 billion Kecamatan Development Program (KDP) in Indonesia, which was active for ten years between 1998 and 2008. KDP provided block grants directly to rural community-based organizations to fund development plans prepared through a participatory process. In this it was very similar to a large 24 number of other community-based projects. Where it differed was in the extent to which it relied on context-specific design, and attention to monitoring systems (Guggenheim, 2006). KDP’s design was based on two key elements: (i) a careful analysis of existing state and community capacity drawn from a set of studies of local institutions; and (ii) a deep understanding of the history of community development in Indonesia. Implementation involved creating a tiered network of motivated and trained facilitators who created a feedback loop to facilitate learning and worked with engineers to supervise the quality of construction. Villagers took control of expenditures and procured goods and services on a competitive basis. They also formed monitoring teams that checked the delivery of material and the quality of construction and reported their findings to the village forum. In addition to participatory monitoring, audits were conducted at the sub-district (Kecamatan) level. Moreover, independent NGOs and journalists were contracted to monitor and report on the quality of the project on a random basis. These innovations in monitoring were supplemented with more conventional quantitative tools such as a carefully designed Management Information System (MIS), several qualitative and quantitative evaluations, and case studies (Wong 2003). Most importantly, the project emphasized an honest system of communication, which allowed observations, both critical and complementary, to constantly inform innovations in design and implementation. Unfortunately, KDP is among a very small group of World Bank funded participatory projects that have made an effort to build effective monitoring systems. We reviewed the monitoring and evaluation systems in place in all participatory projects in the World Bank from 1999 to 2007. In our review, we examined a range of project documents, including the Project Appraisal Document (PAD), which is among the main documents assessed by the World Bank’s Executive Board before approving a loan. The PAD should ideally include a detailed account of 25 the monitoring system and the manner in which the project will be evaluated. We also looked at Implementation Status Reports (ISRs) and Implementation Completion Reports (ICRs) for a random sample of 20 percent of these projects to assess the effectiveness of proposed monitoring and evaluation (M&E) systems in the PAD. The former are typically prepared by the project manager after every supervision mission. The latter are self-evaluations of projects screened by the Independent Evaluation Group. Finally, we assessed information from project supervision documents that synthesized the results of regular project visits by Bank operational task teams. The analysis also encompassed a survey of project managers (task team leaders, in World Bank parlance), to gather information that was not available in project documents. The PADs were striking in their similarity to each other. Often their language was cut and paste from one project to the next, suggesting a singular lack of attention to context. The documents and the survey also revealed a pervasive inattention to monitoring systems. Only 40 percent of PADs included a monitoring system as an essential part of the project design, and a third did not even bother to mention basic monitoring requirements such as a management information system (MIS). When monitoring was mentioned, it usually involved collecting extremely imprecise indicators, and even this was done irregularly. The most important reasons given for the paucity of effective monitoring were poor human and technical capacity and the lack of sufficient funding. On the crucial question of whether monitoring and evaluation data are a guide to learning by doing, only 14 percent of the projects explicitly outlined procedures that would be implemented if the monitoring and evaluation data indicated that a project had gone off track. The majority of project managers participating in the survey stated that the Bank’s operational policies do not provide any incentives for effective monitoring and evaluation. While 26 most believe that M&E is a stated priority for the Bank’s senior management, most believe that in practice this is not the case. Project managers highlighted several major constraints, from an institutional perspective, that limit investments in effective M&E systems. The majority (75%) believe that the Bank’s operational policies do not provide the right incentives to engage in systematic M&E. Further, a majority (66%) of project managers believe that Bank M&E requirements and Bank supervision budgets are not tailored to project size, project complexity or country context. Finally, only a third believe that the standard timeframe for projects (an average of 5.5 years) is sufficient for realizing participatory objectives. An open and effective monitoring and evaluation system requires a tolerance for risk, flexible project design and adequate resources. These seem to be absent for the most part. A major problem, highlighted by the survey respondents, along with lack of management support, was the lack of an adequate project supervision budget. Most task managers also believe that governments see monitoring systems as a box to be checked in order to qualify for a loan, rather than as an instrument to help improve the effectiveness of projects. Given the belief that county counterparts have little incentive to implement good monitoring and evaluation systems, explicit support from the Bank may be all the more critical. Conclusion We do not believe that Sen’s vision of a more inclusive, deliberative and participatory form of development is an idealistic dream. We are hopeful that, with significant changes in structures and incentives within development agencies, induced participatory development can be made much more effective. There is some evidence that things can change. In 2004, the authors of this paper conducted a review of the literature on participatory development (Mansuri 27 and Rao, 2004), in which we found a singular lack of attention to both monitoring and evaluation. Eight years later, we find that the gap in evaluations has been addressed to some extent but the lack of attention to monitoring, unfortunately, persists. This is coupled with inflexible institutional rules that do not internalize the complexity inherent in engaging with civic-led development. The ignorance of context in design is another major problem. If this situation addressed, we believe that participatory development projects will become much more effective. Several things need to be done at the World Bank and other donor agencies. Patience is a virtue. Project structures need to change to allow for flexible, long-term engagement. Projects need to be informed more seriously by carefully done political and social analyses, in addition to the usual economic analysis, so that both project design and expected outcomes can be adapted to deal with the specific challenges posed by country or regional context. Monitoring systems need to be taken far more seriously. For one thing, projects need clear systems of facilitator feedback and effective participatory monitoring and redress systems. The use of new and more cost effective tools, such as SMS based reporting, could assist greatly in this effort. Most importantly, there needs to be a tolerance for honest feedback to facilitate learning, instead of a tendency to rush to judgment coupled with a pervasive fear of failure. The complexity of participatory development requires, if anything, a higher tolerance for failure. This requires a change in the mindset of management and clear incentives for project team leaders to investigate what does and does not work in their projects and to report on it. Instead of focusing entirely on inducing participation, policy would also be well served by thinking through effective ways to ride waves of organic participation. There are several ways in which this can be done: By inviting civic activists to help design and monitor 28 participatory projects; by creating an enabling environment for civic activists to be agents of change; by creating spaces for public deliberation in local governments; and by working with governments to create incentives for agents of government to be responsive to the needs of citizens. Local participation does not work when it is merely the ad hoc, myopically directed, creation of a project. It works when it has teeth, when it builds on organic movements, when it is facilitated by a responsive center, when it is adequately and sustainably funded, and when interventions are conditioned by a culture of learning by doing. References: Fox, J. (1993): The Politics of Food in Mexico. Berkeley: University of California Press. Gaventa, J., and G. Barrett ( 2010): "So What Difference Does It Make? Mapping the Outcomes of Citizen Engagement," IDS Working Paper 347 Goodin, R., and C. Tilly (2006): "It Depends," in The Oxford Handbook of Contextual Political Analysis, ed. by R. E. Goodin, and C. Tilley. Oxford University Press, New York, pp. 3-32. Guggenheim, S. (2006): "The Kecamatan Development Project, Indonesia," in The Search for Empowerment: Social Capital as Idea and Practice at the World Bank, ed. by A. Bebbington, S. Guggenheim, E. Olson, and M. Woolcock. Hartford: Kumarian Press. Koopmans, R. (2007): "Protest in Time and Space: The Evolution of Waves of Contention," in The Blackwell Companion to Social Movements, ed. by D. A. Snow, S. A. Soule, and H. Kriesi: Blackwell Publishing. Mansuri, Ghazala and Vijayendra Rao, 2004, "Community-Based and -Driven Development: A Critical Review," World Bank Research Observer, 19, 1-39. Mansuri, Ghazala and Vijayendra Rao, Localizing Development: Does Participation Work? World Bank Policy Research Report, World Bank Publications, forthcoming 2012 Amartya Sen, The Idea of Justice, The Belnap Press of Harvard University Press, Cambridge MA, 2009 29 Amartya Sen, Development as Freedom, Alfred A Knopf, New York, 1999 Amartya Sen, The Argumentative Indian, Farr, Straus and Giroux, 2005 30 i This paper draws heavily on: Mansuri, Ghazala and Vijayendra Rao, Localizing Development: Does Participation Work? World Bank Policy Research Report, World Bank Publications, forthcoming 2012. ii Mansuri and Rao, op cit. page 25. In the book we review over 500 studies, focusing at greater length on large scale participatory projects that have been evaluated using representative samples of target populations with good counterfactuals. While much of the literature on participation tends to use a case study approach, which is undoubtedly quite invaluable for understanding the process of change, only representative sample studies of large- scale projects with good counterfactuals can generate enough data to make statistically valid judgments on large- scale policy interventions. In other words, we focus on studies that have a valid control group for the communities that have been targeted (or “treated”) by the intervention. Without an adequate counterfactual, there is no effective way of attributing the observed changes to the intervention rather than to changes that might have occurred outside the intervention. Generally speaking, this means that our findings derive from econometric analysis. However, we do draw on several case-studies to develop ideas and to illustrate our conceptual framework. Ideally, this econometric work should be complemented by good qualitative work, which can help to illuminate the processes that resulted in the observed impact. There is an unfortunate dearth of such work. We refer the reader to the book for a detailed synthesis of the evidence and the bibliography. 31