Institutional supply, public demand, and citizen capabilities to participate in environmental programs in Mexico and India

: In this paper, we argue that, like a three-legged stool, participatory programs require three elements for stability: a supply of participatory institutions, a demand from citizens to participate, and citizens with capabilities for participation. We illustrate the importance of these three elements using case studies from forest management in central India and southern Mexico, and use the evidence from these cases to suggest hypotheses for future investigation. We argue that when participatory programs are implemented in places where demand for citizen engagement is weak and citizens lack the capability to engage, participation is unlikely. On the other hand, where people demand to participate and have the capability to do so, they are sometimes able to overcome obstacles to utilizing participatory institutions. Individuals’ agency for citizen engagement is developed in interaction with the structures of participation; this means that there may be a long-term synergy between the supply of participatory institutions and the development of skills and demand for citizen engagement. Our work implies that designers of participatory programs should pay equal attention to cultivating participatory capabilities, providing incentives that enhance demand for participation, and building institutions to open spaces for participation in governance. Building participatory environmental governance may thus require long-term, sustained attention to both citizenship and institution-building. Acknowledgement: We humbly acknowledge the help of many informants in India & Mexico who shared their life experiences with us. We received funding from the National Science Foundation (SBE 0721745 and CHE-1313932 to CRS, GRF-2007054263 to FF) and the USDA National Institute of Food and Agriculture, (McIntire Stennis Project # 1013165 to FF), and support from Dartmouth College’s Environmental Studies Program and the Texas A&M Department of Ecosystem Science and Management. We received helpful cri-tiques of the ideas in this manuscript from Paromita Goswami, Kalyan Nayan, Hal Fischer, Seth Frey, and audiences at Indiana University, Uppsala University, and the University of Minnesota, as well as from the anonymous reviewers of this journal.


Introduction
There is a growing body of evidence that public participation in decision-making is an important factor in the success of a wide variety of environmental programs in developing countries (Chhatre and Agrawal 2008;Agarwal 2010;Persha et al. 2011;Fleischman et al. 2014;Rodriguez Solórzano 2014, 2016. This has led many governments, NGOs, and donors to promote participatory reforms as elements of environmental programs. Yet evaluations of the impact of formal participatory programs demonstrate that participatory programs designed by governments, NGOs, and donors often fail to achieve meaningful levels of participation (Speer 2012;Mansuri and Rao 2013). This paper explores the paradox of people participating in public decision-making even when formal opportunities to do so are absent, yet not participating when formal opportunities are provided to them.
Two divergent streams of research address this paradox, both of which emphasize the role of institutional structures in determining participatory outcomes. As we will see in this paper, neither of these streams are adequate to explain the rise of locally-based political participation that we describe. The first stream attempts to improve the design of formal participatory programs by developing and testing institutional fixes that favor improved participation (Fung et al. 2003;Innes and Booher 2010;Gregory et al. 2012;Nabatchi 2012;Fung 2015). The second questions the underlying motivation for formal participatory reforms, arguing that they are often designed to coerce people, to co-opt pre-existing informal mechanisms, or are simply standardized elements added to all programs with the goal of satisfying donor demands, with no intention of facilitating participation (Cooke and Kothari 2001;Hickey and Mohan 2004;Larson and Ribot 2007;Gaventa and Martorano 2016). These arguments focus on the way that institutional designs for participatory programs stack the deck against those negatively affected by environmental programs -particularly poor people in developing countries. These scholars are skeptical that institutional reforms will enable the oppressed to engage in meaningful participation in the context of oppressive post-colonial states. In place of participatory programs, many of these scholars call for more "radical" reforms, which often boil down to vague calls to reform the institutional structure of the state and/or civil society organizations to favor the needs of the oppressed (Cooke and Kothari 2001;Hickey and Mohan 2005). The unspoken assumption in much of this literature appears to be that a more democratic society is possible if oppressive colonial and postcolonial institutions can be disassembled (e.g. Kothari 2009). Our cases largely fit the pattern of participatory programs that appear to stack the deck against the marginalized -yet we find that in some circumstances they contribute to long-run empowerment.
Our argument in this paper is that the two research traditions outlined above are limited by their focus on institutions. Building on long-term historical case studies of public engagement in environmental decision-making processes in India and Mexico, we argue that institutional design can interact with individual agency to produce citizens who are capable of participating and demand to participate and can thus take advantage of favorable institutions. Like a three-legged stool, three elements are jointly necessary for a stable participatory program: A supply of institutions that enable and provide benefits from participation, a demand from citizens to participate, and a citizenry with the capability to participate. Over time, citizens who have opportunities to participate may develop greater capabilities to do so -and may begin to demand a role in decision-making. As we show in our case studies, this occurs through changes in the skills and expectations of individuals with regards to participatory processes. Furthermore, we show that when skilled citizens demand participatory opportunities, they are sometimes capable of creating them, or forcing higher level officials to develop participatory programs. On the other hand, when citizens' efforts to express their voice are repeatedly stifled, they may learn that participation is not worth their time, and future participatory opportunities will be less likely to succeed. This framework helps explain our paradox, contributing to our understanding of both success and failure in participatory program management: While failures can result from inadequate supply, demand, or capabilities, ill-conceived programs that foster the development of demand and capabilities may enhance possibilities for long-term participatory success.
In making this argument, we build on the mainstream of political science research on political participation in developed democracies. This research has not been widely applied to understanding environmental governance, particularly in developing countries. If radically disassembling colonial structures is necessary for permitting more genuine participation, the open, stable democracies of the developed world should offer models for better participation. Yet research from the developed world shows that even where the supply of participatory institutions is not a major constraint, most people do not participate in political processes not only because some are excluded, but also because many lack interest in participation (Verba and Nie 1972;Brady et al. 1995;Schlozman et al. 2012). The resulting political decisions therefore tend to favor the interests of organized elites (Schatttschneider 1975;Schlozman et al. 2012;Gilens and Page 2014) except when citizens build strong organizations, embedded in existing social networks, that enable them to challenge elite power (Han 2009(Han , 2014Han et al. 2011).
The findings of this literature draw into question the assumption, underlying many participatory programs, that people will participate in environmental decision-making if appropriate opportunities are available to them. In developed democracies, those who participate have greater access to resources (Verba and Nie 1972;Brady et al. 1995;Brady 1999;Schlozman et al. 2012) and/or deep personal commitments (Han 2009). Han finds that political activists in the United States usually experience a process of education that enables them to grow into effective political advocates, a process that is far from automatic and is often facilitated by involvement in voluntary associations that provide formal or informal training (Han 2009(Han , 2014. Collectively, this research points to the importance of training, experience, and pre-existing inequalities in influencing individual decisions about participation, holding institutional supply constant. In this framework, a supply of participatory institutions is neither a necessary nor a sufficient condition for citizens to participate in governance activities -in fact, many of the political activists studied by Han (2009) participate even though there are no formal political processes enabling their participation, and/or the deck is stacked against them by histories of injustice (for example, African Americans in post Hurricane Katrina New Orleans). Our argument builds on and extends Han's work, which focuses on the cultivation of political activists, by examining how the cultivation of political activists interacts with the development of new political institutions over longer time frames than examined by Han.
Since capabilities and demand are developed primarily through participation in politics, a supply of participatory opportunities may be a prerequisite for the development of capabilities and demand. However participatory institutions that are supplied to a population without the capabilities to take advantage of them are unlikely to foster participation, for the obvious reason that the people do not know how to participate. We argue that this catch-22 lies at the heart of the widespread failure of participatory reforms. Nonetheless, our case material shows that escape from this catch-22 is possible when concerted efforts are made to engage citizens in learning how to participate in governmental processes over long time frames. In contrast to Mohan's (2004, 2005) view that participation's potential can only be realized within the context of a radical politics of development, we show that these escapes can happen even within the context of participatory politics that participation's critics would view as incomplete and co-optive. Being subjects of government-led participation can sometimes enable people to become more free.

Defining the outcome: participation and successful participatory programs
We define public participation or citizen engagement to include activities by individuals or groups that aim to influence public actions and outcomes Schlozman et al. 2012). A participatory program or reform is a program which aims to increase public participation in decision-making. In addition to those programs formally advertised as participatory, this includes many reforms categorized as social accountability (Fox 2015;Brinkerhoff and Wetterberg 2016), decentralization (Coleman and Fleischman 2012;Smoke 2015;Lund et al. 2018), community-based (Mansuri and Rao 2004;Dressler et al. 2010), rights-based (Kashwan 2013), or deliberative (Chambers 2003;Gibson and Woolcock 2008). Although much scholarship in environmental policy focuses on participatory government programs, political parties, NGOs, and social movements also play crucial roles in reform processes (Houtzager 2003).
The outcome of interest in this paper is whether participatory programs enable citizens to participate in decision-making. We consider participatory programs to be successful when citizens are engaged in a sustained fashion in decision-making and perceive that their preferences have been taken into account. This does not require that every individual speak in participatory forums or that decisions favor their preferences -people may be satisfied if their representatives have spoken on their behalf, and it is not possible for a political process to leave all participants satisfied with the outcome. Furthermore, our definition of success does not require that the participatory process lead to better material outcomes for those involved. Even if participation does not improve material outcomes, it is desirable because participation and democratic citizenship are in and of themselves a core element of political freedom (De Tocqueville 1945;Ostrom 1994;Sen 1999;Ribot 2008). There is evidence that participatory processes often lead to better decisions (Speer 2012;Mansuri and Rao 2013;Bixler et al. 2015;Fox 2015), yet these improved outcomes require that the intermediate step, citizen participation, be achieved.

The three legged stool of participatory stability
We argue that three elements are jointly necessary for the success of participatory programs. The first is a supply of participatory opportunities. As discussed above, literature on participatory reforms in the environment and development arenas have focused on institutional supply. Institutions, in this literature, are understood as the "rules of the game" (North 1990, 1), which constrain and permit human behavior Ostrom 2005). The institutional focus makes sense because in many developing countries participatory opportunities are limited by centralized decision-making and elite domination which have their roots in a colonial past. Unfortunately, many so-called participatory institutions serve to reinforce the power of post-colonial elites by incorporating formal bureaucratic procedure into previously informal processes, or by privileging technical inputs in terms not accessible to the general public (Ojha 2006;Ribot et al. 2006). Such institutions do not provide the access and standing necessary for the possibility of citizen influence (Senecah 2004;Walker et al. 2006;). As such, these participatory institutions may be seen as rules that elites impose on society in order to perpetuate their political and economic domination (Moe 1990;Knight 1992).
Institutional reformers in both developed and developing countries work towards finding ways to open participatory opportunities to a wider audience However, supplying participatory opportunities will be insufficient to create public participation if people's willingness and ability to use the opportunities are missing. Given that research, discussed above, shows that citizens frequently do not participate even when opportunities are available, institutional supply is clearly insufficient.
The second leg of the stool is demand for participatory opportunities. Demand refers to the expression of interest in participation. People who demand to participate will show up at meetings, and if they are excluded, may protest their exclusion and claim new spaces for participation (Cornwall 2004). Many people do not participate in public decision-making processes because they are not interested in the process or because they are not aware of how to participate . Participation in decision-making can be costly in terms of time, and also in terms of potential exposure to public scrutiny. Elites may block demands for participation if those demands threaten their power, and if interventions are designed to reinforce elite power, participants may have little motivation for participation (Larson and Ribot 2007;Ribot 2008).
In order to overcome these barriers, participants need motivation. Motivation may be supplied by material rewards that come from engaging in decision-making, such as access to resources or preventing unwanted decisions. While this implies an economic calculus involved in the decision to demand participation, not all motivations for participation are financial -studies in India show that people vote because it gives them a sense of involvement in the community, as well as of personal efficacy and leadership (Carswell and De Neve 2014). While disempowered people may be skeptical of the ability of political processes to bring them rewards, well-off people may also find participation too costly relative to small benefits. This calculus may vary in different decision-making venuessomeone who eagerly participates in a village-level government may be reluctant to speak in a regional assembly because they feel shy or out of their area of comfort. Another person may feel that the potential benefits of participating in a powerless village government are too low but may engage in a regional assembly where they believe they can obtain tangible benefits. Demand can be learned -a recent study found that citizens who experienced positive responses from their political participation were more likely to participate in the future (Sjoberg et al. 2017).
The third leg of the stool is the capability to participate. We draw the concept of capabilities from the work of Sen (1987Sen ( , 1999 and Nussbaum (2000), and their many followers, who use the term to refer to "what people are actually able to do and to be" (Nussbaum 2000, 5). While the concept is similar to the idea of human capital, it differs in that it emphasizes not only the instrumental role of human behavior in achieving economic ends, but also the value of that behavior in enriching human lives (Sen 1997). Thus, referring back to our definition of participatory success, the capability to participate in political processes is valuable not only because participation leads to better outcomes, but because participation enriches human lives and is a core element of democratic freedoms.
Participation requires specific capabilities: time, money, and/or specific skillsets. Many people are constrained by childcare obligations or long working hours from even to obtaining information about participatory opportunities (McNulty 2015). Others lack the money to travel to participatory venues or otherwise engage with participatory processes. These costs can be substantial -McNulty (2015) cites a World Bank report that full participation in Peru's participatory budgeting process required an expenditure equivalent to nearly 1/12th of a minimum wage income every year. Participation may require organizational skills, procedural understanding, and networks with media, government, and similarly aggrieved groups. Governmental processes require literacy in official languages and understanding of accounting and legal procedures. Those who are best prepared to participate are often those who work in government or other white collar occupations, whereas manual laborers may be poorly prepared Moodie 2013). Many people are excluded from participatory processes by their inability to develop these capabilities -for example, in many societies, women's mobility and voice outside of their homes is restricted (Agarwal 2001). Differential capabilities may contribute to elite capture.

Methods
We selected India and Mexico for this study of participatory environmental programs because these two countries have widely discussed participatory environmental programs with contrasting histories (Kashwan 2017). Kashwan argues that in the early 20th century both countries had strongly centralized governance over natural resources. Mexico's unusual history of peasant revolts and comprehensive national land reforms meant that by the late 20th century, Mexico's rural population had far more economic and political power over natural resources than in most other developing countries (Bray et al. 2003(Bray et al. , 2009. By contrast, postindependence India increased central control over natural resources (Gadgil and Guha 1995). In the last 30 years, India has introduced new participatory programs into the natural resource sector with very mixed results (Lele and Menon 2014). Drawing on the institutional theories that dominate the study of participatory programs, we would thus expect to see very different dynamics of participation in these two countries. At the same time, recent participatory reforms in both countries were initially driven by the kinds of top-down initiatives that are widely critiqued in the literature, making both of those cases least-likely cases (George and Bennett 2005) for successful participatory reform, according to theories presented by prominent critics such as Cooke and Kothari (2001) and Hickey and Mohan (2004). In spite of this, we find some elements of successful participation in both countries, which is strong evidence for the insufficiency of earlier theories.
Our arguments in this paper were developed through extended fieldwork in India and Mexico, and thus, were developed through a process of dialogue between field observations and theories about participatory governance that is widely discussed among ethnographers as abductive reasoning (Van Maanen et al. 2007;Agar 2010;Watson 2012), and among case study researchers as heuristic case studies (George and Bennett 2005). Our research initially focused on the importance of participatory institutional designs for explaining the outcomes of environmental programs (Fleischman 2014;Rodriguez Solórzano 2014). However in both countries we saw that theories focused on institutional designs were insufficient to explain the processes of participatory politics that we observed. Furthermore, in spite of the two countries' very different institutional development, we saw deep commonalities in the processes that led to successful participation. Although these commonalities are suggestive, our limited number of case studies means that we cannot claim broad generalizability for the theories we develop.
Field research in Central India was conducted by the first author on repeated visits to the region between 2009 and 2014, in which the first author conducted semi-structured interviews with over 300 social activists, scholars, and government officials (Fleischman 2012(Fleischman , 2014. The focus of these interviews was on national and regional level policy-making. Although the first author learned of the cases presented through this fieldwork, the accounts here rely primarily on secondary sources, including newspaper articles, which verify the author's own observations and tie them back to other published sources. The second author conducted field research in repeated visits between 2007 and 2015 in the Calakmul Biosphere Reserve located in southern Mexico. The data collection entailed structured and semi structured interviews with over 300 individuals, including a nested random sample of farmers from 35% of the communities in the region who were interviewed once in 2007 and a second time in 2014-2015, as well as key informants, such as community and regional leaders, and representatives of government and non-governmental organizations. The focus of these interviews was on understanding the relationship between local politics, conservation, and climate adaptation (Rodriguez Solórzano 2014; Rodriguez Solorzano and Fleischman 2018). The interviewees provided information about participation in decision-making forums at the community and regional level. The second author also personally observed interactions in these councils. As with the Indian case, we draw on secondary literature to provide broader context for understanding participation in Mexican forest management.

Participation supply in a low demand, low skill environment: failures of India's forest institutional reforms
Our first example of participation reveals a common story in participatory reform: New institutions are created to foster participation without considering whether people demand or have the capabilities to participate. India's forests are owned and managed by state forest departments created during and after the colonial era through a process that involved expropriation of rights held by rural people to utilize forests to support their livelihoods (Gadgil and Guha 1992). There is a long history of resistance against forest department authority, including occasional allowances for increased decision-making by forest villagers (Baker 1984;Guha 1989;Sastry 1989;Agrawal 2005;Sivaramakrishnan 2009). However legal participation in forest governance was mostly limited to forest officials. Villagers continued to use forests to produce their livelihoods in ways that contradicted legal mandates, and this brought them into conflict with authorities. The forest departments were often ineffective at enforcing new laws, creating an openaccess regime that contributed to forest degradation (Bromley and Chapagain 1984;Guha et al. 1984;Pathak 1994).
Since 1990, two reforms aimed to address these conflicts. First, Joint Forest Management (JFM), introduced at a large scale in the early 1990s, allowed village committees to participate with forest officials in making decisions about local forests. Villagers were expected to protect forests from overuse, in exchange for some combination of usufruct rights and revenues from timber harvests (Poffenberger and McGean 1996). Second, The Scheduled Tribes and other Traditional Forest Dwellers (Recognition of Forest Rights) Act, 2006 (FRA) allowed certain classes of people to claim rights to use and manage forests. Although this reform is generally understood to be about rights, rather than participation, the claiming and operation of rights is supposed to be through a participatory process (Duraiappah and Muñoz 2012;Maharashtra CFR-LA 2017). While both reforms, and particularly the FRA, were the result of popular mobilizations (Kumar and Kerr 2012;Maharashtra CFR-LA 2017), the reach of these mobilizations in village societies was limited (Kashwan 2013;Barnes et al. 2016).
Nearly all evaluations of both JFM (Sundar et al. 2001; Springate-Baginski and Blaikie 2007; Lele and Menon 2014) and the FRA (Saxena et al. 2010;Bose 2013;Kashwan 2013;Kumar et al. 2015;Maharashtra CFR-LA 2017) have found that neither programs' participatory promise has been met on a wide scale. Although several reasons have been identified for this failure, including weak institutional design, power imbalances, and problems of implementation, two elements recur in many explanations. First, many villagers were not interested in the opportunities presented by these participatory programs -they did not demand participation. For example, villagers in Tamil Nadu showed little interest in JFM because the forests they were given the rights to co-manage contained few valuable resources (Matta 2003;Matta and Kerr 2007). In principle these forests might improve and produce valuable products decades in the future, but this exceeded the discount rate of the villagers. In other cases, although forest resources were valuable, the actual rights assigned to local people were insufficient to generate real interest on the part of villagers (Sundar et al. 2001;Springate-Baginski and Blaikie 2007;Lele and Menon 2014). Along similar lines, Kashwan (2011) describes how village leaders in Gujarat who benefitted from existing JFM arrangements often opposed the granting of more extensive rights under the FRA, which might be beneficial to the village as a whole, but would undermine their personal power, illustrating how the interests of elites can conflict with the interest of villagers and thereby suppress political participation.
A second shortcoming was that villagers lacked capabilities to participate fully in the decision-making process. Kashwan (2015) describes how village JFM leaders failed to resist a recentralization of power over JFM resources. These village leaders believed that the power of the forest department was so much greater than theirs that to resist would mean both certain defeat and loss of patronage benefits they obtained by remaining department allies. Similarly, Balooni et al. (2010) describe how inter-elite conflict in a village in Haryana made the village unable to sustain collective action necessary to participate in the JFM program. While it is probably accurate to say that the primary constraint on the success of forest policy reforms in India has been poor institutional supply, the problems illustrated here imply that even an improved institutional design will not automatically lead to improved participation.

Demand and capabilities can drive institutional supply: the case of Mendha
On the other hand, several case studies have emerged from India demonstrating that even when participatory opportunities are limited, well organized communities that demand participation and have high levels of capabilities sometimes organize to overcome limits in institutional supply. Although similar examples have been documented in other regions of India (Sundar et al. 1996;Singh 2001Singh , 2002D'Silva and Nagnath 2002;Sarin et al. 2003;Van Laerhoven 2013, 2015), we focus on the village of Mendha in Maharashtra. One of the key lessons from this case is that in the context of imperfect institutions such as the JFM and FRA described above, which failed to foster effective collective action in most situations, collective organization and external networks -two aspects of what we term capabilities -may be a prerequisite for effective participation. This is similar to arguments made by Han (2014) about the development of activists in the United States, however the villagers of Mendha live in a far more resource poor environment, and thus their external networks are even more crucial. Many authors have noted that this type of strong collective organization remains rare in rural India (Harriss-White 2003;Chatterjee 2004Chatterjee , 2011Corbridge et al. 2005Corbridge et al. , 2013Witsoe 2013; Harriss-White and Prosperi 2014). Mendha's level of collective organization is unusual in the Central Indian context, and thus the experience of the village should be seen as an illustrative, albeit not unique, exception.
The village is surrounded by government forest land which produces valuable timber and non-timber forest products. Influenced by village leaders who had been involved in regional political activism, this community began working together to make community decisions in the early 1980s. The village's high levels of collective action attracted the attention of Gandhian social activists and small NGOs who began visiting the village regularly in the early 1990s, both to learn from the village's success, as well as to share information and resources with the community as part of an organized study circle (Ghate and Chaturvedi 2004;Pathak and Taraporewala 2008;Tofa and Hiralal No date;Singh No date).
One of the problems villagers in Mendha identified was degradation of the forest surrounding the village. In 1987 they began an independent campaign to protect the forest from overharvesting and overgrazing, efforts that brought them into direct conflict with the forest department (Ghate and Chaturvedi 2004). Through their informal network with local NGOs, villagers came to know about Maharashtra's JFM program when it began in the mid-1990s. Because the village was already working to protect its forest, villagers believed that they were a perfect fit for the JFM program. However the forest department refused, because at that time Maharashtra's JFM rules only allowed the forest department to make JFM arrangements in degraded forests. Because Mendha's villagers had already been protecting the forests for several years, the forests were no longer degraded.
In a village with less capability than Mendha, this would have been the end of the discussion, however Mendha successfully mobilized to oppose the denial of JFM in their village, both by drawing on their network of social activist allies, as well as by directly threatening the forest department: they announced that if it was necessary to have a degraded forest in order to obtain benefits from the government, the villagers were willing to degrade their forest so that it would qualify for JFM. After an extended struggle in which villagers successfully translated strong collective action into political power, the villagers were able to work with sympathetic forest officials to create an official JFM committee in the mid-1990s (Ghate and Chaturvedi 2004;Pathak and Taraporewala 2008;Tofa and Hiralal No date;Singh No date). According to the written account of Tofa and Hiralal (respectively the moral leader of Mendha and an external NGO ally who has worked in the village since the late 1980s), Mendha did not "naturally" have a high level of collective capability, but developed its political capabilities through concerted work on the part of villagers, who participate in regular study circles, discussing among themselves and sometimes with members of their external network, how they can improve the quality of life in their village while maintaining cultural traditions (see also Ghate and Chaturvedi 2004).
This initial triumph was followed 20 years later by a similar triumph with regards to the implementation of the FRA. Although the FRA promised to allow communities which met certain criteria to claim rights to manage, harvest, and sell non-timber forest products within the forests which were traditionally held by the village (Government of India 2007), this provision was not widely advertised in many states, including Maharashtra (Saxena et al. 2010). However, due to its network with regional activists, Mendha was aware of and wished to take advantage of the provision because they wished to control the valuable bamboo harvest which was conducted annually by the forest department with no benefit to the village. Their attempts to claim community rights over bamboo management were stymied for several years until they and their allies were able to mobilize the national minister of environment and forests to come to their aid, again illustrating the reach of the villagers' capabilities (Das 2011). In 2011-2012, as a result of the newly granted rights, villagers in Mendha were able to auction the bamboo for a price of over 10,000,000 rs (roughly US $200,000) (Subramanian 2012). In subsequent years, many villages in the surrounding district have learned from Mendha's example, and attempted to sell their own bamboo. These villages also had to fight with the forest department and struggle to learn how to conduct auctions, however by 2017 most bamboo sold in the bamboo-rich district was auctioned by communities who have rights under the FRA, leading to significant profits for well over 100 communities (Maharashtra CFR-LA 2017).
The experiences of Mendha with JFM and the FRA illustrate that well-organized, highly capable rural people who demand participation may be able to overcome limits on the supply of participatory institutions through political activism. It is important to point out, however, that Mendha's residents needed some supply of participatory institutions to validate their collective action. Without JFM and later the FRA, the villagers would have had no formal rights to the forest they were protecting. Under the pre-reform regime, the forest department would have considered the village's forest ripe for timber harvest -destroying the work of the villagers, and giving them no benefit. The villagers of Mendha wanted to be a part of JFM because it insured formal government recognition of their work. Furthermore, the institutions of the FRA were a necessary precondition of their bamboo harvesting, even if it was not sufficient without the organizing efforts of the community. At the same time, the benefits the villages have received may be limited if participatory rights are not embedded within a broader context of market and political power. According to interviews conducted by the first author in 2014, after 2012 local bamboo buyers formed a syndicate to artificially suppress auction prices, making it difficult for the villagers of Mendha to profit from their rights. Additional collective organizing across multiple villages was necessary to overcome this barrier, and while this existed in Mendha's district, neighboring areas have been much less successful at exercising forest rights (Maharashtra CFR-LA 2017). This points to the need for participatory initiatives to be embedded within broader agendas of democratic deepening which may be driven by bottom up collective action combined with top-down institutional reforms.

Long-term synergies between institutional supply, demand, and capabilities: institutional reform in the Calakmul biosphere reserve
In the long-term, the three legs of the participatory stool are synergistic. Institutional supply can help communities that lack demand and capabilities to gain by showing them how they can effectively utilize participatory opportunities. Once communities have developed these skills and demands, they are more likely to turn the tools available to them into participatory instruments. We illustrate this long term synergy by examining the history of participation in the Calakmul Biosphere Reserve, in Campeche Mexico.
Calakmul is an unlikely case for successful participatory governance. Most of the people currently living in Calakmul arrived as impoverished settlers from other parts of southern Mexico in the 1970s and 1980s (Haenn 2005). Taking advantage of Mexico's land reform (Assies 2008), they established small communities and practiced swidden cultivation of maize for subsistence and chili for commercial purposes (Turner et al. 2004). Given the marginal nature of land in the region, most settlers were people who had few other options in life, had limited capabilities, and had few ties to others in the region apart from the fact that they had happened to choose the same place to settle. Furthermore, formal government institutions were largely absent from the region. Unlike many other parts of Mexico, where the 1980s saw growth in civil society activity (Fox 1994(Fox , 1996, and where subsequent years have seen the growth of strong community forestry institutions (Bray et al. 2003(Bray et al. , 2006, Calakmul lacked a history of peasant organizing or social solidarity based on a shared cultural history. Until recently, local politics were dominated by the Institutional Revolutionary Party (PRI), which used its control over rural populations to maintain its hegemonic status both within Campeche state and the nation as a whole.
Mexican policy towards Calakmul was transformed in the late 1980s and early 1990s by four changes (Rodriguez Solorzano and Fleischman 2018). First, the Calakmul Biosphere Reserve was created in 1989. Mostly unsettled land was designated as a nucleus zone, off limits to future land clearing or extractive activities, while land already allocated to a community was designated as part of a buffer zone. This decision was made in the top-down fashion typical of the time, without local participation (Ruiz-Mallén et al. 2015;Pfaff et al. 2017). Second, the Regional Council for Agriculture and Silviculture of Xpujil (CRASX for its Spanish acronym, Xpujil is the principal town in Calakmul) was created in 1991 (Acopa and Boege 1997). This Council was created through a federal government initiative that aimed to foster governance outside of the boundaries of formal local government. CRASX gave local leaders management authority over money flowing into the region from the central government, as well as from international aid and conservation organizations interested in biodiversity conservation. Third, laws passed in 1992 ended the land distribution program, helping to cement the government's new commitment to land conservation. Finally, a new council named CRIPX (Regional Popular and Indigenous Council of Xpujil) and Calakmul municipality were created in 1995 and 1996 (Haenn 2005;Calakmul 2012;CRIPX 2012). The municipality overlapped the biosphere reserve boundaries and provided a formal local governance structure. CRIPX emerged as a group of indigenous people who claimed to be excluded from CRASX and who demanded participation in the distribution of resources CRASX controlled.
CRASX was the first institution created in Calakmul that fostered participation in collective decision-making. It was created to bring development and foster natural resources conservation in the region based on a model of inclusion, although fostering clientelistic relationships between the PRI, the biosphere reserve, and local leaders was also a goal (Haenn 2005). CRASX allowed leaders from the communities to make decisions about the use of resources coming to the region that had previously been made by remote government officials. These new leaders had no previous experience participating in decisionmaking or managing funds outside of their communities. Don Deocundo Acopa, appointed by the central government to make CRASX work, was aware of these limitations, and thus the activities of CRASX were designed to include capability-building activities and to teach people about the importance of conservation. Acopa reflected on the experience in a book chapter, "In general, these organizations (CRIPX and CRASX) served to transmit information, provide technical assistance through contracted staff, create a united front during negotiations over forest policy, supply the organizational vehicle for funding by government and NGOs, promote all of the activities mentioned above, and provide a forum where peasants can learn parliamentary techniques and build their negotiation skills -even if negotiations do sometimes end up in fights" (Acopa and Boege 1997, 95). Thus CRASX supplied participation, but also increased the capabilities of those who participated.
CRASX pursued participation but it was highly imperfect. In fact, CRASX had a goal of fostering support for the ruling party and the biosphere reserve, and excluded people who lacked political connections or were skeptical of the conservation and development model favored by Acopa. Because people saw that those who participated in CRASX were receiving benefits, people who had not initially participated in CRASX demanded benefits as well, through a series of well organized protests. That demand led to the creation of CRIPX, the erosion of power and influence of CRASX and the creation of a formal municipal government. These events are described in much greater detail by Haenn (2005), who details how the creation of CRIPX and later the municipality were in part reactions to organized protests against the power of CRASX, and in part also attempts by PRI leaders to move away from the conservation model favored by Acopa. The organizers of CRIPX specifically drew on resentments against the leaders in CRASX, thereby showing how the expansion of participatory opportunities increased demand for participation even among those who lacked formal participatory opportunitiesand how these demands in turn led to the creation of new participatory programs. The creation of a municipal government decreased CRASX' power and thereby ended its training element, but created a more formal participatory opportunity through the democratically elected municipal government.
CRASX' influence faded away and CRIPX was never as influential as CRASX after the creation of the municipality in 1996. Yet, the community leaders whose participatory capabilities were enhanced through their experience and training in these organizations remain the dominant political actors in the now-formalized politics of the municipality. Their participation in these councils taught them the benefits of participation and enhanced their capability to be involved in decisionmaking and to find spaces for participation. Many of these leaders serve at the local level as formal or informal leaders of their ejidos. Some have also occupied elected or appointed positions in the municipal offices, or have served as leaders in the Municipal Council for Sustainable Rural Development (CMDRS), established in the early 2000s as a multi-stakeholder forum for spreading information and resources in the region (GTZ, Ayuntamiento, and CONANP 2006). The lasting impact of this early supply of participatory opportunities and training has been to create a cadre of leaders who continue to provide the backbone of political activity in the region. Thus, while the initial achievements of the program in the 1990s were limited by the creation of the municipality and subsequent withdrawal of local budget management authority and learning opportunities for local people, its long-term consequences on public participation in decision-making have been significant, particularly as the Mexican state has become more broadly responsive to democratic political demands (De La O 2015;Díaz Cayeros et al. 2016;Rodriguez Solorzano and Fleischman 2018).
Together the Mexican government and local people in Calakmul have put the three legs of the stool together. The government put the first leg using foreign and national resources to supply participatory institutions -as well as through a broader period of reform which saw Mexico move from a strongly authoritarian state to an imperfect but functional political democracy. Aware of local people's limitations, the government reinforced the supply with capability building, which provided a second leg and increased the chances of successful participation. Local people, who realized participatory benefits were large and who have developed the ability to participate, put the last leg on the stool, demanding the creation of further participation opportunities, nurturing subsequent councils and a now vibrant local political scene. The Calakmul municipality, and the ejidos within it, hold politically contested elections in which leaders compete through offers to improve the provision of public goods to rural communities. As a result of this political competition, the delivery of water, education, health, and social welfare programs has increased dramatically in the region, a fact commented on by many of our informants. This political scene allows for far more participatory opportunities than were provided in the past and provides robust empirical evidence about the evolving nature of participation and the synergies between participation supply, demand and capabilities.
Our recognition of the great steps that Calakmul has taken does not mean that participation in Calakmul is complete or perfect. Individuals and communities that have been more effective at playing the participatory game have received more benefits, resulting in rising inequality among the formerly uniformly poor population. Certain subgroups remain excluded (Navarro-Olmedo et al. 2016). Many government programs are plagued by corruption. And the biosphere reserve frequently makes management decisions without involving the community or elected government. For example, UNESCO recognized Calakmul as a mixed world heritage site in June 2014. In a survey of a nested random sample of 280 households from 28 communities (34% of the communities in the region) conducted a few months after this event 80% of the interviewees had heard about this recognition. Most of them received this information only after the regional television news covered the Mexican president's post-award visit to the site. Despite of the awareness, 52% of the interviewees said they had no idea what the recognition meant and only 34% thought the recognition would benefit them.

Discussion
Both the arguments of the enthusiasts and critics of participation fail to explain our cases. All of the institutional designs we examined in this paper were deeply flawed, yet in circumstances of sustained political mobilization, these flawed designs sometimes contributed to political empowerment. In contrast with widely cited arguments, we found that under certain circumstances, very imperfect, nonradical participatory reforms enabled significant increases in citizen engagement.
We argue that attention to the participatory stool helps to explain this divergence. Our case studies suggest that institutional supply, demand for participation, and capabilities to participate may be jointly necessary for stable and effective participatory programs. When institutions were supplied without demand or the existence of a capable public, as in the case of participatory forestry reforms in India, they were not successful at fostering participation. By contrast, when villagers demanded and had the capability to participate, as in the case of Mendha, they forced the government into allowing them to participate, albeit in programs that already existed but had been poorly implemented. The two key distinctions between Mendha and the many unsuccessful JFM villages were, first that Mendha's forests possessed sufficiently valuable resources that the villagers demanded that they obtain benefits, and second, that Mendha's villagers had capabilities -in the form of strong collective organization and linkages with influential external actors -that enabled them to make those demands heard. In Calakmul, several decades of participatory reforms which gave participants training and access to decision-making authority that was meaningful, have led to high levels of public participation in decision-making. Once again, valuable resources have helped create demand for participation, while a concerted effort of capability-building has enhanced the capabilities of local residents. Like Han's (2009Han's ( , 2014 American activists, political leaders in India and Mexico learn from political experience to be more effective political activists. Our argument extends Han's work, however, because we emphasize the ways that state institutions, and not merely participation in NGOs or political organizations, can provide the training for the development of these capabilities and demands. Legacies from past interventions have a strong influence on opportunities for future participatory reforms. The limited participation opportunities afforded by JFM in most parts of India convinced many local leaders that the most effective strategy of working with forest officials relies on close collaboration and avoiding political challenges, a legacy that has carried over to affect the implementation of the FRA in some parts of the country (Kashwan 2015). By contrast, past successful experiences with participation underlie the success experienced in Mendha and Calakmul. These people gained tangible skills that they deployed when new institutions arose. While some of the early participatory opportunities, including JFM and CRASX, were limited and paternalistic, their existence helped build the capability of these communities to engage with larger scale participatory opportunities. While citizen engagement might have been enhanced more rapidly with more radical reforms, this finding supports our claim that participatory reforms need not be dramatic or radical to have the potential for long-term positive effects. In fact, while many more radical-seeming reforms around the world have faltered, modest reforms sometimes set off a cascade that led to radical political changes (Corbridge 1998). At the same time, it is not clear why villagers in Mendha were able to develop these capabilities and demand while so many other villages, including the village immediately next to Mendha (see Ghate and Chaturvedi 2004) did not.
Our cases suggest several hypotheses that, if supported by further testing in a broader array of contexts, could serve as guidelines for facilitating participation in environmental programs. First, it appears that participatory programs which provide motivation in the form of substantial benefits to those who participate are more likely to succeed. In the case of JFM in India, attempts have been made to fund "entry point activities" -often unrelated to forestry -with the goal of providing villagers with incentives to participate. The problem with these activities is that interest in participation dries up as soon as funding disappears. Those villages that have sustained long-term investment in JFM are frequently those, such as Mendha, which can generate meaningful revenue from the harvest of timber and non-timber products over long time frames, or who have strong cultural connections to the forest. In Calakmul people who participated in CRASX and its successor councils and government structures have sustained participatory interest because they control funds that are sufficient to make a significant difference in the livelihoods of individual farmers and communities. Today, participants in the CMDRS, which has limited direct control over resources, are motivated because participation in the council provides information about what other programs are available that can bring such benefits, and enables people interested in politics to build their political networks.
Second, our cases suggest that participatory capabilities can be taught, even to illiterate and impoverished people. In the case of Calakmul, CRASX, CRIPX and the CMDRS have made efforts to educate people how to run and participate in meetings and understand government budgeting guidelines. In Mendha, where villagers had autonomously begun working together cooperatively, involvement by local social activists and NGOs, who were more educated than the villagers, helped the villagers to recognize barriers to full participation and led, for example, to more full inclusion of women in decision-making processes (Tofa and Hiralal No date). Mendha's neighboring villages learned from Mendha's example and have also emerged as leaders in participatory forestry. A wide variety of actors, with a wide variety of political goals, may play important roles in this educational process.
Although participatory capabilities can be taught, people have past experiences and an existing supply of capabilities which affect their reaction to any particular participatory intervention. People who were involved in CRASX and CRIPX were well situated to take advantage of -and demand -further participatory opportunities. Thus, past participation can create a virtuous cycle in which people who have participated learn that participation is valuable, demand greater opportunities, and thus help to create an enhanced supply. Both of our cases take place in countries undergoing longer term processes of democratic deepening, and this may help reinforce the virtuous cycle. This virtuous cycle has its counterpart in a vicious cycle which reinforces exclusion: people who have not had participatory opportunities will not demand participation, and may fail to take effective advantage of new supplies of participatory opportunities, leading to what Mansuri and Rao (2013) call "civil society failure." The vicious cycle may be further reinforced when so-called participatory programs require substantial investments, but bring few benefits, as has been the case with many JFM programs in India. People learn from these programs that they should not demand participation.
Third, along with teaching participatory capabilities, creating networks that connect local leaders with people with higher levels of power and capabilities may be an important element of capability enhancement. Many of the successes in Mendha are the result not only of the village's collective action, but also of its strong networking with educated activists in nearby towns and in distant state and national capital cities. Similarly, in the case of Calakmul, networks between national political parties and local leaders were crucial in the original development of CRASX and CRIPX, as well as in the continued delivery of opportunities to local participatory programs.
Fourth, a supply of participatory institutions may be necessary for successful participation, but in cases where people demand and have capabilities for participation, imperfect institutional designs can in part be overcome. Cornwall's (2004) distinction between invited and claimed spaces is relevant here. While citizen engagement may be more straightforward in the context of spaces in which citizens are invited in by the state, when demand is high, groups of people may claim new sites for citizen engagement. The case of Mendha demonstrates that an imperfect institutional design can be adapted by a determined group of citizens to serve their purposes. CRIPX, one of Calakmul's important participatory spaces, emerged directly out of protest. While it is probably the case that a complete absence of participatory institutions makes it unlikely that participation can be achieved, in imperfect cases, citizens who are empowered by their demands and capabilities can engage in the contentious politics of resistance to remake the participatory institutions, while providing participatory institutions alone may not quickly result in participation if people lack demands and capabilities.
Fifth, there appear to be interactions over longer time frames between the three legs of the participatory stool. While supplying participatory institutions may not quickly create participation, if those who participate initially receive some formal or informal instruction on how to do so successfully, they may be more able to participate in future efforts. If others see that those who participate receive tangible benefits, demand may increase in the future -in fact, this is precisely what happened in Calakmul. Initial efforts made by CRASX led to broader increases in participation as early participators benefited and gained skills and those excluded demanded inclusion once they realized participation was worthwhile. Our argument here is similar to that made by Lund and Saito-Jensen (2013) regarding elite capture: snapshot views of participatory processes may obscure long-term dynamics which can have either positive or negative consequences. There is an essential interaction between structure and agency in participation, and while a supply of participatory institutions is most certainly necessary for long-term success of participatory programs, an engaged and active citizenry struggling to express themselves is also a necessary condition for success (Ribot and Peluso 2003;Brockington 2008). Yet most evaluations of participatory programs, including those we have critiqued, focus on short-term outcomes, and thereby miss the longer development of citizenship that participatory programs can support -or inhibit.
This interactive process may be a crucial element in the process of making citizens -and may thus be analogous to arguments about the role of government institutions in re-shaping subjectivities (Agrawal 2005) as well as Fung's (2012) iterative process of "continuous institutional innovation." A participatory program that works well in one context may need to be reshaped as capabilities and demands are changed by participatory experiences. Examining this longer time scale requires scholars to look more deeply at cases of participatory failure to understand what citizenship skills are being learned through participation. Short-term failures to improve service delivery, as are common in participatory programs (Fox 2015), may be offset by longer term learning -although these failures may also reinforce a non-participatory equilibrium.

Conclusion
The case studies examined here support the idea that successful participatory governance requires not only a supply of institutional opportunities for participation, as emphasized in recent literature in development studies, but also demand and capabilities for participation from citizens, factors largely ignored in literatures on environment and development, but strongly emphasized by political scientists studying developed democracies. Supply, demand, and capabilities thus form the three legs of the participatory stool -required to gain citizens a seat at the decision-making table. Our case studies illustrate that there is a strong potential for synergy between these components, but also show that participatory efforts can flounder in the short-run if they do not attend to the need for citizens to demand and have capabilities to participate. They further show that participatory programs can have an educational component -often ignored in literature on participatory programs -providing people training to make them more effective citizens in the future -or conversely, reinforcing their own disempowerment and exclusion. Further research is needed to sort out these dynamics. While our cases are suggestive, they represent a limited set of contexts, and we are not able to test the theory we have generated from these case studies, nor to evaluate its generalizability to other contexts. Future research on participatory governance should focus on understanding the interactions between institutional context and the agency of individuals.
The participatory stool can be used to improve the incorporation of participation into development programs, but practitioners should proceed with caution. Consistent with arguments made by Fox (2015) with regards to social accountability, participatory programs appear more likely to achieve success if the opportunities they supply provide benefits that people find valuable, and are built with goals that move beyond simple participation to empowering people with capabilities that will allow them to be more effective participants. Put in other words, participatory programs that value citizenship may be more likely to achieve genuine participation than those focusing exclusively on the instrumental goals of participation. Valuing citizenship means placing real authority in the hands of citizens, and providing them with opportunities that help them build their own stock of capabilities for participation. Programs that do this successfully may not always lead to the goals that environmentalists prefer because empowered citizens may disagree with environmentalists about both the means and the ends of society. In our cases, while the Indian JFM reform was pushed by donors including the World Bank and the Ford Foundation, the successful case of Mendha relied on an alliance between organized villagers and activist NGOs. Similarly, while early participatory interventions in Calakmul were initiated by political leaders as means to enhance their political authority, they also took advantage of the abundant support from international organizations for participatory conservation in the region.
A further challenge to the simple application of the participatory stool lies in the problem of history. Just as we illustrated that a virtuous cycle can emerge in which the legs of the participatory stool reinforce each other and increase participation, they can also create a vicious cycle, in which lack of participation reinforces disinterest and weak institutions. We suspect that such vicious cycles may be more likely in authoritarian states. Furthermore, the dynamics of power within and between communities shape the outcomes of participation (Lund and Saito-Jensen 2013). Local people bring this history of their past experiences and local political dynamics to the table with them when they engage in a new participatory process. Environmental programs designed without awareness of these historical dynamics are likely to have surprising -and potentially very disappointing results. On the other hand, our cases show that while radical reforms sometimes fail to bring about radical change, modest reforms may sometimes lead to radical change.