Unpacking ‘gender’ in joint forest management: Lessons from two Indian states

Gender inequalities and social exclusions in community-based forest management have garnered attention, particularly in South Asia. Yet, framings that homogenize women and marginalized groups fail to capture the nuanced processes by which such exclusions occur. Despite provisions for women in local community man- agement institutions, numerous constraints hinder their active participation in forest governance. Understanding participation in JFM requires attention not only to gender, but also to the diversified interests and experiences women hold and the unequal power relations in which they are enmeshed. Based on 85 semi- structured interviews with women and men farmers, JFM committee members, local authorities, NGO staff and Forest Department officials, we explore emic perspectives of how social differentiation shapes participation in JFM. We compare the situation in Karnataka, a wealthier Indian state that is considered exemplary for JFM, with that in Madhya Pradesh, a poorer Indian state inhabited by tribal populations, where JFM is poorly functional. We show that exclusions in Uttara Kannada occur along gender and caste lines, whereas among tribal groups in Mandla, women of certain ethnicities are particularly disadvantaged in JFM despite their extensive forest use. Grouping marginalized groups into homogeneous categories (e.g. as Scheduled Tribes or Scheduled Castes), as do Indian laws addressing tribal issues, deters focus from the inequalities that occur among groups, and from their relevance in shaping local experiences. Place-based environmental and political economic histories further shape local interests and participation in JFM. We argue that a focus on gender is necessary but not sufficient to understand social exclusions in JFM, and that gender must be understood in relation to other factors of social differentiation.


Introduction
Since the mid-to late 1980s, there has been a decentralization of forest governance globally, with increasing co-management rights and responsibilities devolved to forest-dependent communities (Agrawal et al., 2008). Collaborative governance arrangements have emerged, typically premised on the idea of a sharing of power, responsibilities and benefits between governments and resource users, with the latter participating in decision-making processes (Ansell and Gash, 2008). Such arrangements are intended to enhance not only the efficiency but also the inclusiveness and equity of natural resource management processes (Ribot et al., 2010).
As the promise and perils of these arrangements have come to the fore, so too has women's global underrepresentation in the user groups that co-manage forested areas (Sunderland et al., 2014). The capacity of different actors to influence how local forests are managed hinges on their social position, which is shaped by their gender, caste or ethnicity, age, socio-economic status, and other attributes (Varughese and Ostrom, 2001;Blaikie, 2006;Mukherjee et al., 2017). Notably, social structures that mediate gender roles, expectations and relations pose constraints to women's active participation in local governance. Hence, even when women are formally represented in user groups, their presence often remains nominal and their participation limited and lacking influence (Agarwal, 2010).
India is home to the largest forest-dependent population in the world, and is considered a forerunner in developing policies to engage forest dependent communities in the management of forest lands. It has created formal measures to provide space for women and marginalized groups to participate in forest co-management schemes, including through its Joint Forest Management (JFM) programme. In JFM, efforts T to engage women and historically marginalized groups are manifest in rules that endow membership in forest management institutions to all adult residents of forest-dependent communities, and quotas for these groups in the Executive Committee (EC) of forest user associations. In India and beyond, however, the active engagement of women and marginalized communities remains the exception rather than the norm (Nightingale, 2002;Agarwal, 2010;Persha and Andersson, 2014). Understanding this process of exclusion is critical to foster more equitable arrangements that safeguard the rights of, and offer benefits to, marginalized groups.
A significant body of research has called attention to gender and social inclusion within the JFM context. Yet, this work, and the practice on which it reflects, has often conceptualized women as a unified group, rather than diversified actors with differentiated interests and experiences. As Godbole (2002, p.3) indicates, gender concerns in India's JFM programme remain a 'local and depoliticised issue and related to an undifferentiated category called women'. The rhetoric about women's role in JFM is minimally present within implementation, and some authors suggest that there is an intentional lack of clarity regarding women in the programme (Jewitt, 2000;Agarwal, 2001;Sarker and Das, 2002), similar to the lack of clarity regarding marginalized forest dwellers. In only few cases (e.g. Agarwal, 2000;Nightingale, 2002) are caste and ethnicity acknowledged as other important social factors interacting with gender to shape women's interests and engagement in JFM, and empirically-based analyses of these processes are scarce. The relative silence around intersectionality in the forestry sector stands out against the rich engagement with questions of caste and gender-and their enmeshment-in an Indian context (e.g. Purkayastha et al., 2003;Rao, 2003;Rege et al., 2013;Devika, 2010). It critically limits our understanding of the processes of social exclusion that characterize JFM and the management of other common property resources. A situated, emic perspective on why local women and men participate in JFM can shed light on how social position, as experienced through the daily life of participants, affects interest and capacities to participate, and how individuals exercise their agency to open up spaces of participation.
Hence, this paper draws on the narratives of local women and men to examine how social differentiation shapes participation in JFM in the Indian states of Karnataka and Madhya Pradesh. In Karnataka, which is considered an exemplary state for JFM, we demonstrate that caste and socio-economic status, which are linked to landholdings, play a critical role in determining livelihood strategies and rural dwellers interests' in the forest and in JFM. In Madhya Pradesh, home to a large number of tribal peoples, where JFM is poorly functional and forest areas increasingly degraded, uneven relations among different tribes and castes underpin exclusions in JFM. In both states, gender significantly affects capacity to engage in JFM due to norms that relegate public affairs to men and exclude women, particularly when lacking formal education. In conclusion, we argue that the focus on gender is necessary but not sufficient to understand differentiated interests, constraints and opportunities in JFM.

JFM in historical perspective
Current day perspectives on and participation in JFM must be understood within a historical trajectory of appropriation of the forest by the state dating back to colonial times. In the late 19th century, colonial systems that declared uncultivated commons as state-owned forest land were focused on commercial exploitation and control of the forest, and not on forest-dependent communities (Gadgil and Guha, 1993). In 1878, the Indian Forest Act, which gave way to the present-day Indian Forest Act (IFA) (1927), classified the landscape into 'Reserved', 'Protected' and 'Village' Forests (Agarwala, 1985). Reserve Forests offered no rights to local people, unless otherwise specified; Protected Forests offered them rights to pursue all activities, unless explicitly prohibited; and Village Forests were designated to meet local needs (though no lands were set aside under this designation during the colonial period) (Guha, 2001;Sarin et al., 2003). At this turning point in India's history of forest management, the colonial forest department reserved more than one fifth of India's forests, appropriating land from customary users through a forest 'settlement' process, which recorded the settlement rights of certain forest users (Gadgil and Guha, 1993).
The post-independence period saw the continued capture of forests by the state, justified in terms of conservation and wildlife protection goals, despite unabated state-led commercial felling (Guha, 2001;Springate-Baginski et al., 2013;Kumar et al., 2015). Valuable nontimber forest products (NTFPs) were nationalized, and the central government consolidated power over forest management in relation to individual Indian states (Sarin et al., 2003). The Forest Conservation Act (FCA) of 1980 shifted the power to grant tenure rights to customary users from individual states to the central government. Where customary rights had not yet been recorded, lands were listed as government forests, and 'illegal encroachers' were put at risk of eviction (Sarin et al., 2003).
In the late 1970s and 1980s, 'farm forestry', 'social forestry', and more participatory forest management systems emerged amid pressure from social movements and global environmental concerns, rapid deforestation and degradation, and heightened awareness and advocacy around land ownership and forest rights. Departing from previous policies focused on maximizing forest-based revenue for national interests, the Indian National Forest Policy of 1988 represented an attempt to reconcile forest conservation and livelihood goals, recognizing forest dwelling communities as vital to forest landscapes (GoI, 1988;Sundar et al., 2001). The policy was meant to foster the participation of forestdependent people in the management of state-appropriated forest lands, and gave way to the JFM approach in 1990 (Pratap, 2010;Springate-Baginski et al., 2013).
JFM offers no legal sanction to secure the land tenure and rights that are central to community management of natural resources (Saigal, 2003;Springate-Baginski et al., 2013;Kumar et al., 2015). The government retains legal ownership over the forest, and communities have limited usufructs rights to specified forest products for subsistence and sale (Murali et al., 2002). Hence, JFM co-exists uncomfortably with more progressive constitutional and legislative mandates to strengthen local rights and decentralize governance among forest-dependent communities. Such mandates include the Scheduled Tribes and Other Traditional Forest Dwellers (Recognition of Forest Rights) Act (FRA) (2006), which addresses the rights of Scheduled Tribes and Other Traditional Forest Dwellers (OTFD) to access traditional forestlands, use and sell non-timber forest products, conserve and manage community forest resources, and which stipulates the right of habitation and 'any other right traditionally enjoyed by these communities' (GoI, 2006). They also include the Panchayats (Extension to the Scheduled Areas) Act (PESA) (1996), that provides for local self-governance in Schedule V (tribal majority) areas. PESA imbues governing authority over a community's development and resources, including in communitybased forest management, to the Gram Sabha; the constituency of all adult voters of a self-defined community (GoI, 1996). In contrast, JFM establishes new village forest management or protection committees under the supervision of the Forest Department and does not refer to PESA (Sarin et al., 2003). These contradictions and historical and political trajectories contextualize local communities' current interests in jointly managing forests with the state.
in Madhya Pradesh-residents jointly manage delineated forest areas with the state. JFMCs or VFCs comprise a general body formed by all adult villagers who seek membership, for a symbolic fee, and an Executive Committee (EC) of 10-15 elected village representatives and a State Forest Department (SFD) official. In theory, the EC is the main decision-making body at the community level and is responsible for daily management decisions, drafting a forest management plan, and ensuring fair benefit sharing both between communities and the SFD and among local residents.
To promote equity in terms of decision-making and benefit sharing, the JFM handbook stipulates general requirements for the representation of socio-politically marginalized groups in the EC: if SCs and STs represent more than 10% of the village's population, each group must have one representative on the EC. If the two categories together represent 15 percent or less of the total population, the EC should include at least one SC or ST member. At least one landless member should be elected, and women should make up at least one third of the EC (GoI n.d., p. 9). There is considerable variation in state-specific JFM provisions, however, and several states have quotas within quotas, specifying the minimum number of women from marginalized groups who should occupy reserved seats. 1 Yet, quotas alone do not adequately address equity concerns. Scholars have questioned the value of 'threshold criteria' or reservedseat polices on the basis that they can encourage a tokenistic presence (Locke, 1999;Sundar, 2000;Mohanty and Sahu, 2012). A focus on numbers can conceal the hierarchies and social norms that privilege the participation of upper castes in public decision-making, while the interests of marginalized groups, who are usually the most forest dependent community members, remain unheard (Nightingale, 2002(Nightingale, , 2011Persha and Andersson, 2014).
In what follows, we draw on Agarwal's (2001) concept of participation as a six-rung ladder that moves from nominal or passive participation at its lowest end to high and active participation on the other. Moving up the ladder, participants' capacities to understand, have a voice in, and influence decision-making processes increase. An inclusive and empowered participation characterizes the top rung of the ladder, where institutional barriers are dismantled and diverse individuals and social groups engage in decisions that affect their lives. Based on this notion, Agarwal (2001) describes 'participatory exclusions' as the paradoxical situation whereby exclusions are experienced even within seemingly participatory institutions.

Conceptualizing intersectionality in an Indian context
Feminist theories on intersectionality can add analytical purchase to our understanding of inclusive participation in JFM. Intersectionality refers to how multiple axes of social differentiation, such as gender, age, ethnicity or caste, and socio-economic status, among others, intersect and co-constitute each other to create unique social locations (Crenshaw, 1989). The subjectivities thus created are contextually specific and dynamic, varying across time and place (Hankivsky, 2012). As systems of domination, oppression or discrimination cross, they are compounded, such that the grouping of 'the poor' or 'lower castes' or 'women' under-estimates the disadvantages of multiply disadvantaged women. Mollet and Faria (2013) calls for analyses that engage with intersectionality so as to generate a more complex and messier notion of gender, which explicitly accounts for race, racialization and racism. Analyzing the processes by which multiple relations of marginalization intersect and shape participation in community forestry requires us to 'specify the particularity that each inequality brings to each instance' (Walby et al., 2012, p. 235). For instance, in the United States, Latino and African American women considered that despite their inseparability, in many instances 'race' played a more significant role in shaping their marginalization than gender (hooks, 1990;Kobayashi and Peake, 1994). In a Nepalese (Nightingale, 2011) or Indian context, several scholars emphasize the intersection of gender, caste, class, religion, and age, challenging the conception of 'casteless gender' and 'genderless caste' (Rege, 2006).
Already one hundred years ago, Ambedkar (1916), the father of the Indian constitution, argued that strict adherence to endogamy (or marrying within classes), which maintains the caste system, is enforced through the control of women's sexuality and mobility. As Chakravarti (1993, p. 579) notes, The purity of women has a centrality in brahmanical patriarchy 2 […] because the purity of caste is contingent upon it […] The safeguarding of the caste structure is achieved through the highly restricted movement of women or even through female seclusion. Women are regarded as gate-ways-literally points of entrance into the caste system.
Restrictions on women's mobility are particularly prominent among 'upper caste' communities, which are considered the purest and most at risk of being polluted (Rao, 2003). Particularly among wealthier upper caste women, the withdrawal from non-domestic economic activities is a status marker and reinforces women's economic dependence on men (Kandiyoti, 1988). The social honour that accompanies such practices make women complicit in their subordination (Chakravarti, 1993).
Indian Scholars also underscore the overlap of caste with class, as the social structure that maintains people in hereditary (caste) positions and occupational classes has material consequences. For so-called 'lower caste' groups, discriminatory institutions limit not only economic mobility, but also educational achievements and participation in governance structures (Deshpande, 2000, Deshpande, 2006. Attempts to redress these inequalities have led to official designations by the Government of India for various historically disadvantaged castes, referred to as Scheduled Castes (SCs) and Adivasi (indigenous) peoples or Scheduled Tribes (STs), which entitle listed peoples to certain forms of government support as per the Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, 1989[Act No. 33 of 1989, (GoI, 1989). Such entitlements include affirmative action in education as well as in governance; or reserved seats for lower castes and women, in some cases with constitutionally mandated proportions of Dalit or tribal women (Stephen, 2012;Haritas, 2016).
Yet, the effectiveness of these reservations for levelling power relations and life chances is a matter of debate (e.g. Bryld, 2001;Ban and Rao, 2008). Movements such as Dalit feminism underscore that the historical oppression of so-called backward castes and minorities continues to be reflected in modern institutions and organizations (Rege, 2006;Rege et al., 2013). They advance the need to center analyses of caste and gender discrimination on the experiences of Dalit women, whose subjectivities are distinct from those of upper caste women and feminists (Rege, 1998).
In this study, we focus on how intersecting relations of gender and caste or tribal affiliation, which are also strongly associated with landholdings and education, position women in the study sites. We show that these enmeshed relations, and the places and environmental histories that embed them, mediate local interest and engagement in forest governance. Following Thompson (2016) and Westholm (2016, p. 514), we demonstrate the 'situated specificity of social life', and the importance of place in shaping social inclusion and exclusion in communal forest management.

Study sites
This study is situated in Uttara Kannada District in the state of Karnataka, and Mandla District in the state of Madhya Pradesh, in south-western and central India, respectively. The two sites offer distinct geographical, socio-cultural, and political-economic contexts for studying exclusions in the context of JFM. As explained below, the first presents exclusions on the basis of intersecting relations of caste and gender, whereas the second demonstrates that exclusions also occur among different tribal communities, which are often grouped together in government classifications referring to disenfranchised populations. Situating the study in these contrasting contexts, in which JFM has followed separate trajectories, offers insight into the pervasiveness of social exclusions across distinct manifestations of JFM, the plural processes by which these occur, and the different meanings they hold for local women and men.
Uttara Kannada is located in the Western Ghats, which is covered by a lush and biodiverse tropical rainforest (Myers et al., 2000). An area of about 170,000 ha, or 80% of the District, is under forest cover (Forest Survey of India, 2017). Karnataka adopted JFM in 1993 and is hailed as an example of successful JFM (Assis et al., 2010). Uttara Kannada District now counts over 500 VFCs. Home to the Appiko movement-a civil protest initiated in 1983 against illegal logging and in support of sustainable forest management (Feeney, 1998)-the area is known for local people's custodianship and protection of the forest (Pinto, 2013). Yet, rapid population growth and infrastructural development, such as dam and road construction, have significantly increased the pressure on forest lands and resources (UNDP, 2005).
The population of Uttara Kannada mainly consists of native Hindus (70%) of different castes, with 8.1% of the population coming from SCs and 2.4% from STs (GoI, 2011). These social groups show differences in terms of landholdings and forest dependency, as well as in education and socio-economic and political opportunities. Havik Brahmins are on the whole the wealthiest group and own larger landholdings. They have been farming betel nut (Areca catechu) intercropped with various spices and fruit trees since colonial times and were ceded exclusive privileges to gather mulching material for their spice gardens from adjoining lands under British rule. These lands, which are particular to Uttara Kannada, are known as 'soppina-betta' or 'betta' lands. Although some Havik Brahmins are poor, many families have prospered primarily from the sale of spices and betel nuts. Havik Brahmins tend to have better access to education and stronger social and political networks than other ethnic communities in Uttara Kannada (Hegde et al., 2017).
The Khare Vokkaliga are the second largest farmer group and among those living in the region for the longest time. They, and other resident groups such as the Marathi, Poojari, and Sherugars, are classified as an Other Backward Class (OBC) 3 in Karnataka. They are predominantly small-scale and subsistence farmers who cultivate paddy, kitchen gardens, and some betel nut palms. They depend on common property forest lands for fuelwood, poles, fodder, food and medicinals. They also work as waged laborers on the betel nut farms of larger landowners and rely on additional income from the sale of NTFPs collected in the forests (Rai and Uhl, 2004;Assis et al., 2010).
The Siddhis and Naiks are the most populous STs in the study site. They have only marginal landholdings and mostly live on 'encroached' lands in hamlets located in the forests. Although they work for wages, forests play a vital role in their livelihoods, with women especially spending substantial time collecting fuelwood and NTFPs for subsistence and sale (Grosse, 2016).
In Uttara Kannada, ethnicity and identity are closely linked to the caste system and the social norms it embeds. As noted above and observed in the study site, Brahmanical patriarchy places particular restrictions on Brahmin women's movements and behaviours. Patrilineal descent and inheritance give men rights to land (among landholding families) and livestock (Feeney, 1998). Independent of their ethnicity, women are generally responsible for domestic work and caregiving. During the seasons of paddy planting and harvesting as well as betel nut harvesting, their workload increases with seasonal agricultural tasks. These tasks vary depending on women's ethnicity, caste, marital status as well as socio-economic status, and whether the household is organized in a traditional extended family or as a nuclear family. Given social sanctions pertaining to their mobility, Havik Brahmin women generally carry out activities in their homes and immediate surroundings, including on betta lands where they collect tree products. In contrast, women from lower castes travel to or live on the farms of their employers, particularly during the harvesting season, and enter the forest more frequently to collect firewood and NTFPs (Grosse, 2016). Mandla District, the second study site, is located in the Satpura Hills of south-east Madhya Pradesh. Dry deciduous, biodiverse forests cover 44% of the district (Forest Survey of India, 2017) and carry high value products including teak (Tectona grandis) and sal (Shorea robusta) for timber and numerous NTFPs. Since the 1970s, the creation of wildlife sanctuaries, national parks, and protected areas have increased pressure on the livelihoods of forest-dependent communities (Mukherjee, 2009;Véron and Fehr, 2011). Outside of protected areas, forests used for subsistence living are often degraded (Debnath and Dasgupta, 2006). Madhya Pradesh adopted JFM in 1991 (Buckles, 1999). At the time of the study, Mandla District counted 285 JFMCs, although according to the District Forest Officer, only 30 of these were active (pers. comm., December 1, 2015). All study participants from the district characterized JFM and relations between the Forest Department and local residents as conflict-ridden.
The state of Madhya Pradesh has the largest population of STsapproximately 14 million people -living in the forest and in forest fringe areas (GoI, 2011). As in India's other highly forested tribal areas, it harbours a particularly high concentration of poverty (World Bank, 2016). Mandla District is populated by a majority of STs (58%), primarily of Bhil (37%) and Gond (35.6%) descent, OBCs such as the Panka and the Ahir 4 , and members of the General Class. 5 The district's main religions are Hinduism (80.5%), followed by 'Other Religions and Persuasions' (16.5%), and a small number of Muslims and Christians (together roughly 3%) (GoI, 2011).
Mandla's population mainly consists of small-scale farmers or landless labourers who sharecrop. Forest dependency decreases as the size of landholdings increase, but all groups are highly dependent on NTFPs as well as on scarce fuelwood and water resources (Véron and Fehr, 2011). Overlapping land claims by communities and the government contribute to the vulnerability of local communities (Chaturvedi et al., 2018). In Madhya Pradesh as a whole, communities have secured 'Individual Forest Rights' (IFR) titles (pattas) to only 766,938 acres of forest fringe areas, or 13 percent of the area with 'titling potential' (CFR-LA, 2016). Interviews conducted in the present study show that Hindu and Gond communities have land titles over small agricultural plots, whereas more forest-dependent communities (Baiga, lower sub-caste Gond, and other STs) lack secure titles to land and forests. Villagers living in the forest fringe have nistari or domestic use rights to forest products, but these do not grant titles. Moreover, unclear boundaries and lack of knowledge of rights by the Forest Department and communities contribute their denial (Winrock International India, 2005 Many people in Mandla district have a secondary occupation, working as hired labourers and seeking (off-farm) employment opportunities, such as those offered by the Mahatma Gandhi National Rural Employment Guarantee Act (2005) (MGNREGA) -although study participants considered these jobs difficult to obtain. Increased tourism, population growth and poverty levels as well as restrictions on forest access, management and use, pose significant challenges for local forest-dependent communities (GoI, 2011).
Hindu and Gond communities are the wealthiest in Mandla and dominate agricultural production. The Panka are the predominant Hindu group in the study site and often fulfil important community roles, for example, as blacksmiths. The Gond have diverse cultural and social practices and hold a historically important role in the region as powerful landowners and political figureheads under colonial rule. They are both agro-foresters and pastoralists. In the study site, they are colloquially referred to as 'the farmers', and are regarded as relatively powerful; although in a complex system of tribal hierarchies and subtribes, their socioeconomic and landholding status varies widely (Patel, 1998).
Of note in Mandla is the presence of the Baiga, who are categorised as one of three Particularly Vulnerable Tribal Groups (PVTG) in Madhya Pradesh. 6 The highest concentration of Baiga in Madhya Pradesh and India is spread across Mandla (at 3% of the district's population) and surrounding districts (GoI, 2011). The Baiga are often entirely forest dependent and highly marginalised. They practice a system of shifting cultivation on forest slopes under rainfed conditions, and plough cultivation when they can avail arable land (Patel, 1998). In two of the five study villages, the Baiga own no cultivable land and are fully reliant on the forest and on sharecropping for food. They mostly live in 'encroached' forest fringes, 7 and harvest bamboo for use and sale in local markets (Chakma et al., 2014) as well as other high value NTFPs. They worship the forest and its spirits and have a plethora of forestrelated songs, myths, dances, and medicines (Soni and Pradhan, 2015).
The Baiga have a centuries-long history of inhabitation in Madhya Pradesh and neighbouring Chhattisgarh. The Baiga, like other forest tribes, have had conflictual relations with the state dating back to colonial times (McEldowney, 1980). 8 Considered the 'wildest and most isolated tribal people' of the area, the British sought to transform their way of life from nomadic forest cultivators to 'civilized' settled farmers (McEldowney, 1980, p. 416). Given the high value of the (sal and teak) forest they inhabit, the state attempted to capture these lands for colonization and state revenue (McEldowney, 1980). Continuities in postindependence state policies and pejorative attitudes towards forest tribes maintain discrimination and tensions in relations between the Baiga and the state (Sarin et al., 2003).
As in Uttara Kannada, interviews in Mandla show that tribal women in the district are primary collectors of forest products for household consumption and sale. Patrilineal land tenure systems in Mandla give men customary inheritance rights to land and livestock (FAO n.d.). Limitations on women's mobility do not apply as they do among upper caste, wealthier Hindu women, and tribal cultures generally have more liberal norms regarding marriage. Like lower caste women, tribal women have greater independence to pursue economic activities, and educated tribal women are increasingly gaining respect and leadership roles within their communities (FAO n.d.).

Data collection and analysis
Data were collected through in-depth interviews carried out in Uttara Kannada and Mandla districts between August and December 2015. Interviews with 45 participants were conducted in six villages in Uttara Kannada District, and with 41 participants in five villages of Mandla District. In each district, villages were randomly selected among a set of 25 villages that were retained for high levels of forest cover, high dependency on the forest among low income communities, low economic indices, high threats to the forest, and presence of a VFC or JFMC.
Interviewees included farmers from different ethnic, socio-economic and age groups, presidents and representatives from the Executive Committee (EC) of the VFCs/JFMCs, and members of the local administrative council (Panchayat) and of women's self-help groups. Adults in all households of the study villages-and thus all research participants-were members of the general assembly of their VFC/JFMCs, which were formed in the late 1990s and early 2000s. In Uttara Kannada, participants included 22 women and 23 men aged 27 to 70. Of these participants, 15 were Havik Brahmins, 19 belonged to the lower castes, and five to STs. Additionally, three Forest Officers (range and division level, all men) and three conservation NGO workers (one woman and two men) were interviewed. In Mandla, 20 women and 19 men 25 to 75 years of age were interviewed; 19 were Gond (ST), 10 Baiga (PVTG), and 8 Panka and sub-castes, 1 Ahir, and one person selfidentified as Muslim. In addition, two male Forest Department Officers (District Officers) and one female conservation NGO worker -all Brahmin Hindu -were interviewed. Interviewees were identified through snowball sampling. A gender balance was sought but was not possible among SFD officials in both districts or among VFC presidents interviewed in Uttara Kannada, who were all men. Interviews were conducted in Kannada in Uttara Kannada and in Hindi in Mandla, with translation to local dialect taking place when needed in Mandla. Questions focused on the participation of women and lower caste or tribal groups in two formal decision-making spaces: general assembly meetings of the VFC/JFMC, to which all members are invited, and the more frequent meetings of the EC, which only elected representatives attend.
Data were coded using NVivo 10 and analyzed for recurring themes based on a mixed deductive and inductive approaches. Quotes were selected to illustrate perceptions that found recurrent expression among participants with shared social positions linked to gender, caste or ethnicity, and other social identities. The qualitative methodology adopted, which privileges an in-depth and situated exploration of the experiences of socially differentiated groups, precludes the ability to claim the representativeness or generalizability of the data gathered. Yet, it allows for a rich analysis of processes of exclusion in JFM in relation to collective identification, or 'the claims and attributions that individuals make about their position in the social order of things, their views of where and to what they belong as well as an understanding of the broader social relations that constitute and are constituted in this process' (Anthias, 2008, p. 491).

Approaching JFM through a place-based lens
In Uttara Kannada, place-specific, spatial factors pose barriers to attending VFC and EC meetings. Households are geographically dispersed within the study villages and scattered within the forest. The village centre, when there is one, is located several kilometres away from some houses, which are connected across hills only via foot or motorbike paths. Although some better resourced men have a motorbike, others -men and women -must walk multiple kilometres to attend meetings. For the labour community, this can add to long distances travelled daily by foot to reach the fields of their employers. 6 PVTGs are a sub-group of the STs who have particularly low development (poverty, literacy, nutrition, etc.) indices. They are eligible for specific state benefit schemes and programs. 7 While land holding patterns are supported by customary governance regimes, the lack of formal tenure rights renders the Baiga 'encroachers' in the eye of the state. 8 Yet, British colonial administrators expressed both strong disapproval and admiration for Baiga customs, practices and beliefs (McEldowney, 1980).
Older people or those who are unwell cannot travel these distances, which also adds considerably to the time commitment of attending meetings. A related problem is the inadequacy of communication channels and limited circulation of information, particularly among those who do not have mobile phones. This means that to inform some residents about meetings, the forest guard or VFC executive members must visit their house when they are home. These spatial particularities are not reproduced in Mandla District, which is relatively flat and where houses are physically grouped together. Yet, despite these conditions, Uttara Kannada is considered an example of successful JFM due to relatively favourable relations between the SFD and local communities. An NGO staff explains that this was not always the case: building relatively congenial relations between communities and the FD took time and involved active struggle: You know, here in the area of Sirsi we are very lucky. We managed to build good relationships with the Forest Department, which is very different from other areas in India. It took us about 20 years to reach there, where we are now, and many fights. The Appiko movement was very important for showing to the Forest Department that the people are there and that they care about what is happening to the forests.
The Appiko movement recurs in local people's -especially men'snarratives of forest management. An increase in awareness about the importance of conserving the forest is reported to have caused palpable changes in forest management. An elder Khare Vokkaliga man from Haldur village 9 , among others, considers that several factors contributed to this awareness: The establishment of the VFC was one of the most important events. We also had some good leaders in the village that helped to create awareness. We were also in the centre of the Appiko movement, which was definitely important. Two representatives from the movement were invited once here to the village for a meeting. I remember they gave a great speech about the importance of the forests for the people.
Participants provided several examples of collaboration between the SFD and the VFC and/or local people, citing the beneficial role of particular individuals, such as an influential District Forest Officer who stood up for local people's rights. An SFD official describes the role of NGOs in bridging differences between the SFD and communities, and in changing the mindset of SFD officers. Today, he states, they [SFD] have to share some information and 'Local people are informed about the issues and their rights. The conflict is not about fights anymore but more about discussion and negotiation to achieve the best possible solution.' Alternative sources of income in the area stemming from betel nut sales allow residents to more sustainably manage their forest extraction activities. Moreover, an NGO staff and an SFD officer explain that higher overall levels of education in Uttara Kannada District (relative to other parts of the state) allow many government and NGO programs to be successfully implemented. Economic and educational aspects also lessen power differentials between the SFD and educated, more resourced members of local communities, favouring more harmonious relations among them.
These conditions are not found in Mandla, where interviewees describe a dysfunctional JFM system. Both women and men convey a severe lack of livelihood options, high reliance on the forest to survive, and a sense of helplessness with respect to their forest's degradation. They cite few, if any benefits from JFM, and very few regard the forest as being in better condition or better managed due to JFM. On the whole, community members lack knowledge about JFM, and unlike in Uttara Kannada, a lack of interest to participate is manifest across ethnic and gender groups.
Historically conflictual relations between the Madhya Pradesh SFD and local villagers, embedded in ethnic, socio-economic and educational differences that accentuate highly skewed power relations, are at the heart of the matter. As an SFD official explains, the FD does not see local people as viable partners in forest management: There are insufficient funds to provide employment to the communities. The only option is to take action against people, and fullfledged government support to take over the forests. […] No. I refuse to think that [the SFD and villagers working together] is an answer. They have no thought to protect the forest. The villagers do not understand the cause of why people like us want to protect the forest.
In turn, local women and men perceive the SFD as a policing force against which they must manoeuvre for basic subsistence. The conflict between the SFD and villagers results in and reinforces local people's low sense of ownership over their forests, as this quote by an impoverished Gond woman from Sumreri village illustrates: There is a really big jungle. It's the government's jungle. The government has captured the forest. If you go to the forest with an axe, they don't let us. The government cuts the forests. The forest guards take the logs away with them. If we quarrel with them, they don't give us the permit to get inside, not even to get fuelwood.
[…] Even if we collect firewood, they won't let us take it out of the forest. We are now going here close by to get fuelwood. How are we going to live? They scare us. They also do this with the men.
In fact, participants lack knowledge about their rights, including the rights of traditional forest dwellers (SCs and STs) to claim for individual and community land ownership in the form of pattas (land titles) through the FRA. Hence, a middle-aged Gond man from Bandholi village explains that people 'are under pressure.
[…] the Deputy Ranger tells us that "the jungle is not your land and so you have to do what we say or else we will kick you out."' This lack of knowledge-and lack of formal rights to land-makes people reticent to engage and express dissent through JFM. Given the lack of incentives or sense of ownership over the forest and its management, several participants believe they should be paid to participate in JFM or to serve on the JFMC's EC, as 'no one is going to work for free' (male Gond farmer, Karki village).
Hence, villagers report that their JFMCs are not very active, if at all, and that if or when JFMCs carry out meetings, many village residents are not aware of them. There is a sentiment that: The meetings are basically just for the people of the committee, they are just a formality. The Forest Department does all of the administrative stuff. Sometimes some of the village people say something, but the Forest Department doesn't really listen to them (elder Gond man, Bandholi).
Participants-male and female-further indicate that they have too many other problems to cope with, such as a water shortages and unemployment, and that people who do attend meetings make requests but 'No one comes here to help' (Gond woman, Bandholi).
JFM is thus received differently across the two study states, based on distinct political economic and historical conditions. In the Mandla case, social exclusions relate first and foremost to an incomplete and inadequate devolution of rights and responsibilities, including autonomy in decision-making, to all community members. They reflect the lower levels of agency, dating back to colonial times, among impoverished tribes in Mandla than among upper caste betel nut farmers in Uttara Kannada in relation to the state. Yet, in both sites, exclusions are accentuated by uneven power relations within communities across several axes of social differentiation.

Approaching JFM with a gender lens
Perceptions about women's participation in JFM in Uttara Kannada are mixed. Most participants in Uttara Kannada indicate that usually father and/or son attend annual general VFC assemblies, but some (e.g. female Havik Brahmin VFC board member, Navalgi village) suggest that 20 to 30% of participants can be women, mostly EC members. In the study villages, Havik Brahmin women are occupying the (at least two) VFC seats reserved for them in the EC and SC or ST women occupy seats reserved for women of their particular caste or ethnic group, but women are not nominated to fill non-reserved seats.
Yet, the number of seats women occupy tell a limited part of the story. In several villages, few, if any, women VFC members reportedly attend EC meetings or general assemblies. According to an NGO staff from Uttara Kannada, participation 'depends on the issue. This is not only for the participation of women but also more in general. Some issues are more important for villagers. For example, when it comes to the sharing of revenues, then everyone will come'.
Most participants report low levels of women's participation at meetings due to several barriers. First, in the two study sites, participants refer to women's labour constraints: 'See, we have family. We have lots of work to do. We go to the market. We have our limits' (Gond female JFMC president, Bandholi). Domestic responsibilities are especially heavy for women living in nuclear (versus traditional extended) families, which are increasingly prevalent and in which several women cannot support each other in accomplishing household responsibilities (Havik Brahmin woman, Jagadal).
Second, women's engagement in forest activities, and related knowledge, is low for certain caste groups. In Mandla, women from all resident ethnic groups are highly involved in forest-related activities. In fact, conflicts with the SFD, particularly over fuelwood, often involve women collectors. Several women describe standing up to the forest guards to retain the fuelwood they need to cook for their family and describe conflicts with the FD with such statements as: 'We heard someone killed the forest guard -good!' (elder Baiga widow, Deolond). In Uttara Kannada, however, Havik Brahmin women reportedly rarely frequent the forest. This is particularly the case when forests are located some distance from the homestead, as women don't ride motorbikes or drive cars and report safety concerns when travelling alone. Frequent references were also made to the normative undesirability of Havik Brahmin women leaving their house. Moreover, for more resourced Havik Brahmin households, betta lands rather than forests serve as a source of forest products.
Third, and closely related, there is a sense that women who are not involved in forest related activities (i.e. Havik Brahmin women in Uttara Kannada) lack interest in forest management. As the male Havik Brahmin VFC president of Haldur village indicates, 'To be a member of the VFC EC and to attend the meetings is a service to the community. There are no benefits from it. It is not a job. We invest our free time and sometimes our own money to be there'. Hence, one must be motivated by the cause. In line with this, in Uttara Kannada, the few women who regularly attend meetings are reportedly mainly from non-Havik Brahmin communities, whose livelihoods are intimately tied to forest use.
Barriers posed by the manner in which meetings are organized further impede women's participation. Meetings are often held at inconvenient times for them, such as in the evenings, when they are busiest with childcare, dinner preparations and milking cows. Although they might attend meetings if these were scheduled earlier, a Havik Brahmin woman from Haldur believes that, 'it doesn't make sense to have the meetings then. It would not be possible for the men to participate because in the afternoons they are working as labour or in the forests.' In both sites, women's mobility also poses challenges; with particular limitations applying to Havik Brahmin women, as noted above.
Importantly, women's lack of participation in JFM is also a function of cultural norms that give men more voice and responsibility in the public domain. In both sites, participants stress that public affairs are relegated to men, and that it is men's role to attend VFC/JFMC meetings. Hence, a male NGO staff from Uttara Kannada indicates that: Women are mostly [at VFC EC meetings] to sign the records.
[…] It is a male dominated society. The main responsibility of women is to take care of the family, do the cooking and all the other housekeeping activities inside the house. This is what many believe women can and should do.
As a result, women are largely silent, and silenced, in formal JFMC processes. As an older Baiga woman from Deolond village states, 'No woman speaks about the forest. I don't speak, my daughter doesn't speak, my grand-daughter doesn't speak. It is like that. No one tells us not to speak-we just don't speak.' A woman NGO staff from Uttara Kannada adds that, 'Women feel like they are in the wrong place in a VFC meeting. If problems are there, then women trust that men will come up with some solutions.' Norms favouring men in public life were even more strongly articulated in Mandla, where female participants largely reported that, 'when the women speak, the men tell the women "Shut up, you don't know what you are talking about", and they say "don't speak in front of everyone"' (Panka woman, Deolond). Speaking out at meetings can be perceived as a sign of disrespect for their male counterparts, and meeting attendance in itself can be a sign of disobedience towards a woman's husband. According to a male farmer from Bandholi, 'I feel like: there is our great-grandfather and there is our grandfather. That is why they [women] shouldn't and don't speak. They are respecting the elder men. They are taught to be like that'. There is also a perception that women do not to have important ideas to contribute. A Gond female JFMC president from Sumreri states that, 'my husband tells me 'you don't utter a word.' […] They say the women know less and the men know more. The men are physically stronger and are smarter than the women.' Restrictive gender norms are reinforced by the SFD. For instance, a male participant from Mandla reports that although his wife is the one who sits on the JFMC EC, when there is a meeting, the SFD invites him to attend in her place (Gond community leader, Bandholi).
Low levels of formal education, which are particularly prevalent among women, further discourage women's participation in JFM. Illiterate women and men are dismissed, and often write themselves off, as being ignorant. A common sentiment among women and men in both study sites is that educated women (and men) are more likely to attend and participate in meetings. In Uttara Kannada, several participants elucidated a link between education and awareness and interest in conserving the forest, self-confidence, and capacity to participate in public fora and decision-making processes. Educated women are thus considered more active not only in VFCs, but in public life more generally.
Lack of knowledge and awareness about JFM are both a cause and consequence of women's lack of participation. Many women interviewed in Uttara Kannada did not know about the VFC and its functions. Others did not know whether their household was part of the VFC and were not informed about when meetings are held. In Mandla, even the two women VFC presidents lacked an understanding of their JFMC and their role within it, and were unaware that they had been nominated as president during the elections. This lends weight to the allegation that women's participation in JFM is still for name's sake.
Nonetheless, participants note exceptions to these patterns. In Mandla, some report that while women who attend general assembly meetings apparently do not speak up, they listen carefully, and participation essentially depends on the individual (Gond man, Antri). A middle-aged Panka woman from Karki believes that times are changing, and 'Earlier the women were confined, but now the women are smarter and they are speaking more'. Other interviewees in Mandla recognize Baiga women's extensive knowledge of the forest and of forest products, and their stake in forest management. In Uttara Kannada, Jagadal's (male) VFC president apparently encourages women's participation by asking women EC members to express their opinion before men; and a female Havik Brahmin EC member from Tungal feels that, 'I have the ability to influence the decisions taken by the VFC to assign some money from the revolving funds to the women's self-help groups.' Others believe that although few women attend general assembly meetings, 'when they go to such meetings, they tend to be engaged' (young Siddhi woman, Chimmad).
These examples show that some women and supportive men are able to open up spaces for women within JFM despite the structural barriers that hinder their participation. Just as the general climate for JFM has improved in Uttara Kannada, an NGO worker believes that: [T]hings have improved a lot […] The condition of the forest and in general of the environment have changed, the Panchayat system has changed too so there are different and more opportunities for women to participate. A big contribution also comes from the women's self-help groups that have been established in many villages. Those help women follow some income generating activities and take their own decisions about what they want to do and to learn.

Approaching JFM with a caste or ethnic lens
Aside from and intersecting with gender, caste or ethnic affiliation strongly shape participation in JFM. As is the case for women, in terms of representation, there are seats reserved for marginalized communities in the VFC or JFMC, some of which are reserved for women from those communities.
In Uttara Kannada, representatives of ST/SC and OBC communities (none of whom hold office bearing roles) mainly expressed that they do participate and are heard in matters of forest management. As an older Khare Vokkaliga man from Jagadal states, 'if we [Khare Vokkaliga] come up with our opinions during general meetings, most will think that we have a good point. Our feelings are also valued.' Members of the Havik Brahmin community, including many VFC EC members and presidents, also consider that ST/SC and OBC members participate well in JFM, especially as they have the largest vested interest in maintaining and generating income from the forest. According to the elder Havik Brahmin president of Chimmad VFC, The dependency on the forest influences the level of active participation of villagers in the VFC meeting. The landless, and people belonging to the Knights and Marathis use the forests especially for [non-timber] forest products and so are very interested in participating in the meetings. They attend because they need the permits to harvest the forest products from the VFC.
Members of marginalized groups attend general assembly meetings in large numbers especially when NTFP extraction permits and rules of NTFP collection are discussed. In some villages, to promote inclusiveness, VFC decisions may be taken only upon consultation with all communities in the village (male Havik Brahmin VFC president, Tungal).
Yet, a general climate of inequality within the villages permeates JFM. Power is concentrated in the hands of the Havik Brahmin community, and particularly of its male members, as is village-level decision-making authority. An SFD official believes that: [VFC] failures are more common when there is no homogeneity in the community. In that case, it is the majority that takes over. Particular set ups in the politics inside the communities also have an influence. Big differences in economic and social status, as we can see them in India in particular, and in education make the implementation more difficult. Power relations inside the villages should not be forgotten.
Several Havik Brahmin responses insinuate the perception that 'they' (the ST/SC community) are to blame for cutting down the forest and that 'they' lack education and awareness to meaningfully participate in JFM.
On a practical level, OBCs and ST/SCs face other barriers in participating in the VFC EC and general assembly meetings and decisions. A main barrier is competing work schedules, as members of these communities are land-poor or landless and work for daily wages. In some cases, meetings are held in the evenings to facilitate the participation of (male) waged labourers (which is prioritized over women's participation, as noted earlier), but time poverty complicates their participation (NGO staff, Uttara Kannada). Another barrier is the lack of formal education, which is pronounced among marginalized communities in both study sites. As mentioned above, in Uttara Kannada, physical isolation and communication barriers can also be particularly challenging for landless families that live in remote areas of the forest.
In Mandla, exclusions occur instead among different groups of STs and SCs, and particularly affect the Baiga, who experience high levels of discrimination. Participants from all groups state that there are no conflicts related to ethnicity or caste in their village and that all live in harmony. Yet, discomfort around the issue and pejorative language are evident in their responses.
Participants from different resident groups and the SFD agree that the Baiga are the most dependent on the forest, which is coherent with self-identification by Baiga interviewees. Many blame the Baiga for the current state of the forests, although both the Gond and Baiga are named as the communities that are gathering the majority of green/live wood, cutting full branches for timber and firewood, and unsustainably harvesting NTFPs. A female Baiga JFMC president from Antri believes that the Baiga are unfairly blamed, however, as: [E]veryone says that the Baiga are destroying the forest -but everyone is destroying the forest. Everyone is doing these things. People here are poor. Baiga are more poor so people blame things on us. […] Baiga women were vocal about conflicts with the SFD and noted that there were no Baiga forest guards.
A number of interviewees indicated that the Adivasis (Baigas and Gonds) are well represented in the JFMCs. Yet, in at least two villages, their participation at meetings is low compared with the Panka, who tend to have more formal education. As a Baiga man from Antri notes, 'The Panka speak the most. Baiga, we don't know why, we don't speak. We are lower caste people, we are jungle people. We are considered the lowest caste.' Similar social and participation hierarchies (e.g. Ahir followed by Panka, Gond, and Baiga) are described in other villages. A lack of formal education and illiteracy are cited as principal reasons why the Baiga and the Gond do not participate actively in JFMC meetings. In one village, however, a Gond participant indicates that the Gond dominate JFMC meetings, again indicating that inter-ethnic dynamics are case-(village-) specific.

Where gender meets ethnicity
The aforementioned gender and caste or ethnic relations do not operate independently, but rather intersect to shape exclusions in JFM. For example, in Uttara Kannada, a Siddhi woman from Chimmad confirms that, 'There is one woman that I know [in the VFC EC]. Her name is Yenki. 10 She belongs to the Siddhis too. That's why I know her. The others I don't know, they are from another caste.' Village-level social relations are structured by ethnicity, within which gender plays the core structuring role. There is little inter-mingling of women across ethnic groups, whose forest dependency, responsibilities and livelihood activities-and thus interests in JFM-both overlap and differ; substantiating the difficulty of separating gender from ethnicity.
In Uttara Kannada, the experience of Havik Brahmin women 10 Pseudonym.
particularly contrasts with that of women from other ethnic communities, who are more involved in forest-related activities and reportedly in JFM, despite facing important barriers. As noted above, some SC and ST women are reportedly very active and participating in the general assembly and VFC EC given the relevance of the forest to their livelihoods. In contrast, interest in JFM is low for Havik Brahmin women, who fill VFC seats reserved for women of the 'General Category'. 11 Stephen (2012) elucidates that powerful upper caste men have fielded their own women (wives, daughters, daughters-in-law) in reserved seats to maintain a hold over governance instances in India. The case of a Gond EC member's husband being invited to attend meetings in her stead illustrates similar occurrences among other ethnic groups. Rather than disrupting prevailing power relations between women and men and among women, these quotas may thus consolidate the dominance of men in rural social structure; and of upper caste men when those seats are reserved for women of the General Category.
In Mandla, there are inequalities in participation across ST and SC communities in the general assembly and the EC. The Baiga in general are the most readily blamed for forest degradation and least vocal in JFM. When asked about women's participation in her village's (Antri's) JFMC, the Baiga JFMC president replied that, 'We are taught from the beginning to shut up, so since the beginning of time we don't talk. It's ok if the ladies talk less.
[…] The Baiga, we speak less, and the other people speak up.' In this comment, which juxtaposes gender and ethnicity, she points to women's exclusion not only because they are women, but because they are Baiga women-subject to norms and other structures that discriminate against them on the basis of both ethnicity and gender. Grouping marginalized groups into homogeneous categories (e.g. as STs or SCs), as is the case in Indian laws addressing tribal issues, deters focus from the inequalities that occur among these groups, and from their relevance in shaping local experiences (Rao, 2018).
In some cases, reserved seat policies for SCs as heads of village councils have been successful in causing psychic and behavioural changes in India (Chauchard, 2017); and in India and Nepal, having a critical mass of women with implicitly shared interests in community forestry institutions has improved outcomes for the disadvantaged (Agarwal, 2015). Yet, Tyagi and Das (2018) argue that by overlooking socio-economic and cultural differences among women and the structural causes of social exclusions, provisions for 'women' in India's forest policies have done little advance their inclusion in forestry. Indeed, the 33% quota for women loses some of its potency in light of the significant heterogeneities that exist within the group. This is not to negate the contribution quotas can make in opening up space for participation, but to underscore that numbers alone can divert attention from other critical issues related to participation and representation (Locke, 1999;Agarwal, 2010). Purkayastha et al. (2003, p. 505) state that: 'women qua women make up an empty theoretical category'. Given social inequalities and differences in their everyday experiences, as well as limited interactions among members of different caste groups, can Havik Brahmin women appropriately advance and defend the interests of Siddhi women? The intersectionality of gender and caste or tribal belonging raises questions about who is entitled to speak for whom (Spivak, 1988) and problematizes assumptions of shared 'women's' interests. It highlights instead the need to consider how 'women from different groups interact, strategise and ally in the political project of representing interests that are at times common to them as women and at times contradictory in the context of the differences of caste and class amongst them' (Haritas, 2016, p. 35).
Participation in JFM must also be contextualized within the places that embed these social relations of marginalization. This is evident when comparing the narratives of women and men in Uttara Kannada and Mandla, which present drastically different experiences with JFM due to the regions' distinct political economic histories and agro-ecological systems. Dependency on the forest is inversely related to the alternative sources of livelihoods available to local residents, which are more prevalent in Uttara Kannada than in Mandla. The Appiko movement instilled an awareness and ethic of conservation in Uttara Kannada and promoted favourable relations between communities and the SFD. In contrast, in Mandla, deep-rooted, highly conflictual relations between communities and the SFD, a lack of local knowledge about forest rights and a low sense of ownership over the forest, as well as a sense of powerlessness within JFM deter nearly all local men and women interviewed from wanting to participate in JFM. Emphasis on enhancing the participation of women or of specific ethnic groups will be in vain without addressing the profound sentiment of distrust and lack of faith and interest in the system on behalf of the FD and local communities.

Conclusion: Beyond gender in JFM
We have argued that although participation in JFM is a highly gendered process, the role of gender in shaping exclusions in JFM must be understood in relation to caste or ethnicity, which are associated with other factors such as landholdings, education, or class. The ways intersecting axes of social differentiation and marginalization position women in relation to the forest, the state, different groups of men, and each other, have implications for achieving inclusive forest governance. For example, unless JFM offers opportunities to address Havik Brahmin women's interests, the extent to which they will feel motivated to be actively involved-and indeed the relevance their participation, which entails opportunity costs-is questionable. In contrast, Baiga women's lack of participation is motivated by ethnic-based discrimination and lack of faith in the system, as well as low self-confidence related to a lack of formal education. In both cases, norms that relegate public affairs and representation to men, as well as other structures that discriminate against women in the public arena, underpin exclusions and must be addressed.
Rethinking the appropriateness of the JFM model and its modalities, its capacity to serve the interests of the poor, and the constraints that hinder the active participation of the community as a whole is pressing. The case of Uttara Kannada demonstrates that change is possible, and that of Mandla urgently calls for change to protect both the communities and the forests. Rather than encouraging local forest users to participate in a top-down system that does not serve their interests, such change will require meaningfully and equitably engaging them in envisioning a system that does. As Sarin et al. (2003, p. 61) note, 'Meaningful devolution requires nurturing democratic, self-governing [community forest management] institutions with clear communal property rights and empowerment of forest-dependent women and men to make real choices for enhancing sustainable livelihoods in accordance with their own priorities.' In Mandla and Madhya Pradesh, the implementation of the FRA offers provisions to recognize the rights of traditional forest users to their resources and can make important strides in that direction (CFR-LA, 2016).
Place-based environmental and political economic histories matter and must shape the strategies adopted to enhance equitable participation in (any model of) forest management. In Uttara Kannada, certain measures at the VFC-level have shown progress in terms of the engaging women and lower-caste members in JFM deliberations and decisionmaking. These measures can be replicated to foster greater equity in other villages, states, or governance instances. Inclusive local forest governance can foster sharing of knowledge and opinions, and negotiations towards common visions among different resident communities. By challenging some culturally-imposed divisions, such processes can engender wider social transformations.