Social Forestry - Why and for Whom? a Comparison of Policies in Vietnam and Indonesia

Community forestry or social forestry (henceforth referred collectively as SF) programs have become new modes of forest management empowering local managers and hence, allowing integration of diverse local practices and support of local livelihoods. Implementation of these initiatives, however, face multiple challenges. State-prescribed community programs, for example, will remain isolated efforts if changes in the overall economic and social governance frameworks, including the devolution of rights to local users is lacking. Financial sustainability of these measures remains often uncertain and equity issues inherent to groups and communities formed for SF, can be exacerbated. In this article, we pose the question: Whose interests do SF policies serve? The effectiveness of SF would depend on the motivations and aims for a decentralization of forest governance to the community. In order to understand the underlying motivations behind the governments' push for SF, we examine national policies in Vietnam and Indonesia, changes in their policies over time and the shift in discourses influencing how SF has evolved. Vietnam and Indonesia are at different sides of the spectrum in democratic ambitions and forest abundance, and present an intriguing comparison in the recent regional push towards SF in Southeast Asia. We discuss the different interpretations of SF in these two countries and how SF programs are implemented. Our results show that governments, influenced by global discourse, are attempting to regulate SF through formal definitions and regulations. Communities on the other hand, might resist by adopting, adapting or rejecting formal schemes. In this tension, SF, in general adopted to serve the interest of local people, in practice SF has not fulfilled its promise.


Introduction
Throughout Southeast Asia, some 140 million people are dependent on forests for their livelihoods and have developed their own systems of managing resources based on traditional knowledge, practices, rules and beliefs for generations ('customary use') (Chao, 2012). The Vietnam Forest Administration (VNFOREST) estimated that in 2008, around 247,030 ha of forestland were managed in a traditional way without legal title in Vietnam (Nguyen, 2009;Sikor and Nguyen, 2011). Of this, 175,395 ha is forest and 71,634 ha is bare land. In Indonesia, 1,500 villages with an area of more than 11 million ha are located within the state forest area, supporting 80 to 95 million people (Ministry of Forestry and Bureau of Statistics, 2009) and many more live in areas bordering forest.
These millions of people developed forest practices according to local conditions and culture, supported by local governance structures that have been effective in maintaining forest and providing local livelihoods. Yet these practices and structures are often invisible to the State, beyond state control and or simply ignored (Scott, 1998;Soriaga and Walpole, 2006). At the same time, people may choose to be at the forest margins (or the periphery) to avoid State control such as the administration of taxes or land control (Scott, 2009). When States claimed rights and control over the forest, people in remote forested areas continued practicing traditional ways of forest and land use (Arnold, 2001), while at the same time incorporating and adapting with the overall political and economic system of the State (Scott, 2009;Herriman and Winarnita, 2016). Today, people in these areas are also the main target for the latest drive of formal state structured social forestry development.
Community Forestry or Social Forestry emerged in the late 1970s, when concern over increasing deforestation rates led to a questioning of state control over forest resources and its capacity to effectively protect and manage forests sustainably. Social or Community Forestry was seen as an alternative to solving forest management problems (Gilmour, 2016), and forest conflicts (Purnomo and Anand, 2014). Over time, ideas of democracy and justice (Brosius et al. 1998;Larson, 2005), coupled with neo-liberal ideas (McCarthy, 2005) and increasing evidence that traditional practices can achieve positive outcomes for conservation (Chomba et al., 2015) shaped how community forestry and social forestry programs developed in different countries.
Although definitions vary, here we use both Community Forestry and Social Forestry, denoted as SF, to mean "any situation which intimately involves local people in a forestry activity" (Arnold, 1992). We follow an accepted understanding that SF includes both: (1) the use of forest resources by local people, on an individual or household basis, for consumption and sale; and (2) the community management of forests, which refers to a collaborative enterprise conducted by a group of local people who manage forests either independently or with outside support for the production of resources for consumption and sale (RECOFTC, ASFN and SDC, 2010).
Vietnam and Indonesia are two countries experimenting with decentralization, including devolution of forestland to communities through SF as part of the state regulatory framework. As part of the CIFOR project ASEAN-Swiss Partnership on Social Forestry and Climate Change (ASFCC), the two countries provide an interesting comparison given their rather different histories of democracy and levels of forest abundance and pressures. Implementation and outcomes, however, have been disappointing in the sense that devolution has not included the full suite of rights to accompany the responsibilities of forest management (Maryudi, 2012;Maryudi et al., 2012;RECOFTC 2014;Chomba et al., 2014;Chomba et al., 2015;Gilmour, 2016;Schusser et al., 2015).
We ask the question: whose interests do SF serve? Understanding the effectiveness of SF would first require clarity on the underlying motivations of governments in adopting SF, which would then allow for a proper assessment of how SF policies are implemented. In this article, we apply the 4I (institutions, interests, ideas, and information framework) by Brockhaus and Angelsen (2012) to analyze the underlying power relations, economic incentives and discursive practices. We examine and compare the national policies in Vietnam and Indonesia, changes in policies over time and shifts in discourses to see how SF has evolved, been interpreted and implemented in these two countries.

Social Forestry discourse and evolution
Over the more than 40 years of social/community forestry development, different interpretations of SF have emerged, often influenced by different global discourses and often not well adjusted to the local contexts. Understandings of, and perspectives on SF evolved from seeing local people as the problem in driving deforestation and degradation to local people as the solution, from local people being 'allowed' to participate in forest management towards a local governance approach where the management of forests are fully devolved to local communities. Though State motivations for SF might differ, and ideas of justice are rather absent in these, the spirit of SF has been one of devolution and democratization.
Though different discourses emerged at different points of time, core elements in these discourses with regard to problem definition, identified appropriate policy responses and overall objectives of SF remain surprisingly constant over time. One dominant narrative since the 1970s is SF as a solution for poverty, which emerged at the same time as the shift in development discourse where rural people were considered 'poor' and development was aimed to fulfill their 'basic needs' (Pollock, 1973Streeten, 1981Todaro and Smith, 2012). Discourses on SF further include ideas such as devolution and restoration of rights; better governance; development of entrepreneurship (RECOFTC, ASFN and SDC, 2010;Sam and Shepard, 2011;de Jong, 2012); empowerment (Maryudi, 2012;Maryudi et al., 2012;RECOFTC, 2014); a mechanism for reducing the fiscal burden of the state; or improving management of natural resources through applying local knowledge (Meinzen-Dick et al., 2008;Arnold, 2001). This mix of discourses and rationales has been adopted by most governments of Southeast Asia and applied in various forms and using different terms: community forestry, participatory forestry, village forestry, community based forest management or social forestry. Projects are aimed to reforestation and afforestation. Forest conservation CF discourse -Initiated by conservation NGOs -Focus: Deforestation as a worldwide issue needs to be halted, and 'community forest management through sustained harvesting and sales of NTFP is a more profitable use of tropical forests and at the same time contributes to tropical forest conservation'. -Implication: conservation-development projects working with communities to identify marketable NTFPs and establish the related value chain. Community enterprise CF discourse -Development cooperation, forestry agencies and rural development NGOs.
-Focus: CF as an abstract objective of communities who manage forest estates, mostly to produce forest commodities. -Implication: promotion of CF enterprise following the western enterprise models with their related hierarchy and business management approach. Forest management is manifested in forest management plans and annual operation plans that are prepared according to legally prescribed norms.
While SF is diverse, governments define three main objectives: 1) to alleviate poverty of forest dependent people, 2) to enable communities to have secure access to, and ownership of, the resources and their benefits, through empowerment and building of capacity for forest management, and 3) to improve the condition of forests (Blaikie, 2006;Maryudi et al., 2012;Chomba et al., 2014;Gilmour, 2016). Reconciling these multiple objectives of conservation, livelihoods, and governance expected from SF can be daunting. In many cases, conservation goals were attained at the expense of local empowerment, showing that the three objectives might not necessarily mutually reinforce one another (Kellert et al., 2000;Arnold, 2001;Ribot et al., 2010;Chomba et al., 2015).
Empowerment appears to be the most challenging. Empowerment implies that local forest users and/or forest dependent communities have autonomy in setting objectives, local control in forest management and utilization, and ownership of the benefits of the forest (Kellert et al., 2000;Blaikie, 2006). It therefore also implies the right to control the use of forestland and resources within their jurisdictions. Yet, empowerment or devolution of power is constrained by conflicts between state and communities to gain control over forestland and resources. Adiwibowo et al. (2016) uses the term 'contested devolution' in which access and control over the forests and ensuing benefits are reformulated within the interaction and negotiation process among parties concerned as they are implemented. The adoption of neo-liberalism (McCarthy, 2005) further drives the discourse that in order to alleviate poverty, SF needs to be managed as a business enterprise (De Jong 2012), thereby reducing forests and forest resources to a commodity. During processes of devolution, negotiation and change, local elites have frequently managed to grab control over benefits at the expense of broader community interests Schusser et al., 2015).
There appears to be a disconnect between the "spirit" of SF and the policies -is SF really meant to be a devolution of rights for local empowerment and management of forests? Or, as Maryudi (2012) claims, a way for the State and or timber companies to regain control? Whose interests do SF policies in Vietnam and Indonesia serve? To study these questions, we adopt the 4I-Institutions, Ideas, Interest and Information-framework (Brockhaus and Angelsen, 2012), which aims to explain how these factors interact and contribute or actually hinder major policy changes across multiple levels of governance, with a multitude of actors and often conflicting politico-economic interests and objectives. The first I refers to the institutions within a particular policy arena, described as the norms, regulations, and formal and informal arrangements which have shaped the policies. Stickiness of particular rules that aim to govern land and land use practices and persistence of existing arrangements, despite new and emerging evidence can often be attributed to those actors with most power and an interest in maintaining the status quo. A second I is related to interests, and while Brockhaus and Angelsen (2012) refer to this I as material, we would like to stress here also non-material interests, shaped by culture and tradition which are also highly relevant to SF. This links to the third I, Ideas, which refers to ideologies, political beliefs, and mental models, and in fact builds on the other two I's by highlighting the particular role of discourses to frame and shape policy actors' perceptions of policy problems and what is perceived as a solution. Ideas or ideology are important as what actors believe might be just as important as what they want (Vanberg andBuchanan, 1989 quoted in Campbell, 2002). Discourses and particular ideas of actors are also linked to the information available, the fourth I in the framework, and how and by whom facts are selected, interpreted and reinterpreted.
Applying these 4 I's allows us to capture the political economy context of social forestry in South East Asia, which often gets overlooked in literature. With this framework we can explain whose interests and ideas in an existing institutional context are reflected in final decisions over design and implementation of SF and what this would mean for the possibility of change in the wider policy arena.

Methods
In our analysis we focus mainly on three of the 4Is, institutions, interests and discourses (ideas) and we operationalized the 4 Is framework in our study through a mix of methods, in particular policy content analysis, literature review, semi-structured interviews between 2012 and 2016, and participatory observation.
In analyzing the main current social forestry policy documents in Vietnam and Indonesia we focused on how SF is defined and regulated, by whom. This gave us insights in the institutions-I and, by looking at policy changes over time, provided information on diverse interests and underlying power relations. Linking this to literature, we analysed how broader discourses around forestry have influenced and driven diverse forms of social and community forestry in the two countries. We then explored to what extent Vietnam and Indonesia have reconciled the interests of the government in terms of control with a rights based approach through SF, or in other words, how institutions deal with the contested devolution.
To provide more grounded discussions, we complemented the review with our personal observations in both following and engaging in the activities of social forestry working groups tasked with developing national policies in Vietnam and Indonesia. Finally, data was also gathered from informal interviews with key respondents and direct observations during field visits to Son La and Nghe An provinces in Vietnam and West Kalimantan province in Indonesia in 2012 and 2016 as part of the CIFOR ASFCC project.

Pathways of CF policy developments
4.1.1. Vietnam SF practices were common in Vietnam until the 1960s. After the central state took control of forests through collectivization campaigns, local systems were replaced by cooperatives (Sunderlin and Huynh 2005) and state forest enterprises (SFEs). In 1991, reforms towards a more market based economy included realization that the SFE system was not effective and was failing financially. As a result, the strong state control of forests was rescinded and approaches for privatization of forestry were piloted, mainly though donor funded projects (Wode and Huy, 2009). Villages and or households were designated to replace SFEs as the managers of natural forests (Sikor and Apel 1998). To mobilize people in forest protection and reforestation actions, the Vietnamese Government adopted a series of policies to promote community participation in forestry (Sikor and Nguyen, 2007). Vietnam implemented national programs to (i) prioritize forest land allocation to households and household groups, particularly to the poor and to ethnic minorities whose livelihoods are closely linked to traditional forest management; (ii) decentralize forest management; and (iii) develop pro-poor mechanisms targeting groups involved in innovative forest management solutions. However, it was only with the Law on Forest Protection and Development in 2004, that the concept of 'Community Forest Management (CFM)' was officially recognized. CFM was defined as "any managerial arrangement in which local people share collective responsibility and benefits from managing natural forests, inside their community boundaries, for which they have long-term customary and/or legal rights of entitlement" (Wode and Huy, 2009;RECOFTC, 2014). SF in Vietnam was also closely linked to the government's Forest Land Allocation Program (FLA) (Sunderlin, 2006). SF in Vietnam can be grouped into two main types. The first is village forest management (VFM), where forest is managed collectively by a forest management group to which all households in a village belong. The second is forest management by groups of households, which is formed by a small number of households -a subset of a village. The HHG model is applied where village population is dispersed in different settlements. Each HHG often manages a patch of forest near their homestead and is characterized by small groups of not more than 20 households and who are usually fairly homogenous in terms of ethnic composition and/or interest. The VFM model, on the other hand involves the whole village population. SF in Vietnam can be characterized as a practical approach in linking poverty alleviation and forest management goals. While there appears to be some tolerance for inclusiveness and diversity in how SF is implemented, it does reinforce notions of the community or village as homogeneous social structures.

Indonesia
In Indonesia, the state designates forest area to be managed by the state. Failure of effective state management of forests led to the introduction of Social Forestry in the late 1970s and became a government policy in the late 1980s (Lindayati, 2002). SF was defined as a system of management of forests (on either state forest or private forestland) that involves local communities with the goal to improve their wellbeing and realize sustainable forestry (Hakim et al., 2010). Under this definition, SF was not fully meant as a policy for decentralization of rights but rather, was perceived more as a "development" program for villages located in forest areas, and its implementation became the responsibility of the forest corporations holding official exploitation permits in forest concessions. Over the subsequent two decades, SF remained an insignificant part of forestry practices, adopted on small scales by concessionaires and the state forest company primarily as an approach to solve local tenure conflicts (Lindayati, 2002). The forest management practices of local people who lived within or around state designated forest land remained largely invisible.
In the absence of dedicated higher regulations on SF, the ministry developed various community forestry schemes through ministerial decrees. Between 1991 and 1998, several such decrees tried to regulate SF in concessions (HPH-Bina Desa Hutan), in special use zones (KdTK) and in the general forest estate (HKM). The new forestry law of 1999 included provisions for the role of local communities in management of forests (e.g. articles 68-70 and article 34) throughout the forest estate.
The implementing regulation of Law 41, 1999 (Government Regulation (GR) 34, 2002 later replaced by GR 6, 2007, laid out different formal schemes of SF, i.e. community forest, village forest, people's plantation forest and collaborative management, with specific access and use rules. Although explicitly stated to aim for community empowerment, all these formal legal schemes are based on permits to be requested by local communities and granted by the state. Interest in SF since then has waxed and waned according to government priorities but also according to the dominant discourse. In 2002, the empowerment discourse moved SF as an umbrella program but when ministers changed, other priorities took over. SF progressed regardless, but at a very slow pace with only few permits requested and granted.
The global movement for indigenous rights, linked closely with issues of environmental change and sustainable development, gained increasing momentum throughout the 2000s driven by the establishment of AMAN (Alliance of Indigenous Communities of Indonesia) in 1999 (Ardiansyah et al., 2015). The Midterm National Development Plan 2010-2014 set a target to devolve 7.9 million hectares of state forest land to local communities (Praputra et al., 2015). In 2014, the new government, continued this policy commitment to SF by promising to assign 12.7 million hectares to local community management by 2019 (Dirjen PSKL, 2015) and has also framed SF as part of broader initiatives for agrarian reform (Presidential regulation 88, 2017). The stated objectives remain focused on poverty alleviation, envisioned as a lack of access to land. However, it is the pressure of land conflicts, demands for recognition of rights and agrarian reform that has given SF its contemporary urgency in Indonesia.

Vietnam
Although the Forest Land Allocation (FLA) is not a SF program per definition, SF was initially framed through FLA policies with the stated objectives to increase forest cover, improved forest quality, and contribute to hunger eradication and poverty reduction in impoverished upland areas (To and Tran, 2014). The two models mentioned above were the offshoot of FLA which was implemented since 1983 and later strengthened by the Land Law of 1993 (Decree No. 02/CP on forest land allocation), and the revised Land Law of 2003. The Land Law confirms state ownership over forests and defines the rights of land users in terms of land use, transfer, concession, lease, mortgage and contribution of capital on the basis of land value. The law further provides use rights for up to 30 ha of production and protection forests for up to 50 years per user/community and was intended to stimulate private investment (Sunderlin, 2006;Wode and Huy, 2009).
FLA itself, is however, not exclusively about allocating land to households and individuals. To and Tran (2014) distinguishes three main forms of FLA: (i) forest land allocation to state organizations, mainly forest companies; (ii) forest land allocation to households, individuals, and communities; and (iii) contract-based allocation of forest land to households and individuals either by forest companies or the state.
There have been many assessments of the impacts of forest and land allocation for households and groups of households with varied results. Helvetas Vietnam (2002) claimed that FLA resulted in improved livelihoods of poor farmers. Yet it is also claimed that households and communities did not gain any actual control over local forests: they still have to seek permission from relevant agencies to harvest trees and use forestland, and they often have limited understanding of their rights and duties (Nguyen et al., 2008). To and Tran (2014) observed the improvement of forest quality but skewed improvements in livelihood where some people were able to disproportionally capture most benefits. Wode and Huy (2009) mentioned the problems with mapping and limited participation resulting in uncertainties of the allocations and disproportionate sharing of benefits.
In 2004, the Vietnam Law on Forest Protection and Development legalized the allocation of natural forests for communities to manage, and thus formally recognizing SF, thereby recognizing indigenous norms and systems of organization (Sikor and Apel, 1998), and encouraging local people to increase the revenue base for community forestry regimes (Sunderlin and Huynh, 2005). Meanwhile, the Vietnam Forestry Strategy 2006-2020 sets out a clear priority to allocate forests for communities to protect and to benefit from the forests. By the end of 2009, local communities held tenure rights in up to 26% of the total forest area in the country, either as individual households, household groups, or village collectives (Sikor and Nguyen, 2011). Vietnam has also linked different payment for environmental services, such as the PFES (Payment for Forest Environmental Services) and REDD+ (Reducing Emissions from Deforestation and forest Degradation and enhancement of carbon stocks) to forms of community-managed forests. The PFES program has been implemented nationally since Decision 99 in 2010. Article eight in this Decision states that village communities, who are allocated forestland for long-term usage, are the beneficiaries of forest payment. As an incentive to encourage communities to get involved in forest protection and management, the PFES scheme compensates communities for forest protection activities. Moreover, the implementation of PFES acts as a catalyst to move forward forest allocation to communities.
In 2012, the Vietnamese Government issued Decision 799/QĐ-TTg on 27/6/2012, which approved the REDD+ National Action Plan. According to this, communities are at the forefront of forest protection and stand to benefit from REDD+ payments. Both PFES and REDD+ are designed as incentives for community forestry management and protection.
Forestry regulations are also often implemented in different ways in different provinces or regions. Some provinces are more flexible and progressive than others and can set up trial sites for community forestry before the central government legally acknowledges this form of forest management (Pham et al., 2012). SF practices might therefore not fit neatly into the government schemes as shown in Table 2 for Vietnam. Indeed, in practice people often ignore government programs or adapt the programs in accordance to local context.

Indonesia
Law 41, 1999 on Forestry and its derivatives (in particular government regulation (GR) no 6, 2007 and its revision (GR) no 3, 2008, provided an early elaboration on how local communities can gain access to state forest, i.e. through permits to manage community forests (HKm), village forests (HD) or small scale plantations (HTR) in accordance to the forest zoning, i.e. production and protected forest (See Table 3). A latest iteration of implementing regulations is Ministerial Decree P 83/2016. Collaborative management in conservation areas is regulated by yet another decree (P.49/2008). Outside the State Forest land, forests on privately owned land are managed by the rights holders as private forest (Law 41, 1999) or customary forest (MK 35, 2012) although no further special regulations currently exist.
In general government schemes allow only rights of use linked to responsibility to maintain the forest. The different rights are described in Table 4 below. SF is the legal mandate of the MoEF and therefore primarily regulated by forestry laws, regulations and ministerial decrees. However, as a category of land use, SF is also bound by the law on spatial planning, the land law, the law on regional government and the law on villages, all mandated by different authorities. Further, SF should defer to the national and regional development plans, and the use of land for 'strategic' purposes such as mining and roads. At local level, this plethora of laws and regulations leads to contradictory and uneven implementation.

SF within decentralization policies
Vietnam and Indonesia are two countries that have experimented broadly with decentralization processes, including through SF programs. The Vietnam SF policy, for example, is closely linked to Forest Land Allocation (FLA) but has also been driven by decentralization processes such as Decree 29/1998 and its amendment Decree 79/2003 on Grassroots democracy aimed to enhance wider public and social participation of different non-state actors group including communities. In Indonesia, administrative decentralization policies have gone further than Vietnam, although actual devolution of management rights of forest to local communities has been slow. Targets are ambitious, 4 million ha of the 15.2 million ha forest land in Vietnam (RECOFTC, 2014) and 12.7 million ha of the 143 million total forest land in Indonesia are to be devolved to local people by 2020 (Dirjen PSKL. 2015).
Devolution of forestland and resources from the state to local people is assumed to result in improved forest as well as improved livelihood. Implicit is empowerment of people. As the Vietnam policy states: "land allocation makes forests officially owned, attaching rights to responsibilities, and offering favorable conditions for local people to protect forest, who find it secure to manage, invest, and develop and allocate forest." (MARD's master plan for forest lease and allocation, 2007: 5 quoted in To and Tran, 2014).
In practice, both countries are very centralized with regard to forest governance . In Indonesia, SF is simply considered a permit to manage forests with implementation rigidly regulated and or tied within a web of complex regulations around forest products harvesting, transporting and selling (Ministerial Decree P.83, 2016). As found elsewhere, the rural poor under well intended SF policies must compete on "an uneven playing field of … social inequities and economic hurdles" (Larson and Ribot, 2007). Indeed, it might seem that the official SF policy is an attempt to re-centralize, to re-assert control over forest resources by the government (Maryudi, 2012;Schusser et al., 2015;Adiwibowo et al., 2016). Simply put, state-led land and forest reform strategies often do not succeed because "bureaucratic modalities cannot accommodate the varying meanings of land, plural notions of property, and diverse politicaleconomic contexts" (Sikor and Müller, 2009).
As mentioned earlier, local people are also subject to policies from other sectors, which are often contradictory or inconsistent, and add to the complexity of SF. In the Vietnam Forest Protection and Development Law of 2004, for example, communities are recognized as legal entities and allowed to enter forest protection contracts. Yet, in the Land Law and Civil Code 2005, communities are not allowed to enter any legal contract, unless registered in the commune as a group of households (civil code) or as cooperatives (circular 07/TTLT MARD and MONRE).
In Indonesia, the new law on village governance (Law 6, 2014) gives villages autonomy to manage their assets, including village owned forests. Yet the Forestry Law sets state authority over all forests at national level. Furthermore, it remains unclear to what extent state forest within the village area is a village asset. On the other hand, in practice, rights of a village to the exclusive use of major forest products from forests in its vicinity have always been de facto recognized.
In the context of overall forest governance, the plethora and complexity of laws and regulations pertaining to local land use that appear unrelated and contradictory is a main issue. Because of this complexity, local communities have difficulty in following the process and the local government tends to treat SF as an administrative issue, an exercise of ticking the boxes, reducing SF permits to merely confirm de facto use of forest land by local people without real attempts to improve governance of forests or empower local people. Under lack of clarity, forest administrators tend to merely follow the letter of the law and forget the spirit of the law.

Discussion
In the following sub-sections, we discuss the (in)effectiveness and adequacy of existing SF policies as an instrument for recognizing community rights, empowering local forest communities, reducing poverty, and for improving forest governance based on the literature review and complemented with findings from our field research.

The spirit of SF laws: competing or complementary values and discourses?
Like general laws, SF laws and regulations are shaped by the political economy of a nation as well as global discourse on environmental and rights issues (Brosius et al 1998;Li, 2002). Although on the one hand, SF emerged out of global ideals of democracy and local rights (Chomba et al., 2014), as McCarthy (2005) claims, SF has also developed from the confluence of environmental governance, neoliberal policy agendas and responsiveness to contingent historical and geographical factors.
These social political contexts at different times gave rise to different discourses which influenced how SF was perceived. The perceived failure of the forest industry model of the state on the one hand, and the increasing rate of deforestation and forest degradation on the other hand (Gilmour, 2016) led to a discourse of resource scarcity. The belief that community control over forests is or can be more ecologically friendly with properly crafted rules and norms governing use led to SF for tropical forest conservation. The rising dominance of the neo liberal economy in the 1990s, gave rise to the discourse of community forestry enterprise (de Jong, 2012), assuming that communities who are able to profit substantively from the forests will be inclined to manage them sustainably.
While these SF discourses are somewhat sequential (de Jong, 2012), they are not exclusive and overlap and mix with parallel discourses. For example, the social justice discourses driving participatory development and 'Bottom up Planning' approaches of the early 1980s and the movement for rights of indigenous and other local communities to exercise control over their traditional natural resources have contributed to the development of SF as a rights-based approach.
The discussion above highlights the influence of discourses and values and interests of the national institutions on the objectives and regulations on SF. However, all regulations and policy, highlight that SF intends that forest should be for the people by the people. Yet, institutional stickiness and attachment to power has made full devolution of forests an elusive dream. Time and again, powerful actors remain reluctant to give up management power and the benefits to be gained (Edmund and Wollenberg, 2003;Dahal and Capistrano, 2006;Sunderlin, 2006;Schusser et al., 2015;Adiwibowo, 2016).
There remains a tension between governments (and powerful non-state actors) trying to exert control over forestland and local people claiming or reclaiming rights over resources they consider their heritage. There are tensions between processes of recentralization and decentralization, and between processes of gaining access and exercising access. An example is how in Kalimantan, the decentralization in 1999, allowed local communities the right to log their forests. This led to a logging boom benefiting local people but also leading to some environmental damage. In 2002, this right was re-centralized to the province and communities were no longer allowed logging rights. In Vietnam, there are cases where large, often state owned companies took over the land already allocated to individual households. Although the grassroots democracy decree promises protection of local rights, State interest takes precedence in practice. Indeed a contested devolution.

SF as a rights based approach?
At the same time, in both Vietnam and Indonesia, SF has become one approach to solve conflicts and to recognize rights of indigenous or local communities. Yet, rights are allocated rather than recognized, and limited to use and management only rather than ownership or control. The constitution of Vietnam states that all land belongs to the state and citizens have only use rights. The government of Indonesia does not claim to 'own'(memiliki) all land but instead uses the word 'control'(menguasai) thereby claiming the right of disposal and thus the right to allocate land or 'permit' others to use the land.
In the context of SF, forest management is considered a communal effort and land is allocated to groups or communities. However, traditionally, while the rights to a forest or landscape might be communal, there is often a nested mix of communal and individual rights, e.g. in Kalimantan the tembawang agro-forest is a communal property with individual rights to certain trees (Peluso 1996), and in Sulawesi, forest farmers groups were formed to gain a CF permit but in practice, little patches of forest are used and managed individually. This is in contrast to some cases in Vietnam where forestland is allocated to individual households while the community might in fact rather manage it communally (Sikor and Nguyen, 2011). The disconnect between rights allocation and household and community forest use strategies could disrupt local governance structures and how benefits are captured. It is therefore necessary to better understand the interlinkages between communal and individual rights and how communities operate in relation to forest as part of SF policy design if objectives of empowerment and devolution of rights are to be taken seriously.
In both Vietnam and Indonesia, legal rights of use to forestland and resources come with restrictions: they are usually time bound, allow only certain uses, and with only part of the bundle of rights (e.g. no alienation or conversion to other uses) are transferable. These rights also come with associated responsibilities: to manage the forest in accordance with technical requirements often for forest protection and rehabilitation rather than production for local livelihood needs, and includes submitting management plans and budgets often beyond the ability of local people.
Local communities might therefore reject these programs, either because they are not interested in managing forest or because it does not recognize their historical rights, as occurred in one community forest area in South Sulawesi (Moeliono et al., 2015b). Instead, it reinforces what is 'given' by the state and therefore, can also be revoked by the state.
Although, SF is positioned as a solution within the rights based approach, it is also recognized that, even when communities desire SF, tenure rights alone do not guarantee its success (Sikor and Nguyen, 2011;Larson et al., 2010). Secure tenure leads to desirable environmental, economic, political, and cultural outcomes only if local communities can realize the rights given to them in legislation. For example, tenure transfers have little meaning if forest regulations and logging bans severely restrict the concrete rights accorded to people to use and extract timber and other resources. Transfers also possess little value if they emphasize protection obligations over rights to forest use and management. In addition, legal tenure rights often do not translate into real rights on the ground if local authorities and communities do not have the capacity to implement legal stipulations Barry et al., 2010).
At the same time, focus on tenure has been at the expense of the more technical matters of good forest management (Moeliono et al., 2015b). Often communities are given areas that are less productive or poor quality forest, or are expected to restore degraded land without adequate technical input (Sunderlin, 2006;Wode and Huy, 2009;Le, 2010;Pulhin et al., 2010). As Dressler et al (2010) also found, governments tend to consider their task done when policies are made, communities informed, land allocated or permits issued. There is little effort to help communities understand their rights and responsibilities regarding the allocated forests. In most instances, local people accept their responsibility to protect or manage the allocated forest, but have little understanding of the related responsibilities of others, particularly state actors, to support them in realizing their endowed rights  Furthermore, the institutional aspects of governance are often neglected. Permits or rights were at times allocated to groups where such groups are not well established nor have a communal identity (Indonesia) or to individuals where traditional systems are communal (Vietnam). Rather than strengthening governance, the introduced schemes can cause disruption and conflict and lead to disempowerment (Blaikie, 2006;Moeliono et al., 2015b;Maryudi et al., 2016).

SF as a new mode of governance?
SF might also be considered as a new mode of governance, where the failure of the state to manage forests sustainably and or solve forest land conflicts peacefully, calls for redistribution of power and decision making over the use of forest with the expectation that this would lead to more sustainable forest management. Although as in Indonesia, while the law mandates community empowerment, it is unclear as to what it implies and who should do it. 'Empowerment' is thus simply translated as participation or involvement of local community in forestry activities. Some argue that participation does not always empower as often only the elites in a community are able to participate. In that case, participation might even be disempowering (Agrawal, 2001, Cooke andKothari, 2001) Community empowerment can be perceived from two aspects: firstly, it relates to the ability of an individual or group to exercise agency, to make decisions, and take ownership of their lives. The second aspect relates to institutional structure, both state and non-state, that are more responsive and accountable to people and thus enables agency. Empowerment is thus located at the intersection of asset-based agency and institution-based opportunity with transformative processes of democracy and transfer of powers, respectively (Chomba et al., 2014). However, in the practice of SF, empowerment is understood as the state providing access to forestland and resources, thereby simply assuming that access will automatically lead to better management of forest and improved livelihood.
There is no acknowledgement of the ability of communities to manage forests traditionally. There is no recognition that a "community" comprises a complex set of actors, with different social, economic, and political characteristics such as wealth, gender, age, ethnicity, and castes (Agrawal and Gibson 1999), with different access to resources and power, and different interests in, and claims over, various environmental goods and services (Leach et al., 1999, Nygren, 2000Chomba et al., 2016).
Power relations are often the most resistant part of institutions to change (Wilshusen, 2009;Sneddon and Fox 2007;Wong, 2010), perpetuated through political ties, family assets and land holdings, family networks, and religious affiliations (Dasgupta and Beard 2007;Lund and Saito-Jensen 2013;Schusser et al., 2016). In the processes around SF, the State usually deals with the village institution, and this institution is often also an apparatus of the State. Indeed, as we observed in rural settings in Vietnam and Indonesia, local institutions led by a village head is often the only bridge of communication between the community and external interventions. The village head is usually nominated by the state and although might be voted for local people, but they still need to get formal approval from the party to be formally recognized as the village head. The power of access and information allows the local elites to "exert disproportionate influence over a collective action process" (Beard and Phakpian, 2009). Elites tend to position themselves as decision makers and promote their own agenda (Thoms, 2008). Many assessments on SF include reports on how elites in a community are able to capture most of the benefits (To and Tran, 2014;Maryudi et al., 2012). While these examples all refer to power relations at the local level, when placing SF into a wider political economy framework, the lack of changes in power relations beyond the local becomes very visible: the established power relations between state-and non-state actors; state, private sector and civil society; and national and local level power relations all have changed so far very little over the course of the different waves of SF, if at all. SF interventions have contributed to inequality and elite capture, especially where local elites dominate the village enterprises (often a State or State-initiated institution) assigned to manage village forests in Indonesia (Moeliono et al., 2015a). This same problem can be exacerbated when future benefits tied to community forestry practices (such as PES or REDD+) come into play. Studies from West Kalimantan, suggests that the implementation of government development projects or REDD+ without intimate knowledge of the political landscape of resource struggle can easily lead to benefits appropriated by local elites (Kallio et al., 2016;Eilenberg, 2015) Policy documents often imply that local communities are weak and incapable of managing forest sustainably without prescriptive rules. Yet proponents of SF often mention the weakness of the state and the inability of government institutions to manage the forest. Thus on the one hand, the state ignores the different local institutions and their different interests and on the other, an often mentioned constraint is the lack of government institutions responsible for SF.
In Vietnam, there is no official government institution that is responsible for community forest management at both central and local levels. High ranking policy makers in MARD are not convinced about the important role of SF due to small areas of only 3.76% of forest that are currently managed by communities, and the mixed results of pilot programs which shows SF is not always effective1. However, this figure only refers to land allocated to communities for communal management and overlooks the 25% of total forests that are under community management (RECOFTC, 2014) allocated to households, as a large proportion of individual land is managed at communal level.
In Indonesia, the new government of Joko Widodo merged the Ministry of Forestry with the Ministry of Environment in 2015, and formed a special Directorate General to manage SF. The Directorate General is expected achieve the target of allocating 12.7 million hectares for SF by 2020 (RPJM). However, SF is challenged by a small budget, (about half of what is allocated for the forest production sector), and this is also the case in Vietnam.
The SF assemblage. We have discussed how a national-led reform process does not translate well at the local level. Indeed, community forest management is what Li (2009) terms an 'assemblage', where competing and overlapping responsibilities, discourses and interests among provincial government, district government, the various Technical Implementing Units of the Ministry (e.g. SF, Watershed Management, Park management), different non-government actors and local communities have made the cohesive governance and management of forest so difficult. Changes at national level and the establishment of new agencies to manage SF, as occurred in Indonesia, have led to confusion and insecurity. Unless mandates and roles are clarified, a coordinating mechanism is established and funding allocated accordingly, targets will likely not be achieved.
Despite the decentralization processes associated with SF in Vietnam and Indonesia, the State has remained a dominant presence. The governmentality of SF can be seen in two ways: firstly, SF is perhaps not so much meant to seriously transfer authority to communities, but to maintain control over forests in a different manner, for example, through complex permitting applications and reporting obligations. Decentralization also offers an opportunity for States to transfer their unresolved forest problems and financial deficits, while maintaining central control over valuable assets and resources. This is certainly the case for Vietnam in terms of state failure and budget deficits in forest management, and Indonesia in terms of appeasing local conflicts. Hence, what seems empowerment of and increased autonomy for and communities at first glance, is actually driven by the central state to dispose of its wicked problems (Arts, 2014).
Secondly, SF is not about increasing decision-making power for ordinary people, but is really about shaping of the community as environmentally responsible subjects and the creation of mutual consent around local forestry problems and objectives (Agarwal 2005).
Given these two perspectives and state failure to transform SF as a diverse and complex system of local and nested forms of governance, we should be "cautious about accepting too optimistic or too naïve governance claims of devolution, autonomy and empowerment." (Arts 2014).

Conclusions
Although based on different contexts and for different reasons, SF was adopted by the governments of Vietnam and Indonesia as a government program to achieve the multiple objectives of improving livelihoods, empowering communities and improving forest governance. The governments regulate SF by legislating multiple policies and regulations that have had at times contradictory effects to their stated objectives. Based on a literature review and data from the field, we find that conflicts over resources, incomplete allocation of rights, misalignment between formal SF rules and local governance systems, and actors involved have rendered SF ineffective in general.
Several interpretations can be identified. First is that the government itself has no serious interest in the implementation of SF as it not only would lose control over the resources, but also must undergo transformational change. Old habits die hard and institutional stickiness and vested interests is a common problem . Secondly, either there is a lack of trust, or communities are indeed unwilling to manage forests on these terms. Local people do not hesitate to change their livelihood system if they see economic opportunities (Feintrenie et al., 2010) but incentives need to be clear.
Thirdly, empowerment needs more than SF, it would need respect for the local people and structural reform to provide clear rights and responsibilities allowing local people autonomy to exercise power. It requires providing understanding and skills that allows local people to govern resources. Equally important, it also requires an understanding that the interlinkages between the community and forest are not static, but an adaptive and dynamic practice of governance to changing development, movements of people, market, policy and environmental processes. As the experience from Vietnam and Indonesia suggest, SF in its current forms fail to meet up to these requirements. In line with other authors (Adiwibowo, 2016;Maryudi, 2012) it may be that SF is indeed not fully serving the interests of the people, but perhaps that of the State.

Acknowledgement
This study was conducted under the ASEAN-Swiss Partnership on Social Forestry and Climate Change (ASFCC), funded by the Swiss Agency for Development and Cooperation (SDC) -Switzerland F-07476.02.04 (Social Forestry, Swidden Systems and Social Networks: Lessons for Designing REDD+ in Southeast Asia, Contract n 0 810245657).
The research was led by the Centre of International Forestry Research (CIFOR) in collaboration with National University of Laos (NUOL), the Vietnamese Academy of Forest Sciences (VAFS), and the University of Tanjungpura (UNTAN). We are grateful to the ASEAN Working Group on Social Forestry (AWG-SF), and our local partners and informants for their instrumental support, advice and patience during the course of this research. We especially thank the respondents for their valuable time and rich insights on which this paper is based. We also thank the anonymous reviewers for their constructive and very helpful contributions to this paper.
Additional funding was provided by the Norwegian Agency for Development Cooperation; the Australian Agency for International Development; the European Union; the International Climate Initiative (IKI) of the German Federal Ministry for the Environment, Nature Conservation, Building and Nuclear Safety (BMUB); the United Kingdom Department for International Development; and the CGIAR Research Program on Forests, Trees and Agroforestry.