Gender Mainstreaming Principles in Indonesia ’ s REDD + Program A Document Analysis

This paper analyzes publically available, published documents to conduct a systematic review of the inclusion of gender mainstreaming principles in documents related to one of the largest, most important international forest-preservation projects, Reducing Emissions from Deforestation and Forest Degradation (REDD+). Although more than 60 countries have adopted REDD+ initiatives, this document analysis focuses specifically on REDD+ in Indonesia, both because of its early adoption of REDD+ and as one of the most ecologically diverse and environmentally important countries in the world. Of the 383 documents reviewed, only 88 were found to include any mention of gender, and of those, very few included gender mainstreaming principles to a level that could be considered integrated and substantive. The distinct lack of these widely touted but seldom implemented principles should be of concern to international and national policy makers, and taken under serious consideration during the forthcoming Conference of the Parties (COP) 21 meetings in Paris. Moreover, this analysis is the first step to identifying where women’s voices may still be silent, and lays the groundwork for more extensive, on the ground research of REDD+ project sites to evaluate the extent to which gender mainstreaming is or is not incorporated into actual project implementation in Indonesia.


Introduction
This paper examines the prevalence of gender mainstreaming principles in Reducing Emissions from Deforestation and Forest Degradation (REDD+) initiatives in Indonesia.As one of the largest and most important international climate-related, forest-preservation projects, the future of REDD+ will be discussed by leaders and representatives from all over the world at the 21 st Conference of the Parties (COP 21) in Paris, France December, 2015.REDD+ was designed to curb climate change through the protection and management of forests in developing countries, and incorporates over 60 countries as either partner or financing countries.It has been a major international initiative since 2007, but it is slated to end in its present form at the end of this year (UN-REDD, 2011a).Although it is unlikely to be completely dismantled -indeed a 2016-2020 strategic framework for the initiative has already been drafted (UN-REDD, 2015a) -serious questions about the financing, effectiveness, and protections in place for REDD+ are being asked and will necessarily need to be addressed at COP 21 (Sills et al., 2014).As such, it is an opportune time to take stock of both the strengths and weaknesses of REDD+, and to identify areas that need more attention in these initiatives going forward.
One area that demands explicit attention during COP 21 is the incorporation of gender mainstreaming principles into REDD+ policies and projects.Gender mainstreaming is a mandate of virtually every international program (UN Women, 2015), including REDD+, but progress in its accomplishment remains controversial.This paper examines REDD+ related documents for Indonesia, an ecologically important country and early adopter of REDD+.We apply social-ecological systems, feminist political ecology, and the capabilities theoretical perspectives to analyze the extent that gender mainstreaming has been incorporated into REDD+ planning and policies.The analysis demonstrates that gender mainstreaming principles are largely absent from this REDD+ documentation, and when gender and gender issues are considered, they remain superficial and non-substantive.

Background
First envisioned as a clean development mechanism (CDM) under the Kyoto Protocol, the REDD concept was originally developed in a 2005 document submitted to the UN Framework Convention on Climate Change (UNFCCC) by Papua New Guinea and Costa Rica on behalf of the Coalition for Rainforest Nations.These countries saw REDD as an opportunity to expand the protections of the CDM to include deforestation of tropical forests (Venter & Koh, 2012).It was intended to "create a financial value for the carbon stored in forests, offering incentives for developing countries to reduce emissions from forested lands and invest in low-carbon paths to sustainable development ( UN-REDD, 2015b).Despite strong lobbying efforts, however, REDD was never incorporated as an official CDM, and was instead adopted by the UNFCCC, not the Kyoto Protocol (Venter & Koh, 2012).REDD was expanded to REDD+ in 2007 as part of the Bali Action Plan (at COP 13) to increase the emphasis on conservation, sustainable forest management practices, and to enhance forest carbon stocks.Today REDD+ acts as an incentive-based program aimed at helping developing countries protect, manage, and utilize their forest resources by making standing forests more profitable than felled forests through carbon offsets for developed countries (UN-REDD, 2015b).It is intended to take the economic, environmental, and social roles of forests into consideration, and advocates for sustainable management of forests, not an end to all forest harvesting (Venter & Koh, 2012).
Although specific requirements for development and implementation of national REDD+ programs are largely left to individual countries themselves, by participating in a UN program, countries are expected to follow certain guiding principles, of which gender mainstreaming is one.First introduced at the 4 th Conference for Women held in Beijing in 1995 as a strategy for promoting gender equity around the world, gender mainstreaming has gone beyond a best practice and is mandated in nearly all major international organizations and programs today (UN Women, 2015).Although there is no one definitive definition of gender mainstreaming, it is generally understood to include "explicit, systematic attention to relevant gender perspectives" (UN, 2002, p. vi) at every stage of every policy, project, and program -from development to evaluation.The ultimate goal of gender mainstreaming is gender equality: "gender mainstreaming entails bringing the perceptions, experiences, knowledge and interests of women [and] men to bear on policy-making, planning and decision-making … situating gender equality issues at the centre of analysis and policy decisions, medium-term plans, programme budgets, and institutional structures and processes" (UN, 2002, p. v-vi).Gender mainstreaming is recognized to have impacts beyond gender equality, however, and broader development goals are understood to be more sustainable and, ultimately, successful through the inclusion of gender mainstreaming principles.The possibility for such far reaching impacts are particularly important to programs such as REDD+, which not only attempts to address environmental and forest degradation and climate change, but which was also envisioned as a catalyst for addressing social issues, including poverty and inequality, facing the developing countries in which it is implemented.

Theoretical Framework
Despite the fact that virtually all international organizations have recognized the importance of promoting gender rights and equality in development and climate change projects and use gender mainstreaming principles as a key strategy for achieving these goals (UN, Women 2015), the extent to which these principles are translated into action appears to be quite limited (Moser & Moser, 2005;Mikkelsen, Freeman & Kelly, 2002;Derbyshire, 2002;Valk, 2000).Not only does there continue to be confusion about what gender mainstreaming is and how to implement it, but on a more fundamental level, policy makers, researchers, and the general public still do not understand the connection between gender and climate change, nor how including gender considerations in climate change policies can impact the sustainability and effectiveness of those policies.Social-ecological systems theory, feminist political ecology, and the capabilities approach make this connection most directly, and this research draws heavily on all three to inform its research questions and methodology.
Both social-ecological systems theory and feminist political ecology perspectives can speak directly to climate change issues broadly and forest-sector changes specifically through their emphasis on the interconnections of the biophysical and social environments.They show how each impacts the other, and highlight the role of economic, political, and gendered social structures that influence how ecological policy is made and implemented (Ludwig, Mangel & Haddad, 2001;Hovorka, 2006;Merchant, 1980).Climate and environmental changes, as well as the policies developed to mitigate their impact and slow their progression, are processes that occur within social structures, and purely scientific strategies to address ecological issues will be ineffective if they do not consider these broader social and political contexts.The concepts of resilience, sustainability, and vulnerability are key in both ecological and social systems, and dealing with them as separate processes creates an artificial barrier between these two interrelated systems (Levin, 1999;Berkes, Colding & Folke, 2003;Gunderson & Holling, 2002).Rather, effective policies must necessarily consider the social structures in which ecological problems are embedded in order to ameliorate the negative ecological and social impact of those structures.According to Berkes, et al. (2003), social-ecological systems are comprised of ecosystems, local knowledges, people and technology, and institutions.Issues related to resilience, sustainability, and vulnerability have to be addressed at each level in order to ensure the health of the entire system.Feminist political ecology brings a specific gender perspective to this interconnection, however.Local knowledges, for example, are themselves created within gendered structures, and the structures that create knowledge have to be considered as much as the knowledge itself (Rocheleau, Thomas-Slayter & Wangari, 1996;Mitchell, 2000).Climate change policy that fails to consider the social structures, and particularly the gendered processes that constrain individual capabilities within those structures, in which climate change occurs address only half of the problem, and in the end are unlikely to be sustainable or effective.
Both of these theories are helpful in understanding how and why gender mainstreaming principles can enhance the effectiveness of policy and programs aimed at a specific ecological sector, like forestry.Recognizing the connection of social and ecological systems is an important first step in addressing the interrelated issues of gender equity and climate change, but this recognition alone is not sufficient to incorporate gender considerations in policy or women into the decision-making process.The very structures that have created gender inequality and excluded women from decision-making bodies also create significant barriers for women to become active participants in policy development and implementation even when a space for them is created.The capabilities approach, first conceptualized by Sen (1985) and often applied to gender and development issues, asserts that institutional and social structures constrain the functioning, capabilities, and agency of individuals, and those institutions and structures must be changed in order to remove those barriers.Applying such an approach to gender equity and climate change highlights the ways in which the very structures and gendered power relations in which ecological problems and policies are embedded have also constrained women's capability to adapt to a changing climate and to be actively involved in climate policy decisions.The capabilities approach suggests that it is not enough to "make space" for women in decision-making processes about climate change adaptation and mitigation if the structures that constrain their capabilities and agency are still in place.
For the principles of gender mainstreaming that are espoused in policy to be effectively implemented, the direct connection between the ecological system of interest -in this case, forestry -and the social and gender structures in which it is embedded must be clearly and explicitly understood by policy makers.If they do not understand the connection or the importance of gender equity in the sustainability of environmental policy, the likelihood that they will adopt and pursue gender mainstreaming principles is low.Moreover, the internal and external barriers to women's active participation in policy and program development and implementation must be carefully analyzed and specific steps to overcome these barriers must be taken in order to fully integrate women -and gender mainstreaming principles -into REDD+ initiatives.Understanding the structures and overcoming barriers will necessarily vary from country to country.In Indonesia, for example, the national language -Bahasa Indonesia -does not have a word for "gender."A lack of similar terminology and a somewhat different understanding of the concept from the broader international community require an entirely different kind of education regarding the basic principles of gender mainstreaming before any cultural or social barriers to women's inclusion is even considered.

Indonesia
Although a study such as this one could be conducted within the context of any REDD+ country, Indonesia provides a particularly salient case for this research.First and foremost, as an island nation, Indonesia is particularly vulnerable to rising sea levels and the increased frequency and intensity of natural disasters due to climate change (Hijioka et al., 2014).Moreover, as one of the most populous countries in the world, Indonesia's vulnerability to climate change represents a direct threat to a substantial proportion of the world's population.The importance of these factors cannot be overstated and are, in and of themselves, sufficient reason to examine any internationally adopted climate change response within the context of Indonesia.As it is, however, Indonesia's position as a leader of the Association of Southeast Asian Nations (ASEAN) states, as well as its recently reformed political and economic systems, and substantial tropical forest stands and other natural resource reserves provide even greater justification for an analysis of climate change policy to be conducted within this national context.In regards to REDD+ specifically, Indonesia was one of the earliest adopters of the initiative and represents one of the most active national REDD+ programs in the world.
Indonesia also has a long history of both state sanctioned environmentalism, and of co-opting women's work and loyalty in the pursuit of sustainable development, which has focused primarily on forest resource policy (Tickamyer & Kusujiarti, 2012;Tickamyer, Kusujiarti & Wornell, 2014).Going back as far as President Suharto and the New Order, and then strengthened significantly post-Reformasi and the Susilo Bambang Yudhoyono (SBY) presidency, environmental activism has been a mainstay in Indonesia's tumultuous political landscape (Peluso, Afiff & Rachman, 2008).Although sound environmental policies and a commitment to sustainable development have dominated the Indonesian climate change agenda, on the ground reality remains disconnected from these policy ideals, and corruption, inequality, and environmental degradation are still the norm (Tickamyer et al., 2014).Moreover, by many accounts the current administration, under President Jokowi, has downgraded the prominence and importance of the international environmental and climate change commitments of the SBY administration by reorganizing and restructuring several national-level agencies, including the national REDD+ agency, which may weaken the position of environmental concerns in policy as well as in practice.The role of women in national level politics is similarly contradictory, with the government relying on their -sometimes coerced -loyalty and labor to promote the development agendas of the state while simultaneously reinforcing their connection to the domestic sphere and increasing their vulnerability and subordinated status (Tickamyer & Kusujiarti, 2012).This uneasy, co-optive relationship between the Indonesian state and women is largely reproduced in the policy regarding gender mainstreaming in REDD+ initiatives, as we outlined below.
Due to the scale of the forest, which covers nearly 70% of the country's land area, and importance of the forest industry to Indonesia's economy, REDD+ plays a major role in directing the country's national forestry, climate change, and sustainable development policies (REDD Desk, 2015).With funding from the World Bank, European Union, individual European countries, and several UN agencies, and through a key partnership with Norway -which committed US$1 billion to Indonesia for progress achieved in the reduction of deforestation and forest degradation by the end of this year -Indonesia's REDD+ program includes over 70 activities and projects in various stages of development and implementation (REDD Desk, 2015).Indonesia finalized its REDD+ National Strategy in 2012, and several national and sub-national agencies have been developed to support its development and implementation.Touted for its theoretical ability to address serious issues facing the country on multiple fronts -from forestry to inequality and poverty -the national REDD+ program in the country has been criticized for its sluggish implementation, land tenure concerns, and questionable inclusion of marginalized groups, including women (Sunderlin et al., 2013), and the National REDD+ Strategy itself only includes a superficial consideration of gender issues and gender mainstreaming principles.

Methodology
The document analysis reported here examines the policies, project documents, and public releases that are associated with REDD+ in Indonesia for their inclusion of gender mainstreaming principles.As this exact type of analysis on the substantive inclusion of gender and gender mainstreaming principles in REDD+ documents has not been done previously, the methodology for analysis is based on content analysis, which has been used in other studies of REDD+ policies (Sadath & Krott, 2013;Wibowo & Giessen, 2015), and rooted firmly in the three theoretical frameworks outlined in section three.

Document Selection
Indonesia has one of the largest number of REDD+ projects of all the REDD partner countries.Because these projects are supported and funded by a variety of different agencies, there is no one comprehensive list of REDD+ projects in the country, which necessitated an Internet search in order to identify both the projects and the documentation associated with the projects.A similar Internet keyword search methodology has been used with other forestry sector and REDD+ related research (Wibowo & Giessen, 2015).The search for documentation of REDD+ projects in Indonesia was conducted in May 2013 using the search engine Google.Any website that was listed using the search terms "REDD Indonesia," "REDD+ Indonesia" "REDD projects Indonesia," "REDD+ projects Indonesia" was opened and scanned to ensure that it was truly related to REDD+ in Indonesia.Searching was continued until saturation and no new websites or REDD+ projects were identified.
After identifying the main websites relating to REDD+ in Indonesia, each individual website was examined for documentation relating to REDD+ projects or policies.These documents could include official policy, project documents at any stage, public release documents (including but not limited to brochures and news briefs), and "other documents" (including but not limited to research documents related to REDD+ in Indonesia and supported by a stakeholder organization or agency).Each document was searched for the inclusion of key terms: "gender," "woman," "women," or "female."References to a female animal were not included.If they key terms were found at least once in the document, it was labeled "includes gender" and included in the analysis.Using this search method, 383 documents representing 50 projects/policies were identified and analyzed.The majority of documents were in English (329 of the 383), and any documents in Bahasa Indonesia were translated and analyzed by a native speaker on the research team.The translation of these documents was also analyzed by the English-documents coder to ensure inter-coder reliability.

Document Analysis
Each document was coded into one of three "levels."Levels two and three are directly related to the concepts outlined in the theoretical frameworks that inform this research.Level two includes documents that make the connection between the ecological and social/gendered systems.These documents outline why gender is an important consideration for REDD+ initiatives, but do not go as far as specifying how to best include women and gender considerations into them.Documents that look at either the barriers to incorporating women into these policies and projects, or those that give specific details of how to increase their participation -how to increase their capability -are coded at level three.Based on the recognition that any consideration of gender in these REDD+ related documents is a step in the right direction, however, level one includes documents that have some reference to gender, but which fail to meet the criteria for either of the other two levels.The definition of each level, and examples associated with it, can be seen in Table 1.• Lists of potential partners, organizations, policies, etc.
• Acknowledgement that gender should be considered, but without any explanation of why or how Level 2: superficial inclusion • Explanation of best practices or definitions of concepts without specific details or plans of how they will be integrated into the policy/project • Statements that women were included or gender was considered, but without details of that participation • Acknowledgement of why gender should be considered and/or the importance of gender inclusion, but without details of how it has been or will be integrated into policy or projects Level 3: integrated inclusion • Discussions of the potential barriers to women's participation, and how those barriers will be addressed • Explanation of how gender considerations has been or will be included in the policy or project Of the 50 projects and 383 documents, 88 documents include gender at some level, 299 do not include gender at all.Once the documents that included the key terms were identified, each inclusion of these key terms was coded into one of the three levels.If all instances of the keywords were coded at the same level, the document itself was coded at that level.In the cases where one document included several levels -which happened frequentlythen the entire document was coded at the highest level included within the document.

Findings and Discussion
On the most basic level this analysis shows that gender is not mainstreamed or even seriously considered in REDD+ initiatives in Indonesia.Of the 383 documents related to REDD+ in the country, a mere 88 of them -23% -contained any mention of gender or women at all.Of those documents that include the key terms, nearly half of them were coded at level one, meaning that gender considerations were non-substantive.Only 15 documents that were analyzed included gender in a way that could be considered substantive, integrated, and mainstreamed.That amounts to less than 20% of gender-inclusive documents, and a mere 4% of all the documents in the analysis.No matter what way you look at it, serious inclusion of gender issues occurring in only 4% of nearly 400 REDD+ related documents does not equate to mainstreaming.A breakdown of the number and percentage of documents that were coded at each level can be seen in Table 2.A closer look at how gender is included at each level provides further evidence that gender mainstreaming remains a political ideal rather than substantive, on the ground practice.
Documents that fall under level one may include the key terms, but their inclusion of actual gender consideration is essentially nonexistent.Documents at this level often had lists of organizations, of which a women's or gender organization was one of many, but why their participation was important for either policy formation or REDD+ projects themselves was never addressed or considered.Alternatively, these documents may have reported on the number of women and men living in a particular village of interest, giving the semblance of gender disaggregated data, but without providing a fuller context through which that data could or should be interpreted.More often than not, this information was provided as very basic demographic data, but was not connected to any potential outcome variables of interest.The following quote from a design document for the Rimba Raya Biodiversity Reserve Project in Central Kalimantan demonstrates another commonly found use of the keywords at this level: "[h]istorically, and through to the present day, the palm has had significant economic value for the Bahaur people, especially women, who process the palm water into brown sugar" (Lemons et al., 2011, p. 51).
Although this document clearly indicates that women are direct participants and, therefore, stakeholders in this forest activity, it does not actually treat them as such.Rather, it simply reports women's involvement without any additional analysis of how women and men have influenced changes in the forest in similar or different ways, or how changes in this forest have had differential gender impacts in the communities that derive their lives and livelihoods from this forest.Moreover, the document goes on to analyze and discuss important changes in the communities that are located in Rimba Raya zone -including issues of land ownership, economic activity, providing for basic needs, and views about the future of the forest -but fails to consider how these important aspects of community life may vary by gender or have a differential impact on women and men.Although documents at this level may include a tacit recognition that gender exists, they offer little else, and simply mentioning women in statistics, background information, or lists of organizations and laws is inadequate and fails to meet the basic criteria of gender mainstreaming.Nearly half (roughly 48%) of the documents that included gender at some level were coded at level one, making a non-substantive inclusion of gender the dominant type in these documents.The second most prevalent level of inclusion, at 36% of documents that include gender at any level, is at level two.Although more substantive than level one, gender considerations in these documents are still superficial.
The vast majority of these documents directly reference gender mainstreaming as a best practice, either indicating that these principles were observed or that they should be, but details of that inclusion are altogether lacking.They demonstrate a general awareness of why gender considerations are important, often linking ecological and gendered social systems, but without details of how to incorporate these considerations.Without these details, one is left wondering whether the authors truly understand the importance and buy into the practice of gender mainstreaming, or whether they are reciting the necessary catch phrases in order to be in compliance with the gender mainstreaming requirements.Take this quote from the official REDD+ design roadmap, developed to outline how Indonesia will remain compliant to the benefits and incentives distribution system, for example: It should also be noted that REDD+ benefit distribution needs to incorporate a gender perspective, to ensure that the needs of women, who frequently form a marginalized group in the forest sector, are taken into account, and that REDD+ can act as an impetus to improved gender equality.Thus, benefits made to households and communities should include safeguards to ensure gender equality.(UN-REDD, 2011b, p. 20) Although this quote does indicate some elements of why a gender perspective is important -because women often make up a marginalized group, their needs are often ignored, and because, if done properly, REDD+ is thought to be an impetus to gender equality -and provides a very general idea of how this can be accomplished -by ensuring that benefits promote gender equality -its inclusion is entirely superficial.It is the only mention of gender in the entire roadmap document, and it comes at the very end of the document, tacked on like an afterthought rather than integrated into the discussion about how benefits and incentives will be distributed in Indonesia's various REDD+ projects.Moreover, it relegates women entirely to members of a marginalized group, without any recognition of their role as important stakeholders, users of the forests, or important members of their community and country -a point we will return to shortly.Finally, although the document asserts that a gendered perspective is important for the national REDD+ strategy, there is no indication of how a gender perspective will be included in these projects or how to ensure that benefits and incentives will ultimately promote gender equality.Nor does it address any potential barriers that such a perspective may encounter.This type of gender inclusion -one based primarily on rhetoric rather than real understanding -was typical for documents at this level.
Very few of the documents (15 of 383) demonstrated a true and thorough understanding of gender mainstreaming principles.For documents to be coded at this level, gender had to be integrated throughout the document, and details regarding how gender was or will be considered had to be specific.These documents clearly go beyond a superficial and formulaic inclusion of gender, and make a concerted effort to explore either the barriers to gender mainstreaming and/or how to effectively incorporate women and their experiences, concerns, and ideas into projects and policies.Women are not defined entirely or even primarily as "vulnerable" in these documents, but rather they are recognized as full and valuable members of society and stakeholders in forest management and REDD+ initiatives.In the 2011 semi-annual national REDD+ program report (2012), for example, the keywords are found in virtually all of the section throughout the report.Moreover it provides specific examples of what has already been tried and accomplished in regards to gender mainstreaming in the national REDD+ program -including a national communications strategy to inform the general public about the link between gender issues, climate change, and REDD+, and a meeting held to inform government official and non-profit representatives of the same thing -as well ideas and goals for future progress.This document indicates that a team has been established to discuss the inclusion, concerns, and roles of men and women in terms of land rights and access to knowledge and economic activities, and identifies the outside non-profits that the team is and can be working with to ensure their goals are met.
Although this national semi-annual report -which is prepared for review by UN-REDD -was coded at level three, it is interesting to note that the National REDD+ Strategy itself -which directs the national level REDD+ program -was coded at level two.Although gender issues were definitely present in the Strategy, it failed to go beyond a recognition that gender should be included and why: Within the context of social safeguards, the application of principles, criteria, and indicators is meant to ensure a strong basis for the full restoration of the public's right and the overall governance process.Within this context, it is imperative to specifically design the social safeguards framework for the protection and benefit of vulnerable groups, including indigenous peoples and local communities living in and around forests, whose livelihoods depend on forest resources; women, who face the full brunt of changes in family incomes; and other societal groups, whose social, economics, and political status put them in a weak position in terms of fulfillment of their human rights.(Indonesia REDD+ Task Force, 2012, p. 29) This quote represents the best inclusion of gender issues in the National REDD+ Strategy.It highlights one reason why it is important to include women in the process, their potential vulnerability, but fails to provide any guidance as to how this inclusion might be accomplished.It also serves to highlight the differences that we found between the agencies responsible for the writing and publication of documents, as shown in Table 3. Documents produced by intergovernmental agencies, primarily the UN, showed the most consistency across all levels of inclusion, although level three does have the lowest percentage.National and subnational level documents, on the other hand, are most likely to include gender considerations at level two, and least likely to include them at level three.Alternatively, although international NGOs, like the Center for International Forestry Research (CIFOR) and the World Wildlife Fund (WWF), produce the most number of documents related to REDD+ in Indonesia, and are instrumental to the implementation of projects, the majority of their documents provide only a cursory, non-substantive inclusion of gender considerations.This suggests that the closer agencies are to the actual implementation of REDD+ projects, the less compliance with the gender mainstreaming mandate they demonstrate in their documentation.Unsurprisingly, intergovernmental agencies appear to have the best grasp and understanding of both what gender mainstreaming is and how it can or should be implemented, as reflected in their relatively high number of documents coded at level three, and their consistency across all the levels.The high percent of level two documents produced by Indonesian agencies predominately reflects the inclusion of best practices which, far from demonstrating an actual understanding of the importance of gender mainstreaming or how to incorporate it, highlights the disconnect between policy mandates and actual implemented practice.NGOs, on the other hand, are arguably the closest to project implementation.They help conduct the Free, Prior, and Informed Consent (FPIC) process, and have connections -and sometimes representatives -within the communities themselves.They work on the ground to move REDD+ initiatives from policy to implementation, yet the vast majority of their documents reflect very little understanding of either what gender mainstreaming is, or why gender considerations might be important to the success and these projects.
The quote from the National REDD+ strategy cited above also highlights one of the two persistent problems across many of the documents in this analysis.Although it includes a consideration of gender issues, it also suggests that women should be included in REDD+ due entirely to their status as a vulnerable group, rather than as relevant stakeholders in forestry policy and management, sustainable development, and community participation.Indeed, this women-as-vulnerable theme was the most dominant one across all the documents, regardless of level.This one-dimensional characterization of women is problematic for several reasons.First and foremost, at its best gender mainstreaming takes into account the "perceptions, experiences, knowledge and interests of women [and] men" in policy and project development and implementation (UN, 2002, pp. v-vi).
Limiting the participation of women entirely by their experience as a group vulnerable to changing political, economic, and ecological environments only reflects one type of experience with these environments.Women are also important natural resource managers, forest users, and members of their communities (Dankelman, 2002;Fulu, 2007;Carvajal-Escobar, Quintero-Angel & Garcia-Vargas, 2008;Nelson & Stathers, 2009;Glazebrook, 2011).They are vulnerable to a changing environment and degraded and deforested forests, but they are also contributors to these processes.Likewise, women have experience adapting to and ideas about mitigating these changes.Policy and projects that fail to include women's perceptions, experiences, concerns, and ideas beyond their experience as a vulnerable group fail to capture and benefit from the entirety of the female experience with and in the forests in Indonesia.Similarly, although women (and children, the elderly, and the disabled) are frequently recognized as a vulnerable group in these documents, the documents rarely make mention of the ways in which men in these forest communities may also experience vulnerability, nor how the experience of vulnerability may impact men and women differentially in the context of REDD+ in Indonesia.
A second persistent problem that was found across many of the documents was the way in which barriers to women's participation was addressed.Namely, it was not.Even documents coded at level three largely failed to acknowledge that there may be cultural, social, economic, or political considerations that keep women from participating in policy development and project implementation, even if and when there is space made for them.
Of the few documents that acknowledge this, most do not attempt to address, let alone suggest ways to overcome, these barriers: "given the cultural context and socio-political structure of the community in the study area, it proved difficult to interview female informants.Hence, gender-related concerns are inadequately captured in this study" (Yasmi, Guernier & Colfer, 2009, p. 102).Directly related to Sen's capabilities approach, it is vital to the successful and substantive integration of gender mainstreaming principles into REDD+ policy and projects that they include a thorough assessment of the structures that act as barriers to women's full participation in these processes.A fully fleshed analysis and discussion about the barriers to women is particularly important in a country like Indonesia, where women's relationship with the state has been somewhat contradictory.Although women have often been recruited into sustainable development initiatives and are often in positions of power within the government structure itself, this inclusion is not equally accessible to all women, and most women in these positions come from the political and social elite classes (Tickamyer & Kusujiarti, 2012).The simple fact that women do have positions of leadership and power, however, may lead some to assert that gender has been mainstreamed and is no longer an issue that needs to be addressed in the country.The results of this analysis clearly demonstrate that this is not true.

Policy Recommendations and Future Research
Document analysis only provides part of the picture, however.To truly understand how gender and gender issues are or are not included in REDD+ initiatives additional research must be conducted at the sites of their implementation and with the people who are most directly affected by REDD+ projects.Such on the ground research would be able to examine the ways in which women and men are involved in project development, implementation, and evaluation.It could investigate their experience with REDD+ and their views and opinions regarding their inclusion and the extent to which they feel heard and incorporated into these projects.
The benefit to understanding these issues through this type of document analysis, however, is not limited to Indonesia, and future research should conduct a similar analysis in all REDD+ partner countries throughout the world.Not only would such an analysis be able to speak to each individual country's experience with gender mainstreaming principles on a policy level, but a comparative analysis between these countries could also highlight both the achievements and consistent difficulties that countries have in incorporating gender mainstreaming principles within REDD+ policies and projects.Knowing what these strengths and gaps are can help guide future international policy decisions regarding both REDD+ and gender mainstreaming.
Making policy recommendations for such large, international programs like REDD+ is admittedly tricky.These recommendations must necessarily be targeted enough as to be easily understood and implemented in a variety of national contexts, but broad enough to be flexible and culturally sensitive.It is important that UN and other international programs continue to steer clear of the purely top-down approaches to policy that have defined them in the past, but it is also important that they provide enough direction, guidance, and oversight to fulfill the mandates and expectations they outline.Across agencies, even within the UN, there is little standardization regarding what gender mainstreaming is, how it is measured, or best practices for achieving it.More effort at the international level must be put into developing a standardized set of criteria regarding the definition, measurement, and implementation of gender mainstreaming principles if such a mandate has any hope of creating actual, substantive change through REDD+ or any other international program.Such standardization does not imply rigid requirements that are imposed on every partner country in exactly the same way, but they would provide more guidance for countries in terms of identifying what gender mainstreaming is and why it is important, explicit suggestions about how to achieve it, and how to measure whether or not is has been achieved.This is particularly important for countries like Indonesia, where the word "gender" does not even exist in the national language, and agencies must simultaneously develop the concept while attempting to achieve gender mainstreaming principles.
Beyond a call for standardization around gender mainstreaming, we offer three policy recommendations that should be considered during the reauthorization of REDD+ this December: gender disaggregated data, community profiles, and gender sensitive community forums.Although the collection of gender disaggregated data is considered "best practice" for gender mainstreaming in REDD+, there is relatively little emphasis on this point or direction regarding how such data should be collected.As is evident from the documents included in this analysis, many agencies have apparently interpreted this activity to mean reporting on how many women and men live within a given REDD+ community.Although this is a start, it certainly is not enough.Gender disaggregated data must be collected at a variety of levels (national and subnational) and across a range of independent variables if it is to be of real, applicable use.Moreover, a detailed, comprehensive community profile of every community potentially impacted by REDD+ initiatives should be conducted prior to the implementation of projects and again after implementation.These community profiles -which would include basic socio-demographic and formal and informal economic data, all disaggregated by gender -will provide a basis for understanding what impacts REDD+ projects actually have in the communities where they are implemented.Without thorough baseline data and follow up analysis, it is impossible to truly understand the scope and depth of impact -either positive or negative -that REDD+ has on the communities it is intended to help.Finally, as part of the Free, Prior, and Informed Consent (FPIC) process of REDD+, community members are supposed to be consulted regarding the costs and benefits of the project to their community.In terms of gender mainstreaming, however, such a broad approach is unlikely to capture important gender dynamics due to cultural and gendered norms that govern who attends, who speaks, and what is discussed at such events.In order to develop a fully fleshed understanding of the gender dynamics and issues in the community, as well as to ensure that women are full and active participants in the process and that their concerns and ideas heard, these community-based discussions should include women-only and men-only community forums.Without a directive to speak directly to women in these communities the barriers to including them in mixed-gendered community meetings may very well exclude them from the process entirely, as we have already identified.

Conclusion
Exclusion from the process of development, implementation, and evaluation is just the scenario that gender mainstreaming was developed to eliminate.Analyzing documents that are related to REDD+ initiatives provides important insights into whether gender mainstreaming principles help guide implementation of these projects as well as serve as an indicator of how they are viewed, understood, and accepted by the agencies responsible for such implementation.Moreover, basing this analysis on social-ecological systems theory, feminist political ecology, and the capabilities approach offers a frame for understanding why gender is an important consideration for the sustainability and effectiveness for REDD+ and suggest paths forward to incorporate greater understanding and emphasis on gendered structures.Not only is a gender consideration important to foster social equity, but the success of environmental programs such as REDD+ rely on the incorporation of all stakeholder voices and concerns, including women.Indonesia's size, vulnerability to climate change, stores of natural resources, political and environmental history, and experience with coercive and marginalizing inclusion of women in its politics make it a logical starting point for just this type of analysis.As this analysis highlights, however, major obstacles to integrating these principles into REDD+ initiatives continue to persist, and less than 4% of nearly 400 documents analyzed fully and substantively incorporate gender mainstreaming principles.This utter lack of gender mainstreaming into a major international climate change initiative can be considered nothing more than a failure to incorporate these principles, a fact that demands consideration when the future of REDD+ is discussed at COP 21 at the end of this year.

Table 1 .
Gender inclusion by mainstreaming level • Generic or background stories about women

Table 2 .
Gender inclusion in REDD+ related documents by level and language

Table 3 .
Level of inclusion by responsible agency