A policy network analysis of the palm oil sector in Indonesia

The palm oil sector has been targeted by NGOs for its alleged negative environmental and social impacts. In this regard Indonesia represents a major challenge because it is home to some of the largest tropical forests in the world. A recent wave of corporate sustainability commitments peaked with the New York Declaration on Forests in September 2014, which emerged amidst the development of other standards and initiatives toward sustainable palm oil production. This process has made this field very complex, especially in Indonesia. The present study aims at clarifying the positions taken by the various stakeholders and assesses the level of political support and the functioning of policy networks.Results from our Policy Network Analysis based on the survey of 59 institutions representing all types of stakeholders (e.g. government, corporate, NGO) at all levels (international, Indonesian and local) show that standards and initiatives for sustainability have contrasting visibility and impact among stakeholders. In this context, RSPO stands as a reference, with the efforts by the Government of Indonesia to promote its own standard with ISPO yet to gain traction. While IPOP was a well-appreciated initiative and a symbol of zero-deforestation commitments, opposition to it by the government and conflicting interests have resulted in its disbandment. Overall, the lack of progress for sustainable palm oil practices on the ground, in the view of respondents, seems to be caused by political and legal barriers rather than technical challenges or economic losses at a country level.


Executive summary
In their campaigns, and in a context where tropical deforestation remains an unsolved problem, NGOs have increasingly targeted those brands and companies that leverage the growing consumer sentiment in developed countries that the tropical commodities they consume -such as palm oil, soy, beef and timberhave contributed to deforestation, loss of biodiversity and social dislocation. These campaigns have led to the emergence of corporate commitments to sustainability, and more specifically to the concept of zero deforestation, by a multitude of actors along the commodity supply chains from growers to processors, traders, consumer goods manufacturers and retailers. The production of soy, palm oil, timber and beef is a major driver of tropical deforestation, and producers and buyers in these sectors have been under particular pressure to eliminate deforestation from their production practices and supply chains. 1 A wave of pledges culminated with the New York Declaration on Forests in September 2014, when governments, private companies and NGOs endorsed the goal of halving the rate of loss of natural forests globally by 2020 and ending it by 2030.
The palm oil sector in Indonesia has figured prominently, as it has expanded exponentially and remains a priority for the government in its support for economic growth and job creation. While Indonesia's public policies toward the palm oil sector have been sending mixed signals over the last few years -from a moratorium on the conversion of primary forests and peat land and regional green growth strategies on the one hand, to subsidies to produce biofuels with the biodiesel mandate or the unconducive legal framework for forest conservation within concessions on the other hand -NGOs have put considerable pressure on the main groups in the palm oil industry and trade to adopt higher sustainability standards than those required by government. This resulted in a series of initiatives and standards to develop proper methods to foster the implementation of commitments and to set priorities between social and environmental aspects.
The Roundtable on Sustainable Palm Oil (RSPO) has been around for more than a decade but its alleged shortcomings have justified, in the eyes of civil society and the private sector, their efforts toward the creation of alternative (and complementary) mechanisms. This process led to the creation of the high-profile Indonesian Palm Oil Pledge (IPOP) in 2014, which gathered six companies in order to make progress on their sustainability commitments through collaborations, communication with the government and other parties, support to smallholders, and others. At the same time, other initiatives led by specific companies emerged with their own approaches to sustainability and associated methods of implementation, and the government pushed its own sustainability standard created in 2011 with the initial purpose of ensuring the legality of company operations, namely the Indonesian Sustainable Palm Oil (ISPO).
These concomitant initiatives, standards, methodological innovations and mechanisms have made this field very complex -and all the more so with a high number of stakeholders participating in the debates and contributing to forging the solutions. Besides, the financial stakes as well as the political sensitivities of these issues have resulted in complementary, overlapping or opposing strategies that have added to the confusion about the paths toward sustainable palm oil production. Therefore, the present study aims at clarifying the positions taken by the great variety of stakeholders with respect to sustainable palm oil, and their perceptions about the various initiatives and standards. It also assesses the level of political support for the most prominent and promising initiatives, and the functioning of the policy networks.
Our analysis is based on a survey of 59 institutions representing the main stakeholders in this field and including central and subnational governments, private sector and civil society organizations, international bodies, research institutions, consulting organizations and donors. In addition to the collection of information in a systematic manner with a questionnaire that addresses a variety of issues including the perceived influence of actors and the collaborations and informal interactions in the social and policy networks, we also built on our experience and participation to various fora and processes in order to put data into perspective.
Results show that debates have been structured around three main initiatives with contrasting characteristics, political support and credibility among stakeholders. IPOP crystallized the zerodeforestation movement with high rates of support, which became so controversial that it was finally disbanded in 2016 in the face of untenable political resistance in Indonesia. This outcome underlined the absolute necessity for the private sector to secure close collaborations with the political elite, which has many stakes in the palm oil business. Public threats to bring IPOP to court under the accusation that it exhibited cartel behavior and public statements pointing to the risks posed by zerodeforestation commitments to small-scale actors resonated to some extent with the perceptions of stakeholders. However, while many of them mentioned a risk that smallholders and medium-sized planters would be left out of the supply chain if zero-deforestation commitments were applied, it was also reckoned that this risk might affect in priority growers who applied contestable and often illegal practices. Last, it is important to note that IPOP was seen as providing positive opportunities in terms of image, environmental services and, noteworthily, opportunities that could translate into increased competitiveness.
A second initiative is ISPO, a state-driven standards system, which has evolved over the years from a legality standard to becoming a vehicle that condenses the sustainability views of the government in the palm oil sector. While the extent of its adoption remains limited, it is expected to play an increasingly important role in the definition of rules in the sector. It has also been tackling the issue of smallholders, with efforts toward their mapping and legalization. Yet it still faces skepticism among stakeholders, many of whom put more faith in RSPO, the other main initiative and, as of today, the most prominent sustainability standard in the sector, which relies on agreed multistakeholder principles as well as on an independent monitoring system, which gives it some credibility. In addition, we observe potentially significant innovations and decisions originating from this body, which seems to have gained sufficient visibility and credibility among stakeholders to make it a reference for sustainable palm oil in the future. Some of the methodological innovations put forward by zerodeforestation commitments such as High Conservation Value (HCV) assessments are progressively integrated into the more ambitious, progressive and voluntary arm of the standard called RSPO Next.
Our survey was straightforward and underlined that political and legal barriers are mainly responsible for the slow progress toward sustainability as opposed to technical challenges or adverse economic consequences for the country. This is reflected in two prominent examples. First, is the ambiguous status of set-asides based on HCV assessments (the high carbon stock approach remains out of reach in current political circumstances). Second, law enforcement is an absolute prerequisite for major progress toward sustainability, notably with the regularization of tenure rights of small-and mediumsized producers, in order to increase the confidence of markets in the sustainability of upstream suppliers. Law enforcement will also require, as a preliminary step, the finalization of the One Map approach to clarify legal concession boundaries and to prevent the proliferation of company operations without proper permits and oversight. These objectives might only be attainable if there is sufficient political will to also tackle the issue of illegal and unsustainable small-and medium-sized producers, the facts of which are uncertain. Because our survey pointed to the government as the most influential actor to advance the sustainability agenda, but also to its insufficient formal and informal interactions with the other stakeholders, the effectiveness of focusing campaigns on the private sector and of the resulting zero-deforestation movement remains to be demonstrated, since end outcomes will also likely depend on specific state interventions.

Corporate sustainability commitments convey new approaches to address tropical deforestation 1.NGO campaigns increasingly target brands, not governments
While deforestation has remained high on the agenda as a major environmental problem, the many attempts so far to address the issue (e.g. in relation to climate change) have had disappointing impacts on the ground (Hansen et al. 2013). The loss of forest cover is mainly caused by the production of agricultural commodities (Gibbs et al. 2010) by corporations or smallholders. Developing and emerging countries are a matter of concern because they often exhibit high rates of deforestation (FAO 2015), and because tropical ecosystems are of critical importance in terms of climate change mitigation, erosion of biodiversity and provision of ecosystem services (Reid et al. 2005). On top of this, social issues such as land conflicts are also commonly cited as sources of concern (Sunderlin et al. 2008).
Meanwhile, global economic integration and deregulation have diminished state control or containment of corporations (Le Galès 2010). This has contributed to the push by many civil society groups for alternative 'self' and 'multistakeholder' regulatory approaches to managing corporate conduct (Doh and Guay 2004). As branding, reputation, financing and alliances have become increasingly tied to corporate values, nongovernmental organizations (NGOs) and advocacy groups have begun to use consumer awareness campaigns and activism to tackle environmental and human rights issues (Hobbes 2015). These strategies are meant to push companies to acknowledge their responsibility for social and environmental impacts, and not only economic performance. While this movement started in developed countries, similar initiatives are appearing in key emerging and developing countries. It is supported by the rise of new technologies that help produce information (e.g. cheap satellite imagery) and transmit it rapidly or even instantly (e.g. social media) (e.g. ZSL 2015).
This phenomenon might be seen as a manifestation of 'market sovereignty,' as NGOs -whose capacity to represent consumers may admittedly be debated -try to impose their views on how goods should be produced in order to ensure a sufficient level of sustainability; however, this is not necessarily a shared concept among all stakeholders. We summarize in Figure 1 the process by which markets, consumers and NGOs have progressively and increasingly aimed at changing investors' and producers' decisions and practices to reflect society's expectations. This process relies on the assumption that businesses understand the risks they face if they use 'dirty' practices, with potential disruption of activities, e.g. when conflicts or fires spread on the ground, or reduced access to markets, e.g. when effective boycott campaigns are launched (The Munden Project 2012).

Figure 1. Consumer and financial market drivers of corporate sustainability commitments: A hypothesis.
Source: Lawry (2015) NGOs have understood that putting pressure on brands might hold more potential than pressuring governments, mostly for accountability reasons and the capacity to affect profitability through market campaigns. This is perfectly summarized in Burgos (2013, 23): "NGOs have come to realize that anti-corporate demonstrations, organized boycotts, and protests can be far more effective and powerful than anti-government campaigns, particularly when targeting established, reputable global brands. In response, corporations have attempted to identify and select the available areas and opportunities to cooperate with NGOs in order to cement fruitful and self-reinforcing relationships".

Indonesia and palm oil in the spotlight
Indonesia is a major country of application for the deforestation-free movement -a major outcome of NGO campaigns -as it hosts some of the largest natural (tropical) forests in the world (FAO 2015). Given that Brazil initiated a large and effective set of policies to reduce its own deforestation rate a decade ago with highly effective results (Arima et al. 2014;Nepstad et al. 2014), Indonesia has recently taken over as the country with the highest rate of deforestation (Margono et al. 2014). Despite a number of initiatives and public policies, such as a moratorium on the conversion of primary forests and peat lands (Murdiyarso et al. 2011), a high-profile attempt to harmonize and register land rights and land use licenses (Lubis et al. 2015), or specific regulations to improve land use planning and address fires during forest conversion processes, Indonesia lags behind other countries in mitigating the impact of agriculture on its forests.
Weak law enforcement, local vested interests and the belief that sectors primarily responsible for deforestation are key to economic development and employment targets (e.g. the government has pledged to double palm oil production by 2020, Jakarta Post 2010). These have all worked against full implementation of the strategic and promising policies issued to appease outside observers globally, of which the moratorium stands as an emblematic example. In this context, the oil palm sector was seen as one of the main agents running against the application of sustainable development principles on the ground (Carlson et al. 2013). It is indeed a priority for a government that boasts of its capacity to accelerate economic growth and develop rural areas; it is also a recurrent and controversial topic, as the sector is regularly accused of environmental and social violations despite substantial positive impacts on poverty alleviation and the economic development of rural areas (Rival and Levang 2015).
In this context, palm oil has become a sector in which trade-offs between economic development and environmental conservation are difficult to avoid; and hence, its continued expansion issues challenges as to whether sustainable development or green growth concepts would be applicable in practice (Jupesta et al. 2011;Anderson et al. 2016). It has logically attracted the attention and interest of many initiatives by civil society aimed at intensifying the pressure on corporations. For instance, the environmental organization Rainforest Action Network (RAN), previously known for its successful campaign in the 1980s against Burger King's beef supply contracts in Central America, developed an online tool that reveals which banks and investors are engaged in large-scale forest destruction in Southeast Asia (Forests & Finance n.d. ). This tool allows users to easily access assessments of the main banks from around the world with respect to their internal policies and commitments; scores are then provided.
This Forests & Finance initiative is not fundamentally different from a project by the Global Canopy Programme (GCP) that is called Forest 500. The GCP has also developed an online tool that identifies and ranks the 500 powerbrokers that have large-scale influence over forest risk commodity supply chains, including 250 companies, 150 investors and lenders, 50 jurisdictions and 50 other powerbrokers (GCP 2015). It keeps track on a yearly basis of policies and operations for each entity and gives a score. Palm oil is one of the most represented commodities along with soy, beef and paper. In addition, a broader initiative tracing the uptake of sustainability commitments by companies at different stages of the supply chain and sectors, labeled Supply Change is also available and tracks corporate public commitments and progress toward eliminating deforestation related to five forest-risk commodities, namely palm oil, soy, timber, pulp and paper, and cattle.
The prominence of palm oil in Indonesia is illustrated further by other more specific initiatives such as the Sustainable Palm Oil Transparency Toolkit (SPOTT) developed by the Zoological Society of London (ZSL). This convenient online tool (ZSL 2015) assesses 50 of the world's largest palm oil producing companies against 50 indicators and makes concession maps available on the website. A more prominent initiative is led by the World Resources Institute (WRI) and builds on the Global Forest Watch tool that has been providing access to information on the evolution of forest cover with convenient maps on open access. Its latest innovation is called GFW Commodities and has focused on assessing risks of deforestation associated to crude palm oil mills that are established in forest landscapes.
These examples are illustrative of a broader trend toward developing online tools that track and update information on company operations and make them available, almost in real time. By doing so, they apply constant pressure and hold these companies accountable in order to trigger changes.

Corporate sustainability commitments have peaked with the NY Declaration on Forests
Corporate sustainability commitments have multiplied over the last few years and include burgeoning deforestation-free pledges, not least because companies could hardly afford to take the risk of reputational damage . These pledges were made by growers and processors of agricultural commodities but also by other actors further down the supply chain, such as traders or consumer goods manufacturers. They are often labeled as 'zero-deforestation commitments,' but also often include 'no peat' (no development on peat land) and 'no exploitation' (meeting satisfactory labor standards) principles (Pirard et al. 2015). Although commonly presented as being pragmatic, companies are expected to take appropriate measures to stop unsustainable practices presumably with third-party verification; however, it is fair to recognize that these commitments are not necessarily concrete and ambiguity might remain with respect to both targets and deadlines (Bregman et al. 2016). Not meeting these public commitments could have severely negative consequences for a company's image and access to markets, notably in Europe and North America.
In 2010, Nestlé became the first company to make such a pledge in the wake of the 'deforestation resolution' by the Consumer Goods Forum -a network of CEOs and senior management of over 400 retailers, manufacturers, service providers and others -to achieve zero net deforestation among its members. It has been followed by over 50 other commercial giants, including Asia Pulp & Paper (APP), L'Oréal, McDonalds, Proctor & Gamble and Walmart (Beckham et al. 2014) as well as other international companies and traders comprising 96% of the global trade in palm oil (Butler 2015). Companies that have announced commitments to date include those that produce agricultural commodities such as palm oil, beef and forest products; commodity processers and traders; food companies; consumer goods manufacturers; retailers; and financiers.
Beside corporate commitments, it has been very meaningful and certainly a landmark occurrence that all prominent stakeholders convened in New York in September 2014 in order to agree on and sign the so-called NY Declaration on Forests. This document is remarkable in its capacity to gather around the table a diversity of stakeholders with different views, interests and activities who eventually agreed on a common vision. It was endorsed by national and subnational governments, private companies, indigenous peoples' representatives, and local and international NGOs. This vision translates into at least halving the rate of loss of natural forests globally by 2020 and ending it by 2030. It also makes special reference to the private sector and agricultural commodities producers, whose already existing sustainability commitments are acknowledged and should be supported.
This declaration can be viewed with a critical eye because it is collective and does not clearly assign responsibilities to specific actors in case of not delivering properly, which in turn means that signatories do not legally commit to implementation. 1 Yet it remains an important step in the process of negotiation of desirable collective ambitions because of the diversity of stakeholders involved and recognition of the critical role to be played by corporations; it will be assessed annually and the second report was published in 2016 (Climate Focus 2016).

Standards and initiatives for sustainable palm oil have multiplied
The first standard to address the environmental and social aspects of the sector was the Roundtable on Sustainable Palm Oil (RSPO), established in 2003 and informed by inputs from civil society and public interest groups. Today, the RSPO has about 1500 members worldwide and covered 18% of global production in 2014, expanding far more rapidly than other commodity-based standards (RSPO 2014). While it has been slow to gain traction among a broad range of Indonesian growers, RSPO's membership continues to grow, in particular with downstream stakeholders such as retailers and manufacturers. However, it received criticism for weak compliance among some of its members, as well as insufficient enforcement (Laurance et al. 2010), which in turn has largely contributed to the emergence of other standards that are discussed in this report.
Other certification standards have also emerged for palm oil. These include the International Sustainability and Carbon Certification (ISCC) standard, a certification system used to demonstrate compliance with the European Renewable Energy Directive, and the Sustainable Agriculture Network (SAN), an NGO-led standard that has tackled oil palm in the last three years. Besides, the Palm Oil Innovation Group (POIG) seeks to go above and beyond the RSPO by setting ambitious standards that break the link between palm oil and negative environmental and social impacts.
These various non-state, market-based standards for tropical commodities do not remain unchallenged, especially at the national level. On the same basis that Indonesia developed and launched the Forest Law and Governance Voluntary Partnership Agreements (FLEGT VPAs) for Indonesian timber legality verification (SVLK), it also created the Indonesia Sustainable Palm Oil (ISPO) standard. This selfproclaimed national oil palm sustainability standard is based on existing Indonesian legislation, but is third-party audited, and was mandatory for all oil palm companies by the end of 2014 (Daemeter Consulting 2015).
However, NGOs and consumers worldwide (especially in Europe and North America) continue to press for more rapid changes in the production of palm oil and the No Deforestation commitments have led to the design of the High Carbon Stock (HCS) approach pioneered by The Forest Trust, Golden Agri-Resources (GAR) and Greenpeace in order to identify no-go areas for plantation development (Greenpeace 2014). This move is likely to have played a very influential role in how conversations developed around the topic of zero-deforestation as it displayed a rather successful collaboration between stakeholders with allegedly contrasting views on sustainability. Perhaps even more important, it had a concrete outcome with a methodology that allows implementation based on objective and verifiable criteria.
In mid-September 2014, a separate group of major palm oil producers known as the Manifesto group announced a voluntary moratorium on clearance of HCS areas. It is awaiting empirically valid thresholds for emissions reductions from different socioeconomic scenarios, hence adopting a different view from the Greenpeace-initiated HCS approach that is now being driven by the multistakeholder HCS approach steering group.

Box 1. Main standards, commitments and initiatives in a nutshell
We present in this box a number of initiatives for sustainable palm oil. They form a heterogeneous group, as some are strictly standards, others are partnerships or are more methods oriented, but they all share the characteristic of being designed to improve the sustainability of the sector.
RSPO was established in 2004 and is the pioneer in the field of sustainable palm oil. It aims at involving stakeholders in the definition of rules with as many as 2000 members from over 75 countries as of 2015. It develops and monitors the implementation of global standards in the spirit of a roundtable in which all members have equal rights, from private companies involved at all stages of the value chain to banks or NGOs. Despite its popularity and ever-increasing adoption with growing amounts of certified estates and production (more than 3 million hectares and more than 5 million tons of certified crude palm oil in 2015, representing around 10% of the total), it faced criticisms for not being up to the environmental challenge. However, it is constantly undergoing reforms, with stricter rules and novel approaches, such as certification at jurisdictional levels and voluntary commitment to RSPO Next.
IPOP (Indonesian Palm Oil Pledge) was formally created in 2014 as an initiative or partnership involving five large-scale palm oil companies with zero-deforestation commitments with support from the Indonesian Chamber of Commerce and Industry. Its objectives were manyfold, from promoting members' commitments to gathering resources to do research in order to facilitate implementation, or playing a more political role so as to represent members' views to the government and improve communication. A high-profile initiative that embodied zero-deforestation commitments in Indonesia and beyond, it was eventually disbanded in mid-2016 after two years of a tense relationship with the government and threats to bring the organization to court under accusations of cartel activity.
ISPO (Indonesian Sustainable Palm Oil) was established in 2011 as the government legality standard for palm oil production and its job is to guarantee that all laws and regulations have been respected. It continued on next page

Research question and justification
With a myriad of initiatives and stakeholders very active in this field, it is definitely challenging to keep track of each actor's position and of the value added or specificities of new initiatives. It is also difficult to identify most prevalent and influential positions and where this dynamic sector is moving. Yet this is critical, as palm oil remains a threat to sustainability in Indonesia, and the most promising avenues, standards and policies should be supported: for instance, one question is whether ISPO is the way forward according to a majority of stakeholders. Standards, initiatives and commitments have proliferated and terms are used in different and potentially misleading ways (e.g. sustainability). Therefore, while the variety of active stakeholders is valuable in attracting attention to a comprehensive set of relevant issues to improve practices in the palm oil sector, it also contributed to a high level of complexity of the debates, with associated risks of delays for key decisions on the standards or initiatives to support.
This study is designed to make sense of the complexity and to deconstruct the debate in order to find the best ways to move forward on the sustainable palm oil agenda. Based on data collected with a questionnaire administered to a sample of stakeholders, literature review and active participation of the authors to various fora and processes involving government and the private sector, we aim at documenting the following questions: How are corporate sustainability commitments for palm oil in Indonesia received by stakeholders and are the social networks, as well as the political context in which they operate, supportive? has been presented as a sustainability standard in an effort to compete with other emerging international standards and commitments. Despite its less ambitious objectives, especially in terms of forest conservation requirements, its adoption has been slow, with less than 200 companies being ISPO certified by early 2016.
SPOM (Sustainable Palm Oil Manifesto) is an initiative started in 2014 between several private companies with activities at several stages of the palm oil value chain. While being more Malaysia oriented in its membership, it is of interest in this analysis as it promoted research on innovative methods and tools for the implementation of zero-deforestation commitments. In particular, an ambitious study was undertaken in 2015 on refined tools for high carbon stock assessments, known as the HCS+. It was generally considered as less strict on environmental criteria but more comprehensive on social aspects. POIG (Palm Oil Manifesto Group) is an initiative that gathers private companies (e.g. Musim Mas) and NGOs (e.g. WWF and Forest Peoples Programme) and that issued a charter in November 2013. It attempts to cover environmental aspects with the application of HCS and HCV assessments and avoidance of peat land for expansion, as well as social aspects such as the application of free, prior and informed consent (FPIC) or respect of workers' rights. Its specificity is also to explicitly aim to support RSPO and to provide innovative and more ambitious methods in order to raise the bar for sustainability objectives.

A policy network analysis framework 2.1 Analytical framework
Our research questions will be informed by a combination of methods that were developed mostly as part of the policy network analysis (PNA) framework, which aims to describe and understand the significance of social interactions on given issues and in relation to policy processes (e.g. Scott 2000). Theory on which PNA is based suggests that policy processes are affected by networks of actors and coalitions, which serve as ways to promote actors' objectives through their participation to events, direct lobbying or campaigns, and interactions among themselves (Matti and Sandström 2011;Österblom and Bodin 2012). This approach reflects a more realistic understanding of policymaking processes than those assuming that decisions are made by formal authorities, such as central government, in isolation from the rest of society (or from the vibrant dynamics of a sector such as palm oil in Indonesia, as we would argue). Indeed, the body of research on non-state actors associated with multilevel governance approaches has led to the assessment that decision-making processes commonly take place at, and involve, various levels from international to national and local. Besides, it points to the increased role and participation of non-state actors and suggests that decision making should be analyzed in terms of complex overlapping networks rather than discrete territorial levels (Mwangi and Wardell 2012).
Under some circumstances, information and views are expected to circulate better when networks are broader and with a high number of connections between actors; most importantly, the centrality of a given actor within several networks largely determines its capacity to impose views or negotiate solutions, sometimes also relying on brokers that make connections (Gallemore et al. 2014).
The influence that actors have acquired in a policy arena can be measured in different ways, including with the concept of reputational power or perceived influence, whereby respondents indicate actors with higher levels of influence (Krackhardt 1990). Once all the responses are aggregated, we can measure the level of influence of each actor based on statements. This measure can be complemented by the assessment of the in-degree centrality, which is informed by the number of interactions with other actors (and specifically mentioned by the other actors) (Prell 2012).
How do this framework and related concepts help to answer the research questions and how do they connect to the issue of sustainable palm oil? Change (and its direction) are assumed to be dependent on the dominant views, which are defined both by the number of actors adhering to these views and their aggregated reputational power. A given policy may have little chance to succeed if its support is limited and coalitions opposing it are powerful. This applies to standards or commitments for sustainable palm oil that convey specific values and objectives, get support from a number of actors and face opposition by others. Dominant views on sustainability norms and the role of corporate commitments to sustainability in the palm oil sector will emerge from the analysis, as well as from existing networks and their respective powers of influence.
While this specific framework will provide us with some means and tools by which to understand better how actors operate and influence one another, and to what extent corporate sustainability commitments represent their views, we need to go beyond that in order to document questions on the political context and support for their implementation. Hence, additional information will be collected to inform these aspects and to provide a more comprehensive analysis of the state of the debates and the ways forward to support higher levels of sustainability.

Questionnaire
Beyond policy and social networks, our research questions call for the collection of many other types of information that help us to analyze in depth the orientations of the diversity of actors and the expectations on a number of policy aspects. To do so, the questionnaire is structured with six sections (see Annex 3).
• Characteristics of the sample. Each respondent provides information on the status of the institution, such as government, producer, NGOs, standards and many others. To complete the description, we also ask the respondent to specify all types of activity, from advocacy to production, auditing, law enforcement and others. In addition, the respondent specifies the level of its operations, from international to local and for how many years these operations have been undertaken in the field of sustainable palm oil. • Vision and definition of sustainable palm oil. This section is made up of three questions that serve to identify the respondent's vision of sustainability, the likelihood of its realization in the medium term and the most critical definitions of sustainable palm oil production. The latter question is based on rankings, with the choice of five options among a predefined set of options in order to identify absolute priorities as well. • Assessment of sustainable palm oil initiatives. Because of the myriad of standards and initiatives under development in Indonesia, we chose the five most prominent ones and asked similar questions for the sake of comparability: relevance, alignment with NGO or government objectives, foreign influence, feasibility, impacts on smallholders, and others.
Then an additional and more general question was asked on the impacts of deforestation-free commitments. • Networks. This section is specific to the policy network analysis, as it covers the actors with most interactions (and their nature) with each of the respondents, those that the respondent wants to have more access to, those with highest and lowest levels of agreement on the issue under consideration, and actors that most hinder transformational change toward sustainability. • Perceived influence/reputational power. This section is made up of two questions to identify up to five actors with the highest perceived influence for each respondent and the reasons for this assessment. • Main issues, opportunities and policy requirements. In this last section, we cover several topics: the main obstacles to effective implementation of sustainability standards in general, main risks or opportunities for Indonesia resulting from IPOP members' commitments, and the priority policies that should be promoted by the government for sustainable palm oil.
The questionnaire is available in the Annexes.

Sample
We interviewed 59 representatives from different organizations, mostly face-to-face yet with a minority of them by phone or email. A total of 85 organizations were contacted, giving a success rate of about 70%, which is due to our intensive efforts for direct contact and previous interactions in this field. We mostly failed to reach informants involved in the central government and the financial sector, respectively, because of the sensitivity of the topic at the time of survey, difficult access with lack of availability, and lack of sustainable finance corporate policies ( Figure 2).
Interviews took place with individuals having sufficient knowledge of their organizations and being representative enough to ensure they reflect the organization's viewpoint. While we are aware of the limitations, as individuals can hardly be expected to not share their own viewpoints or activities to some extent -either in terms of interactions with other stakeholders, assessment of initiatives or knowledge of the field and visions of sustainability -we tried as much as possible to identify those respondents best representing their organization. For instance, sustainability directors in private companies, program managers in NGOs or department heads in governmental bodies. We could rely on the networks of the five enumerators from two organizations with extensive experience in this field (CIFOR and Daemeter Consulting), and complementarily with Indonesian and foreign individuals, which allowed us to interview respondents in the most appropriate language (English and Indonesian).
Having a majority of face-to-face interviews also provided opportunities to ensure that respondents understood questions correctly and the importance of representing the views of their institution rather than their own.

Source: authors' survey
More information about the sample and related issues of representativeness are available in Annex 2.

Stakeholders have different levels of knowledge and appreciation of sustainability initiatives…
One specific feature of sustainable palm oil is the relatively great number of parallel standards and related initiatives and approaches that partly overlap and partly diverge in their principles, objectives and rules of implementation if not in their nature. There are many connections between them and the field is evolving constantly. For instance, the High Carbon Stock approach has a steering committee involving companies and NGOs and aims at improving the HCS methodology for the implementation of commitments; it is acknowledged by the Palm Oil Innovation Group (POIG), which includes private companies and NGOs (often the same ones) and also develops its own sustainability indicators; some of the POIG members also belonged to the disbanded Indonesian Palm Oil Pledge (IPOP), which stood as a very prominent group of producers with zero-deforestation commitments and eventually disbanded in July 2016 after facing threats of legal action by the government on charges of constituting a cartel (see Section 6); most of these producers also have RSPO certification and aim at getting ISPO certification.

Box 2. Forest definitions and methods to identify no-go areas
Core to the corporate sustainability commitments, forest conservation has to rely on methods and tools that enable the identification of no-go areas where companies should refrain from expanding their plantations. This is a strategic issue as methods will determine the extent and location of plantation expansion and of set-aside areas. Two approaches are proposed: High Conservation Value forests. HCV has a long history, as it was adopted first in 1999 for the Forest Steward Council certification scheme. It is now well structured and widely accepted among stakeholders with a HCV Resources Network, which helps strengthen the application of the method. HCV builds on six kinds of values, such as biodiversity, cultural values, critical ecosystems and contributions to larger landscapes. Its application can be problematic if it takes too much time (e.g. field assessments) or consumes too many resources (e.g. substantial associated costs). Besides, as it is somewhat subjective it is subject to contestation. In the context of zero-deforestation commitments, it was viewed as insufficiently focused on conservation challenges in a forest conversion context, and thus led to the birth of new methodologies such as the High Carbon Stock approach.
High Carbon Stock approach. The HCS method was developed with a focus on carbon as a proxy for the condition of forests. It was the fruit of a collaboration between Greenpeace, the company Golden Agri Resources and The Forest Trust back in 2011. Its name is somewhat misleading as it also deals with biodiversity and social aspects. It is assumed that vegetation classes with a number of environmental properties are correlated to carbon stocks. HCS can thus obviously be adapted to the local context, and thresholds for Indonesia would not be relevant in other contexts with different forest types. Recently, the Sustainable Palm Oil Manifesto has promoted a new vision in HCS+ that shifts the balance from environmental to social aspects.
There are no clear lines of division among the flurry of standards and initiatives, which not only share membership but also indicators and principles. The previous few examples are only meant to illustrate the complexity of the field and thus the need to gather views from the stakeholders to make sense of this complexity. Significant initiatives gained strength initially (e.g. IPOP) before disbanding eventually (see Section 6), whereas others have continued to improve and develop new approaches (e.g. RSPO Next). In our study, we collected data on what we consider the most important and emblematic approaches based on either their political support (ISPO), wide application and recognition (RSPO), market share and visibility (IPOP) or innovativeness (POIG and SPOM).
An initial observation from the survey is that the Palm Oil Innovation Group (POIG) and the Sustainable Palm Oil Manifesto (SPOM) are two initiatives that tend to fly under the radar and are little understood by stakeholders in this field. During interviews, we always suggested that informants not respond to questions where knowledge was insufficient, as this would lead to poorly informed responses. This likely explains why we obtained a rate of response of less than 50% for both POIG and SPOM (and almost no response by subnational governments), which is to be compared to high rates of response for ISPO and RSPO. IPOP stands as an intermediate case, with 10-15% of no response depending on the questions, but usually because the topic was said by respondents to be too sensitive (government representatives mostly) and not because they lacked knowledge or awareness. As reflected in Table 1, IPOP and RSPO are widely praised and attracted high numbers of positive responses. In particular, they are justified, timely and relevant, and they usually align with the sustainability vision of the respondent's institution (note Indonesian Sustainable Palm Oil (ISPO)'s rather similar responses on these aspects).
A remarkable result of this survey is the mixed appreciation of ISPO. A lot of efforts and resources were (and still are) put into its promotion and to extend its application to a higher number of producers, yet so far with limited success (see Box 1). A similarly high number of respondents as for RSPO think it can be satisfactorily implemented by 2020 across the sector, but the challenges may be different: RSPO has more ambitious requirements, while ISPO is reported to face more bureaucratic hurdles.

… but zero-deforestation commitments have moved the lines
These initiatives that have developed and operated in parallel have some attributes in common, such as membership or methods to identify conservation areas, but diverge on others, such as the various weights allocated to environmental and social aspects or the thresholds for set-asides. Thus, the specific contribution of zero-deforestation commitments to sustainable palm oil since the early 2000s remains unclear.
Our survey provides insights on this question, among others (see Table 2), and found that these commitments moved the lines. It also provides a new understanding of what defines sustainability in the palm oil sector (it is worth noting that the totality of central and subnational government representatives picked this response). This revised approach to sustainability has not been consensual and conflicts have resulted from the introduction of zero-deforestation commitments; yet, even if definitions may have been manipulated for political purposes, in the end these definitions are also largely based on scientific evidence according to respondents (we must note disagreement here by government). One contentious issue lies with the impacts of these commitments on economic development in Indonesia, which have been frequently put forward by the central government in the media (along with the exclusion of smallholders). In our survey, responses vary significantly depending on respondents' categories … but not as expected. Indeed, the survey does not show clearly that government would have adopted a clear position. Moreover, the subnational government representatives (a more meaningful survey with more respondents in the sample) never validate the view conveyed through the media. Quite surprisingly, while the standards admit that this risk is real, overall and across all respondents it appears to be a marginal view and there is consensus that economic development in Indonesia would not suffer from the implementation of zero-deforestation commitments.
The same does not hold for the exclusion of third-party actors, of which smallholders are one element alongside medium-and large-scale private owners that operate over hundreds of hectares sometimes (Jelsma and Schoneveld, 2016). This exclusion is widely perceived as having resulted from zerodeforestation commitments by all categories of respondents. Note that this is not necessarily a 'risk' (or a problem) in the minds of some respondents who say that this might be justified in the case of unsustainable or illegal practices undertaken by planters whatever their size and in order to clean up the sector. However, for probably the majority of them, a solution will have to be found in order to avoid their sustained exclusion from supply chains.

No-deforestation, legality and social aspects characterize sustainable palm oil
Another way to look at the extent to which zero-deforestation commitments reflect stakeholders' expectations is to explore stakeholders' visions of sustainable palm oil. This also allows us to determine whether there are consensual views and, if not, what are the discriminating aspects and who supports what. Sustainability is an extremely flexible concept so that we may anticipate a certain level of variety in the approaches as expressed by respondents. To increase the chances that respondents express their priorities and provide more specific responses than vague environmental, social or economic considerations, we asked an end-closed question and put together 20 options among which to make choices (see questionnaire in Annex 3, Question 2.3).
Results show that all 20 options were picked at least twice and the most popular response ("protection of HCV areas") was picked 30 times. Responses are rather evenly distributed. "Inclusion, empowerment and capacity of smallholders" was very popular, while "proper waste management" and "maintenance of clean water" were at the low end.  3). "Deforestation related" includes a, b, c; "various environmental impacts" includes d, e, f, g, h, i, j; "legality" includes k, l, m; "various social aspects" includes n, o, p, q; "economic aspects" includes r, s; and "transparency" includes t.
a "inclusion, empowerment and capacity of smallholders" and "FPIC for local communities" are by far the two main components cited by respondents, and make up respectively 43% and 35% of the category "various social aspects." To make the interpretation easier and to identify patterns, we made broader categories and tried to look at the first-ranked priority answers (see Table 3). One remarkable result is the very high attention paid to legality aspects, especially when focusing on the top priorities for respondents. As much as a third of them declared either "legality of plantation boundaries," "compliance of business operations with laws and regulations" or "completion of One Map" as their very first choice.
In other words, being in compliance with the legal framework is an absolute condition for sustainability to materialize in this sector; another way to see it is that little production is entirely legal as of now (because legal compliance is part of RSPO, it might also mean that the problem goes beyond Indonesia). The One Map policy is a requirement that gives clarity about the areas eligible for development, the existence of community rights, as well as coordination between all levels of the state for the issuance of licenses. It was not obvious at the time the survey was undertaken that legality would have such an important role in the minds of a majority of stakeholders, as the debates on standards and initiatives in this field do address practices that go beyond existing laws and regulations. However, it is noteworthy as it makes the point that legality is a prerequisite that remains to be met.
When taking into account all responses by each respondent, we see that not a single aspect of sustainability strongly dominates the others, and social aspects (or even economic aspects to a lower extent) remain important issues to address. Having said that, the further aggregation of responses with broader categories shows that the environment at large (39% of all responses) must be considered as the number one priority, even if not in isolation from the others.  3). "Deforestation related" includes a, b, c; "various environmental impacts" includes d, e, f, g, h, i, j; "legality" includes k, l, m; "various social aspects" includes n, o, p, q; "economic aspects" includes r, s; and "transparency" includes t.
a "inclusion, empowerment and capacity of smallholders" and "FPIC for local communities" are by far the two main components cited by respondents, and make up respectively 43% and 35% of the category "various social aspects." Note 2: A = National government; B = Subnational government; C = Standards; D = International organization; E = National organization; F = Private sector + related associations.

Source: authors' survey
This is only part of the story and it remains to be seen whether these considerations hold equally across categories of respondents. Hence, we disaggregated responses as shown in Table 4. Results are not obvious and few patterns are discernible. One pattern is that the private sector (with striking similarities between growers, mills, traders) focuses heavily on the "protection of HCV areas." As companies in our sample are all on the sustainability side and have close links to RSPO (which acknowledges the HCV approach), and respondents are from sustainability departments (so that there might be a bias here), these considerations make sense. Yet the total absence of reference to HCS is surprising, as these companies are also mostly engaged in zero-deforestation commitments. Therefore, one interpretation is that the application of HCS is still being developed and field tested and thus still has a long way to go, whereas HCV has already established criteria and guidelines and has been implemented for some time. Thus, it appears that utilizing HCV might be the most realistic way of ensuring that unwanted deforestation does not occur. Besides, the private sector aligns with legality considerations (the category "legality" gets 25% of citations), including "completion of One Map" and "compliance of business operations with laws and regulations." Standards matter, as they define the scope of sustainability through RSPO, IPOP members and others.
Their main preoccupations lie with the inclusion and increased capacity of smallholders (the same as for international organizations including NGOs) as well as transparency. But they more or less cover all aspects, which tends to illustrate some balanced assessment of what sustainability takes in the sector (except that their representatives are the only respondents to never cite options from the category "various environmental impacts").
It is also important to observe the strong differences of view between representatives from central and subnational governments (keeping in mind the limitations of the small sample size, with respectively three and seven respondents). While the absence of consideration for deforestation issues is not very surprising from the central government -indeed the context exhibits disagreements over the objectives of some standards to stop forest conversion (see Section 5) -the subnational government level provides very favorable responses (one-third of responses fall into either no deforestation, HCV or HCS protection). This particular finding merits further investigation.
Overall, we see few clear patterns in terms of opposing views, and respondents across categories usually cover environmental, social and economic aspects even if with various weights. Legality remains a key aspect of sustainability for all categories, which is an important result that underlines the challenges ahead for corporate sustainability commitments. It also suggests the need to adopt objectives with limited ambition in the short to medium term, with priority action on preparing the ground for law compliance and strengthened land tenure and licensing systems (the One Map policy is an absolute requirement).

A variety of activities and governance arrangements to address sustainability
To put sustainability principles and commitments into practice, representatives from the private sector, NGOs, consultancy firms and various levels of government have started to implement policies and to undertake activities on the ground. These policies and activities require coordination between stakeholders, for technical and capacity-related reasons (e.g. disseminating good practices to smallholders), for the innovation and design of rules (e.g. refinement of standards), to upscale traceability systems and smallholder legalization and mapping (e.g. jurisdictional approaches), for assessments of high conservation value forests and conservation areas at subnational levels, and others. In this section, we aim to provide a few examples of these activities as well as the related governance arrangements to illustrate (i) the concrete implications of the corporate sustainability commitments and the related debates that were discussed previously in this report, and (ii) the various ways in which stakeholders join efforts to make it happen.

The GAR -Greenpeace partnership for High Carbon Stock set-asides
One high-profile and impactful collaboration was the development of the HCS methodology involving Greenpeace, Golden Agri-Resources and The Forest Trust. This story also offers useful insights into the process that led companies to engage in pro-active policies with zero-deforestation commitments, as it started with a high-profile report by Greenpeace titled "Illegal forest clearance and RSPO greenwash: case studies of Sinar Mas" (Greenpeace 2009) that pointed to the limitations of RSPO certification. Indeed, a number of Sinar Mas group companies were found to be involved in illegal deforestation in West Kalimantan Province at the end of the 2000s at the same time as two prominent companies of the group, including PT SMART, were RSPO certified. These loopholes are believed to have driven a change in the NGOs' approaches to combatting environmental destruction, and probably justified the new requests for specific commitments that would go beyond existing sustainability standards such as RSPO, with their own methods and monitoring processes negotiated and developed in partnership with NGOs.
The next step was the announcement in February 2011 of a new forest conservation policy by GAR (and its subsidiary PT SMART) based on the elaboration and application of the novel High Carbon Stock (HCS) method with support by Greenpeace and The Forest Trust (TFT). This support would provide the credibility sought in order to prevent future campaigns against the operations of GAR companies; in addition, the process of elaboration involved multistakeholder meetings for feedback and validation. The method is focused on carbon to determine thresholds for set-asides and areas eligible for oil palm development but is usually presented as an approach that goes beyond carbon considerations to account for biodiversity and social aspects.
It was field tested in 2013 in the same locations where the group was accused of converting natural forests in relation to the operations of the concession holder PT Kartika Prima Cipta. One of the important issues was the involvement of local communities and the satisfactory application of FPIC approaches to prevent abuses of local rights and unilateral decisions for a land management plan within the concessions. It appeared that this process was not perfect, as a prominent NGO -the Forest Peoples Programme (FPP) -released a report with a rather negative assessment pointing to the lack of popularity of the process among local people and undue control of community lands for conservation purposes (Colchester et al. 2014).
This experience is interesting in several respects. Not only does it show the impacts of NGO campaigns and their capacity to impose new methods for the implementation of sustainability policies, but they also exhibit the challenges posed by field conditions, the involvement of more stakeholders over time, and the conflicting conservation and social objectives that altogether form the core of sustainability commitments. Perhaps more importantly for our analysis, it highlights not only the dynamism of multistakeholder processes but also the potential gap between statements on sustainability policies and their materialization -whether because of methodological issues, contradictory objectives or an unsuitable legal framework. Another less obvious obstacle, yet recently highlighted by NGOs, is the lack of robustness and credibility of verification and auditing processes that undermine guarantees of sustainable operations (EIA 2015).

The long-term endeavor of jurisdictional approaches
NGOs have been (and will remain) very instrumental in the design and implementation of sustainability commitments, not only by their contributions to standards and methods used by the private sector as suggested in the previous example, but also by carrying out field activities in partnership with local governments and companies. One meaningful illustration of this is provided by the work of an Indonesian NGO -Institut Penelitian Inovasi Bumi (INOBU) -that operates at the local level to support one of the most promising approaches to enabling sustainability standards to materialize at scale, namely jurisdictional approaches.
The concept emerged a few years ago, with discussions around the implementation of Reduced Emissions from Deforestation and forest Degradation + (REDD+). At the time, a number of attempts were made to move forward with jurisdictional programs in different countries. A recent assessment of eight programs across continents concluded positively on their relevance, but the challenges are very significant and the proof of concept remains to be made to a large extent (Fishbein and Lee 2015). With this in mind, there is a general consensus that this is the way forward for changing scale and mobilizing sufficient resources (financial, technical, entities) to tackle the challenge of sustainability commitments.
In the case of sustainable palm oil, the concept of jurisdictional approach resorts mainly to the traceability of supply chains, lower costs of implementation, and political buy-in. In this regard, we must note the promising recent moves by several district-level officials who announced their full dedication to the development of jurisdictional approaches in their districts. Among them, the district heads of Seruyan and Kotawaringin Barat in the province of Central Kalimantan have committed to this development, following assistance and facilitation by the NGO INOBU, which organized the visit of several district heads to Brazil in 2013 so they could learn from more mature jurisdictional attempts in this country. One early effort was the facilitation of the provincial roadmap to low deforestation development, followed by more specific and concrete collaborations at the district level to conduct the mapping of oil palm smallholder locations and support their land registration.
In fact, these activities are at the intersection of several initiatives that symbolize the complexity of sustainable palm oil. The jurisdictional approach is officially presented as serving the purpose of RSPO certification at district level. The two pilot sites in Seruyan and Kotawaringing Barat were also selected as pilot sites for the RSPO initiative for jurisdictional approaches in mid-2015, as the latter may become one of the priorities for RSPO to increase its coverage of palm oil production across the world. But the local government, through its local plantation agency, is also keen on presenting these activities as serving the purpose of ISPO development as the mapping and registration (legalization) of smallholders are key priorities for ISPO. Besides, the local government assumes that concessionaires have already made decisive moves toward sustainability; hence, action at the level of smallholders becomes a convergence point for ISPO and jurisdictional approaches.
Civil servants from the district plantation agencies contribute to data collection, and it was reported that smallholder mapping also benefits from the support given by private companies such as growers and consumer goods manufacturers. So far, achievements remain modest as less than 2000 ha had been mapped as of June 2016 in each of the districts, and uncertainties remain regarding the registration and legalization of the majority of smallholders that operate on the forest estate.
Collaborations go further, with the NGO facilitating the training of smallholders with the best practices available in the company concessions to increase their productivity. In addition, interactions between concessionaires and the local government ensure that extension services will be able to take over at some point. There have also been discussions between INOBU and the local governments to put in place a system of online monitoring of concessions to increase transparency and reduce opportunities for illegal expansion.

Regional High Conservation Value assessments to support sustainable palm oil regulations
There is also scope for tripartite collaborations between government, NGOs and the private sector for assessments of high conservation value forests in order to support the implementation of local regulations on sustainable plantations. As paradoxical as it may sound, with the repeated statements by high-level officials against set-asides in concession areas and the removal of the HCV concept from the latest revisions of ISPO rules (see Section 5), decentralization in Indonesia has allowed subnational levels of the government to move forward with those of their regulations that (partially) govern the palm oil industry.
A prominent example is the province of Central Kalimantan, a pioneer in this field, with a regulation on sustainable plantations issued in 2011 (Perda No. 5/2011) covering many aspects including HCV, which were later addressed specifically by a governor's regulation (Pergub No. 41/2011). It requires concessionaires to identify HCV areas and to develop management plans to ensure their preservation. The concessionaires are held responsible for any damage occurring on these concessions. One main problem, however, is the absence of a clear definition of HCV and of the criteria to consider in the assessments, so that the implementation is weak according to key informants in the province. Yet it is worth noting that NGOs are an integral part of this process as they provide direct contributions to the drafting of these regulations.
The initiative by the Central Kalimantan Government has led to a rather similar regulation in the neighboring province of South Kalimantan (Perda No. 2/2013), and to the submission of yet another regulation for discussion by the regional Government of East Kalimantan. In the latter case, several NGOs and development agencies have joined forces to support the process and have agreed to undertake an assessment of HCV areas at the provincial level. Depending on the legislative process, there is a possibility that for the first time a regulation will provide indicative maps about HCV areas where plantation expansion should be avoided.
These various examples are intended to show the interest aroused by the concept of HCV that can be applied at different scales with different levels of accuracy and details, as its application at the regional level implies a lack of primary data collected on the ground and a reliance on secondary data. More importantly, for the purpose of this section, we can see that a variety of stakeholders have started to provide their own inputs and to collaborate, beyond private companies with sustainability commitments and including local governments and NGOs that seek to guide land use planning decisions.

Networks and central actors in the web of interactions and formal collaborations
The previous section presented a description of key activities that involve a variety of stakeholders with some of them collaborating in various ways. To complement this information and to have a more systematic coverage and understanding of the networks, we collected data from all respondents to our survey about their three main partners in formal collaborations as well as the three actors with whom they interact most frequently informally, e.g. through phone calls, email exchanges or meetings (see Question 4.1 in the questionnaire in Annex 3). We only account for interactions with more than six occurrences over the last six months. Figure 3 provides information related to the direction of the interaction, the relative level of perceived influence, and the relative intensity of in-degree interactions (for one given actor, in-degree interactions are those revealed by other actors: if A indicates interaction with B, then this is an out-degree interaction for A and an in-degree interaction for B). Therefore, the diagram enables us to identify which actors are central, which interactions are prevalent, and whether influence is correlated to the intensity of interactions. We must underline that the size of the curves representing the intensity of in-degree interactions has to be read from the perspective of in-degree interactions (interactions for one actor cited by the other actors); in other words, the diagram does not inform on the relative intensities of out-degree interactions, i.e. whether actor A cites more or fewer interactions with actors B and C. This is because categories of actors have contrasting sizes in our sample, and some key ones (especially the national government) are very small, as explained in the methodological section.

Figure 3. Network based on informal interactions.
Note: Curves represent the direction of interactions according to the respondent making the citation (clockwise, e.g. subnational government cites private sector); the size of the curves represents the intensity of in-degree interactions; the size of the circles represents the relative perceived influence for a given category of actors; the color of the circles represents the relative intensity of in-degree interactions (the darker the color the more citations of that circle by other categories). Ideally, the most influential actors should be those with the highest in-degree centrality, as other actors in the network would try to share their own views and promote their objectives with these influential actors, and also try to get knowledge from these prime sources of information. This is partly reflected by our data, with standards, national government and the private sector at the center of the game. The overlap is not perfect, though, as the national government is the most influential actor but lags behind the other two categories in terms of informal interactions. Moreover, it is important to note that the national government enjoys relatively few in-degree interactions (and no interaction declared by respondents from the private sector), which might be considered a problem and might arguably be related to the fact that debates on sustainable palm oil in Indonesia -and especially on IPOP -have shown conflicting views and statements. These might be due to the relatively low level of interactions (that is, communication) between the national government and other actors if one assumes that differing positions can at least be partially brought closer together with frequent exchanges (admittedly, the causal relationship might be inverted and limited interactions might be due to conflictual views if there is no willingness to collaborate).
This result might have to be interpreted carefully because key informants tended to indicate frequent interactions between the private sector and the government, and these were not captured in our survey. Two possible explanations come to mind: either these interactions are not the most important or frequent ones in the minds of respondents who had to pick three organizations only; or they preferred to keep silent on behind-the-scene negotiations that would contradict the official and simple story about virtuous companies facing resistance from an unaccommodating government. We must also note that small-and medium-sized producers were not included in the sample except for the smallholders' association.
However, replies to another question in our survey tend to give credit to the declared lack of interactions. To the question, "Who would you benefit from having more interactions with?", respondents disproportionately cited the national government (Ministry of Agriculture being number one and Ministry of Environment and Forestry being number two, but also including for instance the National Land Agency). Other responses were very scattered, and even the Indonesian Palm Oil Association (GAPKI) and NGOs were very far behind. This lack of exchanges is to the regret of other stakeholders (according to the survey results), who would welcome having more opportunities to sit together and address pressing issues.
Regarding actors of lesser importance in terms of (perceived) influence, we see that international and national (to a lesser extent) organizations enjoy a fairly high number of in-degree interactions that give them a far from negligible role in debates and decisions about sustainable palm oil in Indonesia. They mostly differ in terms of privileged actors for their interactions, with respectively private sector/standards/national government for international organizations and subnational government for national organizations (both directions). This logically reflects the level of these organizations, but it must be acknowledged that all of them also have a diversified set of interactions that is probably explained by their diversified nature (see Annex 2 with an analysis of the composition of the sample).
Another important observation is the marginal position of the subnational government. With limited (perceived) influence and relatively few interactions with other actors, its role seems secondary. This is a surprising finding because district (mostly) and even provincial levels of government are supposed to have considerable powers in the decision-making process with, among others, the allocation of licenses, and the design of regional land use plans. Their interactions are mostly with national organizations and growers, and quite surprisingly with research institutions probably for research and development, e.g. improvement of seedlings, or in the framework of development projects. Note: Curves represent the direction of interactions according to the respondent making the citation (clockwise, e.g. subnational government cites private sector); the size of the curves represents the intensity of indegree interactions.
The survey also covered formal collaborations and partnerships, as illustrated in Figure 4. Findings once again put the standards at the center, with by far the highest in-degree centrality: private sector, national and international organizations, and other stakeholders focus their partnerships with standards. Figure 5 shows a disaggregation of this latter category and provides enlightening results, as RSPO appears to be the point of attraction. Others such as SPOM and IPOP are on the sides, and ISPO, HCSA or POIG are not even part of these formal networks. 2 This finding shows a particular preference by stakeholders for the most established standard, i.e. RSPO -it has proven its capacity to be adopted by producers at scale and tends to be accepted by the government. On the contrary, IPOP is probably too politically sensitive for many stakeholders to move forward with formal partnerships (and lasted for about a year only; hence, it presented only limited opportunities to establish partnerships), and ISPO is seen as too weak in terms of sustainability outcomes for stakeholders to invest in its application (see Section 3 for an analysis of the stakeholders' viewpoints on the variety of standards and sustainability initiatives). Yet, there are on-going initiatives that aim at strengthening its rules, with the participation of many parties.
International organizations are worth considering in this respect. While they declare formal collaborations with a wide range of stakeholders, and hence seem to hold some potential to act as intermediaries or brokers in this field, they give priority to standards (and to RSPO more particularly). They also entertain substantial collaborations with the private sector compared with relatively few collaborations with either national or subnational levels of government whose limited involvement in informal networks is confirmed with limited formal collaborations with other stakeholders. This is 2 We must reiterate that our sample does not cover all stakeholders and captures only the three main self-stated partnerships for each respondent. It implies that other partnerships exist and might involve entities such as ISPO, SPOM and POIG. Furthermore, at least one result was found to be misleading, as UNDP declared a partnership with the Ministry of Agriculture, which focuses on ISPO but was not recorded as such. important to underline, because the hypothesis that respondents would not properly declare informal interactions with the government is weakened by the fact that formal collaborations (hence, assumed to be public) are not declared either in the survey. Overall, it supports the other hypothesis that the government is rather isolated in the networks reflected by our sample, which does not include many of its main constituencies such as medium-scale domestic producers or smallholders. Note: Curves represent the direction of interactions according to the respondent making the citation (clockwise, e.g. subnational government cites private sector); the size of the curves represents the intensity of indegree interactions.

Political context in which sustainability commitments operate
Going one step further than networks of interactions and collaborations as discussed in the previous section, we now move to the political economy question in more detail. Commitments may enjoy a number of initiatives and activities to support their implementation, but the political context in which they operate is critical to their success. This does not only refer to official statements in favor of or against given progressive moves in the sector, but also to more subtle differences of opinion and interests within governmental circles or regional executive bodies, and the evolution of laws and regulations as well as governance frameworks. We will address these aspects in this section with data from interviews with key informants and from the systematic survey, to substantiate and quantify them whenever possible and relevant (for instance with estimations of the influence of the main actors). Regarding the case of the prominent but short-lived IPOP, we will explore it in greater depth in the following section, as it provides a meaningful and enlightening case study.

Legal framework: Limited expansion of concessions … but less conservation within land banks?
The palm oil sector is subject to a great variety of laws and regulations and to oversight by public bodies as it touches on land use, forest management and the forest estate, agriculture, biofuels, community rights and others. We will focus here on the main legal documents and changes, and their implications for sustainable palm oil and the capacity of private companies to apply higher than legality standards. 3 The sector is subject to Plantation Law UU No. 18/2004, which underwent revision a decade later with UU No. 39/2014. This widely anticipated revision has important implications for company practices. Apart from articles dedicated to the respect of community rights and support for better partnerships between communities and companies, which is a key priority of the president and government for development and the use of land and forest resources, we must note provisions more directly linked to zero-deforestation commitments. In particular, the law states that entire concession areas must be developed whenever there are no technical obstacles such as steep slopes or unsuitable soils. This reflects a clear opposition to companies undertaking their own assessments of High Conservation Value Forest (HCVF) or HCS areas with resulting set-asides.
These so-called "neglected lands" have a longer regulatory history as they were already stipulated in the Neglected Lands Act (Government Regulation No. 10/2010), which allows the government to take back land that has not been converted and planted, and to re-allocate it to other companies willing to invest in their development (Pirard et al. 2015). Of course, all depends on the willingness of public authorities to apply this legal provision, and on assessments of neglected land. Past experience shows that the threat is credible, even if the possibility remains for companies to appeal decisions to withdraw licenses. For example, a case involving the London-Sumatera company ended successfully at the Supreme Court with the cancellation of the decision to withdraw the license after the judges considered that (i) the company had already negotiated with the communities about the status of the land classified as HCV, (ii) it is part of RSPO and the conservation of HCV is one RSPO requirement, and (iii) it has managed the HCV land which is thus not "neglected land" strictly speaking and according to the interpretation of the judges. This legal decision was made in 2013 and was followed two years later by an internal directive from the National Land Authority (Badan Pertanahan Nasional, BPN) to regional government heads. It encourages concessionaires (through local BPN agencies) not to convert HCV areas into plantations. In addition, it requests regional government heads to refrain from issuing licenses on these lands and not to revoke licenses for companies setting aside HCVF. Interestingly, these instructions are literally justified in the internal directive by the need to avoid the boycotting of Indonesian palm oil products.
Yet the legal implications of such documents are unclear, and they face stronger regulations. Besides the revised Plantation Law, the ISPO mandatory certification standard was also revised recently with the ministerial regulation Permentan No. 11/2015. As it is already in effect, a translation of the body of laws and regulations governing palm oil in Indonesia, with guidance for the verification of their enforcement as part of a broader 'certification system', its content is critical. The important changes are the removal of required identification and protection of HCV areas, and a new criterion on "neglected lands" that reinforces the risk of license removal when parts of the concessions are not converted.
Although these represent strong signals and legal provisions for less conservation within concessions and rebuttal of zero-deforestation commitments, other parts of the revised ISPO regulation reinforce protection based on criteria such as riparian zones, and steep slopes or floodplains, and reinforce the spirit of the moratorium on primary forests and peat land. In this respect, we must also note the ongoing discussions within the government about a new moratorium specifically on new licenses for oil palm and mining. This move will give even more importance to the issue of land banks and dormant licenses that are suspected to cover very large areas across the Indonesian Archipelago. In a context where setting aside natural forests for conservation purposes within concessions may not be an option for the legal reasons explained above, the extent and nature of the land banks will probably be decisive for sustainability in the oil palm sector.

Governance framework: Toward enhanced legality?
Legality means little if governance is weak, and we need to explore the state of the institutions that will support, or undermine, the capacity of the legal framework to operate effectively and to ultimately advance the sustainability objectives for palm oil.
Overlaps between concessions issued by various levels of government are a reality in Indonesia. There are also recurrent conflicts between companies and local populations because of uncertainties regarding property rights (Srivinas et al. 2014); illegal encroachments by migrants are also increasing in intensity and are so far poorly addressed by authorities. There is reason for hope, though. The One Map initiative, in particular, is associated with requirements to implement REDD+ and aims to synchronize and establish a single reference, standard, database and geoportal for all geospatial information in the country. Targeted information includes administrative boundaries, land ownership, land use and various kinds of thematic environmental maps. By providing a single point of reference, it is expected to resolve geospatial inconsistencies across government and institutions as well as land tenure and land use conflicts, to inform land use planning, and ultimately to solve once and for all the overlaps between permits issued by different bodies. Nineteen ministries and institutions are involved in the making of the maps and the Geospatial Information Agency (BIG) is responsible for the technical aspects of the policy. A presidential regulation (Peraturan Presiden) issued on February 2016 declared the acceleration of the One Map policy as part of an economic stimulus package and aims to finish all thematic maps at a scale of 1:50,000 by the end of 2019.
The lack of clear and accepted boundaries for concessions is a very serious problem that cannot be tackled by laws only, but that requires better governance and greater political will. A recent study in Riau on the propagation of fires in the province has provided evidence of the scale of the problem with a mismatch between legal concession boundaries and effective operations on the ground (Gaveau et al. 2016). Not only are one-third of concessions occupied by farmers, but also almost one-third of operations by registered companies take place outside of the legal boundaries -and this is a very conservative estimate as it does not include medium-sized companies that fly under the radar altogether.
These issues go beyond One Map and relate to other governance aspects such as the fight against corruption affecting the licensing process and the challenge of acknowledging customary land. It is plausible that licenses can sometimes be obtained through bribes at the district level, as suggested in an assessment of permits allocated to land-based industries, including mining, and tree or agriculture plantations (ICEL 2013). This assessment concludes there are frequent if not systematic violations of law in the allocation of permits, which carry not only the risk of inconsistencies and overlapping of boundaries but also expansion in areas that should be preserved. Another problem related to governance is the incapacity to access practice documents that are supposed to be public. In this context, opacity remains and the absence of transparency makes the challenge of enhancing sustainability more acute. This was indeed identified as a key aspect of sustainability by respondents to the survey, as little progress can be expected as long as this governance problem is not solved or at least improved (see Table 3 in Section 3.3).

An overall defiant government: Beyond press statements, what do stakeholders think?
The road to sustainability is not an easy one and official statements by the government, apart from the various legal aspects analyzed above, have not helped much so far if one considers corporate sustainability commitments to be relevant. Producing countries such as Indonesia have reacted with strong statements about national sovereignty and their absolute right to decide on what rules must be applied on their land (Jakarta Globe 2015). While one may argue that a diversity of views might coexist within the government, it is fair to say that the 'hardliners' who defend their strong views against private commitments have never been publicly challenged. This gives the impression that stakeholders are offering resistance, according to our key informants.
The most common arguments publicly voiced against sustainability commitments refer to foreign interference in domestic affairs, because of trade and competition interests by OECD countries that are believed to be at a disadvantage in the face of what is presented as being like a 'miraculous crop' (e.g. at a meeting held in Jakarta to inform the Indonesian diaspora in 2015). Other arguments refer to neo-colonization through the promotion and mainstreaming of foreign values about the environment (e.g. as espoused by a former ISPO director at the ICOPE conference in Bali in 2015), and expected economic losses with limited expansion. Yet, as recapitulated in Table 2 (Section 3.2), these statements are not widely shared by stakeholders: 15% of respondents (and 0% of subnational government respondents) believe that zero-deforestation commitments are going to reduce economic opportunities for Indonesia, and only a third believe that these commitments result in norms from abroad being imposed.
But increasingly, the official discourse has focused on the risk of exclusion of third-party actors from supply chains because of the requirements for supplies set up by committing companies. This threat is real and acknowledged by the main companies (interviews with key informants), as well as by half of the respondents in our survey (and 83% of the respondents from the private sector). It is commonly accepted that a great number of medium-sized companies are engaged in unsustainable and often illegal activities all over the country and they would be automatically excluded from the supply chains as far as companies with higher standards are concerned. Moreover, smallholders are often involved in activities that may not qualify as sustainable or even legal if one considers the validity of the land certificates that they usually hold. We do not discuss here whether these smallholders should be considered legal, which is a point made by those arguing that smallholders' rights to land should have been recognized by the state in the past; instead, we state that a strict interpretation and application of the rules of commitments would de facto lead to exclusion even if most recent developments have tended to relax these requirements according to a principle of reality (see examples of collaborations for smallholder mapping in Section 4.1).
The role of the government is illustrated by our survey data, as we asked a question specifically on actors hindering change. 4 Results are straightforward, as the main stakeholders in this field -the Ministry of Agriculture and the private sector -stand out as the opposing forces named by the vast majority of respondents (Table 5). There is thus a consensus that the main public institution in charge of regulating the sector and the companies operating on the ground are those resisting change toward more sustainable palm oil. This is further reflected by the fact that the Ministry of Agriculture, national government, palm oil producers and GAPKI (Indonesian association of palm oil producers) make up about half of all responses.
While the presence of producers in this list might sound surprising at first sight and considering the wave of sustainability commitments (and RSPO) pushed by actors from the private sector, it refers in fact to another subcategory of producers that expressed disagreement with more sustainable practices. It shows how heterogeneous is the private sector, and this may probably be applied to the Ministry of Agriculture as well, where there might not be one single position with respect to sustainability initiatives. Note: The question asked was about "institutions". Although some responses might not appear, strictly speaking, to be about institutions, we report answers as provided by respondents.
Despite extensive power over the issuance of licenses and monitoring of practices on the ground, subnational authorities have a much lower profile in the minds of respondents in terms of hindering change. This is a good sign for the future if answers reflect the direct experiences of respondents dealing with these authorities.
Apart from these primary actors, respondents cite a large number of institutions, individuals or even phenomena such as corruption. But another category, namely national NGOs, receives quite a few citations. This deserves our attention because these actors may usually be seen as progressive and are heavily involved in advocacy activities for the improvement of practices both for the environment or local populations and smallholders. The explanation might be found in the source of these quotations; indeed, almost all of these citations come from respondents from the subnational government category. A few respondents elaborated on the fact that the type of advocacy was counterproductive or "bad." It is not possible to draw conclusions about this aspect as it may deserve further exploration, but the likely explanation might be twofold: there is a widespread perception by local governments (and often companies) that advocacy NGOs are negative in a systematic way; and some NGOs (isolated cases?) might sometimes have counterproductive attitudes locally in their interactions with local governments, notwithstanding the relevance and rightness of their positions and activities.
We also asked about the main obstacles to an effective implementation of "sustainability standards in general" (as formulated in the question). The question is closed-ended, but respondents can make their own suggestions if they wish and if they feel that their own views are not well-represented by the predefined list of answers. Besides, the three obstacles selected by each respondent are ranked in order of importance, which allows us to identify whether some obstacles represent outstanding problems in the minds of stakeholders and across categories.
Results (see Table 6) point to a lack of political support including in terms of regulatory/legal frameworks. Moreover, this is also by far the main notion that prevails in the minds of respondents as the number one obstacle to an effective implementation of sustainability standards. This result is absolutely critical for our analysis as it dismisses the possibility that technical issues and reluctance from the private sector would be the main barriers to changing practices on the ground. Our survey conveys a clear message and points to the role of politics and regulations, which is in line with the previous discussion of both the official statements by the government with respect to emerging commitments and the new regulations that impede their implementation. It is worth noting that technical aspects such as monitoring with reliable traceability systems, the development and availability of methodologies for an application of the concept (e.g. HCV and HCS), or more generally lacking rules of implementation of the commitments, are not identified as significant issues. Whether this is true just for standards or for commitments as well remains to be explored in detail, because respondents may have a different understanding of the scope of the question depending on their level of expertise and knowledge. For instance, some aspects or tools are applicable specifically to zero-deforestation commitments, such as traceability systems or HCS. However, it looks like the future of sustainable palm oil, according to stakeholders' perceptions, is in the hands of political forces rather than standard developers. To put it differently, at least in a first stage and as a necessary condition, an unsupportive government would ruin the efforts of most progressive actors in this field.
We also grouped answers into broader categories as also presented in Table 6. It allows us to confirm the trends with the overwhelming prevalence of political, legal and regulatory aspects that represent two-thirds of the first choices by respondents. This also allows us to identify economic and financial aspects as also significant, as they make up most of the rest of the first choices by respondents. These results reinforce the connections with the previous question on forces opposing change: indeed, it certainly explains why (part of) the private sector and producer associations are seen as reluctant to engage more pro-actively in sustainable palm oil. The associated costs might be the main reason for this, unsurprisingly, even though there is no comprehensive study on this to the best of our knowledge (note that results are sometimes surprising, as illustrated by timber certification schemes with debated -and sometimes profitable -cost implications over the medium to long term).

Actors of influence and expected action items to advance the sustainability agenda
Influence was assessed through perceptions of stakeholders as respondents were asked to list the five most influential actors. Results thus provide an assessment of perceived influence rather than an objective measure of their actual influence. One first observation is that responses are varied and as many as 83 different actors were cited spontaneously. This ranges from very specific individuals such as Leonardo di Caprio because of his visit to Aceh Province at the time of the survey, to media or the financial sector in general. Therefore, to make the best use of these data and to identify patterns, we coded the information and used categories, as presented in Figure 6. Source: authors' survey Three groups stand out: government, private sector and standards. This is a remarkable result because it shows a balanced landscape where influence exists and is applied in various ways and at various stages: legal framework and law enforcement (government), practices on the ground (private sector) and definition of sustainability standards (standards).
The national government is perceived as being much more influential than the subnational government. Results are in fact probably even more in its favor as there is a bias in the sample, with twice as many representatives of the subnational government … who cite themselves very heavily (13 times out of a total of 20 citations). Therefore, this shows very clearly that action at the highest level of the government is the one that predominantly guides the destinies of the oil palm sector according to respondents. More specifically, the Ministry of Agriculture is ahead of the Ministry of Environment and Forest with respectively 26 and 18 citations.
Because representatives of the private sector in our sample belong to progressive companies with the desire to reach sustainability, and their direct experience is assumed to be valuable in determining who is influential, we looked at their responses with great interest. They point to central government and to standards as their priority (there was not a single citation of a subnational government), likely for their negative and positive influences, respectively. For consumer goods manufacturers and retailers specifically, the private sector and related associations are also cited: this is due probably to the fact that they are downstream in the value chain and therefore consider that their suppliers' actions make a difference.
Regarding standards, a few were cited: IPOP, RSPO and ISPO with respectively 17, 14 and 8 citations. This puts IPOP at the level of GAPKI, but they tend to pursue contradictory interests. Admittedly, this is true to some extent only, as even though GAPKI was listed as one of the main institutions opposing change, we must underline that it includes companies with various objectives and includes IPOP members; the latter clearly form a much more homogeneous group linked together by clearly defined common goals.
Looking at who cites whom, there is a clear divide between national-and international-level institutions, with the former members neglecting standards in their citations to the benefit of government, especially at the subnational level. This is probably due to their operations taking place mostly at the local level.
The previous results about the national government being the main actor that hinders change as well as being the most influential, need to be put into perspective by examining answers to the question, "Who would you benefit from having more interactions with?". Indeed, respondents disproportionately cited the national government (Ministry of Agriculture being number one and Ministry of Environment and Forestry being number two, but also including for instance the National Land Agency). Other responses are very scattered, and even GAPKI and NGOs are very far behind. Therefore, there appear to be many bottlenecks on the way to sustainability, which will not be solved until more opportunities are created to sit together and address pressing issues in a constructive manner.
Hints at which issues matter and how to address them are provided through the question on priority actions and policies to promote sustainable palm oil. This is in fact another way to ask the question about obstacles, as the right policy and action items would be the ones capable of helping overcome barriers. Results presented in Table 7 are clearly headed in the same direction, as legal, regulatory and governance aspects are flagged and they can be viewed as pre-conditions for sustainable palm oil to materialize in Indonesia. Related answers make up more than half of the top-priority policy by respondents and the bulk of all citations altogether. For instance, the One Map policy remains an important ingredient of the sustainability recipe in Indonesia and shows that there is still a long way to go.  When grouping answers in broader categories, results suggest that a number of financial and economic supportive policies could be helpful in the form of subsidies or specific support to early replanting efforts that the Indonesian Estate Crop Fund (BPDP) is currently undertaking. Yet this is obviously a matter of debate, as the palm oil sector has contributed greatly to the wealth of a number of investors owing to comfortable profit margins. Smallholders are also a potentially important area of action, whether by including them in supply chains or providing support to improve their practices.

From creation to (rapid) dissolution
An important event took place right after the survey was completed, namely the disbandment of IPOP. Not only did this initiative embody the zero-deforestation commitments and have a major influence on the movement and the setting of rules, but it had a very high visibility and became the target of the government. While this story is still to be fully written and its far-reaching consequences are yet to be fully captured, we provide some early elements and analysis in order to get a sense of what that means for the future of sustainable palm oil in Indonesia.
First, we need to make it clear that the end of IPOP might not imply the end of zero-deforestation commitments, as its member companies retain their objectives individually. Their sustainability contract with civil society, NGOs and consumers remains to be fulfilled with or without IPOP as a conduit by which they advocate their cause with the government and find the most promising and feasible means of implementation.
The Indonesian Chamber of Commerce and Industry (KADIN) was initially interested in the promotion of business-as-usual practices in the palm oil sector, but the joint effort with the World Business Council for Sustainable Development (WBCSD) gave birth to the regional chapter of the Indonesian Business Council for Sustainable Development (IBCSD) in 2011 with further emphasis on land use activities.
As a result, and in a context where palm oil was a toxic issue with lots of pressure from civil society, as described in the introduction of the report, it was decided to focus on this sector in order to move forward the sustainability agenda. IPOP was eventually established as a result of this vision in 2014 with five member companies.
The initiative pursued a number of objectives represented by the four pillars of the pledge: improving environmental stewardship, strengthening policies and regulations, expanding social benefits and improving competitiveness of Indonesian palm oil. These cover a range of issues from conservation of HCV/HCS land to improvement of smallholders' productivity/access to finance or support to implementation of One Map. IPOP also mixed lobbying activities with the government, engagement of stakeholders (including research organizations) with participation and communication, research and development based on pilot cases in provinces, or connection to on-going processes such as One Map (IPOP 2016a).
It must be recognized that the emphasis was progressively put on the enhancement of smallholders and land tenure, with a lower profile on conservation issues in order to stick to the government's priorities and lower the risks of conflict in a context that faced increasing tensions over time. But one likely significant event was the official support provided by the US Embassy, along with the signature on an agreement between IPOP and the United States Agency for International Development (USAID) in September 2015, which might have underlined the foreign component of the whole initiative.
There seem to have been ups and downs in the relationship between IPOP and the government of Indonesia during the two years following its establishment, and the general impression that emerges from our interviews with key informants is that IPOP thought that most legal challenges could eventually be overcome to sustain its existence and operations. In particular, the issue of set-asides that became a symbol of sustainability pledges was probably seen as solvable in light of a number of existing legal provisions, its mention in the ISPO regulation (before the latest revision as we explained in Section 5.1) and the efforts toward the improvement of smallholders' practices as one high-profile activity put forward by IPOP.
Despite this attention paid to communication and the professionalism of its staff, it is possible that the pressure put on staff members by NGOs to make rapid progress resulted in frictions with highlevel officials who may have preferred a slower tempo to make reforms in such a sensitive sector with many vested interests and connections between the worlds of politics and business. The possibility of a significant reduction of the expansion in the Papuan provinces, where the presence of good condition forests would make plantation development highly incompatible with the strict application of rules promoted by IPOP, has probably been instrumental in its demise.
Tensions reached a peak on 17 February 2016 when the Ministry of Agriculture, through the voice of the director of plantations, officially gave an ultimatum to IPOP to dissolve and threatened to ban its member companies otherwise. This was the result of a number of attempts by the ministry to force IPOP companies to resume purchases of fresh fruit bunches (FFB) from smallholders, yet with medium-sized companies in mind as reported by our key informants. The main legal accusation was that IPOP violated anti-trust regulations in the country by forming a cartel accused of excluding other players from the market.
While these accusations may have been challenged in courts by the incriminated companies, the cost of such a process would have been extremely damaging for their image and relationship with the government. This lose-lose strategy was logically rejected by IPOP and it issued a press statement announcing its dissolution (IPOP 2016b). Incidentally, IPOP staff were reported to have seen a leaked draft regulation that was about to be submitted to the government in order to ban IPOP. Whether this would have happened remains uncertain, but the risk was certainly too high to take the chance.

IPOP received appreciation from the majority of stakeholders
Our survey gathered opinions on IPOP, as reflected in Table 1 (Section 3.1). A first rather surprising result is that a majority of respondents considered the initiative to be aligned with government policy despite our account of tensions and ultimate dissolution in response to threats by the Ministry of Agriculture (alignment with the NGO's objectives and requests is notably much higher, with 83% of responses). Furthermore, it was not viewed as counterproductive by two-thirds of the sample.
Overall, the respondents had a positive appreciation that is reflected by their view that IPOP's existence was relevant (80%), timely (two-thirds), or justified (81%). Besides, the foreign influence was cited by two-thirds of respondents but not always as a negative aspect, as some respondents mentioned the need for Indonesia to get inspiration from sustainability aspirations coming from abroad.
The most significant negative comment relates to the risk of smallholders' exclusion, which is perceived as credible by a (small) majority of respondents, and which echoes public debates, IPOP's work agenda and the government's criticisms.
To complement these findings, we asked two questions specifically on the risks and opportunities for Indonesia generated by IPOP members' commitments (Table 8). Risks revolve mostly around financial and economic aspects, which take mostly the form of limited access to markets for non-IPOP producers and to a lesser degree of higher costs of production. These are indeed the two main issues that trigger criticism from the government and the share of the private sector that has not made similar commitments. Accusations made by the Government of Indonesia that IPOP would represent a cartel of producers with the intention of dominating a large share of the export market in the long run is indeed perfectly reflected in this result. We must also notice that this view is shared by all categories of respondents who consistently rank this as the number one risk, except for standards and international donors. Regarding higher costs of production as a consequence of the commitments being implemented, several explanations can be proposed and are open to discussion. It sounds like a reasonable assumption that respondents expect stricter standards of production will induce less development of land, higher management costs and a number of transaction costs that can eventually increase the overall costs of production. But others could argue that this particular effect is faced by producers or other agents along the value chain, so that Indonesia as such would not suffer particularly much, especially as other producers without commitments might not be affected. Consumers could also agree to pay a premium for more sustainable products, but evidence so far tends to prove that it is not the case, at least not at a large enough scale to absorb a significant share of the total production. Yet another interpretation could be that respondents anticipate some sort of spill-over effect, whereby the majority of the sector would eventually have to align with these higher standards of production.
Market fragmentation is another way to look at these results because it is reflected in two popular responses: "less access to market for non-IPOP producers" and "smallholders' livelihood negatively affected" received altogether two-fifths of all responses. A recurrent criticism against IPOP was indeed that third-party actors and particularly smallholders might not be able to raise their standards and comply with requirements set by the zero-deforestation commitments so that they would be excluded from the value chain. This worry is amplified by the fact that the sector is controlled by a few large-scale processors and traders, so that consequences of higher standards would be faced by most producers.
It is interesting to observe that "loss of credibility for the government" is the third most cited answer (in fact, at the same level as "higher costs of production") and it must be noted that half of the respondents chose it as one of their three choices. It means that, for a majority of stakeholders, the government has been playing a dangerous game with its negative reactions to the development of IPOP, which is appreciated and credible across the variety of stakeholders as we showed in previous sections. This risk seems not be a minor one, especially when related to previous findings that the Ministry of Agriculture and more broadly the government are viewed as the main forces opposing change toward sustainability. In other words, the recent disbandment of IPOP may sound like a Pyrrhic victory for the government. A second question looked at the brighter side of the story, with the benefits and opportunities created by IPOP members' commitments for the country (Table 9). The main message is straightforward and expected: these commitments with better practices would mitigate negative impacts on the environment and improve the management of natural resources overall. One related aspect is the enhanced provision of environmental services, including reduced carbon emissions, which are widely cited.
But even if environmental aspects are ahead of the others, results suggest that IPOP members' commitments may also have substantial positive economic impacts apart from the expected higher costs of production as identified in the previous question on risks (Table 8). This is indeed an important result, in our opinion, that almost a third of first choices by respondents point to positive economic consequences from the application of commitments. And "increased competitiveness" stands out, which could be interpreted as the fact that these commitments would give Indonesia the opportunity to position itself more perfectly on a world market that might become more competitive in terms of environmental and social standards. In other words, IPOP might be seen as a good opportunity to lead the race in a context when other regions in the world (Central and West Africa, South America) might progressively enter into production and put Indonesia's dominant position at risk.
"Better image" is also widely cited and this reinforces our line of reasoning that these commitments are viewed as an opportunity for the country to change the perceptions of markets and consumers about the impacts of the oil palm expansion. Furthermore, benefits would extend much beyond and embrace possibly even more critical governance aspects, such as "clarified land tenure" which is ranked fourth, but also "less corruption". In other words, commitments are seen as ways to cleanse the sector and trigger even more fundamental reforms such as those related to land tenure.
Keeping in mind all the limits associated to the method applied in this PNA analysis -e.g. a relatively low representation of the central government in our sample, and the biases associated to individuals as respondents vs. the views of their institutions -we were able to gather a large amount of information that supports a number of conclusions and sheds light on the future of the recent wave of corporate sustainability commitments in the Indonesian palm oil sector.
Standards and initiatives have contrasting visibility and impact among stakeholders -but they have collectively moved the lines and enhanced debates and knowledge around the issue of sustainabilityand RSPO stands as a reference in the domain of sustainable palm oil. The efforts by the Government of Indonesia to promote its own standard with ISPO are still to gain traction. IPOP was a well-appreciated initiative and a symbol of zero-deforestation commitments, but the opposition to it by the government and conflicting interests have resulted in its disbandment.
The lack of progress for sustainable palm oil practices on the ground, in the view of respondents, seems to be caused by political and legal barriers rather than technical challenges or economic losses at a country level. The legal framework is not clearly making progress to accommodate pledges for enhanced sustainability, and the risks (as perceived by the majority of respondents) related to the exclusion of third parties (smallholders and medium-sized companies involved in ineligible practices) have not enabled a wide political support to materialize.
In this context, it was not surprising to have the government being mentioned as particularly influential beside standards and the private sector. But the lack of stated interactions especially in the form of formal collaborations and partnerships, as well as the perception of respondents in our sample that it tends to hinder change, do not necessarily lead to optimism. Communication among stakeholders will have to improve, not only to raise the bar of sustainability beyond existing ISPO requirements and to provide a more conducive environment to corporate sustainability commitments, but also to enhance the rule of law and improve governance. Indeed, legality and law enforcement are absolute prerequisites for cleansing the sector of its worst players and practices.
The relative absence of consideration for High Carbon Stock approaches and the prevalence of High Conservation Value assessments as paving the way for sustainable palm oil to materialize suggest that priorities may need to be aligned on the latter. When this result is associated to the identification of legality and good governance as critical factors, it pleads for increased efforts to enforce regional initiatives to support HCV assessments to guide decisions on land use planning. This is all the more important when considering emerging moratoria that might limit the expansion of concessions across the country yet without addressing existing concessions ('land banks') that are partially covered by standing natural forests.
We are now at a crossroads. The disbandment of IPOP means that an opportunity for the frontrunners of the private sector to engage with the government, at the highest level, was missed. Moreover, while the organization was involved in concrete operations on the ground with other stakeholders, including heads of local governments, the future is now unclear. The state certification system ISPO is still lacking international credibility as a sustainability standard despite on-going efforts to strengthen its rules and widen its application by companies and smallholders. POIG and SPOM enjoy little recognition among stakeholders. The popularity and relatively high level of adoption of RSPO may promise a future for the emerging RSPO Next, except if it turns out to be adopted only by a small group of companies. This new voluntary standard is in line with on-going initiatives supported by collaborations between NGOs, local government and the private sector to develop jurisdictional approaches and assist smallholders with mapping, legalization and training with the aim of attaining sounder practices and enhanced productivity.

Annexes Annex 1. Sample
The sample exhibits a slight over-representation of NGOs both national and international, as they cover about a third of all respondents. But when looking at their activities, it becomes clear that they constitute a heterogeneous group in itself if we consider the great diversity of their activities (see table below). While it is not surprising to see advocacy and community support as main areas of activity, results show that their involvement goes way beyond and can even appear to be substantial in the design or laws, regulations or policies. This is actually an interesting information in itself as it points to their direct influence on operations through legal drafting, which contradicts conventional wisdom according to which NGOs would usually be sidelined by governments. Another observation relates to advocacy as its meaning might be ambiguous. Indeed respondents might have different interpretations of the term if we refer to the fact that two-thirds of those practicing advocacy also undertake independent research, and vice-versa. Its meaning might therefore be very broad, and almost two-thirds of all respondents across categories declare involvement into advocacy activities.
Overall, it must be noted that the sample is made of relevant stakeholders with a deep understanding of the sector as half of the respondents have been working on palm oil for more than a decade and three-quarters of them for more than five years. It indicates that the sample if sound and almost only includes institutions with substantial experience (10% had less than two years of experience).
Over the course of the data analysis we had to try a series of ex post coding for a number of questions in order to improve the interpretation. This could either apply to categories of respondents in our sample or answers. Whenever necessary they will be explained, but we can already present the most widely used across the analysis with nine categories of respondents. The palm oil sector has been targeted by NGOs for its alleged negative environmental and social impacts. In this regard Indonesia represents a major challenge because it is home to some of the largest tropical forests in the world. A recent wave of corporate sustainability commitments peaked with the New York Declaration on Forests in September 2014, which emerged amidst the development of other standards and initiatives toward sustainable palm oil production. This process has made this field very complex, especially in Indonesia. The present study aims at clarifying the positions taken by the various stakeholders and assesses the level of political support and the functioning of policy networks.
Results from our Policy Network Analysis based on the survey of 59 institutions representing all types of stakeholders (e.g. government, corporate, NGO) at all levels (international, Indonesian and local) show that standards and initiatives for sustainability have contrasting visibility and impact among stakeholders. In this context, RSPO stands as a reference, with the efforts by the Government of Indonesia to promote its own standard with ISPO yet to gain traction. While IPOP was a well-appreciated initiative and a symbol of zero-deforestation commitments, opposition to it by the government and conflicting interests have resulted in its disbandment. Overall, the lack of progress for sustainable palm oil practices on the ground, in the view of respondents, seems to be caused by political and legal barriers rather than technical challenges or economic losses at a country level.
CIFOR Working Papers contain preliminary or advance research results on tropical forest issues that need to be published in a timely manner to inform and promote discussion. This content has been internally reviewed but has not undergone external peer review.

Center for International Forestry Research (CIFOR)
CIFOR advances human well-being, equity and environmental integrity by conducting innovative research, developing partners' capacity, and actively engaging in dialogue with all stakeholders to inform policies and practices that affect forests and people. CIFOR is a CGIAR Research Center, and leads the CGIAR Research Program on Forests, Trees and Agroforestry (FTA). Our headquarters are in Bogor, Indonesia, with offices in Nairobi, Kenya, Yaounde, Cameroon, and Lima, Peru.

DOI: 10.17528/cifor/006528
The CGIAR Research Program on Forests, Trees and Agroforestry (FTA) is the world's largest research for development program to enhance the role of forests, trees and agroforestry in sustainable development and food security and to address climate change. CIFOR leads FTA in partnership with Bioversity International, CATIE, CIRAD, INBAR, Tropenbos International and the World Agroforestry Centre.