The FPIC Principle Meets Land Struggles in Cambodia, Indonesia and Papua New Guinea

Social and environmental safeguards are now commonplace in policies and procedures that apply to certain kinds of foreign investment in developing countries. Prominent amongst these is the principle of free, prior and informed consent (FPIC), which is commonly tied to policies and procedures relating to investments that have an impact on ‘indigenous peoples’. This paper treats international safeguards as a possible manifestation of what Karl Polanyi called the ‘double movement’ in the operation of a capitalist market economy. Our concern here is with the way that the FPIC principle has been applied in struggles over the alienation of land and associated natural resources claimed by indigenous peoples or customary landowners in three developing countries—Cambodia, Indonesia and Papua New Guinea. Case studies of recent land struggles in these countries are used to illustrate the existence of a spectrum in which the application of the FPIC principle may contribute more or less to the defence of customary rights. On one hand, it may be little more than a kind of ‘performance’ that simply adds some extra value to a newly created commodity. On the other hand, it may sometimes enable local or indigenous communities and their allies in ‘civil society’ to mount an effective defence of their rights in opposition to the processes of alienation or commodification. The paper finds that all three countries have political regimes and national policy frameworks that are themselves resistant to the imposition of social and environmental safeguards by foreign investors or international financial institutions. However, they differ widely in the extent to which they make institutional space for the FPIC principle to become the site of a genuine double movement of the kind that Polanyi envisaged.


Introduction
This paper explores the way that international norms do or do not make their presence felt at the margins of the world system, and the kinds of agency that are involved in the process of fostering or hindering their translation from the centre to the periphery. More specifically, we are concerned with those international norms that have come to be known as safeguard policies, which are intended to protect marginal or vulnerable people from economic activities or forms of 'development' that threaten their livelihoods. Within this body of international soft law, we are especially interested in the translation of one particular norm, the principle of free, prior and informed consent (FPIC), through the different institutional frameworks that prevail in three countries-Cambodia, Indonesia, and Papua New Guinea-and with specific reference to the distribution of rights to renewable resources on customary land or land occupied by indigenous peoples.
The set of international norms that includes the FPIC principle appears to constitute a globalised version of what Karl Polanyi called the general principle of 'social protection', in what he described as a 'double movement' that sets some limits to the application of free market principles in the development of industrial capitalism [1] (p. 138). Of course, Polanyi could not have foreseen the specific constellation of safeguard policies that has emerged in recent decades, let alone the global institutional architecture in which they have been framed. However, the idea of a double movement still seems relevant to any form of political ecology that does not simply assume the triumphant march of neoliberalism across every field of public policy [2,3]. What are now framed as principles of social and environmental protection are still motivated by a recognition that some things in life cannot be turned into pure commodities without threatening the continued existence of the social and natural worlds in which markets operate [4][5][6]. One of those things is the 'false' commodity called land.
Polanyi treated land as a false or fictitious commodity because it has natural qualities that are not the products of human labour. For this reason, land cannot be wholly extracted (or abstracted) from social relationships that are beyond the reach of the market. However, it is not necessary to adopt an old-fashioned labour theory of value in order to appreciate the point that human beings can be 'attached' to land-and more broadly to 'nature'-in ways that lead them to resist the process of commodification. Indeed, questions about the capacity of market mechanisms to solve a range of social and environmental problems have been central to debates about the definition, measurement and achievement of 'sustainable development' for the past half century. While Polanyi built the idea of a double movement on the necessary imperfections (or externalities) associated with markets in things that are not produced for the purpose of being exchanged as commodities, he also detached the contest between social protectionism and economic liberalism from the economic interests of different social classes. In the context of the current world order, the contest is further detached from the class struggle since it is no longer confined to the level of the nation-state, or any particular level of political organisation, but has a vertical, as well as a horizontal, dimension, in which global, regional, national and local actors take up different positions for all sorts of different reasons [7][8][9].
At a global level, the safeguard policies of the World Bank Group may be construed as the focal point of a double movement within an organisation that is often portrayed as a single-minded bastion of neoliberal theory and practice [10][11][12][13]. The Bank initially sought to disseminate these policies to client governments through the mechanism of loan conditionality [14,15], and then tried to 'sell' them to other financial institutions and multinational corporations as global pillars of 'corporate social responsibility' [16,17]. However, the Bank's critics have questioned the success or sincerity of both of these endeavours [18][19][20][21][22][23].
Those who doubt the possibility of a genuine double movement within the corridors of the World Bank might instead find it amongst the various agencies of the United Nations (UN), since these bodies have generated a much larger number of safeguard policies that are meant to be ratified and implemented by all of the member states, and not just those that need to borrow money from the Bank. However, this should not lead to the assumption of a simple distinction between hard and soft forms of international law, since governments can easily evade or ignore their commitments to international conventions, just like the multinational companies that volunteer to adopt them as separate measures of their own corporate responsibility.
Like many members of his generation, Polanyi saw capitalism collapsing under the weight of its own contradictions, including those contained in the double movement [2], yet this does not mean that current versions of the double movement are simply echoes of the global contest between socialism and capitalism. The persistence of various forms of social and environmental protectionism is only one part of an argument that contradictions of various kinds survived the end of the Cold War. Another part of the same argument would be that (economic) neoliberalism has not turned out to be the only other power in a bipolar world [24][25][26][27][28].
Even if neoliberal globalisation does constitute one side of a contemporary double movement, there is no set of international norms that provides a single and coherent alternative to its global reach. That is one reason why we have limited our present discussion to the FPIC principle alone. Furthermore, our present concern is not to analyse the political process through which this principle has achieved its current standing amongst the different global standards with which it seems to be aligned. Instead, our aim is to reflect on the way that one particular standard has been applied to land struggles in different national contexts, and what kinds of double movement have either been encountered in this process or explain its outcome.
While it is possible to think of each instance of a double movement as a contest between groups of protagonists with different policies or ideologies, our preference is to think of it as one of the internal contradictions that account for a change in their mutual relationship. The two sides to a contradiction are not to be conceived as two groups of actors with opposing interests, even if such groups exist, but as the terms of an institutional paradox or dilemma that impels the continuation of a political argument [9,[29][30][31]. In the present case, this means that we are not looking to evaluate the results of different land struggles so much as to track the movement of the terms in which their causes and effects are being represented.

The Principle of Free, Prior and Informed Consent
Articulations of the FPIC principle have generally been tied to the identification of a specific class or category of people whose attachment to land is liable to be disrupted by specific forms of capital investment. This point is clearly articulated in Article 32(2) of the UN Declaration on the Rights of Indigenous Peoples (UNDRIP): 'States shall consult and cooperate in good faith with the indigenous peoples concerned through their own representative institutions in order to obtain their free and informed consent prior to the approval of any project affecting their lands or territories and other resources, particularly in connection with the development, utilization or exploitation of their mineral, water or other resources' [32] (p. 12).
Indigenous (or tribal) people have commonly been featured as the subjects from whom FPIC needs to be sought because they have been defined as people for whom land 'is life', not a commodity, as people who 'belong to the land', instead of having rights to dispose of it, or as people who do not subscribe to the modern (yet possibly false) disjunction between nature and culture that has enabled the conceptual (and practical) separation of land from people. The paradox or contradiction here is that the principle requires them to agree that this attachment should be broken in order for 'their' land or 'their' resources to be alienated and exploited. This contradiction helps to explain why different iterations of the principle have come up with a range of different ideas about the content of such an agreement, and this in turn has opened up further debate about the identity of the people to whom it should be offered [33][34][35][36][37].
While 'indigenous and tribal populations' were identified in the original version produced by the International Labour Organization (a UN agency) [38], very few governments paid any attention to the principle until the World Bank started to produce its own safeguard policies in the 1980s. The Bank incorporated a version of the FPIC principle into its policies on 'indigenous peoples' and 'involuntary resettlement', primarily in response to complaints about the negative impact of some major infrastructure projects that it had funded. These were both conceived as environmental policies because of the way that the 'environment' was represented by the American environmentalists who led the campaigns [10] (p. 631), but their separation raised an obvious question about the rights of people who are evicted from their land without having a truly 'indigenous' attachment to it. Another peculiarity of this version was that the 'C' was taken to stand for consultation rather than consent, mainly because the assignment of a right of refusal to indigenous people, or people threatened with eviction, would constitute an affront to the sovereignty of national governments [39,40].
It was only in 2012 that the Bank began to adopt the stronger version of the FPIC principle, and even then, it was only adopted by the Bank's private investment arm, the International Finance Corporation (IFC), which does not lend money to governments, and only applied to 'indigenous peoples' [41]. The two arms of the Bank that do lend money to national governments did not adopt the stronger version until 2016, but when it did so, it elaborated on the definition of 'indigenous peoples' by remarking that there are countries in which they 'may be referred to by other terms, such as "Sub-Saharan African historically underserved traditional local communities", "indigenous ethnic minorities", "aboriginals", "hill tribes", "vulnerable and marginalized groups", "minority nationalities", "scheduled tribes", "first nations", or "tribal groups." As the applicability of the term "Indigenous Peoples" varies widely from country to country, the Borrower may request the Bank to use an alternative terminology for the Indigenous Peoples as appropriate to the national context of the Borrower' [42] (p. 106). While this statement seems to expand the range of people to whom the FPIC principle might apply, it also concedes a point formerly made in the Bank's defence of the weaker version of the principle-that its translation is a matter for negotiation between the Bank and its clients. Indeed, some commentators believe that the Bank is now more reluctant to impose its safeguard policies on client governments because of the availability of other sources of international finance with no such strings attached to them [23].
Adoption of the FPIC principle by foreign investors in developing countries has partly reflected the translation of the IFC's own 'performance standards' into what are now known as the Equator Principles-a set of safeguard policies that banks and insurance companies are meant to apply to their corporate clients, and which the clients can then adopt on their own account in order to satisfy their shareholders as well. However, the principle has also been adopted in the construction of sectorspecific guidelines by which the producers of specific commodities seek to satisfy the demands of their consumers-especially consumers in developed countries. Examples would be the guidelines first formulated by the Forest Stewardship Council (FSC) in 1993, or those first formulated by the Roundtable on Sustainable Palm Oil (RSPO) in 2005. This second form of promulgation might therefore be assigned to a 'market governance paradigm', as opposed to a 'corporate responsibility paradigm', but that would assume a separation of consumers from shareholders and other 'stakeholders' that is mediated and dissolved through multiple forms of communication between them.
If we ask how different 'external' actors have sought to identify the marginal or vulnerable people from whom FPIC should be sought, there seems to be no inherent limit to the definition of this class of people. It could even be extended to include all the local people (or 'landowners') who are liable to experience the negative social or environmental impacts of some kind of 'development' project [43][44][45]. However, when the external actors are foreign investors, or aid agencies, or nongovernmental organisations (NGOs), there are limits to their capacity to make up their own definition and strike a deal with some of the people whom they have identified as suitable subjects. In the act of translation, different internal actors, including national governments, are not only liable to contest the definition of the people who have this peculiar kind of human right, but also to place their own interpretation on the meaning of its different components-'freedom', antecedence (or 'priority'), and 'consent' (or 'consultation') [46][47][48][49][50].
If that sounds like a recipe for confusion, we should not forget the contradiction remaining at the heart of the FPIC principle. However, the human subjects might be defined, and whatever the manner in which their rights are recognised or activated, the point of the whole exercise is to release some area of land or quantum of natural resources from their ownership or control. If we consider the full range of actors that have subscribed to different versions of the principle, it should be clear that these things are generally being turned into commodities or taken to new markets. What this means is that a demonstration or proof of FPIC is one element of their market value, even if that is only because the 'stakeholders' with an interest in their consumption care about human rights or environmental justice [51][52][53][54][55][56].

Methodology
This paper does not present the findings of a single research project that was designed to compare applications of the FPIC principle to land struggles in the three countries under consideration. It is instead based on a comparison of the experiences of the three authors in their own investigation of land struggles in each of these countries. It has been possible to make this comparison because the authors have been members of the same research group for many years and share a conceptual framework derived from political ecology, within their disciplines of human geography and social anthropology. Each of the authors has studied conflicts over the use of customary land in specific rural communities where these conflicts have arisen from specific forms of external intervention in the shape of agricultural development projects, forestry projects or conservation projects. Our selection of these three countries enables us to consider the way that variations in the nature of political regimes in the Asia-Pacific region might serve to explain the differences we observe in applications of the FPIC principle to rural land struggles.
Our own adoption of a case study approach to the questions we have posed reflects the political ecology framework that has guided our research on the larger subject of land struggles (or struggles for control of renewable resources) in rural areas. This in itself would not prevent us from counting things if the existing evidence allowed for such things to be counted. However, the task of quantification is not only constrained by the limitations of the evidence; it is also due to the fact that our case studies are not all case studies of the same kinds of things.
Instead of seeking to base our cross-country comparison on one type of case-for example, cases in which a national government agency has either supported or opposed some application of the FPIC principle by a foreign donor or investor-we have tried to identify different points in the institutional landscape where the FPIC principle has made a difference. Some of these cases would count as 'projects', but others are practices, procedures or institutions. What we have excluded from our analysis are case studies of things like workshops or talking shops where representatives of aid agencies, government departments and (sometimes) civil society groups meet to discuss the incorporation of the FPIC principle into a national or sectoral policy framework. There are numerous examples of this kind of activity, and we have been participant observers in some of them, but we have chosen to focus on cases where the principle has actually made contact with what happens 'on the ground'-hence the title of this paper.

FPIC in Cambodia, Indonesia and Papua New Guinea
The many differences between Cambodia, Indonesia and Papua New Guinea include differences in the way that the FPIC principle has or has not been recognised in national legal and policy frameworks, and also in the way that it has been applied by foreign investors operating with their own standards of market governance or corporate social responsibility. For each country, we briefly outline the relevant national legal and policy frameworks, and then proceed to discuss specific case studies to illustrate particular applications of the principle.

FPIC in Cambodia
Although Cambodia voted in favour of UNDRIP in 2007, no subsequent changes were made to national laws and policies to reflect this commitment, and no existing law comes close to the FPIC principle in terms of granting local consent or veto rights. A climate of intimidation rather than 'free consent' is the norm in Cambodia [57], and this poses a distinctive challenge for any attempt to incorporate the principle into projects that entail close collaboration between foreign investors or aid agencies and Cambodian private sector or government partners.
Although the Cambodian constitution provides for citizens' rights, sectoral laws contain specific provisions for local consultation that generally fall short of FPIC. Article 31 of the National Constitution contains a commitment to the UN Charter, the Universal Declaration of Human Rights, and subsequent UN conventions on human rights, women's and children's rights, while Article 35 recognises the right of citizens to actively participate in the political, economic, social and cultural life of the nation. Two relevant sub-decrees or regulations have been issued under the terms of the Land Law of 2001. The first, on Economic Land Concessions (2005), stipulates three preconditions for the grant of such concessions: a process of environmental and social impact assessment, 'solutions' to any resettlement issues, and some form of public consultation (but not FPIC). The second, on the Registration of Land of Indigenous Communities (2009), provides for the mapping and registration of communal land for indigenous groups seeking to obtain formal recognition and legal registration from the Ministry of the Interior.
Article 15 of the 2003 Law on Forestry recognises 'customary user rights', while Articles 15 and 31 make broad references to a process of 'community consultation' in relation to the allocation of logging concessions and associated road construction. Article 4 of the 2008 Protected Area Law likewise provides for some form of local participation in decisions about sustainable management and biodiversity conservation in protected areas. Finally, the 1996 Law on Environmental Protection and Natural Resource Management has been supplemented by a set of guidelines for the process of environmental impact assessment that include a requirement for some form of public consultation (but not FPIC).
Two case studies serve to illustrate the ways in which the FPIC principle has been mobilised in this national context. The first one involves foreign investment in an agro-industrial rubber project that failed to trigger any of the standard global safeguard policies. The second involves application of the principle in the context of market certification schemes, specifically the voluntary carbon market.

Socfin's Rubber Project
Socfin's rubber plantation in northeast Cambodia has a complex history that has evaded formal FPIC requirements. Cambodian companies initially held a majority stake in the three concessions that comprise the 11,964-hectare estate, but these are now held by Socfin Asia and its subsidiary Plantation Nord-Sumatra Ltd SA. National government approvals for such investments by domestic companies routinely avoid the consultation and environmental or social assessment requirements of the Sub-Decree on Economic Land Concessions, as well as the environmental impact assessment guidelines. The involvement of Socfin, a French company, might have invoked safeguards if it had sourced funds from the IFC or from banks that subscribe to the Equator Principles, but the company's annual reports suggest that it used other sources of finance for this purpose.
In the absence of formal national or international safeguard requirements, Socfin's operations were only subject to 'soft laws' such as the Guidelines for Multinational Enterprises adopted by the Organisation for Economic Cooperation and Development, and to the company's own voluntary corporate social responsibility principles. The latter are described as follows on the company's website: 'Socfin adheres scrupulously to land ownership and environmental legislation in the countries where it operates, as well as to the principles and standards with which it has decided to comply: the performance standards of the World Bank, the RSPO Principles and Criteria for its oil palm plantations, and the criteria of the Sustainable Natural Rubber Initiative for its rubber plantations.' The statement goes on to outline a broad commitment to social and environmental responsibility, particularly in relation to indigenous peoples, while emphasising the voluntary nature of such commitments.
When forest clearance for the rubber plantation commenced in 2008, the indigenous Bunong community faced sudden exclusion from its traditional lands [58] (p. 181). The razing of burial forests in late 2008 intensified opposition from the community, encouraged by civil society groups. This opposition was initially quashed by the government, acting on behalf of the company, but when protest did not dissipate, the company set up a 'tripartite' (company-community-government) committee in 2009 to resolve the conflict. It also appointed a community liaison team to oversee compensation settlements [59]. In 2017, some eight years after the event, many Bunong were still outraged about the bulldozing of their burial sites, sharing photographs of the human remains and artefacts that were uncovered with one of the authors of this paper.
Given the absence of formal safeguards in this case, advocacy organisations used various international soft laws to place pressure on the company, focusing their case on local people's rights to FPIC regarding use of their customary lands and on appropriate compensation for their displacement. European NGOs were able to engage in direct advocacy because the company had a European base. A detailed investigation of the case by the French-based International Federation for Human Rights found that the project had violated the UN Global Compact and the UN Framework and Guiding Principles on Business and Human Rights, as well as the OECD Guidelines [60]. By 2017, a French NGO supported a network of Socfin-affected groups from Africa and Southeast Asia to directly lobby the company at its European headquarters. From an initial emphasis on the flouting of the FPIC principle, their argument later shifted to compensation and grievance mechanisms.

Voluntary Carbon Markets in North-Eastern Cambodia
This case of a voluntary carbon market project in north-eastern Cambodia, the Seima Protected Forest, illustrates the application of a market-based FPIC requirement, similar to those used in the certification of timber (by the FSC) or palm oil (by the RSPO). Here, the performance of FPIC is required to satisfy the Verified Carbon Standard (VCS) that certain social and environmental standards have been met in the creation of a product that consists of fungible units of forest carbon [61]. The VCS requires community consultations under its 'risk management' provisions since the permanence of carbon conservation may be at risk without local consent. An optional set of higher standards for carbon sequestration projects, known as the Climate, Community and Biodiversity Standards (CCBS), provides a greater return per unit of carbon if they are met. These standards require project proponents to recognise the knowledge and rights of indigenous and local communities and their 'full and effective' participation, following FPIC principles [62].
As with the preceding case, community engagement is closely tied to formal or informal land claims. Here, project implementers planned to apply the FPIC principle to simultaneously satisfy both the VCS and CCBS standards. Nevertheless, this case shows how the standards could be met without deep local engagement. FPIC activities were rolled out across multiple villages on a large scale by project facilitators, with significant attention to VCS requirements for documentation and evidence of compliance. These certification requirements have underpinned the project's emphasis on the performance of FPIC, such as gathering villagers' thumbprints to demonstrate their consent to the scheme, and a signing ceremony where village leaders signed a legalistic agreement that few understood. Although these activities have enabled the project to market carbon credits to a multinational corporation, local engagement and empowerment, along with the higher order actions needed to counter the political and economic drivers of forest loss, have been absent [61]. Weaknesses in the process of consultation and local engagement have also been observed in forest carbon schemes in other parts of Cambodia [63].
Market-based FPIC standards are in some respects akin to the safeguards of donor agencies and the IFC, but with one important difference. The failure to meet market-based standards can undermine product 'value' and market share. In Cambodia, a high-profile carbon offset scheme involving Virgin Airlines fell apart in this way, when evidence of social and environmental dissonance became public, causing Virgin to withdraw. Unlike the public disclosures and complaints directed at Socfin, market-based FPIC requirements appear instead to nurture a culture of silence [61,63].

FPIC in Indonesia
There are a number of ways in which the FPIC principle has been applied in Indonesia, but the main focus here will be on forests and oil palm, and especially on the Indonesian part of the island of Borneo (Kalimantan). This region was historically rich in forests, but policies adopted during Suharto's New Order period (1967-1998) saw large areas leased out as logging concessions or cleared to create new settlements for 'transmigrants' from other parts of Indonesia. The prioritisation of the 'national interest' over the needs of local populations demonstrated 'an almost categorical lack of appreciation for local and traditional resource management' [64]. The 'forest zone', as defined by the powerful Ministry of Forestry, occupied 73% of Indonesia's total land area, and 82 % of the four provinces of Kalimantan [65] (p. 381). 'Forest people' and local communities were vulnerable to the impacts of logging, with insecure tenure and struggles over land rights, and this legacy constitutes the problem to which the FPIC principle has since been applied. The logged-over forests were partially replaced by plantations of fast-growing trees, and later by oil palm, or were simply burned in large fires, especially in areas with peat soils. At the same time, failed transmigrant food-crop schemes were being converted to rubber or oil palm smallholdings. The collapse of the Suharto regime in 1998, closely followed by a process of decentralisation in 2000, devolved much of the responsibility for land and natural resource management to district-level authorities within each province. These district authorities, especially in the provinces of Central and West Kalimantan, have since granted large numbers of leases for oil palm and timber concessions. Disputes with local people have proliferated.
Indonesia's National Constitution treats all natural resources, renewable or otherwise, as the property of the state unless they have been legally assigned to private owners. It only recognises customary rights 'so long as they still exist'. The Basic Agrarian Law of 1960 treats customary rights as rights to the use of state land that could be taken away, without consent or compensation, whenever the land was required for some form of large-scale development. The Basic Forestry Law of 1967 specifically denied the right of 'customary communities' to block the establishment of transmigration sites. A critical report on the government's large-scale transmigration scheme argued that the World Bank should withdraw its financial support, citing both high levels of forest destruction and dispossession of local communities [66][67][68]. The Bank did not accept this argument, even though 'meaningful consultation' was mandatory for Bank projects by the late 1980s, and by 1992 'meaningful stakeholder participation' was being understood to include the right to reject an intervention [12].
While the Suharto Government refused to acknowledge the existence of a separate category of 'indigenous' people within the Indonesian (non-Chinese) population, it did recognise the existence of 'isolated tribes' who were remote from 'civilisation'. In some cases, this was taken as a pretext for their forcible resettlement. For example, in the province of East Kalimantan, about 10,000 Dayak families were resettled during the 1970s in an exercise financed by taxes on the logging companies that were granted concessions over the land the Dayaks formerly occupied [67,69].
Amendments were made to the Basic Forestry Law in 1999, following the collapse of the Suharto regime. Article 67 of the amended law made provision for 'customary communities' to manage 'customary forests' once their existence had been officially recognised, but still defined customary forests as 'forests with no rights attached'. A study carried out in Indonesia for the FSC in 2002 concluded that the weak legal recognition of customary and indigenous rights, combined with state repression and the marginalisation of forest-dwelling communities, made exercise of the right to FPIC 'near impossible' [70].
The Indonesian Government has continued to deny the existence of a separate category of 'indigenous' people, which may explain why it was quite happy to ratify the UN Declaration. A body called The Alliance of Indigenous People of the Archipelago (AMAN), which was established in 1999, made a successful application to the Constitutional Court for a judicial review of the Basic Forestry Law, and in 2012 was rewarded with a ruling that 'customary forests' do not belong to the state, even though they are still part of the national forest estate [71]. However, the power to recognise the continued existence of their customary owners has been devolved to district governments, which have generally been unwilling to exercise it [72]. Furthermore, the ruling has done nothing to enhance the rights of people living outside the official forest zone, those whose lands are already designated as conservation forests or concession forests, transmigrant populations, or people who already have some management rights through social forestry schemes [73].
The case studies chosen for analysis here relate to applications of the FPIC principle by investors in the oil palm industry, especially those who are members of the RSPO.

Oil Palm's Double Standards
The RSPO was established in 2004 by the Worldwide Fund for Nature (WWF) and Unilever, 'an alliance between environmentalists and the food industry' [74]. Its initial set of 'Principles and Criteria', produced in 2005, were considerably influenced by Indonesian NGOs such as Sawit Watch, which served on the RSPO Executive Board and maintained an 'insider/outsider' role, working together with the international Forest Peoples Programme (FPP) in an attempt to ensure the inclusion of indigenous rights and control over land [74][75][76]. However, farmers' comments in two surveys undertaken in 2005-2006 revealed a total lack of transparency on the part of companies in their dealings with local people: 'falsified promises, infringed agreements and fraudulent application of the relevant regulations' [75,77].
The FPIC principle was explicitly mentioned in three of the RSPO's original criteria, where it applied to both the acquisition of land and to compensation for the loss of existing rights. This made it seem like a technique designed to reduce the loss of land to oil palm and provide some measure of compensation for the conversion. The revised Principles and Criteria of 2013 supplied more details of the rules under which companies should operate: a proper investigation had to be undertaken of the legal and customary rights of landholders in the area of the planned development, prior to any clearance and generally with participatory mapping, while local people should also be allowed to reject the company's proposal [78].
Supporters of the principles have acknowledged the cost of FPIC procedures in both time and money [79] (p. 17) but claim that the social engagement tools have been of great value in reducing social conflicts and enhancing the operational environment. A subsequent study of oil palm estates across Asia and Africa, which aimed to assess the application of the principle and to strengthen procedures where necessary, included several examples drawn from Kalimantan [80]. Examples of 'best practice' were meant to be emulated, but case studies of plantations owned by leading corporations such as the Wilmar group and Golden Agri-Resources (GAR), both based in Singapore, revealed continuing problems for local communities.
Questions remain as to whether FPIC principles and procedures really reduce the incidence of conflict, let alone deal with underlying problems such as poverty or inequality within local communities, given the constraints imposed by state laws and policies [81]. McCarthy [82] (p. 1879) notes that the RSPO has little power or influence at the district level in Indonesia, where most negotiations over land take place, and there are few incentives to make sure that the rules are followed. A study in Central Kalimantan has found that 'negotiation processes triggered by the RSPO … fall short of local actors' expectations and claims; thus leading to numerous cases of deadlock' [83] (p. 69). When compensation was limited to small amounts of cash or the opportunity to participate in a smallholder scheme linked to a big plantation, such 'rewards' did not satisfy villagers with deep emotional attachments to grave sites or particular landmarks. Critical flaws have also been discovered in the reliance of the RSPO on auditors whose own assessments have sometimes been 'woefully substandard', who have sometimes colluded with plantation companies to disguise violations of the rules, or who have been unable to do their job properly because of their financial dependence on their clients [74,84].
The RSPO became unpopular with Indonesian palm oil companies when it required FPIC to be obtained from communities before new plantings could begin, instead of a few years later. At the eighth annual conference of the RSPO in 2010, Indonesia's minister for agriculture announced that the government was about to launch a mandatory national standard for Indonesian Sustainable Palm Oil (ISPO). While there are some similarities between the two sets of standards, the ISPO was to be based on existing Indonesian laws, especially the regulation governing environmental impact assessment, and was meant to be simpler. The RSPO's 'commitment to transparency' and 'responsible development of new plantings' were missing from the ISPO rules, as was the FPIC principle and mention of 'high conservation value' (HCV) forests [85]. The latter term, borrowed from the FSC, refers to ecological and social landscape attributes-not only biodiverse forests containing endangered species, but also special set-asides for the provision of basic community needs and the maintenance of cultural identity [86]. Six months after the official launch of the ISPO in 2011, the Association of Indonesian Palm Oil Plantation Companies (GAPKI) resigned from the RSPO and joined the new scheme instead.

Grievance Mechanisms at Work
Between 2007 and 2011, a consortium of local, national and international NGOs, including Sawit Watch and the FPP, made a series of complaints to the IFC's Compliance Adviser/Ombudsman to the effect that the IFC had failed to comply with its own performance standards when providing a series of loans and loan guarantees to the Wilmar group of companies between 2004 and 2007 [87,88]. Although these investments were not specifically linked to the development of oil palm estates in Indonesia, it was argued that the Wilmar subsidiaries that were involved with such schemes in West Kalimantan and West Sumatra were still part of a single conglomerate, and the IFC was therefore bound to ensure that they also complied with the performance standards. The complainants argued that the World Bank Group as a whole should apply the stronger version of the FPIC principle to the acquisition of land for oil palm estates because the IFC was already a member of the RSPO, and that body had already adopted the stronger version in which 'consent' had been substituted for 'consultation' [89] (pp. 4-7).
The Compliance Adviser found enough merit in these complaints to initiate the usual investigations and attempt some form of mediation between the operators of the oil palm estates and the aggrieved members of what were generally agreed to be 'indigenous' communities [87,90]. In 2009, the World Bank's president announced a moratorium on investment in the oil palm sector pending the completion of a consultation process and the preparation of a new investment strategy that would take more account of the issues raised by the complainants. It appears that this process was one of the factors that led the IFC to adopt the stronger version of the FPIC principle in 2012.
The RSPO established its own Complaints Panel in 2010. In its first five years of operation, this body dealt with 37 complaints from Indonesia, of which 41 % were based on violations of the FPIC principle [76] (pp. 159-60). One of these was a complaint lodged by the FPP against GAR and its Indonesian subsidiaries in 2014, which related to 18 different concessions in Kalimantan with a combined area of 300,000 hectares. This complaint was upheld by the panel, which ordered the company to compensate the affected communities as a condition of its continued membership of the RSPO [76] (p. 160). The company's response was to set up its own grievance mechanism and ask another NGO, the Rainforest Alliance, to evaluate its implementation [91] (pp. 22-25).

A New Kind of Forest
The most recent application of the FPIC principle in Indonesia has been associated with the recognition of 'high carbon stock' (HCS) forests. The HCS idea was initiated by GAR and its subsidiary, Sinar Mas Agro Resources and Technology (SMART), as part of a new attempt to produce 'sustainable palm oil'. RSPO member companies such as Nestlé had stopped buying palm oil from GAR but resumed when GAR began to implement a new 'forest conservation policy' in association with an NGO called The Forest Trust. This policy was described as: 'Not developing oil palm plantations on areas that have High Conservation Value (HCV) and areas of peat regardless of depth and not developing forest areas with High Carbon Stock (HCS); obtaining free, prior and informed consent from indigenous and local communities; and complying with all relevant laws and internationally accepted principles and criteria' [92].
An independent review of the policy's implementation in the remote Kapuas Hulu District of West Kalimantan was highly critical of its potential local impacts, especially a lack of community understanding of the way that HCS forests had been defined. In referring to local swidden (or shifting cultivation) practices, the reviewers observed that 'the imposition of (forest) categories based on their current carbon content breaks up a dynamic system of land use and regrowth' [93] (p. 43). The authors recommended detailed community mapping and studies of local farming systems before any attempt was made to set this type of forest aside [94]. However, the idea of HCS forests, with the FPIC principle attached to them, was recognised as being useful in fragmented landscapes with moist tropical forests, and as being applicable to rubber, pulp and paper, and other commodities [95][96]. It was also seen as a potential tool to help palm oil companies implement their 'no deforestation' commitments for the European market.
The Indonesian Government had previously rejected a move by a group of corporations such as Wilmar, together with some NGOs, calling themselves the Indonesian Palm Oil Pledge (IPOP), to unilaterally proclaim a policy of 'no deforestation, no peat, no exploitation', in which the FPIC principle was included [97]. Ministry of Agriculture officials saw such a pledge as an affront to Indonesia's sovereignty, while the 'foreign' companies were accused of operating as a cartel. IPOP was disbanded when it was threatened with court action, and this has led some commentators to conclude that 'the high carbon stock approach remains out of reach in current political circumstances' [98].
Nevertheless, the RSPO has since claimed the high ground, both socially and environmentally, by requiring protection of both HCV and HCS forests in the latest version of its Principles and Criteria [99]. Critics of the RSPO standards have argued that the national ISPO standards, now applied to roughly 30 % of the country's large-scale oil palm schemes [100], do more to protect the interests of smallholders by means of a number of specific regulations, even though the willingness and ability of smallholders to comply with these regulations is clearly problematic [101,102]. For its part, the RSPO has responded by creating a 'simplified FPIC approach' and a 'simplified combined HCV-HCS approach' that is meant to enable independent smallholders to progress through a number of 'milestones' on the way to meeting standards specific to their own form of economic activity [103]. This appears to have been a response to the European Commission's decision to phase out the use of palm oil as a source of biofuel if its production causes deforestation, because smallholders who produce palm oil 'in a sustainable way' are specifically exempted from this ban [104].

FPIC in Papua New Guinea
The authors of the World Bank's latest set of safeguard policies were probably not thinking about the Pacific Island region when they made up a list of alternative designations for 'indigenous peoples'; otherwise they would have added 'customary landowners'. Throughout this region, customary land rights are accorded a degree of legal recognition that is rarely found in any Asian country, even when they are not formally documented, and the native inhabitants of Pacific Island countries, who normally account for a majority of the population, tend to regard their possession of such rights as a key component of their national identity [105][106][107]. Papua New Guinea (PNG) is one of several Pacific Island countries whose governments have not bothered to ratify UNDRIP because they seem to think that it affords no legal rights that their indigenous citizens do not already possess. It is generally assumed that 97 % of PNG's total land area is still subject to customary tenure, and the government is even prepared to countenance the existence of customary rights to secure some benefit from the development of the remaining 3 % that was alienated during the colonial period [108].
The National Constitution of PNG makes one reference to the 'indigenous inhabitants of the country', but only for the purpose of defining the term 'custom', which is otherwise deemed to be the property of 'traditional villages and communities'. While other laws occasionally repeat this constitutional reference to 'indigenous inhabitants', the term 'customary landowner' or 'customary owner' is far more common. For example, Section 46 of the 1991 Forestry Act says that the 'rights of the customary owners of a forest resource shall be fully recognized and respected in all transactions affecting the resource', while Section 10 of the 1996 Land Act says that the minister of lands 'shall not acquire customary land unless he is satisfied, after reasonable inquiry, that the land is not required or likely to be required by the customary landowners or by persons on whom the land will or may devolve by custom'. Recent amendments to the Land Groups Incorporation Act and Land Registration Act were justified as legal measures 'to empower customary landowners in Papua New Guinea to realize the now locked up economic potential which their customary land has' [109] (p. x).
In the years immediately preceding Independence in 1975, PNG's indigenous political leaders took offence at a belated attempt by the Australian colonial administration to establish a new legal mechanism for the systematic registration of customary land rights because they thought that this might undermine the basic principles of customary tenure [110]. The preferred alternative was a new body of legislation constructed on the assumption that customary rights could only be vested in customary groups, that these groups should be free to decide whether or not to 'incorporate' themselves as legal entities and register titles to their land, and should then be free to lease-but not to sell-portions of their land to outsiders [111,112].
In its original form, the Land Groups Incorporation Act of 1974 did not require the applicants for incorporation to do anything more than fill out a form to show that they represented a genuine 'customary landowning group' with a set of rules to determine its membership and govern its operation. They were not required to provide evidence that the group owned any particular portions of land, let alone that the members had agreed to dispose of their land in any particular way, so PNG government officials dealing with the application were not required to assess the question of consent [113]. This might well have been required if a group, once incorporated, had applied to register its land rights, but this additional step could not be taken because government officials were unable to produce the additional legislation. The World Bank and the Australian Government provided some 'technical support' for efforts to overcome this obstacle, but this only provoked a political campaign against the perceived infringement of customary rights [114,115]. National policy makers then took additional measures-and a good deal more time-to remove all signs of foreign interference before they were eventually able to remove the legal obstacle [116][117][118].
In the meantime-for a period of more than 30 years-the legal black hole was partially filled by the deployment of new provisions in the Land Act inherited from the colonial administration that made it possible for groups of customary landowners to lease land to the state on condition that it was then leased back to these very same groups or else to what are known in PNG as 'landowner companies'. The so-called 'lease-leaseback scheme' was thus designed to create the legal titles that could not be obtained directly by incorporated land groups, and to enable customary landowners to either use their land as security for bank loans or issue legally valid subleases to tenants or investors.
The case studies chosen for analysis here include a recent contest over the lease-leaseback scheme and a longer process of forest policy reform in which the World Bank was heavily involved between 1989 and 2002. There is a separate discussion of the peculiar concept of 'landowner awareness' in PNG.

Abuse of the Lease-Leaseback Scheme
In the first 20 years of its operation, the number of leases issued under this scheme was quite small, as were the areas of land that each of them covered, and it is not clear what measures were adopted to ensure the consent of the customary owners [119]. In 1998, PNG's palm oil industry began to make use of this scheme to establish 'mini-estates' around the nucleus estates that had already been established on land alienated during the colonial period. For this purpose, they engaged former (white) government officials with a thorough knowledge of PNG's land legislation to conduct land investigations, facilitate the incorporation of land groups, arrange for leases to be vested in the hands of these same groups, and finally to negotiate the terms of the subleases by which the companies would gain access to their land [120].
This was a time-consuming and expensive business, but it did succeed in partially alienating about 30,000 hectares of customary land over the course of the following decade [121], which was more than the total area of land converted to 'special agricultural and business leases' of this kind in the two previous decades. This endeavour turned out to be significant for three reasons. First, it provided evidence that PNG's palm oil industry was already committed to FPIC when it joined the RSPO. Second, it may have encouraged other developers to take advantage of the lease-leaseback scheme, since more than 5 million hectares of customary land-almost 12 % of PNG's total land area-was alienated in this way between 2003 and 2011. Yet the industry's consultants also knew from their own experience that a lease covering tens of thousands of hectares could not possibly have been created with the same due care and diligence that they had taken in creating the mini-estates. So it is not surprising that they played a key role in persuading the national government to institute a judicial inquiry into what had begun to look like a massive land grab [122]. The terms of reference for this inquiry included an explicit requirement for the lawyers involved to find out whether each of the leases under investigation had been granted with 'prior consent and approval by customary landowners'. The evidence revealed that most of them had not [123]. The lease-leaseback scheme was suspended, and several of the larger leases have since been revoked.
The evidence also revealed that most of the leases over areas greater than 10,000 hectares had been granted to landowner companies for the development of so-called 'agro-forestry projects', in which the 'development partner' would harvest a large area of native forest and use the revenue from log exports to fund the creation of a large-scale cash crop scheme [124]. This was a source of additional concern for the existing palm oil industry because the crop of choice was oil palm, and the newcomers showed no interest in the RSPO standards-if indeed they had a genuine interest in agricultural development at all, rather than an interest in escaping the constraints of PNG's own standards of 'sustainable forest management'. There was something deeply ironic-if not downright paradoxical-about this discovery, because the start of the new 'land grab' had coincided with the PNG Government's removal of the World Bank from a process of forest policy reform that the Bank had been invited to coordinate back in 1989, and which had led to the creation of the management standards that were now being subverted [125].

'Sustainable Forest Management'
The Bank's engagement in this policy process was the result of an earlier judicial inquiry whose focus had been the incidence of corruption in the log export industry. One of the discoveries of that inquiry was that landowner company directors could not be trusted to represent the interests of other customary landowners when they entered into 'private dealings' with logging companies [126]. The possibility of such dealings had been created by the Forestry (Private Dealings) Act of 1971, which, like the Land Groups Incorporation Act, had been intended to free customary landowners (or their representatives) from the constraints of colonial paternalism. One of the results of the inquiry was the repeal of this legislation and the creation of a new legal regime in which customary landowners could only alienate their timber harvesting rights to the state, as had been the case in the colonial period, unless they could demonstrate some capacity to manage a small-scale logging operation on their own account [127,128].
Here was another paradox. In order to provide evidence of genuine landowner consent to the 'forest management agreements' by which the state was now to acquire the timber harvesting rights, national policy makers decided to adopt the mechanism of land group incorporation. This choice was partly motivated by a second decision to enlarge the size of logging concessions so that each one could be selectively logged over a period of several decades without reducing the prospect of a 'second cut' at some point in the future. However, it would clearly not be possible for government officers to conduct a detailed investigation of customary land and resource rights in such large areas, so it was thought that the best way to mitigate the risk of misrepresentation would be for government officers to assume responsibility for supervising the act of incorporation, and to make sure that landowner company directors were kept out of it [129]. Between 1994 and 1998, officers of the National Forest Service facilitated the incorporation of more than 5000 land groups, and a consultant funded by the World Bank produced a 100-page manual that told them how it should be done [130].
Under this policy regime, it does not matter that most (if not all) of the foreign logging companies investing in PNG have shown no interest in any application of the FPIC principle, since they are legally separated from the customary owners of the forest by a government agency that transmits the timber harvesting rights from one side to the other by means of two different kinds of agreement. The companies are only obliged to abide by the regulations that tell them how to conduct a 'sustainable' harvest.
The 'development guidelines' attached to the Forestry Act said that a 'landowner awareness program' should be undertaken in advance of a forest management agreement in order to enable landowners to make their own assessment of 'the likely costs and benefits, impacts and responsibilities associated with a forest development project' and enable them to 'truly participate in the project formulation process and ensure that it is sensitive to their needs and concerns' [131] (p. 4). The larger package of forest policy reforms also included plans for a 'landowner awareness and support project' whose design and implementation became the responsibility of the German Government's aid agency [127] (pp. 271-2). On both counts it looked as if the 'information' component of the FPIC principle was getting its fair share of attention, but since the whole package was directed to the creation of more protected areas as well as to the regulation of the logging industry, there was no chance that the stakeholders in the policy process would agree about what the landowners should be made aware of. There is very little information, let alone a manual, on the awareness programs actually conducted by officers of the National Forest Service in the process of creating a forest management agreement [132] (pp. [11][12]. Indeed, the only manual to emerge from the whole reform process was one that told NGOs how to practice a version of participatory rural appraisal when persuading landowners not to sign one [133].

A Note on the Concept of Landowner Awareness
This last manual was one of the products of a biodiversity conservation program, funded by the Global Environment Facility, whose only lasting legacy has been the establishment of an NGO whose main mission is to make customary landowners aware of the evils of capitalism in general [134]. While the World Bank played a key role in securing the funds to support this program, the NGO that was its final product was one of those that sought to make landowners aware of the Bank's intention to help the government steal their land, at a time when the Bank's staff and consultants were doing their level best to prevent the government from unwinding the web of regulation that had been spun around the logging industry [125] (pp. 42-45).
As we have seen, this is a battle that the Bank eventually lost, but what has survived is the idea that 'awareness' is a peculiar sort of thing to which customary landowners have a right, and which outsiders therefore have an obligation to provide [135,136]. This is commonly represented in the nation's political discourse as a kind of performance-perhaps even a kind of public ceremony-that is demanded or delivered as part of any development proposal or policy process that might have some effect on customary rights to land or natural resources. It has also been conceived as something that university students should do with (or to) their less educated relatives during their long vacations, just as they did when mobilising public opinion against the World Bank's purported land grab in 1995. It is not confined to the routines of participatory rural appraisal; it is not the special preserve of civil society; nor is it connected to specific applications of the FPIC principle. When government lands officers were asked to explain their failure to conduct land investigations before granting special agricultural and business leases over huge areas of customary land, one of their excuses was that 'an awareness had been done'.

Discussion and Conclusion
In all three countries, there are cases in which big companies, foreign aid agencies, or international financial institutions have been embarrassed by accusations of failure to abide by their own public commitment to the FPIC principle. These accusations are often voiced by an alliance between national and international NGOs, sometimes known as a 'transnational advocacy network' [82,137], speaking on behalf of local communities affected by some particular project or investment. When they focus on transgression of the FPIC principle, rather than some other human right, such accusations gain weight from an assertion that the victims are 'indigenous people', albeit with some latitude in the construction of this definition. If the perpetrators are sufficiently embarrassed by the negative publicity, they are liable to quit the project or investment altogether, leaving a gap that may be filled by actors who are not so easily embarrassed. Otherwise, they may defend themselves through the deployment of a 'grievance mechanism' that airs the absence of consent. Either way, the dialectic of market governance or corporate responsibility is one where consent appears as one element in the value of the land or natural resource that is being appropriated, with variable quantities of risk attached to its acquisition, while the process of acquisition itself commonly takes on the character of a 'performance' in which local or indigenous actors take part [138][139][140].
At first sight, this logic seems to operate quite independently of national political regimes, but there are two reasons to question the idea that the nature of the regime makes no difference to what happens in practice. First, the audience for any local or national performance of FPIC is not simply a global one, but is unevenly distributed between different countries, and is not completely blind to the local or national context in which the performance is conducted. This point is exemplified by the contrast between the politics of the palm oil industries in PNG and Indonesia. The original investors in PNG's palm oil industry were of European origin, and since the bulk of the output has traditionally been sold into the European market, the producers have been especially keen to demonstrate their compliance with RSPO standards. This explains why they have put pressure on the national government to limit the entry of new producers who do not share this interest because their eyes are set on less discriminating Asian markets. First World markets have only recently become significant destinations for Indonesian palm oil, which explains why the producers were initially less concerned with the validity of their 'performances' within the paradigm of market governance, yet growing pressure from European governments and NGOs operating within the paradigm of corporate responsibility has brought them to a point at which application of the FPIC principle has become a bone of contention with a government wary of international norms that might infringe its national sovereignty.
This takes us to the second and more significant reason to bring the state back into consideration. Aside from the common force of legal and bureaucratic inertia, governments may resist the application of international norms for different historical reasons, and the history in question may open up more or less space for indigenous (or local) communities to halt or reverse the removal of their customary rights within the current legal framework of the nation-state. In Cambodia, the principle is akin to what James Scott has described as a 'weapon of the weak' [141], since local communities (indigenous or otherwise) are confronted with state institutions that offer no meaningful avenue for consultation or compensation in the grant of economic land concessions, despite the letter of the law. In PNG, by contrast, the power of the 'customary landowner' is not only embedded in a raft of national laws and policies, but also exercised in a variety of legal and illegal activities that constantly threaten the viability of externally funded investments in the realm of customary land [108]. In this case, the national government's resistance to the FPIC principle is itself a sign of weakness, since it reflects the inability of state actors to manage such threats with the means already at their disposal.
Even so, we do not suggest that international norms pertaining to human rights are simply liable to cascade down to national and local struggles in accordance with the capacity of different groups of actors to exploit, control or contest their application at a smaller scale. Instead, we need to make allowance for the fact that these international norms have also emerged as mediations of double movements already taking place in specific national and local contexts. When we set out to compare the application of the FPIC principle in different countries, we should therefore ask whether specific struggles in those countries have contributed to the formation of a specific version of this principle, in the same way that struggles over the construction of India's Sardar Sarovar dam in the Narmada Valley contributed to the formation of the World Bank's safeguard policies in the 1990s [10,142]. The campaign against the IFC's association with the Wilmar group, including its operations in Indonesia, appears to be a case in point.
We do not claim that the case studies explored in this paper are entirely representative of the land struggles that have taken place in each of three very different countries. Instead, we have tried to show the extent of variation between cases where the double movement is barely present at all, because the FPIC principle is merely used to enhance the value of a new commodity, like forest carbon, and cases in which the principle is one of several weapons applied to the type of land struggle that reveals the deeper contradictions of capitalist underdevelopment. In this respect, we have explored the spectrum that exists between what Goodwin [9] calls the hard and soft versions of the double movement itself-the one constituting a threat to the viability of various forms of capital investment, while the other merely creates the sort of problem that can be resolved, or even concealed, by the practice of 'mitigation' or 'remediation'. The location of any given contest within this spectrum is a function of the balance of power between the local, national and international actors engaged in it, and also the balance of power between the state, capital, 'civil society' and local or indigenous communities under different political regimes. However, the nature of the regime itself is not the sole determinant of this balance of power, and that is why the double movement is a somewhat unpredictable affair.