From Confrontation to Mediation: Cambodian Farmers Expelled by a Vietnamese Company

Abstract

Concessions granted to investors in Cambodia have generated a deep sense of insecurity in rural forested areas. Villagers are not confined to a passive “everyday resistance of the poor,” as mentioned by James Scott, insofar as they frequently engage in frontal strategies for recovering land. Such has been the case in the northeastern provinces, where indigenous livelihoods are recurrently threatened by foreign and national companies. But what happens when a land conflict ends up in a stakeholder dialogue? The article intends to follow such a story that occurred for the first time in Ratanakiri, in a vast territory inhabited by several ethnic groups. After gruelling hostilities with the encroacher, dispossessed farmers finally accepted, encouraged by international/national NGOs, to comply with existing mechanisms associated with international law regulations and World Bank procedures. It ends up in an institutionalised mediation, technical and apolitical, which turned to the disadvantage of the people, with evident power imbalance. Our analysis, while portraying the trajectories of national/international actors involved in the mediation process, reveals the effects on this mediation on local sociopolitical organisations.

Over the last two decades, Cambodia’s northeast has become an El Dorado for investors and agro-industrial companies, but a challenging situation for some indigenous people already living there. This coincides with economic liberalisation, compounded by a rising global demand for rice, cassava, sugarcane, corn, and mango from the central plains and latex, cashew tree, and pepper from the eastern provinces. The demand for rubber from Vietnam has spurred agro-industrial projects in Ratanakiri Province (Oxfam, 2016), mounted by foreign investors linked to international companies and domestic elites standing to gain from lucrative procurement and construction contracts, won thanks to approachable government officials. ¹ In the Mekong subregion, where governments depend mainly on Foreign Direct Investment (FDI) for development funding (Socheth, 2012), the granting of Economic Land Concessions (ELCs) ² is instrumental in attracting FDI in the land/agricultural sector where external investments remain insufficient compared with other sectors. Most financing for the agriculture sector comes from companies’ equity funds and regional lenders from Vietnam, Thailand, and China. The World Bank (WB) through its private sector agency, the International Finance Corporation (IFC), and other worldwide conglomerates also provide loans to the companies.

These financial institutions attracted scrutiny when independent analyses demonstrated non-compliance with international laws on investment, human rights, and state sovereignty (Fox and Brown, 1998; Szablowski, 2007) and collusion between mega-development projects and their financial backers (Barros et al., 2003; Human Rights Watch, 2015). These revelations tarnished their reputation. Predictably, both Asian Development Bank and WB adopted safeguard policies and provided advice and money for tenure guidelines governing international land deals in their development model (Krever, 2011). Worried about their credibility, international agencies declare the free, prior, and informed consent of indigenous populations as a “fundamental requirement” ³ to be obtained on projects that may cause their displacement or affect their land, territories, resources, or cultural heritage (World Bank Group, 2010).

The World Bank Group now proposes a set of ethical, environment, and social considerations to be respected in relation to people’s well-being, failing which the investor could be refused the loan. IFC can withdraw the loan based on its own supervision without any compliance audit. If already granted, complainants can request a compliance audit. The loan can also be retracted in case of non-compliance, provided negotiations failed, followed by an audit in favour of the affected populations. Such new regulations theoretically open the world to new mechanisms capable of challenging the powerful companies by directly lobbying not only the state – often proven inadequate (weak judiciary system) and oriented (co-opted) – but international financial institutions. Communities affected by development projects can complain directly to multi-national donor agencies through established institutionalised mechanisms.

This article explores such a story involving the Vietnamese private company Hoang Anh Gia Lai (HAGL). Present in Ratanakiri from 2013 onwards, the much-lauded initiative of mediation metastasised into a deceptive, “wait-and-see,” technically biased mediation, and gradually into a nightmare for most villagers who, at the time this article was written in December 2018, were left misled, confused, and disorganised by the majority of external interventions supporting their cause.

Theoretical, Methodological, and Ethical Considerations

This article investigates a land conflict followed by a mediation process and the effects of this mediation on local sociopolitical organisations in an area known to the researcher since 1994 (Bourdier, 2006). An engaged theoretical approach to political ecology insisting on power imbalance is pursued (Batterbury and Horowitz, 2011). In addition to questions of social justice and ecological imbalances, actors – government, local authorities, villages, NGOs, company, funding agencies – are examined. Divergent positions were adopted when dealing with land conflict and conflict resolution which turned out to be technical and apolitical. A long-term anthropological approach permits an overview prior to and during a mediation process that occurred after a contested ELC was granted by the national government to three Vietnamese enterprises, all backed by HAGL, which, in its hegemonic model of agro-industrial expansion, was accused by villagers of not having respected the WB safeguard policies.

Fieldwork Orientations

The methodology is based on multi-sited ethnography (Marcus, 1995). It allows to follow the webs of power connecting indigenous communities, local and international NGOs supporting them, the government, and other transnational actors such as the World Bank Group. Field enquiries in Ratanakiri focusing on the early social mobilisation started in 2010, when the movement developed into an open land conflict. Interviews were conducted in eight of the fourteen affected villages inhabited by four dominant ethnic groups (Jarai, Kachoh’, Lao, and Tampuan). Other villages (Kreung and a remaining Kachoh’) were only visited three or four times. Fieldwork continued after 2014 when the mediation started. They lasted from one to two weeks. They relied on direct observations, cross interviews, and informal dialogues. Spontaneous focused discussions relied on key individuals and small groups. Some conversations were completed by phone. Most enquiries concentrated on traditional and administrative village headmen, middle-aged men and women eager to raise their concerns, affected households, and elected village representatives. The final surveys occurred in late 2018. While Ratanakiri and Phnom Penh government officers rarely objected to giving their point of view, the two international/national NGOs were reluctant to exchange information, particularly when the mediation started. Yet, the three indigenous associations gave their time. The appointed mediator and her national counterpart welcomed debate. Two Washington-based IFC members and a consultant from the mediation office replied to emails.

In order to discover the fissures in dominant categories of knowledge production, a special attention was given to deconstructing the financial agencies. The heavy weapons of deconstruction that are aimed at the powerful may indicate an oriented research activity in favour of the marginalised affected communities. Following Hale (2006), I respond by arguing that a similar methodological orientation has been extended to the other actors (NGOs, provincial and national authorities, donors, experts, etc.) whose narratives, whenever accessible, have been integrated in the research. Moreover, such a position in favour of poor peasants, while maintaining scientific rigor, recuses “neutrality.”

Social Engagement as a Preliminary Ethical Decision

The ethical considerations chosen to be highlighted are not the conventional ones (informed consent, anonymity, etc.), which, of course, deserve respect. More important is the scholarly engagement and the social impact of the research. Anthropology has already established a foothold in the public sphere – one that harnesses novel forms of communication to reach beyond the traditional “ivory tower” and offers anthropological perspectives on the issues of our time in accessible and innovative ways (Checker, 2009). Since 2015, members from affected villages have requested alternative visions dealing with the unsolved fight against HAGL. Others wished that authoritative claims be made on their behalf and that their story being told “everywhere” [sic]: in official circles, public places, and the media.

They expected help in the service of their struggles. Some wanted to know about original strategies since simply criticising the supremacy of an organised system controlled by leading worldwide agencies mastering the art of the discourse could mislead them by engendering false or unrealistic expectations. Such a request is not incompatible with scientific work. How can we continue to strictly subscribe to conventional research in an unconventional world?

Marcus has already signalled that sociopolitical engagements are becoming very common (Marcus, 1995). Due to the uncertain and inadequate worldwide politics that deserve to be scrutinised and criticised, the anthropologist can adopt an active position. The previously encouraged “ethnographic distance” is outdated, insofar as it separates the academic dimension, entrenched in an “ivory tower,” from the challenges contained in the research. In a perilous situation for the people with whom they work, anthropologists and other scientists are morally obliged to abdicate from an illusory neutral stance, unless they surrender and comply with the sources of power.

Such an alignment with marginalised groups indicates a shift of scientific interest towards an appreciation of the complexity of subaltern policies and politics related to access to, and control over, land. Nothing can prevent the scientist from proposing to the concerned actors (NGOs, populations, civil servants) political choices and strategies that they can consider, evaluate, and, perhaps, accept. For the villagers, understanding socio-economic and legal mechanisms that they can retroactively use, and a depiction of the form of abuses and exploitation they have been victims of, constitutes an important contribution beyond the tedious concept of “providing technical services.” Land is a villager’s life: an engaged researcher cannot view their dispossession as a mere analytical exercise.

Once socially engaged researchers reject the illusory “neutral” position and affirm the intersubjective character of social science research (Hagberg and Ouatarra, 2012), one might expect “the floodgates opening to a growing interest in activist research methodologies and their endorsement as legitimate scholarly activities within the discipline” (Hale, 2006). For those who affirm the idea that knowledge is produced through dialogues among politically situated actors, incorporating this process in research is a logical step that can be done by sharing a political alignment with the subjects of the study, as advocated elsewhere (Borras, 2016: 23–40). However, blind social engagement can be counterproductive (Low and Merry, 2010) and alienate local people (Edelman, 2009). A study revealed how corporate mining industries incorporated external critical discourses into their future strategies for their own benefit, but to the detriment of the affected population (Kirsch, 2014).

Despite the fundamental need to be aware of, and to anticipate, the impending tensions of such a scientific engagement, this is not a reason to reject it. An anthropologist who lived in Brazil with the Yanomami was categorical in declaring that

the social engagement of the ethnographer cannot any longer be seen as a personal political or ethical choice, optional and foreign to the scientific project. It becomes an explicit and constituent element of the ethnographic relationship. The anthropologist’s observation is no longer merely ‘participant’; his social participation has become both the condition and the framework of his field research. (Albert, 1997: 57–58)

Land Conflict: Starting with Confrontation

Prominent Marxist (Hall, 2013; Hart and Negri, 2000), postmodern (Escobar and Alvarez, 1992), and political ecology (Akram-Lodhi, 2007; Li 2014) scholars have demonstrated that the principal adversaries of indigenous territorial and land claims throughout the world are the protagonists of a neo-liberal economic model that has impoverished and dispossessed major sectors of rural societies, blocked the improvement of locally based production (subsistence and commercial agriculture), and promoted capitalist expansion by excluding populations. Besides, anthropologists have revealed struggles and displayed various exposures and mechanisms of resistance (Hale, 2006; Li, 2014; Scott, 1976; Stocks, 2005).

The Primacy of National Development

Since the early 1990s, Cambodia has followed this ideology, inspired and enhanced by Western countries (Hibou, 2004). The impact of government policies on the poor towards land and forests has been criticised (Bourdier, 2009; Global Witness, 2009; NGO Forum, 2015; Un, 2013). The 2001 Land Law, with additional legislations sanctioned in 2005, offered legal tools for granting ELCs to national and international companies, even if Article 29 of the Land Law states that no authority outside the community may acquire rights to immovable properties belonging to indigenous communities. In 2004, Vietnam, Laos, and Cambodia ratified a Master Plan, of which Ratanakiri is the epicentre. ⁴ Once a remote area inhabited by non-Khmer populations (Bourdier, 2006), the province became a destination for landless migrants, speculators, absent landowners, and foreign companies, due to its geostrategic position in this borderland region and fertile basaltic soils.

National laws exist to protect forests and limit the size of land concessions. They require consultation with local communities over land use. They are exceptionally enforced (Neth et al., 2013). In 2013, the NGO Licadho, advocating on land rights, claimed that more than two million hectares, equivalent to half of the country’s arable land, were handed over to ELCs. ⁵ The government denied these calculations, but scientific studies corroborated them (Neef et al., 2013; Neth et al., 2013; Young, 2016).

Ratanakiri is coveted by numerous ELCs. Most concessions overlap with villagers’ farmland and common forests, leading to a rapid increase in rural landlessness and social disruption within the villages (Bourdier, 2008). ELCs are granted, in an opaque way, by three ministries – not all in conjunction with each other – without any knowledge and consent of the native peoples (NGO Forum, 2015). Customary land tenure is unrecognised, and traditional land uses are misleadingly labelled as wastelands and considered underdeveloped and untapped (Ironside, 2017). Most villagers, lacking official land titles, can be considered “illegal occupants.” The authorities can expel them without adequate compensation and steal portions of their ancestral agricultural land (Global Witness, 2009). Such is the violence of the law: ELCs are issued on Private State Land where farmers have no legal claim in virtue of the 2001 Land Law mentioning that possession must have been established beforehand. Put simply, any household clearing land after 2001 is deemed to be an “illegitimate occupant.” Indigenous populations are theoretically protected by specific provisions of the Land Law and have a legal claim to their customary land. But the political reality is different. As we will see below, none of these considerations has prevented massive deforestation and land grabbing with the “three-in-one” ELC granted to HAGL in a territory populated since time immemorial by indigenous groups. Before going further, preliminary information about this company deserves attention.

HAGL has created havoc previously in Vietnam and Laos (Kenney-Lazar, 2012). Global Witness, a London-based NGO, has documented the devastating impact of Vietnam’s rush for rubber on local communities in Laos and Cambodia (Global Witness, 2013). Investigations showed how international financiers like Deutsche Bank, Crédit Suisse, and IFC backed these land grabs, contravening binding rules proclaimed in their policies. At the heart of this story lies a Vietnamese company, operating in Ratanakiri through a web of subsidiaries concealing the identity of the true owner. This “Rubber Baron” ⁶ and its subordinate affiliates have carried out forced land grabs since 2009 without providing compensation. HAGL has also close ties to Cambodian elites belonging to the Cambodian People’s Party, which has ruled the country for more than thirty years (Global Witness, 2013). Up to 2013, HAGL’s financial backers turned a blind eye: the company continued to receive investments from European banks and IFC via intermediary funds, like Dragon Capital Group, a Vietnam-based investment group specialised in emerging financial markets. These investments spawned human rights abuses; threatened national integrity; and breached the most basic legal, social, and environmental standards. Some investors changed their positions, but belatedly: four years after HAGL took the land, Deutsche Bank divested following the 2013 Global Witness revelations. The Crédit Suisse stopped in 2016, seven years after the ELCs’ granting. Dragon Capital sold its shares soon after local villagers filed the complaint to the World Bank Group. Nowadays, HAGL’s financing depends on Vietnamese banks. ⁷

Birth of an Independent Social Movement

Villagers in 14 rural settlements, with an average of 85 households, discovered the intrusion of the company in 2009 when tractors and bulldozers, protected by private militia and Cambodian forces, chopped down the forest and appropriated some of their cultivated and fallow lands. For people practicing swidden agriculture, fallow land is part of the parcel, not abandoned land. HAGL had received 18,952 hectares for growing rubber. None of the fourteen villages, composed of Jarai, Kachoh’, Tampuan, Lao, and Kreung, living inside or on the fringes of the ELC, had been consulted or informed, either by the government or by the company. Only vague meetings occurred: they were told about a national development-oriented modern agricultural project. Its location in their ancestral lands was concealed. ⁸

Villagers reported that the company has been logging in huge chunks of their territory, in complicity with national enterprises and local individuals. They were aware that similar loggings also took place in protected areas and in a nearby national park. ⁹ Physical violence, harassment, and forced evictions of families living in their swidden fields have been reported. ¹⁰ Hundreds of households have been deprived of their livelihoods, including forest products, water sources, and subsistence agriculture.

Testimonies from Tampuan, Jarai, and Kachoh’ attest that most villages responded promptly in unexpected ways when the company started cutting the forest, destroying cultivated lands, and putting up fences. The situation was intolerable. It was the first time that villagers of both sexes had become so vocal and defiant against orchestrated land grabbing. ¹¹ Of course, not everybody reacted with the same intensity, and not everywhere with the same conviction. Some households who were not directly affected did not participate in the movement but were advised to join the protests, mostly in Jarai and Lao settlements. They were told that it could concern them sooner or later. In some cases, communities federated, combined their action, and asked for updated information from local human rights groups. Access to justice via the provincial court was denied, under the pretext that it was out of the scope of the tribunal. ¹² Complaints submitted at the district level were silenced. Letters sent to HAGL never got a reply. Dispossession caused by the company continued, strengthened by physical threats, intimidations, and other forms of structural violence, including refutation of legitimate existence of land tenure for people practicing swidden agriculture, forced enclosure, and arbitrary interdictions prohibiting entry into the ELC zone. ¹³ At this time, provincial/district authorities acted as the agents of the company, warning the villagers not to obstruct the development project, which would generate well-being, employment, health, and social services to them.

Time passed, people were left unsupported, and could not maintain their stand with their initial vigour. In the five villages I visited in March 2012, concerns over survival bypassed the eroding feeling of fight. To the best of my knowledge, only some Lao and Kachoh’ (men and women, young and old) living close to the HAGL office kept on trying to recover portions of land and were adamant about preventing the company from cutting trees and extending its territory. A typical strategy employed by quite a few Lao, Kachoh’, Tampuan, Jarai, and Kreung families consisted in establishing their farmland nearby the HAGL concession (and sometimes even inside, in the not yet occupied territory) in order to prevent the expansion of the plantation.

In 2013, five NGOs ¹⁴ joined the villagers and provided legal tools but not all the tool-givers had juridical qualifications, nor solid sociocultural knowledge of the area. Following NGOs’ field enquiries, it was confirmed that HAGL’s concession overlapped with land traditionally occupied by indigenous communities. A human rights impact assessment was conducted (Bugalski and Thuon, 2015). The size of the ELC was found to be illegal (more than 10,000 ha). In breach of the right of indigenous peoples to self-determination, included in the 2014 Declaration on the Rights of Autochthonous peoples signed by Cambodia, ¹⁵ the company made no attempt to consult communities or seek their free, prior, and informed consent to conduct operations. All eight village elders interviewed [by me] corroborated this. The compulsory ecological and social impact assessment never occurred. ¹⁶

Inclusive Development International (IDI), the only international NGO directly involved for providing pro bono legal assistance to communities, conducted financial tracking and discovered that the WB’s private sector arm did not lend directly to the company but relied on financial intermediary entities to deal with HAGL (IDI, 2017). Such acrobatic exercises are a common practice in the WB phalanstery for being invisible and escaping responsibility for the fiduciary company’s actions.

Since the concessions were granted, affected communities have experienced loss of communal and individual property. The company did not allow the people to enter “their” domain. More than a hundred households experienced severe loss of assets such as orchid lands, paddy fields, animals, crops, and forests used for collecting non-timber products. Affected families could no longer practice their cultural and spiritual traditions. Some farmers revealed their family burial places, located in their cultivated land, had been physically destroyed: this sacrilege prevented them from conducting their traditional second funeral allowing the spirit of the dead person to quit forever the human world. ¹⁷

Despite their struggles, affected people have been denied the right to an effective remedy for human rights violations. Personal anthropological investigations confirm that the introduction of large-scale agribusiness has reduced local food security and sovereignty and shrunk existing livelihoods, as has been shown elsewhere in Cambodia (Prachvuthy, 2011) and in others countries (Li, 2014; Rosset and Martınez-Torres, 2007; Stock, 2005). Occupying people’s land means depriving villagers from land tenure, but more fundamentally it treats with disdain a socio-territorial order, negotiated rules and social fabric, all historically constructed, on which a local citizenship relies to regulate access to land.

Paradoxically, early in 2013, a timid effervescence prevailed. No solution had been found, and no real organisation had been planned; however, the villagers, at least in some of the eight communities, foresaw possible encouraging results, not only thanks to the NGOs but from the seeds they had sewn themselves. ¹⁸ The company did not expand its territory, and villagers attributed this to their tenacity. The situation was far from stabilised, and internal/external conflicts generated by the land encroachment were rampant. But some villages had been socially united, despite ethnicity, status, age, and sex differences. They developed, without systematically displaying any external violence, a sense of solidarity in a situation of crisis. For instance, when the provincial authorities tried to identify hypothetical leaders of a concerted action which led to the destruction of six hectares of recently planted rubber trees, a group of five interrogated villagers replied, “We are all leaders, you cannot divide us.” No ringleaders could be arrested, simply because there was no one. “If you want to arrest us, you have to arrest us all.” ¹⁹ Being vocal was essential. Some local emissaries had journalists and well-connected people’s phone numbers. Others disseminated stories via Facebook and welcomed articles published in the Cambodian press. International visibility was expected as a redemptive strategy. But with the increasing interventions of external actors, a new configuration was about to alter the situation.

Entering Mediation: The Demise of a Social Movement

The proposal for an alternative approach came from the international NGO who shared it with national/local NGOs, then with the villagers. Previous experiences in Cambodia attested that embarking on legal action was unrealistic. According to advisors supporting the affected villages, the national judicial system was in cahoots with the political party in power, leaving no chance that a court case would be handled with integrity by the judges and their counterparts.

The Establishment of New Regulations: Other Actors, ‘Original’ Initiative

The involvement of a third party, as mediator, was proposed as a realistic option. It was decided to call upon the Compliance Advisory Ombudsman (CAO) to mediate. It is the IFC’s watchdog, created in 1999 by the World Bank Group, board to hold IFC management accountable to its safeguards. ²⁰ However, both IFC and CAO are in the same building in Washington DC. It obviously calls into doubt the independence of CAO as a neutral moderator. And the content of the CAO website is of no interest for the subject if not properly scrutinised.

Did the decision to call for international mediation arise from the sincere hope that it would effectively work? Or was it chosen as the “last option,” because other recourses would not fit, or were “unthinkable” in the geopolitical context of the country? In any absence of written documents about the decision process, NGO members provided evasive answers. Some said that they were willing to “try” but without any absolute expectation, while others remained sceptical of the capacity of the CAO to handle the case properly. Others were confident that it could create a precedent in the country. On the villagers’ side, no real alternatives were considered: residents were repeatedly coached and encouraged into admitting that viable options were limited, even if participating NGOs guaranteed their freedom to abandon the mediation process for engaging in another kind of confrontation. ²¹

A commitment appeared: IDI verified whether the IFC’s policies on financial intermediaries had been breached (a prerequisite for lodging a complaint to the CAO). The NGOs Equitable Cambodia and Highlanders Association addressed the question of the affected villagers and explained to the people the complexity of the CAO mechanisms. Each village elected at least two representatives (sometimes up to five or seven) for the mediation process. 2013 was devoted to external investigation, village workshops, financial tracking, local advocacy, and technical training. Early 2014 saw the realisation of a written complaint co-edited by villagers and NGOs. The document was sent to the CAO, which transmitted it to the incriminated company. ²²

The procedure moved quickly. The complaint was sanctioned by the CAO. HAGL did not deny the contents. Optimism prevailed. David had dared to confront Goliath. This important step was unanimously considered as a positive outcome. A second current was prevailing in the villages, even if not shared by everybody: some were suspicious of the encouragement of the NGO Equitable Cambodia and the dispassion of an indigenous NGO. Others wanted immediate action, because they were landless, with the trauma of dispossession being extremely acute. Besides, the concept of mediation was not clear to everybody.

Meanwhile, community-elected representatives formally sat face-to-face with the company in January 2015, under a CAO dispute resolution process. HAGL delegates declared the company’s commitment not to carry out any further deforestation and amplification of its ELC. They promised to return the forests and the lands which had not yet been exploited. HAGL gave the impression of being sorry and repentant. Hoang Anh Oyadav ELC agreed to reduce its size from 16,849 ha to 13,720 ha (Work, 2016: 6), though still illegal. Being Private State Land, the 3,129 ha were to be returned after approval by the concerned ministry. A villager reminisced that, “the HAGL administrators’ attitude during the first third party’s meeting made us believe that they truly recognised their mistakes and would accept a peaceful conciliation.”

Personal observations and enquiries in Tampuan, Kachoh’, and Jarai villages show, however, that deforestation had already been perpetrated on a large scale. No more precious wood and resin trees (previously tapped) could be inventoried. Luxury timbers were exported by a Cambodian tycoon to Vietnam. ²³ And the retroceded land was largely composed of meagre sections of degraded forests with weeds and invading shrubs.

NGOs seemed confident of having created an adequate social environment for approaching the conflict: village-selected representatives, mostly young men, acting as local interlocutors, had constituted village committees who were apparently “conquering” the case. For the majority of the interviewed peasants and most indigenous officers, it was a conclusive proof justifying the adopted strategy with the CAO. The audacity and tenacity of the complaining party was praised. But, in the two years following the oral promises made by HAGL, the optimism in the villages would soon be reversed.

External Non-Advocacy at the Village Level

It was under these good auspices that the negotiation process was about to continue: a “neutral mediation” facilitated by the CAO, supposedly non-aligned, whose role was to try to reach an agreement between the parties, one that responds to the grievances of the complainants. The CAO’s mandate is to resolve disputes using alternative conflict resolution approaches that include mediation, joint fact-finding, and facilitated dialogue between two parties (CAO, 2017).

Moreover, the mediation process shielded HAGL from unpleasant options such as international media attention, a summons to appear in court for illegal activity (farms destruction), or a would-be arbitration in Europe (with the Deutsche and Swiss banks’ complicity). When entering negotiation, the company announced that it would accept the CAO’s mediation, provided the communities remained discreet and did not publicise the case. Implicit messages for the villagers were retain a low profile based on mutual respect and do not spoil the company’s reputation. ²⁴

Returning to the short-term excitement in 2014 about a fair and comprehensive “conflict resolution,” the company confirmed a series of promises, including genuine compensation. Indigenous grass-roots NGOs and national and international NGOs continued mobilising the population, training highlanders in the intricacies of the process so they could be village interlocutors and communicate with the outside world, even if “sensitive” details of the procedure had to be kept secret (a mediation “obligation”). The training allowed an understanding of how the system was working, along with its juridical, legal, and sociopolitical mechanisms. However, personal enquiries in 2016 and 2017 show that not all indigenous representatives, trained to use the system to fight the system, were willing to remain as mere mediators. Some used their prestigious status to decry the traditional authorities, arguing that the elders were useless and could no longer understand those complex procedures. Internal tensions started brewing, leaving several inhabitants uncomfortable about being passive recipients who merely received information from the elected representatives. Others misunderstood the role of the representatives, believing that the election had been a trick to replace the elders, whose words have to be traditionally respected. Some newly appointed young village representatives were, according to elders, pretending to know and decide everything. Surreptitious infighting, lack of communication, disagreements, and factionalism infected some communities.

Power Imbalance, Disorganisations, and Delays

With the mediation process, an evident power imbalance took place: some deprived villages totalling over 6000 inhabitants supported by five NGOs versus one of the biggest Vietnamese private conglomerates, backed by a battery of skilled managers, communication specialists, technical consultants, human resource advisors, jurists, lawyers, and well-connected partners like high-ranking Cambodian officers (Global Witness, 2013), private/national banks, and intermediary funds (IDI, 2017).

By mid-2017, after three years of negotiation, what had happened? Nearly nothing concrete for the villagers. Cut trees hadn’t been replaced, the rubber plantation was growing, and cultivated land seized from the farmers had not been returned. In a Tampuan village, some impoverished families forced to sell their land to well-off families remained incredulous at the mediation process being unable to solve their urgent need to find a place for growing rice. Joined families became more individualistic. Anybody going to Kachoh’ villages close to the Sesan River could attest to the socio-environmental desolation: ghost households in almost empty settlements replaced the previous long-houses hosting fifteen nuclear families from the same lineage. I asked an elder where they all had gone. “They moved elsewhere,” she whispered. How could unmarried girls find suitable grooms in this socioecological desert? Youngsters who agreed to work as daily-wagers for the kronmum (company) were in trouble, because they couldn’t fulfil their pivotal role with their in-laws by working on the family farm. Other nearby Jarai villages were unrecognisable: no trees, and the farms, if still existing, were extremely distant. A disabused man said that he would become a cement factory supplier with the emerging market demand for fences and factory buildings.

During the mediation, a CAO adviser was replaced after being accused of bias in favour of the communities. CAO staff from Washington came for short trip visits to the field, taking photographs when playing football ²⁵ in the provincial capital with the “locals” and HAGL employees; appointed consultants, some highly paid, came for providing community mapping training, discussing past experiences and disseminating information and awareness; stakeholders’ meetings with villagers continued minus the HAGL delegates, who, for over a year, stopped attending. The latest CAO mediator appointed in 2015, a lively South American indigenous woman lawyer, tried her best to put order and restore hope amidst the mess where so many villages were paddling.

A lot of money had been spent to keep the CAO mechanisms alive (Work, 2016). ²⁶ Five NGOs provided assistance, consultants (CAO mediator, two national “focal point,” a CAO staff from Washington, some technical experts) made punctual interventions, HAGL and provincial government commissioners occasionally joined meetings and fieldworks. A careful estimation of the number of actors involved, part-time or permanently, ranges from 90 to 120, and may be more. In the villages lived 6,000 insecure and exhausted individuals with the symbolic benefit of being part of a mediation process over which they had gradually lost control because of mutual disagreements, internal fights, and impatience against the slow process. ²⁷

NGO advisors worked to strengthen the communities and prepared terms of negotiations. The mediator undertook extensive trips to the villages in 2017, trying to make the villagers more confident. She stimulated communities to draw their territories. ²⁸ This social mapping, in vogue in Brazil (Almeida, 2013), allowed people to reappropriate their social space.

Moreover, negotiation with national and provincial authorities, deemed vital for the mediation process, was slow to function beyond the village level. In the present case, once rural peoples reach a certain level of anger, protest, and mass mobilisation, national authorities first objected to external interventions, in a harsh way, under the guise of sovereignty, but finally accepted national/international agencies to negotiate, form joint study commissions, and engage in various forms of diplomatic conflict resolution. ²⁹ The Cambodian government, due to its weak institutions, has for long been transferring its responsibility for gathering information, monitoring and evaluation to some NGOs, provided they remain apolitical, not inquisitive, do not infringe the national sovereignty, and propose a win-win solution. The efflorescence of NGOs’ activities stems in part from the failure of the State to regulate multi-national corporations, a policy outcome of economic models emphasising development and the free-market ideology that is central to the Cambodian neo-liberal agenda (Hughes, 2006; Young, 2016).

The CAO mediator finally succeeded in bringing together government officials, villagers, and HAGL delegates for localising and measuring spirits forests and heritage sites. Early in 2017, a provincial working group accompanied the other actors to the eleven villages for demarcating these areas. Once accomplished, HAGL agreed to return burial places and tiny portions of spirit lands located on hills (all of them totally deforested by the company), provided the government georeferenced identified places with the indigenous peoples. Government officials produced a report to the Governor and State Land Management Committee recommending that these newly demarcated areas (mostly spirit mountains, cemeteries, and river zones) be removed from the ELCs and returned to the communities. The joint statement of the dispute resolution declared that from now onwards (but after eight years of strict interdiction), local inhabitants would be allowed to enter the concession if they needed to reach their cultivated fields, sources of water, burial grounds, and sacred places. This land retrocession ³⁰ would be made effective once the responsible minister endorsed the decision.

Until the second semester of 2017, the ongoing procedure showed, to the disappointment of local inhabitants, that HAGL was not willing to honour its promise of significant land restitution and real compensation. Cemeteries and spirit forests were not sufficient, and most of them, specially the spirit forests, were already devastated. The company offered a forgiveness ceremony in each village (buffalo sacrifice with jar rice-wine drinking), but the ritual, supposed to console the spirits of the land and of the forest having suffered from the company’s undue intervention, was sometimes interpreted by interviewed Jarai and Kachoh’ as a subtle way to befog the villagers, diverting them from their quest for real justice (land return, robust compensations for the most vulnerable families, improved livelihoods, and quality of life). Other communities perceived this forgiveness ceremony too insufficient for cooling down the spirits’ anger insofar as the latter still continued, in December 2018, sending misfortune to the affected villages (sudden death, mortal disease, epidemic, bad harvest, social disharmony, lack of rain). Villagers also deplored the company’s non-response to their written requests, and non-attendance at the meetings supposed to promote dialogue. Arguments percolated in the villages: some individuals were willing to pursue the mediation, while others were opposed. ³¹ In time, HAGL requested the provincial government to address the villagers’ grievances, given that they had been granted a concession by the national government, responsible for the welfare of its citizens. ³²

With undeniable dexterity, the company prevented this toxic affair from being excessively publicised. In the Cambodian national English and Khmer Press, news about the conflict decreased from 2014 onwards and virtually disappeared by 2015 during the mediation. By contrast, the July 2017 episode, with the announcement of token reimbursements, was publicised in both languages and interpreted as a step forward demonstrating HAGL’s clemency. Nonetheless, nobody mentioned the rampant social disorder in some villages – a situation that NGOs could neither foresee nor restore. Meanwhile, the company had preserved its reputation, promoted its aptitude for fair play, and managed to keep most of its stolen cultivated land.

Notwithstanding advisors’ commitments and mediator’s professionalism, most of these ambivalent results depended on interventions by the NGOs, along with the CAO mechanisms. Moreover, such a laudable initiative was inevitably destined to facilitate the games of the most powerful. Given the two parties’ undeniably unequal power, mediation has, likewise, generated “economic imbalance.” While impoverished villagers donated their time, some NGO personnel were substantially funded. During (my) inquiries, only a few practitioners agreed to reveal the amounts their organisation had received. The majority felt surprised, and irritated, when the demands of Lao and Kachoh’ individuals expecting them to be accountable were repeated [by me]. Such a request was understandable, given the fact that they received neither a piece of arable land back nor a single dollar of compensation for the time and energy spent for recovering, calming internal dissensions, training, travelling, organising internal meetings, and other innumerable services.

Mediation Versus Confrontation: Submission and Structural Impairment

The CAO affirms that, “mediation encourages consensual behavior: building empathy, cordiality, and friendly relationship” (CAO, 2017). All types of joint activities with the three parties help villagers and the company to understand each other and build trust. Mediation is perceived as allowing the disputing parties to make their own decisions, thus controlling the content of the agreement. If the appointed mediator personally insists on villagers’ empowerment, the CAO website explains that mediation restores and maintains business relationships.

Envisioning mutual business relationships means for affected people to get their land back, along with respectable compensation for natural resources lost. Anthropological investigations have already demonstrated elsewhere that the rules of the community – with land, since time immemorial, transmitted to the past, present, and future generations – have been violated (Edelman, 2009; Hale, 2006). Among the many sociopolitical aspects neglected by the solicited actors, one is that an indigenous person without land loses his dignity. He is a dead man walking. This notion, deeply rooted in the culture of Ratanakiri highlanders, ³³ needs to be understood.

Obtaining a practical, fair, and workable solution relies on these minimal considerations. But how can they go along with a company that had shown no interest in the well-being, even existence, of the inhabitants identified as obstacles, not as partners? Moreover, structural impairment infiltrates the imported concept of negotiation, supported by the fabrication of informed consent activated by a third party. In the Cambodian context, and indigenous peoples are not an exception, the inability to voice discontent or to engage in any sort of negotiation is an indication of the extent to which the model of gift-giving departs from the reciprocity of customary relations of patron-clientelism (Hughes, 2006). It is clear that, “the most common form of resistance to these kinds of efforts at regimentation has simply been flight”. Reference to rights and law is essentially a bluff, since even a recourse to external assistance is likely to depend on declarations of allegiance rather than on the formal provisions of law (Un, 2013). Insufficient institutionalisation of state roles and poorly qualified local officials generate uncertainty for any intervention at lower/higher levels. Villagers don’t know where to go other than human rights organisations ³⁴ taking the risk of being accused by the sovereign power of incitement and indicted for shaping a “color revolution” aimed at destabilising the country.

In time, confidentiality, implicit in the mediation process, blocked alternative ways of expression. Public advocacy regressed. Villagers unable to reach an agreement became confined in their universe. External interference was discouraged (no coverage depicting the progress of the mediation, no independent monitoring, no additional support). NGOs could join the negotiation, provided they complied with the rules, avoided openly criticising the system, and neither denounced the power imbalance nor the structural impairment of the process.

Supporting NGOs are filling the vacuum created by a retreated state, but not always in a substantial way. Certainly, international and national NGOs encounter great difficulties in Cambodia. They must comply with the national ideology. Open confrontation is forbidden. Any criticism is dangerous because top politicians will associate it with disorder, instability, and war threats. Under these difficult conditions, NGOs’ unintended influence on the HAGL case has been to annihilate the political anger that was anchored during the confrontation period. Notions of aid or benevolence replaced what people ought to have by right, therefore turning people into dependent victims who cannot do anything without external interventions, while blunting the edges of political resistance. The WB mediation has depoliticised the residual resistance that persisted when the complaint was prepared. By interfering with local peoples’ movements, well-minded NGOs became conciliators, interpreters, and facilitators. They did not try to reactivate a disillusioned social movement. Some indigenous people, but not all, have nevertheless managed to persevere, thanks to these external actors. And their temerity added more weight during the mediation process. But no real empowerment could be traceable as a whole.

The sociocultural disparity of the affected villagers has been exacerbated. They never constituted a homogenous group (Bourdier, 2008), but recent socio-familial dynamics show that the mediation process has amplified inequalities and tensions and intensified internal and external conflicts in many areas. ³⁵ Inter-village complications occurred between severely impacted settlements (Jarai, Kachoh’, and Lao) and less impacted ones (Tampuan and Kreung) having diverse priorities and strategies (survival vs livelihoods improvement). Complications arose between district authorities having to respond to provincial decisions and between commune council/village administrative units solicited by deprived households in need of an urgent solution. Intra-village problems took place when the traditional council of elders felt disavowed by the administrative headmen, when young men denied the capacity of old people to deal with the mediation procedure, and when poorly affected households didn’t show any enthusiasm for the mediation. Some village representatives confessed their ambivalent feeling for responding coherently to various queries made by their kin and other relatives. Quarrels emerged even at the family level when men discouraged young women from stopping domestic work for attending meetings and when husbands forced their wives to resign their position of village representative. All these disruptions might be niggling, but altogether they created a damaging environment.

What about neutrality and mediation? Experiences show that disputes between transnational entities and indigenous communities are hampered by structural inequalities, which makes it easier for corporations to take advantage of juridical–social assessment (Leclerc-Olive, 2017). And such assessment can subsequently be used to silence opposing voices. Moreover, the self-proclaimed rules of mediation stipulate non-divulgation of the ongoing process. It does not prevent information but selects its disclosure. Each joint resolution statement is encouraged publicity when the mediation is exhibiting a positive outcome. But non-technical analyses are avoided, specifically if they question the expected untroubled evolution of the process. Any conspicuous attempts to interrogate the relevance of elaborating a “depoliticized institutional apparatus that is by turn legal, financial, bureaucratic, and techno-scientific” (Brosius, 1999: 278) is received with suspicion, if not labelled as reactionary and counterproductive.

Still, NGOs in Cambodia are under pressure and can be closed without justification under the 2016 NGO Law. They are not immersed in political contexts, as in South America and the Philippines, characterised by strong socio-historical movements sometimes supported by local public institutions (Leclerc-Olive, 2017: 15-28). Supporting NGOs managed to block tens of millions of dollars in investment to HAGL: the company became mired in financial difficulties, unable to expand its overseas operations, having to abandon and sell off overseas assets. Regrettably, this achievement has no direct effect for the dispossessed Ratanakiri villagers. Nonetheless, external independent media can have an impact. The government closed a number of journals and radio stations and limited freedom of expression. But communication, in the age of the social media, continues. Information dissemination shape public opinion, connectivity can act as a counterpower, and the circulation of images is in itself empowering (Brosius, 1999).

This shady picture is counterbalanced by interesting outcomes. Dispossession and encroachment debates became tied, for some villagers, to broader struggles for democratisation, access to alternative development, and citizenship. The language elaborated by several interviewed indigenous Jarai, Kachoh’, and Lao turned into an effort to challenge present-day structures of domination and normative discourses. Other individuals, mostly the elected village representatives, started to use ecological terms (biodiversity, climate change, pollution, etc.), challenged destructive resource extraction practices during public gatherings and at the village level, and denounced the “selling of the country” to business firms. New discourses had been intended to empower the historically disempowered indigenous communities but, according to what villagers confessed, these changes arrived too late, and it remains difficult for translating them into persuasive mantras during the negotiation that never converted into a collaborative force, in the sense given by Lassiter (2005).

Elsewhere in Ratanakiri, autochthonous people adopted non-institutional strategies. A Tampuan village near the Sesan River resisted the establishment of a small rubber plantation in 2015. Dreading power imbalance, negotiation was not envisaged. People started burning down company housing and didn’t hesitate to address threats against developers and their Phnom Penh-based Cambodian government counterparts. Federated against a common challenge, villagers kept on occupying parts of the planned concession to prevent the trees from growing. Being aware of impending complications (arrest warrants, physical and psychological harassment), an ad hoc committee played a subtle game with the political party in power, not for mere support, but to show the villagers’ allegiance to its political line by stimulating endogenous development. Recent research (Baird, 2017) indicates that villagers have been successful in stopping the company from moving ahead.

Epilogue: Annihilation or Suspension of a Social Movement?

The underlying logic of the still ongoing mediation process highlights the tense relations between government and NGOs and the impact of this mediation on local sociopolitical organisations. In the course of time, once the initial moment of mobilisation passes, the momentum is lost, and the demands activated by supporting NGOs are not strong enough and the promises made by the company are not kept. Villager’s organisations become less effective at achieving their demands after being less confrontational. They are less operational when they take more conciliatory positions and invest their energies in dialogue. Resistance, according to an indigenous NGO, still exists but remains dispersed, heterogeneous, and ineffective. The consequence of this biased situation occurring in Ratanakiri is not unique. Such as been the diagnosis elsewhere in the world (Borras, 2016; McKeon, 2013; Piven and Cloward, 1978). A methodology elaborated by foremost neoliberal agencies can give a false appearance of human and ethical concern but cannot lead to a genuine future for the oppressed. An approach in favour of the oppressed, not driven by top institutions like powerful international agencies, has yet to come in Cambodia. The negotiation organised with the affected villages by HAGL is no longer a struggle. It has reproduced and reinforced the very structures of oppression the villagers intended initially to resist. Transnational networks are aware of these risks in South America, where social movements have a special distrust, based on bitter experiences, of methods that channel and “calm” dissent: that is, of “conflict resolution,” “stakeholder dialogue,” WB “consultations,” and “participation” (Rosset and Martinez, 2007).

The present article followed the evolution of an uneven mobilisation under a mediation process which provoked disarray in local sociopolitical organisations. Ethnographic enquires confirm that micro-societies became split and fragmented not only before but more again after the mediation started. Besides, one may wonder whether such institutionalised mediation, with its vitrine of sympathetic discourse for the poor, can avoid being complicit with institutions like the WB supporting, with substantial benefits, financial activities contributing to the “global land grab” (Borras, 2016). And this is the point threatening the reputation of international agencies, that engaged researchers should reveal, with the aim of altering such a depraved worldwide exchange system.

Before and just after the IFC partially funded the project, critics failed to coalesce to claim that the company’s enclosure was illegal. It happened later, but it became much more complicated: what can be done, once the damage is done? NGOs accompanied a section of the population with training, tools and materials (village land maps, smartphones, technical knowledge, access to the financial scene, etc.). But nobody encouraged the villagers to lead their campaign under the banner of, for instance, powerful transnational core ideas persuasively promoted by prominent scientists, La Via Campesina, and other worldwide militant groups, like “food sovereignty” and “self-governance with alternative rules and procedures” (Borras, 2016; Martinez-Torres and Rosset, 2010). Such concepts are neither mere slogans not directives but flexible operational procedures allowing marginalised societies to control their destiny with their world visions and perceived priorities. These procedures, non-existent in Cambodia, cannot be implemented in a short time but do represent a significant envisioned societal projects implying the deployment of meaningful sociopolitical mobilisations in which populations can affirm their identity, clarify their concerns with others, get out of their isolation, lobby for their rights on a broader scale, and join transnational peaceful movements if such networking is compatible with their intentions.

One can object that this is “too” a vision for the future. In the meantime, in 2012, the State promulgated a directive (so-called Order 01) to initiate a vast revision of ELCs including the titling of land inside ELCs where rural peoples have rightful claims, and an evaluation of ELCs to identify whether there are in breach of their contractual requirement with the Government. This directive took place in 18 provinces, including in Ratanakiri. As a result, some ELCs had been cancelled or reduced. Such has been the case for HAGL, compelled to reduce its size (some thousands of hectares) in two districts. But the cancelled lands are Private State Lands: they have been retroceded by the company to the government, not to the people. And so far, the national authorities have not yet decided whether they will transfer the confiscated lands, or at least some sections, to the villages traditionally using them. If properly instigated after HAGL’s encroachment, dispossessed villages should get their lost territories back by Rights, not with a mediation process which is nothing more than a sticking plaster over a major wound.