Brokering justice: global indigenous rights and struggles over hydropower in Nepal

ABSTRACT This article explores the dynamics of brokerage at the intersection between the justice conceptions enshrined in global norms and the notions of justice asserted in specific socio-environmental struggles. Using the case of a small hydropower project in Nepal, we trace the attempts of an indigenous activist to enrol villagers in his campaign against the background of villagers’ everyday negotiations with the hydropower company. The study shows how global norms, such as indigenous peoples’ rights, may fail to gain traction on the ground or even become sources of injustice in particular contexts.


Introduction
Shyam is having a hard time convincing the villagers of Bargachhi to support his campaign directed at a small hydropower project. 1 The local chairman of the Nepal Federation of Indigenous Nationalities (NEFIN) has a point: the one meeting held by the project developer with villagers eight years ago does not meet the requirements of the indigenous peoples' rights to consultation set out in International Labour Organisation Convention 169 (ILO 169), ratified by Nepal's government. Shyam's problem is that ILO 169 has little traction with villagers, not only because few know about the convention but also because of its focus on procedural rights. Villagers' attention is on distributive matters: they request remedial actions for avoidable harms caused by construction, claim compensation for unavoidable damages and want a share in the project's benefits.
Shyam encounters a more general issue with regard to the linkages between global indigenous rights and socio-environmental struggles in specific sites: the conceptions of justice enshrined in global norms may be different from people's notions on the ground (Upton 2014;Sikor and Hoang 2016). In Bargachhi, villagers emphasise distributive issues because to them, justice is about avoiding harm and deriving benefits from the project. Their demands resonate with Nepal's regulations on hydropower projects, which stress distributive concerns over procedural rights. In contrast, NEFIN's stated agenda highlights indigenous peoples' self-determination and demands collective procedural rights as operationalised in the Principle of Free, Prior and Informed Consent (FPIC) in correspondence with ILO 169 and the United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP). Shyam finds himself caught in the middle: between the distributive conceptions dominant on the ground and in national policy, on the one hand, and the emphasis on recognition (that is acknowledging indigenous peoples' distinct identities and histories and eliminating any forms of cultural domination) in mobilisations and global norms on indigenous peoples' rights, on the other (Doran 2017)). He faces the problem pointed out by Merry (2006a) with regard to the promotion of global human rights: "[l]ocal communities often conceive of social justice in quite different terms from human rights activists" (1).
Discrepancies between global and local notions of justice may be a common feature of socio-environmental struggles around the world (Martin et al. 2014). Contestations in particular sites are able to invoke a rapidly expanding set of global norms, including those highlighting environmental concerns and others seeking to protect particular marginalised groups such as indigenous peoples. Similarly, where marginalised people have successfully mobilised against powerful companies or state agencies, they have been found to do so by crafting temporary alliances with activists and supporters at national and global levels (Sneddon and Fox 2008;Bickerstaff and Agyeman 2009). Such alliances require the involved activists to create overlaps in multiple problem framings, such as shared beliefs, ideas about collective action and strategies (Tarrow 1998). Part of these framings are ideas about justice, we suggest. Affected people and activists explicitly invoke or implicitly operate on the basis of particular notions of what justice means in specific contexts (Upton 2014;Sikor and Hoang 2016). Consequently, negotiations over notions of justice are a critical component in alliances between people and activists.
Intermediaries of various sorts, such as environmental activists, community leaders and the staff of nongovernmental organisations (NGOs), play a key role as builders of these temporary alliances. They may be rooted at the local level, engaging simultaneously in national and transnational mobilisations. Or they may focus on building transnational linkages and coalitions, yet retain some involvement in local and national struggles (Della Porta and Tarrow 2005;Caouette 2007). Given a certain charisma, intermediaries may help to form global justice networks (Routledge, Nativel, and Cumbers 2006). They may also help redefine and adapt global ideas to a particular historical and cultural context through the process of "vernacularisation" (connecting and creating meaning in a variety of ways) (Merry and Levitt 2017;Goldstein 2015). They also play critical roles on the ground, as observed in the struggle against a mining company in India (Kumar 2014) and the resistance against unequal exposure to industrial pollution in England (Bickerstaff and Agyeman 2009).
The intermediaries' role resembles the work performed by brokers in development (Olivier de Sardan 2005). Brokers help to bring together actors operating in different spheres, forging relationships across existing divides and generating shared framings of the problem. Brokerage helps constitute global norms, policies and projects (Mosse and Lewis 2006) by translating norms into the languages of diverse actors, recruiting supporters (Mosse 2005) and aligning actors' interests with global arrangements.
Drawing on these literatures, this article explores brokerage at the intersection of global indigenous rights and socio-environmental struggles in specific sites and facilitating crossscale alliances. We examine site-specific and global notions of justice and how brokers mediate between them when ideas about justice conflict. Thus, we contribute to a better understanding of how brokerage refashions global rights agendas for local contexts and reframes local grievances in terms of global rights principles and activities (Merry 2006b, 39).
Our case is a small hydropower project in western Nepal. In particular we examine a specific intermediary, Shyam. This bottom-up microscopic approach admittedly neglects important elements such as the role of national movements and struggles among global activists, yet it promises unique theoretical insights on the dynamics of brokerage in specific contexts.
Our data were collected through a variety of methods: a household survey; semi-structured interviews; key informant interviews; and a review of documents. The survey used questionnaires with 37 households, selected out of a total of 179 through stratified random sampling, to describe their livelihoods and experienced project impacts. Nine semi-structured interviews helped us understand how different kinds of villagers (from a wide range of socio-economic backgroundsclass, caste, ethnicity and genderand engaged in different relations with the project) evaluate project impacts, the actions undertaken by the company to address the impacts and NEFIN's demands. The analysis also draws on follow-up expert interviews and the co-authors' longstanding familiarity with Nepal and the study area.
We begin with a short review of hydropower politics, ethnic mobilisations and local politics in Nepal. Following a brief introduction to the study site, we analyse everyday negotiations over project impacts and benefits to demonstrate how villagers and company institutionalised a particular, distributive notion of justice in line with Nepal's hydropower policy. This sets the scene for our analysis of Shyam's attempts to relate villagers' concerns to NEFIN's agenda, which emphasised the recognition of indigenous peoples' collective identities over distributive concerns. We conclude with a discussion of brokerage, its mediating effects on the influence of global norms in specific socioenvironmental struggles and the particular challenges faced by indigenous peoples' activists.

The politics of hydropower and ethnicity in Nepal
Democratisation has had profound effects on Nepal's hydropower sector over the past two decades and a half. The new space available for public debate about the role of hydropower in development has shaped hydropower policy and politics in a narrow sense. Ethnic mobilisations and the associated politics of indigeneity have influenced the exercise of state authority over natural resources, including hydropower development. Democratisation thereby has facilitated the emergence of different politics of justice, one centred on the equitable distribution of hydropower benefits and the other focusing on the recognition of indigenous nationalities' self-determination.

Hydropower policy and politics
Nepal's development planners have long pinned their hopes on hydropower development as a source of foreign revenue and driver of national development. They have asserted that the country's 6,000 rivers had the potential of generating 83,000 megawatts (MW) of hydroelectricity although current installed capacity remains at less than one per cent of that (Dixit and Gyawali 2010). In the spirit of the dominant developmentalist vision (bikas in local parlance to refer to socio-economic development; see Pigg 1992), the government created the state-owned entity Nepal Electricity Authority. It gave the Authority the monopoly over power generation, transmission and distribution but also invited domestic and foreign investors into the hydropower sector. The 1992 Hydropower Policy and Water Resources Act paved the way for private investors to get involved through the "build, own, operate and transfer" model.
The hydropower projects proposed for various parts of the country in the 1990s provoked resistance from civil society. Civil society organisations and movements flourished in the newly democratic Nepal, taking up issues in a wide range of sectors including hydropower (Gellner and Hachhethu 2008). They asserted the importance of conserving the natural environment, protecting local people's livelihoods and rights and safeguarding the economic and social sustainability of hydropower projects. The Arun III project became an iconic case of civil society opposition in the 1990s, especially after the resistance caused the World Bank to withdraw its support (Dixit and Gyawali 2010). At the same time, civil society attention to hydropower issues contributed to the emergence of a regulatory framework in the 1990s that was later found to overlap significantly with the recommendations on responsible dam investment issued by the World Commission on Dams in 2000 (Dixit and Gyawali 2010).
Since then, while civil society activism has relatively diminished in Nepal's hydropower sector (Rest 2012), strong regulation has remained in place to protect the interests of local populations affected by hydropower projects by securing them a share in overall benefits and granting them a right to information. Other global trends and standards (for example, social and environmental standards incorporated as part of development aid and corporate social responsibility becoming increasingly prevalent in business operations) have also shaped new discourses on benefit sharing. 2 Even though the central government assumes a central role with regard to ownership, licensing and benefit distribution, the 2001 Hydropower Development Policy foresees that half of all royalties are shared with the affected development region (38%) and District Development Committee (DDC) (12%). The policy calls upon developers to employ local people in construction, provide suitable training for human resource development and implement corporate social responsibility projects. Since the Environmental Protection Regulations of 1997, 3 most projects have to conduct an Environmental Impact Assessment (EIA), identifying biophysical, economic, social and cultural impacts. The project developers are required to implement impact mitigation measures, present the draft EIA report to the local population in a public hearing and compensate affected people for all harms and damages (Dixit and Gyawali 2010).
However, this mechanism has come under pressure recently as Nepal's government has tried to attract new private and foreign investment. The government introduced various measures to overcome local resistance against hydropower projects, simplifying the process of land acquisition, waiving the EIA requirement temporarily and limiting the benefits local people could expect to receive from new projects.

Ethnic mobilisations
Nepal has an extremely diverse population. The 2011 population census recorded a total of 125 social groups, including Hindu caste and ethnic groups (CBS 2011). The previous census of 2001 had put the number of ethnic people at 36 per cent of the total population (Hangen and Lawoti 2012). In 2002, the Nepal Federation for the Development of Indigenous Nationalities Act established the term "indigenous nationalities" (adivasi janajati) as a legal category, listing 59 officially recognised groups.
Since 2002, ethnic mobilisations have formulated outwardly political demands (Hangen 2010). Indigenous nationalities have demanded access to economic and political resources to redress the historical control of national politics and economy by the members of the traditionally dominant high-caste groups. In 2006, the year of the so-called second people's movement following years of Maoist conflict, indigenous activists supported wider demands not only for a multi-party democracy but for autonomy and self-governance of ancestral homelands. 4 They influenced the drafting of the 2007 interim constitution, took on an active role in the 2008 and 2013 Constituent Assembly elections and were vociferous in raising their demands, such as for ethnicity-based federalism, albeit not met, in the drafting of the 2015 constitution. 5 NEFIN has been the dominant organisation in the indigenous nationalities movement (Hangen and Lawoti 2012). It has successfully claimed a central role in national politics by staging demonstrations, mass rallies and other forms of public protest. NEFIN actively participated in the movement to curtail the powers of the King in 2006 and signed an agreement with the interim government in 2007, which signalled an implicit recognition of its role as representing indigenous nationalities throughout Nepal. In contrast, its efforts to increase its presence outside Kathmandu, for example by establishing district coordination councils, have been less successful (Hangen 2010). In such context of considerable distance and disconnect, intermediaries like Shyam play an important role in linking NEFIN's broader activism with local issues.
The ethnic mobilisations have influenced Nepali politics and the exercise of state authority by asserting justice claims centred on indigenous nationalities' demands for recognition. In 2002, the Parliament passed a bill for the establishment of the National Foundation for Development of Indigenous Nationalities to fund welfare programmes for ethnic groups. Government budget planning started to make special provisions for support to ethnic groups. The new constitution has identified the rights and specific entitlements of indigenous groups as a matter of social justice. However, these successes of ethnic mobilisations have caused a relative decline in attention to the plight of other marginalised groups, particularly the Dalits.
ILO 169 and UNDRIP have played a critical role in ethnic mobilisations in Nepal, giving additional weight to the emphasis on recognition over mere distributive demands (Bhattachan 2012). Nepal was the first country in Asia to ratify ILO 169, requiring government and private sector to consult affected indigenous peoples on policies, programmes and projects (Jones 2012). ILO 169 and UNDRIP provide activists with a specific tool to implement self-governance: consultations based on the principle of Free, Prior and Informed Consent (FPIC). 6 Natural resources "have become a key battleground" (Jones 2012, 627) for self-determination. NEFIN advocates a maximalist demand that natural resources should be considered under indigenous peoples' collective ownership, a demand that is rejected by virtually all major political parties (Jones 2012).

Distributional coalitions in local politics
Local politics in rural Nepal often remains dominated by "distributional coalitions": local powerbrokers team up to control the local population's access to services provided by District and Village Development Committees (VDCs) (Pfaff-Czarnecka 2008). The powerbrokers typically include government bureaucrats, politicians, businessmen, local elites, high-caste groups and important school teachers (Sharrock 2013). Particularly from 1998 to 2017, due to absence of local elections, local government officials and politicians were appointed by higher-level leaders. 7 Lacking accountability downward, they often collude with businessmen to control the provision of benefits, channelling services, contracts, jobs and so on to "one's own people" (afno manchhe) (see Bista 1991). Challenges arising from new political parties, NGOs or other social groupings are addressed by inviting serious contenders into the coalition and including them in the distribution. The influence of the coalitions even extends to presumably civil society institutions, such as public concern committees (PCCs) established around development projects (Sharrock 2013).
Hydropower projects are subject to these patronage politics, since they are a significant source of finance, jobs, contracts, services and other opportunities (Rai 2005). They allow local politicians to offer the promise of electricity connections to their constituencies, which have significant leverage in a country that experienced prolonged power outages. The jobs and contracts for the local population represent opportunities for distributional coalitions even where benefits are disbursed by private developers and not through local government. A common vehicle for the operation of distributional coalitions outside local government are the PCCs. Additionally, once they come online, hydropower projects make significant contributions to the local government budget through the royalties shared with the DDC and transferred to VDCs.
In sum, hydropower development in Nepal witnesses the confluence of hydropower policy and politics in the narrow sense with ethnic mobilisations, shot through with entrenched local distributive politics. Hydropower policy (emphasising distributive concerns) and ethnic mobilisations (demanding recognition of indigenous nationalities' self-determination, including control over natural resources) offer competing conceptualisations of justice with regard to hydropower development. The procedural rights granted to indigenous peoples under ILO 169 and UNDRIP go beyond the consultative procedures required by Nepal's regulation (Jones and Langford 2011). Whereas global indigenous rights emphasise the "sincere desire to reach consensus" (Jones 2012, 627), Nepal's regulation limits participation to information-sharing. The national hydropower policies emphasising royalty sharing and local employment, for example, coupled with local desires to development, also act as a classic means of co-opting local villagers to development projects. These provisions become cultural lenses through which the villagers interpret their situations (Merry and Levitt 2017). FPIC becomes a secondary frame of reference while issues of benefit-sharing become the priority (Lord 2016). Thus, hydroelectric actors' and indigenous activists' ideas about socially just hydropower development are far apart.

The study site: a river valley in western Nepal
The Bargachhi VDC is located in western Nepal along a river that a prominent Nepali development planner referred to as a "golden egg": because it is snow-fed it provides a fairly reliable flow throughout the year, making it ideal for hydroelectric run-of-theriver projects. In the VDC, six villages are directly affected by the project. They include 179 households, of which 65 per cent belong to indigenous groups (Gurung, Ghale, Tamang and Newar), 18 per cent to high-caste groups and 17 per cent to Dalits.
Smallholder agriculture is the main source of subsistence and cash income for the villagers. In our survey, nearly two thirds of the sample households specified cultivation and livestock husbandry as their main source of food and income. Yet, just one half of all households had their own fields while one out of six households worked as sharecroppers on land owned by others. Consequently, only one out of 20 households reported selling agricultural produce, the rest consumed all produce at home. Mirroring the situation across Nepal, Dalit households found themselves at the lower end of the distribution, the average households possessing significantly less land. Living standards were low; some villagers barely met their own subsistence requirements. Half of all households lived in houses with brick walls and corrugated iron or tile roofs. Nearly half of all households reported that their annual cash income remained below 50,000 Nepalese Rupees (NPR), whereas a quarter declared a cash income above NPR 100,000 (1 USD = NPR 100). Five out of six Dalit households had an annual cash income less than NPR 50,000 and lived in houses made from thatch or mud.
Construction of the hydropower project began in 2012, under a license for installed capacity of 50 MW and a period of 35 years. The run-of-the-river project does not require the building of a massive reservoir but channels the river flow into a five kilometre-long tunnel through the mountainside. The project is a joint venture between a foreign and Nepali partner, which we refer to as "the company". The company agreed to invest an estimated NPR 10 billion in the project under Nepal's "build, own, operate, and transfer" policy. It signed a power purchase agreement with the Nepal Electricity Authority and started generation in 2017.
The construction works affected villagers, although impacts were relatively small in comparison with those caused by hydropower projects involving dam construction (for example, Rai 2005). The project did not require much land, minimising the need for permanent resettlement and the loss of agricultural land. It might not even compete with local agriculture for water, since villagers were mostly using secondary streams for irrigation. However, they reported other impacts of construction, as anticipated in the project's EIA prepared in 2005: heavy dust, cracks in structures due to tunnel-related detonations and decreased drinking water. In all, 15 households had to be relocated temporarily, and one permanently.

Everyday negotiations over hydropower
Once the construction started, villagers engaged the hydropower company actively. Many acknowledged the practical benefits the project brought to them, in particular the construction of a new bridge and road improvements. Some professed pride in living next to the country's first road tunnel, which the company had to blast through the mountainside for their vehicles. Since the project signified socio-economic developmentbikasto them, opposition to the project in its entirety was unconceivable. Nevertheless, all villagers we talked to voiced certain demands with regard to the hydropower project: they expected the company to undertake remedial actions against avoidable losses, pay compensation for unavoidable losses and provide further benefits to the local population.
The dust generated by the company vehicles was the problem that annoyed villagers most. Villagers were concerned that the dust would damage their health and resented its effects on shops, houses and agricultural fields along the road. They requested the company to sprinkle the road on a regular basis, which the company usually complied with by sending water trucks down the road. When the company forgot to do so, or when the road got too muddy because of too much sprinkling, villagers spoke directly to the truck drivers. In a few instances, when company drivers did not follow their requests, villagers simply blocked the road for a few hours until the company sent the next water truck, or the road dried up.
The second-most important concern to villagers were the cracks that they saw develop in some of their houses due to the explosions set off by the company. As soon as villagers noticed new cracks in their houses, they reported them to the company. Company staff sent to assess the damage would take photos for storage in a project database. The photos served as evidence for the company staff, a technician of the Environmental Management Unit, a policeman and the District Technical Officer to decide about the validity of the report and significance of damage. If approved, the company offered villagers compensation for the damage, which people appreciated in principle despite complaints that the amount was too low.
The assertion of claims and payment of compensation was never a clear and straightforward process, as illustrated by four households that had to vacate their houses for the heaviest construction period. On the one hand, the affected villagers did not tire of stressing and inflating the costs and losses imposed by the temporary relocation. As an elderly lady told us, after spending part of the compensation payment to build a temporary shed, her household would spend the remainder "on purchasing vegetables because we don't have a kitchen garden here." At the same time, she noted that "we have deposited the remaining amount in a bank". On the other hand, company officials complained in unison that "people are never satisfied" and "demands are always exaggerated", although the compensation was very good. "If Nepalese would run the project", an environmental officer pointed out, "the compensation amount would not be so high". Yet, the company had to give in to villagers' demands, in part because it was under pressure to finish in three years and because it could not afford to generate local hostility, since it was to stay in 318 T. SIKOR ET AL.
operation for the next 35 years. In the end, company and villagers always met somewhere between their initial positions. Another area of heavy negotiations between villagers and the company was labour hire. Villagers generally welcomed the generation of new jobs through the construction. Some 400-500 people from Bargachhi and its vicinity found employment with the hydropower project, close to 200 foreign workers and 200-300 workers from other areas of Nepal. However, villagers were also deeply suspicious whether the company had done enough to generate local jobs. Despite requirements under Nepali regulations and the EIA to give preference to local workers, many villagers were convinced that the company preferred to hire workers from elsewhere in Nepal, and neglected to provide training for local labour. A man explained to us that he was not satisfied with the low pay and thought "it would be better to go to Gulf countries for employment".
In stark contrast to the general dissatisfaction with labour hire, villagers were positive about the neighbourhood support programme, even though the company started it only in 2014. The agreement was that Bargachhi would receive more than half of the support because it was the most affected VDC. The company committed a total of NPR 40 million for three years and declared that in line with the EIA, the budget was available to support local schools, health posts and whatever would be deemed as critical for socio-economic development. 8 An immediate use of some of the funds was the construction of a water tap in a village where construction activities had affected a drinking water source, according to villagers.
Together with the neighbourhood support programme the company established a PCC in early 2014. According to the company's public relations officer (who is from a local Gurung indigenous community), the PCC's main task was to "mediate between people and the project" and "facilitate smooth project operations". Once the PCC was set up and registered as an NGO, the company expected all dealings with villagers to run through it. It had the mandate to allocate funds in the neighbourhood support programme and to recruit workers for the company. The PCC's 53 members are diverse, including Dalit, women and people from various ethnic groups, as its Chairman Ram pointed out. An influential school teacher and a powerful local party leader, Ram belonged to a major indigenous group in Bargachhi. More importantly, as the company's public relations officer noted, "the PCC was formed through multiparty consensus", including all five major parties active in Bargachhi.
Two months after the PCC's establishment, villagers' suspicions about its nature and actions ran high. Several people told us that many villagers did not feel represented by it, which they saw as responding to company concerns only. A Dalit lady identified the Committee narrowly with its Chairman Ram and said that Ram only implemented what the company told him to, and did not listen to villagers' concerns. She complained that the concerns of Dalits in particular are ignored by others, mainly because they are considered a different group and also "because they live in the village periphery". Many villagers had noted that PCC members hired local workers through their personal and party networks, and that the members often preferred hiring workers from other areas since they could charge them a commission.
Thus, villagers engaged in intense negotiations with the company. They demanded remedial action for avoidable harms, compensation for unavoidable losses, and a share in project benefits. The villagers' assertiveness led an NGO activist to comment "there is a perception that in hydro-projects people can make any demands". At the same time, people paid surprisingly little attention to the project's wider environmental, social and cultural impacts, due to concerns with immediate impacts people experienced in their daily lives, such as dust. Nor did villagers' demands extend to future revenue-sharing or the payment of royalties for the use of sand and other natural resources, stipulated by Nepal's regulation as major means for socially just hydropower development. 9 Villagers did not know of these regulatory stipulations because there was no one who told them, the company did not make the EIA report public and the required visits by Ministry of Environment inspectors never happened.
During these negotiations, villagers generally came to accept the parameters set by the company for negotiations over justice, just as they assented to the view that the hydropower project brought bikas to their valley. As a woman explained, "the people are not in favour of closing the project but in favour of getting the compensation". Similarly, an environmental officer commented that any obstructive activities undertaken by villagers, such as temporary road blockages, were "a way of demanding"and not an indication of opposition to the project or the company. "People have the idea that if a strike is done, they can get compensation". Next to jobs, compensation became the currency by which villagers and the company operationalised and measured a notion of justice squarely centred on distributive matters. This implicit consensus found illustration in villagers' reactions to the death of a child who died crossing a creek that had been diverted by the project: they demanded compensation for the parents and the construction of a pedestrian bridge.

Brokering justice in a hydropower project
These everyday negotiations set the scene for Shyam's efforts to relate the demands of indigenous activists to villagers who were mostly indigenous peoples themselves. Stressing the significance of ILO 169 and UNDRIP, Shyam demanded villagers' collective consultation following the principle of FPIC, thereby forwarding NEFIN's recognition agenda that was quite different from villagers' primary concern with compensation. At the same time, Shyam's demand for the application of FPIC was also justifiable under Nepal's EIA regulations. Those required a public hearing on the EIA report and active solicitation of villagers' feedback. In contrast, when the Nepali company conducting the EIA organised a public hearing in Bargachhi in June 2005, which was attended by some 200 people, it distributed the report's executive summary only, and there was no way that villagers could grasp the significance of the proposed project within a couple of hours. Nor were they asked to declare their consent to the proposed project.
Shyam's first action was to report the violation of the FPIC principle to Nepal's Human Rights Commission. He filed a complaint with the Commission as a representative of the Gurung and Tamang ethnic groups. The form stated that the project was "carrying on infrastructural construction […] without informing local people and thereby violated the right to FPIC of the local Gurung and Tamang indigenous communities". Shyam also reported that the project affected indigenous peoples' cultural rights because it "started construction works in the cremation area". In addition, Shyam notified international indigenous rights activists in Asia and Africa about the case.
Shyam was aware of the higher stakes involved in his demands. On the one hand, the demand for FPIC applied not only to the specific hydropower project but had wider relevance to hydropower development and political representation in the district. Noting that the district "has been extraordinary in licensing hydropower", Sham alluded to the various projects being implemented or planned; these projects raised important issues during their construction and heralded the transfer of significant budgetary resources to the DDC in the form of royalties. 10 On the other hand, Shyam denied that the established political parties had any legitimacy in representing indigenous peoples because "political parties represent the state", whereas "we in NEFIN represent indigenous peoples".

Shyam's challenge
Shyam encountered significant challenges in his campaign for FPIC. A critical challenge was personal: how Shyam saw himself, and how he liked to be seen. Shyam had been a member of NEFIN for many years and had played a critical role in establishing its coordination council in the district. He was known as an outspoken advocate for indigenous peoples' rights and had been active in training other NEFIN staff beyond the district. Shyam was motivated to find new fields in which to exercise leadership, and identified the hydropower project and hydropower more generally as such. As a self-promoting civil society representative, he had been trying to insert NEFIN into any conflicts between the project-affected communities and the company. He had voiced the demand to initiate FPIC consultations already two years ago, given two training sessions on FPIC, ILO 169 and UNDRIP and sent a written request to the company a year ago. Shyam felt deeply frustrated that the company "did not listen to this".
Other challenges encountered by Shyam were more of a political nature. He was both the chairman of the NEFIN district council and an active participant in district party politics. He was nearing the end of his term as NEFIN chairman, and witnessed attempts by other members to undermine his leadership. In district politics, he had recently joined the newly formed Federal Socialist Party, which had the recognition of indigenous rights on its agenda. His new party competed directly for the left voter base with the established United Marxists and Leninists party, which already claimed to represent indigenous peoples' interests with regard to hydropower and had some influence on the national NEFIN leadership. Shyam's bid for a seat in the Constituent Assembly elections was unsuccessful, impacting negatively on his standing in district party politics. All these events left Shyam in a precarious position both within NEFIN and district politics.
In addition, Shyam knew that he would encounter serious opposition from most local leaders in Bargachhi. The project had intelligently enrolled many local party leaders in the PCC, in line with the entrenched distributional coalitions common in rural Nepal. The leaders of the established political party that competed with his new party controlled the PCC through Ram, the influential school teacher and party member living in Bargachhi. As one member commented, "whatever the leader says, that will happen". Another member said that in Committee meetings, "the voice of [ordinary] PCC members [was] not heard". Another member voiced her suspicion that she was invited to join the 13-member secretariat only because the regulations required female representation. Although the PCC included members of various ethnic groups, NEFIN was not invited to send a representative. In Shyam's view, the PCC was "a product of biased divisions between political parties, and not really representing the affected population".
In addition, Shyam was convinced that the project was "clever enough to buy the leaders" by allowing them to derive personal benefits from their involvement. Several PCC members seized the opportunity to influence labour recruitment for their own political benefits. Ram managed to hire out some 10-15 dozers, vehicles and tractors to the company, although he did not own any of them. As Shyam found out, Ram sourced the vehicles from their owners and levied a surcharge of 50 per cent on the rental fees. Ram also rented part of his land for the establishment of a military post, charging the company a fee that was more than double the amount paid to another villager and that far exceeded the typical rental price for local agricultural land.
Thus, Shyam understood well that it would be hard to generate local support for his demand. He reacted by pointing out that "we don't want to stop the project" but just get the project to respect villagers' rights. He also made it clear that his demand would not only benefit indigenous peoples but take account of "local people". Furthermore, he extended his demand for FPIC to address the villagers' distributive concerns by calling for "direct payments to local people".
Shyam responded to the situation by adapting NEFIN's self-determination agenda so that it incorporated villagers' distributive demands. He also started to engage in specific activities to support villagers' demands and, in turn, to enrol them in his campaign, as we show in the following.

The labour strike
Shyam made common cause with a labour union leader who mobilised company workers for a strike against their employment conditions. Shyam expressed NEFIN's support for the strikers' grievances, added NEFIN's demands for FPIC and adherence to ILO 169 to the list of demands and provided practical support to the strike organisation. On 24 February 2014, virtually all local workers refused to work further until the company met their requests. They gathered at the project site together with many other villagers and sent a delegation to the district centre to meet the DDC. The strike brought all construction to a halt and caused the company to send home the workers it had hired from other parts of Nepal. In the following nine days, local workers and other villagers continued to hold meetings in the villages. The strike was eventually resolved through mediation of district-level government officials, politicians and the PCC when the company promised a small salary increase and improved benefits.
When we talked to some of the key players barely a month later, they agreed that workers, villagers, labour union and NEFIN had made common cause for the strike. However, their accounts of the strike's aims and how it unfolded varied widely. To Shyam, NEFIN was the main organiser of the mass gathering on the first day, which included 1,000 participants: "all the workers, local people, indigenous people and mothers' groups". Shyam emphasised the extension of the strike to the district centre, and that he had caused Nepal's Human Rights Commission to send representatives to the construction site and meet with the company and affected villagers.
Other accounts agreed that the strike had involved workers and many other villagers, and that the demands included improved labour conditions and the consultation of affected people. However, they attributed a much smaller role to NEFIN and indigenous activists' demands than claimed by Shyam, and challenged his account of a big rally in the district centre. The delegation sent to meet the DDC was a small one, as most strikers stayed in Bargachhi. This was not a minor detail because it reflected the relative significance of the labour demands (directed at the company in Bargachhi) and the consultation demand (addressed to the DDC).
The events and their recollection illustrate Shyam's precarious position between NEFIN's agenda and local politics. Shyam tried hard to connect NEFIN's key demand for the consultation of affected indigenous peoples with villagers' claims. However, when the delegation sent from Bargachhi met the DDC, Shyam was not invited to the table. Similarly, Shyam was dissatisfied with the resolution of the strike since it met none of his demands. "The outcome was not just", he concluded. It was not just within the justice agenda promoted by NEFIN, that is, the focus on indigenous peoples' recognition and the right to FPIC. In contrast, local political leaders, some of whom were also in the PCC, had not supported the strikers' cause but instead pressured the workers to accept the settlement. As a result, the company had to make only minor concessions because it "bought the loyalty of some local leaders" and because it succeeded in dividing the labour union and NEFIN.

Resettlement
Another issue that came to Shyam's attention was the standoff between the company and the village residents on resettlement. The EIA had deemed it necessary to relocate some of the village households temporarily to avoid harm from the construction works. It had classified the 24 village households into three categories according to their exposure to potential harms, proposing that the 11 considered at high risk would be relocated for 15 months against compensation of NPR 500,000 each. The other 13 households would not have to move and receive compensation of NPR 10,000-100,000. The villagers were deeply dissatisfied with the proposal and countered with the demand that every household in the village should receive compensation of NPR 4,500,000.
Shyam saw this standoff as an opportunity for NEFIN to support villagers and simultaneously establish NEFIN as a player in the negotiations between villagers and the company. Shyam mentioned the dispute in the report to the Human Rights Commission, depicting villagers' demands as a request by ethnic Gurung to "guarantee adherence to international law on human rights including ILO-169". He also included the demand for equal and fair compensation in the agenda of the labour strike. He stressed villagers' demand for equal compensation to all, de-emphasising the company's focus on the material impacts of relocation, and instead arguing that villagers were "fearful that their culture will be disrupted because of the relocation".
The first few months of 2014 witnessed intense bargaining over the terms of relocation between the villagers and company. The company's response to villagers' request for collective compensation was very minor increases in the compensation amounts, which infuriated villagers. They in turn wrote to the company demanding payment of NPR 1,846,000 and the guarantee of employment for two people per household. To lend force to their demand, villagers started to block the construction works near the planned tunnel. They continued the blockade for six days, engaging in various displays such as banging plates together or holding up placards.
The villagers and the company eventually came to an agreement regarding the relocation. Ram and other leaders from the PCC played an important mediating role. The company raised the compensation amounts for all households. The households to be relocated were promised NPR 1,000,000, and other households were to receive between NPR 100,000 and NPR 250,000. In addition, Ram invited two people from the village to join the PCC, one as member of the Secretariat, the otherfemaleas a regular member. The company did not make any concession on employment, except the general promise to give priority to skilled workers from the village when hiring.
The agreement effectively divided the villagers among each other. The 11 households that received NPR 1,000,000 appeared pleased in our conversations with them but clearly did not want to reveal that in light of other villagers' frustrations over the agreement. They emphasised that they had to accept the settlement, even though it was not good for them, because otherwise they were in danger of not getting any compensation at all. The other households remained concerned that they may suffer from the detonations. Pointing out that some households were relocated whereas their neighbours were not, a Dalit woman asked how it was possible that "two households so close to each other [fell] under two different categories".
The labour strike and the resettlement dispute reveal several problems limiting Shyam's ability to enrol villagers in his campaign. One was his relatively precarious personal standing and the marginal position of his organisation, NEFIN, in district politics; another was the company's successful recruitment of local leaders into a distributional coalition, as practiced elsewhere in Nepal (Rai 2005;Sharrock 2013). The NEFIN agenda itself restricted the discursive repertoire available to Shyam: its focus on indigenous self-determination and on FPIC limited the appeal of the NEFIN campaign to villagers, with their interest in distributive issues. Although Shyam adapted the NEFIN agenda to villagers' demands, the distributive conception of justice specified in Nepal's regulations and implemented by the company remained more attractive to both villagers and leaders.

Conclusion: brokering justice
Our insights highlight the brokerage operating at the intersection between different notions of justice in specific political contexts. The conceptions specified in global norms, enshrined in national policy and held by villagers are likely to differ from each other. In our study site, the critical difference was between NEFIN's recognition demands in congruence with global indigenous rights and the distributive notions established in villagers' everyday negotiations with the company in line with Nepali hydropower policy. In other socio-environmental struggles in Nepal or worldwide, differences may take other forms, yet asserted notions of justice are more likely to be plural than converge on a single meaning. The presence of multiple notions, therefore, opens up and simultaneously constrains the political grounds for brokerage. Brokerage helps to constitute certain notions of justice as relevant, dominant or even hegemonic by relating them to specific socio-environmental struggles and aligning involved actors. Yet it also helps to keep notions of justice subject to ongoing political contestation and re-constitution, including those originating from global norms.
In our account, we have highlighted the significance of brokerage and practices of brokering justice, which is different from speaking of "brokers of justice". Brokerage does operate through the practices of particular intermediaries who assert, adapt and subvert plural notions of justice. Nevertheless, these intermediaries are not simply "brokers" since brokerage is typically only part of what they do (Mosse and Lewis 2006). In our case, Shyam also participates in party politics, personally cares about indigenous peoples' rights and engages in other activities relevant to his campaign. The loss of standing in district party politics that he suffered after losing the elections also affected his efforts with regard to the hydropower project. Moreover, justice is only one field of brokerage, which concurrently forges relationships and generates shared framings in other fields. Shyam's efforts to convince villagers were not only about how justice should be defined and realised with regard to the hydropower project, but also involve other political manoeuvres, such as stories about local leaders' corrupt dealings, claims about cultural significance, practical organisational support and the use of his personal political networks.
This embeddedness of justice politics in wider political dynamics raises the question of how notions of justice enable and confine practices of brokerage by affecting the positions of involved intermediaries. We suggest that intermediaries derive advantages from their linkages to multiple notions of justice for the discursive and material resources associated with them. NEFIN's agenda provides Shyam with a powerful justificatory framing and access to mobilisational resources. In other instances, references to justice have been found to bolster claims for state support, the state remaining the key guarantor of rights and most critical source of finance (Wolford 2003). Appeals to justice afford privileged access to external support and knowledge through contacts to activists in other locations, countries or at the global level (Lindell 2009). Additionally, notions of justice afford the intermediaries means to make sense of messy social realities and to define as what kind of person they want to be recognised (Mosse 2005).
At the same time, brokering justice is also a source of vulnerability to intermediaries. As Shyam's efforts demonstrate vividly, intermediaries have to persuade people with grievances to accept certain notions of justice and need to convince states, companies, donors and other actors to supply financial and political support in response to such demands. Notions enshrined in global norms, invoked by higher-level mobilisations or imported from other places may not fit the notions, historical experiences and cognitive frameworks of people involved in particular socio-environmental struggles (Wolford 2003;Upton 2014;Merry 2006b). Worse, hegemonic notions of justice may limit the repertoires of ideas and practices available to intermediaries in particular contexts, as highlighted by our case.
Brokerage sheds new light on the importance of global norms in providing discursive support to marginalised people and giving them access to material support. Global conceptualisations of justice may effectively bolster site-specific struggles, as has been observed with anti-dam protests in Thailand and Mozambique (Sneddon and Fox 2008). Global norms may even do so in cases where their justice conceptions do not fit neatly with the notions asserted in site-specific demands, as mobilisations for justice have successfully accommodated discrepancies (Upton 2014). Conversely, in particular contexts, global norms may work to restrict the discursive options available to marginalised people in specific socio-environmental struggles. In other cases, multiple vernacularisations can result in competing understandings of rights and justice, which may differ greatly from their intended transnational meanings and values (Goldstein 2015). Global ideas about justice may even displace alternative visions of social justice due to "the vastly unequal distribution of power and resources that channels how ideas develop in global settings and are picked up or rejected in local places" (Merry 2006a, 4).
Our findings highlight some of the vexing issues encountered in indigenous peoples' mobilisations and the implementation of global indigenous rights. On the one hand, the focus on global and national mobilisations and internal power relations may limit their responsiveness to indigenous peoples' specific demands in particular settings, such as the distributive demands observed in our study. On the other, the emphasis on indigenous peoples may work to discriminate against other marginalised groups, such as the Dalits in our study. The relative success of indigenous peoples' global advocacy, therefore, challenges activists to "indigenise" their demands for justice, as Shyam has attempted in Bargachhi. If indigenous rights activists fail to do this, their mobilisations may fail to gain traction or may even become a source of injustice. Notes 1. Personal and place names in this article are pseudonyms, to protect the anonymity of our informants. 2. Policy provisions and practices for hydropower project benefit sharing in Nepal include: a royalty mechanism, local share offers, support for local livelihoods (such as employment and training), investment in community development and local infrastructure, and allocated budget for environmental enhancement activities (Shrestha et al. 2016;Lord 2016). 3. Available at: http://mowss.gov.np/assets/uploads/files/Environment_Protection_Rules.pdf 4. After years of conflicts arising from bad governance, underdevelopment, widening inequality and discrimination, Nepal has undergone a number of political changes since 2006. These include: the peace agreement between the Maoists and the government (2006); declaration of a republic following the end of the Nepalese monarchy (2008); Constituent Assembly elections (2008 and 2013); promulgation of the new constitution by the Constituent Assembly (2015); and local, state and national elections under a federal governance set-up (2017 and 2018). 5. The consensus today among major political actors is that Nepal cannot become a modern nation-state without acknowledging and addressing layers of exclusionary practices based on class, caste, ethnicity, region and gender. These sentiments are broadly documented in Nepal's 2015 constitution (see, for example, Article 18). 6. See Articles 7.1 and 7.3 of ILO Convention 169 (1989) and Articles 1-4 and 25-30 of UNDRIP (2007). ILO 169 is legally binding, whereas UNDRIP is not. 7. Local elections were held in Nepal in 2017 after a gap of almost 20 years. These were the first to be held after the promulgation of 2015 constitution. 8. With this, the company met the regulatory requirement that hydropower developers spend one per cent of the construction budget on the provision of compensation and benefits to the local population. 9. On the latter, the company had already agreed with the DDC on the payment of NPR 2.2 million, of which 35 per cent would have to be transferred to Bargachhi VDC according to Nepali law. 10. Just for this one project, we estimate that the royalties expected by the DDC are significant, contributing an additional 5 per cent to the district budget in the first 15 years and a staggering 75 per cent thereafter.