Nonjudicial business regulation and community access to remedy

Redress for communities harmed by transnational business activity remains elusive. This paper examines community efforts to access redress for human rights-related harms via recourse to transnational nonjudicial mechanisms (NJMs) – a prevalent but widely debated instrument of transnational business regulation. Drawing together insights from theoretical debates surrounding nonjudicial regulation and evidence from a major empirical study of human rights redress claims in Indonesia and India, the paper explores the conditions under which NJMs can support community access to remedy. Three conditions are shown to be central in enabling some degree of NJM effectiveness: the institutional design of regulatory strategies, the institutional empowerment of regulatory institutions, and social empowerment of affected communities and their supporters. While all three conditions are required in some measure to underpin effective NJM interventions, these conditions can be combined in varying ways in different contexts to underpin either top – down or bottom – up pathways to redress. The former derives its primary in ﬂ uence from institutional authority and capacity, while the latter relies more heavily on diffuse societal leverage in support of community claims. These ﬁ ndings have signi ﬁ cant implications for theoretical debates about the capacity and limits of nonjudicial regulatory approaches to support human rights redress within decentered contexts of transnational regulation where both regulatory power and agency are widely diffused.


Introduction
An intense spotlight has been placed in recent years on an array of human rights violations associated with transnational business activity, including violations of labor rights, threats to community livelihoods or access to land, and compromised health and safety of workers or communities. Among the multiple governmental and nonstate governance initiatives that have emerged in response, one of the most significant has been the formulation and promotion of the United Nations Guiding Principles (UNGPs), which were endorsed by the UN Human Rights Council in June 2011 (Ruggie 2011, p. 22). The UNGPs are organized around three pillars, encompassing responsibilities of businesses to respect human rights, of states to protect human rights, and of all parties to facilitate access to remedy for those affected adversely by transnational business activity. During the process of developing the UNGP framework, significant emphasis was placed on the potential of nonjudicial alongside judicial mechanisms of human rights remedy, stimulating renewed interest in exploring the distinctive capacities and limitations of nonjudicial approaches in supporting redress for human rights grievances (Ruggie 2011).
Nonjudicial mechanisms (hereafter NJMs) designed to support human rights remedy can take widely varying forms, including independent accountability mechanisms of international financial institutions (Scheltema 2013;Altholz & Sullivan 2017); mechanisms established by home country governments to receive complaints about the conduct of companies domiciled in their jurisdiction, such as National Contact Points (NCP) created under the framework of the OECD Guidelines on Multinational Enterprises; and grievance mechanisms associated with multistakeholder standard-setting bodies such as the Ethical Trading Initiative or Roundtable on Sustainable Although the UNGPs have helped to generate previously unprecedented levels of interest in the potential and limitations of nonjudicial regulatory approaches, their effectiveness has been repeatedly questioned. Advocates of nonjudicial regulation have argued that NJMs can have "a legitimate and compelling role to play" in addressing human rights-related disputes between companies and affected people (Rees 2010, p. 22). Such analysts have emphasized their potential to "complement and support" broader forms of state-based regulation (Lukas et al. 2016, p. 326), particularly in handling rights claims involving qualified and competing rights, or where there is room for contextual interpretation and varied understandings of remedy (Rees 2010). Others have expressed significant skepticism of such claimsinstead advocating strengthened legislative or administrative mechanisms backed directly by state authority (Altholz & Sullivan 2017). Building on broader critiques of alternative dispute resolution (Cedillo 2013), critics have highlighted weaknesses of NJMs in countering power imbalances in bargaining processes (Fiss 1984;Coumans 2012), their frequent failure to protect rights or broader conceptions of justice, and their tendency to individualize recurring grievances, inhibiting promotion of systemic change (Edwards 1985;Capulong 2012). Recent Non Governmental Organisation (NGO) reports have reinforced such concerns, finding shortcomings in the performance of nonjudicial processes to be pervasive (e.g. Accountability Counsel 2016; Daniel et al. 2015). 1 Against the backdrop of these persistent and often polarizing debates, the central aim of this paper is to explore under what conditions (if any) NJMs can make useful contributions to empowering community access to remedy. Such contributions may include support for individual redress claims (e.g. by facilitating mediation of company-community disputes) and/or support for broader processes of learning and systemic change (e.g. through processes of investigation and audit that inform ongoing regulatory reform). While NJMs can be constituted at national or transnational scales, our focus is on the use of NJMs as instruments of transnational business regulation.
Analysis draws on an extensive field-based empirical study, which investigated 10 case studies of community grievance claims associated with transnational business activity across a range of sectoral and jurisdictional contexts. By examining a range of case studies, our aim was to explore the dynamics of causal complexity within cases while exploring varying configurations of enabling conditions for success (and failure) across diverse contexts. Such exploration of varying causal "recipes" (Rihoux & Ragin 2008) or pathways would not have been possible through within-case analysis alone. Over the five years of our research project, we carried out 587 interviews with 1,100 individuals across our 10 cases. Interviews and field visits were supplemented by secondary research, including online media articles, civil society reports, and company websites.
Each case focused on a particular dispute (or connected series of disputes) between a specific company and workers or communities affected by the company's activities. Although each case is thus anchored in an individual grievance, its spatial and temporal boundaries were broader. Each case was spatially constituted at the multiple subnational, national, and transnational scales over which the grievance was contested and was temporally constituted around the dynamics of the particular grievance. The spatial and temporal boundaries of each case were fluiddriven by changing patterns of claim-making and contestation and often evolving to encompass an interacting and contested array of individual and systemic claims.
In selecting cases of individual grievances, we considered a number of practical and theoretical criteria. First, we identified individual cases of grievances in which communities had attempted to engage one or more transnational NJMs to support their redress claims. From this list, we purposively selected 10 grievance cases with the aim of generating various NJM institutional designs, economic sectors, and geographical locations. To ensure appropriate variation in NJM institutional design, our cases were selected to include cases of five transnational NJMs, covering each of the most widely used categories of NJMs (linked to international financial institutions, multistakeholder standard-setting bodies, and home country governments, respectively). 2 To ensure appropriate variation in the market and political environments into which NJMs intervened, cases were selected to cover three different sectors (agribusiness/land-based sectors, mining, and garments and footwear), each of which is characterized by highly globalized production and trading systems, widely recognized human rights risks, and active presence of both state and civil society actors engaged in contesting regulatory processes. Cases also focused on grievances of communities in two countries: India and Indonesia, both of which were linked to significant transnational business activity in the three focal sectors. 3 In what follows, we first draw on theoretical insights from a diverse body of regulatory scholarship debating the potential and limits of nonjudicial regulation to develop a framework for analyzing the enabling conditions for constructive NJM contributions to community remedy. This framework focuses on three factors that we expect to be associated with some degree of NJM effectiveness: the institutional design of regulatory strategies, the institutional empowerment of regulatory institutions, and social empowerment of affected communities and their supporters. Second, by examining patterns across all 10 cases, we show how these enabling conditions can combine in varying ways to support NJM contributions to community redress. This analysis highlights a particular contrast between top-down and bottom-up pathways to redress: the former derives its primary influence from institutional authority and capacity, while the latter relies more heavily on diffuse societal leverage in support of community claims. Third, we present a more detailed within-case analysis of the two cases that most clearly illustrate the dynamics of top-down versus bottom-up pathways to remedy.
While the conditions required to enable NJMs to effectively support community remedy are shown to be demanding, our findings provide some space for modest optimismdemonstrating how NJMs can, under conducive social and institutional conditions, make productive contributions to community human rights redress. By highlighting the interplay between institutional and social power in constituting these enabling conditions, our analysis responds to concerns of critics of nonjudicial regulation who seek to defend normative concerns for rights in the presence of deeply entrenched power imbalances. Yet our analysis cautions against excessive emphasis on institutional design and regulatory strategy as a basis for NJM effectiveness. Instead, we highlight the broader sources of social power in which nonjudicial institutions are embedded and the corresponding importance of diffuse sources of regulatory power and agency in supporting access to remedy for marginalized workers and communities affected by transnational business activity.

Enabling conditions for NJM effectiveness: From institutional to societal empowerment of community claims
Nonjudicial approaches to regulation have been theorized and debated across a range of regulatory scholarship, encompassing theories of "pragmatic" or "problem-solving" regulation (Simon 2004), "experimentalist" governance (De Búrca 2010), private regulation (Abbott & Snidal 2009b), and "new governance" (de Sousa Santos & Rodríguez-Garavito 2005;Rodríguez-Garavito 2005;Ford 2010). This diverse scholarship examines the role of nonjudicial approaches within regulatory processes of rule-making and compliance promotion, as well as complaint handling or dispute resolution. 4 Such perspectives share a common interest in regulatory institutions that do not rely exclusively or primarily on the exercise of hierarchical or coercive authority to achieve their desired effects. 5 Our goal here is to draw out specific strands of this work that have focused on identifying the enabling or scope conditions under which nonjudicial regulatory approaches are most likely to be effective: in the words of Ford (2010, p. 442), the "essential preconditions to making [such regulatory approaches] work." Below, we identify three key clusters of enabling conditions that we expect to support the effectiveness of nonjudicial regulatory interventions, relating to: institutional design, institutional empowerment, and social empowerment.
2.1. Institutional design of nonjudicial systems: Problem solving and accountability Many advocates of nonjudicial approaches have highlighted institutional design as a central determinant of contested regulatory outcomes. This orientation is strongly associated with "new governance" literature, which focuses on exploring "the significance of institutional design and culture for effective and legitimate regulation" (Lobel 2012, p. 65). Such an orientation emphasizes the agency and strategic choices of regulators, identifying institutional procedures and governance arrangements adopted by regulators as key determinants of contested regulatory outcomes. Such scholarship explores numerous dimensions of institutional design. To aid our analysis, we have broken these dimensions down to distinguish between institutional design of regulatory strategy and institutional empowerment. Here, we focus on two specific approaches to designing regulatory strategy that have been argued to influence the effectiveness of nonjudicial regulation: we characterize these as institutional orientations toward problem solving and accountability.
Scholarship on nonjudicial regulation has often emphasized the value of institutional designs oriented toward collaborative, problem-solving approaches to regulation, resting heavily on consensual processes of negotiation (De Búrca 2010;Schouten et al. 2012). Assessments of the potential and limits of nonjudicial redress systems have similarly focused on the qualities of consensual mediation processes as determinants of NJM effectiveness. An institutional design focus is evident in the formulation of the Effectiveness Criteria presented as part of the United Nations Guiding Principles on business and human rights 6 and in the assessments of specific NJMs (Wallace 2017). 7 Such problem-solving approaches have often been contrasted with rights-based accountability approaches to institutional design, in which rights, responsibilities, and associated remedies are viewed as authoritatively determined in accordance with recognized normative standards (NeJaime 2009). Such approaches have commonly been endorsed as appropriate when claims or disputes are not amenable to consensual deliberative problem solving, whether because of entrenched conflicts of identity or interest or insurmountable disparities of power between parties (Simon 2004;Cohen 2008;NeJaime 2009).
Accountability approaches have been presented primarily as alternatives to nonjudicial approachesnecessary to address power imbalances through more law or command and control regulation (Simon 2004;Solomon 2008). Debates concerning the relationship between nonjudicial and legal regulation have sometimes distracted attention from the significant role that accountability-oriented mechanisms can also play within the design of nonjudicial institutions. In fact, there is significant scope for NJMs to incorporate mechanisms of evaluation and accountability for rights-based standards despite lacking powers of binding adjudication. For example, NJMs sometimes have specific powers and responsibilities concerning fact finding, standards-based auditing, and/or determinations regarding violations of agreed standards (Sabel & Simon 2004).

Institutional empowerment: Authority and capacity
Also of potential significance in enabling NJM effectiveness are the authority and capacity that regulatory institutions draw on in support of their efforts to facilitate community redress: we refer to these as dimensions of institutional empowerment. 8 One important dimension of institutional empowerment is the degree of authority that an NJM is able to establish over the parties to a given grievance. We use the term authority here to refer to a relation in which regulated parties defer to or comply with authoritative demands at least in part because of recognition of the authority's right to issue such demands (Hall & Biersteker 2002;Krisch 2017). The authority of an NJM is potentially important in giving disputing parties reasons to cooperate with processes such as fact finding or audit and to accept and respect any negotiated or determined outcome, even where they may disagree with the substance of the decision.
NJMs can establish authority through social, political, or rhetorical practices that do not depend on formal backing from legal or state-based institutions (Hall & Biersteker 2002;Emerson et al. 2011). 9 Technical, cultural, or political expertise, or other relevant professional knowledge and competencies possessed by NJM staff or contractors, can provide important sources of authority. Recognition of authority may also be grounded in the NJM's compliance with procedural standards, such as transparency or publicity, or may derive from the prior authority of a multistakeholder initiative or international financial institution in which the NJM is embedded. In practice, a company's willingness to comply with processes or findings of an NJM is likely to depend on interactions between these various sources of NJM authority and more instrumental reasons for corporate engagement associated with wider market and social pressure.
A further dimension of institutional empowerment that we expect to influence an NJM's contribution to community redress relates to the institutional capacities of the NJM itselfby which we mean those resources and functions that the NJM commands independent of its interactions with parties in a given grievance case. These may include financial and institutional resources, alongside human capacities embodied in the skills and experience of NJM staff and contractors (e.g. see Simon 2004, p. 177;Lukas et al. 2016). The specific capacities required for NJM effectiveness vary with the mix of problem solving and accountability strategies adopted. For example, problemsolving approaches demand strong mediator capacities to engender awareness among parties of their "interlocking interests" (Lobel 2004, p. 379) and to encourage balanced mediation through support for weaker parties (Anderson 2003;Menkel-Meadow 2008). Accountability approaches require capacities to conduct independent investigations, audits, or monitoring and to establish perceptions of independence and expertise supportive of adjudicative and investigatory functions (self-reference). In either scenario, capacities to facilitate community redress often require strong local contextual knowledge and relationships, together with sufficient financial and organizational resources to support active community outreach and capacity building where required (Edmunds & Wollenberg 2001;Lukas et al. 2016, pp. 244-246).
2.3. Social empowerment: Community leverage and control over defining remedy Despite widespread emphasis on institutional design and empowerment as determinants of regulatory effectiveness, some have questioned the apparent assumption that "if only the institutional designs can be constructed just right, then gross imbalances of power in the contexts of these institutions will be neutralized" (Fung 2003, p. 259). Instead, such accounts have highlighted the role of broader "configurations of non-deliberative power" in enabling and constraining nonjudicial regulatory processes (Fung & Wright 2003, p. 23), emphasizing the potential for diffuse social actors to mobilize sources of "countervailing power" as a basis for bolstering the effects of regulatory interventions (Rodríguez-Garavito 2005, p. 210;Fung 2003;Melish & Meidinger 2011;Maru 2010).
We refer to such dynamics as processes of social empowerment and focus on two key elements: community leverage and community control over defining the terms of appropriate remedy.
We use the concept of "leverage" to refer to an actor's ability to influence the imposition of costs or withholding of advantages from those they seek to influence. In the context of our cases, such influence relates to specific bargaining, mediation, or investigation processes involving companies. Leverage can importantly underpin both problem solving and accountability-based grievance-handling strategies. Within mediation processes, leverage provides incentives for actors to engage with NJM processes and compromise within processes of interest-based bargaining (Fung & Wright 2001;Simon 2004;Kovick & Rees 2011). Within accountability processes, leverage provides a basis for nonlegal sanctioning or incentivization of actors to comply with agreed rights-based standards (Locke et al. 2009;Simon 2004).
Many communities and workers occupy a position of significant structural weakness, deriving from multidimensional sources of socio-economic marginalization and dependence on particular companies for their incomes or livelihoods (Van Rooij 2010). The capacity of communities and their supporters to access grievancehandling systems in the first place (Anderson 2003) and then to exercise direct leverage over companies or NJMs in support of their grievance claims is thus often very limited, depending on their own collective organization and mobilization capacities and support from local and transnational networks of unions, NGOs or consumer associations, or allied politicians or government officials ( Community leverage may be augmented by support from nonjudicial institutions, when these institutions in turn can establish sufficient leverage to exercise meaningful influence over the companies concerned. While nonjudicial regulatory leverage has often been viewed as reliant, at least in part, on backing from law or other forms of state authority (NeJaime 2009), 10 NJMs may also build leverage by harnessing and weaving together multiple sources of social pressure, including both social and organizational pressure from communities and their supporters, and economic pressures originating in market incentives (Newell 2005;Melish & Meidinger 2011).
Another important dimension of social empowerment is a community's capacity to define the nature of the grievance and associated terms of redress. A grievance may be framed in widely divergent ways: as a human rights violation, an instance of regulatory noncompliance, a symptom of structural regulatory failure, or simply a distributive dispute between private parties. Such framing processes can wield significant influence over the forms of redress that are viewed as appropriatewhether involving apologies, restitution, compensation, punitive sanctions designed to provide redress for specific grievances, or systemic interventions designed to prevent recurring patterns of grievance through wider agendas of regulatory or policy change. Given such a diversity of (often contested) possibilities, community control over framing an appropriate remedy can importantly determine the extent to which they can "defin [e] what the realisation of their human rights or remediation for harms caused should mean in practice" (Rees 2010, p. 7).
A community's influence over processes of problem framing is likely to depend in part on its capacity to establish clear shared objectives regarding desired redress outcomes and to deliberate over what costs and risks they are willing to tolerate. This in turn depends on information about available redress options, together with leadership and collective decisionmaking capacities within communities, underpinning a community's ability to manage internal divisions and its organizational resilience during sustained negotiations or political struggles (Garvey & Newell 2005;Calvano 2007;de Jong Edwin & Nooteboom 2010).

Explaining varying NJM effectiveness: Multiple pathways to remedy
The above analysis generated a rich set of propositions regarding the multiple institutional and societal conditions that are likely to influence NJM effectiveness. To explore how these factors interacted in practice across a range of country and sectoral contexts, first, we briefly compared all 10 cases from our study, exploring variation in relation to key elements of NJM institutional design, institutional empowerment, and social empowerment of communities. We then examined the relationship between configurations of these variables and effects of NJM involvement on community redress. To aid this comparison, we first specified criteria for each of these causal and outcome variables and then assigned a score of 1-5 for each variable on this basis. The criteria used are detailed in the Appendix.
We identified 3 of the 10 cases as reflecting examples of (relatively) successful NJM interventions, understood as those in which they made some meaningful contribution to supporting community remedy. To help us explore which (clusters of) enabling conditions were associated with success (or failure), we created visual maps of the varied conditions across cases to help identify which configurations appeared to be associated with positive NJM contributions to redress within the complex causal processes constituting contested redress claims (Rihoux & Ragin 2008).
With regard to institutional design, this analysis suggested that use of rights-based accountability mechanisms was a necessary but not sufficient condition for useful NJM interventions. Use of problem-solving mechanisms was neither necessary nor sufficient, although in two of the three successful cases, problem-solving approaches interacted productively with accountability approaches to facilitate support for community redress. A moderate degree of both institutional and social empowerment was also necessary to underpin effective NJM interventions, but the relative significance of each varied across the three success cases.
Each case of "successful" NJM intervention was associated with a distinct configuration of causal variables, enabling us to identify multiple causal pathways (or configurations of enabling conditions) that were consistent with productive NJM interventions in supporting community redress. Particularly striking was a contrast between two distinct pathways to redress: a top-down pathway, in which NJM authority and capacity played a lead role in facilitating community redress, and a bottom-up pathway, in which community leverage and influence over the definition of redress were the primary conditions supporting redress. We also demonstrated the potential for a mixed (top-down and bottom-up) strategy, in which all four enabling conditions were present to a moderate degree. Figure 1 presents a visual map of these three configurations of enabling conditions, along with a table summarizing our scoring of each case.
While multiple pathways to redress are thus possiblerelying on varying combinations of enabling conditionsweak enabling conditions in some areas can be compensated by strengths in others only up to a point. In seven of our cases, communities achieved little redress of any kind. While enabling conditions may combine in different ways, patterns evident in Figure 2 (illustrating configurations of enabling conditions in cases of NJM "failure") demonstrate that all are required in some measure for communities to be able to access redress through engagement with an NJM. As Figure 2 illustrates, all cases of failure were assessed at the lowest level (1) against one or more enabling conditions. Even where some enabling conditions were strongly present, a very low rating on any one of the four enabling conditions was enough to undermine NJM effectiveness. This reflects the need for NJMs to draw on plural sources of power to support their effectiveness, in view of their lack of coercive authority, and the heavily overdetermined structural power imbalances between transnational businesses and the workers and communities affected by their activities.
By highlighting the interplay between multiple dimensions of institutional and societal power in generating (or obstructing) these enabling conditions, our analysis emphasizes the dependence of NJM effectiveness not only on institutional design but also on more diffuse processes of social contestation and power struggle. To illustrate in greater depth the complex processes through which plural sources of power can be woven together at local and transnational scales in support of redress, the following section presents a more detailed within-case analysis of our two illustrative cases of top-down and bottom-up pathways to effective NJM interventions. 11

A top-down pathway to remedy: Palm oil in Indonesia
We begin by presenting a detailed analysis of a "top down" pathway to remedy, exploring an interlinked series of community complaints to the World Bank Group's Compliance Advisor Ombudsman (CAO) regarding Indonesian operations of the prominent Singapore-based palm oil company Wilmar. Although redress outcomes for communities in this case were very uneven, this was still a relatively "successful" case, in which some aggrieved communities were able to achieve at least some of their demands as a result of CAO-facilitated mediations, while CAO compliance investigations contributed to reform of World Bank policy concerning the palm oil sector. Despite the significant structural challenges that communities faced in their pursuit of redress, this case illustrates the potential for NJMs to contribute constructively to community redressto the extent that enabling conditions are present.
In this case, some partial successes in delivering a remedy for communities were achieved as a result of the CAO's incorporation of both problem-solving and accountability approaches in its institutional design, backed by its relatively strong institutional authority and capacity. We characterize this as a "top-down" pathway to redress. However, consideration of institutional design and empowerment is not sufficient to explain outcomes in this case. Although community leverage and capacity to define the grievance were both relatively weak, societal mobilization and empowerment interacted with these factors in significant waysharnessing structural leverage to underpin the CAO's influence, generating bargaining power to support community claims within problemsolving mediations, and continuing to promote broader processes of systemic change.
3.1.1. Overview of the case The Indonesian palm oil sector plays a significant role in supporting economic growth and employment creation but has been widely associated with human rights concerns, involving the displacement and dispossession of indigenous people and local communities from their lands and violations of social and economic rights of smallholders and landless laborers (Colchester 2006). While palm oil supply chains have become highly transnationalized, production practices are deeply embedded in local regimes of extraction and intensive land use within Indonesia (McCarthy 2010). Regulatory struggles over human rights redress have played out over shifting local and transnational spaces (Pacheco et al. 2018), with "human rights" grievances, in turn, embedded within wider struggles over agrarian and ecological transformations (Hall et al. 2011;Peluso & Lund 2011;Schoenberger et al. 2017).
Between 2007 and 2011, three formal complaints about the company Wilmar were made by a coalition of subnational, national, and international community organizations and NGOs to a prominent transnational NJM: the CAO of the World Bank Group. The CAO is an independent accountability and grievance-handling body designed primarily to manage grievances associated with business activity funded by the World Bank's private sector lending arms, the International Finance Corporation (IFC) and the Multilateral Investment Guarantee Agency (MIGA). As its name suggests, the CAO incorporates three elements: a Compliance auditor, which assesses the IFC's adherence to its own social and environmental policies; an Advisory arm, which advises the World Bank Group on how IFC/MIGA's social and environmental performance can be improved; and an Ombudsman arm, which provides recourse for people affected by IFC or MIGA projects and facilitates mediation between companies, communities, and other affected parties (Compliance Advisor Ombudsman n.d.(a)).
Each complaint brought to the CAO related to alleged instances of rights violations among particular communities, some in different parts of Indonesia, who called on the CAO to help "restore the rights and livelihoods of the affected people". 12 The complaints also addressed systemic concerns about the company's wider patterns of social and environmental harm, including illegal use of fire to clear lands; clearance of primary forests and areas of high conservation value; takeover of indigenous peoples' customary lands without due process or free, prior, and informed consent; and violence and intimidation against communities protesting against companies. 13 Of the 25 locations named in the complaint against Wilmar, 3 were selected by the CAO for individual mediations: in West Kalimantan, Riau, and Jambi provinces. 14 The success of these mediations in achieving negotiated settlements varied significantly across sites. In Kalimantan and Riau, mediation produced settlements that returned some land to original inhabitants or restored livelihoods and provided some financial compensation (selfreference). In Jambi, despite the best efforts of largely well-respected mediators, the mediation processes failed to deliver an agreed outcome.
The systemic dimensions of the complaints were addressed to some extent through the CAO's compliance audit process: an audit report released in 2009 found the IFC had not complied with the IFC Performance Standards in managing its lending relationship with Wilmar. Associated remediation led to a moratorium on palm oil investment in the entire World Bank Group, the development in 2011 of a new World Bank Group strategy for investment in palm oil, changes to the IFC procedure for processing all single-commodity trade finance, and establishment of a new Advisory Services program in Indonesia to support social and environmental improvements in the sector (CAO 2013). These reflected some significant policy changes within the World Bank Group but contributed little to systemic change within Indonesia.

Institutional design
We first consider to what extent a focus on the CAO's institutional design helps explain the mixed but generally positive outcomes in this casepaying particular attention to the significance of problem solving versus accountability strategies.
The CAO's grievance-handling processes incorporate elements of both problem solving and accountabilityalthough with greater reliance on the former. The CAO's Ombudsman mechanism focuses heavily on consensual negotiation. It does not adjudicate the merits of any individual complaint, nor does it impose solutions; rather, it provides dispute resolution specialists who work with disputing parties to "identify and implement their own solutions" (CAO n.d.(a)). Some capacity building is provided to parties to enhance mediation outcomes. Problem-solving approaches targeting systemic change are also explicitly incorporated through the CAO's Advisory function, which is designed to disseminate learning about cases the CAO is involved in within the World Bank Group (CAO 2015, p. 40).
Accountability mechanisms are incorporated in the CAO's design to some degree via rights-protective standards in the IFC Performance Standards, which lay out detailed expectations regarding social and environmental risk management by IFC/MIGA clients (International Finance Corporation n.d.). While companies financed through IFC loans are expected, as a condition of their loans, to comply with the Performance Standards, the CAO is not mandated to conduct compliance audits of companies unless this is agreed to by all parties as part of a dispute resolution process. Rather, compliance audits focus on the IFC and MIGA themselvesseeking to ensure that IFC and MIGA management has carried out appropriate due diligence in their lending relationships. The CAO can recommend remedial or corrective actions but has no powers to enforce compliance. The compliance process can also support accountability for systemic rights violations to the extent that external actors such as NGOs can pressure the World Bank Group to reform its lending and due diligence policies in ways that better support codified standards.
The interplay between problem solving and accountability mechanisms played a crucial role in facilitating (and limiting) the CAO's capacity to support redress for grievances of individual communities. While problem-solving processes helped secure relatively successful negotiated agreements in both West Kalimantan and Riau, the absence of protections for minimum rights-based standards as a basis for negotiations meant that agreements resulted from compromises between communities' claims to rights to land and livelihoods and countervailing company claims to legal land rights and the right to earn a return on investment for planted areas. Economic livelihood rights (relating to plasma arrangements) were thus traded off against restitution of land (and associated cultural) rights. These agreements were criticized by some observers as reflecting "not so much a process for achieving redress or remedy, as a process for bargaining," 15 in which communities traded recognized rights to land for other short-term benefits.
Problem-solving and accountability mechanisms also interacted to create pressure for systemic change. The evaluation of CAO lending against the Performance Standards helped instigate strengthened social safeguards governing the World Bank's lending in the palm oil sector, while such accountability, in turn, helped to catalyze internal processes of problem-based learning within the World Bank Group. At the same time, limits to the scope of the Performance Standards (which did not extend to systemic practices of firms) duly limited the scope of systemic change support by the grievance process.

Institutional empowerment
While the CAO's institutional design thus played an important role in determining the effectiveness of its interventions, these effects were also shaped in significant ways by the CAO's institutional authority and capacity.
A key requirement for the CAO's effectiveness in mediating community grievances was its possession of sufficient authority to motivate the parties in the dispute to stay at the table throughout protracted negotiations and to implement agreed outcomes. Achieving this depended importantly on the interplay between the CAO's authority and capacity.
The CAO's authority over the company was relatively high. Perceived impartiality was one important source of authority, grounded in part in high formal independence of the CAO from the IFC, protected through independent selection of the Ombudsman, direct reporting lines to the World Bank Group President, and ineligibility of CAO staff to work within the World Bank Group following the conclusion of their terms. CAO authority was also supported by high levels of transparency of the CAO's activities, reflected in the maintenance of a detailed website and regular stakeholder consultations. The CAO's institutional location within the World Bank group was another source of authority, although this proved somewhat of a double-edged swordundermining the CAO's authority in the eyes of those NGOs who were ideologically opposed to the World Bank. 16 The CAO's capacity to intervene in this case was also high. CAO processes were backed by extensive organizational and human resources, at least compared with other NJMs we examined as part of our studyfacilitated in part by its positioning within the World Bank Group.
The CAO's relatively strong authority and capacity contributed to achieving negotiated settlements in West Kalimantan and Riau in several ways. Strong resourcing for outreach in which CAO staff travelled to communities facilitated communities' access to the mechanism, while strong authority and organizational and professional capacity helped the CAO convene parties within the mediation process, shape perceptions of interconnected interests among parties, deal with contested evidence, and build trust. The CAO contracted skilled mediators with strong local contextual knowledge and high levels of credibility among local stakeholders, which many viewed as crucial in helping them establish strong links with local governments, to achieve sustained engagement from disputing parties. The CAO's links with government were particularly strong in the Jambi negotiations, in which a partnership developed in 2011 between the CAO mediators and a group of district and provincial government officials drawn from land, plantation, and forestry departments, who collaborated on what came to be referred to as a "Joint Mediation" process.
Despite these significant strengths, the CAO's authority remained fragileheavily dependent on the willingness of the company to continue engaging with the mediation process and on an accommodating political environment, particularly at provincial and district levels where government was most actively engaged with the conflict (IPAC 2014). Following a protracted stalemate in negotiations, Wilmar made a unilateral decision in April 2013 to sell the subsidiary involved in the Jambi conflict, leading to the loss of CAO authority over the dispute and the termination of the CAO's mediation efforts (Rofiq & Hidayat 2013). As the CAO withdrew, government control shifted from the provincial to the district level, where the company enjoyed stronger links with government officials, leading to the imposition of the company's proposed smallholding arrangement on community groups who had been holding out for a more favorable negotiated settlement (IPAC 2014, p. 21).

Social empowerment
While the CAO's institutional design and empowerment played a central role in facilitating (and limiting) redress, outcomes were also shaped to some degree by the capacities of affected communities and their supporters to exert leverage over redress processes and to define appropriate redress.
Community leverage was significantly constrained by the communities' socio-economic disadvantage and political marginalization within Indonesia and weak legal rights to land (Colchester et al. 2006). This stands in contrast to the multiple sources of power that many palm oil companies can frequently draw on as a result of a "politico-legal nexus" that sees them "strongly supported by district and national politicians, ill-equipped bureaucracies, local police and pro-plantation legislation" (Gillespie 2012, p. 267). Such relationships are, in turn, embedded within wider growth-and export-oriented extractive regimes and a long history of land dispossession and conflict (McCarthy et al. 2012).
The capacity of communities to control the framing of their grievances and associated definition of appropriate redress was further weakened by their high dependence on external NGOs for information regarding redress options and for capacity-building support to foster community awareness of legal rights and facilitate deliberation around shared negotiating objectives. In some locations (particularly in Jambi), the collective decisionmaking capacity of communities was weakened further by internal conflicts, associated with tensions concerning the definition of indigenous identity, competing community claims to different areas of land, and disagreements about desired goals and strategies of mediations (such as the pursuit of plasma arrangements versus financial compensation or whether to engage with the CAO at all) (IPAC 2014).
Nonetheless, communities and their supporters were able to harness some leverage from major European buyers of palm oil, such as Unilever, to help pressure Wilmar to engage with mediation processes. In December 2011, while negotiations were ongoing in Jambi, Perkumpulan Hijau, a local Jambi NGO, assisted by the German NGO Robin Wood, took a family from the plantation to Germany to lobby Unilever to, in turn, pressure plantation management to return land and compensate families for displacement and dispossession. They also attempted to leverage Unilever's commitment to sustainable palm oil by invoking alleged violations of the RSPO Principles and Criteria by this Unilever supplier. 17 Wilmar was vulnerable to such pressure not only because of its desire to avoid negative publicity or damaged business relationships associated with this specific complaint but also to legitimize its broader claims to support social and environmental regulation in the sectoras illustrated by its prominent position as board member of the RSPO. Despite such vulnerability, transnational market leverage was constrained by Wilmar's ability to distance itself from the market relationships in which the influence was groundedrepaying its IFC loan early as a way of concluding formal contractual obligations to adhere to IFC Performance Standards and then selling the subsidiary involved in the Jambi conflict (Balaton-Chrimes & Macdonald, ).

A top-down pathway to community redress
While some direct social empowerment of communities was required to facilitate access to grievance processes and to exert ongoing pressure on the company to engage in mediation, redress outcomes in this case relied, to a significant extent, on the institutional design and empowerment of the CAO. We characterize this as a top-down pathway for resolving disputes. Communities accessed some redress but were unable to exercise significant control over the terms of the remedy. Rather, the top-down pathway entailed high dependence on the CAO's own initiative, agenda, and understanding of its own mandate. To the extent that community aspirations to secure compensation overlapped with the CAO's capacity and willingness to conduct individual mediations, community redress was supported. Nonetheless, there was little scope for remedy in the form of either accountability for individual rights violations or systemic redress for corporate and government practices external to the World Bank Group.

A bottom-up path to remedy: Mining in India
Next, we examine the case of a contested mining project in India, which illustrates what we characterize as a bottom-up pathway to redress. In this case, the partial redress accessed by communities was facilitated primarily by their ability to harness a transnational nonjudicial redress process in support of sustained strategies of protest and grassroots mobilization. A more minor role was played by interventions initiated by the transnational NJMthe UK National Contact pointwhose authority and capacity remained relatively weak.

Overview of the case
This case involved claims by tribal communities affected by a proposed bauxite mine in the Niyamgiri Hills in the eastern Indian state of Odisha. The mine was proposed as part of an aluminum production complex led by the company Vedanta, which was primarily based in India but is also listed on the UK stock exchange. Vedanta's mining operations raised significant human rights concerns, particularly concerning project impacts on public health, the environment, and land rights of the indigenous Kondh people (Saxena et al. 2010).
Communities and coalitions of national and international supporters engaged in sustained campaigns against the mine and pursued redress through various channels. These included several administrative reviews and court cases in India, international NGO campaigns targeting Vedanta shareholders, and a complaint to the UK National Contact Point (NCP)an NJM created under the OECD Guidelines for Multinational Enterprises. In 2009, a successful claim against Vedanta was brought before the UK NCP. Although Vedanta refused to engage with the NCP processrefuting all allegations while refusing to submit any evidence to substantiate its claims and rejecting the NCP's offer of mediation 18the NCP determination against Vedanta was a major reputational blow for the company, leading to well-publicized disinvestment by a small number of shareholders, primarily European investment funds. 19 While influence over administrative and judicial decisions in India was limited, the NCP's decision, combined with pressure from the company's investors, helped catalyze a number of policy changes within Vedantastrengthening internal capacity to manage human rights impacts of the company's operations (self-reference).
After multiple court cases at state and national levels, the Supreme Court of India eventually determined that the decision about whether the mine would proceed should be passed back to the relevant local decisionmaking bodies, the Gram Sabhas. Following unanimous votes against the mine at all these forums in 2013, the Ministry of Environment and Forests (MoEF) revoked permission for the mine (Saika 2014). Although reports continue of communities experiencing harassment, and the Indian government continues to support ongoing land acquisition (Chakravarti 2015), sustained mobilizations have resulted in concrete legal and political outcomes that have significantly altered the course of this project.

Institutional design
The transnational NJM used in this case was the UK NCP, whose institutional design rests centrally on problemsolving approaches. According to the OECD, NCPs are encouraged to consider themselves "informal problemsolvers in corporate responsibility disputes" (Organisation for Economic Co-operation and Development 2012, p. 42) and "[t]he benchmark of success is the ability of NCPs to facilitate mediation and dialogue" (OECD 2014, p. 41). Despite the official emphasis on problem solving, in practice, such approaches played a small role in shaping redress outcomes in this case as Vedanta refused to engage with the NCP process.
The NCP system also encompasses some scope for accountability processes, insofar as complaints to NCPs make reference to allegations of noncompliance with the OECD Guidelines for Multinational Enterprises, and NCPs are empowered to carry out investigatory activities and make determinations as to whether or not the Guidelines have been breached. While NJMs are not empowered to impose authoritative sanctions, such determinations can still generate significant reputational costs for companiesas in this case, where the UK NCP's determination that the company had breached the OECD Guidelines provoked a strong response from company investors, enabling civil society networks mobilizing against the mine to strengthen both market and political pressures on the company. In this sense, accountability elements of NCP institutional design were significant but very indirectgenerating effects primarily through their interaction with broader sources of community leverage.

Institutional empowerment
The indirect character of the NCP's influence over this grievance also stemmed from weaknesses in the NCP's institutional empowerment. The UK NCP lacked sufficient authority to motivate Vedanta to recognize or engage with their processes, and their status as a foreign entity undermined their authority in the eyes of affected communities and NGOs in Odisha, who were wary of outside intervention that would reduce the centrality of affected communities in decisionmaking about the campaign (Macdonald et al. 2017). The NCP's capacity to influence the course of the dispute within India was also constrained by its lack of an operational presence in India and of relationships with Indian regulatory actors that might have facilitated some form of regulatory collaboration.
Despite such weaknesses, the NCP was viewed as highly credible by key European shareholders and institutional investorsdrawing authority from perceptions of its impartiality, its location within the UK government, and its role in implementing internationally recognized standards of responsible business conduct. The divestments that followed the NCP decision were also widely recognized within India, and the Norwegian Pension Fund's decision in 2007 to sell its entire stock in Vedanta was specifically mentioned in a 2007 Supreme Court ruling on the case that same year. There is also evidence that the Environment Minister in India was aware of the NCP determination and the publicity surrounding it, and this may have bolstered his resolve to block the mine (Macdonald et al. 2017).

Social empowerment
To the extent that the UK NCP was able to support community remedy, such contributions depended crucially on the NCP's interactions with affected communities and their supporters and the capacity of the latter to exercise leverage over contested redress processes.
Community leverage was structurally constrained by the socio-economic, political, and cultural marginalization of affected communities. These forest-dwelling communities are among the most vulnerable groups in India, facing persistent problems of land alienation, indebtedness, and involuntary displacement due to development projects (Mishra 2011). Community resistance to Vedanta's project was further impeded by support for the project from a powerful network of government and business actors, including corporate executives, investors, senior figures within the Odisha State government, and officials at local levels of government (Xaxa 2012;Kumar 2013).
Nonetheless, communities were able to draw on multiple sources of organizing capacity and leverage at local and transnational scales. The Kondh communities affected by the planned mine were unanimous in their desire to stop the mine and were able to build strong collective organizing structures based on pre-existing governance systems within the tribal communities. At the same time, they remained relatively open to support from outside actors facilitating links with broader Odisha-based activists who had strong traditions of acting in "solidarity" with tribal and forest peoples (Swain 2000;Kumar 2013). Shortly after Vedanta's memorandum of understanding (MoU) was signed in 2003, Odisha-based activists began attending local protest events, mobilizing villagers, and making connections with other activists involved in anti-industrialization and propoor struggles within Odisha (Kraemer et al., 2013, pp. 834-835).
As the Vedanta conflict unfolded, the networks of local movements opposing the mine gradually gained support both within Odisha and nationally (Xaxa 2012;Kumar 2013). Support from elites at the national level was particularly important as a basis for securing legal and administrative decisions that responded to community demands. Grassroots campaigners were supported by a number of state and national civil society entities, including Lok Shakti Abhiyan, a loose network of antineoliberal activists at the Odisha state level connected to the National Alliance of People's Movements. The remote forest location of the tribal communities and the associated capacity for the project location to be characterized discursively as "pristine," and replete with "charismatic mega-fauna" (elephants) also facilitated the development of linkages with wilderness movements, including groups like The Wildlife Society of Odisha and The Wildlife Institute of India, the latter of which performs an advisory function to the Ministry of Environment and Forestry. Lawyers at the national level provided particularly important support, possessing strategically important procedural and technical knowledge, as well as access to informal networks of elite decisionmakers. As such networks grew in influence, they also drew in powerful supporters such as Rahul Gandhi, Chairman of the Congress Party, one of two major political parties in India, and an influential political figure. Other key political figures from Congress added their weight to opposition to the project at various points, including Jairam Ramesh, Minister of Environment and Forests from 2009 to 2011, a strong political ally of the Gandhis (Macdonald et al. 2017).
Such local sources of leverage were further bolstered by support from international NGOs and public figures in the UK and Europe and by harnessing market pressure from international investors (Xaxa 2012;Kumar 2013). Transnational advocacy often focused heavily on discursive frames emphasizing environmental protection and indigenous rights in contrast to the emphasis of many local activists on tribal rights (self-reference). Nonetheless, transnational networks provided consistent support for the community's central aim of stopping the projectfacilitating strong community control over defining the terms of appropriate redress.

A bottom-up path to remedy
In contrast to the top-down pathway to community redress, where problem definition and the terms of associated responses were controlled by the NJM, communities in this case drew on the authority of nonjudicial institutions selectively, in ways that supported their own framing of the grievance, and associated demands for redress. In this bottom-up pathway, the weaving together of social and institutional power in support of remedy thus depended crucially on the agency of affected communities and their supporters, who were able to mobilize multiple sources of social, political, and market leverage, drawing strategically on NJM authority where this could support broader struggles for community remedy. Despite the advantages of this bottom-up pathway for community empowerment, it relied heavily on protracted social and political struggleimposing significant costs on communities. Communities may be willing to endure such costs when the alternative to a protracted struggleacquiescence to a projectthreatens their whole livelihood or way of life, but in many cases, such sustained struggles will be unappealing or simply inaccessible for affected communities.

Discussion and conclusions
Taking critical debates surrounding nonjudicial regulatory approaches as our central point of departure, our aim has been to explore under what conditions (if any) NJMs can make useful contributions to empowering community access to remedy.
Our analysis provides some support for the widespread view that institutional design choices play a central role in enabling (and limiting) the potential of NJMs to support community remedy. However, while emphasis has often been placed on the distinctive value of problem-solving approaches as a basis for supporting nonjudicial redress, we have also highlighted the significance of nonjudicial accountability mechanisms as means of supporting compliance with recognized rights in the presence of persistent power imbalances between the companies and communities involved in contested grievance claims. We have further argued that, in the presence of such structural power imbalances, effective nonjudicial processes also depend crucially on both institutional and social empowerment as a basis for establishing the plural sources of "countervailing power" (Fung 2003) required to provide backing for their interventions.
As we have seen, dynamic interactions between these plural sources of social and institutional power give rise to different possible pathways to NJM effectiveness. On what we have referred to as a "top-down" pathway, constructive NJM interventions are grounded centrally in the institutional capacity, authority, and regulatory agency of NJMs. In other cases, the countervailing power required to incentivize companies to engage with nonjudicial procedures and determinations is derived primarily through "bottom-up" social leverage, relying centrally on the structural position, alliances, and organizing capacity of affected people. Within each pathway, countervailing power is woven together in response to context-dependent configurations of social, economic, and institutional power at local and transnational scales. Which pathway is likely to emerge will depend on the structure and dynamics of global supply chains, the complex histories and political economies of individual host countries, and more contingent factors relating to the agency of participants in particular regulatory struggles.
By focusing on the significance of power dynamics as determinants of NJM effectiveness, our findings provide support for critiques of voluntarist regulation that stress the importance of tackling structural power imbalances as a basis for countering domination and protecting rights (Ford 2010;Solomon 2008). However, by focusing on the varied sources of social and institutional power that can provide this backing in the absence of state-based legal remedies, our analysis extends a normative concern for rights and power beyond the context of debates surrounding law and new governanceadapting this analysis to the fragmented and interactive global governance processes that characterize the contemporary transnational business and human rights field (Ruggie 2017;Melish & Meidinger 2011). In this interactive regulatory environment, no single source of social or institutional power can hierarchically dominate the others; instead, regulatory outcomes are determined by a complex and dynamic interplay between social and institutional enabling conditions. Although the forms of power and legitimacy associated with these plural sources of countervailing power are very different to those underpinning legal institutions (and cannot substitute for them), we have nonetheless argued that, in a decentered governance setting, social and institutional sources of power need to be taken more seriously in our analysis of enabling or scope conditions for productive nonjudicial regulatory interventions.
Recognizing the plural sources of power that can back the functioning of NJMs also has implications for how we think about regulatory power and agency within decentered regulatory processes. The agency of nonjudicial institutions within interactive processes of business regulation remains important: particularly within top-down pathways to redress, NJMs can play a central steering role within processes of contested redress. Yet it is not always authoritative institutional regulators who are best positioned to steer or orchestrate contested regulatory processes. Under the right conjunctures of enabling conditions, more diffuse social actors can also play significant roles in weaving together multiple sources of social and institutional power in support of community redresssometimes appropriating the resources and capacities of NJMs in ways that these institutions themselves did not intend.
Regardless of how they combine through dynamic and context-specific pathways to remedy, the enabling conditions that we have identified are demanding, and difficult to establish in the kinds of political, economic, and social contexts in which human rights violations associated with transnational business activity are most prevalent. In this sense, our theoretical analysis of the complex conjunctures of conditions required for NJM success helps us to account for pervasive patterns of NJM failure. At the same time, however, our analysis enables us to acknowledge the limitations of nonjudicial regulatory approaches while also informing pragmatist agendas interested in exploring what NJMs can contribute to constructing "actionable paths to cumulative change" in spite of such structural obstacles (Ruggie 2017).
By highlighting the multidimensionality of many community grievances and the corresponding complexity of redress-handling processes, our analysis has also highlighted significant uncertainty regarding the prospects for success in any given case, as relevant enabling conditions interact in changing ways, in response to shifting social and political alliances, social and economic power, and associated patterns of social organization and claimmaking at multiple geographical scales. Even where a primary purpose for appealing to an NJM fails to be achieved, sometimes, the normative or institutional changes associated with the claim-making process itself can feed into future campaigning and alliance building to support further systemic change. The dynamism and unpredictability of such processes may help to make sense of why NJMs continue to be used by communities despite their significant limitations. By providing a more explicit theoretical rationale for understanding these observations, our analysis can perhaps offer at least small seeds of optimism within the rather bleak terrain of opportunities for human rights redress confronted by communities around the world affected adversely by transnational business activity.
19 Following the NCP's determination against the company, the Church of England sold its 3.8 million pound stake in Vedanta; in 2007, the Norwegian government sold its $13 m stake; Martin Currie Investment Management sold its £2.3 m stake in 2009; and BP's pension fund reduced its holdings in Vedanta because of "concerns about the way the company operates" (Hopkins 2010).
international organization actors), agreement of the parties to medium-or long-term involvement (e.g. to monitor implementation of the agreement), and absence of direct discursive challenges to NJM authority. Indicators of NJM capacity include: evidence of financial and institutional resources and capacities required for mandated functions (whether problem-solving or accountability-oriented); human capacities embodied in the skills and experience of NJM staff and contractors; ability to draw on the expertise, resources, and capacities of external networks of actors at transnational or local levels, such as local professionals, civil society, or government actors; and capacity to engage in the medium or long term. Taking these multiple considerations into account (and weighing them according to our judgement of their respective salience in the context of each individual case), we assigned scores of 1-5 for institutional empowerment of the relevant NJM in each case.

Social empowerment
Indicators relate to both direct evidence of the exercise of community capacities and evidence of community possession of underlying resources and capabilities providing them with potential capacity. Indicators regarding community control over terms of possible redress were based on assessments in each case of: changes in the discursive framing of grievances in line with community representations of the underlying problem requiring redress; community organizational, mobilization, and collective decisionmaking capacities; community access to information about available redress options; and community capacity to network with national and transnational supporters. Assessments of community leverage focused on evidence of the capacity of communities or networks of community supporters to impose costs on companies as a means of creating incentives for companies to engage with NJM processes and respond to community claims, whether via direct access to or influence over regulatory or political authorities at local or transnational scale and/or as a result of structural power relations constituted within global and local political economies in particular places and sectors.
Taking these multiple considerations into account (and weighing them according to our judgement of their respective salience in the context of each individual case), we assigned scores of 1-5 for social empowerment of the relevant communities and broader networks of supporters in each case.