Abstract

Transparency is now a core principle in environmental and resource governance. Responding to calls for a clearer identification of pathways from transparency to effective change, this article identifies three “Theories of Change” for governance-by-disclosure and applies them to the Extractive Industries Transparency Initiative (EITI). Among the best known global transparency initiatives, the EITI has used an inclusive multistakeholder governance model and elaborate compliance standards, disclosing trillions of dollars in natural resource revenues. Yet, after two decades, the EITI is still largely without an explicit and proven theory. This study finds that a Theory of Change for the EITI is possible, valuable, and even necessary as the EITI risks becoming obsolete in some participating countries. The proposed Theories of Change provide valuable templates for environmental and resource governance, yet such models need to reflect national contexts, needs, challenges, and objectives to ensure fit and effective implementation, including measures enforcing accountability.

Transparency has become a central tenet of many new environmental governance instruments (Gardner et al. 2019; Gupta and Mason 2014). The effectiveness of “governance-by-disclosure” remains debated, however, with critiques pointing at the limits, conflicts, and counterproductive, even obscuring, effects of transparency-centered governance (Auld and Gulbrandsen 2010; Bauhr and Nasiritousi 2012; Dingwerth and Eichinger 2010). Besides pointing at the lack of evidence, these studies have called for a more explicit “Theory of Change” linking transparency with improved outcomes (Mejía Acosta 2013), suggesting, for example, that the assumptions related to the relationship between transparency and accountability are oversimplistic and generally ill founded (Lujala et al. 2020; Mason 2020; Ofori and Lujala 2015). Here we examine Theories of Change for transparency in environmental and resource governance, focusing on the case of the Extractive Industries Transparency Initiative (EITI) launched in 2003 to improve the governance of extractive sectors in the implementing countries through greater disclosure.

The EITI has been a hallmark of international resource governance efforts and a flagship of “governance-by-disclosure” (Gupta 2008; Haufler 2010). Yet, after nearly two decades of existence, the EITI’s effectiveness is still questioned (Rustad et al. 2017), with regard to both its core objective (improving the developmental impacts of extractive industries) and its long avoidance of dealing with the environmental impacts of extraction (Le Billon and Spiegel 2021). Given the challenges of demonstrating EITI’s effectiveness (Lujala 2018; Sovacool et al. 2016), many researchers and practitioners have called for an explicit “Theory of Change” (ToC) articulating the steps through which the EITI is expected to bring out improvements (Gillies and Heuty 2011; Neumann et al. 2016; Scanteam 2011; Vijge et al. 2019)—a “theorization” that the EITI International Secretariat, for a long time, seemed reluctant to establish (Rich and Moberg 2015).

Based on previous literature and our own observations from nearly two decades of engagement with the EITI, we develop three stylized models on how the EITI can attain its long-term goals of improving natural resource governance and sustainable development outcomes. These are (1) Name-and-Shame, that is, denouncing stakeholders mismanaging or stealing natural resource revenues; (2) Public Debate, that is, increasing knowledge levels among the public and public demand for better resource governance; and (3) Technical Reform, that is, strengthening the procedures and bureaucracy dealing with natural resources. Although the three models are not mutually exclusive and can be complementary, they are useful in structuring reflections and discussions about environmental and resource governance through disclosure.

Using a survey conducted at the 2019 EITI Global Conference in Paris and interviews conducted in Colombia, Ghana, and the United Kingdom, the article presents four key findings regarding a potential EITI ToC. First, there is a need, and a desire among the stakeholders, to develop a ToC for the EITI. Second, the EITI ToC needs to be context specific. Third, the three main ToC models outlined above are applicable to the EITI, although their relative suitability varies from country to country and over time within a country, suggesting that such a ToC needs to be dynamic. Fourth, the EITI is at risk of becoming obsolete in some countries, as it has neither resulted in increased accountability in resource governance nor adequately addressed some pressing issues regarding the extractive sector. Taken together, these results suggest that ToCs for transparency initiatives in environmental and resource governance need not only to set clear pathways to reach defined—and relevant—objectives, but also be contextualized and remain flexible to reflect changing conditions of their implementation and the issues they seek to address.

Transparency in Environmental and Resource Governance

Framed by a broad “good governance” paradigm, transparency has been a major dimension of the environment and resource governance agenda for at least three decades, with information disclosure having an even longer role in the history of environmental activism (see Gupta and Mason 2014). An important discussion within global environmental politics literature is whether disclosure improves the performance of environmental governance and, if so, through which pathways and for whose benefit. In this respect, the literature on transparency and information disclosure points to a wide array of potential avenues by which information availability can propel change (Fenster 2015; Roelofs 2019), each in its particular way and (unintended) effects (Ejiogu et al. 2019; Gupta & Mason 2014).

While a full review of the environmental and resource governance-by-disclosure literature falls beyond the scope of this article (see, e.g., Gupta and Mason 2014; Zalik and Osuoka 2020), the main arguments have been that governance-by-disclosure reduces informational asymmetries and increases the capacity of stakeholders to evaluate performances, but that it falls short in actually enhancing capacity to change the status quo and modify behaviors and the power of incumbent decision makers (Mason 2020). Ciplet et al. (2018), for example, find that disclosure in climate finance implemented under the Paris Agreement did not translate into widespread accountability in emissions, possibly because of the propensity of transparency initiatives to foster technocratization rather than democratization (Gupta and Mason 2016)—with empirical research on transparency identifying the “public” as the weak point of governance-by-disclosure (Fox 2015; Lujala et al. 2020).1

A further challenge is that depending on the specific objective(s) of a governance-by-disclosure initiative and the context in which it is implemented, different kinds and types of information, disclosure avenues, and enabling conditions are required for it to be successful (Fenster 2015; Fox 2015; Heald 2006). For example, restraining extractive industry’s environmental or social impact requires different types of information to be disclosed than the information that is supposed to form a basis for the public to demand reforms in national natural resource revenue governance (Bleischwitz 2014).

As Gupta et al. (2020, 85) recently sum up, the transformative potential of governance-by-disclosure thus reflects not only the design involved but also the “normative and political contexts within which such systems are deployed.” A pessimistic interpretation of this finding would suggest that to be effective, transparency requires preexisting norms of responsiveness to be in place for the disclosed information to effectuate change, something that may not be the case in many countries. Yet, as several studies emphasize, some governance-by-disclosure processes—for example, when carried out through multistakeholder approaches—can empower civil society actors through their inclusion in “constituting and debating” transparency (Vijge 2018, 13), which, in turn, can promote further governance innovation (Arond et al. 2019). More optimistically, and as discussed below, this suggests that in order to be most effective, transparency initiatives should come with a ToC identifying the mechanisms and conditions from information disclosure to the stated objectives and tailored to local contexts, priorities, and promising avenues for more inclusive and effective governance.

Theories of Change for Transparency in Environmental and Resource Governance

There is no consensus around a specific definition of ToC, but broadly, it can be described as a way to understand how an intervention will lead to a specific change (Stein and Valters 2012). The ToC concept has been increasingly adopted as to complement statements of objectives with a “proof of concept” analysis not only spelling out a road map explaining how to get there but also seeking to demonstrate that these objectives are indeed reachable. In practice, a ToC identifies a set of intermediary steps and assumptions through which short-term activities and outputs of an initiative will lead to the achievement of its long-term goals (Funnell and Rogers 2011). Importantly, a good ToC identifies and explicitly states the causal chain and mechanisms between the outputs, intermediary steps, and ultimate goals. Furthermore, it addresses the complexity of change processes, inclusive of the intervening factors outside the control of the project, and identifies measurable indicators that can be used to determine whether an outcome has been achieved (Mayne 2015). Many organizations, programs, and projects use approaches similar to ToC without explicitly calling them a ToC.

A ToC can serve many purposes, such as developing common understanding or strengthening clarity, effectiveness, and focus, as well as providing a framework for evaluation and monitoring, organizational development, communication, and empowering stakeholders. As Stein and Valters (2012) point out, it can be unclear at which level a ToC should be implemented, whether it is necessary at all levels, or whether all levels and parts of an organization should have the same ToC. Designing a unified theory for the whole organization can be an enormous, if not impossible, task and may not be the best option, especially if the ToC is to be implemented in different contexts and levels. A further consideration is ownership of the process, “total ownership” being the ideal in which all stakeholders are engaged in developing and evaluating the ToC (Sullivan and Stewart 2006).

We use the concept of ToC to investigate pathways for how governance-by-disclosure can function in environmental and resource governance. Previous literature on global environmental politics has, among others, discussed these types of approaches translating transparency into accountability in extractive sector management (Epremian et al. 2016), carbon emissions such as the Paris Rulebook (Knox-Hayes and Levy 2014), and climate finance mechanisms like REDD+ (Bumpus et al. 2019). Some of this literature points at the importance of differentiating between types of transparency in order to identify the transformative pathways involved and, thus, the ToC at play. In her recent study, Tienhaara (2020) usefully categorizes the main types of transparency and ways through which they are expected to be transformative (first two columns in Table 1). These five transparency types can be condensed into three possible general trajectories for governance-by-disclosure ToCs in environmental and resources governance (last column in Table 1). These ToCs, in turn, can be used to understand the different paths of changes that transparency can take toward development.

Table 1

Transparency Types, Transformative Pathways, and Theories of Change

Transparency TypeTransformative PathwaysTheory of Change
Conventional transparency Transformation through post hoc access to information enabling accountability Name-and-Shame: Enables accountability mechanisms, such as elections and judicial processes 
Deliberative transparency Transformation through more informed debates leading to reforms Public Debate: Fosters and informs public debates promoting reforms 
Disruptive transparency Transformation through status quo disruption by previously undisclosed information Disrupts status quo and calls for renewed public debates, on the need for reform 
Technocratic transparency Transformation through technocratic reforms informed by disclosed information Technical Reform: Informs and motivates technocratic reforms 
Disciplinary transparency Transformation through technocratic processes improved and disciplined by disclosed information Creates and enforces more rigorous technocratic procedures 
Transparency TypeTransformative PathwaysTheory of Change
Conventional transparency Transformation through post hoc access to information enabling accountability Name-and-Shame: Enables accountability mechanisms, such as elections and judicial processes 
Deliberative transparency Transformation through more informed debates leading to reforms Public Debate: Fosters and informs public debates promoting reforms 
Disruptive transparency Transformation through status quo disruption by previously undisclosed information Disrupts status quo and calls for renewed public debates, on the need for reform 
Technocratic transparency Transformation through technocratic reforms informed by disclosed information Technical Reform: Informs and motivates technocratic reforms 
Disciplinary transparency Transformation through technocratic processes improved and disciplined by disclosed information Creates and enforces more rigorous technocratic procedures 

From Tienhaara (2020) and the authors.

Conventional transparency focuses on limiting corruption and improving governance through disclosure of information on whether decision makers have fulfilled their responsibilities. This gives the basis for Name-and-Shame ToC that focuses on leveraging transparency as an anticorruption instrument and as a threat of exposing unethical behavior or other mismanagement in order to make individuals, governments, or organizations behave in specific ways or adhere to common norms when it comes to issues such as protection of environment, labor, and Indigenous rights. The approach is grounded in the constructivist view with the understanding that the behavior of political agents is socially constructed and formed by the public’s perceptions of them (Koliev and Lebovic 2018). It is based on the idea that reputational damage can be costly to individuals, governments, and organizations breaking collective rules.

In deliberative transparency, the information is also directed toward the public as in conventional transparency, but with the aim that the public engages in discussions and comes with advice on how to better govern the issue at hand. The citizens are the target in disruptive transparency as well, but the aim of it is to “interrupt the secret flow of information within and among powerful institutions” (Tienhaara 2020, 114). Public Debate ToC is thus based on what is probably the most common perception of what a transparency and accountability ToC is: leveraging transparency initiatives as means to increase knowledge levels among the public and as a debate process to improve environmental and resource governance. The idea ties to the deliberative democratic theory, claiming that political decision should be a product of “debate and discussion aimed at producing reasonable, well-informed opinion in which participants are willing to revise preferences in light of discussion, new information, and claims made by fellow participants” (Chambers 2003, 309).

The last two types of transparency have other targets than citizens. Technocratic transparency promotes improved governance through information disclosure to specific expert bodies that can use the information to promote better policies through reforms. Disciplinary transparency, in turn, aims at transformation through improved processes and procedures based on the disclosed information. Both can be considered to promote technocratic changes and are combined in Technical Reform ToC that focuses on leveraging transparency as an informational tool and bureaucratic process to enhance environmental and resource governance through improved procedures, greater efficiency, and reforms.

The EITI and Theories of Change for the EITI

The EITI was officially launched in June 2003. Initiated within an international development and environmental agenda promoting “good governance” principles, including transparency, the EITI was designed as a soft response to corruption concerns relying on a voluntary process expected to promote accountability through the public disclosure of independently verified revenue payments between companies and governments. As such, the EITI represented a blend of our three main pathways for transparency ToCs (Table 1), whereby the intentions of activists to “name and shame” corrupt companies and governments through disclosure were operationalized through “technocratic reform” enabling the verification and disclosure of revenue flows in the hope that this information would then inform a “public debate” pushing for improved resource governance and, ultimately, better development outcomes.

By November 2020, the EITI counted fifty-three implementing countries, including many resource-dependent, low- and middle-income countries, and it had by 2020 helped disclose about US $2.7 trillion of government revenues (EITI 2020). While national-level participation in the EITI is voluntary, the EITI Standard requires mandatory disclosure for the government and all extractive companies along the whole natural resource management value chain, including disclosure on corporate beneficiary ownership, contracts and licenses, and company payments and government revenues from the extractive industries (EITI 2019).

Although a ToC could potentially be useful for the EITI, and was forcefully requested by some donors in the early 2010s,2 the EITI was for a long time reluctant to develop a ToC, the EITI Secretariat’s executive director and deputy head concluding in their book in 2015 that “there is limited benefit in a theory of change and little effort should be made to establish overarching goals” (Rich and Moberg 2015, 67). It was only in 2019 that the EITI provided a ToC, built around the EITI validation criteria (EITI International Secretariat 2019, 51). This ToC, however, does not specify how the change will happen but asserts that as long as a country’s progress is satisfactory in meeting the EITI Standard, implementing the EITI “can contribute” to improved investment and governance and, ultimately, to economic growth and reduced poverty.

Based on our review of the literature and EITI-specific material (e.g., EITI Standards and Principles, literature, and our interviews, discussions, and observations around the EITI and with individuals linked to the EITI for more than two decades), we applied the three ToCs outlined in Table 1 to the EITI and developed three stylized EITI ToC models, taking the EITI information disclosure as the starting point and improved development outcomes as the ultimate goal (Figure 1). Although these models are simplified, seeking to condense the transparency process into a minimum of easily understood steps, they help in thinking key pivoting processes that can lead to change through governance-by-disclosure in environmental and resource management.

Figure 1

Three EITI Theory of Change Models

Figure 1

Three EITI Theory of Change Models

Naming and shaming was forcefully outlined in Global Witness’ (1999) report A Crude Awakening, which sought to denounce the complicity of large oil companies in Angola, like BP-Amoco, ELF, Total, and Exxon, in the plundering of state assets. The initial success in convincing BP-Amoco to commit to publish its payments to the Angolan government was quickly dampened by threats from the government to cancel all of BP-Amoco’s concessions in the country for breaching confidentiality agreements (Le Billon 2005). This reaction led Global Witness to pursue the creation of an international instrument to force extractive companies to disclose payments to host governments, leading to meetings with industry and regulators, the launch of the Publish What You Pay campaign, and the EITI. Name-and-Shame ToC is based on the idea that by publishing data on revenues from the companies to the government, it is possible to pinpoint which companies and government agencies have discrepancies when reporting revenue flows between the industry and state. The discrepancies, or unexpectedly small flows, can be investigated to expose possible tax evasion, corruption, or embezzlement to reduce revenue losses and improve the integrity of resource governance.

Public Debate ToC is based on what is probably the most common perception of what the EITI ToC is and is very much in line with democratic values and the EITI’s ideal of including the public and civil society in the transparency process. This approach also resonates with what Rich and Moberg (2015) refer to as collective governance, in which all three stakeholders—government, companies, and civil society—but, in particular, the civil society together with the public have a key role in ensuring accountability through mobilization to press the state to improve its natural resource (revenue) governance.

In Technical Reform ToC, the mandatory national EITI reporting has a dual role. First, the implementation of the EITI Standard itself, with all its technocratic aspects, delivers change: “A key reason to why the EITI has been successful in many countries is that … [the] implementation [of the EITI Standard] leads to real and measurable results” (EITI 2014, 10). By implementing the reporting procedure, countries are thus expected to achieve some degree of transformation toward better resource governance. Second, the EITI reporting can reveal shortcomings in resource governance systems, and based on these, the national EITI is expected to provide recommendations for improvement, contributing to government’s policy reforms: “In many countries, the most important contribution of the EITI has come about because governments have decided to act on recommendations that have emerged from EITI reporting” (EITI 2016, 3).

In short, while the Name-and-Shame model seeks to expose corruption and undermine misplaced public trust in extractive companies and/or government authorities, the Public Debate model seeks to increase collaboration and trust across civil society, corporations, and governments through informed discussions, and the Technical Reform model seeks to improve the integrity and effectiveness of technocratic processes.3 The three models, however, only present some of the possible steps between information disclosure and development—with social mobilization, litigation, or electoral politics offering other avenues for disclosure to effect change. Furthermore, ToCs are not mutually exclusive and can be combined and connected to each other in different ways. Naming and shaming, for example, can lead to judicial action, resulting in more public debates and more effective bureaucracy. Instead of seeing the three ToCs as separate, linear streams, they should thus be seen as forming a complex system in which they can potentially support and reinforce each other.

Key Findings

Our Analysis

Our analysis is based on a survey, interviews for three country case studies, and the author’s observations about the EITI and EITI process over the last twenty years.4 The survey was conducted among participants to the 2019 EITI Global Conference. The survey sought to map various EITI stakeholders’ thoughts on what they perceived to be key objectives for the EITI, their experience with the theory of change, and what they saw as the main steps to achieving change in the context of the EITI. Our fifty-nine respondents were also asked to indicate which of the stylistic models we had laid out they felt was the most fitting for the EITI. We also conducted semi-structured interviews with twenty-four interviewees in three countries—Colombia (six), Ghana (thirteen), the United Kingdom (five)—and one in Norway with an EITI board member between December 2019 and March 2020. The interviews sought to gain a deeper understanding of national-level perspectives across a range of stakeholders (see Appendix). The interviews followed a common template for all three countries, and all the respondents were presented with the three stylized ToC models.

ToC for the EITI Desirable and Relevant

Most of our fifty-nine respondents surveyed at the 2019 EITI Global Conference agreed (47%) or agreed strongly (12%) that it would be useful to have a ToC for the EITI, suggesting a clear demand for a ToC among the Global Conference participants. Only one respondent disagreed with the statement, while 20 percent of the respondents remained neutral, and 19 percent answered that they did not know or did not answer the question. Our twenty-four interviewees from the three case study countries (the United Kingdom, Ghana, and Colombia) and the EITI Board also saw a ToC to be useful for the EITI, and at least in Ghana and Colombia, there had been discussions about how to link the EITI information disclosure to changes in natural resource governance and broader societal development. Thus, according to both our survey respondents and interviewees, a ToC for the EITI was seen to be desirable.

Public Debate Preferred, but Change through Technical Reform the Reality

When presented with the three stylized ToC models for how the EITI could potentially lead to improved developmental outcomes, our survey respondents chose the Public Debate model as the one that best represented their organizations’ view and also as the one that they personally thought to be the best model for the EITI (Figure 2). Technical Reform was the second most popular model among our respondents and their organizations, although somewhat less so from a personal than an organizational perspective. Based on our limited sample, respondents from low-income countries and less democratic countries tended to express a preference for the Name-and-Shame model rather than the Technical Reform model (see Appendix), suggesting a greater need for “hard” accountability processes and likely a greater distrust in process-oriented approaches, such as Models B and C.

Figure 2

Preferred ToC Based on the Survey Respondent’s Organization or Personal View

Figure 2

Preferred ToC Based on the Survey Respondent’s Organization or Personal View

While the survey respondents had to choose only one model, the interviewees could choose and discuss several and thus provide a more nuanced picture. Many interviewees saw two or all ToC models to fit their countries, especially in Ghana, where most interviewees deemed all three models relevant (Table 2), while several interviewees pointed out that the models work in parallel, often being complementary and feeding into each other.

Table 2

Summary of Model Fit by the Case Country

CountryA: Name-and-ShameB: Public DebateC: Technical Reform
Colombia Least relevant model as other anticorruption processes are in place, and corruption issues are mostly associated with subnational authorities and so far poorly covered by the EITI Most relevant model as EITI reports inform some public debates, but so far very limited impact due to lack of dissemination and limited interest Second most relevant model as the EITI clarified and helped to correct bureaucratic processes and informed reforms 
Ghana Relevant in the sense that it forces the companies to pay their dues (taxes, royalties, surface rentals, etc.) and to comply with the information disclosure requirements Relevant to the extent that the EITI is making a lot of information available to the general public but seen rather irrelevant currently as the link from the disclosed information to accountability is missing Relevant as the EITI has improved revenue collection and informed polices, regulations, and legislation concerning the extractive sector 
United Kingdom Not relevant Could be relevant but of limited impact as the general public is not interested in the issue Relevant as the EITI has been used to push some reforms such as reporting at project level 
CountryA: Name-and-ShameB: Public DebateC: Technical Reform
Colombia Least relevant model as other anticorruption processes are in place, and corruption issues are mostly associated with subnational authorities and so far poorly covered by the EITI Most relevant model as EITI reports inform some public debates, but so far very limited impact due to lack of dissemination and limited interest Second most relevant model as the EITI clarified and helped to correct bureaucratic processes and informed reforms 
Ghana Relevant in the sense that it forces the companies to pay their dues (taxes, royalties, surface rentals, etc.) and to comply with the information disclosure requirements Relevant to the extent that the EITI is making a lot of information available to the general public but seen rather irrelevant currently as the link from the disclosed information to accountability is missing Relevant as the EITI has improved revenue collection and informed polices, regulations, and legislation concerning the extractive sector 
United Kingdom Not relevant Could be relevant but of limited impact as the general public is not interested in the issue Relevant as the EITI has been used to push some reforms such as reporting at project level 

Name-and-Shame ToC was generally seen by our interviewees as initially relevant and motivating, especially on the part of CSOs, which initially pushed for the EITI adoption. There was a sense that “naming and shaming” had helped to secure the participation of the companies—forcing them to “open their books” or have their names published in the annual EITI report—and this way, it had limited intentional and nonintentional tax evasion. Furthermore, it had helped to identify underperforming government agencies not being able to claim companies to pay their taxes and other payments. Regarding outright corruption, in none of the case countries had the EITI been able to expose clear-cut cases of corruption, and in general, the EITI was seen as a support for the state institutions in upholding and enforcing anticorruption and antibribery laws through provision of third-party audited data.

Public Debate ToC was seen as the foundational principle of the EITI, the one that constitutes the core of the EITI approach. Yet few interviewees asserted that the dissemination of information and its contributions to public debates had contributed significantly to improving the governance of extractive sectors and its revenues. Although the EITI was prized to have made information public, the absence of feedback mechanisms from the information disclosure to increased accountability specific to this approach was seen as making the EITI increasingly obsolete. Survey respondents echoed this view: although preferring the Public Debate model, only two respondents evoked the “public” explicitly when asked in an open question about what they saw as the key steps to achieving the EITI’s main goal, while a few more mentioned civil society involvement or open discussions or communication. Furthermore, when asked in more detail about how effective the public debates had been in the respondent’s country (for those representing a specific country) or in general in countries implementing the EITI Standard (for the other respondents), survey respondents reported that public debates had only to some extent been successful in empowering civil society and citizens and improving resource governance, and—to a small extent—in increasing government’s share of resource revenues.5

According to the interviewees, Technical Reform ToC was often seen as the main process through which change had been observed so far. The recommendations from the auditors and steering committees in Ghana and Colombia had helped in improving the revenue accounting, exposed instances in which lack of capacity and adequate fiscal regulations had hampered revenue collection, and influenced legislation, regulations, and policies concerning the sector. In Ghana, for example, the national EITI (GHEITI) influenced the 2011 Petroleum Revenue Management Act 815, in which many of the transparency and accountability provisions are a result of lessons learned from the implementation of the EITI. Yet, several interviewees noted that the EITI does not directly increase accountability through Technical Reform ToC, as there are no enforcement mechanisms that force states to adhere to the recommendations made by the national EITIs.

In general, there seems to be a mismatch between a “preferred” (or “ideational”) way to effect change (i.e., the Public Debate ToC) and what the reality currently is (i.e., Technical Reform ToC). This was aptly shown in our interviews in Ghana, where nearly all interviewees stated that the Public Debate model is important, relevant, and even the foundational one for Ghana, but few came with concrete accounts of how it had (so far) led to change. In contrast, most interviewees were able to state several concrete examples of how the EITI had led to legislative policy and institutional reforms and improved technical processes. As such, it is possible that the preference for the Public Debate ToC reflects an aspiration to see transformative improvements through public debates over what is considered a public asset (i.e., the country’s extractive resources) and what is hoped for (i.e., socially effective and just redistributive policies for revenue spending), while there is an acknowledgment that other pathways can be more effective in bringing incremental change.

The Missing ToC Link: From Information Disclosure to Accountability

Many interviewees saw the EITI as providing highly factual and credible information that is respected by the different parties, with survey respondents confirming that the EITI in their countries had been successful in collecting and making information public and in increasing transparency in the extractive sector.6 Yet, despite successes in producing and publishing high-quality data on the extractive sector, especially on the revenues, interviewees in Ghana and Colombia generally perceived that the EITI had not had impact on accountability and was thus rapidly losing relevance. In fact, there were several indications of increasing frustration among the civil society representatives. For example, a Ghanaian interviewee argued,

We have done everything in the book that we are supposed to do, and yet, we haven’t been able to make that critical transition from disclosures to accountability … we should be asking ourselves, “What next?” … In that process, we are now focusing our attention on the missing link … [that] needs to be identified and addressed. Because EITI international doesn’t actually have that.

According to interviewees, the main reasons for the lack of accountability are that the information does not reach the public7 and the EITI neither equips the citizens with tools to demand accountability, has power to prosecute, nor outlines a mechanism that would ensure accountability, the result being that the governments are still able to manage the sector and spend the revenues as they see fit. A general observation among many interviewees is that the EITI is ineffective when its findings are not considered by the relevant institutions (e.g. revenue agencies and anticorruption agencies). This can be damaging to Public Debate and Name-and-Shame models, but also for the Technical Reform model in the sense that the EITI could yield reforms if other agencies were responsive.

The EITI thus needs to identify what exactly links its information disclosure to increased government accountability and responsiveness and how the EITI can support these mechanisms, for example, through the EITI Standard requirements. It is necessary that an eventual EITI ToC includes these mechanisms and specifies how they are provided for. Some interviewees proposed that the implementing countries should be required to show progress in how accountable different state institutions have become, whether and how they discuss the EITI reports and the recommendations, and how they respond to the issues through budget, policies, and legislative framework. A key message regarding the Public Debate ToC is to carefully think about who the end users of the disseminated information are, how to reach them, how to provide the information, and on what. Equally importantly, the EITI needs to outline and provide tools that citizens can use to participate in the debate(s) and voice their concerns and demands to their leaders about extractive sector and resource revenue management.

Dynamic, Country-Specific ToCs Needed

Many interviewees stressed that each country would need its own ToC for the EITI: to make a ToC relevant and effective, it has to be adapted to a country’s specific political, social, and legal dynamics. Interviewees expressed much concern that a general EITI ToC model would most likely overlook opportunities suited to some countries but not others (e.g. working through courts when these are independent yet rarely used due to lack of funding). Moreover, as conditions in countries change over time, a ToC would need to be dynamic, which would be difficult to achieve with a general EITI ToC. A general ToC could also hinder countries from moving beyond the expectations set in the EITI Standard if the ToC were too closely tied to these and the national EITI was not able to “grow” to include steps not in the Standard. The interviewees strongly felt that, given the country-specific challenges and objectives for the extractive sector, a “universal” ToC would risk misleading EITI processes into focusing on the wrong issues and being a burden rather than a help. As a Ghanaian interviewee summed up, “[if] it’s not reflecting the reality in the country … [it] makes the initiative quite ineffective. If we just do A, B, C, D and cannot do any other thing, it doesn’t speak to your country’s reality,” while another illustrated the point by arguing that it would be “like condemning yourself into a kind of treadmill that does not necessarily address your own concerns.”

Relevance of the EITI ToC

Several interviewees were very critical about the EITI for failing to address some of the most pressing issues in their country, including environmental and social impacts, small- and medium-scale mining, the expenditure side of the revenue management, and subnational issues. A common understanding was that any EITI ToC that does not address these issues risks making the whole EITI process irrelevant for many countries and their citizens.

A few interviewees, especially from civil society in Ghana and Colombia, saw the EITI as a legitimating tool for “extractivism,” arguing that its focus on revenues only points to the “positive” side of extractive industries, while many social and environmental costs are not considered, and fails to follow a rights-based approach bringing transparency on rights to Free and Informed Prior Consent (FPIC) or to a healthy and clean environment. Civil society representatives who had first thought of the EITI as an opportunity to address extraction-related environmental and social concerns, and thus promote it in its early stages, expressed disappointment about the slow pace and/or limits of the EITI (e.g., to include environmental aspects in the EITI Standard).8 They argued that the inclusion of the environmental and social externalities could increase the relevance of the EITI for a general public whose interest is not necessarily about how revenues are spent by the national government but instead about the extractive sector’s environmental impacts. Artisanal and small- and medium-scale mining was another major issue noted by several interviewees in Ghana and Colombia as not being sufficiently covered by the EITI.9

Revenue-related concerns mostly dealt with the relative absence of the EITI’s engagement over the expenditure side of revenue management. As one interviewee from Ghana pointed out, “the problem is in the spending of the revenues. If the EITI cannot address that, it has failed.” In general, interviewees perceived people to be interested in knowing how revenues are spent and what impact, if any, they make locally and nationally.10 Several interviewees in Colombia and Ghana noted that a more profound inclusion of subnational-level extraction is a major area for improvement for the EITI (see also Van Alstine 2014). According to them, fuller integration of the subnational level in the EITI would enable the governments to show in more detail from where the revenues come and for what the revenues are used, and could improve citizen engagement if the issues were “closer” to them. This could potentially improve public debates at both the local and national levels and include other sections of the public than those in the capitals, with more education, or otherwise in more influential positions within society (Kasimba and Lujala 2019).

Conclusions

As a core norm within global governance, transparency has received much critical attention by scholars, including within global environmental and resource politics. Many studies have pointed at the poor translation of transparency principles into accountability and effective change, calling for a more explicit and detailed theorization of “governance-by-disclosure.” In this study, we mobilized key conceptions of transparency and associated transformative pathways to define three general ToCs expected to link information disclosure with broad environmental and development outcomes. We then applied the three ToCs to examine the case of the EITI. Though limited in its scope given the late and still very limited engagement of the EITI with environmental governance as compared to its focus on resource revenues, this research makes four main contributions.

First, we illustrate that ToCs can usefully identify how transparency leads to better stated goals for initiatives and the design of pathways and programmatic activities to implement initiatives. Second, we identify three main pathways for transparency in environmental and resource governance and detailed them further for the EITI (i.e., Name-and-Shame, Public Debate, and Technical Reform ToCs), again noting that these were focused on resource revenues rather than environmental impacts. Each of these models needs to be further elaborated by identifying intermediary steps, implementing agencies, and mobilizing civil society organizations. Importantly, they should not be seen as linear, separate models but rather as processes that support each other and are interlinked in many different ways. A third contribution is that the elaboration of a ToC needs to consider the national or even local context in which it will be implemented. This would ensure that a ToC reflects the specific contexts, objectives, and capabilities so that it suits local conditions. Finally, our last contribution is to point at the need for a ToC to evolve in order not to become obsolete.

We also note that, in some countries, the EITI’s core mandate has been mainstreamed within other institutions, placing EITI decision makers at a crossroads: they can decide that the EITI’s initial mission is largely over and the EITI thus becoming irrelevant, or they can decide to continue pushing for a broader agenda to increase the full accounting of the many impacts of extractive industries and EITI’s relevance for development, democratization of decision-making in resource governance, and how costs and benefits from the extractive sector are distributed.

More generally, the EITI should be more directly involved in fostering innovation and effective implementation across the key areas of concern where information disclosure and transparency can lead to stronger accountability and improved social and environmental outcomes. Clearly outlined ToC models suited to their national contexts and objectives could help in this regard not only to identify and implement pathways toward improvements but also to assess progress. For that, the EITI ToC needs to be grounded at the central level through common goals, and the EITI Standard should still constitute the basis for the EITI compliant membership and apply to all member countries. The Standard, however, needs to allow for variations regarding the pathways through which the EITI can be implemented and have concrete effects on resource governance and development outcomes. Despite the importance of nationally contextualizing and designing ToCs, the EITI International Secretariat has thus a major role to play through its communication around their importance, assistance in the elaboration of national ToCs, and as a clearinghouse for ToC design and implementation experiences.

Related to the role of the EITI, a fundamental question raised by some interviewees was that of the EITI’s limits in effecting broad changes and thus the scope and ambition of its ToC. There was a sense of realism among some interviewees that the EITI cannot be “the fixer of all problems in the extractive sectors” (a Ghanaian interviewee) and that there is a need to (re)adjust its objectives accordingly, delineating what it can and cannot achieve. These interviewees saw a potential EITI ToC to be of a much more limited scope that what we had proposed (i.e., the three ToC models), not ending at contributing to improved development outcomes but rather stopping at the information disclosure stage. Instead, they felt that a country needs first to identify its own ToC for its extractive sector and then leverage specific EITI processes that can help it achieve its objectives.

EITI’s risk of becoming obsolete also points to the importance of more significantly helping to account for, and reduce the negative social and environmental impacts of, the extractive sector at multiple scales. As such, the EITI should not only help recognize, disclose, and settle the local and national environmental liabilities of extractive projects but also support supply side policies seeking to constrain fossil fuel production and associated greenhouse gas emissions (Le Billon and Kristoffersen 2020; Zalik and Osuoka 2020). By broadening its institutional reach and extending its operational capacity, the EITI could not only help bring greater clarity and accountability regarding the responsibility of producer countries and companies benefiting from fossil fuel extraction but also inform efforts to help them pivot away from an increasingly untenable fossil fuel dependence. While such a move may seem a perilous change of direction for an initiative that has so far mostly promoted itself as maximizing revenue transparency for developmental impacts, we argue that it represents a necessary long-term strategy echoing innovative shifts in environmental and resource politics.

Notes

1. 

This can be due to many reasons. For example, the disclosed information provided may not reach the intended public (e.g., poor information accessibility and dissemination), the public many not find the issue or the information relevant (e.g., low prioritization), or it may not have feasible ways of acting or the state may not be willing or able to respond adequately (e.g., lack of agency and political opportunity structures).

2. 

See, e.g., reports by Scanteam (2011), IEG (2011), and WGTOC (2012).

3. 

On trust, transparency, and (resource) governance, see Le Billon (2014) and Roelofs (2019).

4. 

See the appendix for details on analysis material and methods.

5. 

Survey scale: “Not at all; a small extent; some extent; a great extent; a very great extent.”

6. 

Average scores between 2.7 and 2.8 on the scale 0 (not at all successful), 1 (slightly successful), 2 (moderately successful), 3 (very successful), 4 (extremely successful).

7. 

A national survey carried out in 2016 in Ghana with more than 3,500 participants documented that only one-fifth of the general population (i.e., those with no leader position in their communities) had received information from any source about how revenues from oil, gas, or mining had been handled in Ghana in the past 12 months and that a mere 5 percent of the respondents could recall having heard about the GHEITI (Brunnschweiler et al. 2021; Lujala et al. 2020). One likely reason for the low outreach levels was the use of information channels (i.e., internet, newspapers, and meetings in the regional capitals) that do not reflect the most effective ways to reach people in Ghana, i.e., radio, TV, and local community meetings (Lujala et al. 2020).

8. 

Answering to these calls, the 2019 EITI Standard encourages the countries to report environmental payments by companies to governments and disclosure of contextual information related to environmental monitoring, yet it falls far short of exposing the “full costs” of extraction (see Zalik and Osuoka 2020).

9. 

This has come to the attention of the EITI Board and International Secretariat and was included in the Secretariat’s 2020 Work Plan (Actions 45 and 46).

10. 

A survey conducted by Lujala et al. (2020) also documented that Ghanaians listed information on expenditure at the national and local levels among the key issues on which they would like to have more information.

References

Arond
,
Elisa
,
Anthony
Bebbington
, and
Juan Luis
Dammert
.
2019
.
NGOs as Innovators in Extractive Industry Governance: Insights from the EITI Process in Colombia and Peru
.
Extractive Industries and Society
6
(
3
):
665
674
.
Auld
,
Graeme
, and
Lars H.
Gulbrandsen
.
2010
.
Transparency in Nonstate Certification: Consequences for Accountability and Legitimacy
.
Global Environmental Politics
10
(
3
):
97
119
.
Bauhr
,
Monika
, and
Naghmeh
Nasiritousi
.
2012
.
Resisting Transparency: Corruption, Legitimacy, and the Quality of Global Environmental Policies
.
Global Environmental Politics
12
(
4
):
9
29
.
Bleischwitz
,
Raimund
.
2014
.
Transparency in the Extractive Industries: Time to Ask for More
.
Global Environmental Politics
14
(
4
):
1
9
.
Brunnschweiler
,
Christa
,
Ishmael
Edjekumhene
, and
Päivi
Lujala
.
2021
.
Does Information Matter? Transparency and Demand for Accountability in Ghana’s Resource Revenue Management
.
Ecological Economics
181
:
106903
.
Bumpus
,
Adam
,
Thu-Ba
Huynh
, and
Sophie
Pascoe
.
2019
.
Making REDD+ Transparent: Opportunities for Mobile Technology
.
Global Environmental Politics
19
(
4
):
85
117
.
Chambers
,
Simone
.
2003
.
Deliberative Democratic Theory
.
Annual Review of Political Science
6
(
1
):
307
326
.
Ciplet
,
David
,
Kevin M.
Adams
,
Romain
Weikmans
, and
J. Timmons
Roberts
.
2018
.
The Transformative Capability of Transparency in Global Environmental Governance
.
Global Environmental Politics
18
(
3
):
130
150
.
Dingwerth
,
Klaus
, and
Margot
Eichinger
.
2010
.
Tamed Transparency: How Information Disclosure under the Global Reporting Initiative Fails to Empower
.
Global Environmental Politics
10
(
3
):
74
96
.
EITI
.
2014
.
Making Transparency Matter
.
Progress Report 2014. Available at https://eiti.org/files/documents/EITI_ProgressReport_2014_En_Web_Interactif_r.pdf, last accessed April 19, 2021
.
EITI
.
2016
.
From Reports to Reform
.
Available at https://eiti.org/files/documents/from_reports_to_reforms.pdf, last accessed April 19, 2021
.
EITI
.
2019
.
The EITI Standard 2019
. .
EITI
.
2020
.
Global Factsheet: The Global Standard for the Good Governance of Oil, Gas and Mineral Resources
.
October. Available at https://eiti.org/files/documents/en_eiti_factsheet_2020.pdf, last accessed April 19, 2021
.
EITI International Secretariat
.
2019
.
International Secretariat 2020 Work Plan
. .
Ejiogu
,
Amanze
,
Chibuzo
Ejiogu
, and
Ambisisi
Ambituuni
.
2019
.
The Dark Side of Transparency: Does the Nigeria Extractive Industries Transparency Initiative Help or Hinder Accountability and Corruption Control?
British Accounting Review
51
(
5
):
100811
.
Epremian
,
Levon
,
Päivi
Lujala
, and
Carl
Bruch
.
2016
.
High-Value Natural Resources and Transparency: Accounting for Revenues and Governance
. In
Oxford Research Encyclopedia of Politics
.
Oxford, UK
:
Oxford University Press
.
Fenster
,
Mark
.
2015
.
Transparency in Search of a Theory
.
European Journal of Social Theory
18
(
2
):
150
167
.
Fox
,
Jonathan
.
2015
.
Social Accountability: What Does the Evidence Really Say?
World Development
72
(
August
):
346
361
.
Funnell
,
Sue C.
, and
Patricia J.
Rogers
.
2011
.
Purposeful Program Theory: Effective Use of Theories of Change and Logic Models
.
San Francisco, CA
:
John Wiley
.
Gardner
,
T. A.
,
M.
Benzie
,
J.
Börner
,
E.
Dawkins
,
S.
Fick
,
R.
Garrett
,
J.
Godar
,
A.
Grimard
,
S.
Lake
,
R. K.
Larsen
,
N.
Mardas
,
C. L.
McDermott
,
P.
Meyfroidt
,
M.
Osbeck
,
M.
Persson
,
T.
Sembres
,
C.
Suavet
,
B.
Strassburg
,
A.
Trevisan
, and
P.
Wolvekamp
.
2019
.
Transparency and Sustainability in Global Commodity Supply Chains
.
World Development
121
:
163
177
. ,
[PubMed]
Gillies
,
Alexandra
, and
Antoine
Heuty
.
2011
.
Does Transparency Work? The Challenges of Measurement and Effectiveness in Resource-Rich Countries
.
Yale Journal of International Affairs
6
(
6
):
25
42
.
Global Witness
.
1999
.
A Crude Awakening: The Role of the Oil and Banking Industries in Angola’s Civil War and the Plunder of State Assets
. .
Gupta
,
Aarti
.
2008
.
Transparency under Scrutiny: Information Disclosure in Global Environmental Governance
.
Global Environmental Politics
8
(
2
):
1
7
.
Gupta
,
Aarti
,
Ingrid
Boas
, and
Peter
Oosterveer
.
2020
.
Transparency in Global Sustainability Governance: To What Effect?
Journal of Environmental Policy and Planning
22
(
1
):
84
97
.
Gupta
,
Aarti
, and
Michael
Mason
, editors.
2014
.
Transparency in Global Environmental Governance
.
Cambridge, MA
:
MIT Press
.
Gupta
,
Aarti
, and
Michael
Mason
.
2016
.
Disclosing or Obscuring? The Politics of Transparency in Global Climate Governance
.
Current Opinion in Environmental Sustainability
18
:
82
90
.
Haufler
,
Virginia
.
2010
.
Disclosure as Governance: The Extractive Industries Transparency Initiative and Resource Management in the Developing World
.
Global Environmental Politics
10
(
3
):
53
73
.
Heald
,
David
.
2006
.
Varieties of Transparency
. In
Transparency: The Key to Better Governance?
, edited by
Christopher
Hood
and
David
Heald
,
25
46
.
Oxford, UK
:
Oxford University Press
.
IEG
.
2011
.
Multi-Donor Trust Fund for the Extractive Industries Transparency Initiative
.
Global Program Review
5
(
1
). .
Kasimba
,
Sam
, and
Päivi
Lujala
.
2019
.
There Is No One amongst Us with Them! Transparency and Participation in Local Natural Resource Revenue Management
.
Extractive Industries and Society
6
(
1
):
198
205
.
Knox-Hayes
,
Janelle
, and
David
Levy
.
2014
.
The Political Economy of Governance by Disclosure: Carbon Disclosure and Nonfinancial Reporting as Contested Fields of Governance
. In
Transparency in Global Environmental Governance
, edited by
Aarti
Gupta
and
Michael
Mason
,
205
224
.
Cambridge, MA
:
MIT Press
.
Koliev
,
Faradj
, and
James H.
Lebovic
.
2018
.
Selecting for Shame: The Monitoring of Workers’ Rights by the International Labour Organization, 1989 to 2011
.
International Studies Quarterly
62
(
2
):
437
452
.
Le Billon
,
Philippe
.
2005
.
Aid in the Midst of Plenty: Oil Wealth, Misery and Advocacy in Angola
.
Disasters
29
(
1
):
1
25
. ,
[PubMed]
Le Billon
,
Philippe
.
2014
.
Natural Resources and Corruption in Post-War Transitions: Matters of Trust
.
Third World Quarterly
35
(
5
):
770
786
.
Le Billon
,
Philippe
, and
Berit
Kristoffersen
.
2020
.
Just Cuts for Fossil Fuels? Supply-Side Carbon Constraints and Energy Transition
.
Environment and Planning A: Economy and Space
52
(
6
):
1072
1092
.
Le Billon
,
Philippe
, and
Samuel
Spiegel
.
2021
.
Cleaning Mineral Supply Chains? Political Economies of Exploitation and Hidden Costs of Technical Fixes
.
Review of International Political Economy
.
Lujala
,
Päivi
.
2018
.
An Analysis of the Extractive Industries Transparency Initiative Implementation Process
.
World Development
107
:
358
381
.
Lujala
,
Päivi
,
Christa
Brunnschweiler
, and
Ishmael
Edjekumhene
.
2020
.
Transparent for Whom? Dissemination of Information on Ghana’s Petroleum and Mining Revenue Management
.
Journal of Development Studies
56
(
12
):
2135
2153
.
Mason
,
Michael
.
2020
.
Transparency, Accountability and Empowerment in Sustainability Governance: A Conceptual Review
.
Journal of Environmental Policy and Planning
22
(
1
):
98
111
.
Mayne
,
J.
2015
.
Useful Theory of Change Models
.
Canadian Journal of Program Evaluation
30
(
2
):
119
142
.
Mejía Acosta
,
Andrés
.
2013
.
The Impact and Effectiveness of Accountability and Transparency Initiatives: The Governance of Natural Resources
.
Development Policy Review
31
:
89
105
.
Neumann
,
Lutz
,
Stefan
Silvestrini
,
Tove
Strauss
,
Hilda
Harnack
,
Richard
Dion
,
Sefton
Darby
,
Christian
von Haldenwang
, and
Nikolaus
Roloff
.
2016
.
Assessing the Effectiveness and Impact of the Extractive Industries Transparency Initiative (EITI)
. .
Ofori
,
Jerome J. Y.
, and
Päivi
Lujala
.
2015
.
Illusionary Transparency? Oil Revenues, Information Disclosure, and Transparency
.
Society and Natural Resources
28
(
11
):
1187
1202
.
Rich
,
Eddie
, and
Jonas
Moberg
.
2015
.
Beyond Governments: Making Collective Governance Work. Lessons from the Extractive Industries Transparency Initative
.
Sheffield, UK
:
Greenleaf
.
Roelofs
,
Portia
.
2019
.
Transparency and Mistrust: Who or What Should Be Made Transparent?
Governance
32
(
3
):
565
580
.
Rustad
,
Siri A.
,
Philippe
Le Billon
, and
Päivi
Lujala
.
2017
.
Has the Extractive Industries Transparency Initiative Been a Success? Identifying and Evaluating EITI Goals
.
Resources Policy
51
(
1
):
151
162
.
Scanteam
.
2011
.
Achievements and Strategic Options Evaluation of the Extractive Industries Transparency Initiative
. .
Sovacool
,
Benjamin K.
,
Gotz
Walter
,
Thijs
Van de Graaf
, and
Nathan
Andrews
.
2016
.
Energy Governance, Transnational Rules, and the Resource Curse: Exploring the Effectiveness of the Extractive Industries Transparency Initiative (EITI)
.
World Development
83
(
July
):
179
192
.
Stein
,
Danielle
, and
Craig
Valters
.
2012
.
Understanding Theory of Change in International Development: A Review of Existing Knowledge
.
JSRP Paper 1. Available at https://www.alnap.org/system/files/content/resource/files/main/stein.pdf, last accessed April 19, 2021
.
Sullivan
,
Helen
, and
Murray
Stewart
.
2006
.
Who Owns the Theory of Change?
Evaluation
12
(
2
):
179
199
.
Tienhaara
,
Kyla
.
2020
.
Beyond Accountability: Alternative Rationales for Transparency in Global Trade Politics
.
Journal of Environmental Policy and Planning
22
(
1
):
112
124
.
Van Alstine
,
James
.
2014
.
Transparency in Resource Governance: The Pitfalls and Potential of “New Oil” in Sub-Saharan Africa Transparency in Resource Governance
.
Global Environmental Politics
14
(
1
):
20
39
.
Vijge
,
Marjanneke J.
2018
.
The (Dis)empowering Effects of Transparency beyond Information Disclosure: The Extractive Industries Transparency Initiative in Myanmar
.
Global Environmental Politics
18
(
1
):
13
32
.
Vijge
,
Marjanneke J.
,
Robin
Metcalfe
,
Linda
Wallbott
, and
Christoph
Oberlack
.
2019
.
Transforming Institutional Quality in Resource Curse Contexts: The Extractive Industries Transparency Initiative in Myanmar
.
Resources Policy
61
:
200
209
.
WGTOC
.
2012
.
Report from Working Group on Theory of Change (WGTOC)
.
Available at https://eiti.org/files/DFID ToC.pdf, last accessed April 19, 2021
.
Zalik
,
Anna
, and
Isaac “Asume”
Osuoka
.
2020
.
Beyond Transparency: A Consideration of Extraction’s Full Costs
.
Extractive Industries and Society
7
(
3
):
781
785
. ,
[PubMed]

Author notes

*

We thank Kendra Dupuy, Sophie Lemaître, Michael Ogbe, and Aled Williams at U4 Anti-Corruption Resource Centre for their support and comments, as well as Sunniva Hustad and Bintu Zahara Sakor for their research assistance. This work was supported by the Academy of Finland (grants 322097, 309206), the Social Sciences and Humanities Research Council of Canada (grant 410-2010-2035), the Norwegian Foreign Ministry’s Conflict Trends project (grant QZA-18/0227), and Chr. Michelsen Institute.

Supplementary data