Unintended Consequences or Ambivalent Policy Objectives? Conflict Minerals and Mining Reform in the Democratic Republic of Congo

Since the early 2000s, the government of the Democratic Republic of Congo and the international community have initiated a reform process in the artisanal mining sector to break the linkages between mineral extraction and violent conflict. Research on the reforms reaches different conclusions about their impact on conflict, governance and livelihoods, and usually focuses on one aspect of the reforms, accepting the objectives of the policy at face value. This discourse analysis of the objectives of reform policies aims to understand how they represent the problem, the solutions they envision and their proposed plans. The study applied the method of interpretive policy analysis to four policy documents issued between 2010 and 2012: the UN Group of Experts report, the Dodd&#8211;Frank Wall Street Reform and Consumer Protection Act, the OECD Due Diligence Guidance for Responsible Supply Chains of Minerals from Conflict&#8208;Affected and High&#8208;Risk Areas, and the International Conference on the Great Lake Region (ICGLR) Regional Certification Mechanism (RCM) Manual. The study finds that the policies hinge on two seemingly commensurate objectives, varying between conflict&#8208;free sourcing and promoting peace. We find that, in reality, these objectives may not align. We also find that much reform practice is geared towards conflict&#8208;free sourcing, and is far less appropriate when it comes to promoting peace. This includes the tendency to implement the policies in conflict&#8208;free zones, their narrow scope, the reliance on the government and their indifference to the impact of the reforms for poor miners. The findings suggest that exercising due diligence has become a goal in itself. This raises the question of whether giving buyers a clear conscience and developing a traceable and conflict&#8208;free product has received more attention than improving the situation of the Congolese population.


| INTRODUCTION
Publications on the Democratic Republic of Congo (DRC) often begin with the startling contradiction that one of the richest countries in the world in terms of natural resources is also marred by deep poverty, stagnation, the breakdown of governance and ongoing conflict. Since the early 2000s, it has become common to view this situation less as a contradiction than as an inevitable outcome, with control over precious mineral deposits providing both the resources and the incentive for violent conflict. Congolese armed groups-both rebels and public forces-use the so-called "conflict minerals" trade to finance their war efforts, to enrich themselves and to grease the wheels of patronage. A large number of policy initiatives designed to break the cycle of conflict and the exploitation of mineral resources have changed the character of artisanal mining in eastern DRC substantially. In a country where change often seems elusive, the mineral sector has seen a substantial makeover in just a few years as a result of both national and international interventions. This includes changes in the sector's governance and taxation structures, the introduction of co-operatives and the launch of traceability and certification schemes.
The reforms in DRC's artisanal and small-scale mining sector make an interesting case study of the unintended consequences of development initiatives. The introduction of many initiatives in a short time span has resulted in a significant amount of implementation activity, giving rise to a great deal of research on the effects of the reforms. Since the start of the reforms' implementation in 2010-2011, there have been fierce debates concerning the success, viability and-especially-the (un)intended consequences of these initiatives. Policy-makers and non-governmental organizations (NGOs) have highlighted the positive effects of the initiatives on enhanced transparency and good governance (OECD, 2012). There have also been claims that the initiatives have led to a substantial decrease in militarization, from almost 100% in 2010 to 33% in 2014 (Bafilemba, Mueller, & Lezhnev, 2014). These positive assessments have been contested by other research, which has cast doubts on the militarization commensurate objectives, varying between conflict-free sourcing and promoting peace. We find that, in reality, these objectives may not align. We also find that much reform practice is geared towards conflict-free sourcing, and is far less appropriate when it comes to promoting peace. This includes the tendency to implement the policies in conflict-free zones, their narrow scope, the reliance on the government and their indifference to the impact of the reforms for poor miners. The findings suggest that exercising due diligence has become a goal in itself. This raises the question of whether giving buyers a clear conscience and developing a traceable and conflictfree product has received more attention than improving the situation of the Congolese population.

| 455
DIEMEL anD HILHORST and governance effects and pointed to unintended consequences of the reforms, especially on the livelihoods of Congolese mining communities and small-scale Congolese mineral traders, as will be elaborated in detail in the next section.
Remarkably, the different streams of research on the artisanal mining reforms-whether focusing on intended or unintended consequences-seem to have taken the "intentions" of the policies for granted. This does not mean that the design and assumptions of the reform policies have escaped critical discussion. Several scholars have analysed the problematic design of Western policy interventions in DRC, and on conflict-mineral policy in particular. Autesserre (2012) and Seay (2012) have pointed out misconceptions in conflict-mineral policy on the relationship between the mining sector and conflict in DRC. Other researchers have noted the overly technical approach of the reforms (Geenen, 2012), the erroneous representation of Congo's mining sector as criminal, unregulated and militarized (Koddenbrock, 2012;Nest, 2011) and the frequent and incorrect association of state weakness in DRC with the absence of governance (Radley & Vogel, 2015). However, these critical observations have usually served only to provide a commentary on the design of the policies to introduce empirical work that analysed the consequences of a single intervention from the wide range of initiatives.
In contrast, this article takes the design of these policies as the primary subject of analysis. The study explores a key set of policies, posing several important questions: how do the policies frame the problématique of conflict minerals? How do they view its solutions? What objectives are embedded in the policies? What assumptions are interwoven into the policies about the context in which they are intended and expected to work?
This article rests on the premise that, to discuss the intended and unintended consequences of a policy, it is important to understand its "intentions." We conducted a discourse analysis of four policy documents that have been central in the developments making the 3T mineral trade in DRC more transparent: 1 the 2010 UN Group of Experts report on DRC, Section 1502 of the Dodd-Frank Wall Street Reform and Consumer Protection Act (2010), the Organisation for Economic Co-operation and Development (OECD) Due Diligence Guidance for Responsible Supply Chains of Minerals from Conflict-Affected and High-Risk Areas (2011), and the International Conference on the Great Lake Region (ICGLR) Regional Certification Mechanism Manual (2011). These policy documents form the backbone of many ongoing initiatives and are often cited in the conflict-mineral debate. The present article examines their analyses of the problem and proposed solutions. Finding that these policies are less uniform in their objectives than is usually assumed, we explore how this insight can help us to understand some of the alleged unintended consequences of the policies.
The article also seeks to contribute to the growing literature on the unintended consequences of development programmes. This literature has focused mainly on effects brought about during implementation, when the "interventions intrude into a reacting system" (Sherrill, 1984). These unintended consequences result from human mistakes during planning and implementation or from complexities in the realities of intervention (Koch & Schulpen, 2017). Existing work has taken the "intentions" of policy for granted, and there has been far less attention to the policy side of the equation. What little attention there has been has shed light on hidden intentions, and it has been pointed out that unintended consequences may actually have been anticipated and-although undesired-accepted as inevitable trade-offs or side-effects (Koch & Schulpen, 2017). However, little attention has been given to the inconsistencies, contradictions and lack of clarity in the policies that stem from the complexity of policy-making. This complexity may provide at least a partial explanation for some of the observed consequences. Using a detailed case study of the 456 | DIEMEL anD HILHORST artisanal mining reforms in DRC, the present study further explores unintended consequences by highlighting the importance of viewing policy-making as a social process rather than merely a rational exercise linking clearly defined goals to courses of action (Colebatch, 1989). Policy-making is a complex process, full of multiplicity and contradiction, and these find their way into the wording of the resulting policies. Policy development includes an array of actors with diverging backgrounds, perspectives and objectives, and it is influenced by media, public pressure and current politics. This means research into unintended consequences should become more symmetrical, yoking the examination of interventions to an in-depth analysis of the policies that are being implemented.
The first part of the article introduces the artisanal mining reforms and reviews the literature on their effects. This is followed by a section outlining the discourse analysis methods used to interrogate the policies. We then present the findings from the discourse analysis, before revisiting the literature on the conflict-mineral reforms in light of our findings and proposing an alternative understanding of some of the findings of previous work.

| MINING REFORM INITIATIVES IN DRC
In the early 2000s, a stream of reports linked the protracted crisis in Congo to the country's natural resource sector, especially mining. 2 These reports followed influential literature of the 1990s that maintained that greed (for natural resources) rather than grievance is the key driver of conflict (Berdal & Malone, 2000;Collier & Hoeffler, 2000. Advocacy groups, including The Enough Project and Global Witness, started campaigns on the issue of conflict minerals in DRC in the late 1990s and early 2000s, linking Western consumer products to extreme human rights abuses, including sexual violence in eastern DRC. This pressured governments and multinational corporations to stop the looting of Congolese conflict minerals and to prevent the trade in those minerals from financing the war efforts of armed groups. As a result, parliamentary hearings and multi-stakeholder meetings were organized, and policymakers started to look for ways to break through the conflict cycles between Congo's extractive industry and the proliferation of armed groups. Ensuring responsible mineral supply chains became a major priority of the international agenda in the late 2000s. In the following years, various stakeholders, such as the United States government, international organizations, individual donor countries and the extractive industry developed an array of (a) legislation, protocol and guidelines; (b) initiatives that implement supply chain tracking and certification; and (c) financial and technical support programmes.
These interventions differ in terms of whether they are voluntary or mandatory, whether they are initiated by the private or public sector, and whether they operate on upstream mine sites or aim to regulate trade further downstream in the chain. Figure 1 provides an overview of the initiatives that were developed from 2009 to 2015, targeting cassiterite, coltan and wolframite mineral production and trade in Katanga.
The development of legislative documents and guidelines started with the UN Security Council's Group of Experts report 2010/596 (UN Group of Experts, 2010). Mandated since 2001 to investigate the link between conflict and the illegal extraction of natural resources in DRC, the Group reported on large-scale coltan smuggling from eastern DRC and how Western companies and Congolese warlords were implicated in illicit deals. In 2010, the Group published recommendations on exercising due diligence throughout the supply chain, published as a five-step framework.

DIEMEL anD HILHORST
The promulgation of the mandatory Section 1502 of the Dodd-Frank Wall Street Reform and Consumer Protection Act by the Obama Administration followed in July 2010. The Act required US stock market-listed companies that sourced 3T minerals from DRC or neighbouring countries to verify the origin of these minerals and to exercise due diligence on both their mineral source and the chain of custody to avoid financing armed groups. Despite resistance from private industry, the Act was issued, followed in 2012 by a detailed strategy and a mandatory reporting system. 4 The OECD developed its Due Diligence Guidance on Responsible Supply Chains of Minerals from Conflict-Affected and High-Risk Areas in 2011, with the aim of inducing companies to exercise due diligence to prevent contributing to armed conflict. This document provided a five-step framework. Meanwhile, the European Union (EU) worked on its own regulations, agreeing on the legislative text in June 2016. The EU Conflict Minerals Regulations require importers of raw 3T materials from conflict-affected and highrisk areas (not only Central Africa) to set up a due diligence system before 21 January 2021.

DIEMEL anD HILHORST
To implement these various guidelines and pieces of legislation, numerous initiatives were developed to facilitate responsible sourcing. Among these are initiatives enabling traceability through the supply chain, such as the widely implemented ITRI Tin Supply Chain Initiative (iTSCi), which provided a means for companies sourcing minerals from DRC to prove that their chain of custody did not contribute to armed conflict; the creation of the Centres de Négoce by the UN peace mission MONUSCO and its partners; the establishment of predetermined commercial partnerships between mine site operators, traders, smelters and manufacturers (called the "closed pipeline"); and the launch of the Conflict-Free Smelter Program in 2011, with the aim of facilitating electronic and IT industrial actors to foster social, environmental and economic sustainability.
Another significant type of initiative launched from 2009 to 2011 concerns the certification of mine sites, traders and export lots. These initiatives include the German government-initiated Certified Trading Chains launched in 2009 in the Kivus and northern Katanga, as well as the Regional Certification Mechanism, established by the ICGLR two years later and adopted into Congolese law in 2012, effectively replacing the Certified Trading Chains scheme.
The last type of conflict-mineral interventions are financial and technical support programmes, including the 90 million programmes of PROMINES jointly funded by the World Bank and DFID, aiming to build capacity for strengthened artisanal mining sector management; and the Public-Private Alliance for Responsible Mineral Trade, uniting the United States Agency for International Development (USAID) and private partners to give financial support to new initiatives responding to challenges within DRC's mineral supply chain.

| Consequences of the reforms
As it is not possible to review the consequences of all of the above initiatives in the scope of this article, this section summarizes the main issues raised in the academic literature regarding the consequences of the reforms. Many shortcomings have been noted, glitches and new problems that have appeared in the first five years of the implementation of the reforms. Here, we highlight three major concerns.
With regard to the objective of reducing armed conflict, proponents and critics have not yet found common ground on the positive impact brought about by the interventions. A longitudinal data from an International Peace Information Service (IPIS) 5 study on militarization of 3T mine sites in eastern DRC, does point out a downward trend from almost 100% in 2010 to 33% in 2014 (IPIS, 2015). And proponents of the reforms, like the NGO The Enough Project, attribute these security improvements in eastern DRC to the policy initiatives (Bafilemba et al., 2014). Yet, several researchers have cast doubt on the extent of these positive developments. Although the militarization of Congolese 3T mine sites does seem to have been reduced, IPIS itself has argued this militarization has been displaced to the gold sector (IPIS, 2015). Others have questioned whether such reductions can be attributed to the reforms, rather than broader political and military developments (Vogel & Raeymaekers 2016). Finally, existing research has noted the continued influence of armed groups, whose commanders have found ways to continue operating in the project sites by using decoy companies consisting of family members to manage their day-to-day interests (Cuvelier, Van Bockstael, Vlassenroot, & Iguma, 2014).
Many researchers have criticized the unintended consequences of the reforms on the livelihoods of Congolese mining communities and small-scale Congolese mineral traders. Anticipating the introduction of the Dodd-Frank Act, President Kabila issued a temporary ban on artisanal mining in | 459 DIEMEL anD HILHORST 2010-2011. This heavily impacted the miners and the surrounding economies (Cuvelier et al., 2014;Geenen, 2012;Johnson, 2013;Parker, Foltz, & Elsea, 2017;Seay, 2012). In the years following the ban, a large majority of people living in mining communities in Katanga have continued to experience the negative consequences of conflict-mineral interventions on their lives and livelihoods (Bashwira, 2017;Iguma, 2017). Changing mineral trade patterns have not only led to the withdrawal of international mineral buyers from the Congolese market (Cuvelier et al., 2014), but also impeded Katanga's mining communities' ability to benefit optimally from the mineral extraction and trade (Diemel, 2016). The establishment of "closed pipelines" and the emergence of a near monopsony by largescale mineral buyers have restricted access to mineral trade for small-scale actors and hence reduced their income and negotiation position (Diemel & Cuvelier, 2015;Freudenthal, 2017;Iguma, 2017).
A third criticism concerns the limited pace and scope of the interventions during their first five years of implementation, especially in eastern DRC (Cuvelier et al., 2014;Diemel & Cuvelier, 2015;Radley & Vogel, 2015). In 2015, five years after the initiatives began, only 9% of east Congolese 3T and goldmine sites were incorporated into certification schemes (IPIS, 2015), and only 4% of South Kivu's mine sites were covered by traceability schemes (Radley & Vogel, 2015). Mostly caused by a lack of security and financial resources (Cuvelier et al., 2014), this slow pace of implementation risks the policy being mainly implemented in those provinces least affected by conflict, such as the former province of Katanga (Diemel & Cuvelier, 2015). It also drives artisanal miners working outside of the incorporated mine sites into the margins of legality (Cuvelier et al., 2014).
What do these three main issues (the uncertain outcomes of conflict, the effects on livelihoods, and the pace and scope of implementation) tell us about the reform policies? Do they point to unintended effects and misdirected programming, as the criticisms summarized above imply, or are these issues related to the nature of the reform policies? To answer this question, we conducted a discourse analysis of key policy papers that guided the reforms.

| METHODS
This study analysed four policy documents selected as key documents influencing reform policies, based on a literature review and interviews. These four policy documents formed the backbone for the development of the many existing public and private initiatives and continue to guide much of the thinking in the conflict-mineral debate. They reveal the problematic link between the extractive industry and conflict in DRC, propose a way forward by outlining standards and indicate desired roles to be played by various stakeholder groups. Specifically, the documents analysed are as follows: We used discourse analysis to interrogate these policy documents. Multiple types of discourse analysis have been used to examine political documents (Glynos, Howarth, Norval, & Speed, 2009;Jørgensen & Phillips, 2002). Eminent examples include critical discourse analysis (Fairclough, 1995), which focuses on the role of power and social structure in restricting people's free choice of discursive behaviour, and 460 | DIEMEL anD HILHORST political discourse theory, which was developed by Laclau and Mouffe (1985) to examine how discourse historically evolves and delimits social action. This study used a more recent branch of policy discourse analysis-interpretive policy analysis (IPA) (Fischer, 2003;Hajer & Wagenaar, 2003;Yanow, 2000). IPA emphasizes the importance of interpretation and value, aiming to understand the "production of meaning" in policy and how this meaning becomes communicated through discourse (Yanow, 2000). IPA is particularly relevant for the present study, which examines a recent policy domain that has produced many initiatives by different parties and stakeholders.
IPA was developed as a response to mainstream positivist forms of policy evaluation of the 1960s and 1970s, which perceived social "problems" as readily identifiable and objective in nature and assumed that governments were there to solve or at least react to those problems (Bacchi, 2009). These positivist approaches attributed non-achievement of policy goals and unintended consequences to errors in the policy programme, its instruments or its implementation. IPA, in contrast, questions the discourse of the "policy problem" and thus looks at how the policy problem is framed (Bacchi, 2009;Verloo, 2005;Yanow, 2000). A frame gives meaning to reality and shapes how it is understood (Verloo, 2005). Loosely combining Bacchi's (2009) methodological focus on problem framing and Verloo's (2005) focus on attributed roles in such problem framing, our analysis focuses on the "framing" of the policy problem through the interrogation of the policy documents on three questions:

1.
How is the problem represented? How do the documents "see" the cause of the problem, and why is it a problem? Who has contributed to the problem, and who is affected? 2. What solution follows from this problematization? What needs to be done and why? Which objectives are incorporated in this solution, and how are those objectives prioritized? 3. What is the proposed plan of action to achieve the policy objectives? Which instruments are suggested, and who are the problem solvers?
Apart from a discourse analysis of the four policy documents, the article also draws on empirical research concerning governance and the reforms in the artisanal mining sector in Katanga province conducted by the first author as part of a PhD project. 6 This makes some of the observations specific for this province.

| What is the problem, who causes it and who are the victims?
All of the examined policies highlight the problematic link between the continuing conflict in eastern DRC and the country's extractive industry. Co-ordination and interaction between the various policymaking processes have resulted in alignment and reinforced the policy-makers' representation of the policy problem.
All the documents maintain that armed group involvement in the eastern Congolese mining sector is detrimental to peace because the exploitation of conflict minerals provides armed groups with the financial means to finance their war efforts. The documents also stress that the generation of financial means by armed groups has negative consequences for the security situation and leads to serious | 461 DIEMEL anD HILHORST human rights abuses, including sexual violence (OECD, 2011;US Congress, 2010) and the proliferation of small arms and light weapons (ICGLR, 2011a;UN Group of Experts, 2010).
This link between conflict and natural resources, so central in conflict-mineral policy, has been criticized by multiple authors, who have argued that the policies overestimate the role of natural resources in the Congolese conflict. The UN Integrated Bureau, for example, estimated in 2011 that only 8% of all conflicts in DRC were related to access to natural resources (Autesserre, 2012), though this has subsequently been contested by critics. Others have argued that the militarized mineral trade is a symptom of the conflict and of an ill-functioning state rather than the cause of these things (Seay, 2012). In the view of many scholars, the cause of Congo's conflict is to be found instead in a complex range of factors, such as access to land, migration issues and contestation over traditional power (Nest, 2006). Although presenting the mineral sector as the financial motor of the Congolese conflict might be accurate, it is also highly selective. Armed groups also derive income from other sources, including illegal timber and charcoal production in South Kivu, revenues collected at roadblocks (Seay, 2012, p. 19), and poaching, fishing and controlling bush meat markets around Virunga National Park (UN Group of Experts, 2010).
None of the policy documents under study explicitly argues that the mineral sector is the "root cause," "key driver" or "main financer" of conflict, as Salter and Mthembu-Salter have correctly pointed out (2016). However, it is the only driver that receives such large-scale policy attention, and peacebuilding efforts in DRC may be considered unbalanced as a result. Given the complex nature of conflict in DRC, the effect of mineral reform on the restoration of peace is an open question. The actual effect is likely to be indirect and may take some time to materialize.
A second bias in the policy documents concerns the extensive emphasis on "armed criminal networks," whereas the problematic role of "civil/government criminal networks" is, to varying extents, omitted. This emphasis on armed groups calls attention to the fact that the representation of the problem in conflict-mineral policy is based on characteristics of Congo's conflict-ridden provinces of Orientale (Ituri) and North and South Kivu, which are not representative of other Congolese provinces. In Maniema and the former Katanga, for example, mineral extraction and trade are much less militarized. Most mining communities in these regions are confronted rather with extortive taxation practices and mine pit expropriations by state officials, than issues of militarization. Throughout the eastern and southern provinces, it is not only the personal economic interests of high-ranking FARDC commanders that seriously hinder the mining communities' ability to benefit from the sector (UN Group of Experts, 2010). Additionally, former and current ministers, governors and civil servants of mining services such as SAESSCAM are engaged to a significant extent in extortion or facilitating mineral smuggling (UN Group of Experts, 2010) or are allegedly involved in owning mineral co-operatives or hold shares in large mineral buying houses. The dubious role played by Congolese high officials renders access to mineral trade and exploitation subject to political power struggles. Yet the conflicts of interest deriving from these non-military criminal networks are not addressed by conflict-mineral policy, and the politicians and high officials implicated are not held accountable. Rather, the policy documents mention the Congolese government as an important stakeholder, without taking into account the negative role of these state actors.
With regard to the question of victims, we found an interesting and significant variation in the policy documents. All the documents start from the position that the Congolese population is affected by the problem and recognize their suffering from human rights abuses. However, the OECD Guidance and the UN Group of Experts report also draw extensive attention to the fact that (Western) downstream companies risk contributing to or being associated with conflict and human rights abuses. They implicitly view Western consumers as affected, pointing out that consumers should be able to choose conflict-free products.

Varying formulations of the conflict-mineral policy objective
Following on from the representation of the problem casting the mining sector as an economic driver of conflict, the policies take a broadly economic approach to the peacebuilding problem, searching for a solution through changing the economic value chain of minerals. Conflict funding is to be reduced through the regulation of the mineral trade and the close monitoring of commercial activities in DRC. The origin of minerals must be determined to enable companies to avoid sourcing from conflictaffected mine sites or from dubious sellers, such as FARDC elements. Creating transparency, sharing information and increasing oversight over the mineral supply chain are at the core of all four policies.
On closer scrutiny, however, it is less clear what common objective is supposed to be met by the policies. First, we need to distinguish how the policy documents frame the problem-and consequently conflict minerals-from the discourse broadcast by many NGOs. In contrast to this discourse, conflict-mineral policies do not claim to "end conflict" by means of due diligence or transparency, as Salter and Mthembu-Salter (2016) have correctly highlighted. The UN Group of Experts report, for example, is modest about this expectation, stating that "due diligence on its own is unlikely to stop conflict in the area," although it concludes that "it will have a positive impact" (UN Group of Experts, 2010). Likewise, the Dodd-Frank Act proposes the development of "a plan to promote peace and security," in which monitoring commercial activities is just one element, alongside, for example, "stronger governance that can facilitate and improve transparency" (US Congress, 2010).
But if "ending the conflict" is not the objective of conflict-mineral policy, then what is? Our analysis uncovered distinct formulations of the policy objective throughout the four policy documents. These vary from "reducing human suffering" and "promoting peace" to "avoiding conflict contribution" and "conflict-free sourcing." The varying formulations of the objective are in line with each other, and all four policies seem to adhere in one way or another to these different formulations. Yet, we argue that the four policies differ in their emphasis. Highlighting this difference in emphasis is important, as it may help explain some of the seeming incongruencies of the policy implementation as the different emphases result in differing policy scopes and hence different possible impact. To visualize these specific emphases, Figure 2 depicts the different formulations as a continuum with two extremes, along which the four policy documents are placed.
The Dodd-Frank Act and the ICGLR Manual are placed on the left side of the continuum. Even though both documents certainly pay attention to conflict-free sourcing (the Act through its requirement for US stock exchange-listed companies to verify mineral origin and the ICGLR Manual by proposing among others a certification mechanism), their emphasis is on stopping human suffering and the illicit trafficking of armaments. The Act for example explicitly mentions the "extreme levels of violence in the Democratic Republic of the Congo, particularly sexual-and gender-based violence, and the contribution to an emergency humanitarian situation therein" (US Congress, 2010). The ICGLR Manual highlights "the negative impact of conflict on regional stability" and how "the trade in designated minerals can be directly linked to the illicit traffic in, and proliferation of […] small arms and light weapons" (ICGLR, 2011).

F I G U R E 2 A continuum: Varying formulations of the conflict-mineral policy objective
Reducing human suffering C onflict-free sourcing Dodd-Frank Act OECD Guidance ICGLR Manual UN Group of Experts report | 463

DIEMEL anD HILHORST
In contrast, the OECD Guidance and the UN Group of Experts report are placed on the right-hand side. Although the UN groups of Experts' report very extensively discusses the human suffering of the population in Congo's conflict-ridden provinces, both this report and the OECD Guidance focus on conflict-free sourcing and preventing the contribution to conflict. The OECD document is oriented towards "the importance of responsible sourcing" and the need to avoid "contributing to or being associated with human rights abuses and conflict" (OECD, 2011). And the UN Group of Experts states that avoiding contributing to conflict is important not only for the promotion of peace, but also to avoid "the risk of being considered for sanction" (UN Group of Experts, 2010).
Although the continuum helps us to better understand the different policies' main areas of emphasis, it is not immediately clear what difference this makes in terms of the implementation of the policies. The various formulations are clearly in line with each other, and one could argue that there is coherence in the way that one thing follows from another: contribution to conflict should be avoided in order to promote peace and reduce human suffering. However, we argue that the different emphasis placed by the policies makes a difference, especially when they are analysed in conjunction with how the policies speak of the solutions and the different affected parties they primarily address. Portraying the formulations on a continuum highlights an important issue. It clarifies above all that it is difficult, if not impossible, to say what these policies as a whole prioritize. 7 Findings discussed below indicate that giving buyers a clear conscience and allowing multinational firms to avoid bad publicity may have received greater attention than contributing to the situation on the ground in DRC.

| How will the objectives be achieved?
Following on from the proposed solution, which highlights the need to reduce conflict funding, the course of action proposed by the policies revolves around international requirements for companies to exercise due diligence throughout their mineral supply chains. In this process, a large role is given to the private sector, and the Congolese state is expected to provide oversight.
Together, the reform policies should enable companies to source conflict-free minerals. Traceability enables companies to verify the origin of their minerals, and the International Tin Supply Chain Initiative (iTSCi) is a good example of an instrument that helps companies to map their supply chains and to collect and disclose relevant information necessary to comply with international due diligence standards. In addition, the certification of mine sites, export lots and exporters enabled by the ICGLR Regional Certification Mechanism ensures that companies source from conflict-free mine sites and from suppliers who are not engaged in conflict or human rights abuses. The ICGLR's Regional Mineral Tracking Database and Regional Database of (certified) Exporters (ICGLR, 2011a) provide ways to share and publicly disclose information on regional mineral flows and actors, facilitating the inspection of supply chain data by independent third-party auditors.
However, this process of enhancing transparency takes a narrow approach. It singles out one element of the mining sector, namely commercial transactions within the mineral supply chain, leaving out other important elements of the sector: local mining governance, access provision to mineral exploitation and trade, and the levying of taxes on mining and trade activities. Although the transparency mechanisms are impressive, they do not reach into these other aspects of the sector. Hence, statements such as the one made in the OECD evaluation report of 2012 ("The iTSCi project has considerably enhanced transparency and good governance in the 3T mineral sector") tend to overextend expectations 464 | DIEMEL anD HILHORST of these transparency achievements. Achievements in the supply chain should not be mistaken for advances in transparency throughout the sector as a whole.
Although the focus on transparent value chains sounds positive, there is little consideration of the fact that such a reordering of artisanal mineral supply chains is not a neutral process (Diemel, 2018). This process demands that mine sites be listed and subdivided, demarcated and assigned to appropriate licence holders. It demands that supply chain actors be identifiable, preferably by their full name, government registration number and address. Thus, it can only work with registered, formal actors using payments that are "preferably routed through official banking channels" ((ICGLR, 2011, Appendix 4b). Most artisanal miners and small-scale traders (négociants) who formerly worked in the scattered and informal sector do not fit into this highly organized and rigid supply chain model. Consequently, the demand for oversight and governability leads to the exclusion of many of those actors from the sector, or at least to a deterioration of their position.

| Who is responsible for the implementation?
The policies ascribe a large role to commercial actors, who must exercise due diligence in their own supply chains. Importers and processors of 3T minerals and large-scale buyers downstream in the mineral supply chain have been made responsible for setting up and financing the reform mechanisms. The iTSCi traceability scheme was developed as a joint initiative by the tin industry, and a large part of the funding for the implementation of the scheme in Katanga was provided by Mining Mineral Resources, a large-scale Indian mineral buying company (Diemel & Cuvelier, 2015).
Congolese state authorities have been made responsible for monitoring and providing oversight over mineral extraction and cross-border trade. This requires a permanent presence of public mining service agents in close proximity to the mine sites and mineral depots. Their tasks include the administration of the mineral extraction, trade and export volumes, as well as verifying that mineral loads and associated actors comply with the international standards (ICGLR, 2011, Appendix 4b). State mining authorities are expected "to inspect all mine sites within national boundaries," decide upon the conflict status of each of these mines, maintain a mine site database and issue export certificates (ICGLR, 2011b). To fulfil these tasks, the Dodd-Frank Act and the related US strategy highlight the need to develop "stronger governance and economic institutions" (US Congress, 2010) and propose "increas[ing] the presence of mining authorities in mining sites and along trade routes" (US Secretary of State and USAID, 2011).
The policy documents treat the Congolese government as a legitimate and capable actor playing a significant role in providing a solution to the problematic link between natural resources and conflict. However, this stands in sharp contrast to the recognition of state actors as part of the conflict-minerals problem. The central state has installed and extended local government agencies such as SAESSCAM to execute traceability schemes, but it does not sufficiently support these offices financially (Diemel, 2018). The local offices function autonomously, with civil servants relying on local formal or informal taxes to support themselves. According to Autesserre (2012), Congolese state officials "are responsible for the largest part of all human rights violations." Our fieldwork in the mine sites of central Katanga revealed the involvement of local civil servants, the mining police and intelligence service agents in illegal taxation practices and the expropriation of productive mine pits under the pretext of safety risks. 8 The policies expect the Congolese state authorities to play a positive role in mining sector reform, not taking the way that they function in practice sufficiently into account. While extending state authority might improve the mineral trade administration, in a country plagued with a malfunctioning | 465 DIEMEL anD HILHORST state this approach risks enabling state authorities to become the new predators in artisanal mining areas.
To compensate for the lack of capable state involvement, private mineral buyers also take on a role in assisting the Ministry of Mines in the implementation of traceability and certification schemes at mine sites. At a significant number of mine sites, these private actors pay civil servants' salaries (primes de motivation), provide motorbikes to local state offices and organize the transport of the civil servants-by helicopter or car-to the mine sites. Additionally, these private companies play an active role in the functioning of the system, for example through their participation in local 9 and provincial 10 multi-stakeholder meetings on traceability and certification issues, where information exchange and communication take place between state authorities, private industry actors, local communities and civil society.
One could argue that it is only natural for the private sector to take up such a central role in the development and financing of due diligence, as it is their responsibility to "do no harm" and to prove their responsible sourcing practices to the international community and to their consumers. However, in a country characterized by a malfunctioning public administration (Trefon, 2009), where basic services are usually provided by non-state actors such as churches, international NGOs (Bwimana, 2017), rebel groups (Meagher, 2012;Raeymaekers, 2007) or transnational mining companies (Hönke, 2009;Hönke & Thauer, 2014), the private sector playing such an extensive role risks further privatizing mining governance.
Remarkably, this prominent role is not allocated to the private sector as a whole; rather, it seems to be geared more towards large-scale mineral buyers (comptoirs) than to artisanal and small-scale miners or traders. International requirements and the need for traceability demand that mine sites be listed, demarcated and attributed to a single appropriate licence holder. It demands that supply chain actors be clearly identifiable, preferably with full name, government registration number and address, and for mineral trade to be concluded with "non-cash purchases" and "preferably routed through official banking channels" (ICGLR, 2011, Appendix 4b). These requirements all indicate that the creation of a "clear and transparent" supply chain is not a neutral process. It favours large-scale well organized actors, such as comptoirs and co-operatives, with access to financial infrastructure and international debates, over small-scale actors such as artisanal miners and négociants, who have seen their trading position become restricted and their negotiation position deteriorate since the arrival of these new actors (Diemel, 2018).
This demand for oversight risks paying too little attention to the dispersed nature of artisanal mining and to exclude those actors who are not able to adjust fast enough to these emerging international standards. Taking an artisanal mine site in mind, where hundreds of artisanal miners without fixed address or identity cards extract minerals, one can easily see how those actors do not fit particularly well into this new, highly organized supply chain model. From a due diligence point of view, it makes sense to put larger private actors at the centre of interventions aiming for clearer supply chains.
Yet, what stands out from the documents reviewed is the general absence of this large group of artisanal and small-scale mining (ASM) operators, who might make up over 90% of all supply chain actors, except for those cases where they are represented as marginalized actors, victimized either by the problematic link between mineral trade and conflict financing or by the harsh working conditions and low revenues of the artisanal mining sector.
In line with this trend, policy-makers at both national and international level seem to believe that the gradual eradication of artisanal mining is the solution (US Secretary of State and USAID, 2011). The US strategy linked to the Dodd-Frank Act, for instance, proposes to get artisanal miners out of 466 | DIEMEL anD HILHORST mining activities and into alternative livelihoods in order to reduce their vulnerability. The idea of creating alternative economic activities for these people is understandable. Yet, the reality is that two million small-scale miners (World Bank, 2008)-approximately 14% to 16% of the Congolese population-depend heavily on artisanal mining for their survival.
Pushing towards a disappearance of the artisanal sector while neither the malfunctioning Congolese state nor the financially overburdened donor community seems capable of creating alternative livelihoods for this group of actors risks leaving the bulk of artisanal miners and their dependents without a way of fending for themselves.

| ANALYSIS AND CONCLUSION
A review of the literature on the reforms of DRC's artisanal and small-scale mining sector highlighted three issues of concern: (a) the reforms' insecure and variable consequences for conflict, (b) the continuous effects of the reforms on livelihoods, and (c) the pace and scope of the reforms. We asked how to label these concerns: do they point to unintended effects and misdirected implementation, or are these issues largely embedded in the design of the policy? And what do these concerns tell us about the nature of the reforms? This article aimed to answer these questions through a discourse analysis of four central conflict-mineral policy documents.
We found that although there are some inconsistencies between the policy documents, they seem to problematize the issue of "conflict minerals" in a rather uniform way; "revenues from mineral trade seem to fuel violent conflict." Also, in proposing a solution, the policies unilaterally focus on increasing transparency and oversight over the economic value chain of minerals. Our main finding concerns the ambiguous formulations of the objective identified in the reforms. It appears difficult, if not impossible, to say what these policies as a whole hold as a priority, running from "promoting peacebuilding and reducing human suffering" to "avoiding contribution to conflict and hence sourcing minerals in a conflictfree manner." The policy documents move between these two extremes, with differing emphasis. The Dodd-Frank Act and the ICGLR Manual seem to formulate their objective more in terms of "reducing human suffering," whereas the UN Group of Experts' report and the OECD Guidance's formulation of objectives focus on conflict-free sourcing and exercising due diligence. These differences in emphasis are to an important degree due to the difference audiences the policy documents address. Additionally, one could convincingly argue that the distinct formulations of the objective should be viewed as aligned: one has to source responsibly in order to promote peacebuilding and reduce human suffering.
However, the significance of this ambiguity becomes clear when we combine it with the various ways the policies frame the problem of "conflict minerals," and especially in how these documents define the affected parties. In part, the analysed documents view the Congolese population as the party primarily affected by conflict minerals, whereas the OECD Guidance clearly adds international companies and consumers to the group of affected parties, due to possible conflict associations in terms of their choices and purchases, even if they intend to avoid conflict. There is an additional reason for the objectives to be less aligned than is often assumed. Sourcing responsibly might be thought to lead to the promotion of peacebuilding and reduced human suffering at the hands of armed groups, but research over the last six years indicates that the latter does not automatically follow from the former.
Putting the ambiguity of these objectives in sharp relief, we could ask whether giving buyers a clear conscience and developing a traceable and conflict-free product has not received more prominence than contributing to improving the situation of a large part of the Congolese population. As the policies can thus be 'read' in different ways, this may also explain the different interpretations that we | 467 DIEMEL anD HILHORST find in the literature about the reforms, where some authors consider the reforms effective whereas others point primarily to negative effects.
While all policies express a combination of the various formulations of the objective, many of the points raised in our analysis and review of past work, suggest that the evolving practice is especially compatible with the objective of seeking assurance that minerals in international markets are conflict free, while having a negative effect on the objective of reducing human suffering: • There is a tendency for the reforms to be implemented in areas that are free of conflict, where it is easier to ensure that minerals are conflict free, yet not where implementation in terms of human suffering is the most needed. • The reforms focus narrowly on the transparency of the mineral value chain, rather than on the governance of the mineral sector as a whole, which would potentially have a larger positive effect on the mining communities. • The Congolese state is accorded a prominent role as an important part of the solution, while the dubious role played by many civil servants and high state officials tends not to be considered, nor is it recognized that an extension of the Congolese state apparatus risks new forms of predation for the mining communities. • There is a lack of systematic attention to the deterioration of small-scale artisanal operators' negotiation position and access to free mineral trade, as international requirements for clearer supply chains have proved to be preferential for large-scale actors such as comptoirs and co-operatives than for the former.
The ambiguity of the policy objectives revealed by our discourse analysis suggests that there is an urgent need to revisit the priorities of the policies. Although the policies set out to organize transparency in order to reduce suffering, our findings suggest that exercising due diligence in avoiding the funding of conflict has become a goal in itself. If the reforms have indeed become primarily a process geared towards guaranteeing international companies and consumers access to minerals that have been sourced in a conflict-free manner, this raises the pertinent question of what this means for the objective of promoting peacebuilding and reducing human suffering, which was so central to the advocacy campaigns of the early 2000s that spurred the reforms. This should be a central question in evaluating the intended and unintended consequences of the reforms.