Introduction: differing perspectives on military entrepreneurialism in the DR Congo
‘Faced with a gun, what can you do?’ reads the title of a report on the militarisation of mining in eastern Democratic Republic of the Congo (hereafter DRC) from a well-known conflict minerals advocacy organisation (Global Witness 2009). In the popular imagination, the involvement of the Congolese military (Forces Armées de la République Démocratique du Congo – FARDC) in revenue-generating activities in the DRC's eastern Kivu provinces is associated with violent extortion and forced labour at gunpoint. However, over a year of ethnographic field research in the Kivus between 2010 and 20121 revealed that this image presents only one dimension of a wide repertoire of military economic practices. It appears therefore to be more the product of analysis informed by norms than empirically grounded knowledge.
From the perspective of world military history, the notion that armed forces should be financed entirely through public funds has become established only recently. This norm, which is anchored in Organisation for Economic Co-operation and Development (OECD)-inspired conceptualisations of the state, is at present respected in only a small number of mostly Western countries (Brömmelhörster and Paes 2003). Nevertheless, it deeply informs thinking on the security sector in countries that are the object of state-building efforts. This certainly applies to the DRC, where military entrepreneurialism has been framed in the discourse of the criminalisation and privatisation of the state (Bayart et al. 1999).
These framings have influenced mainstream explanations for this phenomenon, which can roughly be divided into three strands. The first ascribes the FARDC's involvement in revenue-generation to greed (e.g. Bafilemba 2010), echoing neoclassical economic theories of violent conflict that are based on rational choice thinking (e.g. Collier 2000). These approaches draw inspiration from widespread international attention to the business activities of ‘generals-turned-businessmen’ such as former FARDC Chief of Staff Gabriel Amisi (e.g. Fessy 2010), or analyses of criminal networks in the FARDC (e.g. United Nations [UN] Group of Experts 2010). A second strand emphasises the FARDC's organisational culture and values. These are seen to be reflected in ‘impunity’, and believed to be the product of the ‘historical legacy’ of their predecessor forces, in particular the brutal colonial Force Publique and the rapacious military of Mobutu (e.g. Davis 2009, Robinson 2012).. A third line of explanations highlights the non- or underpayment of FARDC soldiers (e.g. Ebenga and N'landu 2005, Rackley 2006), reflecting theories of petty corruption and power abuse in state bureaucracies that emphasise the wage incentives of civil servants (e.g. Besley and McLaren 1993).
These explanations ascribe behaviour to incentives, interests and values either at the level of individual soldiers or that of the military institution. Thus, they fail to take into consideration the crucial interaction between the military and the social context in which the armed forces are embedded. Secondly, they tend to emphasise either socio-economic conditions or norms, values and mentalities, often conceptualised as static phenomena. Such an approach ignores the dynamic interplay between and indivisibility of what Giddens (1984, pp. 28–33) has coined structures of domination, signification and legitimisation. Thirdly, they often uncritically project the victim/victimiser dichotomy onto economic relations between the military and civilians, while paying little attention to how military revenue-generating activities are co-produced and instrumentalised by populations (see also Laudati, this volume). This tendency follows from ‘predatory governance’ thinking, which highlights the parasitic nature of the Congolese state. Such analysis neglects the fact that the Congolese administration does, in fact, deliver some necessary public and private services, and may have some – albeit tenuous – legitimacy (Eriksson Baaz and Olsson 2011, Trefon 2009).
Contrary to analyses of military entrepreneurialism focusing solely on the military, this article's point of departure is the FARDC's position within the social order of the Kivus. This should not be taken to mean that the internal workings of the military are of no relevance. On the contrary, they are an important underlying cause of military revenue-generation. However, it is argued that internal military dynamics provide only a partial explanation for a complex phenomenon that is the product of multiple interacting factors. The approach followed herein draws attention to how military revenue-generation is shaped by militarised structures that are in part (re)produced by the routine practices of civilians. It is not intended to suggest thereby that civilians have no agency, for they certainly contest military power, at times in effective ways. In fact, the set of mostly informal norms that guide the FARDC's revenue-generating practices is the outcome of processes of negotiation, not only within the military, but also between the military and civilian actors (cf. Garrett et al. 2009). As these negotiation processes are strongly shaped by power relations and prevailing discourses and norms, their outcomes often, but not always, contribute to the (re)production of the dominant position of elites. However, in accordance with Giddens's (1984) notion of structuration, it is argued that these processes are ultimately contingent, hence their outcomes also reflect actors' agency.
The remainder of this article is organised as follows. In the first part, a brief overview is given of the historical processes leading to the present state of militarisation of the Kivus' social order. The second part describes a number of the core features of military policies and the FARDC's organisation that contribute to the militarisation of the Kivus. The third part, in which civilians' routine practices and perceptions occupy centre stage, will analyse how the military functions as a public and private regulator. This is followed by a discussion of the military's involvement in private protection arrangements with civilians, and how these relate to civilian–military collaboration in the economic sphere. The article then explains how the FARDC's regulatory, protection and security practices, amongst other factors, affect perceptions of the relative legitimacy of its power position and revenue-generating activities. Finally, in the concluding section, the article reflects upon the implications of its findings for policy initiatives designed to influence the military's revenue-generation practices.
The militarisation of the social order in the Kivus
The explosion of violence that occurred in the course of the two Congo wars (1996–1997 and 1998–2003) triggered a radical ‘militarisation’ of the Kivus. The notion of militarisation is defined herein as the process of the increasing imprint of armed actors and violent modes of regulation, action and thinking on various arenas of society (Bernazzoli and Flint 2009). Militarisation affects political and economic structures on the one hand, and social norms, modes of thinking and identities on the other. This last component can also be described as the spread of ‘militarism’ defined as a discourse (Luckham 1994, p. 24). In the context of the Kivus, militarism is understood to describe the discourses that render violence an accepted mode of human conduct, and that legitimise violent actors' exercise of power. As argued by Jabri (1996), violence as a social practice can become institutionalised, indicating the dominant position of (those supporting) violent actors and of discourses legitimising violence. This does not however imply that militarisation is only an elite project. In line with the theory of structuration, it is posited that it is in the routines of everyday life, the day-to-day social activities and encounters of both the military and civilians, that militarised structures are (re)produced. Therefore, militarisation should not be seen as the one-sided imposition of violent actors on society. Rather, it is a process of institutionalisation that is ‘both a medium and outcome’ of the actions of all agents within a social system (Giddens 1984, p. 25).
The Congo wars triggered the rise of a new class of violent actors. These built upon existing social networks, which were consequently transformed and militarised (Vlassenroot and Raeymaekers 2004). In order to establish control, these violent actors needed a modicum of collaboration from civilians. As outlined by Kalyvas (2006, p. 101), civilian collaboration with armed actors in situations of violent conflict is the outcome of ‘a dynamic, shifting, fluid and often inconsistent confluence of multiple and varying preferences and constraints’. One of the most powerful factors within this confluence is undoubtedly coercion, as its impact is not only direct, but also heavily shapes strategies of future action. However, other factors promoting collaboration with armed actors are more related to preferences and opportunities. These include possibilities for income-generation, the obtaining or maintaining of positions of status and authority, and access to public goods and private services. Despite this wide range of motives, the majority of people are strongly driven by ‘survival maximising behaviour’ and display weak preferences (ibid., p. 103). This could clearly be observed in the DRC, where many civilians opted to collaborate with armed actors for purposes of self-preservation or in order to obtain certain benefits. Consequently, interacting and cooperating with armed actors became a part of everyday life and was turned into a ‘routine practice’ (Giddens 1984, p. 60). As the rules and resources that agents draw upon in enacting routine practices are ‘at the same time the means of systems reproduction’ (ibid., p. 19), this routinisation became a crucial element in the (re)production of the militarised order. Civilians' increased collaboration with armed actors was enabled by and led to shifting definitions of licit wealth (cf. Roitman 1998) and transformed representations of economic organisation (Jackson 2002). These processes were already under way in the second half of the 1970s, when the contraction of the official economy and the institutionalisation of ‘la débrouillardise’ or ‘fending-for-one-self-ism’, ensured that trickery, cunning, coercion and deceit became acceptable modes of wealth creation (Trefon 2002, Nzeza Bilakila 2004). The major transformation of the Kivus' economy during the Congo wars, and the ensuing decline in livelihood opportunities, further legitimised unscrupulous and coercive income-generation. Rampant insecurity reinforced the isolation of far-flung rural areas and made markets inaccessible, leading many communities to retreat into subsistence farming. Others were pushed towards the exploitation and trade of natural resources, a sector that was almost entirely controlled by militarised networks. At the same time, youngsters enrolled in militias or turned to violent crime as a livelihood strategy (Vlassenroot 2004). Within this economy of despair and opportunism, violent modes of appropriation and economic collaboration with armed actors became institutionalised practices.
Parallel developments can be detected in the domain of politics. Well before the wars, violent strategies of identity politics were becoming an increasingly widespread mode of obtaining and projecting political power (Jewsiewicki 1998). The process of attempted democratisation in the early 1990s unleashed a vigorous power competition that occasionally took on violent forms. This was most powerfully demonstrated by the outbreak of intercommunal violence in first Katanga and later North Kivu (Mamdani 1998). Militarisation in the war era built upon, accelerated and intensified these existing processes of institutional and discursive change. In the Kivus, armed groups established their own forms of rule and acquired control over a large part of the administrative apparatus, exercising fiscal, justice, policing, and other governance functions. As a consequence, the nature of politics and authority was profoundly reshaped. While the position of elders was undermined by younger generations of armed actors, the majority of local authorities and politicians saw few alternatives to collaborating with armed actors in order to obtain or maintain influence. These processes were both facilitated by and led to shifts in representations of legitimate authority (Vlassenroot 2004, Jourdan 2004).
Through the power-sharing peace agreement concluded in 2002, parts of the militarised political–economic networks that had risen to prominence in the course of the wars were integrated into the state apparatus, including within the armed forces. This enabled some factions to translate wartime gains into post-settlement influence, which helped legitimise their power position. At the same time, the ways in which power-sharing unfolded skewed incentive structures towards the continued use of violence (Tull and Mehler 2005). A policy of ‘many carrots and few sticks’ towards dissident rebel factions and political entrepreneurs who resorted to coercion helped ensure that violence continued to be rewarded, specifically in the Kivus (Eriksson Baaz and Verweijen 2013, forthcoming). As a consequence, violence and collaboration with violent actors remained institutionalised practices. Therefore, the structures of militarisation that had formed during the wars were not significantly transformed. What did change in many places, however, was the identity of the dominant violent actors, as several former rebel-held areas now came under control of the government forces.
Military characteristics and policies contributing to militarisation
In 2003, as a central element of the peace agreement, all ex-belligerents' fighting forces were theoretically placed under the unified command of the newly created FARDC. Furthermore, a process was initiated to gradually merge all troops into new units, while breaking down war-era command structures. However, this so called ‘brassage’ (literally, brewing) process failed to produce a coherent military under a single command chain. As many of the armed factions strove to retain control over established economic networks, spheres of influence and constituencies, they either sabotaged or manipulated the integration process in their own interests. While parts of certain groups refused outright to integrate, others were reluctant to redeploy their troops out of the Kivus (Wolters and Boshoff 2006). Over the years, several of the brassage dodgers reconstituted themselves as armed groups, and their numbers were boosted by FARDC deserters dissatisfied with their ranks and positions. Some of these armed groups however subsequently reintegrated, as the process of military integration was never declared to have been completed.
The ongoing (re)integration of armed groups has had four important effects on the FARDC. First, as shall be explained further below, it has reinforced the army's involvement in revenue-generation. Second, it has contributed to the proliferation of parallel power structures and command chains, as integrated groups continue to function as separate networks within the army (Eriksson Baaz 2011). Third, rebel-military integration has in practice reinforced centrifugal tendencies, given that many integrated factions still maintain close contacts with political–commercial and armed actors outside the military. This is in part the result of the lack of redeployment of integrated troops out of the Kivus (Eriksson Baaz and Verweijen 2013, forthcoming). Sources in the UN estimate that in 2011, over two-thirds of the military deployed in these provinces were native. As locally recruited troops often retain strong links to local social networks, this strengthens the FARDC's involvement in political, economic and social regulation. Such involvement is further facilitated by the fact that the majority of FARDC troops do not reside in military barracks, but instead live among the civilian population (Verweijen forthcoming 2013). A final effect of rebel-military integration as it has been implemented in the Kivus is that some army units are not properly balanced, with a majority of troops composed of members of a particular rebel faction. This phenomenon surfaced after the rapid integration in 2009 of the National Congress for the Defence of the People (Congrès National pour la Défense du Peuple – CNDP) and some smaller armed groups. The deployment of such unbalanced units may aggravate intercommunity tensions and place strains on civilian–military relations, specifically where conflicts are framed as identity-based (International Crisis Group 2010). Given these complex power dynamics and variations in composition, the FARDC is heterogeneous in its make-up, and substantial differences can be observed between one unit and the next (Verweijen forthcoming 2013). Therefore, the observations on the FARDC made in this article should be taken as general patterns rather than applying to the army en bloc. Despite this, what all parts of the military do have in common is that they are strongly shaped by the logics of ‘Big Man networks’ (Utas 2012). These networks operate throughout the Congolese military, intersecting, overlapping and conflicting with the formal hierarchy (Eriksson Baaz and Stern forthcoming 2013). As their functioning is based upon the convertibility of economic resources into political ones and vice versa, the Big Man logics that shape these networks generate a drive for the accumulation of wealth and control over income-generating opportunities. This leads to the use of access to office as a political resource, creating pervasive insecurity of tenure (Bayart 2006). Two important ways in which this insecurity is created within the FARDC are first, the frequent restructuring of the military organisation, and second, the regular redeployment of particular units (Verweijen forthcoming 2013). Both these processes create an ongoing risk that military personnel will find themselves in less lucrative positions and zones. As a consequence, commanders frantically try to reap the benefits of their position as long as it lasts. The fluctuations in deployments and power distribution that result from these dynamics are not only an internal military affair, for they strongly shape and are shaped by external political developments. As the Big Man networks that military elites are part of exceed the boundaries of the FARDC, political, military and economic power in the Kivus are closely intertwined.
As elsewhere, the power of the political–commercial elites who control the Congolese state apparatus is ultimately founded upon their capacity to mobilise coercion. The military, as well as other (in)security services, have always been a key instrument for the exercise of that coercion (e.g. Schatzberg 1988). Furthermore, as will be described later in more detail, political–commercial elites depend upon armed actors for revenue-generation and vice versa. Therefore, there are mutual dependencies between FARDC officers and political–commercial elites. This pattern is reproduced at every level of governance, from the local to the provincial to the national. In the present-day Kivus, it leads to considerable instability as it creates close interactions between ongoing political, economic and military power struggles. The resulting volatility weakens Kinshasa's grip over the Kivus, an area it has historically ruled more via intermediaries than through direct administrative intervention (Tull 2005).
While they are partly dependent upon civilian authorities, there is a constant threat that military elites may become more autonomous of their civilian counterparts or switch sides to different power networks. As a result, incumbent politicians do little to rein in top officers' business activities, hoping to buy off their loyalty in this way (Howe 2001). For the government, this approach also reflects its need for loyal intermediaries who can exercise influence over the Kivus. A similar logic applies to the armed groups that have been co-opted into the incumbents' power network by means of their integration into the FARDC. Most of the time, this co-optation involves the provision of material rewards. The cheapest solution for the cash-strapped government is to make these pay-offs self-financing, by granting integrated armed groups access to revenue-generating opportunities. As a consequence, these groups' continued integration in the military has in some cases become dependent on their ongoing access to such pay-offs (Eriksson Baaz 2011). This makes it difficult to curtail, let alone end, such revenue-generating activities, especially since integrated groups are in a position to reconstitute themselves relatively easily as militarised networks outside the bounds of the state. In this manner, the interdependencies between military and civilian elites further foster military revenue-generating practices, while entrenching the militarisation of the Kivus' social order.
The FARDC as a public and private regulatory authority
Given the porous boundaries between military, political and economic power, as well as the close connections between military and civilian actors, the military's economic practices cannot be properly understood in isolation from the FARDC's position within the Kivus' social order. Of particular importance in this respect is the FARDC's role in social, economic and political regulation of both a private and a public nature. As will be demonstrated below, ‘public’ and ‘private’ regulation are best conceptualised not as binary opposites but as two ends of a spectrum, with many hybrid forms in between. Therefore, some of the underlying mechanisms are the same. One of these is the weakness of local civilian authorities, which can be identified as both a cause and a consequence of the militarisation of regulatory authority.
Civilian authorities' legitimacy and capacities, already heavily eroded before the Congo wars (Van Acker 2005), have not been significantly reinforced in the post-settlement era. In most of the zones where the FARDC has taken over control from non-state armed groups, administrative infrastructure and civilian mechanisms of governance have remained weak (Vlassenroot and Romkema 2007). In outlying rural zones in particular, where the FARDC easily outnumbers the understaffed state administration, civilian authorities continue to have relatively little influence. Local administrators and customary chiefs excel in ‘tracasseries’ (harassment) and are often seen as partisan, corrupt or ineffective. Furthermore, police forces are scarce and underequipped, and their services are expensive, as people have to pay ‘makolo ya leta’ or ‘transportation costs’ for every case. Finally, customary and other local tribunals lack both accessibility and legitimacy, and many of their verdicts are never implemented, as unfavourable judgements are simply ignored (Scheye 2011). As a consequence, justice and dispute resolution are exceedingly complicated and problematic processes, involving negotiations with many different parties and often fuelling existing or creating new conflicts.
The limited legitimacy and capacities of both state and non-state civilian governance actors have increased the propensity of civilians to turn to violent actors, often the FARDC, for forms of regulation. This is a two-way process, as the weakness of civilian authorities allows armed actors to impose themselves with varying degrees of force. In all sites where field research was conducted, it was observed that the FARDC was involved in conflict resolution and arbitration, policing, economic regulation and other governance practices. Since the legal boundaries between the official mandates of different authorities are little known, and often considered to be of little relevance to large segments of the population, civilians engage in ‘forum shopping’ (von Benda-Beckmann 1981). That is, they solicit intervention from the authorities they consider to be most powerful, accessible, legitimate, understandable and effective (Baker and Scheye 2007). A police officer in Lubero territory explained that he could not compete with the military in terms of dispute settlement, as the FARDC offer cheaper, faster and more effective solutions: ‘The law takes time and is uncertain. Violence is fast and easy. You pay the FARDC and you get what you want. And many people want to punish their wrong-doers.’ At the same time, soldiers themselves are ‘shopping forums’ (von Benda-Beckmann 1981), as they usurp jurisdiction over and manipulate disputes from which they expect to gain political advantage and income. While ‘shopping’, the FARDC is often in fierce competition with other state services, for whom ‘fees’, ‘motivations’ and ‘fines’ (paid, for example, for being released from arbitrary detention) constitute a considerable source of income.
Referring to the plethora of solicited and unsolicited tasks he fulfils within his area of responsibility, an FARDC sector commander described his social function as that of ‘Mwami sans frontières’ or ‘customary-chief-without-borders’. As examples of the diverse duties he carries out, this commander mentioned intervening in a conflict about the succession of a local chief; guaranteeing the security of local football matches and the weekly markets; arresting cow herders accused of having poisoned cattle; throwing the losing party in a heritage dispute off family-owned land; and demanding that the son of a notable pay his dowry. As the boundaries between the public and the private are blurred in the DRC, as is characteristic of hybrid orders (Erdmann and Engel 2006), the cases for which FARDC intervention may be solicited include the violent settling of scores (e.g. revenge and rivalries) and accounts (e.g. debts and land disputes) arising from envy and personal or family disputes. Examples are conflicts over heritage, marriage, debts, real estate, love affairs, land, power positions and past wrongdoing. Either upon request of the ‘client’, or in order to effectively fulfil the task and benefit most from it, such private dispute resolution may entail FARDC intimidation, the harming of its clients' opponents, or the use of force to appropriate goods or plots. In such situations, the FARDC turns into ‘private guns’ and ‘enforcers-for-hire’, as Schatzberg (1988, p. 60) called these functions in relation to the Zairian military.
This confirms Kalyvas's (2006, p. 14) observations that civilians ‘use political actors to settle their own private conflicts’, highlighting the joint production of violence in situations of civil war. The resulting ‘privatisation of politics’ makes it difficult to identify the discursively constructed and continually shifting boundaries between private and public. Furthermore, the involvement of armed actors in private regulation often produces public effects, particularly when private disputes concern members of antagonistically defined groups. In such situations, perceived ‘unjust’ adjudications and interventions may trigger communal tensions. For example, in Fizi territory, it was observed how the intervention of an FARDC officer in a private conflict over debt between two individuals from groups with longstanding hostile relations (a Mubembe and a Munyamulenge), created tensions in all of the surrounding villages. The Mubembe individual felt disadvantaged by the FARDC's attempt to collect the debt from him, particularly since the officer involved was a Rwandophone seen as favouring the Banyamulenge, another Rwandophone group. He therefore threatened to mobilise a local Mai Mai group in order to take revenge. This example illustrates how the regulation of interpersonal conflicts can have spin-off effects in the public sphere. Inversely, violent conflict can create private grudges and feuds, particularly as former combatants responsible for past abuses go unpunished, whilst residing in local communities. Therefore, the boundaries between public and private regulation are not always easy to draw, specifically when also taking the indirect and longer-term effects of actions into consideration.
Private protection arrangements in a context of institutionalised insecurity
In many cases, the forms of regulation provided by the FARDC are part of private protection arrangements that are sometimes embedded in the Big Man relationships that are a crucial part of the functioning of the Kivus' social order. Individuals at every level of society seek patronage from influential persons in exchange for loyalty, support and various favours. A major driver of this process is the high level of everyday insecurity faced by Congolese citizens (cf. Erdmann and Engel 2006). Two important causes of this insecurity are the unpredictability of prebendalistic state services and the existence of multiple regulatory and normative frameworks (Trefon 2002). In the specific context of the Kivus, it is also produced by the presence of a wide range of armed factions and widespread banditry. The resulting pervasive insecurity drives people to seek private protection from Big Men who also dispose of the means of coercion. Such protection may be provided either on a transactional basis, involving direct remuneration or other compensation, or be part of longer-term relations that include more comprehensive forms of services and favours, such as brokerage and influence-trafficking. Some examples of private protection services provided to civilians by the FARDC are physical protection (e.g. guarding houses and private property), property rights enforcement (e.g. collecting debts), influence-trafficking (e.g. arranging tax exemptions, exploitation licences, tenders), and shielding or facilitating illicit activities (e.g. protecting brothels against closure, smuggling). Sometimes, private protection is primarily imposed by the FARDC and assumes forms resembling extortion. However, in other cases, civilians actively solicit FARDC protection, as its services are seen as particularly effective in a number of respects. First, military networks can offer physical protection, a highly coveted good in a context of rampant insecurity. Second, both through coercion and its contacts with the political–commercial elite, the FARDC has a significant influence on political and administrative decision-making. This allows them to shield protégés against administrative harassment or control from other state services. Finally, through their access to the means of coercion, the FARDC can offer violent and therefore relatively effective enforcement, appropriation and account-settling services.
Private protection in the framework of clientelistic relations is of specific importance for economic operators. Most of the economic transactions in the Kivus are not fully regulated by formal institutions, and the extent to which they are is commonly a matter of negotiation (Rubbers 2007). Due to this relative absence of (the application of) formal rules, alternative mechanisms of property rights enforcement apply. The smooth functioning of these mechanisms depends either on trust or, where trust is scarce, on private protection provided by coercive actors (Gambetta 1993). As trust among entrepreneurs in the DRC has plummeted from the Mobutu era onwards (Rubbers 2009), economic operators in the Kivus have increasingly solicited protection from violent actors (Raeymaekers 2007), including for the enforcement of agreements, dispute settlement and debt collection.
Military protection services are provided on such a large scale that they have infused the workings of the Kivus' entire social order. This has created a self-reinforcing spiral, as those who do not enjoy military protection feel they are at a comparative disadvantage (Gambetta 1993). For example, through their presence at border-crossings and their influence on customs and other authorities, the FARDC can import and export goods at heavily reduced tariffs (Tegera and Johnson 2007). Many cross-border traders make use of their connections with senior army officers in order to obtain these reductions, leaving their competitors with comparatively lower profit margins. The same logic of comparative (dis)advantage applies to the domain of transport. Those operators who arrange to transport their goods in vehicles owned or protected by the military, have a lower risk of incurring losses through ambushes and robberies. This incentivises other businessmen and women to solicit similar protection. Even at lower levels, this competitive mechanism of protection is at work. In the town of Misisi, a gold mining site in Fizi territory, there is a chaos of ‘étalages’ (displays for merchandise) in the residential quarters. However, officially, goods can only be sold at the market or along the main road. Whereas most of these illegal étalages are owned by wives of soldiers, other civilian women with connections with the FARDC also engage in this practice. It not only allows them to avoid paying the taxe d'étalage at the market, but also gives them a better place for selling goods, as they can directly target passers-by.
In many cases, military protection takes place in the framework of more comprehensive economic collaboration between the military and civilians. Some of the features that make the FARDC attractive as a provider of protection also make it an attractive business partner. Additionally, the FARDC possesses specific qualities as an economic actor. As well as being a combat organisation, the FARDC is an economic network with wide geographical coverage. It possesses a labour force, infrastructure such as communications and transport systems, and has access to large information flows. Furthermore, it is either legally or de facto exempt from taxation, and possesses symbolic capital due to its status as a state actor (cf. Brömmelhörster and Paes 2003). Finally, senior officers have access to investment capital and cash, both of which are highly coveted in an environment where banking infrastructure is almost completely lacking and access to credit exceedingly restricted. These characteristics make it attractive for civilians to do business with the FARDC. Due to the resultant widespread collaboration, civilians and the military have become highly economically interdependent. Given that power in the Kivus is to a large extent obtained, maintained and projected through the accumulation and redistribution of (access to) resources, this has strengthened the FARDC's power position.
The above section has demonstrated that the FARDC's economic activities are enabled by and embedded in their public and private regulatory and protection practices. These practices are shaped by a complex mix of preferences and constraints, demands and offers, persuasion and coercion, both among the military and civilians. They are not only the result of the FARDC's power position, but also help construct it. One way in which they do so, as will be explained in the following section, is by bestowing a certain degree of legitimacy on the FARDC's power.
The relative legitimacy of the FARDC's power position in the Kivus
The field research revealed that perceptions of the FARDC's legitimacy vary considerably across time and space. Such perceptions are intersubjectively constructed by a set of mutually reinforcing factors. Logically, views on the legitimacy of the military's power strongly influence and are influenced by evaluations of the FARDC's economic practices, specifically when these are framed in the discourse of public security. According to Beetham (1991), the legitimacy of power is determined by three elements: first, its conformity to established formal and informal rules; second, the justifiability of these rules in terms of beliefs shared by both dominant and subordinated groups; and third, the demonstrable expression of consent on the part of subordinates, specifically through public actions. In the following section, each of these three dimensions will be explored.
Unsurprisingly, it is fairly easy to encounter civilians in the Kivus who are highly dissatisfied with the FARDC's power position and economic practices. In many cases, their comments reflect an awareness that the FARDC's prerogatives and practices transgress official rules and regulations. Civilian authorities complain about constant unauthorised interference with their work. One locality chief in Fizi territory stated: ‘When they were deployed here, we, the customary chiefs had no work. We simply couldn't do our work.’ Economic operators widely denounce the FARDC's irregular involvement in business activities. For example, a member of the Federation of Congolese Enterprises in Uvira stated: ‘They never pay taxes, they are above the law.’ While farmers throughout the Kivus question the right of the FARDC to levy taxes at the market, fishermen likewise declare the Navy's appropriation of a fixed share of their catch illegal. Yet, despite this widely circulating discursively articulated awareness of law-breaking, the picture of the FARDC's legitimacy becomes much more complex when its authority is placed in comparative perspective, and analysed in relation to the informal norms that are enacted in civilians' everyday practices. Given that vast swathes of the Kivus are under control of armed groups, the FARDC's authority may seem comparatively more legitimate than that of non-state armed actors, although patterns are mixed. The FARDC is strongly associated with the state and the government. It is popularly called ‘jeshi ya serikali’ (‘army of the government’) or simply ‘serikali’ (‘government’), in part to distinguish them from non-state armed forces, generally called ‘jeshi’. This gives it an undoubted edge of legitimacy over these latter forces. Although people in some field research sites declared a preference to the presence of non-state armed forces, in the majority of zones it was the opposite (see also Oxfam International 2012). When inquiring about the reasons for preferring FARDC presence, people often invoked precisely its status as a state actor. As a woman from a village in Uvira territory explained: ‘If something happens, in case of the army we can go to the government, but in case of the rebels, there is nowhere we can go.’ Furthermore, the FARDC does not always compare negatively to civilian authorities, specifically since the legitimacy of the latter has sharply diminished due to what Vlassenroot has called a ‘general crisis of authority’ (2004, p. 56).
Other important elements of legitimacy concern people's representations and expectations of authority as well as their beliefs about the (common) interests that power should support. These issues are part of Beetham's (1991) second dimension of legitimacy, the justifiability of the rules of power in light of shared beliefs. In the context of the DRC, as in other hybrid orders, expectations of authority are influenced by the specific notions of reciprocity and representation that characterise Big Man ties. These do not only turn on the (re)distribution of spoils, but are also related to the demonstration of largesse (Chabal and Daloz 1999). In certain contexts, such expectations impact on the perceived legitimacy of FARDC commanders' power. This especially applies to commanders' home areas, or where they have been deployed for a long time and have built up a large network of clients. For example, the commander of the 323rd Brigade headquartered in Kiomvu, in Mwenga territory, heavily sponsored the Kasmes football club in the capital town of Kamituga. This raised his popularity and standing, thereby facilitating the wide range of business activities in which he was involved.
This example indicates that the relative legitimacy of the FARDC in the eyes of civilians is heavily shaped by beliefs and expectations concerning its tasks and performance. Of particular importance is the extent to which it provides needed public and private services. These services not only concern its role as regulator and private protector, but also its security duties. Logically, the expectations concerning what public and private goods the FARDC should provide differ between social groups. Not all civilians benefit from the private protection arrangements that the FARDC may establish with elites. Yet in certain cases, more private arrangements also generate public effects. For example, where the FARDC improves road safety as a result of a protection agreement with local traders, or in the framework of mutually beneficial commercial relations, wider layers of the population also benefit. These effects can be conceptualised as ‘positive externalities’, which derive from the status of some forms of protection as an indivisible good, from which it is difficult to exclude people (cf. Gambetta 1993, p. 31).
In general, the field data demonstrated that, for the majority of non-elite civilians, in particular the FARDC's perceived performance in the domain of security has a very strong impact on the experienced legitimacy of its power position. This legitimacy also strongly influences perceptions of military revenue-generating practices. Especially where these are justified as contributing to the performance of the FARDC's security duties, or where they enhance people's own livelihoods opportunities, they can come to be seen as relatively licit. For example, roadblocks are less resented in areas where banditry abounds and where the FARDC is believed to reinforce security: ‘Better pay 500 Francs Congolais [FC] to the military than have all your belongings looted by the FDLR’, as a small-scale trader stated at a roadblock in a forest in Fizi that is infamous for frequent ambushes. The importance of security for shaping perceptions of the legitimacy of the FARDC's power position and revenue-generating practices is also illustrated by the impact on such perceptions of the levels and types of coercion used by the military. Logically, where the FARDC engages in violent forms of extraction generally evaluated as illicit, such as ambushing, robbery and looting, its overall legitimacy sharply diminishes.
However, contrary to the ‘faced with a gun’ image depicted in the introduction, the vast majority of military economic practices do not involve direct (threats of) violence. Rather, intimidation is much more subtle, concerning mostly implicit or concealed threats that draw heavily on the imagination, anticipation and strategic calculations of the intimidated. The latter generally believe that ‘the military never forgets’ or that ‘once they meet this soldier on an isolated stretch of road, they will experience an “accident”’. Therefore, they bargain with caution or may refrain from contestation altogether. Another often-advanced belief can be characterised thus – ‘You never know, maybe you once end up in a situation where this soldier is the only one to help you’ – reflecting a desire to keep as many options open as possible. It therefore appears that civilians exercise a type of self-censorship in their dealings with the military, both as they try to manage the current situation and with an eye to the future. This makes it difficult to determine whether people comply out of fear, custom, expedience, strategic calculations or simply because they consider certain claims of the military to be legitimate.
No matter what motives drive them, according to Beetham (1991), public acts of compliance ultimately contribute to making power legitimate. They possess both symbolic and normative force, as they have declaratory power and create mutual normative commitments. When studying the day-to-day routines of civilians in the Kivus and their everyday interactions with the military, it becomes clear that, in general, public compliance with the FARDC exceeds public contestation. This is not to say that the FARDC's power is not openly contested, nor that such contestation is ineffectual. On the contrary, the military's power ultimately has a negotiated character, as it is circumscribed by the power of others. The very complexity and fluidity of the Kivus' social order, the multitude of stakeholders and registers and the transitory and fluctuating nature of power and alliances prohibit absolute domination and crushing oppression. This is a system that creates room for manoeuvre for both the dominant and the subordinated, as both can instrumentalise its disorder (Chabal and Daloz 1999).
The negotiated nature of the FARDC's power is indicated by the substantial variations in behaviour and revenue-generating practices that can be observed between specific military units in various deployment contexts. Analysing the wide range of factors that explain these differences, it was found that civilian practices of contestation are commonly of importance, although not always in a stand-alone fashion (Verweijen 2013, forthcoming). The small town of Kazimia, at the coast of the Tanganyika Lake in Fizi, presents a striking example at the micro level. At the time of visit, Kazimia had a relatively powerful and respected chef de poste (local administrator) as well as a proactive and locally well-known human rights field monitor. These two actors had very little tolerance for arbitrary arrests by the military. Each time a case was reported, the human rights monitor would promptly depart to the FARDC headquarters to plead for liberation, with the chef de poste applying additional pressure. Furthermore, they regularly organised sensitisation sessions for the local population, warning them not to turn to the FARDC in case of disputes, but to submit these to the civilian authorities. As a consequence, the reported number of arbitrary arrests was substantially lower than in nearby Sebele, a roughly similar town controlled by the same battalion. When asked what could explain these differences, many informants pointed to the more assertive attitude of Kazimia's civilian authorities and inhabitants.
Despite the influence of practices of contestation, their cumulative effects on limiting military power appear to be weakened by, amongst other factors, civilians' everyday routines of compliance and their propensity for instrumentalisation. These practices and attitudes are not only found among elites, but concern broad layers of the population. For example, a groundnut seller in Butembo proudly told that he never pays the 200 FC for charging his cellular telephone at a public booth, as he asks the soldier living next door to do it, and the military often refuses to pay for such services. Truck drivers said they condemned people who bought transport tickets half-price from soldiers who are granted free places on every ride, a practice generally regarded as fraudulent. Butchers at a market in Kanyabayonga resented the fact that customers buy cheaper meat from FARDC soldiers' wives nearby, knowing that ‘it comes from the livestock stolen by their husbands’. Village elders in the Ruzizi Plain lamented that their children help the military with committing theft and robbery in a quest for fast and easy money. As most such acts are public, they constitute a show of compliance that ultimately contributes to legitimising the military's power and practices. This, in turn, induces further consent, generating a self-reinforcing spiral of institutionalisation and legitimisation.
Conclusion
The FARDC's vast involvement in sometimes coercive forms of revenue-generation in the Kivus produces, and is in part produced by, an insecure and militarised social order. Military economic practices are both the cause and the outcome of the military's position of dominance within this order. In many contexts, this position has a certain degree of legitimacy, which influences and is influenced by evaluations of the FARDC's revenue-generating practices. One foundation of this relative legitimacy is the military's practices as a regulator, a private protector, a security actor and a business partner. In a context of weak civilian governance, a plurality of normative and authority frameworks and high levels of insecurity, these practices partly correspond to people's beliefs about what public and private goods authorities should deliver and in what manner. One of the causes of this correspondence is that processes of militarisation have substantially altered these beliefs, casting violent practices and actors in a more favourable light.
These findings show some parallels to Roitman's (2005) research on the connections between evaluations of economic practice and the intelligibility of power in the Chad Basin. Analysing how violent modes of appropriation have come to signify licit wealth, Roitman describes the emergence of new frameworks of regulatory authority, in which violent actors have gained prominence. These frameworks constitute what she calls ‘intelligible sites of power’, that provide forms of livelihoods, protection and social mobility to some (pp. 18–22). In a similar vein, this article has demonstrated that the FARDC's power is generally experienced as intelligible, while some of its regulatory, protection and revenue-generating practices are evaluated as licit. Collaboration with and protection and regulation by the military are common-sense practices that have become embedded in the routines of everyday life. It has been argued herein that by engaging in such practices, Kivutiens end up (re)producing the structures of legitimation, domination and signification that underpin the FARDC's power. This does not necessarily imply that they do so in a conscious or intentional manner; rather, the (re)production of structures of militarisation should be understood as the unintended consequences of an aggregate of individual practices. In this manner, individual common-sense acts produce collectively unfavourable outcomes (cf. Giddens 1984, p. 13).
The importance of civilian–military interaction for understanding military revenue-generation, as demonstrated in this article, raises questions about mainstream policy solutions to this phenomenon. In particular, it casts doubts on the possibility of influencing the military's economic practices by focusing only on the incentives of military actors. Policies inspired by such a vision emphasise improving salary payments and human rights training. However, by highlighting the importance of civilian–military interaction, this article suggests that the focus should also be on civilian practices and the incentives that shape these. The more holistic approach that would result from such a reorientation would fully recognise the importance of the interdependencies between the military and civilians. It would need to identify the conditions in which military units are motivated to develop less violent repertoires of revenue-generation, while refraining from constant interventions in civilian affairs. Furthermore, it would need to examine the factors influencing the constraints and preferences that drive the military's regulatory and private protection practices. This draws attention to tenuous property rights, legal pluralism and the absence of civilian governance structures which can provide effective dispute settlement, justice and economic regulation. Finally, it should take into consideration the importance of the discourses that legitimise violence and violent actors' exercise of power. This points to the need to also, or especially, address the question: ‘Faced with militarism, what can you do?’