Corporate sovereignty: Negotiating permissive power for profit in Southern Africa

The growing engagement with sovereignty in anthropology has resulted in a range of concepts that encapsulate how various (non-state) actors execute power. In this paper, we further unpack the concept of ‘corporate sovereignty’ and outline its conceptual significance. Corporate sovereignty refers to performative claims to power undertaken by (individuals aligned to) corporate entities with profit-making objectives within a state-sanctioned space. This contrasts with claims made by other (non-state) actors who operate in a permissive space that (regularly) lacks this legally grounded relationship with the state. By unpacking this state-sanctioned permissive space and highlighting the role of the state as the arbiter, our approach to corporate sovereignty offers a new comparative analytical perspective to theorize how sovereignty is performed and opens ethnographic avenues to explore how sovereignty is negotiated and co-produced across diverse localities. To elucidate our argument, we draw from ethnographic fieldwork conducted on coal mining companies in Mozambique and private security companies in South Africa. By focusing on cases that differ, we want to show the multitude of ways in which corporate sovereignty is enacted and takes shape.


Introduction
To make way for large-scale open-pit coal mines in central Mozambique, multinational coal mining companies have resettled thousands of residents and built new villages that depend on the compensations and services provided by these companies. About 2000 km south in Durban, South Africa, uniformed (armed) private security officers patrol the streets, guard buildings and determine who is included and excluded from a particular space. Although fundamentally different contexts and actors, in both places, we can identify corporate sovereignty: corporate entities that exert authority, claim power and create political communities.
Corporate sovereignty was introduced by Kapferer (2004) to understand the tremendous power that certain transnational corporations have across areas and populations. Kapferer's work is embedded within a growing anthropological engagement with sovereignty, which has been instrumental in understanding claims to authority by a range of non-state actors and processes of legitimization across localities. This focus has resulted in an explosion of concepts such as graduated sovereignty (Ong, 2000), nested sovereignty (Humphrey, 2007), social sovereignty (Latham, 2000) and sovereign agency (Bryant and Reeves, 2020). This multitude of concepts raises the question of how claims to sovereignty differ. We call for more precise clarifications that outline the purpose of adjectives such as nested and how they refer to distinct types of sovereignty. In this paper, we initiate this call by further operationalizing corporate sovereignty and conveying its conceptual and empirical significance.
We define corporate sovereignty as a performative claim to power undertaken by (individuals aligned to) corporate entities with profit-making objectives within a statesanctioned permissive space. This contrasts with claims made by other non-state actors, such as rebels and vigilante groups, who operate in a 'permissive space' (Cooper-Knock, 2018) that regularly lacks this legally grounded relationship with the state. By unpacking this state-sanctioned permissive space and highlighting the role of the state as the arbiter (Rigi, 2007), our approach to corporate sovereignty offers a new comparative analytical perspective to theorize how sovereignty is performed and opens ethnographic avenues to explore how sovereignty is negotiated and co-produced across diverse localities. To elucidate our argument, we draw from ethnographic fieldwork conducted on coal mining companies in Mozambique and private security companies in South Africa. By focusing on cases that differ, we want to show the multitude of ways in which corporate sovereignty is enacted and takes shape. In addition to portraying different levels and forms of power-making, we are specifically interested in analysing the differences of the state's role as the arbiter and how this is shaped by different trajectories of state-making and interactions between state and corporate sectors.
This comparative angle provides insight into the ways corporate power is exercised and the role of the state therein.
In the 'Corporate sovereignty' section, we discuss the conceptual development of corporate sovereignty and how this is embedded within (anthropological) discussions of sovereignty. In the 'The permissive space of corporations' section, we draw from Barkan's (2013) argument that corporations exercise sovereign power by being established by law as well as transgressing the boundaries of law and develop this further by drawing from Cooper-Knock's (2018) notion of 'permissive space' to emphasize that corporate sovereignty is defined by a state-sanctioned permissive space. In the 'Researching corporate sovereignty' section, we outline our methods and the merit of drawing from two different cases, which we then deliberate in the following sections. We conclude this paper by emphasizing the value of our approach and how it contributes to larger debates on sovereignty.

Corporate sovereignty
We understand corporations as legally recognized units that engage in profit-making efforts through the production, distribution and/or trade of goods or services. Broadly speaking, corporations are legal creations that are awarded a fictitious personhood that is recognized by the state or several states (Barkan, 2013). Here we focus on the business corporation, yet we acknowledge that the corporate form comes in various guises, including religious and educational institutions (Welker et al., 2011). Concurring with Welker (2014: 4), we acknowledge the idea of the corporation as a single actor and as a legal creation, but do not regard the corporation as a unified institution. Corporations are complex collectives with their own hierarchies, norms, practices and unclear boundaries (Ho, 2009;Urban and Koh, 2013;Welker, 2014). We see these qualities of the corporation reflected in the different dimensions of how corporate sovereignty comes to the fore, sometimes as a single actor (in the coal mining area in Mozambique, people would for example refer to the mining company Rio Tinto as their neighbour), other times in the shape of individuals who exercise security in the name of a company (such as a security officer patrolling a neighbourhood). When referring to corporations or corporate actors, we do so with this twofold understanding.
Corporations have a long history of governing towns, hospitals and churches (Barkan, 2013). Colonial rule was exercised by companies that controlled territories, populations and resources in many parts of the world (Barkan, 2013;Sawyer, 2006: 29;Stern, 2011), including Mozambique (Allina, 2012 and South Africa (Davenport, 2013). Company rule can also be identified in the phenomenon of company towns, such as the Dutch city of Eindhoven, the 'home' of Philips Electronics (Schippers, 2011) and the Copperbelt in Zambia (Ferguson, 1999). In our contemporary globalized world, under a broader neoliberal transformation of the state (see Harvey 2003), the dominance of corporations has received most attention in the form of capitalist projects of transnational corporations whose power and lack of accountability has been related to the fragmentation, erosion or disassembling of state authority (e.g. Abrahamsen and Williams, 2011;Ferguson, 2006;Mbembe, 2001;Ong, 2000). At its most extreme, this fragmentation is exemplified in the extractive enclave, from which capital is disconnected from its surroundings (Behrends and Reyna, 2011;Ferguson, 2006;Yessenova, 2012). Other scholars have emphasized how corporations have strengthened state authority or rather, how corporate structures and entities have been vital constituents of state-making at large (see Abrahamsen and Williams, 2011). In South Africa, for example, the rise of the private security industry was largely encouraged by the state, which, in turn, strengthened state sovereignty. In Mozambique, the government has welcomed extractive industries to escape donor dependency and to strengthen the economic and political position of (certain) state actors. Corporate power can thus simultaneously strengthen and weaken the legitimate authority of the state.
In understanding these developments and the impact of such corporations, Kapferer (2004) coined corporate sovereignty to pin-point a kind of 'wild' sovereignty that deviates from state sovereignty. While stressing that corporations are state-like structures that have a certain state effect (Mitchell, 2006;Trouillot, 2001), Kapferer (2004: 9) argues that states are territorially based and backed up by popular legitimacy, whereas corporations are rhizomatic and lack this legitimacy. Suzana Sawyer (2006: 24) takes a different approach in her analysis of a lawsuit against ChevronTexaco in Ecuador and explores how 'legal and ethical regimes bring entities into being and establish their worth by endowing them with specific rights and obligations'. Sawyer's approach is akin to Barkan's (2013: 4) recent conceptualization of corporate sovereignty, who argues that modern political sovereignty and corporations are bound together through 'a principle of legally sanctioned immunity from law' (emphasis in original). While Kapferer and Sawyer at times equate corporate sovereignty to corporate power, Barkan offers a more comprehensive understanding and places corporate power squarely within structures of state-led order making. We draw from Barkan's insights, yet develop this further by employing an ethnographic approach that zooms in on the everyday dynamics of corporate sovereignty. Through our ethnographic fieldwork, we were able to identify the continuous negotiations among various actors that shape the nature, scope and scale of corporate power that make corporate sovereignty distinct. In this paper, we draw from a fraction of our empirical material to exemplify this.
During the 2000s, numerous anthropologists reconceptualized sovereignty to understand the violence wielded by non-state actors to produce a particular social order (see Bonilla, 2017;Catellino, 2010;Davis, 2010;Diphoorn, 2016;Grimshaw, 2018;Moore, 2005;Nordstrom, 2000;van Dun, 2014). These anthropological studies prioritize the ability to kill and punish (de facto sovereignty) over sovereignty founded in rules and legality (de jure sovereignty) (Hansen and Stepputat, 2006). This move delinks sovereignty from the state and regards it as a socially constructed source of power that is exercised by multiple actors, repeatedly claimed through public performances, and relative to the assertion and infringement of other claims (Diphoorn, 2016;Hansen and Stepputat, 2006;Humphrey, 2007;Rigi, 2007;Schubert, 2020).
This approach implies that sovereignty is not necessarily linked to control over a particular territory (as a state-centric approach assumes), but that it is a form of power that concerns control over the body (Agamben, 1998;Das and Poole, 2004;Hansen and Stepputat, 2006). Violence is habitually seen as the prime source of sovereign power and with some actors, such as gangs (Rodgers, 2006) or vigilante organizations (Buur, 2006), the use of violence is visible and imperative. Others employ a broader perspective that considers sovereignty as exercising power over 'conditions of existence ' (Comaroff and Comaroff, 2006: 35) and 'the authority to decide on inclusion and exclusion from communities' (Oosterbaan and Pansters, 2015: 125). Sovereignty thus entails the capacity to determine who will live and how.
We argue that corporate sovereignty (in contrast to Kapferer) is based on the use of both the punitive understandings of sovereign power and the 'soft' forms of power linked to agendas of human progress (Buur et al., 2007: 13-14). As we will show, corporate actors may determine, in the case of private security companies, who is arrested and who is protected, thereby creating communities of security. Similarly, extractive companies create communities of dependency by securitization efforts development-oriented efforts such as corporate social responsibility (CSR) (see also Buur and Sumich, 2019;Hönke, 2013;Rajak, 2011). In different ways, through both hard and soft forms of power, corporate entities generate political communities, both through territorial control and everyday security practices.

The permissive space of corporations
Yet corporate sovereignty, we argue, is distinct from other types of sovereignty due to the permissive space in which it operates, namely one that is largely condoned, tolerated or stimulated by the legal and regulatory frameworks of the state. We use Barkan's (2013) understanding of the ambiguous position of corporate power as being simultaneously legally sanctioned as well as formally separated from state power and operating in an area where the law is suspended. However, in contrast to Barkan, we want to emphasize how this ambiguous position is continuously negotiated through formal and informal means and we draw from Cooper-Knock's (2018) notion of permissive space to do so.
Based on ethnographic fieldwork in South Africa, Cooper-Knock (2014: 36) shows how permissive spaces are 'spaces of impunity' wherein illegal violence becomes permissive through unremitting negotiations with numerous actors. Although all claims to sovereignty occur in some type of permissive space, we argue that corporate sovereignty does so in one that is sanctioned by the state, thereby reaffirming the centrality of the state in defining sovereignty (see Bonilla, 2017). To use Barkan's (2013: 19) words, it is the state that gives the corporation special privileges, which he calls 'the sovereign gift'. This gift is not static, but involves complex processes of negotiation with a multitude of actors, with the state as the determining actor.
Due to this gift-giving relationship, we do not see corporate sovereignty as a type of 'wild' sovereign power that emerges 'in the peripheries of powerful state orders', as Kapferer does (2004: 9). Rather, we contend that corporate sovereignty is highly visible and central to state order. One of the key reasons for this visibility is the profitmaking objectives of corporations whose activities are considered (to a certain extent) as legitimate under the legal purview of the state and of which the state somehow also benefits. This thus differs from other non-state sovereign entities, such as the self-defence forces analysed by Pansters (2015) or the shadow networks identified by Nordstrom (2000). This does not mean that non-corporate entities are not concerned with financial opportunities or that corporate entities lack political objectives. Rather, like Cattelino (2010), we emphasize the crucial role of economic wealth to exercise sovereignty and thereby situate the profit-making objective here as a centralized legitimizing feature of corporate sovereignty. This is evident in our empirical cases: the 'extractive turn' has become the dominant way in which the Mozambican government envisions further (economic) development (Schubert, 2020;Wiegink, 2020), and in South Africa, the enormous private security industry is crucial to the country's gross domestic product (GDP; Abrahamsen and Williams, 2011). In both cases, the state's tolerance or encouragement of the industry is framed as benefitting the greater public good and linked to capitalist projections of growth, development and productivity. At the same time, government actors often profit from corporations in illicit ways and sometimes corporate and state actors are hard to distinguish (Buur and Sumich, 2019;Kapferer and Bertelsen, 2009;Reno, 1996). In many parts of the world, we can identify a type of symbiotic relationship where states and corporate actors mutually benefit from corporate efforts, and this is, we argue, largely due to this profit-making dimension that shapes the legally grounded permissive space of corporate sovereignty.
By situating corporate sovereignty within a state-sanctioned permissive space, we aim to emphasize that 'states themselves remain powerful entities' (Bonilla, 2017: 331). However, our aim here is not to focus on the corporate state and how states are increasingly functioning as corporations, as Kapferer does (Kapferer and Bertelsen, 2009). Rather, our aim is to understand how corporate power is exercised and the role of the state therein. This resonates with the work of Coyle Rosen (2020: 4), for example, who argues that the powerful position of the state may even be enhanced by other sovereign configurations. Similar to Rigi (2007: 41), we see the state as a 'final arbiter' that determines, both in legal and illegal ways, the contours of the permissive space. Our notion of corporate sovereignty is thereby distinctive from Sawyer (2006: 38), who contends that corporate sovereigns are increasingly in a position whereby they are 'monitored neither by international law nor by the legal norms of any particular state'. Although we do not deny the immense power that corporations may exercise, we want to debunk the perception that corporate power is inherently unrestricted, unaccountable and unbounded to law (Rhodes, 2016(Rhodes, : 1504Salverda, 2019). Rather, we focus on corporate sovereignty as a continuous negotiation, and at times, struggle, between corporate entities, state actors, institutions and their representatives.
Such negotiations and contestations occur through formal frameworks and procedures, such as contracts, lawsuits, concessions and regulations that stipulate how corporate entities can operate. In South Africa, state legislation dictates the contours of the industry, most notably the Private Security Industry Regulatory Authority (PSIRA), which regulates the private security industry and stipulates how companies and employees must operate. These legal requirements, as we will show later, emerged from corporate-state interactions and negotiations that developed over time. In Mozambique, extractive companies are dependent on state approval for the exploration of concessions, which are guided by for example Environmental Impact Assessments, international foreign direct investment standards and the National Mining and Land Law. The formulation of the Mining Law and other regulations of the extractive industries have been negotiated in close collaboration with corporate actors, to the extent that Schubert (2020) refers to a process of 'co-production' of articulated sovereignty.
In addition to formal procedures, we also include informal negotiations as central to corporate sovereignty. Such informal negotiations may include bribes and threats that can be opaque and superficial, and may occur in the absence or corruption of the law, allowing corporate sovereignty to exist in a situation of 'dirty togetherness' (Podgόrecki, 1987). We also recognize that negotiations over permissive space are shaped by the demands and expectations of other actors. As our case studies show, the clients of private security companies have certain expectations of how private security officers should act and human rights organizations may try to curb the permissive space allowed by the state. However, we maintain that the state is most significant in demarcating the permissive space.
We therefore aim to shed light on how the state as an apparatus, effect and idea acts as the arbiter and how this may come in many guises, as the state's position may be weak or fragile as well as limiting and decisive. The negotiation process may be unequal and murky or tough and restricting or mutually beneficial. In any case, the arbiter position is central to shaping the contours of the permissive space and this makes corporate sovereignty fundamentally different from other types of sovereignty that do not require the permission of the state or that function in direct opposition to the state.

Researching corporate sovereignty
For us to comprehend how this permissive space is negotiated with numerous actors through both formal and informal means, we advocate an ethnographic approach to corporate sovereignty that privileges empirically grounded analyses within specific spatial arrangements. Both authors conducted ethnographic fieldwork to understand the everyday practices of power-making and how individuals give meaning to their everyday reality amidst the presence of corporate power.
The first author, Tessa Diphoorn, conducted 20 months of ethnographic fieldwork in Durban, South Africa between 2007 and 2012 by primarily engaging in participant observation, which entailed accompanying various industry employees in their everyday affairs. She conducted over 160 formal interviews with relevant stakeholders, such as police officers, collected life histories and analysed secondary data. The second author, Nikkie Wiegink, conducted nine months of fieldwork in the surroundings of two coal mines in the province of Tete, Mozambique, between 2016 and 2017. She conducted over 50 interviews with relevant individuals, including employees of mining companies and government officials, spent several weeks in the resettlement villages and joined company community relations officers in their everyday work activities.
We both focused on a demarcated geographical zone from where to examine the complex, everyday negotiations that define corporate sovereignty. With this, we do not aim to reproduce a territorial perspective of state sovereignty that has dominated much of the research on state power. In fact, we want to underline that corporate sovereignty can be identified at multiple sites and levels some of which were accessible for us (e.g. neighbourhoods and resettlement villages) and others were not (e.g. board rooms and government officials' offices). However, we do suggest that there is a spatial dimension to the way corporate sovereignty is enacted and that an ethnographic approach can best uncover the multiplicity of power relations within a particular space.
By drawing from two different case studies, we aim to highlight how corporate sovereignty takes shape in multiple ways. This paper does not draw from a comparative research project and is not the result of a multi-sited ethnography (Marcus, 1995), but emerged from our mutual theoretical interests in understanding corporate power. The Mozambican case entails a transnational industry that centres on resource extraction, a small, highly skilled workforce, land acquisition and the displacement of local populations. Here we can identify corporate sovereignty performed by mining companies that offer services conventionally provided by the state. The ruling elite of Mozambique sets the contours of the 'permissive space' in which corporate sovereignty is performed, largely by granting concessions and by de facto suspending the law. In South Africa, the permissive space of the industry was founded upon collaborative efforts between the state and the industry, where the industry was allowed to flourish to support the state. Here the industry is large in terms of wealth production, yet compromises numerous smaller companies that primarily rely on low-skilled labour to carry out the bulk of the work. Here we see how corporate sovereignty is performed at the individual level, for example, a security officer intimidating a suspect, and how such individual practices have a collective effect in shaping the broader political and social landscape. Yet although divergent, in both instances we can recognize how a corporate actor exerts authority by using soft and hard forms of power within a state-sanctioned permissive space.

The coal mining industry in Mozambique
In Tete province, Mozambique, there are two villages, Cateme and Mualadzi, which were newly built between 2009 and 2014 to house over 10,000 residents who were displaced to make way for the coal mines (Lesutis, 2019). People in Tete would generally refer to these villages by denoting them to the mining companies that built the villages: Cateme was referred to as 'Vale's village', as it was constructed by the Brazilian mining company Vale, and Mualadzi was referred to as 'the village of Rio Tinto', as the village was created by the Australian mining multinational. One of the residents of Mualadzi characterized his living situation by saying: 'we are in the hands of the company', thereby illustrating the people's dependency on the mining company and their expectations of the companies (Wiegink, 2018: 249-250).
Company rule in Mozambique has a long history, as early Portuguese colonialism was largely organized through land concessions called prazos that were granted by the Portuguese Crown to settlers and traders (Isaacman and Isaacman, 1976;Newitt, 2017, 37). The owner of the prazo acted as a landlord, governing a certain tract of land by maintaining their own armies. The prazo owners were involved in all kinds of trade including gold, ivory and people, and paid rent to the Portuguese Crown (Newitt, 2017). The prazo system was followed by larger administrative concessions to companies in the late 19th century. Companies such as the Companhia de Moçambique (British and South African owned) and the Companhia de Zambezia were allowed to collect taxes and to exploit forced labour (Allina, 2012;Isaacman, 2018 [1983]). These companies also assumed the responsibility for pacifying, policing and administrating their concession areas (Newitt, 2017: 99). In the 1930s, the situation changed due to the implementation of a centralized colonial administration, yet much of the colonial crops produce, largely cotton and rice, occurred through concessions to private companies (Newitt, 2017: 124).
After independence in 1975, companiesnow often partially state-ownedwere considered part of the ruling Frelimo (Mozambican Liberation Front -Frente de Libertação de Moçambique) utopian ambition for socialist reform of Mozambique. The partially state-owned coal mining company Carbomoc operative in Tete province was involved in the provision of state-like services to surrounding populations (Schubert, 2020), although it soon suspended coal production due to the civil war. Corporate-state relations changed in the late 1980s when the Frelimo government, in the context of a protracted civil war and with pressure from international donors, implemented rigorous political and economic reform (Pitcher, 2002). The Mozambican state became increasingly dependent on foreign donors and at the same time, a relatively small Frelimo elite benefitted from the privatization and a steep increase of land acquisition and (foreign) investment projects (Fairbairn, 2013;Macuane et al., 2018). The promotion of foreign development investment increased in recent decades with a series of energy and extractive mega-projects (Kirshner and Power, 2015;Selemane, 2010) and the allocation of large concession areas to companies such as Vale shapes the contours in which corporate sovereignty emerges in Mozambique.
We argue that coal mining companies such as Vale are engaged in performances of corporate sovereignty as the companies enforce a kind of rule over the territory and population in the surrounding of the mine. The coal mines are fenced-off and guarded by private security companies who operate in co-optation with state security forces. The latter have repeatedly been called upon to disperse protests against the coal mining companies by using excessive force and randomly arresting people (Human Rights Watch, 2013;Lesutis, 2019). This exemplifies how mining companies are involved in the creation of order through the ability to use violence or the use of violence in their name, i.e. the hard aspects of corporate sovereignty. However, corporate sovereignty is most visible in the soft forms of power that coal mining companies exercise over the populations living in the mining area.
The people who are at the forefront of performing these soft forms of power are the companies' community security officers, who through community projects and other developmental efforts aim to establish and maintain a so-called 'social license to operate'. This is not only understood as the mine being a 'good neighbour' to its surroundings, but is also framed in relation to 'social risk'. Part of the work of community relations officers in Tete is to maintain relationships with people living in the surroundings of the mine or its vital infrastructure. One of these officers commented that 'I have to know these people [living close to the mine] and somehow be part of their daily life. They could protest one day and if I do not have these relationships, I would not know who to approach' (7 June 2017, conversation). His job was part of the company's strategies to secure the production and infrastructure of the mine as well as the reputation of the company, which could be damaged by violent incidents and environmental destruction. The management of community relations in the vicinity of extractive projects through CSR discourses and practices and other forms of proactive social engagement are common security strategies (Buur and Sumich, 2019;Hönke, 2013;Welker, 2014). Through such strategies, mining companies are implicated in governing populations (Rajak, 2011), or as we articulate here, in performing sovereign power by exercising control over people's 'conditions of existence ' (Comaroff and Comaroff, 2006: 35).
These soft forms of corporate sovereignty are particularly visible in the resettlement villages. In Mualadzi and Cateme, the companies provided (at least temporarily) services that are generally associated with the state, such as water, electricity, housing, and the infrastructure of the villages, including schools, health facilities, roads, lampposts, markets, government buildings, and so on. In the initial phases of the resettlement process, the relocated families received one hectare of land, a sum of money to buy a second hectare, a furniture set, and, in some cases, basic food baskets. To contribute to the process of people's livelihood restoration, the mining companies organized capacity building workshops focused on adjusting subsistence agriculture practices to semi-arid soil, since the fields the resettled populations left behind were often more fertile. Through such provisions, the mining companies created communities of dependency over which the companies exercise or are expected to exercise some kind of control.
Whereas the companies' expected performances resemble some of the key provisions of the state, there is at least one fundamental difference, namely the fickle nature of corporate sovereignty. Coal mining companies depend on global commodity markets for their revenues, which often directly affect company budgets for community relations and resettlement. Companies may also practically disappear overnight: this happened with the Benga concessions that Rio Tinto sold quite unexpectedly to the Indian consortium ICVL. After the ongoing resettlement processes came to a standstill, 200 families singled out for resettlement had to remain in their houses, livelihood restoration projects in the resettlement village of Mualadzi were paused, and crucial infrastructure such as water systems were not maintained or repaired. The residents of Mualadzi felt abandoned and in several interviews asked: 'when will they take care of us?' With 'they', the residents implied both the company and the (local) government agencies, illustrating how the responsibility for the governing of the resettlement was ambivalent (Wiegink, 2018). These people experienced corporate sovereignty as a rather capricious form of power, characterized by a volatility that largely stems from the profit-making character of corporate entities. This capricious nature can, however, only exist in a permissive space that is allowed by the state.
Mining projects are bound to concessions that are given by the Mozambican government, i.e. the 'sovereign gift' (Barkan, 2013: 19; see also Coley Rosen, 2020). Such 'subterranean sovereignty' (Luning and Pijpers, 2017: 760) is an example of how access and authority over land and subsoil between mining companies and (local) authorities are negotiated. These concessions entail a kind of demarcation for the mining company's sphere of influence over a certain territory and by extent the people who live there. Negotiations over concessions are conducted between companies and the governing elite of Mozambique. Fairbairn (2013: 337) has called this 'access control' and highlights the mediating role of the domestic elites. In Mozambique, this access control is predominantly located with the Frelimo party elite that dominates Mozambique's political and economic life (Fairbairn, 2013: 343). Former President Armando Guebuza is for example considered to have been directly involved in the negotiation of the Vale mining concession (Macuane et al., 2018). While these concessions are legally defined, the ways in which they are acquired are less transparent and surrounded by (the suspicion of) obscure negotiations and corruption (Macuane et al., 2018).
The contours for corporations' dealings with local populations are shaped by the state's laws and regulations, such as the mining law and the resettlement regulation drafted in 2012, which oblige companies to fulfil certain responsibilities. These laws and regulations are formulated in consultation and negotiation with corporate actors and other stakeholders including the World Bank and international donors (Schubert, 2020: 552-553;Wiegink, 2020). These negotiations have resulted in a stronger emphasis on transparency (Wiegink, 2020: 5) and there are indications of strengthened audit regimes and improved efficiency in the governing of extractive resources (Schubert, 2020: 552-553). Yet Macuane et al. (2018: 432) have pointed out the selective nature of 'pockets of efficiency', resulting for example in reduced tax payments for companies and speculations about party and personal accumulation of rents. In addition, mining companies are by and large not held accountable for failures or flaws in their resettlement programs. When Wiegink discussed Mozambique's norms and laws concerning mining and CSR with a community relations officer of a mining company, he said: 'it is already very positive when the companies, for themselves, without being obliged by no one, remember [the norms and laws]' (20 March 2017, interview). He worked for six years in the mining industry and observed that the mining companies have only sparsely been held accountable for malpractices by the government, a process that is further complicated by the alleged corruption of courts. This suspension of the law as condoned by the state defines the permissive space of corporate sovereignty.
Human rights and other watchdog organizations have attempted to curb the permissive space of coal mining companies by taking companies and the state to court. In October 2019, Vale and the Mozambican state were condemned for the violation of rights of communities affected by the mining exploration. 1 This was a rare legal success, as mining companies have generally been able to go about their business without the strict implementation of the mining law or resettlement regulations (Amnesty International, 2018). However, law cases and awareness campaigns may exert influence in other ways, as shareholder-based companies, such as Vale and Rio Tinto, are concerned about their reputation. Due to previous malpractices elsewhere (Marshall, 2015), Vale has been closely scrutinized by various non-governmental organizations and human rights organizations (Human Rights Watch, 2013;Lillywhite et al., 2015). In the words of one of Vale's community development officers, 'Vale is always in the public eye. While we are not even the worst company' (7 June 2017, conversation). In Tete, Vale has been responsive to criticism and has adapted its performances in CSR and resettlement. Vale's attempts are in stark contrast to the Indian consortium ICVL, which is partially owned by the Indian government and lacks shareholders, and seems less concerned with the critique raised by watchdog actors.
Coal mining companies in Mozambique thus perform corporate sovereignty by exerting control over territories and populations through informal and formal negotiations with a range of actors, including state agencies, shareholders and watchdog organizations. Yet it is the Mozambican state that plays the most defining role in setting the contours of the permissive space in which these companies operate and this state-sanctioned role is also visible in the private security industry in South Africa.

The private security industry in South Africa
During the summer of 2010, Diphoorn strolled around the grounds that would host the community annual fair with Paul, the owner of Armed Reaction. 2 While others were busy setting up the various stalls and necessary infrastructure for the fair, Paul explained how his company would provide security for this well-attended event that residents looked forward to each year. Paul outlined their strategies and the potential security risks amidst the numerous rides, games and other leisurely activities, and explained how this was all a 'community service'. A few years ago, the South African Police Services was responsible for providing security, yet due to a shortage of personnel and resources, this task had been outsourced to Armed Reaction, a locally owned and community-based company that played a prominent role in the everyday security of this predominantly Indian community. Yet, although formally outsourced, the services were provided for free and were, as Paul stated, 'something that I want to do for the community'. Throughout that week, Diphoorn observed how the company acted as the primary security provider at this public event and we argue that this is one of the ways in which (armed) security officers perform corporate sovereignty in Durban, South Africa.
Armed Reaction is one of the 8916 registered companies that make up the private security sector in South Africa (PSIRA, 2017(PSIRA, -2018. With 522,542 active registered security officers (PSIRA, 2017(PSIRA, -2018, South Africa has the largest private security sector in the world, which was valued at ∼2% of the country's total GDP in 2011 (Abrahamsen and Williams, 2011). In addition to its size, the industry is also highly diverse and ranges from ordinary locksmiths to armed security guards that patrol communities, to complex technological systems.
In South Africa, corporate sovereignty is established through both soft and hard forms of power, although the hard forms of power are more visible and prominent and draw from the officers' ability to use violence. Violenceor more importantly, the potentiality of violenceis crucial to the means by which these officers acquire and maintain authority. During the fair, for example, officers purposely displayed their firearms while patrolling the grounds, ensuring that their ability to use violence was visible to all attendees. Diphoorn regularly observed how security officers used (the potentiality of) violence in their everyday practices. This company, Armed Reaction, had a reputation for regularly resorting to physical violence and this was also a key source of their popularity in the area.
Yet the fair is also a fine example of soft power. Although soft power is less prominent than with the mining companies in Mozambique, private security companies habitually donate goods and services to the communities in which they operate. Companies increasingly voluntarily participate in community activities, such as the fair describe above, and assist citizens who are not formally paying clients, such as supporting them with and after crime incidents. These efforts are framed as ways in which companies 'are part of the communities they serve' (marketing manager, 2 July 2010, interview). Such discourses resonate well with clients, who expressed their satisfaction with such companies better serving the community.
Through the use of both hard and soft forms of power, security officers differentiate between insiders and outsiders and construct physical and social borders across urban space, which results in the creation of 'communities of security'. These borders and their maintenance are often highly racialized, whereby young Black, poor men are habitually approached, intimidated, and denied access to these communities (Diphoorn, 2017). This happens not only on daily patrols, but also at larger events, such as the fair, where Diphoorn witnessed how undesired individuals were denied access to the fair or were closely monitored while attending. This was not necessarily a hidden social fact: when speaking to some young men residing in a township neighbouring the community where the fair was held, they shared their disdain at being excluded from such public events, and at times even the community at large. One of them said, 'They will not allow me to go there, treating me like a criminal'. When saying this, he was specifically referring to Armed Reaction, and during other conversations about their experiences with this company, many young men shared their experiences of being targeted and mistreated by the officers. Such sentiments, of abuse experienced by security officers, were echoed by other research participants in other parts of the city.
We can thus identify corporate sovereignty by analysing the individual practices of industry employees, but also in how these practices accumulate and result in the creation of (imagined) boundaries that construct distinct political communities. Despite the visible social and economic differences across the urban realm, the policing efforts of security officers are very similar to the practices executed by vigilantes and gangs in South Africa's urban centres, especially regarding the use of violence (Buur, 2006;Jensen, 2008;Smith, 2019). Yet there is a difference between their everyday practices and this is largely due to the relationship between the state and the industry.
In South Africa, the state has played a leading role in shaping the growth and contours of the private security industry. The private security sector originally operated on industrial sites and slowly entered the urban realm during the 1970s (Brogden and Shearing, 1993). During the 1980s, tasks that had previously fallen under the remit of the police were handed over to the private sector and this was made possible through various new laws, such as the National Key Points Act 102 of 1980, which transferred the responsibility for security provision at crucial sites to the owners of these sites, who primarily recruited private firms. An alliance emerged and many firms identified with the discourse of state sovereignty. This was also facilitated by the fact that most of the firms were established by former policemen or soldiers, creating an 'old boys network' that still exists today, albeit with different faces (Diphoorn, 2016).
The collaborative relationship between the industry and the state was further consolidated by the passage of the Security Officers Act of 1987 and the accompanying Security Officers Board, which entailed setting up a formal regulation system. Originally intended to control the predominantly black labour force, it acted as the first step of state regulation over the industry, thereby legally determining how the industry can and should operate. Important to note is that industry personnel played a large role in shaping this regulation, further strengthening the corporate-state partnership. As a result, the industry experienced a tremendous growth during the political transition period and thereby also played a pivotal role in the security provision of (White) residential areas across the country. The collaboration also explains the initial suspicion held by the post-apartheid government towards the private security sector, seeing it as part of the old (apartheid) order.
This changed, largely due to the expansion of state regulation through the Private Security Industry Regulation Act No. 56 of 2001 and the accompanying PSIRA. Furthermore, in the fight against crime, the government envisioned a 'multi-agency approach' (Singh, 2008: 14) to law enforcement and the private security industry was defined as an ally. For many state officials and industry employees, everyday operations of private security officers are thus regarded as an outsourcing of state coercion that operates within the legal parameters of the state, such as Armed Reaction's role at the community fair. And although PSIRA is flawed and faces allegations of corruption, it does provide the state with a certain degree of oversight and defines the contours of the permissive space in which corporate sovereignty is performed. This differs from various other policing bodies, such as gangs and vigilante organizations, which are (formally) exempt from such government oversight. We acknowledge that there are policing groups across the country that blur such distinctions, including vigilante groups and community policing bodies that have formally registered as private security companies, for example (Diphoorn, 2016;Oomen, 2004). We also identify the state oversight exerted over citizen-led policing groups that operate under the formal structures of community policing forums (CPFs) that were instigated to act as the institutional interface between police officers and local residents (Gordon, 2001). Yet private security firms are provided with the broadest permissive space and this, we argue, is primarily due to the profitmaking dimension of the industry and the ways in which their practices strengthen state sovereignty.
This strengthening of the state and state-led stimulation of the industry is particularly evident at the local level of police stations and on the streets. This occurs not only formally through meetings held at police stations where crime information is exchanged, but also informally through undocumented patrols and active social networks between company employees and police officers, i.e. the 'old boys network'. Such social networks, often between company owners (of which many are ex-police officers) and highranking officers, act as the framework for informal and ad hoc partnerships, which regularly result in acts of impunity and corruption, also a type of 'dirty togetherness' (Podgόrecki, 1987). Police officers, acting as representatives of the coercive arm of the state, tolerate or even encourage (punitive) practices of private security actors that may be formally deemed illegal. Many police officers voiced support for the extrajudicial use of force by security officers, allowing them to 'do our dirty work' so that they did not have to get their hands dirty. This highlights again how the suspension of the law is unpredictable and largely sanctioned by the state.
In addition to the state, clients (governmental, industrial and residential) also dictate how security officers operate. The South African private security industry is highly saturated and companies invest heavily in marketing strategies. One of the main strategies has been to invest in communities, as stated by a marketing manager: 'As a company, you need to differentiate yourself through your service, through making your product stand out (…) So we invest in the CPFs, sponsor local crime newsletters, provide free security training, provide free monitoring to schools, all of that. It costs us a hell of a lot of money, but it pays off (….) People think: they really care for the community, and so I want to be their client' (2 July 2010, interview).
Armed Reaction's security provision at the fair is an example of this marketing strategy. Besides the enormous company banner strategically placed at the entrance, the company's emblems were located on numerous parts of the ground, thereby constantly reminding the attendees of their dominant presence and the need for their security.
Security officers are also encouraged to win over new clients and to operate with a mentality that 'the client is always right' and this ascribes clients a large degree of purchasing power. Clients can choose between companies and when unsatisfied, they can easily switch between security companies and this places immense pressure on the security officers to please them. As the female client of a large internationally owned company once said: 'We pay these companies good money, so if they don't respond on time or escort me home, I'll complain and threaten to go elsewhere ' (10 December 2008, interview).
This ability to make demands is particularly visible in collective security arrangements whereby citizens club together to benefit collectively from a particular service. It is through such security arrangements that security officers are increasingly pushed to operate in the public realm, such as public roads. Clients thus push corporations to expand the permissive space in which they operate, thereby hardening the bounds of their political community through illegal measures. In fact, Diphoorn regularly witnessed how clients encouraged or even demanded security officers to employ physical violence when, for example, arresting suspects within the vicinity of their homes. This key role of paying clients differs from the citizens that condone and/or support the security practices of vigilantes and gangs, thereby reaffirming how the profit-making dimension of companies results in distinct performances of sovereignty.
In the private security industry in South Africa, we can thus identify how individuals claim a socially constructed source of power that is steered by profit-making logics and occurs in a legally sanctioned permissive space. The performance of corporate sovereignty occurs at the individual level through everyday practices of security, but also manifests itself at a larger level, in the way that political communities are created from such practices. Although the contours of this permissive space are negotiable and in flux, they are largely moulded by the legal sanctioning of the South African state.

Concluding remarks
In this article, our aim has been to further develop the concept of corporate sovereignty that was first outlined by Kapferer (2004) and later explored by Barkan (2013) and emphasize how corporate sovereignty refers to a distinct type of sovereignty. Although corporate sovereignty is a 'formation of sovereignty' (Stepputat, 2015) that exists in relation to other competing, converging or coexisting sovereigns, there is something different to corporate sovereignty that demands further ethnographic attention.
For us, corporate sovereignty is the performance of sovereign power that is based on hard and soft forms of power-making, operates with a profit-making objective, and is enacted within a permissive space that is legally sanctioned by the state. This permissive space is not fixed but is incessantly negotiated through formal and informal interactions with various actors, such as police officers, shareholders, clients and other beneficiaries. Yet it is particularly the state, most often as the arbiter (Rigi, 2007), that plays a decisive role in formally and informally shaping the contours of the permissive space. This is not the same as Kapferer and Bertelsen's (2009) understanding of the corporate state that centralizes how states are increasingly operating as corporations. Rather, our notion of corporate sovereignty aims to exemplify the ways in which the state plays a role in the emergence and performance of corporate power. Furthermore, we also do not intend to (re)produce a static understanding of the state as, for example, a 'broker state (Urteaga-Crovetto, 2012) that serves to facilitate the entry of corporate capital or a 'phantom state' (Goldstein, 2012) whereby the state legally condones the presence of corporate entities, but is absent in enforcing its laws and regulations. Rather, we aim to understand how state institutions and representatives play a critical role in the negotiations with corporations. The state may act as a facilitator of processes of dispossession through privatization (Grandia, 2014;Harvey, 2003), as a regulator by setting the limitations of corporate conduct (Schubert, 2020), or as a collaborator through informal power structures and networks. Through our ethnographic lens into these negotiations, we want to alleviate reigning depictions of corporate power as unlimited, uncontrollable, and unfathomable.
Through our divergent case studiesthe coal mining industry in Mozambique and the private security industry in South Africawe have identified how corporate sovereignty is enacted within state-sanctioned spaces in different ways. With the mining industry, we see the prominence of soft forms of power-making through the provision of essential goods and services to resettled populations, while hard forms of power-making are more decisive in the everyday practices of the private security officers. In South Africa, paying clients play a fundamental role in influencing the permissive space, while shareholders and human rights organizations function as key actors in Mozambique. Yet despite these differences, we see how the state acts as the arbiter in both cases. In Mozambique, this role is most clearly seen in the state's allocation of concessions to particular companies and reluctance to enforce mining regulations. In South Africa, state-sanctioning occurs at the formal and national level through state regulation, and at the local and informal level through ad hoc interactions, co-patrols and social networks between police officers and security officers. The state arbiter role is thus crucial to the emergence and performance of corporate sovereignty. In our analysis of two different industries, we have seen a modus of territorialization, either through policing certain neighbourhoods or through fencing off concessions and resettling people. In both cases, corporations are linked to specific spatial arrangements wherein certain services are delivered. However, this territorial dimension may not be relevant for all corporate sectors. Herein lies the need for further analytical work to tease out the applicability of corporate sovereignty beyond the private security and extractive industry. Yet for now, we believe that the concept of corporate sovereignty is ethnographically useful as it draws attention to everyday performances of sovereign power and the negotiations of the permissive space in which corporate sovereignty is enacted across different localities, contexts and sectors.