‘Arms for mobility’: policing partnerships and material exchanges in Nairobi, Kenya

ABSTRACT

This paper analyses two policing arrangements between the state police and several private security companies in Nairobi, Kenya. These arrangements entail that police officers team up together with security officers in their company vehicles. As private security officers are unarmed in Kenya by law, there is a direct exchange of ‘arms for mobility’, an emic term that refers to an exchange of firearms for ‘mobility’, i.e. vehicles and other financial resources. Based on ethnographic fieldwork on policing in Nairobi, Kenya between 2014 and 2018, I analyse how this exchange (re)centralises the state police and the critical role of the ‘arms’ in this process. Drawing from Star and Griesemer (1989),  I see the firearm as a ‘boundary object’ that brings policing actors together, but simultaneously reaccentuates their differences and in this case, reaffirms and repositions the dominant role of the state police in the Kenyan policing landscape. With this argument, I aim to further prompt more in-depth studies on how certain objects define policing practices, and emphasise the merit of ethnographic research as a methodological approach to uncover such dimensions.

KEYWORDS:

• Policing
• materialities
• firearm
• Kenya

Introduction

I was squished in the back of the car during a joint-patrol organised between a private security company and the Kenyan police and attentively listening to Michael, a police officer in his mid-30s. Throughout this entire night shift, Michael was actively chatting away about his police experiences and what it means to be a Kenyan police officer. Riding shot gun, he was undoubtedly in charge, habitually instructing the driver – a private security officer from the company in question – which turns to take, which spots to check, where to stop for a tea break, and so forth. The other more junior police officer and security officer, sitting to my right in the back of the car, were rather silent and sporadically nodded in approval of what Michael was saying.

At one point, of his own initiative, Michael shared his appreciation for this policing partnership, in which private security companies and police officers patrol certain (affluent) neighbourhoods together. He described it as a ‘win-win’, and then went on to elaborate on how ‘it helps us all’. But then, rather abruptly, he turned around, looked straight at me and said: ‘they need us, because of this’ – and he pointed to his gun. He then slowly pulled it up from beneath his legs and narrated how the gun gave him, and his fellow officer, ‘the power’. ‘This [the firearm] is what brings us together … and why I am here, in this car’.

*  *  *

This patrol in May 2015 is one of the several patrols that I accompanied during my ethnographic fieldwork on policing partnerships in Nairobi, Kenya. Similar to other parts of the world, Kenya is characterised by a pluralised landscape of policing, wherein we can identify a range of actors engaged in the provision of security, especially within the larger urban centres (see Anderson 2002, Ruteere and Pommerolle 2003, Rasmussen 2010, Smith 2015, van Stapele 2015, 2016, Price et al. 2016, Skilling 2016, Kimari 2017, Jones et al. 2018, Schuberth 2018, Colona 2019).

Within the scholarly focus on policing, the dynamics and interactions between the private security industry and the state police has received ample attention. These studies outline the diversity of relationships that range from outright competition to fruitful collaboration and how such relationships are defined by the sharing or exchanging of certain services (see Jones and Newburn 1998, Hummer and Nalla 2003, Berg 2004, Minnaar 2005, Manzo 2009, Diphoorn and Berg 2014, Puck 2017). However, in-depth analyses on how certain materialities shape and define these partnerships are scarce. It seems that the material turn – an exploration into the relationships between people and materialities – has not yet found its way to the policing scholarship. This article aims to initiate that by analysing the role that certain objects play in defining two policing arrangements in Nairobi, Kenya between the Kenyan state police and several private security companies.

As can be seen from the introductory vignette, these arrangements (largely) entail that police officers team up together with security officers in their company vehicles. As private security officers are unarmed in Kenya by law, there is a direct exchange of ‘arms for mobility’, an emic term that refers to an exchange of firearms for ‘mobility’, i.e. vehicles and other financial resources.¹ Based on ethnographic fieldwork on policing in Nairobi, Kenya between 2014 and 2018, I analyse how this exchange (re)centralises the state police and the critical role of the ‘arms’ in this process. Inspired by Frossard and Jaffe (2018), I also draw from Star and Griesemer (1989) and analyse the firearm as a ‘boundary object’ that acts a bridge and brings policing actors together for potential collaborative practices. However, in contrast to Star and Griesemer, I highlight that boundary objects simultaneously reaccentuate differences and in this case, reaffirm and reposition the dominant role of the state police in the Kenyan policing landscape. With this argument, I aim to further prompt more in-depth studies on how certain objects define policing practices, and highlight the merit of ethnographic research as a methodological approach to uncover such dimensions.

In the first section of this article, I first discuss the material turn and how it can provide new perspectives to understanding everyday policing across localities and then zoom in on the firearm as object. In the second section, I discuss policing in Kenya and how certain legal and regulatory frameworks have steered the formation of policing arrangements. In the third section, I briefly consider my methodology and then analyse two policing arrangements in Nairobi, Kenya. In the last section, I offer some concluding remarks and further advocate the conceptual opportunities that a focus on materalities and an ethnographic approach can provide for the policing scholarship.

Policing and materialities

The growing work on policing across disciplines has highlighted both its globalised and pluralised nature, meticulously showing how numerous public and private security providers perform a range of security practices across diverse localities. Numerous theoretical frameworks have been developed to analyse the various interactions that shape this plurality, such as the nodal framework (Shearing and Wood 2003), the security network (Dupont 2004, Krahmann 2005), and the security assemblage approach (Abrahamsen and Williams 2011). Combined, these frameworks provide ways to understand how various security providers (co)exist.

A key focus has been on the interactions between the private security industry and the state police, and several studies have outlined the diversity of relationships that range from outright competition to fruitful collaboration (Jones and Newburn 1998, Hummer and Nalla 2003, Berg 2004, Minnaar 2005, Diphoorn and Berg 2014, Puck 2017). These studies have exemplified the numerous ways in which people and knowledge are shared or exchanged and how this shapes certain power structures. As a way of further dissecting such exchanges, Diphoorn and Grassiani (2017) discuss the way in which various forms of capital – social, political and economic – are employed and mobilised by security providers across localities to acquire authority in the field of security. The authors encapsulate this with the concept of securitising capital, which refers ‘the processes of acquiring symbolic capital through other forms of capital in the field of security’ (435). Another popular concept within this field of study is that of ‘hybridity’, which specifically focuses on the entanglements between state and non-state actors and practices. The idea of hybridity, distilled into various concepts, such as hybrid governance (Colona and Jaffe 2016) or hybrid political orders (Boege et al. 2009), aims to understand how state and non-state actors are intertwined and thereby create a ‘new or emergent political formation that is neither state nor non-state’ (Colona and Jaffe 2016, p. 2).

Combined, these frameworks and concepts allow us to make sense of the various ways in which policing actors connect and/or disconnect, yet the role of materialities remains undertheorized. In the policing literature, there are several studies that examine the growing role of technologies and the digitalisation of security work, focusing on crime mapping, information technology (IT), social media, and various forms of surveillance, such as CCTV cameras (see Manning 2008, Kelly and Finlayson 2015, Holstein 2018). Yet these studies tend to frame objects as ‘new technologies’ that are part of a ‘technological revolution’ (Sheptycki 2018, p. 2) in an increasingly digitalised and globalised world, and thereby overlook certain (mundane) objects, such as the firearm, that have existed for much longer. In addition, technologies are often regarded as neutral and impartial, are seen in terms of their potential for solutions (Frossard and Jaffe 2018), or are analysed for what they represent and how they are used. However, an analysis on how objects shape the practices and experiences of humans, an approach that has defined the so-called material turn, has not yet been fully developed and applied in the policing scholarship.

Broadly speaking, the material turn (or the material cultural turn, see Hicks 2010) refers to an increased recognition and analysis on the role that objects, materials and things play in shaping human action.² Largely drawing from Bruno Latour (2005) and John Law (1986) and the building blocks of actor-network theory (ANT) and more post humanist approaches (Haraway 1990, Barad 2003), the main idea is that materials should be regarded as actors – as actants – that do a wide range of things, shape subject formation, and produce a myriad of social relations. This materialist line has emerged differently across disciplines. In urban studies, the infrastructural approach has provided an analytical lens to understanding how urban life is structured (Simone 2006, Larkin 2013). In security studies, scholars have shown how materialities determine security governance. This includes a focus on ‘critical infrastructure’ (Coward 2009, Aradau 2010) in determining systems and relationships of power and authority; on how certain objects become the focus of political controversies (Walters 2014), and more specifically, on how particular technologies mould the security landscape, including airport security scanners (Valkenburg and van der Ploeg 2015) and drones (Gregory 2011, Kindervater 2016).

In anthropology, certain scholars have lead the (renewed) focus on materalities, such as David Miller (2005, 2010), Arjan Appadurai (2013), and Tim Ingold (2007, 2012), to name but a few. The main idea here, too, is that objects do not merely represent or symbolise certain aspects of social life, but that they are meaningful, active, and do something on their own. Daniel Miller claims that ‘objects create subjects much more than the other way around’ (2008, p. 298) and Arjun Appadurai (2013) coined the phrase ‘the social life of things’ to encapsulate how objects have a ‘life’ in and of themselves. Although I recognise the discussion on the (sematic) differences between objects, things, materials, non-humans or artefacts (see Ingold 2007, 2012), in this paper, I use the term objects, as I am particularly interested in the role that (certain) objects play and how they are exchanged. I also echo some of the issues raised within anthropology that we need to be cautious of removing people (i.e. subjects) and their perceptions and feelings from our analysis. I contend that we need to recognise the limitations of the agency of objects and continue to question whether objects can be equated with subjects. I therefore highly appreciate Chelsey Kivland’s (2018) proposal of a ‘relational model in which people and technologies coparticipate in action’ (355, emphasis added). Based on extensive ethnographic fieldwork in Haiti and focusing on the ‘gun’, Kivland argues that objects cannot be separated from the user’s intents, but that they provide spaces of potentiality for certain human actions. She thus claims that the gun – her object of analysis – is both ‘scripted and scripting, both containing a script for human action in its technological design and scripting as it is taken up through scenarios of action that people project onto it’ (364, emphases added).

A specific strand in anthropology concerns the role of materialities, such as documents, in processes of state-making (Sharma and Gupta 2006, Hull 2012, Colona 2019). Matthew Hull, for example, outlines how ‘documents are not simply instruments or tools, but that they act as mediators and are constitutive of bureaucratic rules, ideologies, knowledge, practices, subjectivities, objects, outcomes, and even the organizations themselves’ (2012, 253). Similarly, documents do not merely ‘represent’ things, but have capacities, such as exerting administrative control and creating particular subjects. In understanding the relationship between objects and state power, the role of the firearm has been overlooked, despite estimations that over one billion of them exist, of which 85% rest in civilian hands (Small Arms Survey 2017).

Charles Fruehling Springwood (2007 , 2014) is one of the few scholars that thoroughly explores how the firearm is a persuasive exertion of symbolic power, a cultural object, and a sought after commodity that is widely and overtly purchased.³ In fact, Springwood argues that the firearm is qualitatively different than other objects and carries a unique material role, acting as both a symbol of ‘subversion and domination’ (2007, p. 11). He claims that the firearm cannot be compared to another object, as it is ‘the most highly charged, fiercely desired, intensely feared, and broadly contested material object in the United States, if not the whole world’ (2014, p. 452). One of the paths that Springwood takes in analysing the firearm is an embodied one, demonstrating the crucial embodied relationship between firearms and humans, and analysing this as an ‘enchanted assemblage of performance, control, omnipotence, pleasure, and fear’ (2014, p. 452). In his work on Hezbollah militants, Younes Saramifar (2017) also employs an embodied approach to the firearm, highlighting how ‘the weapon forms a relationship with the body’ (5). This resonates with some of my previous work, where I also reflect on how certain equipment, including the firearm, is experienced as a part of the body for security officers and a crucial constituent of how they exert and claim authority (Diphoorn 2015).

In this paper, I do not take an explicitly embodied approach, but accentuate the close relationship between the firearm and the police officer. In this context, I see the firearm as an artefact that provides space for the state to ‘be imagined, encountered, and reimagined by the population’ (Sharma and Gupta 2006, p. 12) and thereby (re)produces state authority. The firearm is therefore not just a prop that police officers possess while on duty that represents state power or acts as a symbol for the state’s use of force and ability to kill. Furthermore, it does also not single-handedly decide the actions of a police officer. Rather, I argue that, through the presence of the firearm, spaces of potentiality are provided to ‘perform particular acts of power and violence’ (Kivland 2018, p. 355) that (re)enforce the presence of the state and make the state what it is in that particular moment (also see Colona 2019). More specifically, I am interested in understanding how this occurs within two policing partnerships in Nairobi, Kenya and how this is experienced by the individuals involved in these arrangements.

To do this, I draw – as Frossard and Jaffe (2018) do – from Star and Griesemer (1989), and analyse the firearm as a ‘boundary object’ that acts a bridge in bringing policing actors together. In their article, Star and Griesemer focus on the heterogenous nature of scientific work that inherently comprises different actors, perspectives and practices, and thus often produces tension. By using the Berkeley’s Museum of Vertebrate Zoology as their case-study, the authors analyse how translation across various settings and actors is needed and how this can be achieved through the standardisation of methods and development of boundary objects. They describe boundary objects as artefacts that inhabit ‘intersecting social worlds and satisfy the information requirements of each of them’ (393). They accentuate how boundary objects may have different meanings to different groups and in diverse settings, yet that they make sense to those groups that communicate through and with them. Taking this, boundary objects comprise knowledge that becomes shared and thus enables alliances across difference.

In an article written two decades later, Star (2010) further outlines what a boundary object entails by highlighting four crucial dimensions. The first is collaboration: boundary objects have the fundamental objective of allowing ‘different groups to work together without consensus’ (602). The second is that objects are ‘the stuff of action’ (603) and that ‘materiality derives from action’. The third is that an object does not have to be an actual ‘thing’, but is ‘something people act toward and with’ (603). A theory, she argues, can thus act as an influential boundary object. The fourth is that boundary objects are temporal – their function and purpose is not permanent, but emerges from and during specific moments of shared space in which various parties (and their perceptions and practices) come together.

I draw from these two pivotal articles and argue that the firearm is a boundary object: through its presence, various public and private policing actors are provided with the potentiality to interact and share a certain (policing) space. And as will be made clear in the following section, despite there being divergent ideas of how and by whom the firearm should be used in the policing realm (i.e. lack of consensus), the firearm has a bridging function between various policing actors and provides new possibilities for certain actions. However, in contrast to the work of Star and Griesemer (1989), and in line with Frossard and Jaffe (2018), I also want to emphasise that besides bridging, boundary objects also divide and as the term implies, reinforce and consolidate boundaries. Although the firearm (and the vehicle) bring private security companies and police officers together, the firearm (and especially its use of) exemplifies the power hierarchies between the two parties and reaffirms the dominant role of the state police in the policing realm. Echoing what Sharma and Gupta (2006, p. 17) contend on the role of documents: the firearm further demarcates between what is state and what is not the state. This paper aims to flesh out this dual-sided nature and thereby contribute to the (interdisciplinary) work on boundary objects in the field of policing.

‘The firearm belongs to us!’

In April 2018, a newspaper article titled, ‘Private security guards to be issued with guns in new move’⁴ sparked renewed uproar on an ongoing and heated issue in the Kenyan security sector. According to the article, a new proposal to arm private security officers had been approved by the National Security Advisory Committee, thereby going against the legislation that forbids private security officers to be armed. The approval was largely based on a move to better equip security officers in the fight against crime and terrorism.⁵

For decades, Nairobi has faced high crime rates and recurrent terrorist attacks. As a result, the city hosts a myriad of various public and private security providers and ample scholarly research outlines how such actors operate across diverse terrains, from the informal settlements (van Stapele 2015, 2016, Price et al. 2016, Skilling 2016, Jones et al. 2018), to the middle-class housing estates (Smith 2015), to the wealthier parts of North-western Nairobi (Colona and Diphoorn 2017). In addition to the numerous forms of community policing, ranging from formalised community policing forums, vigilante-type groups, gangs, and neighbourhood watches, two main actors, which also lie at the heart of this article, are the National Police Service (NPS) and the private security industry.

Similar to other British colonies, the Kenyan state police emerged under colonial rule and the corps was divided into two sectors: the Administrative Police (AP) and the regular police, also referred to as the Kenyan Police (KP). The AP was known as the ‘tribal police’ and supported provincial administrations and chiefs, while the regular police acted as the main state police and was engaged with more traditional policing duties, such as crime prevention and investigation. Since the establishment of the new constitution in 2010, the NPS has experienced numerous reforms (Osse 2016), and a main objective has been the amalgamation of the two forces under one overarching command. At the time of writing, more dramatic efforts were being implemented to achieve this objective: the majority of the AP officers are being brought under the KP, leaving only very specialised units of the AP intact, such as the Border Police Unit and the Critical Infrastructure Protection Unit (CIPU).

Broadly speaking, the NPS suffers from a poor reputation, described as being inactive (Musoi et al. 2013), incapable to deal with crime, in cahoots with criminals (Omenya and Lubaale 2012), and generally corrupt, ill-trained, and under-equipped. Furthermore, various forms of documentation show that police violence and extrajudicial killings have increased in Kenya throughout the past years (MSJC 2017, Jones et al. 2018). This was particularly so during the presidential elections that dominated the political scene in the second half of 2017. Although different organisations, such as Amnesty International and the Kenya National Commission on Human Rights (KNCHR), have released different statistics on the injuries and fatalities on behalf of the police, there is a general consensus that the behaviour of the police has dramatically worsened, especially in the lower socio-economic parts of various urban centres.

In Kenya, the private security industry originated in the 1960s and has experienced continuous growth since then, with an exponential boom after the Westgate mall attack in September 2013. It is estimated that over 2000 private security companies and 400,000 security officers operate in Kenya (Mkutu and Sabala 2007), although respondents more recently used the figure of 500,000 guards. One of the key dimensions that has dominated the private security scene in Kenya has been regulation. Up till 2017, there was no form of state regulation, so self-regulation initiatives were enforced by the two leading employers’ associations: the Kenya Security Industry Association (KSIA), which comprises the larger, more high-end, international companies, and the Protective Security Industry Association (PSIA), which mainly consists of Kenyan owned companies geared towards the acquisition of government contracts (Diphoorn 2016b).⁶ Combined, the KSIA and PSIA enforced most of the regulatory efforts, yet this changed in 2016 when the Kenyan Parliament finally approved the Private Security Regulation Act after more than 20 years since its first drafting. This was applauded by all sides of the industry, who saw it as a formal recognition of the industry by the state. In 2017, the Authority fully came into play and it has started to play a more influential role in the industry.

One of the key topics that has defined the regulation process (and perhaps the reason for the delay in approving the bill) has been the issue of arming security officers. During interviews and meetings that I attended, this issue habitually resulted in heated discussions. Many private security personnel are adamant about arming guards, citing the risks involved of providing poorly trained guards with a lethal weapon. The claim is also that Kenya already faces an excess of firearms, and that more firearms will lead to more insecurity. In contrast, other members of the private security industry are proponents of arming guards, arguing the need for extra protection to fight crime, as can be seen from the following statement by a company owner:

My opinion has been yes [to arm] … What reason is there not to arm them?  … I have lost more than 40 men who are shot … And those 40 souls were innocently doing their work … And I’ve paid for that compensation very heavily. Why? And yet, at night, we attend alarms, left, right. You get there, you are shot at … So, we are not saying we arm every, Tom, Dick, and Harry … But, we arm those who are doing responses to emergencies. We arm those ones who are in a sensitive place.⁷

Another argument is that armed guards will enhance the reputation of security officers, who are seen as helpless in the fight against crime and terrorism: ‘In Kenya, a guard is a bad thing and the firearm might have changed this’.⁸ This sentiment is particularly voiced from various members of the Kenya National Private Security Workers Union (KNPSWU), who openly express their desire to arm guards. Yet they also highlight that this should primarily occur for specific types of guards in high-risk areas. The idea is that guards are defenceless against highly armed criminals and thus need more equipment to protect themselves.

These proponents were thus disappointed to hear that the legislation towards firearms was unchanged: since state regulation has been approved, private security guards are still not permitted to carry a firearm when on duty, entailing that in the policing realm, the firearm remains a monopolised object of the state police.⁹ Yet despite this, a loophole has been identified which allows some firearms to be issued to private security employees. This was explained to me by a manager who had worked for three different companies in the past five years. Very hesitant to discuss the situation, he explained how individuals transform their personal firearm permits into ‘temporary permits’ (TPPs). Simply put: a licensed firearm owner with a civilian firearm certificate can assign and extend his/her permit to other individuals for a period of one year. Within the private security industry, this primarily occurs with high ranking employees, responsible for ‘large and important clients’, and thus not for the ‘guards at the gate’. In 2016, another male member of the industry stated that ‘there are a few 1000s’ of these cases.¹⁰ He highlighted that there is not a maximum amount of TPPs that one can apply for and that this is ‘one of the many loopholes’. Although this entire process is monitored by the Firearm Licensing Board that was established in 2016, one male owner claimed that it is ‘very dangerous and risky’ and therefore not something ‘you do for everyone’. He then further reaffirmed the secrecy of the issue, saying that ‘we always gloss over it’, mainly because ‘the police wouldn’t like it’.

This idea – that the police do not want armed guards – was reinforced by all of the police officers I spoke to: none supported the idea of an armed private security officer. Although many cited poor training levels and a lack of experience, several police officers openly claimed that this would jeopardise their position of authority. Several police officers shared a sensation of entitlement, arguing that they should be in sole possession of a firearm. As one police officer voiced: ‘the firearm only belongs to us’¹¹ and others reinforced this statement. This feeling of ownership can be easily observed: during patrols, police officers always held their firearms close and they were certainly not shared with the security officers. As a middle-ranking male police officer said to me in 2018:

This firearm is only for us. And this one is mine. I need it when I work. If I could, I would have it with me, always, next to me, ready to take and use. There are many criminals who want to kill us, this is dangerous work, so I need it. (…) So these askaris [guards] cannot have this. They don’t have the training, the skills … they don’t know. So only we can keep it, it stays with us.

This crucial relationship between police officers and their firearms recurred during many interviews with police officers over the years and this was expressed both visually and verbally. The firearm was described as a ‘companion’ and a ‘brother’, and it was stated to give them ‘confidence’, ‘power’, and ‘control’. They discussed at length how this allowed them to ‘kill’ and to ‘deal with criminals’. Even more so, it was what distinguished them from society and gave them power over others. This was echoed by civilians, especially human rights activists, as the following quote from an activist shows: ‘intimidation is their main goal and they use the gun to do so. You know where a policeman stands by looking at how he holds his gun’.¹²

The firearm is thus an object with divergent meanings for different people who perceive it contrarily, both in its meaning and use. Yet despite these differences, it also holds a common identity across social worlds and sites, namely as an object that provides a certain violent potentiality and, as noted by Kivland (2018), an ability ‘to perform particular acts of power and violence’ (355). It is this commonality that allows it to act as a boundary object and play a key role in the creation of various policing partnerships in Kenya.

Policing partnerships in Nairobi, Kenya

The reality is thus that private security officers are unarmed (for now), and if security firms want to provide armed protection to their clients, they must team up with police officers. This process of ‘teaming up’ has occurred in both informal and formal ways in Nairobi, and in some cases, has occurred through support of or in cooperation with community based policing initiatives (see Colona and Diphoorn 2017). In this article, I focus on two formal policing partnerships between the private security industry and two different units of the National Police Service. Both partnerships entail that armed police officers join the private security personnel into their vehicles and operate along a user-pay framework.

I will draw from ethnographic fieldwork on policing conducted in Nairobi, Kenya between 2014 and 2018 on two different research projects. Between 2014 and 2016, my research specifically focused on state regulation and policing partnerships and I spent a total of five months in Nairobi. Between 2017 and 2018, I embarked on a new research project that examines the various institutions and mechanisms aimed at transforming the Kenyan police. For this project, I spent a total of twelve months in Nairobi. Although different topics, there was quite some overlap between the projects, especially with certain individuals, who I regard to be key informants. Combined, both projects resulted in approximately 240 interviews, and these ranged from semi-structures interviews with pre-determined questions to open-ended ones that were guided by a mental topic list and often largely steered by the interviewees.

On top of the interviews, I actively engaged in participant observation, which has often been the means in which anthropologists distinguish themselves from other disciplines. Yet numerous scholars, such as Ingold (2008) and Forsey (2010), highlight that participant observation should not be equated with ethnographic fieldwork: it is a method that largely defines ethnographic fieldwork, but it is one method among many, and almost all anthropologists combine this method with others. Traditionally speaking, participation observation is ‘a method in which a researcher takes part in the daily activities, rituals, interactions, and events of a group of people as one of the means of learning the explicit and tacit aspects of their life routines and their culture’ (DeWalt and DeWalt 2002, p. 1). Furthermore, as argued elsewhere, I define participant observation as a larger tool-kit that also includes systematically recording the observations made, establishing rapport, and being reflexive (Diphoorn 2017, p. 175). For both research projects, this entailed attending meetings and seminars, patrolling with public and private security providers, attending court sessions, and simply ‘hanging out’ with key stakeholders. In addition to a range of informal activities, I engaged in approximately 70 formalised activities during both research projects. And last, but certainly not least, I also conducted secondary data analysis, including MoUs, minutes of meetings, and newspaper articles.

Elsewhere I discuss the difficulties, disadvantages and advantages of ethnographic fieldwork, and more particularly, participant observation (Diphoorn 2013, 2017). Here I want to highlight two main advantages, especially when conducting research in the same place on similar issues. The first is that participant observation provides us with the ability to uncover both what people say and do. During numerous occasions, there were disparities between what people said during interviews and what people actually did. By employing both interviews and participant observation, one can identify and analyse such discrepancies and unravel what these differences mean. The second is longitudinal research, i.e. returning back to the same place and regularly interviewing people over time. This not only allows one to observe changes over time, but it is also crucial in establishing long-lasting relationships and building trust. These benefits will become evident in the following sections on two policing partnerships.

‘A win-win’: the diplomatic police unit (DPU) and alarm response

The first policing partnership is that between the Diplomatic Police Unit (DPU), a police unit of approximately 400 police officers that serves the diplomatic community in Kenya, and seven private security companies that provide alarm response services and are part of the Kenya Security Industry Association (KSIA). The DPU primarily tends to crimes committed against diplomats and the guarding of the property of diplomatic missions in Kenya, such as the US embassy and the residential compounds of ambassadors. The DPU police station is strategically located across the UN headquarters in Gigiri (North-western Nairobi), which is also the neighbourhood where most diplomatic missions reside.

According to the official records, the first meeting occurred on 11 April 2006, when three DPU officers, two UN employees, and seven representatives from the private security industry came together to establish the arrangement, thereafter labelled as a ‘joint security policing partnership’.¹³ The two core components of the partnership were outlined: (1) monthly crime sharing meetings and (2) joint-patrols. The monthly meetings take place at the DPU police station and participants share crime statistics, news reports, and other security-related issues. The partakers include DPU officers, representatives of private security companies, and other relevant stakeholders, such as security officials from diplomatic missions and members of resident associations. The joint patrols, such as the one with Michael in the introduction, entail that police officers join private security officers in the vehicles of the companies and patrol a specific geographical area. The patrols are the heart of this exchange: the private security companies provide the vehicles and the police officers provide the firearms, hence an exchange of ‘arms for mobility’. The vehicle and the firearm thus both act as boundary objects and provide possibilities for collaboration.

Although slightly different per company, the patrols generally occur in this fashion: the private security vehicle, often consisting of a driver and a security officer of higher rank, goes to the DPU police station. The OCS (officer commanding station) determines which (two) officers join the vehicle. Before heading out, they decide which areas to patrol and which static posts to check. Between 2014 and 2016, I accompanied several of these patrols and it became clear that the patrols were very ad hoc and not as formalised as presented. For example, the patrols do not occur from 6 to 6, as stipulated in the original minutes, and they are not as structured as claimed to be: this concerns the allocated nights on which the patrols are conducted, who initiates the patrol, which police officers and private security employees join the patrol, and which areas they patrol.¹⁴

In line with UN policy, the geographical area of focus within this arrangement is along the lines of the Blue Zone, the UN-demarcated area where diplomats are recommended and/or compelled to reside for safety reasons. When the partnership commenced in 2006, company representatives divided the larger Blue Zone area into eighteen other zones and determined who would be responsible for which zone. The allocation was based on where the private security companies were already operating, yet the system has become more flexible and informal since then.

As the DPU officers and private security companies operate in the same areas, they share the same clientele. Although DPU officers will also respond to another situation if it arises, their main concern are DPU posts and this differs from other police stations, who do more random patrolling and park off at particular areas, such as crime hot-spots, and (officially) serve the larger Kenyan public. For the private security companies, this means that they do not only attend to their clients, but often have to check the posts of their competitors. One private security manager discussed this change to responding to non-clients and increased focus on patrolling, signifying that this had resulted in a more pro-active policing approach for the companies involved. This suggests that boundary objects, when used to bridge different social settings, can result in new actions and modes of operandi.

In addition to the exchange of ‘arms for mobility’, police officers also received a financial contribution from the companies after the patrols, which was commonly referred to as a ‘lunch allowance’ and ‘token’. Under Section 45(1) of the Police Act, police officers can be hired for private use and these rates can be found online.¹⁵ The payment of officers operates alone these lines, yet many private security managers were very secretive about the matter. In fact, during an interview with the owner of one of the largest companies in Kenya, he rather aggressively denied that his company provided payment, stating that, ‘Paying cash to an OCS to facilitate deployment, that is outright corruption!’¹⁶

The rates differ slightly per company, yet the average pay is between KSH 500–1000 (equivalent to five to ten euros). Although some companies have a standard fee, the amount is habitually negotiated after the patrol, a process that I personally witnessed, and numerous factors are taken into account, such as one’s rank, the length of the patrol, the activities undertaken, and whether or not the OCS would also be receiving payment, either as a slice of the allowance or as extra payment on top of that. During 2014–2016, several police officers stated that this financial payment is their key motivation to participate in these patrols. As one male officers in his late 20s voiced: ‘It’s all about the money – if they don’t pay me, I stay here [the station]’.¹⁷ Other police officers detailed that they only join the companies that pay more. Furthermore, when police officers voiced preferences for a particular company, this was based on the amount of money they received, and not on policing-related skills, such as patrolling or the sharing of crime intelligence.

Interestingly, during interviews conducted in 2018, a very contrasting opinion emerged, namely that police officers regard these patrols with private security officers as a burden and that the small amount of KSH 500, which often only resulted in KSH 250, is incredibly low in comparison to the money that could emerge from more lucrative practices, such as bribery. The following extract of a (recorded) interview with an ex-police officer demonstrates this:

• I: So how did you feel about going on these patrols?

• R: I tried to refuse, but I couldn’t always. I mean, 250 shillings? Most people never wanted to go in those cars.

• I: You never wanted to go with them?

• R: No, if you go on patrol yourself, you can make so much more money. So why would you be limited into the box of a car?

• I: So these patrols were unattractive?

• R: They were very unattractive … unless I was commanded, I would never get into that car.

This contrast in opinion not only points towards the disparate ways in which such issues can be experienced, but also to the benefits of longitudinal ethnographic fieldwork and interviewing various people about certain issues over time.

While researching this policing partnerships, the most strikingly visible dimension was the hierarchical position of the DPU officers. As was the case with Michael in the patrol in May 2015, the DPU officers determine the entire sequence of the patrols. Although the company owns the vehicle and pays the driver, the DPU officers determine and manage the entire procedure. This hierarchy also exists in the meetings: although the non-state representatives outnumber the DPU officers and the chairman is the manager of one of the largest companies, the DPU representatives are given the last word, determine the content of the meetings, and exercise a type of ‘veto’ power.¹⁸ Therefore, although the partnership is described as a ‘win-win’, the role of the state police is regarded as indispensable by both sides of the arrangement. One police officer mentioned that ‘we do all the work’ and ‘we are the ones risking our lives for them’.¹⁹ Several others supported this claim and when doing so, the firearm was a central feature.

This attitude was supported by another male police officer during another patrol I joined in 2015. He was very interested in my former research on armed security officers in South Africa (Diphoorn 2016a). At one point, he mentioned, ‘but there, they don’t cooperate with the police, like here’. As this is not the case, I refuted his statement and outlined the numerous ways in which private security officers collaborate with the South African state police. He was absolutely stunned and confused by this and during the remainder of the patrol, he repeatedly asked me, over and over again, ‘But if they have the gun, why do they need the police?’ From his perspective, an armed security officer would have no need to work together with a police officer. His view not only points towards a perceived monopoly of the police of the firearm, but it also suggests that the firearm acts as the primary boundary object and that the vehicle, the other object that is exchanged, is irrelevant. Interestingly enough, I heard similar statements from private security employees. When asking them about their relationship with the police, the firearm was habitually mentioned as the sole driver for collaboration. When asking a security manager how he felt about paying the police, he responded: ‘it’s almost a saving, to pay police officers as opposed to your own paid guards, it’s actually cheaper. And with them, you get arms, so it’s even better’.²⁰

Therefore, in this partnership, the two key objects – the vehicle and the firearm – do not carry the same symbolic and practical significance. They both act as boundary objects that both bring the private security companies and the police officers together and accentuate the differences between them. The vehicle allows the under resourced police to conduct patrols and physically be present in certain neighbourhoods, and thereby reaffirms the financial constraints of the state police and the profit-making objectives of the industry. The firearm allows the companies to provide armed protection to their clients, and in the case of a dangerous situation, the actual firepower to protect themselves and fight crime. Yet the firearm, in a much more striking way than the vehicle, reaffirms their differences by emphasising the state police’s capacity to kill. Echoing the claims made by Springwood (2007, 2014) and Kivland (2018), the lethal potential of the firearm ascribes it with a unique quality that sets it apart from other objects. Furthermore, the vehicle is, albeit temporarily, shared by all participants during that patrol, while the firearm is not: it remains solely in the hands of police officers. Although the security officers, the companies, and the clients (and perhaps even other citizens) are affected by its presence, it is not an object that is physically shared or becomes entangled. Even within the shared space of a vehicle during a patrol, the firearm continues to demarcate difference.

‘A requirement’: Administrative police (AP) and cash-in-transit (CIT)

The second policing partnership is between the Administrative Police (AP) and private security companies that provide cash-in-transit (CIT) services. As mentioned, the AP was a separate unit (until recently) and primarily occupied with the security of government buildings and national borders across the country. Cash-in-transit – CIT – refers to the movement of cash from one place to another by private security vehicles. In this formalised arrangement, AP officers join the cash-in-transit vehicles to provide armed security.

This arrangement emerged after the Kenyan government ordered all CIT vehicles to be escorted by armed police officers after a surge of violent attacks on CIT vehicles. Although there were already some instances of AP officers accompanying CIT vehicles, none of this had been officially registered and this resulted in poor documentation and a lack of transparency. A well-known and often cited example is the robbery of a G4S vehicle in 2010 when more than KSH 80 million was stolen. Although eventually both G4S security officers and AP officers were charged, it had been difficult to rely on official records to prove this. Therefore, to prevent such incidents from reoccurring and to compel both parties to maintain accurate records, a formalised arrangement was established.

On 3 March 2010, an MoU was signed by representatives from eight different companies and the then Administration Police Commandant. According to the MoU, the Commandant is responsible for deploying ‘suitable/competent’ police personnel, coordinating the operations, and reshuffling the officers attached to the company. Furthermore, the AP officers, although seconded or attached to the companies, remain under the direction of the Commandant. Similar to the DPU arrangement, it is clear from reading the MoU that the AP is formally in charge.

In turn, the companies are responsible for most of the operational aspects: the companies are compelled to provide the AP officers with ‘induction courses’ (i.e. training), accommodation, transport, and so forth. Similar to the DPU partnership, AP officers also receive an allowance, yet in this case, this is formalised and stipulated in the MoU. The daily subsistence allowances range from KSH 500–1000 per day, depending on one’s rank. In addition, the companies also pay KSH 500 per day to the ‘government as appropriation-in-aid’.

What we thus see is a formalised arrangement whereby the CIT companies bear the operational responsibilities, yet the AP Commandant remains in control over the operation. In contrast to the DPU policing partnership that was characterised as a ‘win-win’ situation and was openly discussed, the CIT arrangement was regarded by many industry representatives as a burden and it was kept under the radar. In fact, people were very secretive about it and this was reflected in the difficulties I faced in collecting the data.

I first heard about the partnership during a meeting of the governing council of the Kenya Security Industry Association (KSIA), when it was mentioned that the members of the CIT committee needed to provide their payments. Additionally, another member asked what the status was with the ‘angry landlord’, which another member answered with the comment, ‘it is being looked into’. The entire issue – in contrast to others – was quickly brushed under the table and the next point on the agenda of the meeting was introduced, providing no space for further discussion.

When I inquired about this matter during follow-up interviews and conversations, many informants exuded discomfort. It was only a few weeks later, after repeated questioning on my behalf, that some people discussed it with me. It then became clear that the entire arrangement was extremely problematic, as is illustrated by the following quote from the manager of one of the companies involved: ‘the whole thing is an absolute headache which we really want to get out of’.²¹ The main reason for this is that the private security personnel felt that they were not getting their end of the deal. This largely revolved around financial issues: the total payment had increased substantially over the years. In 2015, all of the seven companies paid a total monthly amount of KSH 1,700,000, and this figure excluded the additional daily subsistence allowance fees.

Yet the main financial problem concerned the housing for the AP officers, which the CIT companies were obliged to pay. This is rather unusual: in Kenya, most police officers (up to a certain rank) reside in state-provided barracks adjacent to police stations and posts. However, in the MoU, it states that the AP will provide up to a maximum of 600 policemen that would solely be used for this CIT-purpose on the condition that the companies would house them. However, in 2015, there were only 276 police officers dedicated to CIT duties, and this was reflected in the pay slips of the companies that paid every month. For example, one of the largest CIT companies paid KSH 765,000 per month, of which they only received 102 policemen that were lodging in 51 houses. This comes to an average price of KSH 7500 per police officer. All of these costs were becoming excessive for the companies. As one private security employee mentioned, the main motivation for companies was to have ‘direct access’ to officers ‘that were dedicated to you and nothing else!’²²

In addition, several private security representatives claimed that they were not actually paying for the housing, but that the money was going directly into the pockets of the police commanders and that other individuals, such as the landlord, were profiting from the entire arrangement. The claim was thus that the housing was a façade and one company owner sarcastically described it as ‘great revenue for the police station’.²³ At one point, I wanted to visit these houses to speak to the officers living here, yet I was not permitted to do so on the premise of it being a ‘security risk’. However, interviews conducted with other police officers over the years seemed to confirm the accusations of the private security members, and they often laughed at the mere idea of the CIT companies paying for the housing. Regardless of what is true about this case, these allegations of corruption unquestionably influenced how the entire arrangement was experienced.

Yet despite these antagonistic sentiments and accusations of corruption, the involved private security personnel companies felt unable to cease the deal. To carry out CIT, private security companies are compelled to have armed protection, and as the police is the only legally armed security actor in Kenya, they must work with the police, as highlighted by the following quote from a company manager: ‘the only people who carry guns are the police; therefore you have to hire them to protect the cash that you are moving, whether you like it or not.’²⁴ Another company owner described the entire arrangement as a ‘government requirement’ by which ‘we don’t have a choice’.²⁵ Another manager described a case in the past where they had stopped paying the police, out of protest. As a result, they did not receive any police officers, became unable to operate, and lost several large corporate clients, such as primarily banks.

Yet simultaneously, despite these frustrations, the private security personnel also recognise the need of having armed protection: without the firearms, they would be attacked more regularly. One company owner stated: ‘Without the AP in our vehicles, the criminals would be a lot more daring’. The rationale behind this is that the death of an AP officer would lead to retaliation and ‘the thugs know this’.²⁶ Another manager concurred and detailed that:

it’s only on the basis of the way the system is set up that stops you from getting robbed because you know who the police are, you’ve got their IDs, you’ve got the log of their weapons, you know where you’ve picked them up from, you know where you’re assigning them to, so for a policeman to decide to rob a convoy, he’s got to realize that if he gets away with it, the police are gonna hunt him down forever.²⁷

Yet despite this recognition, this policing partnership can better be described as a form of dependency that is a necessity for the private security industry. For police officers, this arrangement is an additional source of income and this largely motivates them. As one high-ranking police officer smilingly said to me in 2018, ‘this is not an arms for mobility, but an arms for cash!’ This perhaps also explains why officers support the existing state regulation and will continue to resist pushes to allow security officers to carry firearms on duty.

This policing partnership provides insight into the use and function of boundary objects. In this partnership, the vehicle does not act as a boundary object: as this partnership is more a relationship of dependency, rather than an exchange, the vehicle does not really act on its own or bridge people and practices together. It provides a shared space, but this is primarily steered by the firearm: the firearm thus solely acts as the boundary object. Combined, this shows how boundary objects have different functions and effects in dissimilar contexts, and that these divergent meanings and usages need to be uncovered in order to understand the dynamics of policing partnerships.

Concluding remarks

In this paper, I have analysed two policing arrangements between two different state police units and various private security companies in Nairobi, Kenya. The first arrangement with the DPU is defined as a ‘partnership’ that is based on a direct exchange of ‘arms for mobility’ and is experienced as mutually beneficial. The second arrangement with the AP, in contrast, is not regarded as a partnership, but as a requirement: the private security members feel they are compelled to participate and allegations of corruption taint the entire arrangement. The differences between the two were also reflected in my ability to research them: whereas the DPU partnership was accessible and easy to discuss, it was difficult to collect data about the CIT arrangement.

Yet despite the differences between these two arrangements, they share the centrality of the firearm. In both cases the firearm acts as a boundary object that brings parties together and simultaneously reaffirms their differences. Furthermore, although the ‘arms’ are exchanged for ‘mobility’, these two counterparts do not carry the same symbolic and practical value and weight. Although DPU officers are in need of extra resources and appreciate the vehicles provided by the companies, they are not as indispensable as the firearms. And with the CIT-AP arrangement, the vehicle plays a very minimal role and does not act as a boundary object. In both cases, we see how the firearm – for what it is, represents, and does – provides the state with its hegemonic position. It is the firearm that, both symbolically and literally, represents and produces the state’s monopoly of force and the ability to kill. Therefore, despite the fact that the Kenyan state police has a poor reputation for being unreliable, untrustworthy, and corrupt, its position as being the sole agent legally permitted to carry a firearm provides it with an inescapable position of authority and power. And with recurrent pushes to permit security officers to carry firearms while on duty, it will be interesting to see how events unravel in Kenya in the near future and how this may impact these existing policing arrangements.

With these insights, I want to conclude this paper with two remarks. The first is a word of caution of overemphasising entanglements and interconnections when analysing policing partnerships and thereby applying concepts such as hybridity to understand them. Although policing is increasingly pluralised and often results in blurry lines of demarcation – issues that I have thoroughly discussed elsewhere – it is equally productive to examine the ways in which pluralisation reproduces distinctiveness. For example, Colona and Jaffe (2016) discuss similar patrols in Nairobi that involve community participation and they define these as ‘hybrid arrangements’ whereby ‘Residents have become so used to this symbiotic arrangement that they recognise the presence of one as the presence of the other … they have come to understand them as an empirical co-presence’ (6). I concur that these patrols entail a form of co-presence, particularly in the eyes of citizens. During my research I also observed how citizens were not always able to distinguish between the two and regularly assumed that security officers were armed, due to the police presence.

Yet I also assert that this copresence is highly temporary, and in some cases, very superficial. Although the joint patrols in both policing partnerships discussed here provide new spaces for certain actions, such as the companies’ increased attendance to non-clients and the employment of more pro-active policing styles, I do not think it is appropriate to label these as new political formations. This is primarily due to the highly temporal nature of the boundary objects (as Star outlines, 2010) that act as the foundation of these partnerships. Furthermore, we need to recognise that for the key participants involved – the security officers and the police officers – there is a fundamental hierarchical distinction between them that must be maintained throughout. In other words: the firearm acts as a boundary object: although it unites and provides opportunities for interconnections, it also, and sometimes more powerfully, divides and thereby prevents, rather than prompts, entangled practices that can be seen as varying degrees of ‘hybridity’. I thus argue that if there is any form of hybridity, it is extremely fleeting and limited due to the profound ways in which key objects (re)produce and consolidate boundaries.

Building on this, my second remark is a call for more in-depth studies on the role of materalities in the policing scholarship. This implies not only granting attention to the so-called technological revolution and implementation of new devices, but also a key focus on how more mundane objects, such as the radio, handcuff, uniform,²⁸ vehicle, baton, and many more define everyday policing practices and interactions. This involves in-depth analyses of how these objects are perceived, used and ascribed meaning across various social settings and actors and how this changes over time. And although this call can include various methodologies, I want to advocate an ethnographic fieldwork approach that prioritises one’s perceptions and experiences, combines an understanding of what people say and do, and further progresses an understanding of the relationship between subjects and objects (or humans and non-humans) and how this shapes various social realities across the globe.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by ERC Starting Grant: [Grant Number 337974]; Netherlands Scientific Organisation (NWO): [Grant Number 016.Veni.175.074].

Notes

1 This term emerged during an interview conducted with a high-ranking DPU police officer and the police liaison officer from the Department of Safety and Security from the United Nations (UN) in March 2015.

2 See Hicks (2010) for an elaborate discussion and historical analysis of this material turn, especially in anthropology.

3 In the few studies on firearm, the AK-47 has received the most attention (see Graves-Brown 2007, Hodges 2007, Saramifar 2017).

4 ‘Private security guards to be issued with guns in new move’, Standard Digital, https://www.standardmedia.co.ke/article/2001277867/private-security-guards-to-be-issued-with-guns-in-new-move, accessed 22 April 2018).

5 More recently, new pleas to arm security guards have emerged after the terrorist attack on the Dusit hotel in Nairobi on January 16, 2019. Although news reports vary, it seems that the Cabinet has approved the Private Security Regulatory Authority (PSRA) to move forward with allowing guards to carry firearms on duty. See, ‘Bid to arm guards is a ploy to push agenda of rich and powerful clique’, Standard Digital, https://www.nation.co.ke/oped/opinion/Arming-private-security-guards/440808-4993084-15fir79/index.html, accessed 26 February 2019.

6 For more information, see the websites: http://www.ksia.co.ke/ and http://www.psiasecurity.com/. In 2015, the Protective and Safety Association of Kenya (PROSAK) was also set up to specifically include individuals that are not aligned to the two other associations. For more information, see the website: https://www.prosak.or.ke/.

7 Interview: 21 April 2015.

8 Interview: private security consultant, 4 August 2016.

9 I recognise that there are many illegal firearms in Kenya possessed by a range of individuals, yet due to the scope of this paper, I will not discuss this dimension.

10 Interview: 4 August 2018. He further stated that the majority of them are used on conservancies where guards are given guns to protect the conversancy against, for example, poaching.

11 Interview: 26 March 2015.

12 Interview: human rights activist, 15 September 2017.

13 In other meeting minutes, the partnership is labelled as the ‘community policing working group’, because, as one security manager explained, ‘it’s actually for the community’ (interview: 14 April 2015).

14 In fact, the patrols occurred far less than they initially did. One reason for this is that many companies have set up similar, yet more informal, arrangements in other neighbourhoods and with other local (non-DPU) police stations. These are often strategic areas for them, i.e. where they have more clients.

15 See: http://www.kenyalaw.org/KenyaGazette/view_gazette.php?title=2086, accessed 25 September 2018.

16 Interview: 27 April 2015, conducted with Dr. Erella Grassiani.

17 Interview: 25 May 2015.

18 Some participants of this partnership argue that this is largely due to the status of the DPU, which is a police unit known for being ‘elitist’ and having ‘the best officers in Kenya’. Some participants even claimed that this partnership could only work with the DPU, yet the existence of similar policing arrangements elsewhere (Colona and Diphoorn 2017) refutes this claim.

19 Interview: 25 May 2015.

20 Interview: 14 April 2015.

21 Interview: 15 May 2015.

22 Interview: 10 April 2015.

23 Interview: 27 April 2015, conducted together with Dr. Erella Grassiani.

24 Interview: 15 May 2015.

25 Interview: 27 April 2015, conducted together with Dr. Erella Grassiani.

26 Interview: 25 May 2015.

27 Interview: 15 May 2015.

28 There is some work on the role of the uniform of police officers (see Bell 1982; Joseph and Alex 1972), yet this work focuses more on what the uniform symbolises and how it is viewed by others, and not on what the uniform does, i.e. the active part of objects.