Rethinking ‘policebuilding’

Introduction

In recent years, the stabilization of so-called failed or weak states has been labelled as one of the greatest challenges of international security. Hereby, the provision of security has come to be regarded as one key prerequisite for statebuilding: ‘Without security, the essential political, social, and economic tasks of state building cannot be accomplished, yet in the wake of conflict, the local institutions necessary to provide this security are often lacking’ (Edelstein, 2009: 81). Whilst security certainly is not the only aspect of statebuilding (Hampson and Mendeloff, 2007), Jackson argues that ‘the provision of security’ is situated right at the ‘very centre of what states are’ (Jackson, 2011: 1804). Against this background, police assistance has moved centre-stage in strategic thinking on international security in the last couple of years (Hansen, 2011). Day and Freedman hold that order and stability in the context of external interventions would be best ensured through ‘policekeeping’ rather than by military force alone (Day and Freeman, 2003). This is, by and large, based on two interrelated premises: first, military forces have become viewed as often being ill-suited for post-Cold War ‘peace operations’, ‘crisis management operations’ or ‘operations other than war’ (Day and Freeman, 2003; Schultz and Merrill, 2007); and second, going beyond operational aspects of the police in external operations, it is widely assumed that the level of effectiveness of the police in effectively dealing with violence, crime and disorder in the aftermath of civil war, armed interventions and regime change will ‘determine whether new governments will be regarded as worth supporting’ (Bayley and Perito, 2010) and thus determine the ‘sustainability’ of such interventions. Such a view has been present in exemplum during discussions over NATO’s withdrawal from Afghanistan, whereby the Afghan National Police (ANP), because it regularly interacts with the population, are viewed as ‘in many senses, the face of the Afghan government’ (NATO, 2010). Greener has described this move of the police from the margins to the centre of debates on international security as ‘a new international policing paradigm’ (Greener, 2011: 183).

Within the new international policing paradigm, one of the key controversies has been about the so-called militarization of the police. Whilst police forces during the Cold War were mainly deployed under United Nations (UN) auspices to monitor peace agreements, they are nowadays fulfilling a much wider range of tasks ranging from traditional monitoring and advising to reforming, training and restructuring of indigenous police forces. In some cases, like Kosovo or East Timor, international police held an executive mandate and thus carried out police duties on the ground whilst rebuilding national police forces from scratch. ‘Policebuilding’ in countries such as Afghanistan or the Democratic Republic of the Congo (DR Congo) has often gone beyond the export of democratic, civilian policing techniques, featuring robust, military elements of policing such as riot control or even counter insurgency.

It is especially the latter type of robust, often militarized, type of policebuilding that has recently caught the attention of scholars and policy-makers alike. According to various observers their added value lies in the ability to overcome ‘the public security gap that invariably confronts – and frequently confounds – peace and stability operations’ by filling ‘a critical gap in capabilities that arises between military contingents that are proficient in the use of lethal force and individual international police, who have very limited force options and may even be unarmed’ (Agordzo, 2009; Dziedzic and Stark, 2006; Kelly, 2009). Critics, on the other hand, have raised concerns over the negative effects a militarization of police assistance can bring, such as a decline in public trust of the police and the misuse of police as ‘cannon fodder’ (Friesendorf, 2011; Greener, 2011; Murray, 2007; Sedra, 2006; Wilder, 2007).

Against this controversy, this article argues that the underlying assumption of a clear military–police divide, according to which the police are understood to be a civilian institution that mainly ‘serves and protects’ while the military ‘breaks things and kills people’, is theoretically and empirically flawed. While the blurring between military functions and police functions might be more bleak and observable as part of international interventions in so-called zones of disorder than in domestic settings, this article suggests that we need to understand the police theoretically as part of a single continuum of state institutions designed to simultaneously serve and protect the population and to establish and maintain state power through (sometimes) forceful, exclusionary means. It is inherent to the ‘Janus-face’ (Merlingen and Ostrauskaite, 2006) of the police, and to ‘policebuilding’, that both functions are competing and complementary; they are complementary when socio-economic and political order is under threat/unstable; they are competing when a certain political and socio-economic order is stable and maintained.

The paper will illustrate the argument by tracing the evolution of European Union (EU) international police power. There are several reasons for the selection of this particular case study. First and foremost, the EU has in recent years become one of the main actors of international police assistance – both in terms of the numbers of personnel and the numbers of missions the EU has committed itself to, as well as in strategic terms with regard to the growing importance of the EU’s civilian crisis management capabilities, of which the police are deemed to be the most important feature. Secondly, police assistance has thus far largely been portrayed as a civilian tool aimed at reforming local police agencies in line with European standards of democratic, civilian policing. Coming from there, the EU has been quick to assert that its very own mode of ‘policebuilding’ does not aim to create militarized police forces. Hence the EU has been regarded, especially in contrast with US-led policebuilding, as an actor that engages in civilian policebuilding. Against these assertions, it can be assumed that EU policebuilding should contain little or no trace of policies that hint towards forceful, militarized or illiberal policing capacities. Findings of such traces would, on the other hand, speak in favour of a theoretical understanding of the police which questions the inherent binaries of police/military and civilian/military, thereby offering insights and explanations previously missed in the literature on policebuilding.

The rise of policebuilding

It has become a common point amongst scholars and policy-makers alike to argue that ‘weak and failing states have arguably become the single most important problem for international order’ (Fukuyama, 2004: 92) in the 21st century. Hereby, it is specifically the inability of states to carry out one of their core functions – to establish and safeguard internal order and security – which has raised security concerns amongst the international community. This notion of disorder, which is at the heart of debates over weak and failed states, has police rather than military connotations. The military’s role in international security has traditionally been understood (often erroneously) as being involved in centrally organized, highly structured modes of conflict between, or within, states. On the other hand, police operations have been portrayed as ad hoc, anomic, reactive responses to disorder within states. However, in the context of the blurring between external and internal security, it has been observed that post-Cold War military operations resemble more and more police operations (Dean, 2006). These changes can be understand as part of a wider transformation of Third World states from possible strategic partners or proxies into what Duffield calls a ‘potentially dangerous social body’ (Duffield, 2002: 1067). Therefore, any form of aid or assistance is now directly concerned with reconstructing and transforming this social body into non-dangerous (that is, liberal) societies. Accordingly, traditional inter-state warfare has been transformed into global police operations as states are increasingly ‘defining their security interests more in terms of crimefighting than in terms of warfighting’ (Andreas and Price, 2001: 31).

In practical terms, however, because of the fact that the international community has never paid much attention to the role of police forces in the context of external interventions, the lack of capabilities to rapidly deploy police forces to zones of conflict has created an enforcement gap, which in turn has forced the military to carry out police tasks (Oakley et al., 1998). Against this background, many scholars have observed that traditional military forces are not well suited to the task of ‘establishing security in precarious political environments’ in the context of ‘peace operations’, ‘stabilization efforts’ or ‘operations other than war’ (Day and Freeman, 2003: 304). A seminal study published at the end of the 1990s has pointed to the fact that military force alone is a ‘blunt instrument’, able only to enforce ‘a most basic, rigid form of order’ (Oakley et al., 1998). Others have stated that, while the ‘insertion of international military forces may suffice to halt open conflict, separate combatants, and begin disarmament’, military forces are usually unable to successfully protect the population from ‘criminals and extremists who seek to undermine the emerging order’ (Dobbins, 2008: 47). What is more is that military forces are often reluctant to engage in policing activities, because of the fact that they are neither trained nor equipped for what are essentially police tasks: riot control, engaging in confrontations between civilians or dealing with ordinary crime (Bayley and Perito, 2010; Marten, 2007; Neild, 2001). One scholar has summed up this argument, found in much of the literature, by stating that ‘police officers were better prepared to work with a population to help it achieve internal security; soldiers were trained, as the saying goes, to break things and kill people’ (Marten, 2007: 242).

It has hereby been noted that the blurring of divisions between the military and the police has not only had a transformative impact on the military, but has also had an impact on the police. Various studies have illustrated a growing militarization of police forces domestically in the US, manifesting itself in the increased usage of heavily armed paramilitary police units (PPUs, SWAT teams) as well as the militarization of the US Border Patrol (Andreas and Nadelmann, 2006; Andreas and Price, 2001; Campbell and Campbell, 2010).

On the international level it has also been noted that civilian police forces deployed in international missions have come to resemble paramilitary forces as they are often deployed in contexts which are characterized by a high degree of hostility (Cline, 2003; Lutterbeck, 2004: 50). A number of scholars have hereby argued that the militarization of police in the context of peacebuilding and stability missions has actually had a detrimental effect on the reconstruction efforts (Friesendorf, 2011; Hills, 2001; Larivé, 2012; Murray, 2007; Sedra, 2006; Wiatrowski and Goldstone, 2010). This is because paramilitary police forces often fail to perform what Bayley and Perito (2010) have called the ‘core functions’ of the police: establishing the rule of law and thereby increasing the legitimacy of the government vis-a-vis the general population. The underlying claim hereby being that it is essentially ‘a well-functioning police force that enjoys civic trust which affords society the degree of order, predictability and accountability needed for the functioning of a market economy in a democratic system’ (Hinton and Newburn, 2009: 5). On the other hand, some scholars have refuted the claim that the police and the military have become increasingly blurred on empirical grounds. Studies by Greener (2009) and Weiss (2011) have argued that the blurring of military and police tasks has thus far been limited to strategic debates led on the macro-level. On the ground, as part of peacebuilding and stabilization operations, it is argued that the police and the military still follow different tactics, and still differ significantly in their values, their organizational cultures as well as with regard to the legal frameworks under which they operate.

When reflecting on the debates over the blurring of police and military functions it becomes apparent that, whatever the merit of a police–military divide for supporters and/or critics of the militarization of police, such a divide ontologically suggests a ‘pure’ starting point, implying that at one point there were distinct police and military spheres. This, in turn, involves the creation of a number of binaries: most importantly police/military, but also military/civilian, domestic/international and war/peace. All of this is, first of all, empirically problematic. The strict separation between the military and the police, which has been at the core of the Anglo–American model of the police and has thus informed the thinking of many scholars, has never been fully distinct outside of Great Britain and the US. In many countries, such as Indonesia, the Philippines, Turkey or South Africa, the police, or at least parts of it, have been military in character, tradition and function. Integrated into the armed forces under the command of the Ministry of Defence and trained by, or even comprised of, military officers, they were heavily armed to quell unrest and to conduct counterinsurgency activities.

Paramilitary police forces, so-called gendarmerie forces, are also to be found in various European countries. ¹ They have military origins in terms of their personnel, equipment and outlook as they ‘served to deal with particularly severe forms of internal strife and turmoil, which in many European countries accompanied the nation building process’ (Lutterbeck, 2004: 47). Hence, while the blurring between military functions and police functions has, at least according to some scholars, been very bleak and discernible as part of international interventions in so-called zones of disorder, it is widely observable in domestic settings, too.

More so, when examining the export of police to colonies and later to so-called Third World countries one finds a similar picture. Established first in Ireland and consecutively spread to other colonies, British colonial police forces mainly functioned as semi-military forces in place of the army, with their primary function being the repression of any kind of opposition to colonial rule. It was, by and large, military in nature, from the ranking system to the uniforms, to armament and to the training of recruits (Chandavarkar, 1998; Deflem, 1994; McGloin, 2003; Vinod Kumar and Verma, 2009). Similarly, other colonial powers also relied heavily on a paramilitary colonial police apparatus to establish colonial rule (Bloembergen, 2007; McCoy, 2009; Thomas, 2005). Armed and with military training, the colonial policeman was viewed as a soldier in a police uniform (Sinclair, 2006: 26). During the Cold War police assistance to countries facing a ‘Communist threat’ became an important part of US foreign policy (Rosenau, 2003: 76) and was mainly directed at the suppression of threats to the colonial order in order to strengthen client regimes against ‘Communist subversion’. This was mainly done by effectively militarizing police forces to become a vital force in counter insurgency (Kuzmarov, 2009: 192).

Secondly, conceptualizing the police as a civilian government institution tasked with upholding law and order, as well as a ‘force for public safety and security that is independent of political agendas’ (Greener, 2009: 25), is theoretically not without contestation either. Overstating the differences between the police and the military in terms of mission, command structure and professional ethos on the conceptual level is seemingly, in many ways, a result of under-theorizing on police and policing. While political scientists, especially in security studies, often use both the terms ‘police’ and ‘policing’ interchangeably and mainly in dissociation from the military, criminological theorizing on policing and police lends itself to a different theoretical perspective on police, policing and thus on ‘policebuilding’ too. Criminologists generally understand policing as ‘a wide range of regulatory practices that serve to monitor social behaviour and ensure conformity with laws and normative codes’ (O’Brien and Yar, 2008: 122).

Such a theoretical understanding highlights a number of important aspects of what ought to be understood as ‘policing’: first, ‘policing’ is not to be restricted to a single set of activities or practices, i.e. fighting crime, but encompasses all ‘regulatory practices’ interlinked with the control of social behaviour in order to ensure conformity with legal and normative codes; second, a wide range of social actors and institutions can be involved in ‘policing’; and third, ‘policing’ can be formally as well as informally organized. In contrast to this definition of ‘policing’, the ‘police’ are a formal state apparatus in charge of upholding order directed by and accountable to the state. Hence, the police are only one of many institutions involved in policing; nonetheless, the police play a central role in the context of the establishment of modern statehood as the means of acquiring and retaining state power (Bittner, 1985; Gordon, 2005: 62; Lynch et al., 1989: 64).

What is more, some authors have argued that our conceptualization of the police needs to move beyond the means of simply acquiring and retaining state power (Dubber and Valverde, 2006). Accordingly the police, because they are one of the central means of acquiring and retaining state power, must be understood as a fundamental condition for the establishment, and safeguarding, of a particular political and socio-economic order (Sheptycki, 1999: 218). In line with such a conceptualization, the police are not to be defined over a fixed set of public roles or activities (i.e. ‘serve and protect’), but through their function as a tool for the fabrication of a particular social, political and economic order (Neocleous, 2000: 58). Conceiving of the police a priori as a ‘civilian’ actor has been criticized as succumbing to a ‘liberal myth’ (Neocleous, 2011: 156). To be sure, the criminalization of disorderly subjects makes for a different logic of action: rather than to destroy the enemy, the aim of policing is to modify the disorderly subject’s behaviour according to the norms of the dominant political and socio-economic order in order to secure it. Hereby, judging social behaviour as ‘criminal’ or ‘disorderly’ is, as no form of behaviour is ‘deviant’ or ‘criminal’ in and of itself, to judge it from the particular perspective of the police ‘as a condition of good or even civilized order’ (Dean, 2006: 189). Hence policing is also a form of power that builds on dominance over the definition of what ‘public order’ is and which behaviour violates ‘order’ and is thus deemed to be ‘criminal’ and ‘disorderly’ (Ryan, 2013).

Seen through such a lens, police work is not to be understood as a neutral, technical activity – quite the contrary. Scholars have stressed the inherently political nature of police work (Yarwood, 2007: 458). This is mainly because ‘all relationships which have a power dimension are political’. Policing is inherently and inescapably political in that sense, because ‘in a society that is divided on class, ethnic, gender, and other dimensions of inequality, the impact of laws, even if they are formulated and enforced impartially and universalistic, will reproduce those social divisions’ (Reiner, 2000: 8). Hereby, many criminologists heavily borrowed from Foucault, who has suggested understanding the ‘police’ as a technology of government that provides for liberty and security, while it does so at the same time via forceful, violent and exclusory practices. Hence Foucault treated the ‘police’ as a ‘governmental technology’ of power ‘peculiar to the state’ with the objective of governing the population (Foucault, 1995).

Against this background, the debate over whether ‘policebuilding’ should entail civilian or more robust policing needs to be reconsidered. A more theoretically informed conceptualization of the police, which defines them not a priori as a civilian agency in disassociation from the military but as a tool to establish and maintain a particular (liberal) social order through a wide range of practices – including at times illiberal (violent, exclusionary) means – points us towards an understanding that sees ‘civilian’ and ‘militarized’ policing not as competing but complementary techniques of governing disorderly populations. It can be assumed that both types are competing when a stable liberal order is established at the same time as the overt use of brute force as a mode to manage disorder has been replaced by a more ‘reasonable’ use of force. Both types, however, can be assumed to be complementary when the state’s monopoly of the legitimate use of force is being challenged. As Ryan (Ryan, 2013: 456) holds: ‘The use of force by the “other” will be always judged irrational or unreasonable. The greater the level of illegitimate force exerted against police, the more reasonable it is for police power to move up the scale of violence in its reaction’. The following part will illustrate this theoretical argument by tracing the evolution of the EU’s ‘policebuilding’ activities.

The Janus-face of ‘policebuilding’: The case of CSDP police missions

Most analysis of Common Security and Defence Policy (CSDP) police missions generally starts with the establishment of EUPM Bosnia in January 2003 (Merlingen and Ostrauskaite, 2006). While the European Union Police Mission (EUPM) Bosnia was the first police mission to be carried out by the EU, discussions within the EU on so-called crisis management missions had started as early as in the 1990s. The Petersberg Declaration of 1992 explicitly mentions the need to enhance ‘capabilities for conflict prevention, crisis management and the peaceful settlements of disputes’ (Council of Ministers of the Western European Union, 1992: 1).

In 2000, member states agreed on four priority areas – police, strengthening the rule of law, civilian administration and civilian protection – in which the EU would need to develop civilian capabilities for crisis management. The Police Action Plan adopted in 2001 called for the provision of a 5000-strong police force for international missions by 2003, of which 1000 were supposed to be deployable within 30 days (Council of the European Union, 2001b: 16).

It wasn’t until 2002, however, that the debates shifted from purely quantitative, operational aspects to the functional aspects of the police (Council of the European Union, 2002b). The EU Comprehensive Concept for Strengthening of Local Police Missions depicts the ‘strengthening of local policing capacities’ as a ‘key function in conflict prevention, crisis management and post-conflict rehabilitation’. The tackling of this objective is then framed as to ensure that ‘local police forces are able to undertake the full range of police functions with strict respect for human rights, fundamental freedoms and the rule of law in accordance with the demands of a democratic society’ (Council of the European Union, 2001a: 20).

In line with this, the EUPM in Bosnia in 2003 was designed as a non-executive, unarmed mission. The EUPM was to achieve three ‘common goals’: to support ‘a stable and peaceful Bosnia and Herzegovina’, to foster ‘Bosnia’s progress to Europe’ through the stabilisation and association process, and to ‘continue the development of the EU’s external identity’ (Solana, 2003). The respective joint action thus sets out the mandate as establishing ‘sustainable policing arrangements under BiH ownership in accordance with best European and international practice, and thereby raising current BiH police standards’ (Council of the European Union, 2002a). In a letter to Amnesty International in December 2002, Solana linked EUPM with ‘best European and International practices’ of democratic policing and the mainstreaming of human rights (Solana, 2002).

Hence, police assistance was first and foremost presented as a way to support the EU’s neighbours in attaining European standards of democratic, civilian policing. This particular approach to ‘policebuilding’, with its main characteristics being ‘monitoring, mentoring and advice’ (Wright and Auvinen, 2009: 188), was later dubbed the ‘Balkan template’. The ‘Balkan template’ was described as a non-executive, civilian mission with a mandate focusing on capacity-building through long-term reform of the police and the wider field of rule of law (Korski and Gowan, 2009: 11). The ‘Balkan template’ was decisive in shaping later missions and became the EU’s default approach for police missions in the Balkans, in Palestine, the DR Congo and Afghanistan.

However, while police assistance has generally been presented as a civilian tool in the context of the EU’s emerging CSDP, the Comprehensive Concept for Strengthening of Local Police Missions envisions two types of future EU police missions: first of all, the strengthening of local police forces; and secondly, the substitution for local police forces (Council of the European Union, 2002b). In line with these two different types, the council called for the development of a broad range of capabilities to deliver police assistance, ranging from training, advising and mentoring to performing executive police functions. In order to be able to carry out both, non-executive and executive, police missions, the Comprehensive Concept for Police Substitution Missions and the Progress Report on Civilian Aspects of Crisis Management, both published as early as 2003, stressed the need for the new police capabilities to ‘comprise both police forces with civil status and police forces with military status of gendarmerie type’ (Council of the European Union, 2003a: 3, 2003b: 9).

The latter were designated as ‘integrated police units’ said to ‘perform executive tasks in order to re-establish the law and order in non-stabilised situations’ (Council of the European Union, 2003a: 2). These so-called integrated police units were to operate, at least temporarily, under the military authority during the initial stage of CSDP missions, keeping in mind that ‘the missions assigned to them are different, but complementary’ (Council of the European Union, 2003a: 10). Hereby, the document draws on lessons learned from previous police missions of the international community, namely the UN’s Brahimi report on future peace building operations, which suggested the establishment of robust, militarized police capabilities for future peacekeeping missions (United Nations, 2000). In line with these more general observations, the development of paramilitary, gendarmerie-type police forces was perceived as ‘a qualitative asset for the European Union’ in the context of international security (Council of the European Union, 2001a: 7).

In many ways, discussions over future missions in ‘non-benign’ territories mirrored the key points of the European Security Strategy (ESS) in late 2003, which held that due to the emergence of a wide array of new, transnational security threats, such as terrorism or failed states, the boundaries between internal and external security had become blurred (European Union, 2003: 7). Coming from there, the ESS argued that it is in the EU’s strategic interest to develop capabilities ‘that foster early, rapid, and when necessary, robust intervention’ across the globe (European Union, 2003: 1). The policy imperatives of this were laid out in exemplum during a speech by Benita Ferrero-Waldner (then European Commissioner for External Relations and European Neighbourhood Policy) in early 2005, during which she described the new raison d’etre of CSDP as the export of stability beyond the EU’s borders, highlighting the importance of stable, functioning state institutions in other parts of the world (Ferrero-Waldner, 2005: 4).

First glimpses on how such a robust, paramilitary conceptualization of ‘policebuilding’ in ‘non-benign’ environments had an impact on actual EU police missions were to be found during the launch of EUPOL (EU Police Mission) Kinshasa in the DR Congo in early 2005. Hereby, the objective monitoring, mentoring and advising of the Integrated Police Unit (IPU) in Kinshasa was framed not so much as a way to attain European best practices of policing but as a tool to ensure ‘the protection of the state institutions and reinforce the internal security apparatus’ (Council of the European Union, 2004). EUPOL Kinshasa’s main objective was to build-up a paramilitary IPU as a ‘neutral force’ to ‘guarantee the security of the government and transitional institutions’ (Council of the European Union, 2004), ensuring that ‘the actions of the IPU are in line with international police best practices’ (Council of the European Union, 2005). The focus hereby was on crowd control and riot control as well as to supply police equipment ranging from batons and tear gas to UZI and AK47 sub machine guns (Deutscher Bundestag, 2010). The overarching strategic aim behind the provision of police assistance to the DR Congo, specifically to the capital Kinshasa, was the facilitation of successful elections. It was assumed that a successful electoral process would in turn ‘enable the peaceful transition to a democratically elected government’ (Solana, 2007).

However, various reports by the UN and human rights organizations indicate that the Congolese police, including, at various incidents, the IPU, were used by the then president of the transitional government, Joseph Kabila, who in 2006 won the elections, to intimidate political opponents. Human Rights Watch for example reported the lootings of Movement for the Liberation of Congo (MLC) party offices (the MLC led by Bemba was the main contestant in the election), harassment, torture and intimidation of MLC party activists as well as local journalists, and also an estimated 500 extrajudicial killings before and directly after the elections committed by state security forces (Human Rights Watch, 2008: 44). A report by the UN’s MONUC (United Nations Mission in the Democratic Republic of Congo) human rights division into violent incidents involving the Congolese police and the Bundu Dia Kongo (BDK), a political–religious group based in Bas Congo that had aligned itself with Bemba, found that during demonstrations led by the BDK against vote rigging in early 2007 police and government soldiers shot or stabbed to death 104 BDK supporters. Furthermore, in March 2008 the police made a pre-emptive strike in anticipation of further protests, in what UN investigators said appeared to be a deliberate effort to wipe out the movement:

The size and composition of the PNC force deployed, the comprehensive geographical dimension of the operations, the type of weapons and ammunition used, the excessive use of force employed and the arbitrary executions that were carried out, the systematic destruction of BDK temples and houses and the large number of arbitrary arrests all suggest that the authorities may have intended to considerably reduce the operational capacity of the BDK movement (Office of the United Nations High Commissioner for Human Rights, 2008: 3).

During the course of events, over 200 alleged BDK supporters and others were killed and the BDK’s meeting places were systematically destroyed. The report specifies that these operations were largely carried out by the Rapid Intervention Police, which were built up by France through a bilateral police reform programme, and the EU-assisted IPU (Office of the United Nations High Commissioner for Human Rights, 2008). While EUPOL Kinshasa, as well as its predecessor EUPOL Congo, were portrayed as missions that would simply provide to the electoral process by protecting the transitional government and its institutions, these reports indicate that the IPU was used by Kabila to intimidate political opponents during and after the elections. Despite all of this, Kabila was openly supported by the EU before and during the elections as the candidate of choice. Diplomats, according to another Human Rights Watch report, preferred to ‘look the other way’ and ‘not to rock the boat’ so as to not disturb the fragile election process (Van Woudenberg, 2006).

Besides the fact that EU policebuilding missions outside the EU differed significantly from the ‘Balkan template’, the reiteration of demands for the EU to be able to launch robust police missions globally commensurated on the institutional level with the establishment of the European Gendarmerie Force (EUROGENDFOR or EGF) in 2004. While the EGF, because of being based on a policing model not unanimously shared by all EU member states, was created outside of the framework of the EU, its founding document clearly establishes the EGF as a direct contribution ‘to the development of the European Security and Defence Policy’ as the EGF is able to ‘provide Europe with a full capability in order to conduct all police missions in crisis management operations within the framework of the Petersberg Declaration, with particular regard to substitution missions’ (EUROGENDFOR, 2004, 2007). The actual tasks to be performed by the EGF are specified later in the document as comprising ‘performing security and public order missions’, ‘training and advising police officers as well as police instructors’, ‘performing criminal investigation work’, ‘conducting public surveillance’ and ‘protecting people and property and keeping order in the event of public disturbances’ (EUROGENDFOR, 2004).

The establishment of the EGF was positively acknowledged in a declaration of the EU Chiefs of Police. The declaration, drawing on earlier calls for the creation of ‘robust’ police forces, welcomed the idea of creating an EGF as a capacity for police missions in so-called non-benign environments (EU Chiefs of Police, 2004: 3). In a speech in early 2007, Solana stressed the need to further develop a wide range of tools at the disposal of the EU to perform a greater operational role in Europe and beyond. Besides the formation of EU Battle groups it is the EGF that gets explicitly mentioned during the speech. The EGF is envisioned hereby as a tool to enable the EU to quickly ‘deploy police forces with military status and full police powers’ (Solana, 2007). In line with this, the ESDP ‘Civilian headline goal 2010’ lists as one of its future objectives the ‘further development (…) of rapidly deployable police elements’, which may be deployed ‘together with military components and temporarily under military responsibility’ (Council of the European Union, 2007: 3). Explicitly drawing on lessons learned from previous ESDP missions, the ministerial declaration published ten years after its formal launch in 1999 called for the importance of increased civil–military coordination and civil–military synergies in CSDP missions, which are thought to enable the EU to undertake more flexible, complex and robust crisis management missions (Council of the European Union, 2009b). In a similar manner, a resolution by the European Parliament depicted the combination of military and civilian capabilities for crisis management as ‘added value’ to the EU’s role in international security. Consequently, the resolution ‘welcomes the development of Integrated Police Units, i.e. robust, rapidly deployable, flexible, and interoperable forces able to perform executive law-enforcement tasks, which, in certain circumstances can also be deployed as part of a military operation’ (European Parliament, 2010). Hence, despite being based on a policing model not shared by all EU-member states and located outside the EU framework, the EGF, as a ‘third-type’ of force situated between military and civilian police, was widely regarded as an asset to the EU’s CSDP toolbox.

The Janus-face of policebuilding became ever more apparent during the discussion over the EU’s police mission in Afghanistan. Largely based on the ‘Balkan template’, EUPOL Afghanistan was frequently criticized for being inadequately designed, under-funded and under-staffed (Feilke, 2010; Gross, 2009; ICG, 2007, 2008; Kempin and Reinicke, 2009). Critics especially stressed the fact that EUPOL’s objective of establishing a civilian, democratic police force was ill-suited in the context of a consolidating insurgency in Afghanistan. Subsequently, police assistance became increasingly linked to the overall NATO-led counter insurgency efforts in Afghanistan. This has been confirmed by statements of police trainers deployed to Afghanistan, who described the training of Afghanistan policemen as largely geared towards counter insurgency (Karioth, 2009). Or as one observer put it: ‘Carrying out state-building operations in a war-like situation poses a particularly serious challenge’ with EUPOL appearing as ‘too sophisticated’ for the reality on the ground in Afghanistan, thus calling for stronger synergies between the militarized ‘boots on the ground approach’ favoured by the US and the civilian approach to police assistance endorsed by EUPOL Afghanistan ‘despite the fact that discrepancies between the two models remain’ (Peral, 2009: 332). And it was exactly a paramilitary police force in Afghanistan that was, in the eyes of NATO, needed to support the transatlantic organization and the Afghan National Army (ANA) in their counter insurgency efforts against the Taliban. Consequently, the EGF was deployed to Afghanistan from 2007 onwards to cooperate with NATO in the creation of an elite gendarmerie unit. The Afghanistan Civil Order Police (ANCOP) was designed to serve as a counter-insurgency force working closely together with the ANA and ISAF (International Security Assistance Force) troops (NATO, 2011). According to a report, ANCOP had, by 2011, ‘firmly established itself as an elite rapid reaction and counterinsurgency force’ (Perito, 2012: 1).

Despite claims over rebuilding a democratic, civilian police force, various studies have illustrated that police assistance has increasingly become dominated by counter-terrorism and/or counter-insurgency agendas with little attention on civilian, democratic policing (Friesendorf, 2011; Heiduk, 2011; Murray, 2007; Sedra, 2006). The militarization of the police in order to achieve short-term gains in the field of counter-insurgency has in turn had a negative impact on the image of the police in the eyes of the population. Human rights organizations have documented a series of alleged violations of human rights and humanitarian law on the part of the police, which had significantly tainted the image and diminished the public’s confidence in the police (Davis, 2009; Martin, 2009; Oxfam, 2011; Vinck et al., 2008). Thus the militarization of ‘policebuilding’ has served as a case in point for those arguing that the idea of a militarization of the police is essentially counter-productive in terms of reaching the stated long-term goals, namely increasing the public’s trust in the state and its institutions to boost the legitimacy of the state and thereby enhance its stability. It has been argued that, in reality, police assistance inherently involves choices over, or at least a certain hierarchy of, these often contending objectives of ‘policebuilding’; that is: if robust security matters should trump those of democratic police reform, or vice versa (Ryan, 2009: 313)? In line with this, a study commissioned by the European Parliament in late 2004 suggested that ‘it would be important to clarify the question of how far police support will be taken out of the traditional context of development cooperation and be turned into a branch of security policy’ (Rummel, 2005: 10).

Rethinking policebuilding

What is by and large understood as a policy dilemma, presenting policy-makers with the challenge of having to choose first of all between two diverging concepts of police assistance (civilian or military), which in turn entails different policy imperatives, requires a more general examination. Essentially, much of the literature on ‘policebuilding’, by contrasting civilian, democratic policebuilding with the militarization of police assistance, builds on a conceptualization of police which understands police simply as a civilian institution designed to enhance public safety and law and order. From this, the police and its activities are viewed as being fundamentally different from the military and its activities.

Yet our analysis of the evolution of CSDP police missions has clearly illustrated that both police functions, civilian and military, must be understood as complementary aspects of international policing. From as early as 2002 onwards, even before the launch of the first CSDP police mission, police assistance was conceived as simultaneously incorporating ‘civilian’ and ‘military’ aspects. On the one hand, CSDP police missions were seen as externalizing a civilian, democratic model of police as part of an overall liberalization process. Hereby police assistance was framed in a language emphasizing support for third states in attaining ‘European standards’ and ‘European best practices’ of policing. At the same time we found a very different conceptualization of police, which revolved around an understanding of police assistance as a form of stability export to so-called weak or fragile states threatening the security of Europe and its citizens. The latter rested on a robust, militarized model of the police; it is framed in a language that stresses the need to stabilize ‘failed states’ and provide order in ‘non-benign’ contexts. Again, both concepts were by and large conceived as two sides of the same coin. This is in line with theoretical insights that have helped us to conceive the police not as an a priori civilian or military actor, but as a tool to establish and maintain a particular (liberal) social order.

How can such an alternative theoretical conceptualization enhance our understanding of the apparent Janus-face of ‘policebuilding’? Firstly, such an alternative, theoretically informed conceptualization of ‘policebuilding’ situates the phenomenon in the context of liberal interventionism. The emergence of ‘policebuilding’ in the last two decades has coincided with a liberal interventionism and its attempts to deeply modify (or at times re-build) social order in so-called zones of disorder. This implies conceiving of ‘policebuilding’ not as a neutral, technical activity but as a political instrument used to replace perceived ‘illiberal’ orders with liberal ones. Because the police function as an important interface between the state and the people, we must understand ‘policebuilding’ as a crucial tool to impose disciplinary norms geared towards promoting certain liberal forms of behaviour to elevate potentially dangerous, disorderly subjects into civilized, productive, harmonious citizens. Secondly, understood from such a theoretical angle, the main concern of ‘policebuilding’ is not the export of Western police models and best practices but the fabrication of an order subordinated to the imperatives of liberal statehood. Theorizing policebuilding as a tool to transform potentially dangerous zones of disorder into non-dangerous (that is: liberal) societies through the use of ‘reasonable force’ (including, at times, coercive, illiberal means) furthermore allows us, thirdly, to challenge the commonplace account of policebuilding as an essentially ‘civilian’ tool in the liberal interventionist’s toolbox. Fourth, we therefore need to understand ‘militarized’ approaches to, and practices of, policebuilding theoretically as part of a continuum of policing approaches and practices that include civilian, democratic, as well as forceful, coercive means. This implies that policebuilding’s Janus-face is an integral feature of liberal interventionism rather than a mere policy dilemma. The police, as well as policebuilding, are therefore not to be predetermined as ‘civilian’ or ‘military’ as their practices are rarely statically found on just one side of the continuum. They can shift over time in line with the overall dynamics of statebuilding efforts, thereby emphasizing more exclusive, coercive practices when the liberal statebuilding project is perceived to be under threat, whilst employing more inclusive, civilian practices in more ‘stable’ contexts. Hence, this then furthermore allows us to understand policebuilding theoretically as part of a wider range of interventions, ranging from interventions in the field of education to police and military reform, designed to simultaneously serve and protect the population and to establish and maintain liberal state power through, at times, coercive (though not necessarily always military) means. That said, policebuilding is indeed aimed at ‘the heart of governance’ (Council of the European Union, 2009a.