Constructing the capable state: Contested discourses and practices in EU capacity building

Capacity building has risen to prominence in the vocabulary of the international community as a way to promote security and development in fragile and post-conflict environments. Capacity building seeks to promote a bottom-up approach drawing on and strengthening existing local capacities. This article argues that capacity building can be understood as part of a broader governmentality that seeks to determine from the outside what constitutes a ‘capable’ subject. However, the effects of these governance practices are not straightforward as they are constantly shaped by the way local actors on the ground engage with these. Drawing on both policy documents and interviews conducted in Bosnia, Kosovo and Somalia, the article examines European Union capacity building initiatives in these post-conflict environments. By examining the rationality and problematisations behind this discourse, the article unveils how such assumptions (in particular, regarding the lack of institutions, power and knowledge) result in interactions and contestation between the local and the international in practice, which lead to new outcomes that neither straightforwardly reflect the existing status quo nor represent a linear imposition of power by external capacity builders.


Introduction
Capacity building has become a key form of interventionary practice in post-conflict zones and transition countries (European Commission, 2011;OECD, 2006). On the face of it, capacity building moves away from previous top-down engagements of the 'liberal peace' towards more bottom-up methods focused on existing local capacities in line with resilience building approaches (Chandler, 2015). The reality, however, often defies the rhetoric (Mac Ginty and Richmond, 2013). Particularly problematic is the fact that international capacity builders continue to promote Western-inspired models of the 'capable state' premised on questionable assumptions concerning a lack of institutions, power and knowledge at the local level. In consequence, capacity building programmes in peacebuilding have faced persistent and well-documented problems of effectiveness, sustainability, local ownership and legitimacy (Edmunds et al., 2018).
In this article we ask why it has been so difficult for international capacity builders to move beyond idealised notions of the capable state and examine what happens when such notions encounter the often-complex realities of governance in the local environments in which they take place. We draw on the experience of recent European Union (EU) capacity building initiatives to show such activities are deeply rooted in, and aspire to replicate, particular forms of state subjectivity. Capacity building discourses and truth configurations shape ideas about what a 'capable state' is, in the form of an institutionalised, liberal state entity, while foreclosing unacceptable forms of state capacity that may be more localised, informal or traditional in nature, constraining the imaginary of what is possible in capacity building. We employ the existing governmentality literature on peacebuilding and statebuilding (Chandler, 2010;Joseph, 2013;Merlingen and Ostraukaite, 2006) to argue that these processes are productive of new forms of governance in the EU periphery, and between that periphery and the EU itself. We also draw on insights from the literature on hybridity in peacebuilding and our own interview data to suggest that the interactions produced through capacity building do not have linear or straightforwardly hierarchical effects and do not represent an uncontested exercise of power by the EU over uncontentious subaltern actors. Instead, they comprise dynamic interactions between EU subjectivities and practices and those of local actors on the ground.
Empirically, the article focusses on EU capacity building programmes in the Western Balkans and the Horn of Africa. These two regions vary in terms of developmental levels and security threats, but also in relation to the nature of EU engagement, with the former benefiting from the prospect of EU integration. Yet, the evidence exposes similar assumptions and practices regarding EU capacity building activities in these two regions. The research is based on our analysis of secondary and primary sources, including official documents and qualitative interviews. Research fieldwork conducted in Bosnia, Kosovo and Somalia between May and November 2016 comprises 98 interviews with EU and international officials, civil society actors and local governmental representatives. For reasons of anonymity, all interviews are coded.
Our discussion proceeds as follows. First, we examine the ways in which capacity building has been understood to date and offer our own conceptualisation based around Foucauldian notions of governmentality. We then move on to analysing EU capacity building discourses, focusing on their articulation in key policy documentation. The empirical sections explore the rationalities and problematisations that underpin these initiatives in more depth, arguing that they are intimately bound up with highly specific and normative assumptions about what kind of state entities are 'capable'. More specifically, we explore the ways in which these assumptions are challenged through interaction with circumstances on the ground, focusing on contestations over institutions, power and knowledge.

Theorising capacity building
Capacity building is an increasingly important strategy for international actors in their efforts to promote security and development. It encompasses a broad range of activities that aim to strengthen the ability of war-torn or fragile states to manage their own security challenges and to achieve development objectives in the context of peacebuilding, statebuilding and security sector reform (SSR). It assumes that local actors are best placed to understand and address their own security challenges and that assisting (rather than replacing) them will produce outcomes that are effective, legitimate and sustainable (Clements et al., 2007: 46). It seeks to help partner states to become more resilient in the face of the various insecurities they confront by providing the material assistance, knowledge and technology to enable them to do so. Capacity building in this sense is understood as a technical, problem-solving endeavour, and an alternative to the more ambitious and costly liberal interventions of the 1990s and 2000s.
While the term capacity building has a long history in management and development studies (Smillie, 2001), it is a more recent newcomer in International Relations. Kühl (2009: 554-555) observes that in its initial iterations, the concept was closely associated with so-called institution-building approaches aimed at improving bureaucratic infrastructures in postcolonial states. Recent approaches are more bottom-up in nature and pay significant attention to human resource and skills development (Haldrup and Rosén, 2013). As Moore notes, capacity building 'includes everything that was covered by the different definitions of "institution building" and much more besides' (Eade, 2005: 1). Notions of capacity of one sort or another are also implicitly present in much of the SSR literature, which addresses democratic change and transformation of security governance. The concept of security sector capacity is rarely explicitly theorised (Denney and Valters, 2015), 1 however, and has often come to mean little more than a technical exercise in organisational restructuring. The limitations of such narrow approaches are captured in a report by the UK Department for International Development (DFID): [They] understand […] the problem being faced as a deficit of technical skills and resources that, once filled, will result in an improved security system. What this misses are the deeper layers of capacity, about how the various components of a complex system work together and relationships between the security system and the communities they serve. It also neglects the fact that dysfunction is often the result not only or primarily of weak capacity, but also of a particular constellation of political incentives, often underpinned by the nature of the prevailing political settlement. (Denney and Valters, 2015: 11) In other words, much of the work which touches on capacity building not only fails to conceptualise this term, but also to theorise the power relations which sustain these practices, both at the domestic and international level.
Building on these insights, we argue that capacity building represents a form of governance insofar as it aspires to constitute particular kinds of subjects through dominant discourses and imaginations about what a 'capable state' is or should be. In this regard, Paul Jackson (2011: 1817 has already argued that contemporary forms of statebuilding seek 'to construct states that are "capable" in a liberal sense, i.e. providing good governance, democracy and security. ' He adds (2011' He adds ( : 1818, that [a]t the heart of this shift is a redefinition of state sovereignty from being an international absolute to a variable one based on state capacity or a state being sovereign only in so far as it is capable of carrying out certain functions. More recently, Jackson and Bakrania (2018) have argued that SSR interventions to date have promoted technocratic 'linear models of statebuilding' based on Western liberal templates and prioritising formal institutions over informal or traditional ones. This article draws on these findings by exploring and theorising what the assumptions underlying such an understanding of the 'capable state' may be and how this process unfolds in practice.
A Foucauldian approach helps us to shed light on how power circulates and how knowledge travels from Western to non-Western contexts through capacity building programmes. The starting point here is Michel Foucault's notion of governmentality (Foucault, 2007). Governmentality is defined as the 'conduct of conducts', in other ways, it is about regulating the behaviour (or conduct) of subjects. Rather than working through direct imposition, it relies on individual freedom and self-governance; governmentality is about 'disposing things' and 'employing tactics', drawing on technologies of power such as partnerships (Sending and Neumann, 2006: 656). Governmentality works in indirect ways by seeking to govern through the consent of those to be governed, through self-regulation and responsibilisation of the subject. It can thus be understood as a form of (liberal) governance which seeks to shape the conduct of others from a distance (Joseph, 2013). In the development area, Phillips and Ilcan (2004: 393) have argued that 'capacity-building efforts often work to govern groups and populations by making them become self-regulating and responsible through new market-based relations'. In a similar way, EU capacity building efforts in the security sector can be seen as an apparatus of rule that works to govern states and populations by making them become responsible actors in line with the 'model' of the liberal (capable) state. It does so by seeking to both shape their capacity to govern themselves and their subjectivity. While Foucault's concept of governmentality had the population as its target, it is also possible to see how this concept can be applied to governance of the state and how global institutions can be agents of governmentality (Joseph, 2009). According to Fougner (2008: 308), states can be understood as subjects of governmentality 'in the sense that they are constituted and acted upon as subjects with a rationality derived from arranged forms of entrepreneurial and competitive behaviour'. Joseph (2009: 427) explains that 'the regulation of states takes place through the targeting of populations'. Hence, capacity building can work at different levels, from the individual to the communities and the state. This is also in line with how the EU envisages capacity building, as will be discussed in the following section.
Capacity building encourages self-government because it acts as 'an affective forcea visceral force that works beneath, alongside, or generally other than conscious knowing-that can serve to drive us towards movement' (Seigworth and Gregg cited in Gabay and Ilcan, 2017: 479-480). It not only has an ability to travel, it can also instil optimism about the future and build confidence in one's newly acquired capacities (Gabay and Ilcan, 2017). Because of these affective qualities, capacity building can act as a productive force shaping subjects and communities. Yet, by focusing on internal capacity gaps and apparently dysfunctional organisations, capacity building approaches shift the attention away from external actors and serve to depoliticise such interventionary activities (Chandler, 2015: 48). It also functions as a way for international actors to responsibilise local actors while denying them the exercise of power (Joseph, 2013). This trend has been reinforced by the increasing association of capacity building with discourses of resilience: that is, the idea of (local) adaptative capacities as a way to deal with increasing complexity and uncertainty (Haldrup and Rosén, 2013).
Our analysis draws on Foucault's notion of political rationality. As argued by Lemke (2002: 55), 'a political rationality is not pure, neutral knowledge that simply "represents" the governed reality. It is not an exterior instance, but an element of government itself which helps to create a discursive field in which exercising power is "rational".' In so doing, rationalities help construct what and who needs to be governed and how. Just as medicine does not only concern itself with curing 'the sick man', but also generates knowledge about the 'healthy man' (Foucault in De Larrinaga and Doucet, 2015: 59), capacity building is not limited to providing recipes for improving the state security apparatus; it also propagates a discourse of the 'model state' as the liberal (capable) state. As we show in this article, the rationality of capacity building is premised on a specific, and deeply rooted, vision of what a 'capable' state is -understood in functionalist, centralised and liberal terms -but also what it is not. Moreover, capacity building takes as a starting point specific problematisations, in this case, the weak or absent capacities in targeted countries (hence the metaphor of 'building'). Here, problematisations are understood as the process whereby some specific issues are rendered as problematic and in need of rectification − and intervention (İşleyen, 2017;Merlingen, 2011). Such problematisations legitimise EU (and other donor) interventions aimed at shaping the governance of states and populations on its periphery from a distance (Merlingen and Ostraukaite, 2006). The article identifies three embedded problematisations of the local that are common to EU capacity building activities: a lack of institutions, a relative lack of power, and a lack of knowledge (see below).
A focus on governmentality should not overestimate its actual effects in practice, however. This is for two reasons. First, as Joseph (2009) argues, there are limits to the effective operation of governmentality beyond the advanced liberal societies where it first emerged. In cases of 'failed governmentality', particular rationalities are imposed because international actors are 'so bound up with the dominant neoliberal rationality that they are unable to see the world outside of this discursive framework' (Joseph, 2009: 421). The same logic and failures of governmentality can be observed in relation to the EU's 'capable state' rationality.
Second, even where it works, issues of contingency and the agency of those encountering and/or resisting capacity building need to be considered. Governmentality work on global governance (e.g. Fougner, 2008;Merlingen, 2011;Sending and Neumann, 2006) has tended to play down contingency and unintended consequences, hence foreclosing opportunities for agency of those subjected to governmentality. Yet, as stated in one of the most well-known quotes of Foucault, 'where there is power, there is resistance ' (1978: 95-96). Moreover, the peacebuilding scholarship has demonstrated how ambiguity undermines the implementation of the liberal peace (Autesserre, 2014;Pouligny, 2006). As put by Zanotti (2010: 19), 'practical implementations of the script of international governance are ridden with ambiguities, indecision and continuous contingent and context-specific negotiations of divergent sets of principles and practical necessities'. This article adds to this literature by showing how ambiguities and contingencies open spaces for agency at the local level. While capacity building as governance is a powerful process -one that has effects -those effects are often not linear or as intended. Instead they are productive of new forms of subjectivity that sometimes differ from the original aspirations of the capacity builders. The article applies these insights to examine interactions and contestation between the local and the international, leading to new outcomes that neither straightforwardly reflect the existing status quo nor represent a linear imposition of power by capacity builders. In what follows, we examine these issues in further depth, looking first at the specific discourses underpinning EU capacity building and then the ways in which they construct, and are imbued with, notions of the capable and incapable state.

The rationality of EU capacity building
Building the capacities of partners as a way to enhance their resilience has been identified as a key EU strategic objective. The EU's Global Strategy states that, through its external action, the Union will ensure 'that [its] security sector reform efforts enable and enhance our partners' capacities to deliver security within the rule of law' (High Representative, 2016: 26). Several Common Security and Defence Policy (CSDP) operations focus on or include a capacity building component (e.g. EUCAP Somalia, EUCAP Sahel Niger, EUTM Mali). Moreover, recent initiatives such as the Capacity Building for Security and Development (Commission and High Representative, 2015) have a capacity building focus.
Similar to other organisations (OECD, 2006;UNDP, 2009), the EU defines capacity as 'the ability of people, organisations and society as a whole to manage their affairs successfully' and adds that it is 'an attribute of people, individual organisations and groups of organisations' (European Commission, 2011: 9). Capacity building thus needs to be based on internal motivation of the recipient actor and local ownership and cannot therefore be imposed by external development partners. This is also in line with the EU's Global Strategy which states that interventions should now be focused at the level of an actor's capacities rather than on the external environment (High Representative, 2016). The role of the EU appears in this way to be that of a facilitator and partner moving away from the liberal peace discourses of external intervention. According to an official in Bosnia, the job of the EU in the country should be 'to enable people to stand on their own two feet', 'enabling ownership'.2 EU capacity building is conceptualised as being holistic in nature, including a wide range of activities. This conception requires a comprehensive or integrated approach (High Representative, 2016). As stated in EU documents, '[f]or any country to ensure its security and development, it must have or acquire adequate capacities in all critical sectors, including security and defence. This will not only stabilise that country but also enable it to contribute constructively to peace, stability and crisis prevention in its region.' (Commission and High Representative, 2015: 3). As argued by an interviewee, one of the most important lessons of the past decade is that 'institutional capacity of local organizations is best enhanced through a holistic and systemic approach involving organizational reforms in governance and other management operational areas than a piece-meal and short-term quick-fix approach'. 3 In sum, capacity building has become a key priority in EU security discourses, to the point that practitioners often refer to it as a new 'buzzword'. The official EU rhetoric emphasises internal capacities as opposed to externally driven processes of reform and the need for local ownership as a key principle both at the design and implementation phases. The official definitions also point to the comprehensive and holistic nature of capacity building. But what is the vision of the 'capable state' that EU programmes seek to promote?

Imagining the capable state
EU capacity building is premised on a vision of the 'capable' state as a functionalist, centralised and liberal state. This vision or imagination can be understood as the political rationality of EU capacity building. As mentioned earlier, a rationality helps delimit the subjects and practices to be governed; it 'shapes what is and is not thinkable, reasonable, practicable and doable in relation to governance' (Merlingen, 2011: 152). Capacity building can thus be seen to constitute a new form of governance in that it seeks to 'normalise' states in the EU's periphery in line with this imaginary. This is also reminiscent of the idea of the EU as a normative power which, according to Manners (2002: 236), describes an actor's ability 'to define what passes for "normal" in world politics'. The EU seeks to set and spread a vision of what a 'normal' capable state should be.
This model state is based on an ideal-typical Weberian form of statehood which requires strong formal institutions and an exclusive monopoly over the use of force. From a technical standpoint, an approach that is premised on rationalist principles of organisational effectiveness may appear unproblematic and neutral. However, as we discuss below, it can often be in tension with established local practices and ways of doing things. Moreover, the centralised structure implied by the Weberian model is far from being uncontroversial. As argued by Schroeder and Chappuis (2014: 138), in many peacebuilding environments and 'in the field of security, control over the state monopoly on the use of force is often contested among different domestic groups' rather than monopolised by the state itself. In other cases, as explained below, institutional power is far more dispersed than the Weberian state model implies (Hoehne, 2013). Hence, in many of the post-conflict contexts where the EU has deployed its CSDP missions and capacity building programmes, this requirement to establish a centralised model has been impossible to achieve. In Bosnia, the failure of the police reform can be explained this way (Juncos, 2011).
Because of the problems involved in promoting this imaginary of the state (institutionalised, centralised, liberal), it is perhaps not surprising that the EU has sought to portray capacity building as a merely technical endeavour. As Jackson (2011Jackson ( : 1804 suggests, '[t]his approach has largely been carried out as a "technical-administrative" exercise with a focus on the technicalities of constructing and running organisations rather than on the politics of creating states'. Similarly, it assumes that capacity building can be measured and quantified by external actors (Jackson and Bakrania, 2018: 19), facilitating monitoring and governance from a distance. In the case of the EU, this technocratic approach finds its roots in the previous enlargement experience and the prominent role played by the Commission in the overall process. EU officials, particularly Commission officials, have sought to detach themselves from 'the political' as they believe that the politicisation of conditionality has a negative impact on its effectiveness.
Yet, the reforms promoted by the EU are not just technical in nature; they promote specific models of political and economic re-organisation which follow closely the Western liberal imaginary of the state. The EU identifies the following principles underpinning capacity building: 'promoting respect for international law, in particular humanitarian and human rights law, gender perspectives, UNSCR 1325, and principles of democracy and good governance is integral to these efforts' (Council of the EU, 2016: 13). Thus, capacity building is concerned with normative questions about how capacities should best be employed and governed against a wider framework of democratic governance, statebuilding and reform. As acknowledged by an EU official: 'this is not just about technical reforms; there is also a commitment to the values that the EU embodies'. 4 According to another official, 'it is not about technical or operational level work; it's about getting the strategy in place'. 5 Hence, what could be seen as a merely functional approach is deeply political insofar as the EU seeks to govern states and populations by making them become responsible actors in line with the model of the liberal (capable) state.
Certainly, at the local level, there is a widely held perception that capacity building is a mechanism for imposing a specific form of governance derived from Western models. As one interviewee noted, 'in general they [the EU] see the situation in Somaliland as chaotic and they want to fix it. The way they want to do it is to support establishing the Western liberal state here.' 6 For many of the interviewees in Bosnia, one of the main challenges was to change the 'mindsets' of those participating in capacity building programmes. 7 According to a local official, 'capacity building is basically all about assimilating the norms, standards and values of the EU […] it is about filling this framework with the essence, and that essence is democracy'. 8 In this regard, he continued, 'we're lagging behind as a country not only in terms of the military and the security sector, but also as a society as a whole'. 9 To a certain extent, EU capacity building shapes how a 'capable state' is perceived at the local level, but it also leads to contestation, as we discuss in the next section.

Encounters with the incapable state
There is seemingly widespread agreement on the problematisation that sustains the EU's rationality of the capable state: a lack of capacity on the part of the 'locals'. Similar accounts of capacity deficits support international (UN, EU) interventions in other cases (Autesserre, 2014;Rosén, 2011). Implicit in the rationality of EU capacity building is the understanding that local capacities are weak and even in some cases absent (Björkdahl and Höglund, 2013: 291). The linear approach adopted by EU capacity building assumes that progress takes place 'down a set path towards liberal peace, and that the only things standing in the way [are] a lack of resources, of capacity, and of international will, in tandem with the presence of predatory local elites' (Jackson and Bakrania, 2018: 17). This is particularly the case regarding Somalia, where such perspectives are illustrated in policy documents including the Joint Strategy Paper, which states that '[c]apacity in all forms is in short supply' and that 'the challenge will be to build and extend local capacity, ensuring ownership of institution-building' (European Commission,n.d.: 19). Another document states that significant challenges include 'the lack of capacity and collective political will of the Federal Government of Somalia' (European Scrutiny Committee, 2015). The perception of a capacity gap also extends to Somalis themselves. According to some interviewees: 'people lack skills', there is a 'lack of capacity, capital and experts', 'low human capacity' as well as a 'lack of expertise locally'. 10 Another stated that 'we have had dysfunctional institutions for a long time, and [capacity building projects] help us to tackle this'. 11 Similar sentiments are present in EU problematisations of the Western Balkans, though here such discourses tend to comprise 'weak' capacities rather than their absence. 12 However, in Bosnia, 'this means starting from scratch because you have the police who were militarised during the war […] And you need some new arrangements in accordance with the Dayton Peace Agreements because now you have to establish common institutions at the state level.' 13 This problematisation of capacity deficits at the local level plays a key role in legitimising various forms of international (and EU) capacity building. It presents the host environment as either an empty canvass waiting to be filled − what Lemay-Hebert (2011) terms the 'empty shell' approach − or a dysfunctional space in need of correction. Such problematisations fit neatly within the so-called 'failed' or 'weak' state paradigm, but often do little to capture the reality of governance on the ground. Indeed, the failed state paradigm itself has long been criticised by scholars on the basis that it privileges a geographically and historically specific model of idealised statehood -characterised by formal institutions and a monopoly on the legitimate use of force -while failing to recognise alternative, often locally legitimate and subjectively effective, forms of governance that may exist elsewhere (Call, 2008). Similarly, the burgeoning literature on hybridity in peacebuilding warns against considering such processes as a straightforward imposition of hierarchical control by powerful internationals over subaltern locals. As Roger Mac Ginty (2010) and others (Hoehne, 2013;Richmond, 2009;Schroeder and Chappuis, 2014) observe, peacebuilding in practice entails multiple dynamics of change and interaction between 'local' and 'international actors', in which each can shape and influence the other in important ways.
The following sections interrogate these relationships with regard to three specific problematisations of the local: an assumption of weak or failing institutions in target states; an assumption that international capacity builders operate in an asymmetrical relationship of power with local actors due to their superior resources and the strength of their own capacities; and an assumption that international expertise offers privileged knowledge on how such weakness can be overcome. In so doing, we show the tensions such problematisations create when capacity building programmes encounter local environments. However, we also show that such encounters are productive and lead to new outcomes that neither straightforwardly reflect the existing status quo nor represent a linear imposition of power by external capacity builders.

Institutions
EU capacity building activities focus overwhelmingly on the formal institutions of the state (Richmond, 2009). Here we use institutions to mean 'agreed rules about ways of doing things' (Leftwich, 2007: 9). They thus concern the mechanisms through which governance is mediated and can be formal or informal in nature (Edmunds, 2009). Helmke and Levitsky (2004: 727) characterise formal institutions as 'rules and procedures that are created, communicated and sustained through channels widely accepted as official'. They can include the bureaucratic and administrative apparatus of states, organisations such as the police, armed forces or courts, as well as the rules and regulations governing how processes such as career progression take place within them. In contrast, informal institutions are 'socially shared rules, usually unwritten, that are created, communicated and enforced outside of officially sanctioned channels' (Helmke and Levitsky, 2004: 727). They can include governance structures based around clan or ethnic ties, kinship and clientelist networks, customary law or non-state security providers such as neighbourhood watch groups or community militias.
Informal institutions typically exist alongside formal ones and sometimes even supplant them (O'Donnell, 1996: 39). In environments in which the formal state is weak or absent, it is often informal institutions that fill the governance gap. Clan structures, for example, may take the place of state bodies in mediating disputes within communities or allocating resources (Hoehne, 2013: 205-207). Customary law may take the place of statutory law in places where courts do not function or are difficult for people to access (Moe and Simojoki, 2013: 396). Petty corruption or bribery may function as an informal means of remuneration when police or minor officials are rarely, if ever, paid (Debos, 2011: 419). In other cases, formal and informal institutions may work together and even be dependent on each other. Underfunded or weak policing structures, for instance, may rely on neighbourhood watch or community groups to manage petty crime and alert them to more serious infractions when these occur. As argued by an interviewee from Kosovo, 'informalities remain much more important than formalities'. 14 While they may be illiberal, inefficient or iniquitous in practice, they can be subjectively effective in the governance of spaces in which the formal state does not or cannot reach (Hoehne, 2013: 199). In this respect, EU capacity building programmes rarely take place in environments where institutional capacity is completely absent. Instead, they function alongside and by implication aspire to supplant existing informal institutions with those of the 'capable state'.
We saw numerous examples of the important influence of such interactions in our interview data. In Somaliland, the clan system was seen to play an important role in determining appointments and promotions in organisations such as the police and coastguard. International interlocutors complained that key officials were chosen because of their influence as clan mediators rather than their expertise in security or their organisational skills. 15 Capacity building reforms aimed at introducing merit and expertise-based promotion and appointment systems, in this case in the coastguard, represented a direct challenge to the informal institutions that had determined these matters to date. They were strongly resisted by local elites and were even rumoured to have led to the transfer of then Head of Mission of EUCAP Nestor 16 to another post elsewhere. 17 Informal mechanisms for remuneration were also common in our cases, where wages can be low and irregularly paid. It was noted by several interviewees that local engagement in EU capacity building initiatives was facilitated and motivated primarily by the payment of per diems to participants. 18 A Kosovar interviewee noted that training activities often 'take place in an Albanian seaside town, so government personnel get their per diems and can go to the beach after work'. 19 Others argued that petty corruption in the police and elsewhere could not be easily eliminated by new regulation or training programmes when their wages were so low. 20 The importance of customary law, either alongside or often over and above, formal law, was also recognised by our interviewees. One international official noted the difficulties of training police in Somaliland in a context where customary law and practices dominate, noting 'we are basically working outside the law. The only law [which is exists is] the 1972 Police Law which is never used.' 21 Another noted that 'they [the Somalis] are very good in applying Sharia law but not other written law'. 22 Yet, from the Somali perspective, the interaction of formal and informal laws was key to relative peace and security of Somaliland compared with other parts of Somalia. One interlocutor suggested the main 'best practice' they would recommend would be 'to make customary law work with formal law'. 23 Another that 'the traditional system is strong; that's why there is stability'. 24 As Gryzmala-Busse (2010: 317-326) observes, informal institutions can replace, undermine or reinforce their formal counterparts and, in the context of our analysis, can play similar roles in relation to international capacity building programmes. Our interview data suggests that all three dynamics were present in our case study countries. Informal systems sometimes replaced their formal equivalents, as was the case with the Somaliland police law. At other times, as with the clan system, they could undermine it and lead to resistance. But these effects were not all negative either. Informal systems were understood to allow organisations and processes to function in the absence of formal counterparts and, as was the case with per diems, encouraged locals to engage with international actors when they might otherwise be reluctant.

Power relations
The discourse of absence associated with the capable state rationality confers an assumption of powerlessness on the part of local actors vis-à-vis their international donors. As an EULEX official noted, 'we deliver monitoring, mentoring and advising, not projects or programmes directly. It's a bit like teaching, only they don't get to evaluate the teacher, at all, ever.' 25 This quote symbolises the asymmetrical power relations between EU and local actors. However, local actors often have significant resources of power to draw on in their dealings with the EU, even in the most apparently dysfunctional of states. At a minimum, they are likely to have a key gatekeeping role, facilitating access to institutions, geographical areas and significant individuals in the recipient state concerned, as well as providing language skills and knowledge of who matters, where and why (Kappler, 2015: 880-883). One Somali interlocutor engaged with an EU programme noted that: 'nobody goes outside of the gates but me. I am the master and the key […]. If I sleep one day, all operation stops.' 26 Such gatekeeping roles are important in terms of directing or withholding access to particularly stakeholders or groups implicated in the capacity building endeavour. They can lead to favoured groups being incorporated and others excluded, with implications both for access to the resources that capacity building initiatives often bring and their ultimate effectiveness. Perhaps more prosaically, they can play a key mediating role in terms of basic communication between local elites and internationals. Thus, one international capacity builder complained that he struggled in his dealings with a key minister because 'his English was extremely poor. He was advised by a younger guy about another agenda, he was told in his own language things that are not true and which were extremely difficult for me to prove wrong.' 27 The importance of local gatekeepers is indicated by the problems that are often caused for internationals when one supportive local partner is replaced by another who is either not interested or actively opposed to their activities. One Somali interlocutor noted, in reference to a UN-funded programme that, 'the project went very well at the time. Then the minister changed, and the new minister didn't know anything about decentralisation. As a result of this lack of interest, the project stopped, and everything rolled back.' 28 Indeed, whatever their relative disparity in terms of resources or technical skills, it is the locals that play the key executive role in the implementation -or frustration -of policy on the ground (Schroeder and Chappuis, 2014: 139). This can take place at the highest levels of government -in the sense of approving legislation or strategy documents -or in organisational processes, such as appointments, promotions or remuneration. These powers can be exercised both in formal institutional contexts, over which internationals may have some degree of oversight, but also, often more importantly, through informal mechanisms such as patronage to which they are effectively blind (Mac Ginty and Richmond, 2013: 778). For instance, EU officials expressed frustration where the EU supported particular reforms and the locals either only implemented some superficial changes or ignored them altogether because they did not match their needs and/or interests. 29 In Bosnia, The EU insisted on police reform, but the reform that took place was not meaningful-it was just cosmetic. Therefore, we still have police systems that are significantly divided […] the reform of the judiciary system is also having serious trouble. They are by their very nature centralised, whereas this is fiercely objected to by lower levels of society in Bosnia. 30 In relation to the establishment of a Bosnian anti-corruption agency, one EU official explained: when the agency first opened, it took a long time to agree the legislation. Once it opened, it was given premises that weren't really suitable […] They are not given the tools to do the job. So, the resistance is continuous, even though for the last four years the EU has had permanent advisors inside the anti-corruption agency. We're pushing against the resistance all the time but it's frustrating and very noticeable. 31 These examples are a reflection of not only the relative powerlessness of international actors in the face of intransigence and delay by key local enablers, but also an indication of the important role of local interests in mediating or resisting apparently technical reforms that may have internally significant political or institutional consequences.
At lower level, the formalisation of human resource structures within organisations -often a key component of institutional capacity building -may create resistance amongst those with a stake in their informal alternatives (Brinkerhoff and Goldsmith, 2002: 33-35). One Somali interviewee noted that 'older guys felt threatened [by organisational reforms in the coastguard that threatened their positions] because there is no pension system in the country. When you stop working you stop getting paid. Naturally, they were protecting their jobs and salaries.' 32 Such expressions of power and interest on the part of local actors can lead to delay, obfuscation and obstruction as discussed above. However, they can also lead to a change in the nature of programme outcomes. One senior Somali official observed that 'we always leave room for negotiating and changing what we think is not a priority for the country'. 33 Another mentioned that 'because the resources we have are very limited […] we sometimes prefer to divert the donor contribution to, for example, staff salaries rather than staff skills development. This creates conflict with the internationals.' 34 These examples demonstrate both the dependence of international actors on their local partners in implementing their programmes, and the manner in which locals can use their power advantages on the ground to shape international activities and resources to their own advantage.

Knowledge
Finally, as mentioned above, EU capacity building is premised on the assumption of a chronic expertise gap amongst local actors. These assumptions may have merit in relation to the kinds of bureaucratic or technical expertise necessary to the functioning of the 'capable state' and the organisational practices therein, including specific skill sets such drafting legislation, administrative procedures, or equipment operation that may be absent in recipient countries (Neild, 2001: 27-29). However, this discourse of absence risks obscuring both key knowledge gaps on the part of the internationals and the interplay of different knowledge forms that occurs on the ground during the interventionary process itself. For example, one of the recurrent criticisms of EU capacity building programmes is that they have been implemented with only limited knowledge of the political and social dynamics of the local environments in which they take place. In the words of a local expert in Somaliland, 'one of the biggest weaknesses of EUCAP Nestor is that they don't understand the local context, the dynamics of the institutions that they work in'. 35 Another interviewee from Kosovo argued that 'there is a lack of deeper analysis and contextual knowledge when applying template approaches from other countries'. 36 Interviewees in Bosnia also complained about the high rotation turnaround of international staff and the lack of local knowledge, including knowledge of the local language, which inhibits the development of trust between the EU staff and local actors. 37 Sentiments such as these represent perhaps the most consistent theme across all our interviewees, particularly in Somaliland. One senior Somaliland official complained that 'there is always a fundamental difference between what donors are willing to provide and what we actually need […]. It is frustrating because they come with their plans [and] do not value the local knowledge. At best they come with diaspora people who do not have a better local understanding.' 38 Others noted that 'it is […] more frustrating when countries who are supposed to help you and focus on your priorities do not want to listen'; 39 that 'external actors behave like they know everything'; 40 that the EU 'disregards local knowledge and expertise'; 41 that '[their] mandate was divorced from the realities on the ground'; and that 'I'm not going to listen to external actors. They lack a basic local knowledge and act like experts.' 42 These tendencies are reinforced by the way the 'capable state' rationality prioritises certain knowledge forms above others (Autesserre, 2014). These programmes often draw on expert-driven notions of 'best practice' that derive their authority from practical experience elsewhere and are assumed to be transferable in a linear manner from one country or professional domain to another. As Bernstein and van der Ven (2017: 554) observe: 'The very act of labelling a set of principles "best practices" connotes widespread agreement, as if all existing practices have been evaluated against some objective criterial and a particular subset has emerged as superior.' Yet, as noted before, there is disagreement as to what the specific content of these 'best practices' is.
Such approaches can not only clash with existing local practices and interests, but may even prevent international staff noticing that they exist. As one of our interlocutors noted, 'Somalis described the behaviour of the EU as imperialistic. The EU treated Somalia as having no history.' 43 Another stated that 'their mandate […] was based on European knowledge that was supposed to be shared with Somaliland […]. They haven't sought advice from local experts at all. What they are doing is exporting their views.' 44 One Bosnian interviewee argued: 'here you have experts who know about a technical issue, but they have no background understanding of Bosnia and Herzegovina. They should have a longer time staying here overall.' 45 In sum, capacity builders' assumptions about a lack of local knowledge and expertise can lead to contestation at the local level and, conversely, a de facto reliance on local knowledge forms and interlocutors in the face of internationals' own knowledge gaps in these areas.

Conclusion
By drawing on the EU's intervention in the Western Balkans and the Horn of Africa, this article has sought to unveil the rationalities and problematisations that underpin capacity building activities. Drawing on a governmentality perspective, the analysis has shown that EU capacity building programmes rely on and aspire to produce particular forms of state subjectivity. Specifically, the rationality of EU capacity building seeks to construct a 'capable state' in the form of a functional, centralised and liberal state, while foreclosing unacceptable forms of state capacity. However, while EU rationalities might reinforce the discourse of a Weberian state, they do not necessarily produce that state itself. In that sense, this rationality is self-legitimising, but is not productive in the way it aspires to be, resulting in a form of 'failed governmentality'. In order to appreciate the (messy) effects of capacity building programmes, one also needs to understand the problematisations that underpin such interventions which are often expressed in the form of 'capacity deficits' at the local level.
The recipients of capacity building are problematised as 'incapable states', weak in institutions, lacking power and agency in the face of the challenges they face, and in need of external assistance and expertise. These problematisations encourage 'one size fits all' solutions that may be ill-suited to local circumstances and push out established practices and expertise in favour of generic international alternatives. They can also overlook existing practices which might be locally effective and legitimate on their own terms, isolate international capacity builders from the societies and practices in which their activities take place and increase their dependence on local actors through gatekeeping roles.
These rationalities and problematisations are integral to the very foundations of the EU's capacity building project. In consequence, they are deeply resistant to change and not easily amenable to technical fixes or incremental improvements. Nevertheless, and as our evidence has shown, this does not mean that capacity building in practice is a sterile or necessarily unproductive activity. Instead, it is dynamic and interactive, with substantive effects. The problematisations of EU capacity building and their often at least partial dissonance from the realities of governance on the ground themselves create the 'ambiguities and contingencies' (Zanotti, 2010: 19) that allow for considerable local agency in the substantive implementation of such programmes. In these spaces, local actors can work to resist, facilitate, ignore or adapt the capacity building projects in which they are engaged. Capacity building activities can and sometimes do fail for these reasons. However, they can often be productive of new outcomes. In these ways, the push and pull at play in the capacity building relationship is more equal, contested and dynamic than the image of international tutors imparting resources and knowledge to unschooled local pupils tends to imply. Instead, capacity building remakes itself -and the outcomes it produces -through the very process of its implementation on the ground and the dynamic interaction between local and international actors this necessarily engenders.