Questioning Anglocentrism in plural policing studies: Private security regulation in Belgium and the United Kingdom

Recent Continental European scholarship has identified a problematic Anglocentric bias running through the field of comparative plural policing studies. It has sought to counter this bias by emphasizing a series of divergent plural policing trajectories between the more market-friendly countries in the Anglosphere and the more state-centric countries in Continental Europe. While acknowledging the significance of this corrective, we argue that it tends to overemphasize the levels of divergence between these two regions. We substantiate this claim by examining the rise of the private security industry and its regulation by the state in the UK (representing the Anglosphere) and Belgium (representing Continental Europe). Interpreting historical and contemporary data through Sabatier and Weible’s advocacy coalition framework, which focuses on the cut and thrust of democratic politics, we observe how in both countries this important dimension of the plural policing landscape is characterized not by counterposed market-friendly and state-centric trajectories, but rather by a complex mix of state–market interactions. In other words, the dynamics of private security regulation are more state-centric in the UK and more market-friendly in Belgium than recent Continental European scholarship suggests. Moreover, we illustrate how, under conditions of post-financial crisis austerity, the overarching pattern is, if anything, one of convergence towards a common set of political dynamics. This is an important finding that not only makes an original contribution towards private security regulation scholarship but also encourages us to question the nature of Anglocentric bias within comparative plural policing studies.


Introduction
It is a truism that policing is no longer the exclusive domain of the police. In most countries, key functions such as patrol, protection, investigation and detention are increasingly carried out by a range of public, private and third sector actors working in collaboration and competition with one another. Over the past couple of decades, scholars have responded to this trend by exploring what is commonly termed the 'pluralization of policing'. It is notable, however, that most contributions to this scholarship draw from the empirical dynamics of single countries. There is a notable dearth of comparative analysis (De Maillard and Roché, 2018: 385;Jones and Newburn, 2006: 1). Moreover, the comparative analysis that does exist is often regarded as Anglocentric in orientation, giving rise to a geographically biased reading of the plural policing landscape (Devroe, 2017;Terpstra, 2017;Van Stokkom and Terpstra, 2018). In an effort to counter this bias, recent years have witnessed a new wave of comparative plural policing scholarship rooted in the experiences of Continental Europe, in particular Austria, Belgium, France and the Netherlands (De Maillard and Zagrodzki, 2017;Devroe, 2017;O'Neill and Fyfe, 2017;Terpstra, 2017;Van Steden, 2017). One of the headline findings in this scholarship is that, whereas the more market-friendly countries in the Anglosphere generally exhibit high levels of pluralization across the public/private divide, the more state-centric countries in Continental Europe usually display lower levels of pluralization across the public/private divide. This divergence is then used to redraw the Anglocentric map of the plural policing landscape.
While acknowledging the significance of this corrective, we contend that it often overemphasizes the levels of divergence between the Anglosphere and Continental Europe. We substantiate this claim by re-examining a key set of variables in this comparative scholarship -namely, the rise of the private security industry and its regulation by the state in the UK (representing the Anglosphere) and Belgium (representing Continental Europe). Interpreting historical and contemporary data through Sabatier and Weible's (2007) advocacy coalition framework, which focuses on the cut and thrust of democratic politics, we observe how in both countries this important dimension of the plural policing landscape is characterized not by counterposed market-friendly and statecentric trajectories, but rather by a complex mix of state-market interactions. Although notable divergences certainly exist, it is apparent that private security regulation is more state-centric in the UK and more market-friendly in Belgium than recent comparative scholarship suggests. Moreover, we illustrate how, under conditions of post-financial crisis austerity, the overarching pattern across these two countries is, if anything, one of convergence towards a common set of political dynamics. This is a significant countercorrective. It suggests that, rather than reproducing a market-friendly depiction of the Anglosphere and then challenging the analytical reach of this depiction by emphasizing incongruencies with a state-centric portrayal of Continental Europe, we should remain cognizant of the commonalities in plural policing arrangements across these two regions.
It further implies that we should continue to embrace an open orientation towards the transfer of analytical and practical plural policing knowledge between the Anglosphere and Continental Europe. In sum, then, we advance an empirically and theoretically novel comparison of private security regulation in the UK and Belgium that questions what we regard as an emergent misunderstanding in the extant comparative plural policing scholarship.
We develop this line of argumentation over five sections. The next section critically reviews this recent comparative scholarship. The subsequent section sketches out our theoretical and methodological approach. The following two sections use the advocacy coalition framework to compare private security regulation in the UK and Belgium. The final section reflects on both the divergences and the convergences in the dynamics of private security regulation across these two countries, before reflecting upon the nature of Anglocentrism in plural policing studies.

Comparative plural policing scholarship
As the pluralization of policing has gathered momentum across the globe, it has become ever more important to identify patterns of convergence and divergence so we can better understand not just the reasons behind these trajectories but also the extent to which we can and should draw lessons between them. Yet only a small proportion of the now quite extensive scholarship on plural policing is explicitly comparative in nature. As De Maillard and Roché (2018: 385) observe: 'it is rarely cross-nationally comparative . . . it rarely relies on a systematic analysis of national realities to identify similarities and differences.' It is further noted that the comparative scholarship that does exist is predominantly Anglocentric in orientation (Devroe, 2017;Terpstra, 2017;Van Stokkom and Terpstra, 2018). This is because the pioneering studies in this field have focused on Canada, the US and the UK (Jones and Newburn, 2002;Shearing and Stenning, 1981), generating a longstanding bias towards this region of the globe. Unsurprisingly, this bias has come to be regarded as an issue. According to Van Stokkom and Terpstra (2018: 418), for instance, the main 'problem' in this subfield is the 'dominance of Anglocentrism'. They continue: [N]ot only do many researchers in the U.S. and England (presumably implicitly) assume that their arrangements and problems are universal, this dominance is so strong that even researchers from other countries may be inclined to accept theories and interpretations from the Englishspeaking world as if they also apply to their own situation, even if this may be unclear or questionable.
For them, this Anglocentric bias has (unintentionally) distorted how we make sense of the plural policing landscape.
In recent years, a new wave of Continental European scholarship has made significant progress in exposing the contours of this bias through an exploration of plural policing trajectories in Austria, Belgium, France and the Netherlands (De Maillard and Zagrodzki, 2017;Devroe, 2017;O'Neill and Fyfe, 2017;Terpstra, 2017;Van Steden, 2017). For present purposes, the most salient dimension of this scholarship is the notable divergences it reveals between these countries and the Anglosphere. These divergences are exposed in a number of ways. One approach is to question the degree to which models of plural policing formulated within the Anglosphere can be applied to Continental Europe. Taking issue with the nodal governance model, for instance, Terpstra (2017: 79) observes how, 'in the continental European context, the state is not only more than "just one node among many" . . . given the historical context of these countries it is seen as very risky to distribute policing and policing powers across a wide range of actors'. Another approach is to draw broad empirical comparisons between countries from each region. In this vein, Devroe (2017: 97) remarks how 'the Belgian case study shows a deviation from the central Anglophilic statement stipulating that industrialised countries are becoming increasingly "pluralised-privatised". Fierce opposition in Belgium has obstructed the trend towards pluralisation that is present in the UK.' A final approach is to draw specific empirical comparisons relating to one particular dimension of the plural policing landscape. Button and Stiernstedt (2018), for example, rank 26 European Union members in terms of how comprehensively each state regulates its domestic private security industry, finding that Belgium occupies first place, leading a pack of Continental European countries, with the UK down in 20th position (Button and Stiernstedt, 2018: 404; see also Scheerlinck et al., 2020). They note how these findings corroborate Button's earlier analyses, which, among other things, 'highlight the disparities between the generally highly regulated mainland European countries and the minimal Anglo-Saxon models in England and Wales, North America and Australia' (Button, 2007: 124).
As such, one of the key messages running through this scholarship is that, whereas the more market-friendly countries in the Anglosphere are generally characterized by high levels of pluralization across the public/private divide (including relatively light-touch models of private security regulation), the more statist-centric countries in Continental Europe are usually defined by lower levels of pluralization across the public/private divide (including more comprehensive models of private security regulation). This means, so the logic continues, we should be wary of transferring analytical and practical plural policing knowledge between the Anglosphere and Continental Europe. This message is persuasive and represents a significant advance in the field of comparative plural policing studies. At the same time, though, we believe this scholarship tends to overemphasize the levels of divergence between plural policing trajectories in the Anglosphere and Continental Europe. It proceeds by reproducing a market-friendly depiction of the Anglosphere and then challenges the analytical reach of this depiction by emphasizing incongruencies with a state-centric portrayal of Continental Europe. In our view, this is an oversimplification. The Anglosphere is often more state-centric and Continental Europe is often more market-friendly than this line of reasoning suggests. There is, in other words, a common ground across these two regions that has been left out of the picture. Our intention here is to bring this common ground into frame, in the process questioning the nature of Anglocentrism in plural policing studies.
To accomplish this task, we focus on a specific set of variables: the rise of the private security industry and its regulation by the state in the UK (representing the Anglosphere) and Belgium (representing Continental Europe). We have selected these variables precisely because they play a central role in the recent comparative scholarship under examination. As the above literature review demonstrates, they are cited as key examples when emphasizing divergences between the market-friendly trajectories in the Anglosphere and state-centric trajectories in Continental Europe. Using these same variables to illustrate not just known divergences in state-market trajectories but also important convergences serves, we think, as an illuminating mode of critique. It meets this recent comparative scholarship largely on its own terms, rather than reshuffling the deck, so to speak. Of course, this also raises the question of how we can examine the same key variables and yet arrive at a different interpretation. The answer lies in our choice of theoretical lens.

Theory and method
The comparative scholarship that ranks European Union members according to how comprehensively each state regulates its domestic private security industry (Button, 2007;Button and Stiernstedt, 2018;Scheerlinck et al., 2020) is not presented through an explicit theoretical lens. But this does not mean it is atheoretical. It is informed -consciously or not -by what Rhodes (1997) terms the 'formal-legal' approach to state regulation. This approach sees 'legal rules and procedures as the basic independent variable, and the functioning and fate of democracies as the dependent variable' (Rhodes, 1997: 67). Understood thus, less state regulation equates to a freer market and higher levels of pluralization across the public/private divide, whereas more state regulation equates to a more controlled market and lower levels of pluralization across the public/private divide -hence the divergence between the Anglosphere and Continental Europe in the above rankings. Although it is of course important to explore state regulation in this way -and we certainly do not deny the existence of the legal divergences it identifies -this approach does not tell the whole story. It lacks an appreciation of the manifold ways in which state regulation is interpreted and negotiated by stakeholders on the ground. This is a notable blind spot because, once this ground-level agency is taken into account, the extent to which state regulation takes on a market-friendly or state-centric character begins to change. In what follows, we put this agency front and centre. It is this emphasis on the cut and thrust of democratic politics -as opposed to the codification of legal rules and procedures -that allows us to arrive at a different interpretation of the same key variables. To bring coherence to this theoretical approach, we draw upon the advocacy coalition framework (ACF).
Over the past three decades, the ACF has emerged as an important alternative to the formal-legal approach. For present purposes, we use a simplified version of Sabatier and Weible's (2007) iteration, which runs as follows. Whenever public, private and/or third sector actors enter into a policy arena -such as private security regulation -they are first guided by their beliefs. These beliefs find articulation at different levels: at a 'deep' level they give expression to a fundamental conviction on any particular issue; at a 'secondary' level they reflect a strategic reading of those policies that are most likely to serve this conviction. To advance their beliefs, these actors team up with similarly minded allies in multi-sectoral networks -or 'advocacy coalitions'. The interplay between different coalitions gives initial direction to the policy process. Over time, this interplay is then further shaped by two other meta-variables: the distribution of resources across coalitions, which includes not just 'legal rules and procedures' but also public support, expertise, cash and charisma; and changes to the political environment both inside the policy arena (for example, localized scandals) and outside (such as economic crises and changes in government). When all these moving parts are considered together, they give rise to a pluralistic, agentic and open-ended reading of the policy arena -one driven not by 'formal-legal' state-centric hierarchies but rather by the interplay between different advocacy coalitions in an ever-changing political environment.
Over subsequent sections, we first use these theoretical propositions to identify three advocacy coalitions in the UK and Belgian private security regulation policy arenas: the 'reformers', who believe in the supremacy of state over market in the policing landscape (at a deep level) and seek to control the private security industry through state regulation (at a secondary level); the 'legitimators', who see a prominent role for both state and market in the policing landscape (at a deep level) and look to enhance the comparatively diminished status of the industry through state regulation (at a secondary level); and the 'free marketeers', who likewise see a prominent role for both state and market in the policing landscape (at a deep level), but view this as best accomplished by opposing state regulation and/or deregulating the industry (at a secondary level) (see also White, 2018). For a summary of these coalitions, see Table 1. We then use these propositions to trace the interplay between these coalitions over time, as they negotiate their way through a series of changes both inside and outside their policy arenas.
This approach, we reason, can be used to shed new light on the dynamics of private security regulation. If we were, for instance, to use as our compass the formal-legal approach, set against the backdrop of the recent comparative plural policing scholarship, we would expect to find the free marketeers controlling the UK policy arena and the reformers dominating the Belgian policy arena. This would fit with the emergent pattern identified in this field of research. Significantly, though, this is not the pattern we reveal over the following sections. Through our application of the ACF, we instead uncover a complex interplay between all three coalitions -and the different state-market balances they represent -over different periods of time, with notable convergences under the present conditions of post-financial crisis austerity. This is an important finding that not only makes an original contribution to private security regulation scholarship more specifically but holds notable implications for comparative plural policing scholarship more broadly, as we discuss later. To clarify, though, we are not asserting that our approach is intrinsically superior to the formal-legal one. Nor that it somehow 'completes' the story of private security regulation in the UK and Belgium -there are other key variables in play, not least different political economic backdrops across these two countries, which also inform the state-market balances under examination (see Esping-Andersen, 1990). We are simply arguing that the cut and thrust of democratic politics -and the convergences it brings into view -should be given careful consideration because of the fresh insights it brings to these discussions. Before turning to our two case studies, however, it is necessary to explain the origins of our comparative dataset. It is made up of two parts. The UK part comprises: archival material from the Home Office, the police and the private security industry covering the post-1945 era; an extensive review of primary and secondary literatures on UK plural policing; and 82 semi-structured interviews with representatives from the UK government, the police and the private security industry conducted between 2006 and 2014 (for more detailed methodological discussion, see White, 2010White, , 2014White, , 2015a. The Belgian part comprises: archival material from the Belgian Private Security Directorate, the police and the private security industry covering most of the 20th century; an extensive review of primary and secondary literatures on Belgian plural policing; and 23 semistructured interviews with representatives from the Belgian public and private security sectors conducted in the first quarter of 2020 (for a more detailed methodological discussion, see Leloup, 2021). These different parts were initially collected through separate research projects (see funder acknowledgements below), which has (perhaps inevitably) resulted in certain disparities when synthesized into a single comparative dataset -as we are doing here. The most robust component is the complementary archival material set against the backdrop of the extensive literature reviews. As a consequence, it is this component we rely upon the most when reconstructing the interplay between different advocacy coalitions in the UK and Belgium. The interview data are less well aligned, with discrepancies in terms of both volume and temporality. Recognizing this issue, we primarily use these data for purposes of behind-the-scenes understanding and triangulation, and for the most part we refrain from using direct quotes from these interviews. We are confident that these decisions enable us to draw reliable comparative inferences from our dataset. With these methodological considerations in mind, we now switch over to empirical matters, turning first to the UK case study and then the Belgian one. Briefly, though, it is important to emphasize that, for ease of comparison, each case study not only focuses on the interplay between the same three advocacy coalitions over time but is also divided into the same five chronologically ordered sub-sections: context, coalition formation, regulation, austerity and state-market balance.

Context
Although localized private security arrangements were commonplace throughout the UK in the 19th and early 20th centuries -with locks, safes, alarms, watchmen and guards readily available to the wealthier and commercial classes (Churchill, 2017) -the large private security companies that dominate the industry today, such as G4S, Mitie and Securitas, did not emerge until the postwar era. By 2013, the industry had an estimated annual turnover of €3.97 billion and employed 364,753 private security officers -more than twice the number of police officers, which stood at 162,324 (Confederation of European Security Services [CoESS], 2013). 1 Since the turn of the 21st century, the industry has been regulated by the Security Industry Authority (SIA) -a public body underpinned by the Private Security Industry Act 2001 and accountable to the Home Office -which is tasked with reducing criminality and raising standards across the sector. To do this, the SIA licenses individual private security officers in line with 'fit and proper' person criteria and training requirements. However, the SIA is not empowered to license companies themselves -a vacuum that places significant limits on its regulatory capacity (White, 2015b). As previous sections have demonstrated, the combination of an established private security industry dominated by large multinational companies and a comparatively light-touch system of state regulation has prompted many scholars to characterize plural policing -and more specifically private security regulation -in the UK as straightforwardly market-friendly in orientation. We view this as a misrepresentation, however. The interplay between advocacy coalitions in this policy arena reveals a series of state-centric ideals and calculations that challenge this depiction.

Coalition formation
The decades immediately following World War II are often referred to as the 'postwar consensus' -a period marked by a widespread belief in the capacity of the social democratic state to distribute goods and services throughout economy and society. Significantly, it was this backdrop that gave rise to the 'golden age' of the police. As Reiner (2010: 68) puts it: 'the relative social harmony and consensus of the midtwentieth century, symbolized by the Battle of Britain and the Festival of Britain, was also the finest hour of the British bobby myth.' This deep-seated cultural attachment to the police -or 'police fetishism' (Reiner, 2010) -held notable implications for the nascent private security industry and the question of regulation. For present purposes, the most salient implications relate to how these sentiments shaped the deep and secondary beliefs of two emergent advocacy coalitions in the private security regulation policy arena: the reformers and the legitimators.
The reformers comprised politicians from all parties alongside senior police officers. At a deep level, they believed in the supremacy of the state in the policing landscape. At a secondary level, they supported the introduction of statutory regulation to control and limit the activities of the private security industry. Their primary concern, as one Conservative MP succinctly exclaimed in an unsuccessful 1969 Bill, was that 'there is no one available to keep an eye on the "private eye"'. 2 The legitimators were also made up of politicians from all parties, but included senior industry executives as well. At a deep level, they believed in a prominent role for both state and market in the policing landscape. At a secondary level, they supported the introduction of statutory regulation to imbue private security companies with the quality of 'stateness' -a highly prized attribute in a landscape shot through with police fetishism. Their agenda is neatly captured by another Conservative MP (who also served on Group 4's board of directors) in a further unsuccessful Bill four years later: We must all learn to take crime prevention more seriously. If that is the case the private security industry will have an increasingly important part to play. This part should be encouraged and the intention of the Bill is to encourage good security firms and to ensure high standards. 3 This means that, although these two advocacy coalitions contrasted in their interpretation of the state-market balance, they came together in their support of statutory regulation.
During the 1960s and 1970s, they both consolidated their identities and accumulated resources, gradually shaping the terms of debate in the private security regulation policy arena, most notably when certain reformers and legitimators teamed up in 1977 to sponsor two more pro-regulation Bills in the House of Commons. 4 At this time, however, they had not quite yet built sufficient momentum to alter the free market status quo.
During the 1980s, the fortunes of these advocacy coalitions started to wane. The neoliberal policy turn -embodied in four successive Conservative governments -privileged the distributive capacity of the market over the state, fostering the emergence of another advocacy coalition: the free marketeers. The free marketeers comprised Conservative politicians and industry representatives, many of whom were former legitimators. At a deep level, they believed in a prominent role for both state and market in the policing landscape. At a secondary level, they championed the unfettered marketplace and opposed regulation. Although they recognized the presence of 'police fetishism', they did not regard it as a constraint that needed to be overcome by enhancing the 'stateness' of the industry through regulation. Moreover, given their roots in the government, the free marketeers were able to keep regulation firmly off the negotiating table. For example, when questioned about evidence of malpractice in the industry towards the end of the decade, members of the Conservative government -including the Home Secretaryrepeatedly asserted their conviction that the rigors of market competition represented the best solution. 5 In the late 1980s and early 1990s, though, a series of scandals realigned the interplay between these three advocacy coalitions. In 1988, the Association of Chief Police Officers (ACPO) uncovered substantial evidence of criminality in the industry (ACPO, 1988). Then in 1989 a Royal Marine barracks in Deal protected by Reliance Security was bombed by the Irish Republican Army, killing 11 marines and seriously injuring 22 others. Reliance Security -and the private security industry more generally -received damning coverage in the media, with one article in The Independent reporting how 'some guards refused to patrol the graveyard because they believed it was haunted' (O'Sullivan, 1989: 2). Lastly, in 1993 Group 4 commenced a Home Office contract worth £9.5 million per year to escort prisoners between police stations, courts and prisons and accidentally allowed seven prisoners to escape in the first few days, drawing yet more criticism. These events re-energized and augmented the membership of the reformers (who saw a need to better control the industry) and the legitimators (who saw a need to enhance the status of the industry through a symbolic association with the state). During the 1989-90 parliamentary session the government was required to provide no fewer than 13 written answers on the matter of private security regulation, and, in the three years after the Deal bombing, four pro-regulation private members' bills were introduced into the House of Commons, with two more following the Group 4 prisoner escapes (though none was successful in securing a second reading). 6 The momentum was moving from the free marketeers towards the reformers and legitimators.

Regulation
Responding to this shift in momentum, in 1994 the House of Commons Home Affairs Committee launched an inquiry into private security regulation. The Committee received evidence from a wide range of stakeholders, with most industry representatives casting themselves as legitimators and almost all other participants -including Members of Parliament, the Police Federation, the Police Superintendents' Association and ACPOjoining together as reformers. 7 Only the Conservative government -still animated by neoliberal free market ideals -bucked this trend by adopting a noncommittal stance. 8 After receiving evidence, the Committee concluded in favour of state regulation. 9 However, the Conservative government chose to remain silent on the issue. It was only after New Labour -now firmly aligned with the reformers -initiated an Opposition Day debate on the matter in early 1996 that the government was forced into action, publishing outline proposals for state regulation (Home Office, 1996).
In 1997, New Labour won a general election on the back of a manifesto promising a 'third way' between state-centric and market-friendly policy paradigms. This approach played a vital role in translating the outgoing government's tentative proposals into concrete legislation because it straddled the reformer and legitimator positions and thus received widespread support. This was evident in the new Home Secretary's address at the 1997 British Security Industry Association (BSIA) annual luncheon, where he pledged to introduce 'proper regulation' to 'get your industry onto a sound footing' and 'restore public faith in your important role in the fight against crime' (Straw, 1997). This rhetoric gave clear expression to the deep and secondary beliefs of the reformers and the legitimators, while paying scant attention to the corresponding beliefs of the free marketeers. This new configuration of advocacy coalitions resulted in the relatively smooth passage of the long-awaited Private Security Industry Act 2001 into the statute books, though it would not usher in long-term stability in the policy arena -within just a few short years, unforeseen events would take over.

Austerity
The 2008 global financial crisis hit the British banking sector hard. In the ensuing turmoil, New Labour nationalized a number of banks, amounting to 'the largest UK government intervention in financial markets since the outbreak of the First World War' (Bank of England, 2008). The 2010 general election unsurprisingly revolved around the issue of how to repair the resulting hole in the public finances. The Conservatives -who came to power as the senior partner in a coalition with the Liberal Democrats -pursued a textbook neoliberal solution: downsize the state and remove market constraints to ease the burden on the Exchequer and stimulate economic growth. For present purposes, this agenda resulted in two notable developments. First, the central government police budget was immediately cut by £2.42 billion (in real terms) over the period 2010/11-2014/15, significantly reducing state capacity when responding to crime and disorder (Her Majesty's Inspectorate of Constabulary [HMIC] 2013: 15). Second, on 22 September 2010, the BBC broke the news that the SIA -alongside 903 other regulatory bodieswas to 'face axe in cost drive' (Campbell, 2010), thereby freeing up market capacity in the same area of activity. In other words, the free marketeers had emphatically re-entered the policy arena with the full authority and resources of the government behind them.
The other two advocacy coalitions were not suddenly cast aside, however. The legitimators in particular were vociferous in their reaction. Two days later, the Chairman of the Security Institute -a high-profile industry body -emphasized that regulation has 'increased confidence and trust in the private security sector on the part of the police and, indeed, the public in general'. He continued: 'Any steps taken which could damage such trust and confidence would, in our view, be a retrograde act' (Security Management Today, 2010: para. 6). The message here is clear: far from representing burdensome red tape, state regulation had been instrumental in enhancing the status of the industry -an especially important consideration given the growing number of outsourcing opportunities coming from the budget-constrained police (White, 2014(White, , 2015a. This was a widely held view, for all the main professional and trade bodies quickly joined forces in the proregulation Security Alliance and sent a letter to the Home Secretary on 13 October 2010 opposing abolition. The next day, the Cabinet Office released a brief statement announcing that, rather than being abolished, the SIA would instead undergo a 'phased transition to a new regime'. At the time of publication (May 2021), no such transition has taken place and the SIA continues to operate in line with the expectations of the reformers and legitimators.

State-market balance
We noted earlier that, if the divergent plural policing trajectories identified in recent comparative scholarship are to hold true, we would expect the private security regulation policy arena in the UK to be dominated by the free marketeers, whose primary and secondary beliefs align with a more market-friendly depiction of plural policing arrangements. Interpreting our historical and contemporary data through the advocacy coalition model, we have now demonstrated that this is not necessarily the case. Although the free marketeers have certainly contributed significantly towards the dynamics of private security regulation in the UK, two other advocacy coalitions have played an equal if not more important role: the reformers, whose primary and secondary beliefs broadly align with a more state-centric depiction of plural policing arrangements; and the legitimators, whose primary and secondary beliefs combine market-friendly and state-centric elements. In other words, this supposedly market-friendly part of the plural policing landscape is more state-centric than recent comparative scholarship suggests. We now turn to the Belgian case study, which similarly runs counter to the portrayal advanced in this recent scholarship.

Context
Similar to the UK, localized private security arrangements first emerged in Belgium during the early 20th century -primarily watchmen and guards employed in the maritime, commercial and financial districts of Antwerp and Brussels -with large private security companies such as G4S, Securitas and Seris (formerly Garde Maritime Industrielle et Commerciale) appearing in more recent decades (Leloup, 2015(Leloup, , 2017(Leloup, , 2021. However, owing to the lower number of market opportunities -in part a reflection of the smaller Belgian economy -these companies have not reached the same proportions as their UK counterparts. By 2013, the industry had an estimated annual turnover of €641.7 million and employed 17,522 private security officers -less than half the number of police officers, which stood at 39,934 (CoESS, 2013). Interestingly, though, Belgium has a much longer history of private security regulation compared with the UK. To begin with, the anti-militia Law of 29 July 1934 and its extension of 4 May 1936 required companies to gain official authorization as 'non-political organizations'. As the industry matured, this makeshift approach became ever more impractical, leading to the Law of 10 April 1990, which charged the Directorates Private Security and Private Security Control -both part of the Belgian Federal Public Service Home Affairs -with licensing individuals and companies in line with extensive quality control criteria. This legislation has subsequently been replaced by the Law of 2 October 2017, which streamlines the criteria enumerated in its predecessor. As earlier sections have illustrated, the combination of a relatively small private security industry together with a comparatively strict system of state regulation has caused numerous scholars to characterize plural policing -and more specifically private security regulation -in Belgium as state-centric in orientation. We regard this as a misinterpretation, however. The interplay between advocacy coalitions in this policy arena brings to light a series of market-friendly ideals and calculations that challenge this portrayal.

Coalition formation
Echoing the UK experience, the immediate postwar decades in Belgium were defined by a popular belief in the capacity of the corporatist welfare state to distribute goods and services across economy and society (De Preter, 2016). This did not, however, precipitate a corresponding 'golden age' of the police. During this period the three police branches -municipal, judicial and military -were locked in often acrimonious competition over resources. To make matters worse, the military police were repeatedly accused of engaging in oppressive actions towards Belgian citizens, leading to numerous protests between the 1950s and 1970s (Van Outrive et al., 1991). As a consequence, the police did not accumulate the same degree of legitimacy as their UK cousins at this time. Importantly, though, cultural attachment to the institution was still sufficiently widespread to shape the deep and secondary beliefs of two nascent advocacy coalitions in the private security regulation policy arena: the reformers and the legitimators.
First on the scene in the 1960s was the Belgian permutation of the reformers, made up of magistrates from the public prosecutor's office and civil servants from the Ministry of Justice. Like their UK equivalents, they believed in the supremacy of state over market in the policing landscape (at a deep level) and sought to control and limit the private security industry through state regulation (at a secondary level). Their immediate concern was how to regulate the burgeoning industry using the Laws of 1934 and 1936, which were not actually designed for this purpose -indeed, these laws are perhaps best viewed as regulations that hold implications for the private security industry rather than bona fide private security regulation. They accomplished this task by adding quality control criteria to the process of authorizing companies as non-political organizationsthough in due course they also started petitioning for new sector-specific legislation. During the 1970s, this coalition was joined by the Belgian iteration of the legitimators, mainly comprising industry executives lobbying through professional bodies such as the Belgian Professional Association of Security Companies (BVBO-APEG). Like their UK contemporaries, they believed in a prominent role for both state and market in the policing landscape (at a deep level) and looked to enhance the comparatively diminished status of the industry through state regulation (at a secondary level). Although they saw benefits in the authorization process set out in the Laws of 1934 and 1936 -which as one industry executive explained in 1978 could be advertised as 'proof that the Belgian state and authorities trust our private security companies' 10 -they also began to lobby for new sector-specific legislation towards the end of the decade. 11 As Belgium too went through something of an early 1980s neoliberal transformation (De Preter, 2016), one further advocacy coalition entered the policy arena -the Belgian incarnation of the free marketeers, principally made up of politicians and industry representatives, many of whom were former legitimators. Like their UK counterparts, they interpreted the rise of neoliberalism as a sign that the industry no longer required statecentric legitimation and could instead trade on its market credentials alone. Their initial objective was therefore to neutralize the Laws of 1934 and 1936 and to bring about a system of self-regulation. In 1981, for instance, the BVBO -which now housed both legitimators and free marketeers -pushed for 'alternative solutions' to these laws such as 'granting us an official statute', which would definitively categorize the industry as 'nonpolitical' and therefore outside the scope of anti-militia legislation. 12 Although the free marketeers represented an important voice during the early 1980s, a series of (disconnected) events around this time coalesced to privilege the pro-regulation agenda of the reformers and legitimators.
To begin with, this period witnessed a spate of violent attacks by far-right movements such as Vlaamse Militanten Orde, casting a shadow over private security companies that, through historical convention, still operated as 'authorised private militias' (Leloup, 2021). This shadow was reinforced by a number of scandals surrounding the abusive practices of private security guards working for Wackenhut (Braeckman and De Kock, 1980). Then in 1980 a Parliamentary Committee of Inquiry was convened to address persistent issues of police corruption, bribery and ineffectiveness. Importantly, this Inquiry questioned the capacity of the police to control the private security industry in the absence of adequate state regulation and, at the same time, fostered debate about whether the industry should be employed to combat rising crime rates, challenging the hitherto dominant principle that policing ought to be the exclusive domain of the police. 13 These events not only prevented the free marketeers from realizing their objective of self-regulation but also galvanized the reformers and legitimators in their pursuit of new sector-specific regulation. In 1983, the Commission of Internal Affairs announced that new regulation would indeed be beneficial, 'since their [the private security companies] resources and activities are becoming more and more important, and their sphere of actions is growing hugely', 14 setting the scene for the introduction of bona fide private security regulation.

Regulation
The first concrete step towards a new regulatory regime was taken in 1986 when the Ministry of Justice introduced the Gol Bill, which included, among other things, a tailored system for regulating private security companies. Whereas the Bill was tentatively welcomed by the reformers, the legitimators enthusiastically anticipated this legislation. For example, the Belgian Professional Association of Security and Alarm Companies (BBAB) -another prominent trade organization with a vocal legitimator membership -issued the following statement: 'Good News! We hope so in any case! . . .. The law is in preparation and is expected to be ratified by the end of 1986. A historic moment that the professional security industry has been eagerly awaiting' (BBAB, 1985: 1). Soon after, however, parliament descended into a political crisis, which lasted 182 days, ultimately stymying the Bill's progress. Towards the end of the decade, a new centre-left government -a coalition of Flemish and Frenchspeaking Christian Democrats, Flemish and French-speaking Socialists and Flemish Nationalists -finally returned to the issue and duly passed the Law of 10 April 1990, which regulated individuals and companies operating under both contractual and inhouse arrangements. Reflecting the state-centric ideology of the government, this new sector-specific law was, in the words of the Minister of Internal Affairs, designed to 'strictly control the private security companies and limit their growth', 15 and therefore reflected the deep and secondary beliefs of the reformers over those of the legitimators and free marketeers. The reformers proceeded to dominate the policy arena throughout the 1990s, amending the law on a number of occasions to enhance its rigour (De Vroe, 2018).
Around the turn of the 21st century, however, events once again conspired to realign these advocacy coalitions. To begin with, a new spate of law enforcement scandals served to further damage police legitimacy. This resulted in the most extensive police reform in Belgian history (known as the 1998 Octopus Agreement) and the introduction of the community-oriented policing and integral security management model, which gave the market a more prominent role in combating crime and disorder (Ponsaers and De Kimpe, 2001). Furthermore, in 1999 a new coalition government led by the Liberal Party came to power espousing a more market-friendly approach to public service delivery. This ideological shift, combined with mounting pressures on police resources, triggered the so-called Kerntakendebat -a further policy initiative that sought to identify 'non-core' police functions suitable for outsourcing to the market (Cools, 2006). This new governmental emphasis on raising the status of the private security industry -as opposed to just 'controlling and limiting' it -augmented the position of both the legitimators (who seized on the moment to lobby for more light-touch private security regulation) and the free marketeers (who continued to champion self-regulatory initiatives). Indeed, capitalizing on these changes in the political climate, free marketeers in the BVBO doubled down on their efforts to persuade the government that they could operate 'self-regulation while upholding high standards and values' (BVBO, 2006: 29). It would, however, require some even deeper ruptures in the Belgian political system before these two advocacy coalitions could make a decisive impact upon the legislative status quo.

Austerity
In 2008, the escalating global financial crisis reached Belgium too, prompting a massive publicly funded bailout of the banking sector and sending the country into a prolonged period of austerity. To compound matters, the long-serving centrist coalition government was soon afterwards forced to resign following allegations of exercising unlawful political influence, provoking a political crisis that lasted for almost two years. It was not until 2011 that a new coalition government -comprising Flemish and French-speaking Socialists, Flemish and French-speaking Liberals and Flemish and French-speaking Christian Democrats -finally returned a sense of stability to the political scene. Reproducing the pattern of the previous decade, this government's now austerity-driven strategy of reducing public spending and promoting public-private partnerships across the policing landscape continued to raise the profile of the legitimators and free marketeers at the expense of the reformers. 16 However, it would take one more change of government for legislative reform to take place.
In 2014, a new centre-right coalition government -made up of Flemish and Frenchspeaking Liberals, Flemish Nationalists and Flemish-speaking Christian Democratsentered office on a market-friendly neoliberal platform. Against the backdrop of a systematic effort to make the state more 'slim and efficient', 17 it evaluated the Law of 1990 with the intention of simplifying private security regulation and creating more opportunities for the outsourcing of police functions to the market. 18 This process was reinforced by the terrorist bombings of 22 March 2016, which concentrated already limited police resources in counter-terrorism units, leaving other units depleted and thus more open to cost-saving measures such as outsourcing. Seeing the tide turning against them, high-profile reformers began to raise concerns, with the Police Union pointedly questioning 'if the objective of the Michel government is to hand over the Federal Police's tasks to private security companies or to deliberately destroy this federal institution?' 19 Legislative reform was imminent. Just over a year later, the Law of 2 October 2017 came into effect, streamlining the now abandoned Law of 1990 and signalling a new approach to private security regulation. Its intention was no longer simply to protect the public from the industry, but rather to empower the industry. As one high-profile manager at G4S Belgium put it: 'this is what the legislator wanted to achieve with the new private security regulation. They don't want to make everything compulsory, but they want to provide a legal framework which the private sector can fill to a certain degree' (Interview, 2020). For its part, the government clearly hoped such market empowerment would, in the words of the Minister for Internal Affairs, 'encourage the establishment of partnerships between the police and the private sector' (BeSafe, 2017). These aspirational shifts in the state-market balance ultimately represent something of a compromise between the reformers, the legitimators and the free marketeers.

State-market balance
We remarked earlier that, if the divergent plural policing trajectories identified in recent comparative scholarship are to hold true, we would expect the private security regulation policy arena in Belgium to be dominated by the reformers, whose primary and secondary beliefs roughly align with a more state-centric depiction of plural policing arrangements. Analysing our historical and contemporary data through the advocacy coalition model, we have now illustrated how this is not the case. Although the reformers have certainly contributed significantly towards the dynamics of private security regulation in Belgium, two other advocacy coalitions have also performed an important role: the free marketeers, whose primary and secondary beliefs broadly align with a more market-friendly depiction of plural policing arrangements; and the legitimators, whose primary and secondary beliefs combine state-centric and market-friendly elements. To our minds, this supposedly state-centric part of the plural policing landscape is therefore more marketfriendly than the comparative scholarship examined above suggests.

Conclusion
Our comparative analysis demonstrates that, when private security regulation in the UK and Belgium is explored through the lens of the advocacy coalition framework, there is considerably more going on in terms of convergence and divergence than the formallegal approach suggests. To recap, the formal-legal approach focuses exclusively on 'legal rules and procedures' and, as a consequence, brings to light divergences with regard to how private security regulation is codified in UK and Belgian statute books, with the former coming across as light-touch and market-friendly and the latter manifesting as heavy-handed and state-centric. In a sense, of course, this is quite true. When ranked in terms of legal comprehensiveness, the UK Private Security Industry Act 2001 is indeed more laissez-faire than the Belgian Law of 2 October 2017. This is not the whole story, however. As we show in our re-examination of the same variables through the advocacy coalition framework, which focuses on the cut and thrust of democratic politics, the dynamics of private security regulation across these two countries have been animated by a distinctive set of advocacy coalitions -the reformers, the legitimators and the free marketeers -each representing a specific state-market configuration. Although there have certainly been divergences in the relative influence of these coalitions over time -a corollary not just of the aforementioned legal contexts but also of different political economic backdrops (see Esping-Andersen, 1990) -there have also been numerous convergences (see Table 2). These have arisen because in each country the three coalitions have responded to shifts in their surrounding political environments at the local (legitimacy and scandals), national (electoral outcomes and legislative reform) and international (neoliberal turn and financial crisis) levels in remarkably similar ways. Significantly, these convergences are most evident under present conditions of post-financial crisis austerity, with the three coalitions -and the state-market balances they represent -holding more or less equal sway in the UK and Belgian policy arenas. Indeed, with this in mind, we can actually push our analysis further by reassessing the legal divergence identified within the formal-legal approach. Although on paper the Belgian Law of 2 October 2017 is certainly more state-centric than the UK Private Security Industry Act 2001, it is also more market-friendly than its predecessor (the Law of 10 April 1990), reflecting the interplay of coalitions at the time of its inception. This has served to reduce the legal gap between the two regimes. In other words, under conditions of post-financial crisis austerity, there has been not only political convergence across the UK and Belgian private security regulation policy arenas, but legal convergence too.
Although these findings make an original contribution to private security regulation scholarship more specifically, they also hold notable implications for recent comparative plural policing scholarship more broadly, which takes us back to our starting point in this article. As we noted earlier, this scholarship has exposed the contours of Anglocentric bias in this field by emphasizing a divergent set of plural policing trajectories in the Anglosphere and in Continental Europe. It has accomplished this by essentially reproducing a market-friendly depiction of the Anglosphere and then challenging the analytical reach of this depiction by emphasizing incongruencies with a state-centric portrayal of Continental Europe. This line of enquiry has undoubtedly led to significant insights, but our comparative analysis suggests that it has a tendency towards overcorrecting for the biases it seeks to address. The preceding sections show that, although there are indeed divergences in plural policing arrangements across these two regions, there are also important convergences. There is common ground. Of course, our analysis is a partial one: it focuses on just one country from each region; it concentrates on a single strand of the heterogeneous plural policing landscape; and it is shaped by a particular theoretical framework. We are not therefore claiming that convergences between the Anglosphere and Continental Europe are ubiquitous or that the advances made in recent comparative plural policing scholarship need to be undone. Nor indeed are we intimating that the Anglocentric bias running through this field of research is not real -it most certainly is. But we are asserting that we need to be careful in how we handle and correct for these biases. In particular, we must not lose sight of the convergences across these two regions and we should, by extension, remain open to the transfer of analytical and practical plural policing knowledge between the Anglosphere and Continental Europe.  (1) Pre -1979- Pre-1980 Reformers (2) Legitimators (1) Reformers (2) Legitimators Coalition Formation (2) 1979-2001 1980-1990 (1) Free marketeers (2)