Abstract
The study of ‘policy transfer’ has been subject to sustained criticism, in particular by critical policy studies scholars. This critique—together with the rather marginal role that policy transfer research has played in criminological debates to date—raises questions about the continued utility of such research in scholarly discussions of crime control and penal policy-making. However, we argue here that such studies can enhance our understanding of the local, national and global influences over crime control policy formation. In particular, the developing interest in comparative criminology, in the political economy of punishment, and in the ‘proximate causes’ of penal change, are all areas to which this work can make a useful contribution. Although we feel that some elements of the critique are over-stated, the critical policy studies notions of ‘mobilities’ and ‘assemblages’ offer important advances that capture more fully the complexities of the processes involved in the cross-national movement of penal policy.
Introduction
The ‘punitive turn’ (inter alia, Downes, 2011; Jacobson, 2006) has drawn the attention of scholars around the world to issues of comparative penal policy. Within this body of activity there has been a significant debate around issues of policy ‘convergence’—the extent to which (and reasons for which) penal policy across a range of industrial democracies became increasingly similar, and in particular more punitive, over the latter decades of the 20th century (Crawford, 2011). In parallel, and linked to this, increasing attention has also been paid to the diversity of policy trajectories in different national contexts, and has sought to identify those factors that might explain why some countries appear to be more resistant to punitive shifts. Such work has drawn on a range of disciplinary perspectives. Distinguished theoretical sociologists have provided sophisticated accounts of the ways in which deep structural, cultural and economic shifts in late modern societies have predisposed nation states towards more punitive policy stances, leading to a degree of cross-national policy convergence (Garland, 2001). Scholars working closer to the tradition of political economy have explored the associations between the nature of penal policy and such factors as political institutions, types of electoral and party systems, forms of economic organization and the structure of the mass media (Cavadino and Dignan, 2005; Lacey, 2011). Others have drawn on anthropological and historical perspectives to map out and explain the dimensions of particular contemporary penal cultures (Melossi, 2001; Whitman, 2003).
This article takes a slightly different path, but seeks to contribute to this body of work by highlighting the potential contribution of scholarship that focuses on the ways in which policy ideas, institutions and practices flow across national boundaries. We feel that these approaches, currently rather marginal to criminological debates, have much to offer studies of penal policy change. In particular, they have the potential to provide more nuanced explanations of the ways in which extra-jurisdictional developments can shape the domestic ‘proximate causes’ of penal policy trajectories (Garland, 2013). Such approaches problematize the notion of ‘policy’, draw attention to its dynamic and emergent nature and, crucially, provide the impetus for detailed and systematic empirical study of policy formation processes over time and space. Such work, we would suggest, is of relevance to key scholarly debates and current theories of punishment and control. In addition, it is clear that—whether explicitly acknowledged or not—much work on penal change is underpinned by a broadly shared normative position, and hence a normative project, the aim of which is to find ways of changing the direction of policy towards greater penal moderation (Loader, 2010). If academic criminology is to become more than a ‘counsel of despair’ (see Zedner, 2002: 361), systematic empirical study of the policy formation process—including the ways in which cross-national influences work to shape crime policy—is a pre-requisite for identifying opportunities to bring about progressive change.
The article has four main sections. First, we provide a brief overview of the key themes in the literature on comparative penal policy change which forms the context for the article. Second, we discuss how ‘orthodox’ policy transfer scholarship has been applied to the fields of criminal justice and penal policy. The third section discusses emergent critiques of orthodox policy transfer studies by scholars interested in notions of ‘policy mobility’ and ‘policy assemblages’. We conclude by outlining some implications of this critique for future work in the study of policy mobilities in crime control.
Theorizing penal change
The field of comparative criminological study has seen significant growth, and, despite all the difficulties of undertaking such work, seems to be becoming more prominent (Aebi and Linde, 2015). It is broad in its focus, taking in such varied topics as criminal careers (Farrington, 2015), violence and democratic institutions (Karstedt, 2015) and criminal justice institutions (Jones and Van Steden, 2013; Nellis and Bungerfeldt, 2013). More particularly, there has recently been increasing interest in the comparative study of policy-making and policy-development—both within and between nation states (Barker, 2009; Lappi-Seppälä, 2008; Miller, 2008). Although some of this work, especially cross-national comparative scholarship, remains at quite a high level of abstraction, the growth of interest in such analyses opens a window for detailed studies of policy worlds.
These developments have been influenced by the considerable criminological attention paid to the rising punitiveness that appeared to be visible across many western democracies in the last decades of the 20th century and the implication, whether explicitly stated or not, that there was a growing convergence between the penal policies of these countries. Influential scholarly accounts emerged which, though they differed markedly in the specifics of their explanations, all focused on the broad structural economic, social and cultural shifts taking place in ‘late modern’ societies and which appeared to conduce such nations towards more punitive policy stances, leading to a degree of cross-national policy similarity (Christie, 2000; Garland, 2001; Wacquant, 1999, 2001, 2009). All these analyses have, in different ways, linked increasing penal severity with the spread of ‘neoliberalism’, although as a recent set of contributions to the journal Social Anthropology demonstrate, there is much that remains contested about this term and the nature of its relationship to penality. Wacquant (2012), for example, contends that the political project of neoliberalism links purposively the promotion of economic freedoms for some with increasing penal severity for others. Wacquant is critical of both the hegemonic ‘market rule’ and the governmentality models of neoliberalism. He suggests that the market rule account of the spread of neoliberalism is too deterministic and utilizes an overly narrow account of the institutional bases of influence. By contrast, he argues that their analyses of the migration and mutation of the ‘technologies of conduct’ lead governmentality scholars to find neoliberalism ‘everywhere and nowhere at the same time. It becomes all process and no contents; it resides in flowing form without substance, pattern or direction’ (2012: 70). Although his analysis recognizes variations in the nature and form of neoliberal governance in different contexts, these are viewed as deviations from an emerging global standard. Thus, he continues to lean towards a position that emphasizes cross-national penal convergence as the key focus for analysis.
Prompted by such models of global penal convergence, a range of scholars have argued that such broad accounts of penal change underplay persistent variations in national and local penal practices and policies. An important body of work has consequently taken penal variation as its starting point and has begun to delineate the factors that shape such variation. Work on the comparative political economy of punishment has suggested key features of the political economy and social organization of different democratic countries that shape distinctive approaches to penal policy (Green, 2008; Lacey, 2006, 2007, 2011; Pratt and Eriksson, 2013; Tonry, 2007). This work has illuminated the complex inter-connections between national cultural dispositions, political institutions and aspects of social and economic organization, which appear to interact in ways that predispose nations to greater or lesser penal severity (Miller, 2016). There are parallels to these arguments in the critiques of Wacquant’s position on the spread of neoliberal forms of penality. Some have argued that Wacquant fails to take full account of the complexity of the geographies of neoliberalization (Peck and Theodore, 2012), and adopts an overly structuralist account of its development and spread (Collier, 2012). Hilgers (2013: 79) argues that whatever the surface appearance of similarity, policy reforms are always ‘idiosyncratic’, and that ‘[a]s soon as they are ingested, they change: they end up transformed, more autonomous, playing a new role’.
The implication is that the study of policy change must ‘go beyond a perspective that contrasts case studies with a ‘pure variant’ (Hilgers, 2013: 79–80) and rather should focus on
the historicity of the spaces in which policies are put into action, the intentional constructions but also involuntary historical formations in which they become entangled, and the transactions, negotiations, associations, working misunderstandings and chains of translation that give them their flexibility and support their deployment.
Campbell and Schoenfeld (2013: 1383) echo this point in their analysis of the transformation of the US penal order. What is needed, they argue, ‘is a theoretical framework that can explain both the consistent, persistent punitive shift and the variation in the extent and timing of that shift’. Their analysis, while offering conditional support for structuralist explanations of penal change (which have been largely dominant in criminology to date), suggests that an understanding and analysis of the political context, partisan politics and the nature of political institutions is vital to any fully developed explanation. Peck and Theodore (2012: 178) go somewhat further, in their argument for a more fully spatialized account of the process of neoliberalization across the globe, and by implication its variable impacts on penal policies, suggesting that: ‘the nonlinear, multidirectional course of real-world neoliberalisation cannot be reduced to a process of enacting a singular, pristine plan or grand design’.
The value of focusing primarily on penal convergence between nation states has been further brought into question by recent work that has explored variations within states. In particular, it has explored the importance of the institutional, cultural, economic and political make-up of sub-national jurisdictions in understanding penal variation (see, for example, Barker, 2009; Lynch, 2010; Page, 2011; Rubin, 2015), and how, for example, the US federal system operates to facilitate or inhibit the movement of policy (Campbell and Schoenfeld, 2013; Miller, 2008). Although naturally the emphases differ between studies, such approaches support Tonry’s (2001: 518) earlier claim that, notwithstanding the influence of broad globalizing accounts of penal change, ‘explanations of penal policy remain curiously local’. This observation highlights the acute challenges of ‘reconciling sensitivity to local difference with the generalizing imperatives of the comparative’ (Zedner, 1995: 518). To this end, work within comparative penal law has started to examine the interaction between ‘globalizing’ forces and national/sub-national patterns. Fourcade and Savelsberg (2006) suggest that we need to acknowledge the ways in which apparently ‘global’ processes are themselves constructed from below, and they highlight some of the ways in which ‘global’ legal changes are reconfigured by particular national and sub-national institutions and cultures (see also Benson and Saguy, 2005).
In this regard, and significantly in our view, the scholar most associated with broad structural accounts of late modern penal change, David Garland, has recently called for work that adds to macro-level analysis by focusing on the ‘proximate causes’ of mass incarceration and related developments. These he identified as the ‘specific forms of state action’ that ‘are obvious but […] tend to be overlooked in the concern to identify the “prime movers” behind such change’ (Garland, 2013: 484). He added that ‘[s]ocial currents may ebb and flow, but they have no penal consequence unless and until they enlist state actors and influence state action’ (2013: 494). Accordingly, he suggested that ‘the character of the penal state and the processes whereby it responds to social forces, translating (or not translating) political pressures into specific penal outcomes, are always the proximate causes of penal action and penal change’ (2013: 494). 1 In Garland’s view, the problem of much extant inquiry is that it has left us ‘with something of a black box when it comes to demonstrating how penal laws and policies are shaped and how penal decisions are made’ (2013: 492). What is now needed, he argues, is work that will ‘focus on state structures and processes and seek to analyze the […] state in comparative perspective’ 2 (2013: 483).
The growing interest in comparativism generally; debates about the extent, nature and causes of penal policy convergence; the increasing focus of attention on the comparative political economy of punishment; and, latterly, the identification of the study of the proximate causes of penal developments as a scholarly priority—are all areas to which detailed analyses of the cross-national circulation of penal policies and practices can make a significant contribution.
‘Policy transfer’ and crime control
There is now a large body of work—drawing upon a number of different disciplinary traditions—that in broad terms addresses the extent to which, and the ways in which, public policies appear to travel across national boundaries, and with what consequences. The concept of policy transfer is linked to a range of related ideas, including ‘policy learning’ and ‘lesson drawing’ (Rose, 1991), as well as the broader ideas of policy ‘diffusion’ (Walker, 1969) and ‘convergence’ (Bennett, 1991). The overly rationalist lesson-drawing literature was superseded by the alternative notion of ‘policy transfer’, proposed by Dolowitz and Marsh (1996), as an attempt to try to capture both the voluntary and coerced elements of the international spread of particular policy approaches (though see James and Lodge, 2003). Work on policy transfer is most associated with the field of political science. There has since been a significant expansion of work focused on, to quote the most frequently cited definition, the process by which ‘knowledge about policies, administrative arrangements, institutions and ideas in one political setting (past or present) is used in the development of policies, administrative arrangements, institutions and ideas in another political setting’ (Dolowitz and Marsh, 2000: 5). Though far from uncontested, this definition provided both a clear focal point around which an identifiable body of empirical research grew, and has also been the starting point for much of the criticism of the approach that has emerged more recently. There are now studies of cross-national policy transfer in a range of policy spheres including (among others) welfare policy (Peck and Theodore, 2001), environmental regulation (Jordan et al., 2003), education (Bache and Taylor, 2003), transport (Marsden and Stead, 2011) and utilities regulation (Padgett, 2003).
Despite the long-standing interest in how crime control models, mentalities and practices travel between contexts (see Cohen, 1988) the empirical study of policy transfer, at least in the terms outlined by Dolowitz and Marsh (1996), remains relatively rare in criminology (Tonry, 2015). One of the first explicit references to the term ‘policy transfer’ in a criminological publication was by Muncie (2001) who examined the influence of restorative justice ideas originating in Australasia and Scotland on youth justice practice in England and Wales. Jones and Newburn (2007) examined three case studies of widely cited crime control policy ‘transfer’ from the USA (zero tolerance policing, two and three strikes sentencing and privatization of corrections). Despite significant transfer-related activity, there was no evidence of comprehensive importation of US models, with most movement being visible at the level of policy ‘talk’. While the initial inspiration for changes had clear roots in US developments, important differences in the political and legal institutional context worked to shape subsequent developments in a distinctively ‘British’ fashion.
Whereas Jones and Newburn (2007) examined international policy transfers from their perspective as outsiders in the policy process—and a similar approach was later adopted by Ogg (2015) in his study of preventive justice—subsequent studies by Blaustein (2015), Durnescu and Haines (2012) and McFarlane and Canton (2014) have provided insider accounts of the complexities and imperfections of criminal justice policy transfer processes. In the first of these studies, Durnuscu and Haines (2012: 905) were personally involved in a British-funded probation reform project and their experiences and observations led them to conclude that ‘simplistic notions of transfer need to be challenged if the development and delivery of criminal justice services are to develop a more constructive route’. Blaustein’s (2015) study of ‘community policing’ implementation attempts in Bosnia and Herzegovina used ethnographic methods to shed light on the ‘translational’ capacity of seemingly disempowered stakeholders, as well as their ability to navigate cultural and structural constraints. Most recently, Brown et al.’s (2015) analysis of the development and implementation of the notion of ‘Justice Re-investment’ (JR) in the USA and later in Australia has challenged straightforward notions of ‘policy importation’ and helped underline the importance of domestic contextual factors in reconfiguring policy ideas and programmes that have their origins elsewhere.
Studying penal policy flows: A marginalized perspective
Despite this rich body of work, studies of this kind have remained rather marginal to debates about the nature and direction of penal change. There are several reasons one might posit for this. First, some of the most influential criminologists working in the field of cross-national comparison have been preoccupied with the analysis of broad structural trends. From the impact of late modern social change to the rise of the risk society, the dominant concerns have been with the changing ways in which the penal systems of nation states are organized, operate and are governed, and how this links to deeper, underlying social, political and economic changes that apply across a range of national contexts. Understandably such work tends to focus on the general sweep of policy rather than its particular detail.
Second, and relatedly, even where there is a more concrete focus on policy, the bulk of criminological attention has tended to be directed mainly to what one might consider ‘outcomes’—the numbers of people incarcerated for example—rather than policy-making processes (but see Annison, 2016; Page, 2011). For instance, the work that draws on comparative political economy certainly presents a more empirically grounded picture than the grand sociological narratives, one that acknowledges both penal policy convergence and variation between nation states. However, the taxonomies of the correlates of ‘penal severity’—often measured by fairly blunt outcome measures such as overall incarceration rates—present a somewhat static focus on the content and outcome of policy rather than engaging with the complexities of its genesis. Such studies by their very nature cannot shed light on the specific mechanisms of influence via which certain aspects of economic organization or political institutions work to shape particular penal policies. In addition, it can be argued that political-institutional analyses of differences in penal trends have themselves focused too much on the nation state as the unit of analysis, and consequently underplay important ‘sub-state’ dynamics in the shaping of local variations in penal policy and practice as well as ‘supra-state’ institutional frameworks at the transnational level (McAra, 2011). As we discuss below, studies of policy transfer tend to be significantly more concerned with the empirical details of the minutiae of policy development and—perhaps somewhat ironically given one of the thrusts of the constructivist critique we later outline—with a range of policy actors beyond those at nation state level. These lines of enquiry sometimes appear to sit uneasily alongside bigger structural concerns and formal policy outcomes. Third, and linked with the two earlier points, the dominant structuralist criminological paradigm has tended to operate at the level of discourses and, as a consequence, has of necessity paid less attention to the work of particular actors—even elite actors.
Finally, the reasons that have now been regularly rehearsed as explaining the relative dearth of comparative studies within criminology also help to explain the marginal position of policy transfer research (Tonry, 2015). These include: the costs—materially and in terms of time—of undertaking such research; the dynamics of the contemporary academy and the growing emphasis on short-term outputs rather than the deep immersion generally required by comparativism (Downes, 1988); and a variety of problems of method (Nelken, 1994). Notwithstanding the growing interest in comparative studies within criminology in recent times (Heidensohn, 2006; Nelken, 2012), each of these barriers continues to inhibit the expansion of such work and, we would argue, in turn, has and is likely to continue to limit the growth of policy transfer scholarship. Nevertheless, it remains our contention that work of this type is of great value for criminology. Before expanding on this view, we turn our attention to constructivist critiques of the policy transfer literature.
Policy mobilities and the constructivist critique
In recent years, questions have arisen as to the broader utility of the concept of policy transfer. A growing body of work has been critical of what has been viewed as the overly rationalistic, occasionally positivistic tendencies of much orthodox policy transfer scholarship. This literature has important implications for scholars interested in comparative penal change and is worth considering in some detail. The first point to make is that scholars working in the ‘policy mobilities’ field continue, in contrast to criminology (Cook, 2015), to view cross-national policy flows as a significant feature of public policy-making in many parts of the globe (McCann and Ward, 2011). Drawing from their research on comparative social policy in South-East Europe, Lendvai and Stubbs (2009) were early advocates of an interpretivist approach to studying what they identified as ‘assemblages’ and ‘policy translation’. They argued that the idea of an ‘assemblage’ was useful as a means of illuminating how social policy is shaped by ‘contradictory, complex and uneven processes’, while ‘translation’ was to be preferred to ‘transfer’ because, they argued, it helped account for the transformative nature of this process (2009: 676). Both concepts were grounded in what they described as ‘a poststructuralist understanding of power, where power is understood as a temporal, fluid and interactive process’ (2009: 676–677).
Similar arguments, developed in parallel by human geographers, including Jamie Peck, proposed an agenda for ‘critical policy studies’ analysis that viewed the concept of ‘policy mobilities’ as better suited to capturing the increasingly peripatetic nature of public policy across the globe. Advocates of the policy mobilities approach distinguish it in a number of ways from what they see as ‘orthodox’ policy transfer literature. While they accept that previous policy transfer literature has provided a useful focus upon key policy actors, institutions and practices involved in cross-national policy flows, they contend that it remains flawed in a number of ways. Perhaps most importantly, the conceptualization of the policy process in ‘orthodox’ studies is argued to be underpinned by individualistic rational-choice assumptions, in which ‘policymakers are maximizers (or at least boundedly rational in their behavior) and that there is a tendency for good policies to drive out bad, in a process of optimizing diffusion’ (Peck and Theodore, 2010: 169). By contrast, the mobilities approach views policy formation as a socially constructed process that involves an ‘intrinsic politics’, that policy actors are not ‘lone learners’ but rather are members of wider epistemic and practice communities (Peck, 2011). Because political science regularly focuses on national state institutions it is held to pay ‘too little attention to agency and the process of policy mobilization and the wider contexts that shape and mediate the agency of various policy actors’ (McCann and Ward, 2013: 6, emphasis added). Orthodox political science accounts are also criticized for presenting the process of policy transfer as one of linear replication rather than non-linear reproduction. There is, it is argued, a tendency for literalist notions of transfer to imply that the policies being moved are fully formed and ‘off the shelf’ rather than being diffuse, fragile and emergent (McCann and Ward, 2013). Finally, the essential ‘spatiality’ of policy-making means that mobile policies should not be seen as travelling across an inert policy landscape, but rather as re-making the terrain as they travel (Peck, 2011; Tenemos and McCann, 2013). In short, interpretivist critics suggest that the orthodox policy transfer literature engages insufficiently with the full range of social, spatial and political elements that are involved when policies travel across national boundaries.
Implications for the study of penal policy mobilities
Notwithstanding its hitherto rather marginal position within criminology, 3 together with the far-reaching constructivist critique of recent times, we think that the study of the movement of policy still has a great deal to offer those interested in the comparative study of penal change. First, whatever one’s position on the critique of ‘orthodox’ notions of policy transfer, critical policy studies scholars have drawn attention to international public policy flows as an important empirical phenomenon. Put quite simply, there appears to be a lot of it about. This alone is a compelling reason why such policy flows should not be overlooked by scholars interested in understanding and explaining the trajectory of crime control policy. Second, we noted earlier the tendency of influential bodies of criminological work to privilege in their analytical gaze the global, national or local levels respectively. The study of international policy flows, by contrast, can begin to illuminate the complex multi-scalar and multi-directional influences over contemporary penal policy-making (see also Edwards, 2016).
That said, to what extent, and in what ways, might the criticisms made of traditional policy transfer approaches affect future work focusing on the mobility of penal policy? First and foremost, the constructivist critique makes clear the necessity of some reappraisal of the terminology used in such scholarship. More particularly, and directly as a consequence of the new critical approaches, both the terms ‘mobilities’ and ‘assemblages’ are ones which seem to us to provide a conceptual advance by pointing to the need for more subtle tools for the analysis of policy development and movement than offered in the early policy transfer literature. Second, while the constructivist adoption of the term ‘assemblage’ helpfully draws attention to the need to treat ‘policy’ as emergent, processual and dynamic, we believe that there is further work to be done if the complexity of policy is to be adequately grasped. We take each of these points in turn.
Mobilities and assemblages
We do not wish to overstate the substantive differences between constructivist approaches and the political science tradition or dismiss the merits of earlier studies of criminal justice policy transfers. Indeed, later work by political scientists (see, for example, Dolowitz and Marsh, 2000; Evans and Davies, 1999) acknowledges the conceptual shortcomings of earlier research and is particularly critical of this literature for its overly rationalist account of the policy formation process (Legrand, 2016; Marsh and Evans, 2013). Furthermore, most policy transfer research is process-oriented (as opposed to outcome-oriented) and the majority of political scientists, regardless of their ontological and methodological dispositions, recognize that policy-making processes are inherently complex. However, the term ‘transfer’ does imply a flatness that belies such complexities. We acknowledge therefore that the concepts of ‘policy mobilities’ and ‘policy assemblages’ are likely to be better suited to capturing the empirical realities of the movement of policy within and across jurisdictions.
Critical approaches have much to offer conceptually, not least in focusing attention on the intricacies and malleability of policy movement. They also highlight the important role that agents and institutions situated above, below and beyond the nation state play in constructing policy meaning and content. In addition, the constructivist critique prompts consideration of what is meant by ‘policy’, though as we shall argue below, such work could go further in its deconstruction of such terminology. Constructivists argue that inquiries be centred on ‘policy assemblages’, a concept they argue is better suited to the study of phenomena that are ‘always in the process of coming together and being territorialized just [as they are] always also potentially pulling apart and being de-territorialised’ (McCann and Ward, 2012, 328, emphasis in original). The term itself originates in the work of the Gilles Deleuze who used it to denote a ‘multiplicity of heterogenous objects, whose unity comes solely from the fact that these items function together, that they “work together” as a functional entity’ (Patton, 1994: 158), though the term has been deployed inconsistently in contemporary scholarship (see, for example, Haggerty and Ericson, 2000; Rabinow, 2003; Schuilenburg, 2015). Nevertheless, in the context of policy mobilities, the term ‘assemblage’ conveys both a necessary spatial awareness and an interest in the dynamics of policy analysis, focusing on ‘the process of arranging, organizing, fitting together… [where] an assemblage is a whole of some sort that expresses some identity and claims a territory’ (Wise, 2005; quoted in McCann and Ward, 2013: 8, emphasis added). It consequently draws attention both to spatiality and to scale—policies being matters that neither emerge fully formed nor move intact. It further highlights the fact that policies are inherently unstable and prompts consideration of a wider range of sites within and beyond the state—together with the state itself—in any analysis of policy developments.
The historical spread of so-called ‘three strikes and you’re out’ sentencing (TSS) offers one illustration of the enhanced understanding that notions of ‘mobilities’ and ‘assemblages’ offer. In particular, the case demonstrates the pitfalls in interpreting the apparent spread of ‘TSS-related’ ideas and terminology across various parts of the globe as a relatively straightforward ‘export’ or ‘import’ of a set of concrete sentencing practices. According to many observers TSS’s emergence is associated with a series of legislative changes occurring in a number of US states in the early 1990s and, most particularly in California (Austin et al., 2000; Zimring et al., 2001). Though captured by the use of the ‘three strikes’ metaphor, the focus in fact is a wide variety of mandatory sentencing provisions that often had little in common beyond their mandatory component (Shichor and Sechrest, 1996). Indeed, ‘three strikes’, rather like its populist punitive cousin, ‘zero tolerance policing’, has tended to be deployed as a political symbol denoting toughness rather an obviously concrete set of penal practices (Jones and Newburn, 2006).
Following the spread of TSS laws across much of the USA, a number of other jurisdictions followed suit, passing their own so-called ‘three strikes’ laws. In England and Wales, a Conservative government passed the Crime (Sentences) Act 1997, which included three sets of mandatory provisions: an automatic life sentence for a second serious sexual or violent offence; a minimum seven-year prison sentence for third-time ‘trafficking’ in Class A drugs; and, a minimum three-year sentence for third-time domestic burglary (Jones and Newburn, 2002). In Western Australia in 1996, amendments to the Criminal Code (WA) meant that an adult or juvenile convicted for a third time or more for a domestic burglary would receive a minimum of a 12-month prison term (or period of detention for juveniles). A year later, in 1997, the Northern Territory amended its Sentencing Act 1995 so that a first-time offender convicted of a limited range of property offences would receive a minimum term of imprisonment of 14 days (Roche, 1999). This was repealed in 2001. In 2012 Queensland passed the Criminal Law (Two Strike Child Sex Offenders) Amendment Act, which mandated life imprisonment with a 20-year non-parole period, for repeat serious child sex offenders. In New Zealand, following a national campaign led by a victim’s advocacy charity known as the Sensible Sentencing Trust (SST), the government adopted its own version of ‘three strikes’ sentencing legislation in 2010 (Sentencing and Parole Reform Act 2010; see Oleson, 2015). Seeking to differentiate itself from the original Californian approach, described by SST as ‘too tough’, New Zealand’s legislation only applies the third-strike provision to ‘serious violent or sexual offences, not minor crime’. It further mandates the maximum prescribed penalty for a third serious offence as opposed to ‘a blanket 25 years to life’ (Sensible Sentencing Trust, n.d.). The following year, the rhetoric of ‘three strikes’ also made an appearance in New Zealand’s Copyright Infringing File Sharing Act (2011) that was introduced as a deterrent to online piracy.
It seems clear that there were a number of structural and cultural conditions in the USA, the UK, Australia and New Zealand that have rendered the promotion of ‘get tough’ penal policies more probable. More particularly, in each of these jurisdictions there was a general political climate that provided fertile ground for the promotion and adoption of certain styles of sentencing policy. It is clear that TSS policy originated in the USA, and the fact that the terminology subsequently travelled to a number of other jurisdictions indicates the possibility of at least a degree of emulation. But analysis suggests that influence was limited and indirect. The differing legal, political and cultural circumstances in those jurisdictions involved in emulation meant that the legislation that emerged in each case was quite distinctive. In both Western Australia and the Northern Territory, for example, the mandatory minimums targeted juveniles in particular (Bayes, 1999), and were quite unlike their British counterpart, or indeed many of the mandatory sentencing laws enacted in the USA. Both within the USA and outside, such mandatory minimums varied in a number of ways, the most important being: the offences to which mandatory sentences applied; the number of previous offences necessary before the mandatory sentence was imposed; and the penalties imposed as a consequence of the mandatory minimum.
Our point, simply, is that these and other differences reflect the variable nature of the policy assemblages which came together to shape the nature and timing of the local manifestations of these ideas. The United States, for example, had a longer and more varied history of existing legislative provisions for dealing with repeat offenders than was the case in either Australia or England, and the existence of such provisions clearly affected the nature of the eventual three strikes laws that were passed (Clark et al., 1997). The process by which three strikes laws came into being varied very significantly, from ballot initiatives in the first examples in Washington State in 1993 and California in 1994, to more straightforward parliamentary legislative procedures in England and the two Australian states.
The political circumstances in which such laws were passed also varied considerably. In some US states, including Washington State and California, three strikes law were enacted in the aftermath of, and were much influenced by, moral entrepreneurial and media campaigns following high profile murders committed by repeat felons. Western Australia’s three strikes burglary laws were introduced in the run-up to a State election early in 1997 (Morgan, 1999) and in England, the new mandatory minimums were in part a product of an attempt by an incumbent government to hold its increasingly popular, law and order-focused opponents, at bay, also in the run-up to an election.
Studies drawing on the ‘policy transfer’ approach can of course begin to explore the ways in which particular policy ideas and practices mutate within contrasting local socio-political contexts (Jones and Newburn, 2007). But it does seem that research within the ‘policy mobilities’ approach can provide particularly helpful insights into the messy realities of policy-making on the ground, by their emphasis on ‘the simultaneous making and moving of policy to develop an analysis that is both global and local, whilst also being close to practice’ (McCann and Ward, 2015: 828). Whatever TSS and its variants might be taken to be, it is clear that it has never been either a stable or unchanging entity. Even in its home setting it refers to a complex array of policies and practices that were, at least to a degree, quite fluid. Such fluidity is only exacerbated when the focus turns to how such ideas, and the practical initiatives to which they give rise, move across time and space. Peck and Theodore’s (2012) call for ‘cross case and cross-conjectural’ theorization in order to understand the uneven and variegated emergence of policy forms, seems apposite here. Moreover, the example of New Zealand’s Copyright Infringing File Sharing Act (2011) also highlights the prospect of studying policy assemblages that span different policy spheres.
While there is still work to be done to fully specify the key dimensions conveyed by the term ‘policy assemblage’, such a notion would appear to overlap in some important and useful ways with the idea of a ‘policy network’ (see Dowding, 1995; Marsh and Smith, 2000). Both denote scepticism about the overly ‘state-centric’ approaches to understanding policy formation, and highlight the need to focus empirical attention on the dynamics of interaction and exchange between a plurality of policy actors in a range of ‘sites of governance’. Though the idea of ‘policy networks’ has been utilized in a number of ways in political science, it is generally deployed as a meso-level concept enabling micro-levels of analysis (such as the interactions of a diverse array of state and non-state bodies in relation to particular policy decisions) to be linked with macro-levels (including the societal structures of power distribution and ‘global’ social forces). In saying this, we are not seeking to suggest that ‘assemblage’ approaches are simply network-oriented political science in disguise. Indeed, we acknowledge some of the key criticisms levelled at policy network studies in political science, not least that too often they appear to be more interested in producing detailed descriptive typologies than illuminating the actual mechanics via which they shape policy in particular contexts. Much of the work has been based on attempts to quantify key variables of proposed networks and their impact. It is clear that this approach to studying policy networks cannot capture the complex ways in which policy meaning is negotiated dynamically in different sites of power. But qualitative studies of policy networks do offer reasonably transparent methods of mapping systematically the core components of ‘policy assemblages’. Accordingly, we argue that representing the actors and institutions that collectively ‘assemble’ policy meaning and content along with the sites or mediums in which this occurs (albeit crudely) is a required first step for producing a richer understanding of their inner mechanics. Finally, in this regard, we note that to date there has been little scholarly interest in conceptual or empirical analyses of the notion of policy process in crime control. By its very nature, policy mobilities approaches require acknowledgement of the emergent nature of policy and an exploration of influences over its trajectories that transcend its immediate geographical situation.
Policy levels
The critical, constructivist position, as we have argued, offers a series of important reminders both about the complexity of mobility/ies, and about the multi-layered and dynamic nature of ‘policy’. Nevertheless, as we alluded to earlier, notwithstanding the subtleties potentially introduced by the idea of ‘assemblages’, or similar, to our eyes it remains the case that there is more to be done to problematize the very idea of ‘policy’, and then operationalize these elements for the purposes of empirical study. In saying this, we are not suggesting that this is not recognized in the constructivist literature, rather that it still requires some elucidation, particularly if systematic empirical research is to be effectively undertaken in this field.
One basic starting point is to return to Jones and Newburn’s (2007) threefold distinction between policy ideas, symbols and rhetoric; policy content and instruments; and the implementation of policy. These dimensions of ‘policy’ correspond broadly to those outlined by Pollitt (2001), who argued that the complexity of public policy necessitates the analytical distinction between distinct levels of ‘talk’, ‘decisions’ and ‘action’. There is considerable evidence from the policy transfer/mobilities literature that it is rhetoric, talk, knowledge and symbols that travel most easily. Work on the diffusion of ‘social problems’ offers many examples of the ways in which claims, rhetoric and tactics have more or less successfully diffused (Best, 2001). Even the spread of ‘claims’ or rhetoric, however, is far from straightforward, the mobility of ideas facing very considerable cultural, practical and political barriers (Jenkins, 1992). This level of ‘policy’ must be distinguished from the more substantive elements, for it is important, as Brunsson (1989: 231) has noted, not to ‘make the mistake of supposing that organizational statements and decisions agree with organizational actions’. Official rhetoric about restorative justice, as just one example, has often centred on claims about victim and community involvement in the criminal justice process. In practice, however, when acted upon, the initiatives introduced vary very significantly indeed in the extent to which, and the ways in which, victim and community representation is imagined in legislation and policy documents (Daly, 2004).
The tendency within much criminological research when discussing ‘policy’ has been to focus attention at the level of documents, instruments and legislation. These formal indicators of policy offer a solid, substantive basis for the analysis of trends, changes and ‘mobilities’ and, indeed, authors such as Bernstein and Cashore (2000) have argued that empirical studies need to focus upon formal policy decisions such as statutes, regulations and statements, because these manifestations of policy capture the ‘actual choices of government’. Important though such work is, even in combination with an analysis of knowledge, claims and rhetoric, it still offers an incomplete picture of the nature of ‘policy’ for there are often very considerable gaps between ‘policy on the books’ and ‘policy in practice’. Even in circumstances where policy agendas (Kingdon, 1995) and the chosen formal solutions to them appear similar, what occurs in practice—policy as implemented—will most usually continue to diverge in important ways.
In our view, while it might therefore be argued that the notion of an ‘assemblage’ potentially encompasses all three policy ‘levels’ identified here, without spelling this out there is a danger that future work in this field will replicate one of the key shortcomings of much extant work, and conflate (or at least, fail to distinguish sufficiently clearly between) these analytically separate aspects of policy. Such a failure almost inevitably leads to misconceived conclusions about the trajectory, nature or impact of particular policy initiatives and, more particularly, to simplistic assumptions about the mobility of ‘policy’.
Conclusion
Comparative studies of penal change within an increasingly ‘globalized’ world have provided a range of different analyses that together have enriched our understanding of the factors shaping policy trajectories. A broadly structuralist sociology has tended to look to deep-lying global social processes as the basis for explaining the trends observed. More recently, and partly as a result of a desire to understand the continuing differences between jurisdictions, greater attention has begun to focus on the role of political, economic and other institutional formations of national polities on penal outcomes. These bodies of work sit alongside comparative scholarship that emphasizes the unique and embedded features of penal policy developments in local contexts. To date, these different bodies of work have tended, implicitly or explicitly, to downplay the importance of cross-national policy flows as a factor in understanding and explaining shifts in crime control. In addition, research on policy formation and development, policy change and policy implementation arguably remains, if not in its infancy, still not long past adolescence in terms of its development within criminology (Sherman and Strang, 2010).
For two main reasons, we think, policy mobilities research offers an important lens through which such work—especially comparative research—can be undertaken. First, critical policy studies scholars—albeit mainly focusing on areas of public policy other than crime control—have demonstrated that cross-national policy flows are an important empirical phenomenon in shaping policy trajectories in different jurisdictions. Their work suggests that public policies are circulating around the globe with increasing frequency and rapidity, and this in itself is an important reason for taking this phenomenon seriously in the sphere of penal policy. Second, the study of global policy mobility offers an analytical approach that—rather than privileging one level of analysis—begins to help capture the complex inter-connections between the local, national and global in a world characterized by increasingly multi-scalar and multi-directional circuits of influence. The term ‘policy mobilities’ is better suited to the analysis of the non-linear movements of penal policies and practices than the somewhat less agile idea of ‘transfer’. We also agree that the notion of a ‘policy assemblage’ is useful for conceptualizing the transformative processes that characterize these movements and through which policy meaning and content are negotiated. It further challenges the notion of a policy as an easily discernible and static entity and calls for more critical engagement with the variegated nature of ‘policy’ and the multiple levels at which it might be argued to exist. McCann and Ward (2012: 325) rightly observed that ‘[t]here is a lot left to understand about how, why, where and with what consequences policies circulate globally’. Criminology, being what Downes referred to as a ‘rendezvous’ subject, is ideally placed to draw valuable insights from a range of disciplines, and to utilize the strengths of both the political scientific and constructivist approaches to the study of policy mobility in its continuing quest to understand the dynamics of comparative penality.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.