CRIMINOLOGICAL POLICY MOBILITIES AND SEX WORK: UNDERSTANDING THE MOVEMENT OF THE ‘SWEDISH MODEL’ TO NORTHERN IRELAND

Northumbria University has developed Northumbria Research Link (NRL) to enable users to access the University’s research output. Copyright © and moral rights for items on NRL are retained by the individual author(s) and/or other copyright owners. Single copies of full items can be reproduced, displayed or performed, and given to third parties in any format or medium for personal research or study, educational, or not-for-profit purposes without prior permission or charge, provided the authors, title and full bibliographic details are given, as well as a hyperlink and/or URL to the original metadata page. The content must not be changed in any way. Full items must not be sold commercially in any format or medium without formal permission of the copyright holder. The full policy is available online: http://nrl.northumbria.ac.uk/pol  i cies.html


Introduction
Between 2012 and 2014, Maurice Morrow, a Democratic Unionist Party (DUP) peer, led a campaign to criminalize the buyers of sexual services in Northern Ireland. An important and intriguing element of Morrow's campaign was his regular use of Swedenthe first country to criminalize those buying but not selling sex-as a reference point. Morrow, like many of his supporters, would often talk of Sweden as a source of inspiration and a place that must be learnt from and emulated. Even after the Assembly voted to introduce Morrow's plan in October 2014-as part of a wider package of measures targeting sex work and human trafficking-Morrow was still keen to stress the links to Sweden. For instance, several weeks after the vote, he stated in a speech to the Northern Ireland Assembly that: [T]he evidence clearly suggests that the approach modelled by Sweden is the best available. It recognises the abuses involved in the prostitution industry and seeks to reduce the core driver for prostitution-the demand for paid sex. (Northern Ireland Assembly 2014a: n.p.) Later that day, Morrow reasoned that having visited criminal justice officials and policymakers in Sweden a year prior he was confident of its success in Northern Ireland. 'If it is as effective in Northern Ireland as the equivalent is in Sweden', Morrow argued, 'the Assembly will have done everyone a favour, particularly victims of human trafficking' (Northern Ireland Assembly 2014a: n.p.). Morrow, of course, is not the only person to have campaigned for the 'Swedish model' to be emulated in other countries.
In fact, several other jurisdictions-including Norway, Iceland, France, Canada and the Republic of Ireland-have all followed Sweden in criminalizing the buying of sex without criminalizing the selling of sex. So while many worldwide have criticized the Swedish model (e.g. Levy and Jakobsson 2014;Scoular 2004;Vuolajärvi 2018), it has been heavily promoted and highly influential.
By exploring the movement and mutation of the Swedish model into Northern Ireland, the paper has two purposes. The first purpose is that it illustrates and enhances some of the ideas mooted in a recent article in Theoretical Criminology by Newburn et al. (2017) who call for a re-imagining of policy transfer within criminological research. In short, they make the case for drawing ideas from the geographical literature on policy mobilities. We strongly support this but also argue that there are other ideas in the policy mobilities literature, not mentioned by Newburn et al. (2017), which can enhance our understanding of the mobilization of criminal justice ideas and policies. In particular, we make the case for a focus within criminological policy mobilities studies on learning, educating and what we call extrospective showcasing.
The second purpose is that it provides an insight into the so-far unexplored international circulation of sex work policy models. This is important because debates about sex work and its regulation in different parts of the world often reference other places and mobile models. It is not just Sweden or the Swedish model that is widely discussed; Aotearoa New Zealand, for instance, is regularly framed as a model for decriminalization, while the Netherlands and Nevada are positioned as models for legalization. All these places-and others-are strategically used in debates; they are held up as being successful or unsuccessful, good practice or bad practice, and policies that should or should not be emulated. So, given the prevalence and implications of drawing on examples from elsewhere in sex work policymaking and advocacy, it is somewhat perplexing why the sex work literature has not closely examined the transnational movement of sex work policy models. In this article, we begin to address this gap.
The structure of this article is as follows. It begins by considering the ways in which criminological work has sought to understand the movement of policies, ideas and models, and then it draws on the policy mobilities literature to set up a framework for understanding this phenomenon. This is followed by a brief methodological note, after which there is an in-depth exploration of the adaptation of the Swedish model by policymakers in Northern Ireland. Let us turn our attention now to the criminological literature.

Towards criminological policy mobilities
As noted previously, a small and varied body of criminological research has steadily emerged providing insights into the movement of selected policies, ideas and models. These include probation (Canton 2006;Vanstone 2008;McFarlane and Canton 2014), victim impact statements (Wemmers 2005), preventive orders (Ogg 2015), football banning orders (Hamilton-Smith and Hopkins 2012), electronic tagging (Nellis 2000;Jones and Newburn 2007) and policies targeting violence against women (Walklate and Fitz-Gibbon 2018). Early work on the movement of policies from the United States to the United Kingdom (Nellis 2000;Jones and Newburn 2007) has been complimented by studies exploring a variety of places and routes of travel. Historically, the work has concentrated on the movement of policies in recent decades, although Vanstone (2008) provides an exception with their exploration of the international emergence and circulation of probation in the late 19th century and early 20th century. There are of course gaps in this literature, not least on the issue of sex work and on the role of the Nordic countries as 'sites of policy import and export' (Geddie 2015: 242).
Criminologists, of course, are not the only academics interested in the movement of policies. There is a long history within political science of research into policy diffusion, policy learning and, most noticeably, policy transfer (e.g. Dolowitz and Marsh 2000;Benson and Jordan 2011). Scholars outside of political science have also explored such issues with, most noticeably, a large body of work in and around human geography emerging since the mid-2000s on what McCann (2008) has coined 'policy mobilities ' (e.g. McCann and Ward 2013;Peck and Theodore 2015;Temenos et al. In press). There are several links between the criminological and political science literature on mobile policies, the most noticeable is the shared use of the term 'policy transfer', a term developed by the political scientists Dolowitz and Marsh (2000: 5) who define it as 'the process by which knowledge about policies, administrative arrangements, institutions and ideas in one political system (past or present) is used in the development of policies, administrative arrangements, institutions and ideas in another political system'.
While a significant number of the criminological articles on policy transfer are empirically focused and cite little or none of the political science policy transfer literature, a smaller number of criminological studies have drawn explicitly and substantially on the conceptual tools developed by Dolowitz and Marsh as well as other political scientists (e.g. Jones and Newburn 2002;2007;Canton 2006;Ogg 2015). An important example is Jones and Newburn's (2007) monograph Policy Transfer and Criminal Justice that explores criminal justice policy transfer from the United States to the United Kingdom, critically examining the apparent movement of 'zero tolerance' policing, two and three strike sentencing, electronic tagging and the privatization of prisons. Jones and Newburn draw especially on the work of Evans and Davies (1999) and Smith (2004) to identify whether policy transfer has occurred or not, as well as Dolowitz and Marsh (2000) to consider what is transferred, where lessons are drawn from, what degrees of transfer occur, what constrains or facilitates the transfer, the success or failure of the transfer, who is involved in transfer and why they get involved.
A decade after the publication of Policy Transfer and Criminal Justice, Jones and Newburn have teamed up with Jarrett Blaustein to advocate a noticeably different way of viewing the movement of criminal justice policies (Newburn et al. 2017). Their inspiration this time is not the political science work on policy transfer; instead, it is the geography-led work on policy mobilities. For them, a policy mobilities approach is more nuanced as it avoids the literalism of the 'policy transfer' label, embracing at the same time the geographical messiness and social construction of policy formation (cf. Dolowitz and Marsh 2012;Marsh and Evans 2012). Newburn et al. (2017) usefully single out two aspects of the policy mobilities approach that can enhance our understanding of travelling criminal justice policies (and are downplayed in the political science and criminological policy transfer literature). The first is the focus on mobilities. Indeed, the policy mobilities literature focuses on the qualitative mobility and immobility of policy knowledge. As part of this, it repeatedly stresses that policy knowledge mutates as it moves and that it is influenced by and influencing the landscapes through which it travels (Peck and Theodore 2015). Here the policy mobilities literature takes its cues from the wider mobilities literature which focuses on the qualitative experiences and ramifications of mobility and immobility (see, for instance, the collection edited by Adey et al. 2014). The second aspect of the policy mobilities highlighted by Newburn et al. (2017) is the concept of assemblage, which was first developed by Deleuze and Guattari (1987) and points to the process of temporarily bringing something-such as a policy-into coherence. While it is not used in all policy mobilities accounts, the concept of assemblage offers a useful lens through which to view policy mobilization (Prince 2010;McCann 2011;Baker and McGuirk 2017). This is illustrated by McCann and Ward (2013: 8) when they say: Policies … are not only local constructions; neither are they entirely extra-local impositions on a locality. Rather, policies and governance practices are gatherings, or relational assemblages of elements and resources-fixed and mobile pieces of expertise, regulation, institutional capacities, etc.-from close by and far away. They are assembled in particular ways and for particular interests and purposes … This concept [of assemblage] is helpful as a frame for policy studies because it emphasizes … that policies are not internally coherent, stable 'things' but must be understood as social processes.
For Newburn et al. (2017), the policy mobilities approach with its emphasis on mobilities and assemblages can be enhanced through a careful consideration of what they call policy levels. Expanding on ideas developed a decade earlier in Jones and Newburn (2007), Newburn et al. (2017) argue that there needs to be clarity in what we mean by policy and what exactly is and is not being mobilized and assembled (see also Lovell 2017a). For them, it is important to differentiate the softer elements of policy in terms of (1) policy ideas, symbols and rhetoric from (2) 'the more concrete manifestations of policy in terms of policy content and instruments' as well as (3) 'the more practical applications of policy in terms of its implementation by practitioners and professionals' (Jones and Newburn 2007: 23, emphasis in original). We concur with Newburn et al. that attention to mobilities, assemblage and policy levels are particularly useful in understanding criminological policy mobilities. Nevertheless, Newburn et al.'s (2017) favourable, amended and brief interpretation of the policy mobilities approach overlooks two issues that are being grappled within the policy mobilities literature, which would provide a useful supplement to the focus on mobilities, assemblages, policy levels and policy mobilization agents. The first is learning and educating, and the second is extrospective showcasing.
The circulation of policy knowledge is intimately connected with, and dependent on, forms of learning and educating. By learning, we echo Dunlop (2009: 296) in viewing it as 'a knowledge acquisition process' and by educating we use de Oliveira and Ahenakew's (2013: 233) definition of it as 'the steering of learning towards particular desirable ends, which are defined differently in different societies, cultures and contexts'. Learning and educating are every-day, power-infused practices (McFarlane 2011). They take multiple forms, take place in person or at a distance, occur at various sites and occasions (formal and informal) and shape the collection, interpretation and use of mobile policy knowledge.
An important concept within the policy mobilities literature that speaks to the relationship between learning, educating and policy mobilization is informational infrastructures. These, McCann (2008: 12) notes, are assemblages of institutions, events and technologies that 'frame and package knowledge about best policy practices, successful cities, and cutting-edge ideas and then present that information to specific audiences'.
Here McCann's definition of informational infrastructures strongly echoes de Oliveira and Ahenakew's (2013) understanding of educating. Informational infrastructures are therefore focused on educating audiences such as policymakers and shaping their learning. It is important, then, to think about the role of informational infrastructures and how they influence policy mobilization. A central task in exploring the relationship between policy mobilities, learning and educating is to consider how learning and education are performed at informational infrastructure events such as study tours, conferences and award ceremonies and with what effects, something that has received attention within the existing policy mobilities literature (Cook and Ward 2012;Wood 2014;Cook et al. 2015;Temenos 2016;Montero 2017).
Turning to extrospective showcasing, this term speaks to the ways in which local policy actors showcase their work and locality to audiences elsewhere, often drawing on wider informational infrastructures to do so. As McCann (2013) notes, policymakers often act extrospectively through drawing ideas from elsewhere and, just as importantly, actively promoting their policies to audiences elsewhere. Such promotional practices are akin to education and would include, for instance, writing about their policies in trade magazines or newspapers, speaking at conferences, providing materials for exhibitions and hosting study tours (Cook 2018). Work by McCann (2013) and Cook (2018) encourages us to critically examine the 'supply-side' of mobile policy knowledge and challenges us to consider how and why local policy actors engage in extrospective showcasing, how such practices are embedded within wider informational infrastructures and how they influence policy mobilization.
In summary, then, the policy mobilities approach can facilitate a nuanced understanding of the geographical circulation of criminal justice policy knowledge and especially sex work policy models. We will demonstrate our argument through an analysis of the movement of the Swedish model to Northern Ireland. As a part of this, we will pay close attention to five aspects of policy mobilization: (1) mobilities, (2) assemblages, (3) policy levels, (4) learning and educating and (5) extrospective showcasing. First though, we will provide a brief reflection on the methodology.  (2012) have suggested a useful methodological approach in studying the complexities of policy mobilization: the 'following' of policy mobility. Here, McCann and Ward speak of the need to follow policies, people and places as they travel as well as the requirement to closely examine the connective sites through which policies are mobilized and mutated. Such a methodological focus on mobility, of course, needs to run alongside an emphasis on the less-mobile and more inward-looking aspects of policy formation (Lovell 2017b;Weller 2017), in addition to a pragmatic acknowledgement that access restrictions, time demands and spiralling costs can prohibit all-compassing following.

Methodology
In 'following' the Swedish model to Northern Ireland, the research project utilized two key methods. The first is semi-structured interviewing with those involved in policymaking, advocacy and the mobilization of the Swedish model in Northern Ireland CRIMINOLOGICAL POLICY MOBILITIES AND SEX WORK (n = 7) and in Sweden (n = 5). The interviews were conducted either in person or via telephone or Skype, and the interviewees have been anonymized in this article. The second is narrative analysis of a variety of texts. These include newspaper articles, governmental and non-governmental policy documents, advocacy reports, information packs, conference presentations, academic literature and social media feeds. Particularly, important sources were the Hansard reports of proceedings in the Northern Ireland Assembly as well as texts (such as the meeting minutes and written evidence) relating to the Committee Stage through which Morrow's Bill-the Human Trafficking and Exploitation (Further Provisions and Support for Victims) Bill-had to pass through. As with the interview transcripts, narrative analysis was used to reveal the arrangement of these textual 'stories' and the characterization of policies, people and places within them (Moore 2014). By using a combination of semi-structured interviews and texts, it is possible to offer an insight-albeit a partial one-into the mobile and less-mobile policies, people and places involved as well as some of the important connective sites through which policy mobilization was shaped, such as the public debates and meetings in the Parliament Buildings in Belfast known as Stormont.

Introducing the Swedish model
On 1 January 1999, it became illegal to buy sex in Sweden. Initially, it was an offence punishable by a fine or up to six months' imprisonment with the maximum sentence revised to one year in 2010. Those selling sex were not criminalized. Prior to the legislative change in 1999, there were no laws in place criminalizing on-street sex work with off-street sex work subject to few criminal laws (Hubbard et al. 2008). The 'sex purchase law' (sexköpslagen) was brought in as part of the Women's Peace (Kvinnofrid) Bill voted in under the Social Democratic government (see Holmström and Skilbrei 2017 for an overview of its introduction). As well as reflecting some of the paternalistic underpinnings of Swedish welfarism, sexköpslagen was also influenced by radical feminism within Sweden (Scoular 2004). Echoing radical feminist ideas (see, for instance, Jeffreys 1997; Raymond 2013), prostitution has been positioned in Sweden as an inherently violent form of patriarchal oppression that, irrespective of the circumstances in which it takes place, is harmful to women both inside and outside the sex industry (Ekberg 2004). It was argued that sexköpslagen was introduced not only to abolish prostitution but also to improve gender equality in Sweden. To achieve both goals, it is believed that the law must target those who are deemed to possess power and choice: the clients as well as those profiting from the sex industry. Sex workers would not be criminalized but offered support and advice on how to exit the industry. As Levy (2015) describes in detail, much of this support has been funnelled through 'prostitution units' in the three major cities in Sweden: Stockholm, Malmö and Gothenburg. These units are funded by local councils and provide a variety of services for sex workers often focused around issues such as welfare, alternative employment and health. Notably though, Malmö focus on harm reduction, while Stockholm and Gothenburg focus on facilitating service users to leave sex work (Levy 2015 2015, it became a crime punishable by a fine and potentially up to one year's imprisonment. This replaced existing laws criminalizing kerb crawling and the purchase of sex from someone subject to force. It also repealed the acts of soliciting and loitering, while keeping a brothel used for prostitution would remain illegal. Section 19 of the 2015 Act also required a programme of support for those exiting prostitution. As we will detail, the measures under the 2015 Act and the rationale for them were clearly informed by Sweden, but they have mutated as the policy was mobilized and implemented in Northern Ireland. During the campaigns for and against the new laws in Northern Ireland, Sweden was positioned as a model: 'the Swedish model'. Indeed, the Swedish model has become a much-used term within debates on sex work worldwide. It is important to unpack this phrase. Drawing on the work of Peck and Theodore (2010), we should view the Swedish model, like all policy models, as a socially constructed, stripped-down abstraction of a messy and prosaic reality. We should avoid seeing the Swedish model as one that has only been 'made in Sweden' and one whose component parts are fixed; instead, it is coproduced by numerous actors in different places and it mutates as it moves.
A central aspect of the Swedish model is, to use Peck and Theodore's (2010) words, its metonymic tagging of a policy to a place-a metonym being a linguistic device 'where a name is used for something with which it is somehow culturally or spatially associated' (Brown 2006: 317). With its symbolic association to a particular nation state, the Swedish model 'evokes a grounded form of authenticity, implies feasibility, and signals an ideologically palatable origin story' (Peck and Theodore 2010: 170). The metonymic tagging of the Swedish model has become more complex as it has been emulated elsewhere, with policymakers and advocates also referencing the 'Nordic model'. The Nordic model is mostly used as a synonym for the Swedish model. Nevertheless, those who use the Nordic model terminology often do so to suggest a common approach in the Nordic countries; this is despite the fact that Denmark has not criminalized the purchase of sex and Finland has only partially criminalized this, with clients who purchase sex from trafficked victims or someone controlled by a pimp being targeted (see Skilbrei and Holmström 2013). The Nordic model terminology is used to demonstrate a transnational transferability of Swedish policies. The emulation of Sweden in Canada, France, the Republic of Ireland and Northern Ireland has further complicated this metonymic tagging. Despite all this, in Northern Ireland and elsewhere, Sweden remains the dominant reference point within the movement of the Swedish/Nordic model.

Showcasing, educating and learning
A sizable informational infrastructure has developed since the late 1990s that seeks to educate people in different parts of the world about what has become known as the Swedish model. Many of the educators have not been directly involved in the development and implementation of prostitution policy in Sweden. Furthermore, much of this education occurs at a distance, disseminated, for instance, via newspaper articles, journal articles, books, blogs and other social media. The experiences of Sweden have been commented on at length by Swedish and non-Swedish activists and journalists (e.g. Demsteader 2012; Meredith 2013). One such UK-based activist is Julie Bindel (2016: 1) who has argued that the 'Nordic model is the best way towards a society where women are not saleable objects'. Many Swedish and non-Swedish academics, too, have been vocal on the sex purchase law in Sweden. As with journalists and activists, the Swedish model has proved extremely divisive among academics; some have criticized the Swedish model (e.g. Levy and Jakobsson 2014;Vuolajärvi 2018) and some have been supportive of it (e.g. Raymond 2013; Coy 2016). Others, meanwhile, have broadly supported it while recommending changes to its modus operandi (e.g. Johnson and Mathews 2016). Those academics and activists who are vocal in their support of the Swedish model are often, but not always, linked to a wider transnational neo-abolitionist movement. This is a movement that includes but is not limited to radical feminism and seeks to abolish prostitution through the punishment of clients and third parties. As Ward and Wylie (2017) note, for many within the neo-abolitionist movement, the Swedish model is the exemplary model.
Of course, much of this education has also involved those working within or alongside the Swedish state. This is unsurprising given that the Swedish state has actively funded the external promotion of the Swedish model (Ward and Wylie 2017) and that at its inception the sex purchase law was intended to make a statement internationally. There is a desire from within the Swedish state to stop sex work occurring outside of Sweden; one interviewee in Sweden who promotes the model internationally reasoned that 'we do not want to have a world where it is OK to buy other people'. This extrospective showcasing has occurred not only at a distance but also face-to-face with curious onlookers from abroad. This has often involved Swedish representatives participating at informational infrastructure events such as conferences, workshops and study tours (we will return to the latter further on). There are numerous showcasers in Sweden who have regularly and proactively promoted the Swedish model to international audiences as well as responded to requests from curious policymakers and practitioners from abroad. Two of the most prominent Sweden-based showcasers are Kajsa Wahlberg (the Swedish National Rapporteur on Trafficking in Human Beings) and Simon Häggström (who works in the Prostitution Unit within Stockholm's Police). Much of Wahlberg and Häggström's showcasing focuses on their experiences of delivering the Swedish model; they can give the 'inside story' of how it works. This partly explains the international interest in Häggström's (2016)

book, Shadow's Law: The True Story of a Swedish Detective Inspector Fighting Prostitution.
It is a book that the American radical feminist Melissa Farley (2017: 1) has praised, calling it 'a much-needed tool for confronting prostitutionharm-denial and for making the links between prostitution, trafficking, and pornography'. In addition to Wahlberg and Häggström, another prominent Swedish showcaser is Per-Anders Sunesson, the Swedish Ambassador at Large for Combating Trafficking in Persons, who has spent considerable time travelling abroad in this capacity promoting the Swedish model, as his Twitter account @PASunesson illustrates.
Whether it is conducted at a distance or face-to-face, by those in Sweden or outside Sweden, most of the pro-Swedish model educators and their educating materials are adamant that the model is (1) successful, (2) morally right and (3) transferable. Although their claims have often been rebuked by critics, these positive and digestible messages have accompanied the Swedish model on its travels. The words and actions of one showcaser, Gunilla Ekberg, are important to note here-primarily because she played an important role in bringing the Swedish model to Northern Ireland. Ekberg worked as a Special Advisor on Trafficking in Human Beings in Sweden between 1998 and 2006. She continues to write academic articles and briefings on the model (e.g. Ekberg 2004;Ekberg and Werkman 2017). In short, she has become very closely associated with the Swedish model and a travelling advocate for it. In what Ellison (2017: 207) describes as 'one of the most unlikely pairings in recent political history', Ekberg-a radical feminist-worked for Morrow and the DUP as an advisor and helped to facilitate the adoption of the Swedish model in Northern Ireland. Among other things, Ekberg helped draft Morrow's Bill and gave oral evidence at length during a Committee Stage meeting at Stormont in September 2013. Ekberg's close links with the Swedish model encouraged Morrow to work with her. As one DUP Member of the Legislative Assembly (MLA) interviewed (#1) reasoned, Ekberg is 'an absolute world authority … on this issue; she has walked the walk, as they say, in legislation and studies [of] other countries, and we listened intently and closely to what Gunilla Ekberg was saying in relation to the Swedish model'.
In Ekberg's writing and public speaking-before, during and after her involvement in Northern Ireland-she has repeatedly stated that prostitution is in almost every case exploitative and abhorrent, dismissed the idea that many sex workers have the capacity to exercise choice and reasoned that sex work is inseparable from trafficking (Ekberg and Werkman 2017). Ekberg has also repeatedly praised the measures put in place in Sweden-and the sex purchase law in particular-which she argues has been a considerable success. Like many advocates for the Swedish model, Ekberg has frequently stressed that there has been a reduction in sex work and trafficking in Sweden. For instance, when Ekberg gave evidence at the Committee Stage of the Bill and was asked whether prostitution had been abolished in Sweden, Ekberg responded: No, of course not. Any social change takes more than 15 years, but what I can say with some conviction is that Sweden is a country where prostitution is much less prevalent and is a country that is not attractive for traffickers… [I]f you look at, for example, Finland, you see that they had 15,000 victims of trafficking per year, whereas we have maybe 200 or 300 at the most. Again, that is because it is not attractive. Think about it. If you were a trafficker, where would you go? Where would I go? I would not go to a place where you risk getting caught. (Northern Ireland Assembly 2013: n.p.) As we suggest later, the reported ability of the Swedish model to deter trafficking rather than its ability to reduce gender inequality has become its most seductive 'sales pitch' to audiences abroad, especially those in Northern Ireland. Ekberg was not the only person giving oral evidence at the Bill's Committee Stage who spoke about the success or otherwise of the Swedish model. In fact, the majority who gave evidence referenced Sweden and the impact of the sex purchase law. Their views were often justified by claims to have read reports about Sweden, visited Sweden or spoken to people from Sweden connected to sex work. Most of those giving oral evidence were from Northern Ireland with several from the Irish Republic and Britain; Ekberg, however, was the only person from Sweden.
Prior to receiving written and oral evidence, it was decided that six members of the Committee for Justice would attend a two-day study tour to Stockholm on Ekberg's recommendation. They were accompanied by Morrow who participated at his own expense (Committee for Justice 2013). Taking place in December 2013, the trip was an important means through which those in Northern Ireland learnt about Sweden. Its itinerary centred on the delegation meeting and hearing presentations from ten professionals whose work focuses on sex work or trafficking, including Kajsa Wahlberg, CRIMINOLOGICAL POLICY MOBILITIES AND SEX WORK Simon Häggström, the academic Petra Östergren and co-ordinator of the NGO Rose Alliance, Pye Jakobsson. The tour report makes it clear that while many of the Swedish 'hosts' presented policy and practice in Sweden as working successfully, some such as Östergren and Jakobsson spoke critically about it (Committee for Justice 2013). Echoing study tours elsewhere, this visit was a way of learning face-to-face from those deemed to have experience and expertise on the issue. For Morrow, the trip also offered him a means of convincing the Committee of Justice to support his proposals-some of whom had spoken publicly about their reservations about clause 6 which criminalized the buying of sex (clause 6 would later become clause 15). The trip would also be regularly namechecked by Morrow when speaking to the media and other audiences about his 'tried-and-tested' proposals.
For some of the other delegates, the tour confirmed the success and transferability of Morrow's proposal. One such delegate was the Chair of the Committee of Justice and DUP MLA, Paul Givan who in a Committee Stage meeting said: The Swedish police service said-and I have no reason to suggest why they would tell me different-that the deterrent value in clause 6 would reduce things by approximately half. That was the deterrent value that allowed them to put their resources into the harder cases. The deterrent value would reduce it by about half. Why would we not have that same kind of deterrent value? (Northern Ireland Assembly 2014b: n.p.) Others delegates, however, took more critical lessons from the visit to Stockholm. Sinn Féin MLA Rosaleen McCorley, for instance, reasoned at another Committee Stage meeting: I was in Sweden and was told by people who work in the sex industry there that they have seen no significant reduction. There is also evidence that it has increased. In fact, there is Eurostat evidence to suggest that convictions for trafficking in Sweden have quadrupled and that trafficking is increasing more there than in other countries in the area. What the women said-this is very concerning-was that life had got more dangerous and that they felt more stigmatised. (Northern Ireland Assembly 2014c: n.p.) The visit to Sweden would go on to shape the ways in which Morrow's Bill was framed and how it was received by policymakers. The trip, however, was not a 'tick-box exercise' as it did offer an opportunity for some critical voices in Sweden to speak directly to the delegation. That said, important questions have been asked about the lack of interest by Morrow and his supporters in learning from academic research on sex work in Northern Ireland, particularly research that presents empirical data that runs contrary to the image of sex work in Northern Ireland as sizeable, exploitative and inherently linked to trafficking (e.g. Huschke et al. 2014;Ellison 2015).
One such author, Graham Ellison (2017)-who submitted written and oral evidence to the Committee-reasoned that the DUP saw little interest in engaging with such research. 'Empirical evidence from research into commercial sex', Ellison argues, 'was either ignored or treated as inferior to that conducted by a number of advocacy groups which was based normatively on feelings, emotions and particularistic moral stances' (Ellison 2017: 309). Academic research was met with resistance from some Committee members such as Paul Givan who reasoned that 'some of us do not need any research or any evidence. For some of us, the principle of purchasing sex from a woman is sexual violence, full stop' (Northern Ireland Assembly 2014d: n.p.). Many Committee members, however, were receptive to evidence given by those also with a neo-abolitionist standpoint such as Women's Aid and CARE (Christian Aid Research and Education). Ellison's sentiments are echoed by Huscke and Ward (2017) who bemoan the way in which their own research (Huschke et al. 2014), commissioned by the Department of Justice into the demographics and experiences of sex workers and their clients, was treated. They reason that it was ignored and dismissed by some MLAs as unnecessary, biased and a rouse by the Justice Minister to delay the Bill. One wonders how academic research informed by a neo-abolitionist philosophy would have fared with Morrow and the Committee.

Policy assemblages, levels and mutations
In Stormont, during December 2014, Morrow's Bill was passed by 81 votes to 10, gaining support from DUP MLAs and many MLAs from other parties. It is a Bill that should be understood as an assemblage, bringing together into temporary coherence three couplings: (1) sex work and trafficking, (2) radical feminism and the Christian right and (3) Sweden and Northern Ireland. Let us explore each of the couplings in turn. A central tenet of neo-abolitionist thinking is that sex work and human trafficking are closely connected to the point where sex work cannot exist without human trafficking (Raymond 2013). Many from a neo-abolitionist perspective would concur with a Swedish police official interviewed who suggested that 'if there was no demand for women and girls in the sex industry, in prostitution, we wouldn't have trafficking for sexual exploitation'. Tackling the demand for paid sex, therefore, is viewed as a necessary tool to reduce trafficking. Although the ways that neo-abolitionists connect trafficking and sex work have been vehemently disputed (e.g. Agustín 2007), these linkages became a repeatedly used justification for the criminalization of buying sex in Northern Ireland. So much so that prior to the vote, Jim Wells, a DUP MLA and supporter of the Bill, said in Stormont that '[w]ithout clause 15 the Bill is meaningless' (Northern Ireland Assembly 2014a: n.p.). These linkages are not exclusively Swedish in origin, but they have become closely associated with the Swedish model. As Svanström (2017) has detailed, although sexköpslagen was not originally conceived as a tool to reduce trafficking-an important difference from the 2015 Act in Northern Ireland-it has become increasingly framed as being so within Sweden and showcased internationally as such also.
This leads onto the second coupling: the bringing together of secular radical feminism with the Christian right. While many from these two groups would strongly disagree on several issues-such as abortion-many within them would agree that sex work is problematic and should be abolished (Ellison 2015;. For instance, many radical feminists would share the sentiments of one DUP MLA interviewed (#2) who argued that 'prostitution is a nasty, evil activity that causes huge hurt to the women concerned'. This coupling most clearly manifested itself in the way in which Morrow worked closely with both Gunilla Ekberg and members of CARE in drafting the Bill (Ellison 2017). In contrast to the introduction of sexköpslagen in Sweden, the Christian right have been central to the reforms in Northern Ireland. Indeed, Morrow stated in Stormont before the vote that 'taking action was very much motivated by my Christian faith and principles' (Northern Ireland Assembly 2014a: n.p.) and his party, the DUP (who are a socially conservative right-wing party), have strong connections to the Free CRIMINOLOGICAL POLICY MOBILITIES AND SEX WORK Presbyterian Church. This coalescence between the Christian right and radical feminism has been strategic, selective and, for some commentators, unequal. Huschke (2017: 201), for instance, has made the case that '[i]n Northern Ireland, the feminist rhetoric is merely used as a way of packaging sex-negative, repressive policy measures based on conservative Christian values, thereby rendering them more appealing'. It is a coupling that, according to Ellison (2017), is much more likely to benefit the religious right than feminists or women in general.
The third coupling is the linking of Northern Ireland and Sweden whereby Sweden has had a significant influence on policymaking in Northern Ireland. Other countries who have adopted the Nordic model played a limited role in policy development in Northern Ireland, the exception being the Irish Republic whose development of similar laws and the appeal of an 'all-Ireland' approach gave Morrow's Bill further impetus. Thinking about the role of Sweden, then, it is beneficial to bring in Jones and Newburn's (2007) notion of policy levels as discussed earlier. We can identify a clear transfer of certain policy content and instruments from Sweden to Northern Ireland. Certainly, the adoption of the criminalization of buying sex, the non-criminalization of soliciting and loitering, and the one-year maximum sentence closely reflect the Swedish approach. The use of fines as punishment travelled too; however, their positioning and calculation morphed. Indeed, while a fine can be used as an alternative to imprisonment in both jurisdictions, a client can be sentenced to both imprisonment and a fine for a singular offence in Northern Ireland but not in Sweden. Furthermore, the calculation of the fine in Northern Ireland does not take into account the daily income of the client as it has does in Sweden; instead, it is based solely on the perceived seriousness of the offence (in line with sentencing guidelines in Northern Ireland).
The influence of Sweden becomes more blurry when considering the incorporation of exiting services in the Act. The original Bill did not include such services, but Morrow has stated publicly that this was revised following lobbying by campaign groups in the United Kingdom and Ireland: Let me be very clear: on reflection, it was not properly catered for in my Bill. It was as a result of listening to what people were telling me about an exit strategy that we decided that it was important, and we needed to introduce it. It was emphasised by [groups…] including Women's Aid and Ruhama. Other groups said that if the Bill did not have an exit strategy, there would be a fundamental weakness in the whole strategy … This [revised] strategy is designed to try to steer them away from that and to give them the support, self-esteem and confidence that they really need. (Northern Ireland Assembly 2014e: n.p.) Every so often those lobbying for exiting services in Northern Ireland would point to similar services in Sweden. Yet Sweden did not seem to act as a blueprint in Northern Ireland. One reason perhaps is that Sweden's exiting services have a different legal and political setting: they have not been written into law but are embedded within a far more comprehensive welfare state (Levy 2015). In Northern Ireland, where the welfare state has been increasingly eroded, lobbyists successfully sought to guarantee such services through the inclusion of a section in the 2015 Act that requires the Department of Health to work with other departments to develop exiting services. The Act goes on to state that exiting services are not compulsory for sex workers and that acting as a witness in criminal proceedings is not an entry requirement. Conversely, in Sweden, although its approach is supposed to enable better access to state services, Levy (2015) suggests that some of those who refuse to stop sex work (despite it not being criminal to sell sex) have had access to support services withheld, while some will not seek these services because of fear of judgement and discrimination. Some aspects of Northern Ireland's new laws, therefore, emulated or echoed those in Sweden, but other policy content and instruments did not make the move west. One example is the powers available to the police in Northern Ireland to follow their Swedish counterparts in engaging in covert surveillance such as phone-tapping to catch those paying for sex (see Häggström 2016). As suggested by representatives of the Police Service of Northern Ireland at a Committee Stage meeting in Stormont, the payment of sex between two consenting adults would not be deemed serious enough to meet the threshold that allows convert surveillance under the UK-wide Regulation of Investigatory Powers Act 2000 (Northern Ireland Assembly 2014b).
Moving onto policy ideas, symbols and rhetoric (Jones and Newburn's second policy level), there are similarities in the way in which sex work and trafficking as well as sex workers and clients have been imagined and represented in Northern Ireland and Sweden. It is also possible to identify an argument used against critics by policymakers in Northern Ireland that was 'borrowed' from Sweden (and explicitly so): that criminalization of buying sex cannot drive sex work underground-where it cannot be policed-because it needs to be advertised in some way. Yet this example aside, it has not been possible to identify a stream of policy ideas, symbols and rhetoric flowing from Sweden to Northern Ireland. Indeed, many of these discursive tools seem to have originated in neo-abolitionist networks that stretch far beyond Sweden.
In terms of implementation by practitioners and professionals (Jones and Newburn's third policy level), more research is required into whether those at the front end of service provision and policing in different parts of Northern Ireland have taken inspiration from Sweden. That said, the relatively low arrest rates for buying sex in Northern Ireland hint that, unlike in Sweden, arresting clients has not become a policing priority (cf. Häggström 2016;McClafferty 2016). In sum, then, it appears that the movement of the Swedish model to Northern Ireland has centred on the selective borrowing and reworking of bits and pieces of policy content and instruments more than anything else.

Conclusion
This article has made the case for an alternative way of understanding the circulation of criminal justice ideas, policies and models. It has called for a more expansive engagement with the policy mobilities literature that originates in human geography. Ideas and concepts that emerge out of human geography do not receive as much attention in criminological studies as they should (for notable exceptions, see Hayward 2012 and contributions in Moran and Schliehe 2017). Acknowledging the value of engaging with human geography, we have supported Newburn et al.'s (2017) call for criminological policy mobilities studies to utilize three policy mobilities concepts (mobilities, mutations and assemblages) alongside Jones and Newburn's (2007) concept of policy levels. This, we have argued, does not go far enough. Here we have called for these issues to be explored alongside other concepts used in the policy mobilities literature (namely learning, educating and extrospective showcasing).
By drawing more expansively on the policy mobilities literature, we are better equipped to understand the circulation of sex work policy models, an issue that has received surprisingly little attention in the academic literature on sex work. By paying close attention to practices of learning, educating and showcasing, for instance, we can see that these practices-which are often ignored in academic studies of policy formation-significantly influence the mobilization of policies, ideas and models. The making of the Human Trafficking and Exploitation (Criminal Justice and Support for Victims) Act (Northern Ireland) 2015, for instance, was the result of policymakers and advocates learning from Sweden, primarily through working with the consultant Gunilla Ekberg and attendance on a study tour to Stockholm. The Act was also influenced by a transnational informational infrastructure comprising actors, events and technologies that sought to showcase the Swedish model and educate onlookers. Ignoring all this means only revealing a small part of the story. That said, a focus on mobilities, mutations, assemblages and policy levels provides a complimentary insight into the geographies of policy formation and circulation. These conceptual tools have enabled us to understand that while the Swedish model is often presented as being a coherent package 'made in Sweden', it has actually been transformed on its travels, adapted to local contexts and brought into wider assemblages when it 'lands'. So, in Northern Ireland, there are clear Swedish resonances in the Human Trafficking and Exploitation (Criminal Justice and Support for Victims) Act (Northern Ireland) 2015, particularly so in the case of policy content and instruments-to use the terminology of Jones and Newburn (2007). Yet there is a distinctive Northern Ireland 'flavour' to its take on the Swedish model that also incorporates influences from other parts of Ireland and the United Kingdom as well as wider ideas (such as those in radical feminism) whose geographical origins are difficult to pinpoint.
There is much more to explore in regards to the movement and mutation of criminal justice ideas, policies and models, with those related to sex work in particular need of further research. There are many stories to tell, here, of models that are mobile in varied ways and those that are not. This research should continue to draw on the policy mobilities literature. More importantly, criminologists can and should play an important role in shaping the future of this literature.

Funding
This article is based on research funded by a Northumbria University Research Development Fund studentship.