The Swedish Sex Purchase Act: Where Does it Stand?1

In this article, we present and discuss the intended and unintended effects of the Swedish Sex Purchase Act, which criminalises the purchase of sex within a context where the sale of sex is legal. Whether or not this means of regulating prostitution is successful, and whether it has negative consequences for people who sell sex, are important questions in international policy and academic debates. This article builds on a scoping study aimed at identifying relevant sources of information as to the consequences of the Swedish Sex Purchase Act, then summarising and discussing these findings. The article offers policy makers and scholars a comprehensive presentation of the evidence and a discussion of the methodological, political and theoretical challenges arising from this.


INTRODUCTION
Prostitution is now one of the most contentious issues in Europe.In international debates on prostitution policy, the case of Sweden, where the purchase of sex is criminalised and the sale of sex is legal, takes centre stage.Experiences from Sweden are used to argue both for and against the criminalisation of buying sex, because there are competing claims about the consequences of the law officially referred to as the Swedish Sex Purchase Act.As countries around the world debate whether to adopt 'the Swedish model', it is important to establish the effects of the Sex Purchase Act.Sweden is also a central case utilised in the scholarship of prostitution, with scholarly debates too suffering from a lack of evidence on the impact of the Sex Purchase Act.This article aims to offer policy makers and scholars a comprehensive presentation of the evidence and a discussion of the methodological, political and theoretical challenges arising from this.

THE SEX PURCHASE ACT IN CONTEXT
The introduction of the Sex Purchase Act was preceded by a long and lively political debate, but the proposal to criminalise the purchase of sex was eventually presented as part of the Women's Peace Bill. 2 The stated aim of the Act was to combat prostitution, which was seen as harmful, both for those directly involved and for society at large.It was designed to reduce prostitution; in the short-term by actively policing it and in the long run by changing attitudes towards the purchase of sex.It was also hoped that the ban would help reduce human trafficking for sexual purposes. 3Though the legislation is formulated in gender-neutral terms, the intervention model is explicitly gendered in that its aim is to target men's actions and attitudes.Gender and power were central topics in the debates leading up to the legislative reform, but earlier debates had also emphasised the link between prostitution and social problems, social inequality and a commercialisation of human relations. 4Social work initiatives have been formulated to help people selling sexual services since the 1970s in Sweden, and the existence of these is key to understanding how political and public opinion on prostitution changed leading up to the introduction of the Sex Purchase Act. 5 Social work was intended to be a central tool in the fight against prostitution following the introduction of the Sex Purchase Act: 'the social interventions will, of course, continue to be important in order to motivate prostitutes to seek help and to get away from the destructive life they lead' . 6he combination of welfare provisions and punishment resulted from the view that one party, the seller, needed help, while the other, the buyer, deserved punishment. 7However, there are some ambiguities in this, since the client was also to some extent represented as deviant and in need of help to stop buying sex. 8n debates, concerns were also expressed that the ban would have unintended consequences that would impact negatively on the lives of people who sell sex. 9It is still unclear whether the law has achieved its intentions or confirmed such concerns, and in this article we will present and discuss the existing evidence.We are particularly interested in the question of whether the law can have the kind of uniform consequences that debaters often seem to assume.There has been great optimism surrounding the power of the law to transform gender relations, with Sandra Walklate, for example, demonstrating how feminists have turned to the law as a mechanism for change. 10In the face of issues such as sexual and domestic violence, turning to the law was a particularly important strategy, with the underlying assumption being that this would accomplish both criminal and social justice.However, Walklate and others have demonstrated that the law might not live up to its promise in the face of these issues. 11We focus here in particular on whether the law can or has lived up to the promise that debaters initially hoped it would.

MATERIAL, METHODOLOGY AND MAIN OBJECTIVES
This article builds upon a review of literature relating to the consequences of the Swedish Sex Purchase Act.The review is scope-oriented, aimed at mapping key concepts, main sources and types of evidence available within this research field.According to Arksey and O'Mally, a scoping study is a technique used to map relevant literature and identify key concepts and findings. 12This type of literature review is relevant and appropriate when addressing broader topics where many different study designs might be applicable.A scoping approach offers the possibility to be as comprehensive as possible in identifying research studies and reviews relevant for informing a central research question.There are different reasons for undertaking a scoping study.The purpose can be, for example, to examine the extent, range and nature of research activity, or to determine the value of undertaking a full systematic review.Yet, the aim of a scoping study can also be to summarise and disseminate research findings and to identify research gaps in the existing literature. 13The aim of this review is to summarise and disseminate research findings concerning the effects of the Swedish Sex Purchase Act, but also to draw conclusions concerning the research activity within this field. 14hen searching for studies on the consequences of the Swedish Sex Purchase Act, we used a variety of sources.The searches involved both Swedish and international electronic databases, in social sciences and public health.We followed a so-called building block strategy: different blocks were constructed thematically, in order to make searches as specific as possible. 15This approach builds on similar and related terms combined into blocks.The search strategy involved building blocks that included terms related to the purchase of sex and criminalisation respectively and then combining the blocks to capture the most relevant articles.The first block comprised the phrases such as 'purchase of sex', 'buying sex' , 'sex client' , 'sex buyer' , 'sex work' , 'prostitution', 'human trafficking' , 'human trafficking for sexual purposes' .The second block comprised terms such as 'criminalisation' , 'prohibition' , 'prostitution law' , 'legislation' , 'sex purchase act' .We then built a third block comprising terms as 'Sweden' , 'the Nordic countries', 'Swedish model' .This resulted in a fourth block which was simply a combination of all the terms related to purchase of sex, all the criminalisation-related terms, and the Sweden-related terms.Consequently, at least one term in the first block had to be combined with at least one term in the second block and one term in the third block for an article to be considered relevant in this search.In the first, second and third blocks, terms were separated by OR and in the fourth block by AND.A total of seven databases have been used in the search process: Academic Search Elite, ASSIA, NCJRS, PubMed, Proquest (PsycARTICLES, PsycINFO and Social Science Premium Collection).We conducted searches through this building block approach regularly from 1 September 2014 until 17 May 2017.We prioritised peer-reviewed articles written in English and Scandinavian languages from 2000 onwards.The result of our search strategy showed that several peer-reviewed articles analyse and discuss Swedish prostitution policy, but very few present empirical data on the effects, consequences or impact of the prohibition on the purchase of sexual services in Sweden.We identified 19-20 peer-reviewed articles that addressed and contextualised Swedish prostitution policy.Among these articles, only ten presented empirical evidence related to the question of the effects of the Swedish Sex Purchase Act.In addition to the searches of electronic databases, we included material that we found through reference lists and so-called 'grey literature' , such as reports commissioned by national, regional or municipal authorities and NGOs.The grey literature material was identified and selected partly through our own knowledge about what is being produced within the field and partly through electronic searches via Google.Through this search strategy we included 13.Ibid.14. Ibid.15.Ibid.
29 reports and papers.Our inclusion criteria concerning this material were that the reports had to present empirical material concerning prostitution in Sweden.In addition, we included two Ph.D. theses and four monographs, based on the same inclusion criteria.
In our research study searches, we sought to answer the following questions: • is there reason to believe that the prostitution market in Sweden is now smaller than it would have been without the Sex Purchase Act? • have norms pertaining to prostitution changed due to the Act?
• has the Act limited the extent of human trafficking for sexual purposes in Sweden?
• are there signs that the Act has had unintended consequences for those who sell sex?
The first three questions relate to the central aims of the Sex Purchase Act: (1) to decrease prostitution in the short run, by deterring prospective clients and disturbing the market; (2) to decrease prostitution in the long run by changing norms and thus making the purchase of sex less attractive; and (3), by way of the first two effects, to prevent human trafficking in Sweden.The fourth question, about unintended consequences, was presented as an argument in the debates leading up to the Act, but has later been even more central in Swedish and international discussions of the pros and cons of introducing this kind of legislation.

HAS THE SEX PURCHASE ACT ACHIEVED ITS INTENDED EFFECTS?
Several initiatives with the aim of producing knowledge about prostitution have been launched since the implementation of the Sex Purchase Act.The Swedish government commissioned the National Board of Health and Welfare to monitor developments in the prostitution market to assess the practical effects of the ban.The board has published three reports and continues to follow developments. 16Sweden's first and only action plan against prostitution and human trafficking was launched in 2008. 17Various educational materials were produced as a response to the action plan, together with a number of studies and reports.Svedin and others published a comprehensive report based on a series of quantitative and qualitative studies of experiences of buying and selling sex and attitudes to prostitution. 18 prostitution units were also published. 19The Swedish Federation for Lesbian, Gay, Bisexual and Transgender Rights (RFSL) conducted a study on LGBT people's experiences of selling and buying sex. 20The National Board for Youth Affairs published a report on young people and their vulnerability to sexual exploitation on the internet, including through commercial forms of exploitation. 21The Ministry of Justice was commissioned to evaluate the ban on buying sexual services and its findings were published in 2010. 22The County Administrative Board of Stockholm was commissioned by the government to survey and gather knowledge about the extent of prostitution in Sweden and published a report on this topic in 2015. 23hese reports and studies are attempts to assess how far the objectives of the Act have been achieved, and in the following we summarise the results of these studies and present research on the same topics, before turning to a discussion of the implications.

A Reduction of the Extent of Prostitution?
A number of studies have investigated the extent of prostitution and human trafficking in Sweden to assess whether the law has had the intended effect of reducing both.Some studies are surveys designed to map how many report having sold or bought sex. 24Other studies focus on particular arenas and groups, collecting quantitative material on street prostitution, online, among young and LGBT people. 25he results of representative sample surveys are often used to compare the situation before and after the introduction of the Sex Purchase Act.In Kuosmanen's 2008 study, approximately 8 % of male respondents report having bought sexual services at some time in their life. 26In Månsson's 1996 survey, 13 % of male respondents report the same. 27uosmanen is cautious about drawing conclusions when comparing these figures, arguing that there may have been underreporting in 2008 when buying sex was criminalised, whereas in 1996 it was not. 28Waltman, however, considers Kuosmanen too cautious and concludes that the figures indicate that the number of sex buyers has declined: 'Considering the changed situation in Sweden after the law was adopted, its deterrent effects are obvious even though conviction rates are not staggering' . 29Waltman argues that their anonymity makes respondents likely to respond honestly, even in a situation of criminalisation, and cites other research on crime with an acceptable reliability level. 30His conclusion is supported by a study by Kotsadam and Jakobsson on the prevalence of buying sex and stigma. 31Based on surveys in Norway, Sweden and Denmark, Kotsadam and Jakobsson conclude that buying sex seems to be most prevalent where it is legal and least prevalent where it is illegal.However, they also raise questions concerning reliability when examining criminal acts: '[T]here may be less willingness to report doing an activity if it is illegal than if it is not'. 32According to Kotsadam and Jakobsson, the reliability problem can be solved by using an anonymous web survey, rather than face-to-face interviews or written questionnaires.They thus consider their results reliable.
The results presented in Kuosmanen's publications are often cited as evidence for a reduction in sex purchases due to the Sex Purchase Act. 33However, Kuosmanen's caution is supported by researchers in the Department of Criminology at Stockholm University, in a consultation response to the evaluation of the Act. 34Their response points out methodological problems in measuring the impact of legislation and the fact that such a rapid decline is not very likely if both studies are representative, as earlier experiences with buying sex are also included. 35here is, however, other relevant information based on experiences of buying sex which may shed light on this development.The Swedish National Police Board report that interest in buying sex persists even though the act is criminalised. 36A qualitative study of men's experiences of buying sex indicates that the law has had no deterrent effect on the informants, but that they now take more precautions to protect their identities. 37Some of the informants in the study say that criminalisation, rather than discouraging them, makes buying sex more exciting. 38 number of studies and reports assess the size of particular arenas or groups, taking the number of sellers as a starting point. 39Assessments of the situation on the street are often based on local social work units' estimates. 40Numbers of women observed in street prostitution by social work units is the basis for the conclusion in the evaluation that street prostitution has decreased by 50 %. 41Svedin and others support the conclusion that street prostitution has declined significantly over the past 10 years. 42The evaluators of the Sex Purchase Act conclude: 'This reduction should be considered a direct consequence of the criminalization of the purchases of sexual services' . 43Similar conclusions had also been presented previous to this. 44However, the lack of an empirical basis for this conclusion has been pointed out, as well as the fact that only visible prostitution (i.e.women in street prostitution) has been counted. 45stimates of the extent of prostitution online are seldom discussed from a comparative perspective, probably because these figures are difficult to interpret in relation to the ban, since the internet was not a major contact arena for prostitution in 1999, when the Swedish Sex Purchase Act was introduced.However, according to a study conducted by the County Administrative Board of Stockholm, there has been a marked increase in the number of escort ads aimed at men who buy sexual services from women in the period 2006-2014, when it increased from 304 to 6,965 ads. 46An increase in the number of ads primarily targeted at the LGBT community, has also been observed.In the period 2010-2014, the number rose from 190 to 702 ads. 47owever, the evaluation concludes that there is no evidence of an increase in indoor prostitution. 48This conclusion has been questioned by a number of agencies, both during the consultation process and in political debates following the inquiry.The critique has focused on the lack of clarity regarding the evaluation's perspectives and definitions as well as being based on limited empirical material. 49Dodillet and Östergren find the data used unreliable, and point out that there are groups being excluded in the estimates. 50The Swedish National Board of Health and Welfare expresses similar concerns about prostitution in less visible arenas, and questions whether it is possible to discern any clear trend at all. 51Even though the evaluation of the ban also noted that knowledge of the extent of prostitution is limited, it concludes that the ban has had an effect on the extent of prostitution in Sweden. 52he criticism directed at the methods and sites used to collect data suggests how challenging it is to assess the extent of prostitution.Even if the above studies were taken to mean that there has been a decrease or increase in levels of prostitution in Sweden since the introduction of the Sex Purchase Act, this is very difficult to link to the introduction of a particular Act.The prostitution market is both highly flexible and international and this means that its development is influenced by many different factors.In such a market, the law has limited scope to produce predictable outcomes.When the official Swedish evaluation of the Sex Purchase Act takes various figures at face value and claims a causal link between these and the law, this indicates an understanding of the law as working independently of other social processes, as if the rule of law operates directly upon the world.

Normative Change?
One aim of the law was to change norms by making people see prostitution as a matter of gender equality and to shift the stigma of prostitution from those selling sex to the client.Fear of stigma is assumed to impact on willingness to buy sex, and is intended to deter people from buying sex. 53Much of the research on normative changes following the 1999 Sex Purchase Act focuses on support for the Act and continued criminalisation.Kuosmanen compared results from a 2008 survey with results from studies conducted between 1996 and 2002 and found that there has been a change in the normative approach to prostitution, though there are some differences in the various studies' methods and formulations. 54In the 1996 study, 32 % reported that they believe that a man buying a sexual service from a woman should be treated as a criminal, while 67 % said that he should not.In the studies of 1999, 2002 and 2008, the responses are different.In 1999, 76 % of respondents supported banning the purchase of sexual services, while 15 % opposed this. 55A 2002 follow-up survey showed similar results: 76 % thought buying sex should be illegal. 56In Kuosmanen's 2008 study, the question about the law was put differently: 'Should we keep the law that prohibits buying sex?' Seventy per cent answered yes.In a recent study, Jonsson and Jakobsson present results from a study that compares attitudes across different prostitution regimes, concluding that citizens in a country where prostitution is legalised or regulated are more accepting of the purchase of sex, compared to citizens in countries where the purchase of sex is criminalised. 57everal of the studies mentioned (including the 1996 one) show gender differences in attitudes to the law.In the last three studies, approximately 80 % of female respondents supported the ban, while men's support for it decreased from 70 % to 60 % over the same period.The findings of a 2012 study match Kuosmanen's results to some extent: 65 % think the law should be kept. 58However, the study demonstrated that the proportion of men supporting the ban on buying sexual services had fallen since 2008.In 2012, barely 50 % of male respondents supported the ban on purchasing sexual services, while the figure for women is stable at around 80 %.
Kuosmanen's study also found that 59 % of respondents favoured criminalising the sale of sexual services, a view that women (66 %) held more often than men. 59Similar results were obtained by Svedin and others who found that 37 % of male respondents and 65 % of female respondents were in favour. 60Jakobsson and Kotsadam found that approximately 63 % believed buying sex should be illegal, and 54 % believed selling sex should be illegal. 61he fact that Swedes are not only positive about the criminalisation of the purchase of sex but also its sale is worthy of further consideration.Svedin and others argue that theirs and Kuosmanen's results indicate that the law has had an impact on people's attitudes in a different way than intended: 'This suggests that the Sex Purchase Act has led to increased support in the population for banning not only buying sex but also selling sexual services' . 62uosmanen suggests a similar interpretation: 'In contrast to many interest organisations, who regard women as a subordinated and vulnerable group, the general public seems to view prostitution in more gender-neutral terms.One plausible explanation is that those who sell sex are not viewed as victims, but more in terms of being equally legally responsible as those who buy sex' . 63ltman interprets these findings differently and argues that as the question about criminalisation in Kuosmanen's study was formulated in a more gender-neutral way than in previous studies, respondents may have believed that the question concerned procuring or trafficking, rather than those selling sexual services. 64Jakobsson and Kotsadam, however, discuss findings similar to Kuosmanen's and conclude: 'a law that criminalizes buyers is likely to affect attitudes toward selling as well, since it puts focus on the issue and signals that there is a problem'. 65aken together, these findings indicate that while there is great support in the population for criminalisation as a tool to combat prostitution, the intended message behind the law -that prostitution is demand driven and a form of violence against women -does not seem to resonate.It is further difficult to ascertain whether the increasing support for using the law to combat prostitution has happened because of the Sex Purchase Act or should rather be interpreted as happening in line with broader developments with an increase in support for rehabilitative schemes and a greater reliance on law and punishment. 66

Impact on Trafficking
One aim of the Sex Purchase Act was to combat human trafficking for the purpose of prostitution.According to the Swedish National Police Board, the number of plaintiffs appearing in human trafficking cases does not represent the scope of trafficking as this reflects police priorities, resources and competence. 67Despite the obvious difficulties of measuring it, the Swedish National Police Board states that the ban on purchasing sexual services has helped prevent trafficking for the purpose of prostitution: 'Sweden has the reputation of being a country in which it is difficult to operate in prostitution, which may have a deterrent effect on those involved in human trafficking' . 68The Swedish National Police Board's conclusions are supported by international studies examining the link between prostitution legislation and the incidence of sex trafficking. 69As mentioned above, Kotsadam and Jakobsson found that there is less sex trafficking in countries where prostitution is criminalised, and more in countries where prostitution is legalised. 70If the sale and purchase of sex are legal but procuring is criminalised there is more sex trafficking than if prostitution is criminalised, but not as much as there is if procuring is legal too. 71hese results are supported by another study by Cho and others, based on more comprehensive data, showing that those countries where the selling and buying of sex, as well as procuring, are legal report more cases of trafficking. 72However, both Kotsadam and Jakobsson and Cho and others stress that such findings should be treated with caution, both regarding the quality of the data and the question of what data should be used.Kotsadam and Jakobsson state that there are several problems with victim data. 73One is that countries with a lot of resources and well-functioning police and legal systems may detect a lot of trafficking even though the problem is not as severe as in other countries.Additionally, the legal definition of trafficking varies between countries, which may result in large differences in official records. 74uestions about data appear to be highly relevant when assessing the extent of trafficking to Sweden and these problems make it very difficult to state that the Sex Purchase Act has decreased trafficking to Sweden.Further, if we had a clear conclusion on the figures it would still be difficult to assess whether the development in such a complex phenomenon as trafficking is due to one particular Act in one particular country, especially since the Act is not broadly implemented. 75

ARE THERE UNINTENDED CONSEQUENCES?
As demonstrated above, it is difficult to assess the evidence relating to the effects of the Swedish Sex Purchase Act.Governmental and NGO reports and research describe findings that can be interpreted in different ways.If one only looks at street prostitution, it is possible to conclude that the market has declined.If one only looks at the indoor market, it seems like the market has actually increased.Similarly, looking only at the finding that support for the law is strong and increased after the inception of the law, it appears that the law strengthened the perspective that clients should be made responsible for the existence of prostitution.But when also considering levels of support for criminalising the selling of sex, it appears that the intended aim of the law to redefine prostitution and ideas around who is to blame has not succeeded.
As mentioned earlier, in the debates leading up to the introduction of the Sex Purchase Act and in discussions about it in other countries, the question of unintended consequences has been important, and we now move on to present knowledge-production relating to this aspect.
The study of the intended effects of reducing prostitution and sexual trafficking and creating normative change has been most prominent in academic research, while limited attention has been paid to the experiences of those who buy and sell sexual services.However, there has been some research on the application and implementation of the law and on social work agencies and welfare provisions that the Sex Purchase Act was intended to complement. 76The experience of selling and buying sexual services has also been examined from various perspectives. 77We have organised the presentation of our findings on this in terms of changes in the market that can be partly related to the Swedish Sex Purchase Act and that affect the lives of people involved in prostitution, as well as changes in stigma and marginalisation for people who sell sex, particularly in relation to representatives of social welfare and health institutions and the police.
5.1 'Spatial Switching' -A Consequence of Criminalising Purchases of Sexual Services?
As described above, several studies conclude that visible prostitution has declined since the ban was introduced. 78According to the evaluation, there is nothing to suggest that prostitution in less visible arenas has increased in the same period. 79In international prostitution scholarship, however, technological developments are seen as shifting where and how prostitution takes place. 80The internet appears to be of great importance for commercial sex today: there has been a sizeable increase in the number of online ads and profiles, and the marketing of sexual services has been professionalised. 81The question is how these changes are linked to the Sex Purchase Act.Hubbard, Matthews and Scoular note the movement of the sale of sex to less visible arenas, calling this phenomenon 'spatial switching' . 82They argue, however, that such developments are not only due to the availability of the internet and mobile phones, but are also linked to how prostitution is regulated within the law.In Sweden, the purchase of sex is criminalised irrespective of where it takes place, so, according to Hubbard and others, the law was not expected to precipitate a move 'indoors' . 83They further claim that the police seem to have targeted visible street prostitution, while giving other arenas less attention. 84n their recent survey, the County Administrative Board of Stockholm addresses this: 'there are many indications that the digitalization of society and the way we trade goods and services in general have affected the reduction of street prostitution.In parallel, police efforts to prosecute the purchase of sexual services in public places, have of course, also contributed to this reduction' . 85hat do we know about how technology and prostitution policy interact in influencing where prostitution takes place?One observation is that heavier policing on the street seems to have decreased the number of buyers. 86This is an intended consequence of the law, but what kind of consequences might this create for people who sell sex?Fewer buyers will mean lower income for many sellers, who are then less able to turn down clients or to negotiate condom use. 87Further, negotiations on the street have to be quick because clients fear getting caught, something which reduces the sex-sellers' ability to assess risks. 88dlund and Jakobsson highlight experiences of victimisation, violence, exclusion and stigmatisation, and how the law affects safety negatively. 89Several informants say they no longer sell sex in their own apartment/room or in a hotel, even though they feel these are safer places.The Swedish National Police Board has observed the same trend, saying that many sex buyers prefer 'out-calls' since they believe that such a strategy decreases the risk of detection. 90Strategies to avoid detection make people selling sex more vulnerable, since such strategies make it harder for them to control who they meet. 91Those selling sex as their main source of income are described as being worst affected: 'those who are in most desperate financial need are not so well placed to pick and choose their clients with such care.Again, those who are vulnerable "survival" sex workers, are those affected' . 92coular concludes that, while the law was supposed to protect and assist women who sell sex, it actually seems to have created worse conditions, especially for the most vulnerable -women in street prostitution. 93t is not necessarily only street prostitution that is affected.There is also an increasing awareness of a police presence on the web among sellers and buyers.Scaramuzzino's Ph.D. thesis on social interaction online describes how both buyers and sellers of sex fear police surveillance, despite the fact that it is only the purchasing of sex that is criminalised. 94hile the studies above indicate that vulnerability and risk-taking have increased, and that prostitution has moved to less safe arenas, the question remains as to whether this is due to the Sex Purchase Act or to other factors, including market developments.Some suggest that sellers' decisions to move from street to online prostitution may have concurrent reasons, of which the Sex Purchase Act is only one.Edlund and Jakobsson argue, for example, that safety is not affected just by the Sex Purchase Act, but also by how the prohibition against procurement prevents people selling sex from working together, something which had served as a protection. 95This might work differently for different groups, par-ticularly since Swedish police seem to prioritise prostitution that is most visible and that involves foreign citizens. 96One of Hulusjö's informants describes how she began selling sex online, due to changes in the street market.She made her decision to go online due to the increase in foreign women on the street.She found that these women were younger and prettier than her, and this, together with the fact that there is more police presence on the street, made her establish her business indoors instead. 97As a Swedish citizen, she had better access to indoor markets operating in seclusion, and she was thus able to make use of her relative privilege in a way that also protects her clients.
It is also argued that the move indoors is not only creating more risk but also has some benefits.Hulusjö demonstrates how operating online offers the advantage of sharing experiences and a sense of community. 98In internet forums, women can warn others against clients who are disrespectful, violent or suspected of being dangerous. 99In addition, indoor prostitution may offer higher financial rewards. 100 shift towards contact establishment through the internet and a shift towards indoor arenas are developments taking place in many countries, including countries with very different laws from Sweden.While the Sex Purchase Act might play a role in how this shift takes place and is experienced in this particular setting, there is no reason to consider the Act as the reason for this development.

Social Stigma and Contacts with Authorities
Experiences of stigmatisation and discrimination are recurring themes in discussions related to the criminalisation of purchasing sex.In the government's evaluation of the Sex Purchase Act, it is stated that people who sell sex report that criminalisation has increased the social stigma linked to prostitution. 101The evaluation concludes that 'as for the people who are still exploited in prostitution, the negative effects of the ban described should be regarded as positive, as the aim of the Sex Purchase Act is to combat prostitution' . 102While the government considers stigmatisation of a non-criminalised party to be a positive consequence of the law, it is necessary to consider what kind of consequences such stigma entails for those who sell sex.
It is particularly important to investigate whether stigma and discrimination occur in encounters with authorities, as this may block access to services and rights.Several studies describe that sex sellers experience distrust, stigmatisation and discrimination in their contact with authorities. 103 with authorities: 'Skepticism towards the police is clearly not unfounded, many of the participants had experienced discrimination, not only by the police but also the law courts and the judicial system'. 104In Levy's study, informants report that they have experienced discrimination by the police, especially when trying to report crimes such as rape or robbery: 'One police [officer] wouldn't take my report because he said: "You're a prostitute and a prostitute can't be raped, because you get money"' . 105According to Scaramuzzino, sex sellers frequenting online forums express a lack of trust in social workers and the police. 106heir main concern is that contact with the authorities will have negative consequences: 'Sex sellers said they were very afraid of becoming disempowered by the authorities and of losing legal custody of their children if the "wrong" official found out they were selling sexual services' . 107 related theme is experiences of being judged by representatives of the welfare system.In Levy's study, informants have experienced that some units only give support under certain conditions. 108One informant explains: 'They make you a victim, and if you don't comply with being a victim, it's like they make you a criminal'. 109Levy and Jakobsson found that the approach of some specialist units is coloured by the view of prostitution the prohibition is based on: 'Thus, sex workers may feel that they have to construct a "victim" status so that they can gain access to service providers' "energy" and resources' . 110anna also argues that a particular understanding of prostitution affects social workers' approaches to those seeking assistance and concludes that the ideology behind the Sex Purchase Act colours social work agendas and services. 111Levy points out that this is truer for some social services units than others. 112A key aim is to reduce prostitution and its impact on services.There are, however, examples of harm reduction approaches within units targeting prostitution as well.
Concerns are also expressed when it comes to accessing the health care system and HIV prevention.In Edlund and Jakobsson's study, most informants say they have no confidence in the authorities and are hesitant to contact the health service.Prejudice and discrimination are central features of these informants' lives.This study also shows that those who have experienced discrimination are less likely to reveal that they sell sex when going for HIV/STD testing.Informants say discrimination inhibits them from asking direct questions about safe sex practices as sex workers, and from asking for regular testing, due to their fear of being questioned. 113n Larsdotter and others' study on LGBT people's experiences of selling sex, several informants describe stigma as a barrier to seeking help and also to talking about selling sex. 114They face an additional barrier in the face of preconceptions concerning prostitution.In the same study, informants say they experience prostitution differently from (assumed) heterosexual women who sell sex to men, and that they come up against an understanding of 'prostitutes' with which they find it difficult to identify. 115The Equality Ombudsman also points to the fact that the evaluation almost exclusively focuses on prostitution involving men buying sex from women and that the naturalisation of this increases the invisibility of other forms of prostitution. 116Larsdotter and others argue that men who sell sexual services are not as inaccessible to social work units and researchers as is often assumed. 117The authorities' lack of knowledge concerning, for example, men who sell sex, is instead due to a lack of competence in how to reach these target groups. 118uthorities' distinctions between different types and forms of prostitution, and the consequences such distinctions have, are discussed in other studies as well.Siring's study demonstrates how police officers and social workers express contradictory attitudes to people who sell sex. 119Siring's informants make distinctions that influence their priorities and focus in the field. 120Another example is the Swedish Alien Act, which has been used by the police to deport women suspected of supporting themselves in a 'dishonest way' (i.e.selling sex). 121This indicates a clear distinction between foreign women and Swedish women; a woman staying in Sweden temporarily selling sex can be deported, while a Swedish woman selling sex is considered to be in need of protection and support.These findings indicate that even though the law is neutral in terms of gender, sexuality and nationality, Swedish authorities do not approach all prostitution in a uniform way.
Even if negative encounters with the authorities are more frequently demonstrated and addressed in the empirical material, there is additional evidence of positive encounters.In Levy's study, social workers and sex sellers also report positive experiences with the police. 122To some degree, the fact that the Sex Purchase Act stigmatises those buying sex also seems to give some protection to those selling it.There are sex sellers who report that the law has increased their safety since they are now able to threaten to report the customer to the police if he does not pay, or if he is threatening or violent. 123Olsson's study also indicates sex sellers' relatively positive experiences of encounters with the authorities. 124In Kjellgren and others' study, informants state that they appreciate being acknowledged in encounters with the municipal units targeting prostitution. 125Informants also say establishing a relationship with a therapist can be a painful and lengthy process, but that they appreciate having access to a specialised unit 'where you can meet someone who actually meets others in the same situation, but who is not hardened because of that' . 126ahlborg and Hulusjö's study also shows that the presence of social workers on the street is appreciated. 127Some informants in Kjellgren and others' study experienced these specialised municipal units as an expression of societal concern. 128As mentioned above, the evaluation of the Sex Purchase Act concluded that increased stigma linked to prostitution should be regarded as positive. 129The aim of the Act, however, was to combat prostitution by shifting the focus from seller to buyer.If the legislation has resulted in increased stigmatisation, reduced trust in authorities and fear of discrimination in encounters with police and welfare agencies, and this at the same time is considered a positive outcome, this does not work towards establishing such a shift.Instead this indicates that people who sell sex still bear the burden of being considered the problem in the realm of prostitution.Stigma, shaming and discrimination are experiences shared by people who sell sex in many different empirical contexts, often with very different legal regimes than that in Sweden.

DISCUSSION
Of the themes that emerged in the literature we surveyed, the first is the difficulty of assessing whether prohibition has affected the extent of prostitution.It is important to note that this is not only a question of counting, but also one of delineation: the question of 'how much' links with what is considered prostitution.As mentioned above, quantitative studies suggesting a decrease in visible prostitution have been met with critique related to the question of delineation: is it possible to draw conclusions on the extent of prostitution solely by collecting data from the street, and/or particular periods online?How is the selling and purchasing of sex defined, and how do these definitions affect the information gathered?Taking the full scope of this problem into consideration makes it evident that the question of 'how much' is much more than a methodological problem.How and who one counts is linked to a particular definition of prostitution.Tiby and Pettersson demonstrate how police and social workers relate to the issue of prostitution, and draw attention to how administrative categories and political perspectives affect what is observed and what is not. 130They find that the police and social services in Sweden have a narrower definition of prostitution than that of NGOs and health care agencies.Siring's research among police officers indicates that it is not prostitution per se that is considered the problem, but rather women's prostitution, and particularly the prostitution of non-Swedish women. 131The question of delineation is also theoretical, in the sense that how prostitution is delineated from other social, economic and sexual relations needs to be seen in the context of societal norms and practices related to sexuality, gender, labour and the body, to name a few.
A second theme is the relationship between law and normative change.Studies of attitudes to prostitution before and after the enactment of the legislation indicate a change in attitudes to how prostitution should be approached legally.Public support for the criminalisation of buying sex is strong, but so is support for criminalising its sale.If the Swedish population unambiguously considers prostitution to be a form of violence the seller is a victim of, which is the argument behind the Sex Purchase Act, they would probably not support criminalisation of the seller.The normative change thus does not seem to be moving in the intended direction.The kind of instrument that law is in contemporary society is a subject for theorisation in several disciplines.Building on Foucault, Hunt emphasises that the law does not act on individuals in a straightforward way, but rather establishes particular acts as normal, not meaning 'average' or 'typical', but 'ideal', and is thus a normalising instrument. 132This is relevant to how the emphasis on law as the solution to the problem of prostitution can take part in normalising criminalisation per se, and can contribute to what is often conceptualised as 'juridification' , as well as take part in the continued construction of women who sell sex as different from other women, as evidenced through the way the intentions behind the law were formulated. 133he third theme is the link between prostitution and human trafficking.As described above, prostitution policies might affect the extent of sex trafficking.The data applied in empirical studies on this are highly contingent on national definitions of trafficking and the resources and competences of authorities.This poses methodological problems for attempts to analyse the effects of legislation.The assumption that lower demand for prostitution decreases trafficking is also conceptually flawed, as argued by Anderson and O'Connell Davidson. 134Trafficking is produced by a range of factors, and more policing of the prostitution market may very well also produce more trafficking, as it creates a market for brokers that sex sellers grow dependent on and are vulnerable to. 135This has political consequences.If Anderson and O'Connell Davidson are right, policy makers should be more concerned with whether prostitution policies create harsher conditions in prostitution that might make people more vulnerable to trafficking.
The fourth theme concerns the unintended consequences of the Sex Purchase Act.We indeed see a shift towards prostitution taking place indoors, but this is also something that is taking place in countries with different legislation.In Sweden, this change is often presented as a result of the Sex Purchase Act, but developments such as internet availability and mobile technology also need to be taken into account.The move indoors creates challenges for the study of prostitution.This also raises a theoretically interesting question: how are This article is downloaded from www.idunn.no.© 2017 Author(s).This is an Open Access article distributed under the terms of the Creative Commons CC-BY 4.0 License (http://creativecommons.org/licenses/by/4.0/).the emergence of national prostitution markets and possibilities of enacting national law affected by large-scale globalised developments?If a shift to more indoor prostitution leads to more risk-taking this is something that challenges politicians and policymakers, whether their overarching goal is abolitionist or harm-reductive.
Even though the Sex Purchase Act does not criminalise the sale of sex, people involved in prostitution report experiences of discrimination and stigma to the extent that it limits access to the support to which they are entitled.These results raise the methodological question of how we can measure stigma and discrimination in relation to legislation.The centuries-old stigma of 'the whore' is linked to a sexual double standard between women and men and a division of women as either 'bad' or 'good' . 136Its expressions are both symbolic and violent, but difficult to discern from other phenomena in practice.The theoretically interesting question in this is how a phenomenon like stigma is produced, and, particularly, what roles policies play in that process.Scoular points to how legal norms are much more than a practical instrument to police a population. 137To the extent they frame a phenomenon, they take part in normalising particular understandings, and as such, authorise interventions within welfare systems and impact social relations.If the Sex Purchase Act has had a negative impact on the lives of people who sell sex, this creates a political dilemma.While the evaluation concluded that a negative impact of the law on the lives of people who sell sex should be considered positive -as the goal is to combat prostitution -the intent of the law, and the continued formulation of it elsewhere, builds on a conception of prostitution as an expression of gendered power that victimises women. 138he wish to combat prostitution has to be weighed against the needs of the vulnerable few for protection, and the ideals behind the law.
New questions arise from how the market shifts its location and how globalisation affects who is part of it.Does this affect the delineation of the phenomenon and does it affect ideas about the relationship between laws and other norms, and assumptions about the relationship between this particular phenomenon and others (in this case trafficking)?The interaction between space, technology, laws and markets, and the issue of how stigma is produced and comes into play, are not exclusive to prostitution, with the above pointing to societal relationships and conflicts between ideas and practices in many fields.Scoular and O'Neill place the development of prostitution markets in a context of a political trend whereby social problems are becoming individualised, in line with a neo-liberal logic. 139hey argue that legal and social interventions targeting individuals obscure the structural conditions determining their quality of life. 140Furthermore, they interpret various forms of criminalisation in the field of prostitution as a way to make individuals responsible for dealing with structural problems.The fact that there seems to be considerable support for bilateral criminalisation can be connected with such ideas of increased individualisation and the expansion of punitive approaches to social problems. 141In addition, this can be an expression of how the law has been repurposed in a time where there are great concerns over security, something which supports Bucken-Knapp, Schaffer and Strömbäck's analysis of the Sex Purchase Act as not only resting on the ideal of gender equality, but also reflecting a securitisation agenda. 142hile the individualised responsibilisation of clients, in the form of criminalisation, and the clientification of sellers build on the idea of prostitution as a uniform phenomenon, several recent Swedish and international studies have shown the complexity and variety of experiences of selling and buying sexual services. 143The surveyed literature also demonstrates that the various groups in prostitution are also affected differently by the law.This supports the critique of the application of the term 'the Swedish Model' , as prostitution policies in Sweden are argued for, implemented and experienced in more than one way.

CONCLUSION
The knowledge base on prostitution in Sweden, as elsewhere, is both patchy and biased.The study of prostitution policy development and its implications for developments in the prostitution market is a growing part of international scholarship on prostitution.There is thus currently no shortage of claims about the relationship between policies directed at prostitution and the development in the market.These are often devoid of real discussions of methodological problems and information about context, such as broader legal cultures.Research is particularly needed to investigate how the relationship between social and legal norms is imagined and enacted; how the link between prostitution and human trafficking is conceptualised and practiced; how wider developments to do with technology, globalisation and gentrification interact with law, and how stigma and discrimination are produced and impact on the lives of individuals.There is empirical evidence and theoretical developments in other fields that can inform scholarship on prostitution and contextualise findings in a way that may increase the impact of research on policy development.

Evaluations of the social and health related work performed by the
Hulusjö's informants describe negative experiences in meeting