Sex Work Policy Worldwide: A Scoping Review

Researchers have conducted a lot of research into policies regulating commercial sex. This study is a scoping review aiming to characterize the research field as well as finding research gaps and suggest directions for future research. Nine electronic databases and a key journal (Sexuality Research and Social Policy) were searched using Boolean operators to identify studies containing “prostitution AND policy” or “sex work AND policy” in the title and/or abstract. A total of 3663 studies were identified, and of them, 351 were deemed eligible after duplicates were removed and the title and abstract had been assessed according to the study’s inclusion criteria. The studies on sex work policy were often conducted in English-speaking countries, the majority of which were about streetwalkers, criminalization of sex work, and trafficking policies. Interestingly, few empirical studies were conducted and stigmatization was frequently mentioned. There is an unmet need for studies addressing the lived experiences of sex workers under the Swedish Model, as well as studies about migrant sex workers and studies from Africa, Asia, and Latin America. Researchers should direct additional efforts into understanding the lived experiences of sex workers under the Swedish Model as well as into studies covering Africa, Asia, and Latin America.


Introduction
After the end of World War II, most countries in the world revised their regulatory frameworks concerning sexual activities to reflect an increased tolerance for individual expression and personhood (Frank et al., 2010). However, the same consensus has not been achieved concerning sex work policy (e.g., Skilbrei, 2019). Regulating the exchange of sexual services for money instead remains a highly contentious policy field. Religious groups may consider sex work an evil vice that policymakers must ban, as it threatens public morals and the health of youth and children (e.g., Grohs, 2020;Sudarmo, 2018). Radical feminist groups may consider purchasing sex an expression of men's violence against women, as both buyers and organizers are seen as exploiters (e.g., Barry, 1995;Hughes, 2002). Therefore, these contenders argue that policymakers must ban the purchase of sex in a bid to abolish sex work and protect women (e.g., Coy, 2016;Niemi & Aaltonen, 2017;Zhidkova & Demir, 2016). In contrast, several human rights organizations-such as Amnesty International (2016), Human Rights Watch (2019), World Health Organization (2015,2016), and The Joint United Nations Programme on HIV/AIDS (UNAIDS, 2012)have demanded that policymakers start protecting sex workers by integrating them into the ordinary labor market and by decriminalizing the organization, selling, and buying of sexual acts between consenting adults. Given the many and diverse policies that have been put in practice, the present study aims to provide a detailed characterization and description of sex work policies worldwide with regard to the (1) countries, policies, and populations studied, (2) methods applied, and (3) themes addressed.

Various Forms of Sex Work Policies
Categorizing sex work policies is a difficult and daunting task riddled with methodological difficulties. It is beyond the scope of the present study to detail all these attempts and related discussions. However, a simple method of approaching the issue is to recognize four main ways of regulating sex work: criminalization, the Swedish Model of criminalizing only the purchase of sex (sometimes referred to as the Nordic model), legalization, and decriminalization. The first two models share the goal of abolishing sex work (Östergren, 2020). Criminalization means that both buying and selling sexual services are criminal offences. This is the most common policy in the world (Vanwesenbeeck, 2017), and is applied, for example, in the United States (e.g., Dewey & St. Germain, 2015;Phillips, 2015), Asia (e.g., Choi et al., 2019;Shen, 2016), and many countries in Africa (Global Network of Sex Work Projects, 2021). The Swedish Model criminalizes purchasing sex but refrains from taking punitive measures against the sellers. This model has recently gained traction and has now been adopted by, for example, Ireland (McGarry & FitzGerald, 2019), Northern Ireland (McMenzie et al., 2019), Canada (Krüsi et al., 2014), Israel (Levy-Aronovic et al., 2020) and France (St. Denny, 2017).
Legalization instead legalizes, or at least accepts, certain forms of sex work, while other forms are banned. England and Wales, for example, accept independent escorting but not brothel keeping, as third-party involvement is banned but not selling and purchasing sex per se (Campbell et al., 2019). Germany and the Netherlands exemplify countries that accept more forms of sex work, for example, licensed brothel keeping (Økland Jahnsen & Wagenaar, 2019). Thus, the number of forms of sex work that are accepted and not subjugated to punitive measures varies across countries. However, the common denominator is a focus on trying to restrict, control, and contain sex work (Östergren, 2020).
Decriminalizing sex work means removing all punitive measures against buying, selling, and organizing commercial sex and shifting toward instead trying to integrate sex work into the labor market (Vanwesenbeeck, 2017). So far, only New Zealand (Abel, 2014), New South Wales of Australia, and Northern Territory of Australia (NSWP, 2020) have adopted such policies.

Research on Sex Work Policies
The research field of sex work policy is vast and diverse. Researchers have been investing substantial resources in understanding the lived experiences of sex workers as well as how different national and local policies affect sex workers. The following paragraphs will briefly introduce the reader to this research field.
From a public health perspective, researchers have criticized criminalization schemes for increasing HIV infections (e.g., Shannon et al., 2015) as well as physical and sexual violence against sex workers; this violence is perpetrated not only by customers, but also by police officers (e.g., Deering et al., 2014). Researchers have further criticized criminalization for increasing stigma and discrimination, escalating the risk of violence, deteriorating working conditions, reducing access to health care services, denying sex workers self-agency, and making it more difficult to leave sex work (Vanwesenbeeck, 2017). The stated success of punitive measures in reducing the scope of sex work and changing the behaviors of buyers has also come into question, for instance, in Sweden (Dodillet & Östergren, 2011) and Northern Ireland (Ellison et al., 2019).
Researchers have also criticized legalization schemes from various perspectives. Tallmadge and Gitter (2018) criticized legalization on an overarching level for being associated with increased rates of human trafficking, and Cho (2016) pointed out that legalization schemes do not seem to sufficiently protect victims of human trafficking. More specifically, Ito et al. (2018) criticized the Senegalese policy of mandatory registrations for placing a stigma on sex workers and reducing their wellbeing, and similar criticism was directed at the corresponding German policy (Hunecke, 2019). Others have pointed out that legalization is mostly suitable for brothels, but not for regulating other types of sex work such as escorting (Jeffreys, 2010).
Owing to the few examples of decriminalization that exist today, researchers have not studied decriminalization as much as other legal schemes. However, in Australia, decriminalizing sex work has not been associated with an increasing number of sex buyers (Rissel et al., 2017), and in New Zealand, the number of sex sellers has not increased since decriminalization in 2003 (Abel et al., 2009). In Denmark, sex work was partially decriminalized in 1999 (Bjønness, 2012;Mathieson et al., 2015), but the proportion of men who buy sex has been stable at around 13 to 14 percent for decades (Lautrup, 2005;Melbye & Biggar, 1992;Schmidt et al., 1989). Thus far, some researchers have demonstrated that decriminalization is associated with improved health and working conditions (e.g., Deering et al., 2014;Shannon et al., 2015). Nevertheless, decriminalization does not eradicate all the dangers of sex work. For example, after decriminalization customers could still victimize 1 3 Sex Work Policy Worldwide: A Scoping Review streetwalkers in New Zealand (Abel, 2010;Abel et al., 2007), and still today, pimps can affect the lives of foreign sex workers in Denmark (Dyrvig Henriksen et al., 2021).
Researchers have conducted scoping reviews (Arksey & O'Malley, 2005) of various subjects concerning sexuality, sex work, and policy. For example, Raine (2021) investigated violence against male sex workers, Kirkman et al. (2013) studied Australian policies on midlife and older adult sexual health, and Carter et al. (2021) examined the research on the sexual health and rights of young people with intellectual disabilities. However, a literature survey conducted prior to the present study did not identify any previous scoping reviews looking at the research on sex work policy. The study aims to fill this research gap.

Aim and Purpose
The present study aimed to provide a detailed characterization and description of sex work policies worldwide with regard to the (1) countries, policies, and populations studied, (2) methods applied, and (3) themes addressed. In addition, the aim was to identify research gaps and make suggestions for future research. To address this aim, a scoping review (Arksey & O'Malley, 2005) of articles on the topic published during the past 30 years (between 1991 and 2021; Table 1) was performed. The following overarching research question guided the scoping review: • How can the research field of sex work policy be characterized in relation to the (1) countries, policies, and populations studied, (2) methods applied, and (3) themes addressed? Which research gaps can be identified in the field?

Methods
Policies are often the subject of scoping reviews, which are suitable for facilitating more focused research in the future (Davis et al., 2009). In comparison with systematic reviews, scoping reviews typically have broader research questions, the inclusion criteria may be developed after the literature search, the quality of the studies is not a priority, and the synthesis is typically qualitative rather than quantitative (Armstrong et al., 2011). The present study was conducted as a scoping review of research on sex work policy worldwide according to the framework formulated by Arksey and O'Malley (2005). Their framework suggests five steps to successfully reviewing the scope of a research field: (1) formulating the research question, (2) finding relevant studies, (3) selecting studies for review, (4) charting the collected data, (5) collating, summarizing, and reporting the results. The following section aims to detail how these steps were applied in the present scoping review.

Table 1
The process of identifying eligible studies for review

Identification of Relevant Studies
The first step in identifying relevant studies was to consult Uppsala University Library (www. ub. uu. se/) concerning the most relevant electronic databases for identifying studies on sex work policy. Nine relevant databases were identified: Web of Science, Scopus, Science Direct, Sociological Abstracts, Applied Social Sciences Index and Abstracts (ASSIA), PubMed, Social Services Abstract, HeinOnline, and Kluwer Law Journals. The second step was to construct Boolean operators to query the databases for all titles and abstracts containing the terms "prostitution" and "policy" and/or "sex work" and "policy." To retrieve a body of research covering both the criminalization of sex purchase in Sweden in 1999 (Ekberg, 2004) and the decriminalization of sex work in New Zealand in 2003 (Abel, 2014), the timespan was set to 1991 to 2021 (see Table 1 for details). The third step was to identify a potential key journal to query by hand, which is suggested by Arksey and O'Malley (2005) because databases may be incomplete. Given its scope and title, Sexuality Research and Social Policy was identified as such a journal, and it was used to search for articles containing the terms "prostitution policy" and/or "sex work policy." In total, a sample of 3663 entries was created. As this was a considerable sample, the scope of the present study was limited to peer-reviewed research articles written in English, thus excluding grey literature and books, although the inclusion of these sources is suggested by Arksey and O'Malley (2005).

Selection of Studies for Review
To select studies for review, all titles and abstracts identified in the performed search were read through to judge whether sex work policy was the subject of the study. Several types of studies were deemed irrelevant and therefore not selected for review, e.g., (1) studies only mentioning sex work policy as a topic, (2) studies focusing on, for example, the viewpoints of sex workers in order to improve a policy (e.g., Benoit et al., 2021), (3) book reviews and commentaries, and (4) duplicates were removed from all databases except Web of Science, as it was the first database to be searched. Studies that were deselected were recorded into a Microsoft Excel 2016 spreadsheet to calculate of the number of duplicates and irrelevant studies. In total, 351 articles were selected for review and downloaded in PDF format (Table 1 and Fig. 1).

Charting the Collected Data
The data were charted by recording information in the form of a set of variables created in a Microsoft Excel spreadsheet. The formation of these variables was an iterative process (cf. Levac et al., 2010): to establish which variables to use, the first 60 results from Web of Science were read through twice. Variables were both added and deleted before the final set of variables was established. When that process was finished and all data had been abstracted from all 351 studies, the spreadsheet 1 3 was imported into IBM SPSS Statistics version 28 to create and code additional variables used to summarize the findings. Microsoft Excel 2016 was used for some calculations.

Data Analysis
The study data were collated, summarized, and reported using descriptive statistics computed in IBM SPSS (version 28). This following section aims to explain and outline the variables coded in SPSS as well as providing some explanations for the variables used in the study when so deemed necessary.
Variables were coded and calculated in SPSS to concern: (1) the number of policies being studied, (2) whether some form of survey was used, (3) whether the survey was used to survey active sex workers, (4) whether interviews had been conducted, (5) whether the interviews had been conducted with active sex workers, and (6) whether the study was about active sex workers. In the case of empirical studies, dummy variables were created for the gender of the sex worker (non-specified gender, female, male, and non-binary gender), and the type of sex workers the study explicitly concerned (streetwalkers, escorts, brothel workers, migrants, and online workers).
The country of the study refers to the actual country or countries being studied, which is not the same as the author's affiliation. To obtain a geographical overview, this variable was recoded into which part of the world the study had been conducted Selected for review n=351 Studies excluded according to inclusion criteria n=3020 Included studies must have had sex work policy as a study object and "prosƟtuƟon" and "policy" and/or "sex work" and "policy" in the abstract or Ɵtle. Further, studies must have been wriƩen in English as well as having been peer-reviewed and published in a scienƟfic journal between 1991 and 2021. in, which was complemented with a variable for whether or not the study was about an English-speaking country (the United States, the United Kingdom, Canada, South Africa, etc.). The most difficult part of this scoping review was identifying the most suitable method of categorizing the studies according to sex work policy. Several suggestions for how to categorize sex work policies have been made over the years (see, e.g., Mathieson et al., 2015;Östergren, 2020;Skilbrei, 2019;Wijers, 2001). However, using the iterative process (Levac et al., 2010) to identify the most suitable method of categorizing studies about sex work policy led to the conclusion that the starting point must be from the perspective of the sellers and buyers of sex. What happens when they buy and sell sex? Are there any punitive measures taken? Are there any laws in place designed to protect the health and wellbeing of those involved in sex work?
Based on this line of reasoning, the most suitable categories were deemed to be criminalization, the Swedish Model, restrictionism/legalization, decriminalization, trafficking, and historical policies. The three first categories are deemed to require additional comments. To begin with, the present study divides abolitionist policies into criminalization and the Swedish Model; this is justified by the fact that only criminalization schemes regard the sex workers as criminals. To continue down this line, the study concludes that both restrictionist and legalizationist regimes refrain from taking punitive measures against buyers and sellers as long as the exchange takes place within the purview of the law. As mentioned in the introduction, however, the permissiveness of the law may vary across countries. England and Wales, for example, outlaw brothel keeping (Campbell et al., 2019), in contrast to most Germanic countries (Økland Jahnsen & Wagenaar, 2019). But still, the overarching aim of these policies is to restrict, control, and contain sex work (Östergren, 2020), which is notably also the aim of Dutch (Outshoorn, 2012) and German (Hunecke, 2019) policies. Conclusively, the present study considers that restrictionist and legalizationist policies share the same overarching logic and consequently exist on the same continuum, rather than being completely separate policies.
The next important conclusion drawn from the iterative process was that each study had to be categorized according to the object of that specific study. Just because a study had been performed in a specific country, it could not simply be categorized as that country's national sex work policy. Consider, for example, the punitive measures taken against streetwalkers in Spain (Villacampa, 2017;Villacampa & Torres, 2013). These studies must logically be categorized as being about criminalization, even though Spain has no punitive laws against sex work on a national level (Oso, 2016). The same goes for the managed zones for streetwalkers in Liverpool (Howell et al., 2008), which must logically be categorized as legalization because the zones were in fact managed, tolerated and accepted, regardless of the general sex work policy in England. In addition, there was one category for trafficking policies and one for historical policies.
However, some studies were not categorized according to sex work policy. Typically, this was the case for studies focused on other aspects of sex work policy. Examples include the victimhood-oriented policymaking in the United Kingdom (Sagar & Jones, 2014), policies on facilitating the exit of women from sex work 1 3 (Sanders, 2009), and specific help for homeless sex workers (Krüsi et al., 2012). Studies that put forth arguments on a more overarching level were also left uncategorized in this regard (see e.g., Karandikar, 2009).
Two variables (policy description and policy outcome) were used to map out whether the study had a policy description and/or a discussion concerning the outcome of a policy. Eligible studies typically had some form of evaluative concern, e.g., by discussing the impact of a certain policy on different types of sex workers. For example, the studies by Vuolajärvi (2019aVuolajärvi ( , 2019b were found to be eligible, as they evaluated the consequences of the Swedish Model for sex workers (and were subsequently coded as describing a policy outcome).
One variable (methodology) was used to discern which methodological framework had been applied by a specific study. Four categories were used: qualitative, quantitative, qualitative and quantitative, and other method. Other methods could include legal studies that formulated arguments based on laws, jurisprudence, etc. One example is Schwinn (2020), who discussed how the First Amendment of the United States Constitution bars Congress from blocking federal funds to fight HIV and AIDS to organizations without policies explicitly condemning sex work and sex trafficking. If the method was qualitative, it was noted whether the researchers had conducted (1) an ethnographic fieldwork, (2) and/or a documentation analysis of, for example, policy documents, (3) and/or an analysis of media representations. If interviews had been conducted, it was noted which type of interview had been applied, i.e., individual interviews or focus groups, and with whom the interviews had been conducted, i.e., with active sex workers, former sex workers, various kinds of stakeholders, etc. If the study was quantitative, it was noted whether surveys had been used and, in that case, who had been surveyed, or whether any previously collected data, such as cohort data, had been used. Some studies were categorized as other method, which was the case, for example, for some economic studies attempting to model various aspects of the sex work business (e.g., Della Giusta, 2010).
There may be various ways of mapping different aspects of the sex industry (see e.g., Matolcsi et al., 2021). After an iterative process (Levac et al., 2010), a decision was made to group the type of sex workers studied in the studies on sex work policy into the following categories: streetwalkers, escorts, brothel/indoor workers, online workers, migrants, and non-specified. The last category pertained to studies that had not specified which type of sex workers had been studied. This variable was also used for studies on overarching discussions concerning sex work policy that did not specify any particular type of sex workers.
Finally, three content themes were selected a priori: victimization, stigma, and empowerment. Windows Explorer in Windows 10 allows the user to search for strings in PDF files. This feature was used to search through all studies that contained any of the following strings: "victimization," "stigma," and "empowerment." All matching files were opened and searched to check (1) that the string was not located among the references, and (2) that the search term referred to the stigmatization, victimization, and empowerment of sex workers, and not, for example, the stigma attached to male buyers (O'Brien, 2015) or the Roma minority (Vidra et al., 2018). The theme of victimization included trafficking victimization and re-victimization. However, studies about victimization discourse did not qualify (see e.g.,

3
Sex Work Policy Worldwide: A Scoping Review Bromfield, 2016). Research has shown that victimization is highly prevalent among sex workers (Sanders, 2016), that stigma is an important factor in determining the health and wellbeing of sex workers (Benoit et al., 2018), and that empowerment is important in boosting security and job satisfaction among sex workers (Pajnik & Radačić, 2020).

Results
To provide a detailed characterization and description of sex work policies worldwide with regard to the (1) countries, policies, and populations studied, (2) methods applied, and (3) themes addressed, a scoping review of articles on the topic published during the past 30 years (between 1991 and 2021; Table 1) was performed. In addition, the present study attempted to identify potential research gaps in the field. The sections below outline the results from this study.

Countries of Study
This analysis demonstrated a steady yearly increase in the number of published studies on sex work policy, with a record high in 2012 with a total of 34 published studies (Fig. 2). Moreover, the four most studied countries among all 351 studies were the United States (12.3%), the United Kingdom (11.7%), Canada (7.7%), and the Netherlands (4.8%; Fig. 3). Multiple countries were studied in 11.7 percent of the studies. Among these, the most common countries were Sweden and the Netherlands with 7 studies each. In total, European countries were studied in 44.7 percent of the studies (including the studies of multiple countries). For the Nordic countries, Sweden was studied in 5.1 percent of all studies, Norway in 1.7 percent, Finland in 1.1 percent, Denmark in 0.9 percent, and Iceland in 0 percent. English-speaking countries were studied in 42.5 percent of all studies. Asian countries were only studied in 10 percent of all studies, which were followed by African and South American countries that comprised 2.8 and 2.8 percent of all studies, respectively.

Sex Work Policies and Types of Sex Workers
Among the six main categories for sex work policy (criminalization, the Swedish Model, legalization/restrictionism, decriminalization, trafficking, and historic policies), criminalization (33.9%) and trafficking policies (31.6%) were the most common (Table 2). Legalization was addressed in 26.8 percent of the studies, the Swedish Model in 19.4 percent, decriminalization in 13.1 percent, and historical policies in 13.4 percent ( Table 2).

Methods Applied
Of all studies (n = 351), 13.4 percent had applied quantitative methods (e.g., surveys, pre-collected data sets, etc.), 34.5 percent had applied qualitative methods (documentation analysis, media representations, individual interviews, etc.), 3.4 percent had applied both quantitative and qualitative methods, and 48.7 percent had applied some other method (e.g., legal studies or theoretical articles) or had not specified a method at all (e.g., articles arguing for something; Table 4).
Additional calculations show that of the studies that had applied some form of quantitative method (n = 47), 61.7 percent had used some form of pre-collected data and 51.1 percent had conducted some kind of survey. Of these latter studies (n = 24), 70.8 percent had surveyed sex workers.
Of the studies that had applied some form of qualitative method (n = 121), 76.0 percent had performed individual interviews, 2.5 percent had performed focus group interviews, 31.4 percent had done ethnographic fieldwork, 30.6 percent had carried out documentation analysis, and 9.9 percent studied media representations. Thus,  Table 3 Crosstabulation over the gender of the studied sex worker and their respective workplace, numbers represent frequencies Observe that one study can be about, for example, both men and women as well as both about streetwalkers and escorts. Thus, the total number of sex worker types refers to the number of studies that had studied, for example, streetwalkers a This figure means that the total number of streetwalkers in this scoping review were fifty-one. However, please observe that one study could be about several of these types of sex workers. One study may, for example, be about both streetwalkers and escorts, and so forth Gender of sex workers individual interviews were prevalent, with sex workers being interviewed in 71.7 percent of the studies that had interviewed someone (n = 66).
To investigate interviews with sex workers further, crosstabulation and Chi-Square Independent Tests were performed to test the hypothesis that use of interviews might differ significantly depending on (1) where the study was performed, (2) which policy had been researched, and (3) which type of sex workers had been interviewed (Table 5). The results are detailed in Table 5. The most interesting result was that significantly fewer interviews had been conducted with sex workers in studies investigating the Swedish Model, (χ 2 (1, N = 66) = 4.000, p = 0.046)).
Of the 92 studies that performed individual interviews, calculations made show that 7.6 percent had interviewed former sex workers and 53.3 percent had interviewed stakeholders in sex work (e.g., various civil servants working at agencies, policymakers, police officers, social workers). Crosstabulations and Chi-Square Independent Tests were again performed for the interviewed stakeholders in sex work. These analyses demonstrated that significantly fewer interviews with stakeholders were performed in studies about migrant sex workers, (χ 2 (1, N = 49) = 10.803, p = 0.001); Table 6).

Variance Over Time
To investigate variance over time, linear regression was performed using dummy variables for (1) whether the study had been performed in an English-speaking country, (2) type of sex work policy, and (3) the method used (interviews or surveys). The number of studies on the Swedish Model increased by an average of 3.32 per year since 1991 (95% CI [1.588, 5.056], p = < 0.001). Furthermore, the number of studies that had performed interviews increased by an average of 1.57 per  (Table 7).

Discussion
By performing a scoping review of articles on sex work policy published during the past 30 years, the present study was able to show an increase in the number of studies focused on sex work policy worldwide. Most studies had applied neither quantitative nor qualitative methods, and many studies on sex work policy were conducted in English-speaking countries, with relatively few from Asia, Africa, and Latin America. Interestingly, the most studied policies included criminalization of sex work and policies against trafficking, and most studies did not concern any specific type of sex worker. However, those that did mostly looked at streetwalkers. Stigma was often touched upon in the studies. Although the number of studies about the Swedish Model has increased over time, interviews have been used more infrequently in these studies than in studies about other policies. Since the mid-1990s, there has been a steady increase in the number of studies about sex work policy published each year. Of the studies written in English, Anglo-Saxon countries can be said to be over-researched, as they constituted over 40 percent of all studies, at the same time as these countries only make up about 6 percent of the world population (United Nations, 2021). European countries were quite often represented. However, given the large heterogeneity in sex work policies in Europe, this part of the world is considered under-researched. There are, for example, few studies from Germany, despite the need to better understand this huge and influential country's experiences from legalizing sex work. Another clearly underresearched country is Denmark, which notably has a model representing a unique policy stance that challenges the notion of a homogeneous Nordic response to sex work. Denmark is, namely, the only country in the world that has decriminalized sex work, except for procuring, the intention being to facilitate the exit of sex workers from the business (Bjønness, 2012;Bømler, 2015). In addition, it should be mentioned that Africa, Asia, and Latin America are clearly under-researched parts of Table 6 Crosstabulation over interviews with stakeholders, English-speaking country, type of sex worker policy, and type of sex worker, Chi-Square (χ 2 ) Test of Independence, p-value *p < .05 a This figure means that the total number of interviews with stakeholders in this scoping review were forty-nine. But please observe that one study could be about several policies or types of sex workers. So, one study may, for example, be both about streetwalkers and escorts, and so forth

3
Sex Work Policy Worldwide: A Scoping Review the world. There is a very urgent need for additional studies from these parts of the world. Although there are some studies from outside a Western context included in the present scoping review, the review could be said to neglect policy discussions from non-Western contexts. For this reason, the generalizability of the present results to non-Western contexts is somewhat limited. Almost half of the studies had used neither qualitative nor quantitative methods and can consequently not be said to be empirical in nature. Studies that were empirical in their approach were dominated by qualitative methods. Few studies were quantitative. Thus, a research gap may be identified of studies using quantitative methods to, for instance, understand the short-and long-term consequences of different policies by using longitudinal registry data. A literature survey carried out for the present study showed that only one longitudinal study has been conducted, namely a study by Day and Ward (2007) that followed British sex workers in London between the mid-1980s and the 2000s.
The most common policy studied in the reviewed articles was criminalization, which is not surprising given its status as the most common approach to sex work policy in the world (Vanwesenbeeck, 2017). The interest in trafficking policies has also seen a steady yearly increase since the 1990s. This may be explained by the increased wariness of human trafficking among the public and policymakers (Majic, 2015;Ward & Wylie, 2017). Fewer studies have focused on other policies such as legalization, decriminalization, and the Swedish Model. It should be noted that very few countries have implemented policies on legalization and decriminalization. At first glance, there seems to be quite a few studies on the Swedish Model. However, regression analysis showed a steady yearly increase in the number of studies about this model. The increased interest among scholars is likely explained by the traction this model has gained recently. It has now been implemented in, for instance, France (St. Denny, 2020), Canada (Landsberg et al., 2017), Ireland (FitzGerald & McGarry, 2016, and Israel (Levy-Aronovic et al., 2020). However, the studies on the Swedish Model had interviewed sex workers more infrequently than other studies had. This constitutes an important research gap, i.e., studying the lived experiences of being a sex worker under a legal model that criminalizes the buyer but not the seller. Many studies did not investigate a particular population or type of sex worker. Among those that did, the variance over time indicated a decreasing interest in studying streetwalkers. This might be explained by the fact that streetwalkers constitute a small part of the entire sex industry (Kuo, 2002). For example, in Thailand less than 1 percent of sex workers walk the streets (Steinfatt, 2002), and the corresponding figure for the United States, the United Kingdom, and Canada ranges from 10 to 30 percent (Benoit & Millar, 2001;Lowman, 2005;Matthews, 1997;O'Leary & Howard, 2001). But despite the decrease over time, streetwalkers may still be regarded as somewhat over-researched, given the large proportions of sex workers offering their services in other settings such as brothels, escorting firms, and online (Weitzer, 2005). Especially the latter category was under-researched, given the growing importance of the Internet for the sex work business (e.g., Sanders et al., 2018). Further, it should be mentioned that migrant sex workers are under-researched, given the large share of sex workers who are migrants (e.g., Gülçür & İlkkaracan, 2002;Kempadoo, 2001;Oso, 2016). Migrant sex workers may be impacted by other laws than those specifically targeting sex work (Vuolajärvi, 2019a).
Finally, stigma was frequently touched upon in the studies, but victimization and empowerment less often. Consequently, there is a gap in this research field and more studies accounting for aspects of victimization and empowerment under different policy regimes should be performed.
One limitation of the present scoping review was that it largely relied on electronic databases, which may contain errors. However, this was compensated for to some degree by the very large number of entries that were assessed for eligibility (n = 3663). Another limitation is that only studies written in English and published in peer-reviewed journals were eligible for inclusion. Hence, grey literature and literature written in other languages were excluded. Further, studies using terms other than "sex work," "prostitution," and "policy" may have been excluded, though they should have been considered eligible for the present scoping review.

Conclusion
The research field of sex work policy may be characterized as often being focused on English-speaking countries, streetwalkers, and criminalization of sex work and trafficking policies. Empirical studies are fewer in number than other types of studies, and stigmatization is often mentioned. Nevertheless, there is still an unmet need for studies addressing the lived experiences of sex workers living under the Swedish