Alone against the danger: a study of the routine precautions taken by voluntary sex workers to avoid victimisation

This article explores the routine precautions taken by sex workers (SW) in Switzerland, a country in which sex work is a legal activity. It is based on approximately 1100 h of non-systematic participant observation spread over 18 months and 14 semi-structured interviews with indoor and outdoor SW. The findings show that SW use a series of routine precautions that overlap with the situational prevention techniques for increasing perpetrators’ efforts or their perception of the risk of offending, reducing the rewards of the crime, and decreasing the provocations and perpetrators’ excuses. Future tests of the efficacy of these routine precautions could help developing specific situational crime prevention techniques for deterring offences against SW.


Introduction
This article analyses the techniques employed by sex workers (SW) to protect themselves from crime victimisation. Sex work is a complex phenomenon, with a wide variety of forms (see Azhar et al., 2020;Harcourt & Donovan, 2005) and legal frameworks (Danna, 2014), and this research presents the particularity of having been conducted in Switzerland, a country where sex work is legal.
Previous research in different nations has addressed the issue of SW crime victimisation (e.g., Abel et al., 2010;Barberet, 2000;Berger et al., 2018;Bungay & Guta, 2018;Brents & Hausbeck, 2005;Földhàzi & Chimienti, 2007;Karandikar & Próspero, 2010;Mitjans Núñez & Molnar, 2016;Molnar 2021;Ros, 2021;Sanders, 2001;Sanders & Campbell, 2007) and the reviews of Benoit et al. (2019) and Ros (2021) conclude that their risk of victimisation is high mainly because of their work conditions. However, only few of these previous studies paid attention to SW's resilience, i.e., the techniques they employ to protect themselves. A noteworthy exception is the study conducted in Spain (N = 50) by Barberet (2000), while the review of Sanders (2001) summarizes the findings of the British studies that had addressed this issue until the turn of the Century. Both authors mention that SW attempt to avoid victimisation by prescreening their customers, asking in advance for the payment of their services, frisking gently their clients for arms, and rejecting those that they consider as undesirable. In the United States, Williamson and Folaron (2003) point out that SW (N = 13) pay attention to the verbal and non-verbal language of their clients and carry little weapons for self-defence. In addition, Brents and Hausbeck (2005), highlight the camaraderie between SW in legal brothels in Nevada, where debutant SW learn defence strategies from the more experienced ones, and they all take care of each other. Furthermore, in the Nevada parlours, SW are closely monitored while working in the premises and strict rules are applied under the principle of safety. Finally, systematizing their findings, Sanders and Campbell (2007) propose a threefold classification of the strategies employed by indoor SW in the United Kingdom (N = 117): managing the environment, mobilising individual protection strategies, and exerting collective control. Moreover, Sanders and Campbell (2007) call for the use of situational crime prevention (Mayhew and Clarke (1980; adapted later by Cornish & Clarke, 2003) to protect SW and, as it will be seen immediately, that call is particularly relevant in the context of our research.
Against that background, this paper advances the issue in three ways. First, as shown by our short review of the literature, the topic of SW's self-protection against crime has received little attention, and the scarce available research dates from more than a decade. Second, the paper is based on first-hand testimonies from a hard-toreach group exercising a high-risk activity but, contrary to previous research, in a context where that activity is legal. Third, it addresses the challenge launched by Felson and Clarke (2010): "one of the main research challenges is to collect more detailed information from citizens on their repertoire of ideas about and experiences with prevention" (Felson & Clarke, 2010, p. 9). This repertoire of ideas is what the same authors labelled as routine precautions.

Theoretical framework
This paper is inspired by the situational crime prevention paradigm, which is based on opportunity-based theories in criminology. Situational crime prevention has its roots in the works of Leslie Wilkins at the Home Office Research Unit in the 1950s and early 1960s, but was formalised in 1976 with the publication of Crime as Opportunity (Mayhew et al., 1976), followed in 1980 by Designing out Crime (Mayhew and Clarke, 1980). From that perspective, the number of crimes depends upon, among other things, the number of targets available, their attractiveness, their ease of access, the risks associated with the potential crime, and the perceived legitimate alternatives. Accordingly, Mayhew and Clarke developed in 1980 a first inventory of eight situational crime prevention techniques. In the following years, empirical research, reviews, and theoretical contributions were conducted on this topic (see, e.g., the volumes of the collection Crime Prevention Studies available at the POP Center 1 ). Subsequently, the inventory grew over time, until stabilizing itself in 25 techniques in the early 2000s (Cornish & Clarke, 2003).
These 25 techniques of situational crime prevention belong to five main categories or strategies. The first consists of increasing the efforts the perpetrator requires by making the targets less vulnerable, controlling the access to facilities, screening the exits, deflecting offenders, and controlling the tools or weapons useful for committing a crime. The second consists of increasing the risks of committing an offence by increasing guardianship and natural and formal surveillance, and by reducing people's anonymity. The third involves reducing the rewards obtained from the offence and entails concealing or removing targets, identifying property, disrupting the markets, or denying the benefits of the crime. Another strategy is based upon reducing provocations by reducing frustrations, stress, and emotional arousal, avoiding disputes, neutralising peer pressure, and discouraging imitation. Finally, the strategy of removing the excuses is related to the neutralisation techniques offenders use to justify their acts that Sykes and Matza (1957) developed. This involves setting clear standards and instructions, activating conscience, facilitating compliance, and controlling drugs and alcohol.
Situational crime prevention techniques must not be confused with the already mentioned routine precautions (Felson & Clarke, 1995: Situational crime prevention techniques are based upon evidence of what works empirically (Cornish & Clarke, 2003); while the routine precautions taken by citizens to avoid victimisation are not necessarily effective, may backfire, and could also lead to crime displacement. Still, both concepts are related because researchers can develop situational prevention techniques by applying science to the routine precautions taken by citizens. Specifically, by placing routine precautions in a theoretical framework and applying the scientific method to study them, one can produce a set of distinct opportunity-reducing techniques that can be tested empirically later.

Aim of the study
The purpose of the article is to describe, analyse and classify the routine precautions taken by freelancer SW in a Swiss city. Although it was not foreseen at the beginning, the proposed classification of these routine precautions follows the structure of the classification of situational prevention techniques developed by Cornish and Clarke (2003). The reason is that, as we advanced in the analysis of the data collected, it became more and more clear that the routine precautions developed intuitively by the SW overlapped to a great extent with the already existing techniques of situational crime prevention, even though the latter are usually implemented by agents of social control and the former are generally self-implemented.

Background
This research is based on approximately 1100 h of non-systematic participant observation spread over 18 months, and 14 semi-structured interviews with indoor and outdoor SW (N = 14) of a city in a Frenchspeaking Swiss canton. The SW interviewed do not work under coercion (from pimps or sexual exploitation networks, for example) and the choice of the city is due to the fact that the field researcher and first author of this paper was already engaged as a social worker by an NGO dedicated to the minimisation of risks related to sex work in that city. At the time of her engagement (March 2016) she did not have the intention of conducting a study; however, one year later, when the Head of the NGO manifested their need to understand the victimisation endured by the SW, it became evident that her background in criminology could be helpful for conducting such a study. Consequently, a semi-structured research protocol was developed-the lack of previous studies conducted in the same context did not allow for a more structured protocol-and approved by the Head of the NGO. The implementation of the protocol in 2018 was largely facilitated by the fact that the field researcher had established a rapport of trust with the interviewees. The research also allowed developing a structured crime victimisation prevention program for SW based on the principles of situational crime prevention (Molnar & Pongelli, 2019;Molnar 2021).

Description of the Field
Since 1992, Switzerland has considered sex work as a legal activity that can be performed by Swiss or European Union citizens who must adhere to municipal, cantonal and federal regulations, and are subject to taxation and social insurance. It is to be noted that Switzerland is a confederation composed of cantons that dispose of a great autonomy in terms of legislation. In practice, this means that sex work is a legal economic activityalthough not a profession-at the federal level, but is regulated in a specific, and often different, way in each canton.
In the canton in question, sex work is permitted both indoors and outdoors. The latter is allowed in a specific perimeter located roughly one kilometre from the city centre, in a neighbourhood that can be described as an inner-city industrial zone composed of large streets with factories and office buildings, but also schools, concert houses and bars, because it is surrounded by residential areas. At night, sex work through street soliciting in that area is allowed from 10 PM until 5 AM. During these hours, the neighbourhood-which cannot be considered as degraded or particularly dangerous-is frequented by SW and their customers by car or on foot, but also by persons who live nearby, as well as passers-by returning home from work or from a concert, or simply traversing the area to go to the city centre. Police patrols also tour the neighbourhood by car and on foot. Customers usually arrive by car and stop next to the SW, who are waiting for them on the sidewalk, to ask about pricing and type of service. If the SW and the customer agree on the terms of the service, the SW will get in the car and perform the service somewhere else, such as in the customer's home, a hotel, the SW's parlour or inside the car.
Indoors sex work is allowed in premises that have obtained an administrative licence for erotic tariffed activities and are known as erotic massage salons. These can be apartments, but there are also bars or clubs with private rooms. They all have a manager that rents the rooms to the SW-usually for the whole day-after checking that they are legally entitled to work. The managers are not authorised to impose any condition on the SW. In particular, they cannot force them to work during a certain number of hours, receive a minimum number of clients, perform specific sexual practices or accept any type of client. Moreover, they cannot take a percentage from the earnings of the SW. Most often, erotic massage salons are open from morning until midnight, with the exception of the clubs and cabarets, which also have a night shift.

Research Strategy
The 14 interviews took place from May 2017 to September 2017. The protocol allowed the inclusion of SW in an illegal situation, but keeping a balance between those in a legal situation, and also between cisgender and transgender SW. Following those rules and keeping theoretical saturation in mind, the field researcher invited the SW to participate in the study, explained to them the goals of the research, and asked their permission to make an audio recording of the interviews, assuring them that it would be erased once transcribed, and that the transcription would not be shared with any members of the NGO. All but one agreed to be recorded. For this participant, responses are based on notes taken manually during the interview. Informed consent to participate was obtained orally because SW were extremely suspicious when asked to fill out a form, provide their full name, and sign it, even if they were reassured that their name would remain anonymous. We found no evidence that the suspicion was motivated by the interviewer, which suggests that it could be due to the general stigma imposed on sex work (see Sprankle et al., 2018) that deters SW from providing personal data or signing written engagements. In the social sciences, this phenomenon has been observed in other situations, in which cultural differences justify the use of oral instead of written consent as the only way to preserve the reliability and validity of the research (Gurzawska, 2017).
The individual interviews were carried out at the times and places chosen by the SW. Usually, we proposed coffee houses and bars because they provide an opportunity of meeting the SW outside of their working hours and out of sight of colleagues or acquaintances. In these cases, the field researcher would pay for the beverages drank during the interview. When the SW could only be interviewed during their working hours, we used the rooms of the erotic massage salons. In these cases, the field researcher brought croissants and offered them to the participants. Interviews were conducted in Romanian, French and Spanish, depending on the mother tongue of the interviewee or on the language in which each of them felt more comfortable. The recorded interviews were transcribed into the same language in which they took place. For this article, quotations have been translated into English.
According to the semi-structured protocol, interviews started with general questions about the SW's situation in Switzerland, focusing later on their experience in sex work and, finally, on the sensitive issue of the related victimisation. As usual in these cases, interviewees were given the possibility of speaking freely, sharing anecdotes or stories, which allowed for follow-up questions to be formulated spontaneously according to these hints. This informal setting promoted an open atmosphere that encouraged the free association of ideas and led progressively-and much in the way grounded theory emerges (Glaser & Strauss, 1967/2000-to establish a link between the routine precautions applied by the SW and the inventory of crime prevention strategies developed by Cornish and Clarke (2003). Interviews were analysed through content analysis (Castleberry & Nolen, 2018). A first horizontal analysis (consisting in going through each transcript) allowed the identification of recurrent topics, which in turn allowed a transversal analysis (consisting in the analysis of each of those topics across transcripts).

Ethical and methodological challenges
It is well known that the study of SW is a sensitive topic (Díaz Fernández, 2019) surrounded by ethical and methodological challenges (Shaver, 2005). First, as pointed out by Shaver (2005), obtaining a representative sample of SW is challenging because the real size of the population is usually unknown. In that respect, this study-as with most qualitative studies-never sought to be representative of the population of SW in Switzerland, but to reveal aspects of their victimisation which, in turn, led us to focus on the routine precautions employed by SW. Second, prostitution is and has been the object of many controversies as well as opposing political and philosophical positions (see Benoit et al., 2019). Our position vis-àvis sex work coincides with the one adopted by the Swiss legislation, which by the way has never been challenged by the initiatives or referendums foreseen by the country's direct democracy system. Naturally, we are aware of the fact that such positioning goes beyond science.
That having been disclosed, it must be stressed that in this paper we adopt a post-positivist philosophy, which consists of describing and analysing the data as objectively as possible, while keeping a reflexive orientation (Phillips & Burbules, 2000). In that perspective, we apply the reflexivity model proposed by Alvesson andSkolberg (2009/2000) to address the three remaining ethical issues identified during our research, which are the following:

Problematising the empirical data
The data were generated through the interaction between the field researcher and the participants. As the field researcher was a social worker relatively well known by the SW before the interviews, one cannot exclude the influence of a social desirability bias (Paulhus, 1984) on the answers provided by the SW. Nevertheless, the attitude and code of conduct adopted by the interviewer during all the months that preceded the interviews should have reduced that bias. In particular, and in accordance with the principles adopted by NGOs in Switzerland, the field worker applied in her professional practice the deontological values of what is known as social work of low threshold (bas-seuil). These values are based on providing unconditional support to the users of the NGO, avoid imposing them conditions or judgements, and guaranteeing confidentiality. The aim is to facilitate the rapport of trust between the social worker and the users while promoting the respect of their self-determination. Hence, it seems plausible to conclude that the respect of those values should have had a positive impact on the relationship between the interviewer and the interviewees.

Research engagement
Raw data were analysed as such because the first author is native or fluent in the languages in which the interviews took place (Romanian, Spanish and French) and the second author is fluent in two of these. For this manuscript, the translations into English are not literal but faithful, trying to deliver the same sense expressed in the interviews to an English-speaking audience. The mastery of the three languages for conducting the interviews was an asset because 12 participants were interviewed in their mother tongue. The only exceptions were two Bulgarian participants who were interviewed in French. All the interviews took place in a cordial setting, with the SW often sharing anecdotes and jokes. We found no evidence of any type of harm caused to them by the interview and received no complaint from the intervieews in the months that followed.

Clarification of the political-ideological context
In Switzerland, sex work is a legal employment, which in practice implies that research focuses on the working conditions of SW and on finding ways of improving them. As mentioned, this was also our position in this study. At the same time, since the field researcher knew the interviewees and had provided them assistance on many occasions in the past-accompanying them to the hospital or the town hall, for example-one could wonder whether her scientific neutrality has been affected. To avoid this bias and try to guarantee neutrality, the following conduct rules were adopted: (1) applying the principles of social work mentioned above, especially the non-judgmental attitude; and (2) discussing the data with the second author who is a scholar and had no participation in the field work. Table 1 illustrates the sociodemographic characteristics of the 14 persons interviewed. Five of the interviewed SW worked outdoors in the prostitution area mentioned above, soliciting customers from 10 PM to 5 AM, four exercised their work-together with other SW-in apartments under the license of erotic massage salons, and five of them combined both. The latter would receive their customers in the erotic massage salons during the day and, at night, they would solicit clients on the prostitution zone and go back with them to the apartment to perform the sexual service. It is also to note that half of the participants were registered by the authorities as sex workers.

Description of the sample
A specific analysis of victimisation and trust in the police goes beyond the scope of this article but we shall mention, nonetheless, that all the members of the sample have been victims of at least one work-related offence. The most common offences are threats, verbal aggressions, and sexual assaults. The perpetrators are generally clients, followed by pedestrians and, less often, other SW, their pimps or managers of the salons for erotic massage. Pedestrians are typically passers-by who insult the SW, aggress them or take their belongings with or without force. Physical assaults are mainly committed by clients and often start when they get frustrated because they cannot reach their climax within the time window agreed (usually 15 min) but are still sexually aroused and not willing to pay for a second period. Most of the SW (n = 5 out of 9) expressed that they appreciated the Swiss police, although five of them also argued that the police are unsuccessful in preventing crime. Table 2 summarises the routine precautions of SW identified in this research. They are presented following the structure of the table proposed by Cornish and Clarke (2003), which places them within five overarching categories divided in five subcategories. Each category is explained in the following sections.

Increase the efforts
Apart from basic prevention routines, such as asking to be paid in advance and hiding the money received, some SW work in tandem with a colleague. In addition, some of the indoor SW avoid working during the night, when clients are more often under the influence of drugs or alcohol:  ing you, for using drugs…" (SW 1, Erika, indoor and outdoor SW). 2

"After midnight, if you open the door, you know that everything can happen. It's because of that that I wouldn't work during the night. The worst clients would come after midnight: for stealing, for bother-
In some of the salons for erotic massage, the managers have installed closed-circuit television (CCTV) on the front door, so they can prescreen the customers before interacting with them. In fact, both indoor and outdoor SW indicated that most of the time they choose their clients. This screening mixes observation with intuition. On the one hand, they check whether the customer is sober or drunk; on the other hand, they rely on their instinct to evaluate if the client seems peaceful or dangerous.
In the salons, SW propose a shower to the client, and frisk their clothes and belongings-in search of guns or other arms-while putting them aside. At the same time, some of them carry pepper spray for self-protection.
Likewise, when accepting to perform the service in an unfamiliar location such as the customer's car or home, they try to control the environment by, for example, checking the surroundings of the clients' home or the backdoor seat of their car in search of other people or arms:

"I look, wherever I go, I look in the surroundings, if I go to his place, I pay attention to the path in order to know how to come back, because you cannot trust anyone" (Eva, SW 11, indoor and outdoor SW).
Another strategy for deflecting offenders is to threaten the customer with calling the police if the conflict escalates. Some physically strong transgender SW do not hesitate to take advantage of their body to intimidate a rude client:

"I made him [the client] leave and I didn't care.
Because if you let a person dominate you, you will have problems, eh? He's going to take advantage of In a similar perspective, a SW who had been raped by a client was offered a self-defence course by the victim protection agency (Centre LAVI) that she found extremely useful: "They paid for my self-defence courses. It was very cool; it helped me a lot afterwards" (SW 13,Isabella,indoor SW).

Increase the risks
Some SW prefer to work in settings where it is possible to ask for help in a relatively easy way. For instance, they rent a room in a hotel-even if that is illegal-estimating that a client will not assault them there for fear that the persons in the nearby rooms could hear the noise and call the police or the hotel reception. This strategy highlights the limits of the current law, which could be adapted to avoid criminalising behaviours motivated basically by the need to increase personal safety.
"So, I used to work either at the client's house or, sometimes, I would join him in his hotel room. Later, for security reasons, I decided to book the hotel room by myself " (SW 13,Isabella,indoor SW).
Furthermore, in front of their clients, other SW call a colleague to inform her/him that they are going to provide a service mentioning, for example, where it will take place. If they cannot count on the latter, they would simulate social control in order to make the client think that they are accompanied. For example, when they receive a client at their home, they previously turn on the television or the radio in the nearest room. Similarly, seeking to reduce the anonymity of the client -and at the same time collect evidence that could be used in the framework of a legal procedure-some SW use email or text messages instead of telephone calls to establish the first contact with a customer.
"By email, yes. I have never given directly my telephone number" (SW 6,Harriet,indoor SW).
In the erotic massage salons, SW are usually protected by the managers or by colleagues who adopt the role of a "guardian", in the sense of Cohen and Felson (1979), and make themselves available for intervening or calling the police if needed. Some of these salons have a panic alarm button in their rooms, although that is not the rule yet. Finally, the visible presence of CCTV cameras-already mentioned in the previous section-also increases the potential offender's perception of the risk of being caught.
"Once in Geneva, there was a customer that started to be a little bothersome and he wanted to beat me, but the girls [the colleagues] defended me and called the police. Then the police came, and they took him, and he didn't want to pay me, and the police took the money from him and gave it to me" (SW 10, Janne, outdoor SW).

Reduce the rewards
As a rule, SW divide their belongings and keep them in separate places in order to reduce their losses if one or the other is stolen. For example, if they work in the street, we have observed in many occasions that they will put the ID card in their jacket and the money in their purse, or vice versa, while in an erotic massage salon they would use different hidings, depending on the architecture of their room: Similarly, as a way of reducing the temptation for the offender, some SW prefer using old cell phone models when working.

Reduce the provocations
When the prescreening mentioned above-based on observation and intuition-gives the SW the impression that the client could be problematic, they reject him politely mentioning as an excuse that they are already occupied (in the erotic massage salon) or that they are waiting for a customer (in the street). Their aim is to avoid hurting the client's feelings or generating a violent reaction, while remaining assertive concerning the choice they made.
If they accept the client, they treat him politely and try to render their work environment as "relaxing" as possible by playing meditation music or lighting candles or incense sticks. Once more, the aim is to anticipate potential sources of tension.
"With the clients it also depends on how you behave. If you behave nicely with them from the beginning, it's OK, but if you're hysterical and talk badly to them, it's normal that they also wonder: "Who does she think she is?". When you speak nicely, there are no problems" (SW 7,Monica,outdoor SW).
SW ask for payment in advance as a way of avoiding fraud or "faked" misunderstandings that may arise later. Despite that, if the client gets violent after the sexual service and wants to be reimbursed pretending that he is unsatisfied, the SW usually preserve their safety and give the money back: "The money, always in advance. One guy told me: "Yes, I will pay you afterwards". "No, brother, the hooker gets paid in advance. I am not kidding; you pay the hooker before if you want to do something. I don't even trust myself; how would I trust you? The job I do makes me not trust anyone"" (SW 7,Monica,outdoor SW).
During the service, if they feel that the client is getting nervous, the SW's routine to minimise the tension consists in continuing to speak kindly, use affectionate nicknames to address the customer (e.g. sweetie, darling), offer a massage, or try to reason him quietly. For example, the following discussion took place when discussing how a SW proceeded when a customer insisted on having unprotected sex: Interviewer When these routines do not work and the tension escalates into a conflict, most of the SW choose consciously to keep a low profile, and to get out of the situation as soon as possible. Similarly, as tension tends to escalate quickly within a group of peers, most SW avoid providing services to more than one client at the same time. In all these situations, they prioritise their bodily integrity as much as possible: "Suddenly, he took off his clothes in the middle of the street and he pulled my hair while putting a knife on my throat. And I told him: "But what do you do? First,the payment";And he said: "No,you blow me". And I told him: "Yes,OK,no problem,but wait until I take a condom" (SW 4,Cecilia,outdoor SW).
Another strategy for reducing provocations consists in avoiding making false promises about sexual practices or lying to the customer. In that sense, they associate the victimisation of some of their colleagues to what they consider unethical work practices: "[Talking about the SW from the neighbourhood] They need to be careful because they propose a lot of stuff that afterwards they don't do. It is not OK and when the client arrives, he pays you and you do only half the stuff, it's normal that he gets angry. Either he asks back for his money, either he hits you" (SW 1,Erika,outdoor and indoor SW).

Reduce the excuses
Knowing that some customers want to obtain services without paying for them, SW set clear boundaries that anticipate the potential neutralisation routines of these clients. In particular, they fix the cost according to the agreed sexual practice and specify the ones they accept and the ones in which they do not engage. The price also depends on the duration of the service, which is agreed from the beginning, too. When they advertise their services online, some SW encourage potential clients to write them specifying their expectations or ask about the latter when they engage in an exchange of emails. They also take that opportunity to clarify the type of sexual practices they offer.
Similarly, they facilitate compliance with sexual health prevention by providing their clients with the basic prophylactic material. To encourage the use of condomsthat some clients prefer to avoid-they use "marketing" strategies for making safe sex more appealing. For example, knowing that most clients would like to see the SW take pleasure during the coitus, a SW would ask them to use a specific type of condom pretending that its special characteristics increase her arousal.
Finally, some SW try to reduce the excuses by rejecting clients who are under the influence of alcohol or drugs. They also avoid drinking alcohol or using drugs themselves while working:

Discussion and conclusion
With the noteworthy exceptions of Barberet (2000), Brents and Hausbeck (2005), Bungay and Guta (2018), Sanders and Campbell (2007), and Williamson and Folaron (2003), academia has neglected the study of crime prevention from the point of view of SW, which is the focus of this article. These previous studies applied a different theoretical framework than ours, but in terms of the strategies applied by the SW, our findings are similar to those of Barberet (2000), Sanders and Campbell (2007)-both in countries where sex work is illegal-and Williamson and Folaron (2003), who conducted their study in Nevada, a state of the United States where SW is legal in some counties. Bungay and Guta (2018) did not place their results in a comparative perspective, but argued that routine precautions are limited in Canada because of the lack of regulation of sex work. These researchers did not present an inventory of the strategies applied by Canadian SW, which prevents us from comparing their results with ours; nevertheless, one can plausibly conjecture that, compared to SW in prohibitionist countries, legal SW in Switzerland are more likely to defend themselves, to mobilise the authorities in case of danger, and to have a better knowledge of their rights in case of victimisation. The only exception would be the case of non-documented migrant SW, who could risk deportation from Switzerland if they enter in contact with the criminal justice system, even as victims.
From the point of view of routine activities theory (Cohen & Felson, 1979), the clients of the SW can be seen as potential offenders, while SW or their belongings constitute a suitable target for violent or property crimes. At the same time, the service provided by the SW puts them in direct contact with clients, and these encounters, by definition, take place in the absence of a capable guardian. This means that exercising as a SW constantly creates what Felson and Eckert (2017) consider risky situations. As the intervention of police officers is not always possible (or desirable), SW are obliged to adopt routine precautions intended to reduce (a) the effects of the absence of a guardian, (b) the vulnerability of the target, and (c) the motivation of the potential offender. For example, in order to mitigate the absence of a capable guardian, they may choose to work in an erotic massage salon or in a hotel where persons who can act as guardians are always nearby, or when they work at home they may simulate the presence of a person by turning the radio on the contiguous room, or they may make a fake phone call to someone who is supposed to know where they are and with whom. On the other hand, to reduce the vulnerability of the target, they ask to be paid in advance and hide the amount received. Finally, to encourage client's compliance, they treat him in a warm way. In revised versions of the routine activities approach, Felson and others (see Eck & Madensen, 2015) focus on three elements: offender, place and target/victim, paying particular attention to the guardians that may intervene for each of them. As the strategies applied for the offender and the target/ victim have already been presented, we will only remind here that SW also try to control the place by, for example, checking the backseat of the car or the surroundings of the place where they are working. The efficacy of all these routine precautions cannot be scientifically tested with the data collected for this article, but constitutes a relevant subject for future research.
The classification of the situational crime prevention techniques developed by Cornish and Clarke (2003) provided us with a useful framework for a parsimonious presentation of the routine precautions taken by SW. Logically, crime prevention techniques go further than routine precautions. For example, reducing emotional arousal by decreasing the sexual excitation of the customer is a crime prevention technique that could not be applied by SW because excitation is inherent to sexual intercourse and probably the reason for the persistence of SW throughout the history of humanity. Similarly, disrupting markets (Cornish & Clarke, 2003) does not make sense when sex work is a legal activity, as in Switzerland.
It is also important to remember that prevention techniques should be applied by a variety of actors-both governmental and non-governmental-and not only by the potential victims. Forgetting that principle could lead to the infamous situation in which victims would be blamed for not implementing such precautions. Examples of state-led prevention techniques in Switzerland can be found in projects funded by the Federal Police to develop effective crime prevention measures to protect SW, but also in the field of law enforcement. For example, in the canton under study in this article, there is a specific brigade whose objective is the protection of SW and the enforcement of the law on prostitution.
We suggest several areas for further study. For example, our observation that some SW of legal status take the risk of infringing the legislation by working in a hotel to feel safer suggest that there is room for improving their protection through an enlargement of the sex work law. Hence, further studies could focus on the criminogenic environment of SW rather than on them as individuals (Ros, 2021) and provide thus useful insights for policymakers. A limitation of our research comes from the fact that it was not originally designed to study the routine activities taken by SW. As a consequence, we did not collect systematic information on each and every one of these techniques. Finally, our sample includes a good part of the voluntary SW that were exercising in the area under study at the moment of data collection, but it would have been useful to increase the number of participants to allow establishing further differences among profiles of SW. Similarly, it would be useful, in the future, to replicate this study in other Swiss cantons. Such kind of studies may uncover nuances among cantons that could be due to differences in the local regulations of sex work.