Violence, dignity and HIV vulnerability: street sex work in Serbia


Address for Correspondence: Tim Rhodes, Centre for Research on Drugs and Health Behaviour, London School of Hygiene and Tropical Medicine, London WC1E 7HT


Sex work can be contextualised by violence, social and material inequality, and HIV vulnerability. We undertook a qualitative study to explore female and transvestite sex workers’ accounts (n = 31) of HIV risk environment in Belgrade and Pančevo, Serbia. Violence emerged as a key theme. Accounts emphasise the ubiquity of multiple forms of everyday violence – physical, emotional, social – in street sex work scenes, linked to police as much as clients. We highlight the salience of emotions in sex work risk management, in which the preservation of dignity is of prime importance. Accounts draw upon narratives of hygiene and responsibility which, we argue, seek to resist portrayals, normative to this setting, of sex workers as contaminated and irresponsible. Findings highlight how the ubiquity of the risk of violence in street sex work scenes reflects institutionalised social inequalities and injustices. Sex workers are inevitably participant in the cycle of symbolic violence they seek to resist. The challenges for HIV prevention are therefore considerable, and require interventions which not only seek to foster safer micro-environments of sex work but structural changes in the welfare, criminal justice and other social institutions which reproduce the cycle of violence faced by sex workers day to day.


Street sex work can be characterised by violence, extreme poverty, gender inequality, as well as HIV vulnerability (Church et al. 2001, Nixon et al. 2002, Jayasree 2004, Surrat et al. 2004, Pyett and Warr 1997). Studies indicate markedly higher levels of rape and mortality among sex workers relative to general population estimates (Potterat et al. 2004, Ward et al. 1999). Recent research on sex work has focused on the behavioural context of HIV risk, with arguably diminishing interest in violence (Farley and Kelly 2000). Violence can be ubiquitous in some sex work scenes (Nixon et al. 2002, Miller 1993, Barnard 1993, Sanders and Campbell 2007). Importantly, violence links with HIV vulnerability (Maman et al. 2000, Epele 2002, Kalichman et al. 1998).

Most studies have concentrated on physical violence linked to sex work (Farley and Barkan 1998, Wechsberg et al. 2006, Jayasree 2004). Physical violence interplays with HIV vulnerability in at least two ways: directly through unprotected sex coerced by the threat or experience of violence (Barnard 1993, Sanders 2004, Maman et al. 2000); and indirectly and over the longer-term by contributing to sex workers’ diminished sense of self-efficacy and volitional control (Romeo-Daza et al. 2005, Campbell 2000).

Recent research has emphasised the psychological harm of sex work, fuelled by social stigma and the risk of physical violence (Sanders 2004, Jackson et al. 2007, Farley and Barkin 1998). Non-physical violence linked to sex work, and how this shapes HIV vulnerability, is under researched. Ethnographic studies emphasise the salience of emotions in the management of risks associated with sex work, through strategies which seek to construct positive social identity and preserve self-dignity (Campbell 2000), usually by separating sex work from private life and ‘body from soul’ (Sanders 2004). A distinction between the private and public is important in how violence is socially organised and legitimised (Richardson and May 1999).

Violence, as well as other practices linked to sex work, are reproduced in a wider social relation of structural and gendered inequality (Barnard 1993, Sanders 2004, Campbell 2000). Structural violence affecting the vulnerable is normalised and internalised (Bourdieu 2001, Farmer et al. 1996), and has symbolic effects (Bourdieu and Wacquant 1992). Symbolic violence is the process whereby symbolic systems (words, images, cultural practices) promote the interests of dominant groups as well as distinctions and hierarchies of ranking between them, while legitimating that ranking such that those dominated come to accept, and thus also contribute towards, the status quo (Bourdieu and Wacquant 1992). This emphasises the multiple layers of violence which extend beyond physicality to violations of self-worth and dignity (Bourdieu and Wacquant 1992, Kleinman et al. 1997). HIV vulnerability is thus situated within a context of overlapping forms of violence linked to sex work – social, emotional and physical (Sanders 2004, Parker and Aggleton 2003).

Serbia is a society brutalised through involvement in recent wars, exacerbated by accompanying sanctions, economic hardship, corruption, and socio-economic transition (Rhodes and Simić 2005, Nikolic-Ristanovic 2000). International sanctions and four regional conflicts in former Yugoslavia reportedly created ‘universal compulsory poverty’ (Gordy 1999), trapping whole social groups on state subsidy while forcing the most able into illegal economies, including sex work (Ministry of Social Affairs of the Republic of Serbia 2003). The adverse effects of economic uncertainty are also gendered, with women more likely to be holding low-wage jobs, including those in hidden economies which lack protection regarding discrimination (Voice of Difference from Serbia 2007). Gendered inequalities in the Western Balkan region have become increasingly visible, illustrated by the rising number of reports of domestic violence against women (Vojvodina Ombudsman 2005), rape used as a weapon of war (Nikolic-Rastanovic 2000), and increased reports of family break-down related to divorce and poverty (Vojvodina Ombudsman 2005). Ethnic inequalities are also highly visible, with sizable proportions (in Serbia, almost half) of Roma living in refugee centres without employment, access to public services, or rights to citizenship (European Roma Rights Centre 2002). Media representations of conflict, much of it brutal, has been accessible to many, with uncertain long-term psychological and social consequences (Olujic 1998). During the 1990s, the state media were linked with ‘violence glorification’, proffering an image of the militant warrior as a positive role model (Voice of Difference from Serbia 2007). There is a history of violence used as a stratagem for managing social unrest, and there are reports of police violence against vulnerable populations (Djenovic 2007, Human Rights Watch 2000). Arguably, a culture of governance lacking in tolerance of social diversity and vulnerability has evolved (Gordy 1999, Djilas 1993). HIV-affected populations are subject to intense social stigma (Bernays et al. 2007).


In mid-2005 we undertook 31 qualitative interviews with sex workers on two sites in Serbia: Belgrade, the capital, and Pančevo, a nearby town in the region of Vojvodina. (Rhodes et al. 2008a). Our primary aim was to enable sex workers, for the first time in this setting, to describe their perceptions of HIV-risk environment. It became apparent that an over-arching theme in accounts was violence. The study had ethical approval from the Riverside Ethics Committee (UK) and support from the Republican AIDS Commission of Serbia.


The study included female and transvestite sex workers living or working in Belgrade or Pančevo who had exchanged sex for money, drugs or other goods in the previous month. The study had set an a priori target of 30 interviews to be reached within a month. The purposive sample set minimum quotas of outdoor and indoor sex workers from two age groups – under and over 21 years – and regarding the inclusion of injecting drug users (IDUs) and Roma sex workers. We found recruitment of sex workers, especially those working in non-street settings, difficult. Sex workers working in non-street locations in this setting are often extremely tightly controlled by pimps and organised gangs, which restricts the feasibility of network-based recruitment (Simićet al. 2006). Recruitment was assisted through collaboration with local outreach projects, and thereafter, through snowballing. Our sample is a convenience sample, with recruitment taking four times as long as originally proposed.

Data collection and analysis

Data collection used a topic guide designed to explore participants’ accounts. Key areas of interview discussion included: initiation into, and current patterns of, sex work; accessibility and use of condoms; drug use; risk management; HIV and STI prevention; and health-service need. Interviews were undertaken in Serbian (by MS and two trained fieldworkers). All participants were reimbursed 10 Euro for their participation. Interviews were tape recorded with informed consent, and lasted between 60 and 90 minutes. In Belgrade, interviews took place in cafés and restaurants (n = 14), a syringe exchange (n = 5), participant homes and living spaces (n = 4), and quiet outdoor spaces (n= 1). All interviews (n= 7) in Pančevo took place at a local non-government organisation.

All interviews were transcribed verbatim. Data coding was descriptive and thematic. Coding worked predominately at the level of participant description. Recruitment and data collection were undertaken in two waves to enable provisional coding to inform the focus of ongoing recruitment and data collection. At the interim analysis stage, violence emerged as a key theme. The themes of ‘hygiene’ and ‘responsibility’ emerged in the final analysis. Extracts below were translated by MS. Any reported names have been changed as a means of preserving anonymity. Translation is as direct from Serbian to English as possible.

Sample characteristics

Our sample comprised 24 female sex workers and seven transvestite sex workers from Belgrade and from Pančevo, who averaged 28 years, and were predominately soliciting sex on the street (n= 25). The remaining six sex workers solicited sex through local newspapers. About half of the sample were Roma (n= 15), all of whom were soliciting sex on the street and among whom were all seven transvestites. Two-thirds of the women in the sample had children (n= 15), and almost all were divorced. Only five sex workers self-reported ever having had a sexually transmitted infection and eight reported injecting drugs, one of whom reported being HIV positive and another three hepatitis C positive. All those interviewed reported previous experience of physical or sexual violence linked to their work.


As data were generated it quickly became apparent that some of the most impassioned accounts of sex workers concentrated on violence. Violence featured strongly in accounts since it was embedded within the everyday. Accounts depicted sex work as a process of ‘juggling’ competing risks, emphasising at once multiple forms of violence (physical, emotional, social) and agents of risk (police, clients):

This is not an easy job in the least. Demanding. You are emotionally fucked. Emotionally. You have to think about your client. You have to think about cops. You have to think about money. You have to think about everything (1, transvestite, Roma).

The ubiquity of the risk of violence

Street sex workers found themselves caught in a cycle of violence, protecting their bodies from physical harm, themselves from emotional harm, and their earnings from theft. We found that violence was at once a form of risk and a means of communication and regulation in relation to risk, with violence perpetuated by multiple actors, and sex workers at once victim and active in this process. We also found that everyday violence linked to sex work was embedded within a broader complex of structural and symbolic violence affecting the lives of sex workers.


Consented sex work was understood as a sexual exchange for money according to agreements made in advance with clients. Yet disputes with clients were common. Where deviations from agreements were coerced by clients, they were described as dehumanising. Clients were described as ‘difficult, primitive men’, who threatened or used physical force to realise their desires for free or cheap services. Client physical violence occurred over disputes regarding money (coerced free sex; attempts of theft) and time (accusations of having rushed clients ‘to finish’; clients being ‘unfinished’ when the agreed time had elapsed; disagreements about time having elapsed). Client physical violence was also generated out of a perceived contempt (unexplained beatings; violence targeting transvestites). Juggling the risk of physical violence is a fine line between defending a service agreement and having to accept client abuse of that agreement through coercion:

I never had a pimp, someone to look after me, so I don't watch the time. It is my [skill] to get him done as soon as possible. But when I see that he can't do it, I leave him and I go away.

Int.: How do they react?

Well, many of them are rude, so to say. Like, ‘Hey, I’ll break your bones’. You know, this or that. ‘You can't break my bones. I told you what your time is, we said exact time I'd be with you, you agreed, and that's all there is to it. OK then, bye.’ And that's it.

Int.: Do they never create problems for you?

Then sometimes I have to stay to finish him off, even if that means a full hour. I don't want to get myself into problems so that someone is beating me. Fuck the time (23, female, Roma).

A sense of being in control in sex work transactions was all important. Control over the transaction protects against violations to dignity and the body. Yet enforced acceptance of emotional harm – such as the ‘giving-up’ of additional time and services for free as above – can be a necessary risk-management strategy. Once physical violence has been experienced, it seemed to have the effect of puncturing a sense of security derived from everyday risk reduction tactics (such as working in familiar places only, checking on other sex workers’ safety, relying upon ‘instinct’ to assess problematic clients). Risk becomes pervasive and consuming:

He was beating me. My eye was blue with bruises. Head, back – all that was blue. He was kicking me. Took all my money away from me – his own and the money I've earned with other men. And that was an older guy . . . And I never thought I'd have any problems with him, but I did, as you can see. And so I get scared every time I get in with someone I don't know. So that I am consumed with the fear (12, female, Roma).

An awareness of the risk of physical violence, and especially previous experience of it, can reproduce a loss of sense of self-control, and an acceptance of risk, even subservience:

How can I do it? How can a woman like myself take on a man? What can I do to him? Stab him with a knife? What?! If I hit him, he’ll batter me even more. What can I do? Shut up, put your head down, get out of there like the worst whore on earth, a scum, out of the car and shut up! Maybe you had it easy after all! (40, female, IDU).

Being coerced into providing sex outside that agreed upon is at once physically and emotionally violating. It is common for sex workers to employ bodily ‘exclusion zones’– parts of the body, or bodily practices, which are excluded from commodification as a means of preserving dignity and separating a private self from work (Sanders 2004). Providing sexual services which transgress such exclusion zones, and which are beyond those agreed with clients, is emotionally scarring as well as an act of physical violence. Here is one example:

He forcefully did that [anal sex]. Nor did he allow me to put cream on, not a thing. [He did it] forcefully and my body burst. Burst from the inside. Anal guts. I was operated on because of that (25, female, Roma, IDU).

The threat of physical violence from clients was often presented as calculated, with men with violent ‘intentions’ presented as damaged in some way, as having ‘personal problems’, and with violence in these cases ‘something a man wants from the very start’. With gendered violence also portrayed as a normative feature of Serbian society, foreign clients were the safest and showed most respect:

Foreigners are best.

Int.: Why?

Well, they are gentle. You know how it is here in our country, in Yugoslavia . . . No one feels empathy towards women. There are some, but . . . It is rare to find good [men]. Foreigners are different. They live in an honest country (23, female, Roma).

Violence towards clients was usually articulated as a form of resistance borne out of struggle and discrimination, sometimes as routine, and also as strategic or tactical to maintain control. For a minority, violence was the only conceivable way of communicating with clients, since clients were viewed as essentially violating. This was the case for those who had experienced relentless violence over many years. Kata, for example, was trafficked into sex work at the age of 11, and was beaten and raped regularly for a year before escaping. She had pride in ‘beating up my clients’, would often have ‘physical fights with my clients’, and would routinely trick clients by providing ‘fake sex’.

Most sought to preserve a sense of self-dignity, and of control over clients, not through physical violence but through various forms of ‘cheating’, usually by employing ‘fake sex’, by ‘finishing-off’ clients by masturbation and as quickly as possible, by over-charging, and sometimes by stealing money. An active engagement in violating clients’ experiences of the sex for money exchange is a strategy which resists emotional harm:

He thinks he is already in the vagina, but he is not in the vagina, but between thighs . . . You are testing him. If he comes, it's an ace . . . If not, you work as usual (25, female, Roma, IDU).

Sometimes I leave them high and dry. I respond to a call, honestly. I say, ‘Sir, we can settle the bill.’ He pays 50 Euros. I say, ‘I am just going to take the money down to the driver.’ When in fact there are no drivers anywhere. The ordinary cab is waiting for me. I say to the taxi driver, ‘Now slowly start the engine. We are going back.’ And we go back. And the man is left without either the money or pussy (19, female, IDU).


The weight of various authorities was felt to be against them, but it was the police in particular who were said to be waging an ‘undeclared war’ on street sex workers. While those working through telephone contacts and bar venues had little contact with the police, street workers had day-to-day contact. There was consensus that the police posed the greatest threat of physical violence (Rhodes et al. 2008a). Whereas ‘you can manage your clients somehow’, ‘I have to be honest, the greatest enemy to us are the police’ (7, female). Accounts are peppered with multiple instances of police physical violence, either threatened or enacted, in order to obtain free sexual services, arbitrary arrests, enforced admissions, and beatings. Sex workers are commonly portrayed as having given up citizenship rights to be protected, and that, by virtue of their unacceptable occupation, are ‘asking for it’, having waived the right to be treated with respect. There was a common perception that the police had the right to beat them:

There they can even beat us. They have a right to beat us.

Int.: Is that so? Why?

Well, because we do this prostitution thing. How do I know? They have the right to beat us (12, female).

The police were routinely feared for causing physical harm, for stealing, and for acts of public humiliation. Police raids enforced individuals to register as ‘prostitutes’, enabling prosecutions. Once registered, prosecutions were relentless. The hassle and violence experienced from authorities, and especially police, was experienced as a form of punishment:

They [police] take us into their office, and one starts kicking you in the legs, the other one in the kidney. Without any reason. They [the police] want to accomplish something, to prevent us from doing something.

Int.: From what?

From doing this work: ‘Why don't you find another job?’. I say: ‘Come on, find me another job and I will do it’. And he goes, ‘Why should I look for a job for you?’, and so on (2, female).

Such ‘moral policing’ was described as an act of ‘discipline’ to ‘bring them [sex workers] to their senses’, acting to remind sex workers that they were matter out of place:

Well, he [policeman] is swearing at us, calling us sluts, whores, saying that our fannies have swollen from the amount of work we do and so on. That the city is forbidden to us. That we are not allowed to walk about, nor into cafés. If I have a boyfriend and I go out to have a drink with him, he [a policeman] comes, he wants to throw me out of the café (3, female).

Relentless police raids had the effect of pushing sex workers to work longer hours, later into the night, and into unfamiliar locations, largely as a strategy of arrest avoidance. Once again, risk management is a double-edged sword, with attempted risk reduction in one domain (the risk of police contact, arrest or violence) inevitably entertaining risk in another (the risk posed by unknown clients):

We work at the city centre itself, where we should not work [according to the police]. We are aware [of that]. But, if we go to another place, there are a lot of problems when it is night time, when it is in the late hours.

Int.: What kind of problems?

Well, the problems are you get beaten. They [clients] take away your money. They [clients] molest you (3, female).

Accounts emphasise that police punishments extend beyond the immediate environment of sex work into private life. A common tactic described was humiliation in the face of ‘normal’ others. These experiences were most violating when disclosure of a sex worker identity was made by the police to friends and family, thus disrupting attempts to preserve a dignified self and relation with others. Here is one example, in which the police undertook a ‘clean-up’ of the sex work scene in the company of invited television journalists:

The inspectors came with journalists, when they filmed me. I didn't want to say anything. They added themselves that I said something. They said, this girl does not want to state her name, they guessed my age, said this, said that, and I said nothing of it. Nothing! And that's how they found out, my husband, father-in-law, mother-in-law, my family, who rejected me. To this day I don't speak to them because of it (10, female).

A sense of generalised powerlessness in the face of the law, and deep mistrust, is contextualised by the historical institutionalisation of the police as a force of ‘obedience’ in settling civil unrest, and of generalised low tolerance to deviance in contemporary Serbian society. The police cannot be relied upon to help:

I came in to the cops, to complain. ‘Come on,’ he says, ‘you don't even know who they were.’ I say, ‘How can I know, man – two cars, eight of them. How can I know the registration plate? I know that it contained BG [Belgrade], but I don't know the whole registration plate. ‘Come on,’ he says, ‘go home. There is nothing wrong with you’. And I was covered in bruises and blood (20, transvestite, Roma).


The full blast of physical and structural violence was felt most intensely among transvestite sex workers. All transvestite sex workers we interviewed were Roma, and most were refugees of the Kosovo conflicts of the 1990s. Roma in Serbia face multiple vulnerabilities and stigma, including lack of official citizenship recognition and access to public services. Being transvestite entertains a higher risk of physical violence in a generalised context of vulnerability:

It is important to earn money, to survive. What could I do, I mean, looking like this? Even if I went to some company, honestly. I look like a woman but all the documents are for a man. We are not accepted. And we should be given a chance to show that we are normal . . . I find it very difficult to get into someone's car. That role has to be played exceptionally well, because life is at stake (22, transvestite, Roma).

Physical violence against transvestite sex workers appeared particularly brutal, and was generally interpreted as driven by contempt. Most clients of transvestites had sex on the basis that they were ‘real women’. Extreme violence would result from clients who felt cheated, and to prevent such violence occurring transvestite sex workers attempted to disguise all traces of maleness on their faces and bodies:

I got into the car and there were five of them, all skin heads. And then they ask me, ‘Are you a transvestite or a real girl?’. I claim that I am a real girl. He says ‘Tell me if you are a transvestite, it does not matter’. And I still claim that I am a bird. He wanted to put a hand into, you know, to grab my breasts. I say, ‘Don't, I'm breastfeeding’. He sees through it. And he switches on the light [in the car]. And I didn't shave that well and here a black line could be seen [shows the jaw line]. He says, ‘Why did you’, he says, ‘lie to us?. Now we will,’ he says, ‘you’ll see what we are going to do to you now’. And fortunately one guy, who liked being with transvestites, says, ‘Don't touch her. She is scared, that you might hurt her’. They take me to the woods. One held a gun against my forehead, and the other is beating me, slaps over the face (6, transvestite, Roma).

Once again, police violence was relentless and acted as a form of moral policing:

The police, how they beat us. They killed everything in me. Killed, killed, killed us with beatings. Just transvestites . . . Arms, legs. Torch into our eyes. A million times I've said ‘Take me away. Have you come to arrest me? Arrest me then. But, do not beat me’ (20, transvestite, Roma).

First, they beat us in the woods. Then they take us in the station. And in the station, he says to us, ‘Hey, freshen up’, and there in the bathroom they beat us (5, transvestite, Roma).

The sex work environment provides a site in which wider forces of structural violence towards transvestites find expression:

I had very bad times at this work. That's street life for you. Here [in Serbia], you can expect nothing that's good. There is nothing [in me] that would be normal for our people here, for our nation, here in Serbia. To them, this is something horrible. Something alien (1, transvestite, Roma).

In search of respect and health responsibility

In his ethnography on crack use In Search of Respect, Bourgois (1995) shows how social marginalisation is experienced through a struggle for dignity as well as for survival. At the same time as describing everyday violence, sex workers narrate a self in search of respect. Notions of hygiene and responsibility featured heavily in accounts. These narratives may resist normative portrayals of sex workers as ‘dirty’ and ‘irresponsible’, presenting the self as having moral integrity.


Hygiene was an important interpretive framework informing understandings of health and other ‘desirable behaviour’. This was especially the case in an absence of biomedical certainty about HIV-transmission risk. Boundaries of health/illness were commonly fused with those of cleanliness/dirt. This next extract fuses notions of hygiene, self-care and a ‘risky other’, in this case ‘junkies’:

I take good care of myself. All that [HIV] comes from the lack of hygiene, to those who don't take good care. And if they, God forbid, sleep with someone with AIDS, without a condom. That is usually junkies – sharing needles, all that – they are full of AIDS. They don't bath. You know what I am getting at? Whereas, I bath regularly. I have everything regular. I eat well. And I've got nothing, absolutely no problems (8, female, Roma).

Protecting personal and physical integrity involved maintaining boundaries of cleanliness, and this involved minimising contact with people known or thought to be HIV infected. Sex workers living with HIV infection were shunned as well as feared by some colleagues:

AIDS is transmitted when like you hang around them [people living with HIV]. You eat with them, drink and all, drink from a glass, if you get forgetful. You drink and that. That is it – the disease. Or spit. Or he [the client] is sitting on your toilet. That's it. I mean. No one is hanging around with her [HIV positive sex worker] (20, transvestite, Roma).

She injected drugs recently with Stefan, with that queer . . . He's got AIDS and they took from the same needle. Now I am afraid of her. And she owes me money, but I can't beat her because I am afraid if she bleeds over me I will get AIDS (8, female, Roma).

Hygiene was a key feature of risk-management strategy, especially regarding client selection. Smell was described by some as a sign of illness, and thus smelly, dirty and untidy clients – along with difficult or dangerous ones – were avoided: ‘When I see that someone is very messy, dirty, I don't go there. I am trying to do the job with clean, normal people’ (7, female). If dirty clients were accepted as a result of economic necessity, the use of condoms was rationalised as a means of protection from dirt as much as illness. Among those who did not use condoms for ‘low risk sexual services’, such as oral sex or masturbation, condoms were used if clients were dirty, and if otherwise suspected of having sexually transmitted infections (STIs):

I have to put a rubber on. I can spot it [STIs]. First, I can sense the smell. You can sense all that, you know. It's obvious.

Int.: Then you put the rubber on?

I do. What can I do? I can't just tell him, ‘Hey, you know what? In my opinion you are infected. Bye!’ What about the money, you know? I do this job for money (14, female).

For a minority, and for them on rare occasions, unprotected sex might be offered for selected ‘reduced risk’ clients, such as those ‘married and clean’:

Clean. It happened . . . even without a condom, but he was married. Clean. Neat. Spotless. Smells nice. Clean. [ . . . ] I needed one more customer [for the day]. I went and did it without a condom, but I saw that he was OK . . . That is, completely normal. He paid more for the service and that was it (7, female).

There is a desire to be associated with the world of the clean. Critical here are examples of ‘being clean’ in the normal world. Accounts feature instances in which the self can be shown to have been judged as clean by others, such as outside authorities, thus providing evidence differentiating themselves from the world of sex work and the ‘not so clean’. Here is an example:

When I went into labour, in the city hospital, I had no problems at all. I was the cleanest woman there. I was the cleanest woman in the city hospital. And when the doctor examined me, just like that I explained what my profession was, you know. And I told him who I was, what I was, you know what I am trying to say? And when he saw the [healthy] state I was in – the man was surprised (7, female).

Being clean is an important gateway to being in the normal world. This was especially important in the context of family. Here, being clean was borne out of a desire to protect family members and significant others – physically and symbolically – from contamination by association with sex work. This boosted a sense of self-respect and enabled the privilege of access to family interactions like any other normal or responsible adult:

I am a family person. I have brothers, I have sisters, all of whom are married. Meaning, they have children. I should be able to kiss those children . . . And not to be thinking, ‘Well, I've worked without the rubber’ (1, transvestite, Roma).


Not using condoms with clients was highly stigmatised among sex workers. Though there were exceptions (see above), most emphasised that they had never had unprotected sex with a client, and if having done so, this was ‘not willingly’, and not ‘irresponsibly’. Unprotected sex was normatively positioned as an outcome of accidents or coercion. In both cases, agency is denied, responsibility absolved, and any negative moral connotation neutralised. The appeal to accidents or to forces beyond the control of the individual, despite best intentions, is a common feature of accounting for socially unacceptable risk (Rhodes and Cusick 2000, 2002). Here, accidents usually constituted condom breakage and clients removing condoms during sex, while coercion included the multiple forms of physical and emotional violence noted above. Almost all reported accidents:

When I started working at the very start, men started ripping off the rubber themselves, and they themselves want it, you know, without the rubber . . . And then he, like, wants to work with me without a condom. And then I am the first to look. And sometimes I put my hand down to see whether it is still there, whether it didn't break. Although sometimes I don't know. Sometimes there are these rubbers which break. They are not greasy, they break on their own (12, female, Roma).

In cases of accidents, most sex workers also accounted for how they managed the situation as responsibly as they could. The emphasis here was upon learning from mistakes, and in accidents reducing in frequency as risk management experience is gained. Strategies included using more reliable condoms, checking condoms during sex, and instructing clients on how to behave during sex to prevent condom breakages:

Since I am in the business four years, maybe it [condom] broke at the most three or four times. So, believe me, these condoms are really good . . . It does not break. And all that also has to do with the men, with, to excuse my language, how they fuck and that . . . They forcefully push in, then it [condom] definitely breaks. Then, you know, I do explain to them how to do it, so that it [condom] in most cases should not break (19, female).

Accounts of coerced unprotected sex tended to emphasise the presence of violence in the extreme; being bound, humiliated, locked-up, beaten for hours, gang-raped, or threatened with weapons:

We agreed the price. He takes me to a hotel. Tied my hands to the bed. I thought, well, let him do it. Let it be smooth, you know. Easier for me later, you know. So that he does not molest me or something like that . . . He is looking at you and masturbates . . . And later he took my money.

Int.: Did he fuck you while you were tied up?

Ahhh. I mean it is really hard to talk about it, but yes, he did.

Int.: Did he use a condom?

No. He didn't use it. I don't know why so much bad luck has stuck to me!

Int.: And you agreed to intercourse only?

Yes, just a fuck. And he agreed it, gave me the money (25, female, Roma, IDU).

He put money into my hands and made me blow him. I didn't want to. He kept on banging my head against the door. And my whole head, face. I didn't know at all what I looked like [ . . . ] After that, he molested me. And I started crying. And he says: ‘You just’, he says, ‘You just dare let any noise out,’ says, ‘this gun,’ he says, ‘I’ll put it in your mouth’. He had a gun on him. I didn't dare say anything. I had to accept everything he asked for.

Int.: And he asked you for a blow job without the rubber?

Without the rubber . . . Of course, a fuck.

Int.: Without the rubber?

Yes. On top of that he asked for the anus.

Int.: Also without the rubber?

[nodding] (25, female, Roma, IDU).

Just as appeals to ‘being clean’ were realised through comparison with a dirty other, responsibility was often invoked relative to an irresponsible other. Personal accounts of inconsistent condom use were rare, but frequently reported of others. Such irresponsible behaviour was linked with older sex workers, those using drugs, those soliciting in particular areas of town, those of a different ethnicity or gender; in short, ‘other’ sex workers: ‘These addicted girls are guilty of it’ (23, female); ‘I have two colleagues working with me, they do it without condoms. Most of the clients I refuse go and get into a car with them’ (1, transvestite, Roma). Appeals to responsible agency seek the legitimation of identity but serve to sustain a cycle of symbolic violence of sex workers.


Sex workers in our study lived day to day in a context of violence. Recent studies emphasise the emotional and psychological harms of sex work alongside the risks of physical violence (Sanders 2004, Jackson et al. 2007). These have been under-estimated previously because research has concentrated on the most visible and vulnerable of street sex work scenes, in which physical violence is more common (Church et al. 2001). We found that the risk of physical violence is ubiquitous in the context of street sex work in Belgrade and Pančevo, Serbia. We found accounts of extreme violence, especially towards transvestites, including that exercised through police authority. It is important to note that we cannot speculate on the generalisability of the findings of this case study to other sex work settings in Serbia. To our knowledge, there are no published studies estimating the prevalence of violence among sex workers in Serbia.

The violences of sex work

There are multiple forms of violence linked to sex work which constitute forms of assault to self-worth and dignity, whether or not physical violence occurs (Sanders 2004). We found the preservation of dignity to be a priority. This was usually accomplished by strategies seeking to separate out, symbolically or materially, sex work from private life, and sex work identities from other identities of self (Scambler 2007). Violence can constitute a transgression of these symbolic and bodily boundaries. Examples included client coerced deviations from service agreements involving the provision of services or granting of access to parts of the body other than those for sale, and police violations through humiliation or shaming involving the revealing of an otherwise undisclosed sex worker identity to ‘normal others’. These actions disrupt the fragile boundaries between the private and public which serve to preserve dignity.

The physical and emotional risks of sex work, therefore, co-exist, albeit at different oscillations in different contexts (Sanders 2004). There is an inevitability of emotional or physical harm when violence is ubiquitous. Attempts, for example, to prevent physically violent reactions from clients or police can demand giving up the self to accepting psychological harms linked to transgressions in service agreements and the preservation of bodily boundaries. Risk reductions of one kind may entertain risk in another domain. This highlights a plurality of simultaneous violences, which can become normative in some contexts (Kleinman 2000). Pervasive social marginalisation is an everyday assault to dignity, described by some as ‘oppression illness’ (Singer 2004), which feeds risk behaviour through reduced self-efficacy, stress and the internalisation of structural violence as self-blame (Parker and Aggleton 2003, Cockerham 2000, Marmot 2005, Pietila and Rytkonen 2008). It is, then, the embodiment of the ‘inhuman’ condition of life which dehumanises, depersonalises and discredits (Shotter 1993, Goffman 1990), with felt-stigma made possible through social and structural inequalities (Scambler 2007). Whether causal relationships can be established between violations to dignity and health status is a key question for research.

Symbolic violence

Characterisations of violence in sex work often portray sex workers as beyond respect and as deserving, given their transgressions to unacceptable behaviour, or as victims of violence perpetuated by others, most often clients, but also by state systems and social or material conditions (Barnard 1993, Sanders 2004, Pauw and Brener 2003). Victimisation is a social construct, contingent upon how violence is legitimatised in certain contexts (Richardson and May 1999). Many features of structural violence – such as social, gender and ethnic inequalities – are normalised, thus rendered invisible, and create inordinate social suffering for the vulnerable they trap (Scheper-Hughes 1996, Farmer et al. 1996). The sex workers in our study are subjects of multiple vulnerabilities. Importantly, social suffering can involve those marginalised misrecognising the structural forces that propel them to self-destructive behaviours, including the physical and emotional abuse of those closest to them (Bourdieu 2001, Bourdieu and Wacquant 1992).

Previous work has highlighted how distinctions between the private and public shape social constructions of violence and victim status, especially regarding gendered and sexualised violence (Richardson and May 1999). Attempts to legitimise violence towards gay men and lesbians, for example, may emphasise maintaining a boundary in which homosexuality is linked to the private, thus preventing ‘trespassing’ into public, heterosexual, territory. In the case of sex work, the purity of the public as a sphere of the objective and non-sexual is transgressed (Carabine 1996). Violence reinforces, and works through, boundaries of the private and public, whether through processes of symbolic and structural violence which protect the purity of public space (including police actions), client violence which denies sex workers the possibility of a deserving victim status, or sex workers’ personal strategies to protect their privacy and dignity from the violences they experience at work.

We also found that sex workers are participants in the cycle of symbolic violence they seek to resist. Accounts depict a self which is active in its preservation of dignity and in the search for respect. In the face of ubiquitous violence, street sex workers would routinely cheat clients as a means of preserving dignity and of ‘taking back’ part of the self given-up for sale. Accounts also drew heavily upon narratives of hygiene and health responsibility (see also, Rhodes et al. 2008b). These narratives seek to resist, but work inside, normative portrayals of sex workers as contaminated and irresponsible. Such claims to self-integrity and responsibility were usually made in comparison to a risky or irresponsible other. Being judged, or passing, as clean or responsible by those of the normal world was especially important. Additionally, physical violence may be met by physical violence and, for a minority, physical violence was the only way through which sex workers communicated with clients.

We make these comments conscious of how research itself can contribute to symbolic violence towards its subjects. It has been noted by UK researchers that a primary focus on street sex work may conjure a ‘worse-case scenario’ account of violence, whereas across different types of the sex work scene – including indoor, managed and regulated venues – the majority of transactions are likely to be made without violent incident (Sanders and Campbell 2007). This, however, does not detract from the ubiquity of physical and emotional violence recounted by sex workers in this setting, or from recognising how these multiple forms of violence reflect institutionalised social inequalities and injustices.

Challenges for HIV prevention

The ubiquity of violence linked to street sex work poses serious challenges for public health. HIV prevention theorists champion the need for ‘social transformation’ interventions (Parker and Aggleton 2003), which foster new and resistant identities among the marginalised towards redefining their position in society. These are interventions far beyond those that increase ‘empathy and altruism’ or ‘reduce fear and anxiety’ towards the stigmatised (Parker and Aggleton 2003). As noted, Serbia is a society brutalised through recent wars and violence, and arguably characterised by economic hardship, socio-economic uncertainty, gendered and ethnic inequalities, intolerance of social diversity, and intense social stigma of HIV-affected populations (Gordy 1999, Rhodes and Simić 2005, Nikolic-Ristanovic 2000). Sex workers – and transvestite and Roma sex workers especially – are highly vulnerable. Interventions should serve sex workers by helping to build upon their search for respect. These include self-help and community organised interventions (Karim et al. 1995), the creation of safer working environments by ‘designing out’ violence from sex work (Sanders and Cambell 2007), as well as large-scale social changes in the welfare, criminal justice and other institutions, which reproduce rather than resist the cycle of everyday violence faced by sex workers (Lopes 2006, Cusick 2006).


We would like to acknowledge the funding support of the UK Department for International Development (DFID), and core funding from the UK Department of Health to the Centre for Research on Drugs and Health Behaviour. We would like to give special thanks to Sladjana Baroš who participated in this project as a fieldworker, and to Bojan Žikić at the University of Belgrade, for comments on earlier drafts of this work. We thank all individuals who participated in interviews or who helped by referring us to others, and especially JAZAS, a non-government organisation working in HIV prevention. This study had the support of the Ministry of Health of Serbia and the Republican AIDS Commission of Serbia.