In most countries, regardless of the legal status of prostitution, interviewing of people in this study was periodically obstructed by others who controlled their lives, whether brothel operators (Turkey) or pimps/boyfriends (USA), or older women hired to guard those in prostitution (Thailand). Thus recruitment was necessarily opportunistic, relying on the discovery and utilization of sites where these prostituted people were away from the supervision of those who pimped them. This opportunistic sampling means that these samples may or may not be representative of the actual populations of those prostituted in each country. Nevertheless, comparisons of the results from the different countries suggest some noteworthy similarities, as well as differences.

Violence marked the lives of these prostituted people from five countries (see Tables 4 and 5). Since not everyone answered every question, the numbers of responses to a given item varied. Averaging across countries, 81 percent reported being physically threatened in prostitution; 73 percent had been physically assaulted in prostitution; and 68 percent had been threatened with a weapon. In Istanbul, 46 percent of these respondents reported physical assaults by police e.g. being kicked, beaten, or hit with a nightstick.

An average of 62 percent of the respondents from five countries told us that they had been raped since entering prostitution. Of those who were raped, 46 percent had been raped more than five times. Of these 475 respondents, 41 percent reported that they had been upset by attempts to coerce them into imitating pornography and 46 percent had pornography made of them while in prostitution.

An average of 54 percent of these interviewees reported that as children they were beaten by a caregiver to the point of injury. And 58 percent reported sexual abuse as children, with an average of four perpetrators.

Of these respondents, 72 percent reported current or past homelessness, with 61 percent reporting a current physical health problem; 52 percent reported a problem with alcohol addiction; and 45 percent reported a problem with drug addiction. In some of the countries, these percentages were much higher (see Table 5).

We found differences in lifetime and current experiences of violence, based on country. There were statistically significant differences in the experience of physical threat in prostitution (chi square = 105.37; d.f. = 4; p = .000); also in the experience of physical assault in prostitution (chi square = 27.30; d.f. = 4; p = .000); and in rape in prostitution (chi square = 17.79; d.f. = 4; p = .001). Similarly, there were significant differences, by country, in report of childhood physical abuse (chi square = 20.73; d.f. = 4; p = .000) and childhood sexual abuse (chi square = 47.43; d.f. = 4; p = .000).

The mean PTSD severities fell in a narrow range from 51 (Thailand) to 56 (South Africa) (see Table 6). Differences between the five countries' mean PTSD severities were not statistically significant (F = 1.33; d.f. = 4, 397; p = .41). Average PTSD severities across the five countries were slightly higher than treatment-seeking US Vietnam veterans (Weathers et al., 1993).

A person must have at least one of the four B symptoms of intrusive reexperiencing of trauma symptoms, at least three of the seven C symptoms of numbing and avoidance of trauma, and at least two D symptoms of physiologic hyperarousal in order to meet criteria for a diagnosis of PTSD (see Table 1). Across the five countries, an average of 67 percent of these 475 respondents met criteria for a diagnosis of PTSD. Of people currently or recently in prostitution, 75 percent in South Africa, 50 percent in Thailand, 66 percent in Turkey, 68 percent in the USA and 76 percent in Zambia met criteria for a diagnosis of PTSD. The differences between these percentages of people in each country with PTSD were statistically significant (chi square = 19.8; d.f. = 4; p = .001). When the Thai respondents, who were administered the questionnaires in a large group, were excluded from the analysis, the differences between the remaining four countries were not significantly different (chi square  = 2.66; d.f. = 3; p = .45).

In order to qualify for a diagnosis of partial PTSD, respondents must meet two of the three foregoing criteria for B, C, and D symptoms. Across the five countries, 85 percent of our respondents met criteria for partial PTSD, which suggests a significant degree of psychological distress. Of people currently or recently in prostitution, 87 percent in South Africa, 72 percent in Thailand, 86 percent in Turkey, 83 percent in the USA and 96 percent in Zambia met criteria for a diagnosis of partial PTSD. The differences between these percentages of people in each country with partial PTSD were statistically significant (chi square = 25.7; d.f. = 4; p = .000). When the Thai respondents, who were administered the questionnaires in a large group, were excluded from the analysis, the differences between the remaining four countries were not significant (chi square 4.24; d.f. = 3; p  = .24).

We investigated differences in PTSD associated with race in South Africa and the USA. There were no differences between racial groups in South Africa (chi square = 1.56; d.f. = 3; p = .67) or in the USA (chi square = 3.98; d.f.  = 4; P = .41).

We also investigated differences in PTSD associated with gender. In the USA, differences in PTSD incidence among women, men and the transgendered were not statistically significant (chi square = 2.48; d.f. = 2; p = .29). In Thailand, differences between women and the transgendered were not statistically significant (chi square =  1.31; d.f. = 1; p = .25). In South Africa, differences between women and men were not significant (chi square = .2 1; d.f. = 1; p  = .65).

In South Africa, 25 of our respondents prostituted in brothels and 43 prostituted on the street. There was more violence in the lives of those in street prostitution than brothel prostitution. We found significant differences in the incidence of physical assault in brothels as compared with street prostitution (Fisher's Exact Test, p = .000) and rapes in brothels as compared with street prostitution (Fisher's Exact Test, p = .000). There were no differences in histories of childhood physical and sexual abuse, based on whether the person was prostituted in a brothel or on the street. We investigated the relation between PTSD and whether the person was prostituted in a brothel or on the street. There was no statistically significant difference in incidence of PTSD between brothel and street prostitution (Fisher's Exact Test, p  = .25).

There were differences in the availability of support services. All of the women at TASINTHA, in Lusaka, Zambia, and most of the women in northern Thailand, were interviewed at agencies which offered support and job training. These agencies not only advocated but actually provided alternatives to prostitution. This level of support and vocational training was not available in San Francisco at the time of this study. Little governmental or nongovernmental funding in the USA is dedicated to services for those escaping prostitution. In the USA, there is widespread acceptance of the notion that prostitution is a reasonable job choice for women, and there is denial of the extent of prostitution in that country. On the other hand, European NGOs are more actively involved in providing support services for prostituted women in Asia and Africa.

There were also very few services for those in prostitution in South Africa. A drop-in center in Johannesburg, the House, advocated escape from prostitution for drug-addicted teenagers, and provided emergency services. SWEAT was a peer support agency in Capetown which promoted both safe sex and the sex industry.

We asked respondents what they needed (see Table 7). On average, 92 percent stated that they wanted to leave prostitution; 73 percent needed a physical place of asylum; 70 percent needed job training; 59 percent needed health care; 55 percent wanted individual counseling; and 49 percent wanted peer support; 47 percent needed child care; 45 percent wanted self-defense training; 38 percent needed drug or alcohol addiction treatment; 24 percent thought that prostitution should be legalized.

In South Africa and Zambia, we asked whether respondents believed that legalizing prostitution would decrease violence in prostitution.' In reply 62 percent of respondents in South Africa and 73 percent in Zambia stated that they did not believe that legalization of prostitution would decrease violence in prostitution. It should be noted that at the time the question was asked in South Africa (1996), there was a national political movement promoting legalization of prostitution.