ABSTRACT

We initiated this research in order to address some of the issues that have arisen in discussions about the nature of prostitution. In particular: is prostitution just a job or is it a violation of human rights? From the authors' perspective, prostitution is an act of violence against women; it is an act which is intrinsically traumatizing to the person being prostituted. We interviewed 475 people (including women, men and the transgendered) currently and recently prostituted in five countries (South Africa, Thailand, Turkey, USA, Zambia). In response to questionnaires which inquired about current and lifetime history of physical and sexual violence, what was needed in order to leave prostitution and current symptoms of post-traumatic stress disorder (PTSD) we found that violence marked the lives of these prostituted people. Across countries, 73 percent reported physical assault in prostitution, 62 percent reported having been raped since entering prostitution, 67 percent met criteria for a diagnosis of PTSD. On average, 92 percent stated that they wanted to leave prostitution. We investigated effects of race, and whether the person was prostituted on the street or in a brothel. Despite limitations of sample selection, these findings suggest that the harm of prostitution is not culture-bound. Prostitution is discussed as violence and human rights violation.

INTRODUCTION

In an effort to document the experiences of women in prostitution, we interviewed and administered psychological tests to 475 people currently and recently prostituted in five countries (South Africa, Thailand, Turkey, USA, Zambia). These people live in social and legal contexts defining them variously as hated and filthy women, criminals and 'sex workers'. We inquired about respondents' histories of violence in childhood, and in adult prostitution. For many, these two historical periods overlapped. Since violence is associated with psychological trauma, we also inquired about the severity of current symptoms of posttraumatic stress disorder (PTSD).

We began this work from the perspective that prostitution itself is violence against women. The authors understand prostitution to be a sequela of childhood sexual abuse; understand that racism is inextricably connected to sexism in prostitution; understand that prostitution is domestic violence, and in many instances -- slavery or debt bondage; and we also understand the need for asylum and culturally relevant treatment when considering escape or treatment options for those in prostitution. The perspective that prostitution is violence against women and other political perspectives on prostitution have been described and critiqued by Jeffreys (1997).

Another viewpoint considers prostitution to be an issue which primarily involves economic and sexual determination (Bell, 1994). Prostitutes' rights advocates understand prostitution as just another job, a vocation that they should have a choice to make, and as sexual liberation. Alexander (1996) commented on the advantage to the prostitutes' rights movement brought about by the AIDS epidemic. HIV has indirectly facilitated the growth of the commercial sex industry by funding outreach programs which provide sex workers with a safesex education, condoms, union-style organizing and by legitimizing prostitution as commercial sex work. Customers' anxieties about contracting HIV from those in prostitution has further created a vast pool of research and education monies. The contribution of this study to these differing perspectives will be discussed later.

Sexual and other physical violence is the normative experience for women in prostitution. This has been clinically noted by all four authors, and reported by others (Baldwin, 1992; Farley and Barkan, 1998; Hunter, 1994; McKeganey and Barnard, 1996; Silbert and Pines, 1982; Vanwesenbeeck, 1994). Noting 'everpresent' violence against 361 prostituted women in Glasgow, UK, McKeganey and Barnard (1996) described a range of violent behaviors against women in prostitution ranging from name-calling to physical assault, rape and murder. Of the prostituted women interviewed by Hoigard and Finstad (1992) in Norway, 73 percent were exposed to acts of violence -- physical assaults, rapes, confinement and threats of murder. The remaining 27 percent spoke of the extreme violence which had victimized their friends. The Council for Prostitution Alternatives in Portland, Oregon, USA, reported that prostituted women were raped about once a week (Hunter, 1994). A Canadian report on prostitution and pornography found that women and girls in prostitution had a mortality rate 40 times higher than the national average (Baldwin, 1992).

The diagnosis of PTSD describes psychological symptoms which result from violent trauma. In the language of the American Psychiatric Association (1994), PTSD can result when people have experienced:

. . . extreme traumatic stressors involving direct personal experience of an event that involves actual or threatened death or serious injury; or other threat to one's personal integrity; or witnessing an event that involves death, injury or a threat to the physical integrity of another person; or learning about unexpected or violent death, serious harm, or threat of death or injury experienced by a family member or other close associate.

In response to these events, the person with PTSD experiences fear and helplessness.

Exposure to any of these events may lead to the formation of symptoms of PTSD. These symptoms are grouped into three categories: symptoms of traumatic re-experiencing (items 1-4 in Table 1); efforts to avoid stimuli which are similar to the trauma as well as a general numbing of responsiveness (items 5-11); and symptoms of autonomic nervous system hyperarousal (items 12-17).

Authors of the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 1994) comment that PTSD may be especially severe or long-lasting when the stressor is of human design (e.g. torture, rape).

The following are two examples of symptoms of PTSD: many years after escaping from prostitution, an Okinawan woman who was purchased by US military personnel during the Vietnam War became extremely agitated and had flashbacks of sexual assaults on the 15th and 30th of each month -- those days which were US military paydays (Sturdevant and Stoltzfus, 1992). Describing symptoms which were ignored by her counselor, a survivor of prostitution from the USA stated:

I wonder why I keep going to therapists and telling them I can't sleep, and I have nightmares. They pass right over the fact that I was a prostitute and I was beaten with 2 x 4 boards, I had my fingers and toes broken by a pimp, and I was raped more than 30 times. Why do they ignore that? (Farley and Barkan, 1998).

The symptoms of PTSD may be cumulative over one's lifetime. Several studies report a positive correlation between a history of childhood sexual assault and symptoms of PTSD in adult women (Farley and Keaney, 1994, 1997; Rodriguez et al., 1997). Since almost all prostituted women have histories of childhood sexual abuse, this undoubtedly contributes to their current symptoms of post-traumatic stress.

Prolonged and repeated trauma usually precedes entry into prostitution. From 55 to 90 percent of those in prostitution report a childhood sexual abuse history (Bagley and Young, 1987; Belton, 1992; Farley and Barkan, 1998; Harlan et al., 1981; James and Meyerding, 1977; Silbert and Pines, 1981, 1983; Simons and Whitbeck, 1991). Silbert and Pines (1981, 1983) noted that 70 percent of their sample told them that the earlier sexual abuse had an influence on the later ,choice' to become a prostitute. A conservative estimate of the average age of recruitment into prostitution in the USA is 13-14 years (Silbert and Pines, 1982; Weisberg, 1985). Any distinction between prostitution of children and prostitution of adults is arbitrary, and obscures this lengthy history of trauma. The 15-year-old in prostitution eventually turns 18, but she has not suddenly made a new vocational choice. She simply continues to be exploited by customers and pimps.

A number of authors (e.g. Barry, 1995; Hoigard and Finstad, 1992; Leidholdt, 1993; Ross et al., 1990; Vanwesenbeeck, 1994) have described the psychological defenses which are necessitated by the experience of prostitution, and which frequently persist: splitting off certain kinds of awareness and memories, disembodiment, dissociation, amnesia, hiding one's real self (often until the nonprostituted self begins to blur), depersonalization, denial. One woman said, 'Only my head belongs to me now. I've left my body on the street' (Hoigard and Finstad, 1992).

Some have criticized the application of any psychiatric terminology to women who have been harmed by the sexism, racism or class prejudice which comprises male supremacy. There is an assumption that the oppression is dismissed once a psychiatric diagnosis is applied. Pejorative terms such as 'masochistic', 'deviant' or 'borderline' have certainly caused pain and sometimes additional harm to women in prostitution.

On the other hand, the concept of PTSD has been important in describing the psychological symptoms suffered by combat veterans, sexual abuse survivors, concentration camp survivors -- and it may also be useful in describing the psychological harm of prostitution. The authors' experience is that when the trauma-related symptoms of PTSD are discussed, survivors of prostitution feel less stigmatized, less 'crazy' and may experience relief at having their symptoms named. Attaching a name to symptoms associated with severe trauma makes it possible for survivors of prostitution to learn about their own and others' experiences. Further, it becomes possible for survivors to organize politically around their own health needs, just as gay men have organized around HIV.

The diagnosis of PTSD is a departure both from the biological viewpoint that psychological symptoms are essentially biochemical in origin and from the psychoanalytic notion that psychological symptoms result from intrapsychic turmoil. The diagnosis of PTSD requires an external stressor, clearly implying that psychological symptoms result from material conditions that oppress women.

The diagnosis of PTSD alone, however, does not completely articulate the extent of the psychological harm caused by prostitution. Over time, the constant violence of prostitution, the constant humiliation, and the social indignity and misogyny, result in personality changes. Herman (1992) described long-term changes in trauma survivors' emotional regulation, changes in consciousness, changes in self-perception, changes in perception of perpetrator(s), changes in relations with others, and changes in systems of meaning. These personality changes have been called complex PTSD by Herman and others. Describing prostitution, one woman said:

It's a process. The first year was like a big party, but eventually progressed downward to the emptiest void of hopelessness. I ended up desensitized, completely deadened, not able to have good feelings because I was on 'void' all the time.

Herman saw these symptoms as resulting from a history of subjection to totalitarian control over a prolonged period, and noted that organized sexual exploitation may be one cause of complex PTSD. The violence of pimps is aimed not only at punishment and control of women in prostitution, but at establishing their worthlessness and invisibility (Dworkin, 1997; Patterson, 1982). The hatred and contempt aimed at those in prostitution is ultimately internalized. The resulting self-hatred and lack of self-respect are extremely long-lasting.

Graham et al. (1994) have also described the psychological consequences of being in prostitution. The Stockholm syndrome -- a psychological strategy for survival in captivity -- is useful in explaining the traumatic bonding which occurs between women in prostitution and their pimps/captors. When a person holds life-or-death power over another, small kindnesses are perceived with immense gratitude. In order to survive on a day-to-day basis, it is necessary to deny the extent of harm which pimps and customers are capable of inflicting. Survival of the person in prostitution depends on her ability to predict others' behavior. So she develops a vigilant attention to the pimp's needs and may ultimately identify with his view of the world. This increases her chances for survival, as did Patty Hearst's identification with her captors' ideology. Graham described other behaviors which are typical of the Stockholm syndrome: extreme difficulty leaving one's captor and a long-term fear of retaliation.

Barry (1995) and Giobbe et al. (1990) estimate that at least 90 percent of prostitution is pimp-controlled. Sexual and physical abuse and torture are used by pimps to keep women from escaping prostitution (Barry, 1995; Dworkin, 1997; Hunter, 1994; MacKinnon, 1993). Pimps in Washington, DC, USA, employ ,catchers' -- thugs who stand guard at the borders of their turf and 'catch' girls trying to escape from prostitution (Michelle J. Anderson, personal communication, 1996).

Houskamp and Foy (199 1) suggested that a primary etiological variable for the development of PTSD in battered women is the extent of violence to which they have been subjected. Giobbe et al. (1990) compared prostitution to other forms of domestic violence. They described methods of coercive control exercised by pimps and customers over women in prostitution which are identical to the methods used by battering men to control women: isolation, verbal abuse, economic control, threats and physical intimidation, denial of harm and sexual assault used as a means of control.

Although the incidence of PTSD has been investigated among battered women, and ranges from 45 percent to as high as 84 percent (Houskamp and Foy, 1991; Kemp et al., 1991; Saunders, 1994) -- the frequency of the diagnosis has not been investigated among prostituted women, who are exposed to the same violence as battered women.

PTSD has been assessed in people from non-Western cultures, such as Southeast Asian refugees, Latin American disaster survivors, Navajo and Sioux Vietnam veterans (Marsella et al., 1996). A recent review of the PTSD literature noted that its major limitation is that many of the most traumatized populations have not been studied (de Girolamo and McFarlane, 1996; Keane et al., 1996). Our study documents, across several cultures, some of the violence and traumatic stress which result from being prostituted.

METHOD

Brief structured interviews of people in prostitution were conducted in San Francisco, CA, USA; in two cities in Thailand; in Lusaka, Zambia; in Istanbul, Turkey; and in two cities in South Africa. These countries were included in the study in part because of the first author's wish to include a majority of women of color, since globally prostitution exploits vast numbers of women of color. Second, all four authors shared a commitment to the project of documenting the experiences of women in prostitution, and to providing options for escape.

If respondents indicated that they were working as prostitutes, they were asked to respond to a 23-item questionnaire which asked about the following: physical and sexual assault in prostitution; lifetime history of physical and sexual violence; and the use of or making of pornography during prostitution. The questionnaire asked whether respondents wished to leave prostitution and what they needed in order to leave. We asked if they had been homeless; if they had physical health problems; and if they had a problem with drugs or alcohol or both.

Respondents also completed the PCL, a 17-item scale which assesses DSM-IV symptoms of PTSD (Weathers et al., 1993). Respondents were asked to rate the 17 symptoms of PTSD (see Table 1) on a scale where: I = not at all; 2 = a little bit; 3 = moderately; 4 = quite a bit; and 5 = extremely. Weathers et al. (1993) report PCL test-retest reliability of .96; internal consistency, as measured by an alpha coefficient, was .97 for all 17 items. Validity of the scale is reflected in its strong correlations with the Mississippi Scale (.93); the PK scale of the MMPI-2 (.77); and the Impact of Event Scale (.90). The PCL has functioned comparably across ethnic subcultures in the USA (Keane et al., 1996).

We measured symptoms of PTSD in three ways. First, using a procedure established by the authors of the scale, we generated a measure of overall PTSD symptom severity by summing respondents' ratings across all 17 items.

Second, using Weathers et al.'s (1993) scoring suggestion, we considered a score of 3 or above on a given PCL item to be a symptom of PTSD. Using those scores of 3 or above, we then noted whether each respondent met criteria for a diagnosis of PTSD. PTSD consists of three kinds of symptoms: persistent, intrusive re-experiencing of trauma (B symptoms); numbing of responsiveness and persistent avoidance of stimuli associated with trauma (C symptoms); and persistent autonomic hyperarousal (D symptoms). A diagnosis of PTSD requires at least one B symptom, three C symptoms, and two D symptoms. We report the numbers and percentages of respondents who qualified for a diagnosis of PTSD in each country.

Third, we measured partial PTSD, following Houskamp and Foy (1991) who investigated PTSD among battered women. These authors suggested that if a person meets at least two of the three foregoing criteria for PTSD, a significant degree of psychological impairment exists. We report the numbers and percentages of respondents who qualified for a diagnosis of partial PTSD in each country.

The two questionnaires were translated into Thai and Turkish. In Zambia, interviewers translated verbally as needed -most participants spoke some English. The authors either administered or directly supervised the administration of all questionnaires.

In San Francisco, we interviewed 130 respondents on the street who verbally confirmed that they were prostituting. We interviewed respondents in four different areas in San Francisco where people worked as prostitutes.

In Thailand, we interviewed several of the I 10 respondents on the street, but found that pimps did not allow those they controlled to answer our questions. We interviewed some respondents at a beauty parlor which offered a supportive atmosphere. The majority of the Thai respondents were interviewed at an agency in northern Thailand that offered nonjudgmental support and job training.

We interviewed 68 prostituted people in Johannesburg and Capetown, South Africa, in brothels, on the street and at a drop-in center.

We interviewed 117 women currently and formerly prostituted at TASINTHA in Lusaka, Zambia. TASINTHA is a nongovernmental organization which offers food, vocational training and community to approximately 600 prostituted women a week.

In Turkey, some women work legally in brothels which are privately owned and controlled by local commissions composed of physicians, police and others who are 'in charge of public morality'. We were not permitted to interview women in brothels, so we interviewed 50 prostituted women who were brought to a hospital in Istanbul by police for the purpose of venereal disease control.

In two of the five countries, respondents were racially diverse. In the USA, 39 percent (51) of the 130 interviewees were white European/American, 33 percent (43) were African American, 18 percent (24) were Latina, 6 percent (8) were Asian or Pacific Islander and 5 percent (4) described themselves as of mixed race or left the question blank. In South Africa, 50 percent (34) were white European; 29 percent (20) were African or Black; 12 percent (8) described themselves as Colored or Brown or of mixed race; 3 percent (2) were Indian; 6 percent (4) left the question blank.

We included transgendered people in this study because they represent a significant minority of those in prostitution. A previous study (Farley and Barkan, 1998) found that transgendered persons experienced the same degree of physical assaults and rapes as did women in prostitution. These authors concluded that to be female, or to appear female, was to be targeted for violence in prostitution. In Turkey and Zambia, all respondents were women. Table 2 below describes gender of respondents in South Africa, Thailand, and the USA.

     Across the five countries, the average age was 28 years, ranging from 12 to 61 years. See Table 3.

Sample Selection Differences and Limitations to Generalizability

The most daunting challenge in cross-cultural research is sample selection. Were the 475 people we interviewed representative of all women in prostitution? We attempted, as McKeganey and Barnard (1996) did in Glasgow, UK, to contact as broad as possible a range of those in prostitution: women of diverse races, cultures, ages, location where working, and including gender differences. However, 'there is quite simply no such thing as a representative sample of women selling sex' (McKeganey and Barnard, 1996). Given the illegality of prostitution in most places, it was necessary to interview those people to whom we had access. In most cases, researchers have access only to people prostituting on the street. We were fortunate that, in South Africa, we were able to interview 25 people in brothel prostitution. Should it become possible to interview those in massage parlors, escort services, stripping, or others who are in brothel prostitution -- the authors would immediately include these people in a future expansion of this work. We will also share our questionnaires with researchers who have access to these groups of people.

There were differences in the ways the samples were selected. In all cases, we interviewed people who were either currently in prostitution or had recently been prostituted. In the USA and South Africa, all were currently prostituting, whereas a higher proportion of respondents in Thailand and Zambia were actively attempting to leave prostitution and find other employment. Respondents in Turkey were interviewed after they were brought to a clinic by police for STD testing. In Istanbul, as elsewhere in this study, women in prostitution were freely offered STD testing, but other acute and chronic health problems were rarely addressed.