Abstract
Paid employment for women has been one factor under consideration as we search for a comprehensive theory to explain women's greater psychiatric and physical morbidities, yet greater longevity than men in countries with established economies [1, 2]. The morbidity difference may be less than has been previously reported due to a number of confounding variables [3–5]. Several writers have linked work conditions to health, sometimes directly, sometimes indirectly through inadequate income and stressful work conditions [6]. The adverse health consequences of living in poverty have been frequently documented [6–9]. Other writers have focused on the particular risks found in work typically undertaken by women [10]. These include physical hazards, psychological pressures, shift work, lack of control, responsibility without authority, boring repetition, sexual harassment and violence [11, 12]. One recent report found that work characteristics, particularly decision-making latitude, more strongly predicted wellbeing than the social support received in the workplace [13]. Social networks and support, self-esteem and stress buffering are closely linked to work conditions and, for many women, paid work is often associated with better social networks [14–17].
One controversial type of work undertaken by women is sex work or prostitution, which appears to have been always subject to regulation by law or custom. The nature of legal and social control over sex-worker activities has varied over time and between countries. In New Zealand where this study was conducted, the legislation is a confusing mix; prostitution itself is not illegal, but its consequences are, which produces the same result. It is illegal to solicit, to run a brothel or to live off the earnings of prostitution of another person. This has been summarized as ‘being illegal to offer sex for money, but not money for sex’. Massage parlours as sex centres are tightly regulated.
Because sex work is quasi-legal, it lies outside the usual social protections for occupational health concerns [18]. This may mean that the medical, legal and welfare needs of sex workers are not properly addressed. Public health action has often been driven by the narrow focus of limiting HIV-AIDS rather than by general considerations of the health and wellbeing of the workers themselves [18–20]. Very few published reports give weight to sex workers' own views of these important themes [21]. Much of the theorizing about sex work reveals the prejudices of the writer; good data to support or refute the claims made are hard to find.
As a marginalized employment group, sex workers might be expected to experience poor health parameters, both because they may be exposed to greater health risks and because they may have problems in obtaining good medical care. We use the term ‘sex worker’ rather than prostitution as it seems more neutral and is the term often preferred by the workers themselves.
We report here a preliminary survey of New Zealand female sex workers, which was designed with awareness of the many methodological problems of previous published reports [22–24]. These included the definition of sex worker, strategies for recruiting subjects, choice of an appropriate comparison group, small samples and anecdotal evidence, as well as the ethical aspects of such research. This study overcame most, but not all, of these challenges.
This is the second of two reports from this study. In the first, we reported that the women undertaking sex work were more likely to grow up in families with multiple difficulties and to recall poor relationships with their parents [25].
Method
Subjects were recruited in two urban New Zealand centres, Dunedin and Wellington, through the New Zealand Prostitutes Collective (NZPC), a support and industry lobby group set up in 1987 by a group of sex workers who considered that there was a need to improve their working conditions. There is no official list maintained of those involved in sex work. The Ministry of Health funds the NZPC. It provides education and support for sex workers with the specific goal of limiting the spread of HIV-AIDS in New Zealand. The NZPC identified the first subjects who then, using a snowball sampling method, suggested others who they thought would be interested in participating in the research. Maintaining subjects' anonymity was a prime concern at all stages of the project. The local Health Funding Agency's Ethics committee, who agreed that verbal rather than written consent would be adequate, granted ethical permission. Further, the women could use their ‘working’ names during the interviews and not disclose their real names.
The results for the sex-worker subjects were compared with community data from the two Otago Women's Health surveys using information from women aged 18–40 years only: the age range of the sex worker sample. For most items, the Otago Women's Health Child Sexual Abuse survey (OWHCSA) data were used. This 1989 community project assessed the prevalence and consequences of childhood sexual abuse on later psychological function by comparing sexually abused and non-abused women, using a two-stage method (postal questionnaire, n = 680 in the 18–40 age bracket, then personal interview n = 258) [26]. The same structured instruments were used in the sex workers study. Physical abuse was defined in the same way as in OWHCSA, categorized into two groups: abuse occurring in childhood (before the age of 16); and those later (16 years or older). Smoking was not studied. For cigarette smoking only, the earlier 1985 OWHS study provided a community comparison figure [27]. This also had used a two-stage (postal questionnaire, then interview) method with a random community sample (at interview n = 314).
KP interviewed the women at the place of their choice in 1994. The interview contained sections covering sociodemographic information, family background, work history, health attitudes and behaviours and abuse experiences. These were abbreviated from the OWHCSA survey, with an expanded section on work history specifically related to sex work. The women were asked to complete the General Health Questionnaire (GHQ-28) [28] to measure the psychiatric symptomatology and the Intimate Bonding Instrument (IBM) [29], to assess the quality of their relationship with their intimate partner, if they had one. The IBM generates two indices, care and control; high-care and lowcontrol relationships are optimal. Self-esteem was assessed with a single item taken from the Present State Examination ‘What is your opinion of yourself compared to others?’ Ratings of overall relationship satisfaction, social support satisfaction, quality of physical health and level of comfort with medical care, were each assessed with a single item, using a six-point Likert scale. The socioeconomic status (SES) backgrounds of the sex workers were determined by assessing the Elley-Irving socioeconomic status of the main family wage earner divided into high, medium and low SES levels on education and income [30]. A combination of open- and closed-ended questions was used to collect information from the women about their experiences working in the New Zealand sex industry.
Statistical analyses
The women's mean GHQ-28 score was compared with that of the comparison women using Student's t-test. The IBM scores were split into high and low scores using cut-offs of 28 or more on the care scale and a six or less on the control scale (Wilhelm K, personal communication). The two subscales for the IBM were combined to generate quadrants, with the high-care/low-control quadrant describing a good quality relationship with her partner. These and other categorical variables were compared with ?2 analyses.
Results
Sample
Twenty-nine sex workers were interviewed, 15 in Dunedin and 14 in Wellington. Their ages ranged from 16 to 47 years with a mean age of 28.7 years (SD = 7.6). The 680 women from the OWHCSA had an age range of 18–40 years, with a mean age of 29.1 (SD = 6.2). There was no difference in the socioeconomic status of their families of origin. Of the 28 sex workers with data available, 15 (54%) came from families in SES classes 1–3, the high bracket, compared with 288 (44%) of the 651 controls (?2 = 0.61, df = 1).
At the time of the interview, 17 women (59%) were working in massage parlours, two on the street and one for an escort agency. The remaining nine (31%) worked privately, either through advertising (n = 5) or seeing regular clients (n = 4).
Partner status
Two-thirds of the sex workers reported that they were currently in a ‘close or intimate’ relationship (n = 19, 66%), which was not statistically different from the OWHCSA control subjects (536/680, 79%). Ten of these women were living with their partners, and four were married. When compared with the OWHCSA group of women, sex workers were less likely to be living with their nominated partner (10/29, 34% vs 423/680, 62%; ?2 = 8.1, df = 1, p < 0.01). Of those who were cohabiting, sex workers were much more likely to be in a de facto relationship rather than married when compared with the OWHCSA subjects (6/10, 60% vs 50/373, 13%; ?2 = 16.1, df = 1, p < 0.0001). Eight sex workers were separated or divorced; this proportion (8/29, 28%) was greater than the control OWHCSA group (41/680, 6%; ?2 = 18.5, df = 1, p < 0.0001).
Motherhood
Half the sex-workers group (n = 15) reported that they were mothers, not different to the control sample. The mean number of children was 2.3, with a range from one to four. Two-thirds of these women had custody of their children the majority of the time. Twelve (80%) reported using their income to support their children.
Accommodation
Most sex workers were living with other people, either partners (35%), relatives (17%), flatmates (21%) or their children (38%). Five (17%) lived on their own. Sex workers, as a group, were more likely than the OWHCSA women to be living on their own (?2 = 5.5, df = 1, p < 0.05) and less likely to be living with their partner, as noted earlier.
Intimate relationships
The mean overall relationship satisfaction score of the sex worker sample did not differ statistically from the OWHCSA comparison group (proportion scoring 3 or more: 24/29, 83% vs 540/666, 81%; x 2 = 0.05, df = 1).
Intimate Bond Measure
There was no difference in the proportion of sex workers and control subjects in an intimate relationship who assessed its quality to be high care and low control (sex worker 8/19, 42% vs controls 116/234, 50%; x 2 = 0.15, df = 1, p = 0.7).
Social support
Over 90% of the sex workers interviewed indicated that they had someone to talk to if anything was troubling them (n = 27). Eight women each named their partner or a female friend (28%) as being their main confidant. One woman indicated that she would talk mainly to a parent and five thought that they would either confide in a fellow sex worker (17%) or another person (17%, other relative, male friend or counsellors). No one named a child or a sibling as their main confidant. There was no difference in the mean social support scores between the sex workers (1 = excellent support, 6 = poor support: mean score = 2.3, SD = 1.2) and the OWHCSA women (mean score = 2.2, SD = 1.2). Over 80% of both groups of women scored a 3 or less, indicating satisfaction with their level of social support. There was no indication that a subject's reported satisfaction with her social group was associated with the number of people outside the sex industry who knew about the nature of her work.
Physical health
One-quarter of the women rated their health as excellent (n = 7) and four-fifths (79%) of the group perceived their general health positively, scoring a 3 or lower. The mean score was 2.4 (SD = 1.1). These did not differ from the comparison results.
Tobacco
Three-quarters of the sex worker subjects (n = 22, 76%) reported that they smoked cigarettes, and a further two said they had quit smoking within the previous month (24/29, 83%). This was a much higher proportion than the 1985 comparison sample, 29% of who smoked (?2 = 20.9, df = 1, p < 0.001). Sex workers were no more likely to be heavy smokers (20 cigarettes or more per day) than the comparison group. There was no association between the age of the sex worker and smoking.
Alcohol
There were no significant differences in the frequency with which the groups drank alcohol, although the sex workers drank more on the occasions when they did consume alcohol. Fifteen of the 26 sex workers (58%) estimated that they drank five or more standard units in a session whereas 144/639, 23% of the comparison group did so (x 2 = 15.1, df = 1, p = 0.0001).
Street drugs
Fourteen women (48%) reported use of street drugs: 10 marijuana only and four a range of other substances. Three (10%) were involved in a methadone programme. No questions were asked about intravenous drug use. There were no comparison figures available.
Mental health
Self-esteem
The majority of the sex workers (90%) reported that they generally felt as good as (n = 21) or better than most other people (n = 5). Only three sex workers (10%) reported low self-esteem, not different to the OWHCSA group (9%). Almost 90% reported that their feeling of selfworth had increased over the past 5 years (n = 18). There was no association between work characteristics (the length of time a subject had been working in the sex industry, the number of hours she worked, the number of clients she saw a week) or income and self-esteem.
General Health Questionnaire
The mean GHQ-28 score for the sex worker group was 5.3 (SD = 5.7), not statistically different from the OWHCSA community sample of 3.5 (SD = 4.7, t = 1.7, df = 30, p = 0.1).
No significant associations were found between the subjects' mean GHQ scores and their age or reported marital status, family SES score, level of education, alcohol or tobacco use. Sex workers with poor physical health had higher GHQ scores than those with good health (t = 8.34, df = 4, p < 0.001). High satisfaction with their available social support was associated with lower GHQ-28 scores (t = 4.49, df = 4, p = 0.008). Women working in massage parlours, or privately with regular, known clients, had lower mean GHQ-28 scores than women working in other areas of the sex industry (t = −2.12, df = 27, p = 0.04).
Relationships with health professionals
Almost 90% of the women interviewed reported that they had one particular doctor whom they usually visited (n = 25), and for 13 sex workers this doctor was female. Over half of the women (n = 15) rated their level of comfort discussing their general health problems with their general practitioners as ‘very comfortable’ (score of 1 of 6), with most (24, 83%) scoring 3 or less. However, one-third of the women said that their doctors did not know of their involvement in the sex industry (n = 9). One-third said they would approach their general practitioner if faced with a sexual health problem, one-fifth would use an independent sexual health clinic. Half the sex workers interviewed in the area where the NZPC itself employed a doctor would consult this professional. Almost half of the subjects had seen a counsellor at some point in their lives (n = 14) and one-third had seen a psychologist (n = 9). One-fifth of the subjects (n = 6, 20%) reported that they had seen a psychiatrist at least once; a proportion which did not differ from the OWHCSA subjects (14%, ?2 = 0.40, df = 1, NS).
Adult abuse experiences
Physical abuse
Over half of the sex workers (59%) reported that they had experienced at least one incident of physical abuse after they had reached the age of 16 years (n = 17). This was significantly higher than the 20% of the OWHCSA comparison group who had experienced similar abuse (?2 = 17.2, df = 1, p < 0.001). The majority of these 17 subjects had been deliberately hit, pushed or kicked by one or more individuals (n = 15), and five subjects had been hurt with a weapon or an object. Only two sex workers reported being physically assaulted by a client (7% of the sample).
Sexual assault
Over half of the sex workers interviewed (16, 55%) reported that someone had tried to have sex with them against their will after the age of 16 (n = 16). Adult sexual abuse was higher among the sex workers than the OWHCSA comparison group, of whom 12.8% had experienced this type of sexual assault as adults (?2 = 30.15, df = 1, p < 0.001). Half these sexual abusers in adulthood were previously known to either the subject or her family (n = 10). Five assailants were boyfriends or partners. Five women reported that they had been sexually assaulted by one individual previously unknown to them and two women had been assaulted by a group of strangers. Four sex workers (14%) reported that a client had sexually assaulted them.
Sex work experience
Subjects started sex work at a mean age of 23 (SD = 7, range = 15–40 years). Most began sex work before age 25 (72%) with one-third before 19 years. Prior to sex work, one-third had held another paid job excluding study; when tertiary study was included, all except four (86%) had worked in areas other than sex work. Seventeen women knew another sex worker prior to starting sex work and eight responded to her suggestion that they too should undertake this work. Three women were introduced into sex work by their boyfriends, who arranged their first job. Of the 29 women interviewed, 24 (83%) were actively working in the sex industry at the time of the study.
Most subjects gave financial reasons as their major work incentive (n = 23). One woman had been pressured into beginning sex work by her partner, and the remaining five women gave a variety of reasons for their initial decision to begin sex work, including curiosity, boredom with a current occupation and enjoyment of sex. Half the currently active sex workers also had concurrent other occupations or responsibilities (n = 12); these included childcare, study and a range of partand full-time jobs.
The sex workers estimated that they worked a mean of 33 h (SD = 17) a week, with 40% less than 30 h and 20% more than 50 h.
Most (n = 26) saw fewer than 20 clients per week, with 12 averaging between 11 and 20 clients a week. Thirteen were using their income to support someone, usually a child or children (n = 12). None reported supporting a partner or a parent. Nine reported saving regularly.
Stopping and restarting sex work was common. Most had ‘taken a break’ from sex work for 3 consecutive months or more (n = 15). Reasons given included pregnancy (n = 3), becoming fed up with sex work (n = 3) and no longer needing the income that they received from sex work (n = 3). Two women were pressured by a partner to stop work and two women felt they needed a break because they were emotionally and/or physically fatigued. The main reason for returning to sex work after a long break from it was the same as that given for entering the sex industry initially: financial.
Most subjects considered that they were well informed about the laws regulating sex work (n = 21, 72%). Over half of the women said they had received information about the laws governing sex work from the NZPC (n = 16). Co-workers were also a source of information for 45% of the subject group (n = 13).
Most sex workers interviewed thought that there were difficulties associated with working in the sex industry (n = 24). These included specific aspects of sex work, such as clients being uneducated about safe sexual practices, and also more general problems arising as a result of the marginalized position of sex work and sex workers in society. Eight women reported difficulties with their personal relationships arising from their work, seven negative attitudes or ignorance (n = 2) from the general public as a problem with sex work. Thirteen noted threats to personal wellbeing, health or safety.
Five women reported that most or all of their family and friends knew about their work. However, these women seemed to be the exception rather than the rule. All but two of the women who were in relationships (n = 17) reported that their partners knew that they did sex work and almost 80% of the subjects (n = 23) said that their close friends knew about their work. Seven reported that their mothers or fathers knew. Only two sex workers reported that their children knew. The basic reason that women gave for non-disclosure was their assumption that the general public had a negative attitude towards sex work and sex workers.
Discussion
The theoretical framework of this study was that sex work is authentic work engaged in by some women as one occupation among a number available to them. We were interested in the impact that working in a stigmatized job might have on sex workers' physical and mental health. On balance, the data presented here do not provide any convincing evidence that these sex workers have poorer physical or mental health, lower self-esteem or impaired social relationships. They did report experiencing more assaults both physical and sexual as adults and drank and smoked more. Although a number of sex workers made positive comments about their work, the majority identified difficulties associated with working in the sex industry, its illegal status and the associated stigma.
Little previous comment on the health of sex workers has been informed by data. This study has overcome some of the deficiencies of previous research. Reliable instruments were used wherever possible. The large comparison data sets permitted comparison of the sex worker data with a robust sample randomly selected from the general community. The small number of sex workers in the present study (n = 29) raises questions about the statistical significance of some results; a type two error may be present. We estimate the power of this study, with only 29 sex worker subjects, but a 1:23 ratio to the control subjects, to be able to detect a 25% difference when the base rate is 50%, and an alpha of 0.05. Future research projects should aim to recruit a larger representative sample of sex workers while retaining the control group methodology.
The major potential limitation in this study comes from the non-random sampling of sex workers. A truly representative sample of sex workers will be difficult to assemble, as long as sex work remains marginalized and often outlawed. From a qualitative research perspective, this is less important, as the data represent the reality for those who chose to participate. Care must be taken therefore, in generalizing these results to all female sex workers. It is possible that by using workers linked with the NZPC, we obtained a systematic bias. This is hard to evaluate; our sample may be better informed about risks of HIV-AIDS and safe sex practices because of their link with NZPC. Future research could benefit by using prospective designs and following a carefully delineated cohort of sex workers over time. The higher GHQ-28 score for sex workers not working in a massage parlour or with regular clients suggests they may be a subgroup with particular problems and should be involved in future health research.
In the earlier companion paper, this group of sex workers was reported as entering adult life carrying increased burdens of greater childhood disadvantages (single-parent families, parental disputes, childhood physical and sexual abuse, poor relationships with parents), early age of leaving home and first pregnancy [25]. Yet, it seems that the sex workers may have overcome these potentially significant disadvantages to a significant extent by adult life. There were no objective indicators of impaired intimacy with current partners on the IBM, no statistical differences in the GHQ-28 scores, or selfesteem, and no increase in consultation rates with a psychiatrist. The sex worker sample rated their physical health identically to women with a similar age profile. Overall, despite any personal problems arising from the particular nature of their work, these women described as adequate relationships with partners, friends and work colleagues. Two papers have suggested that sex workers have high rates of psychological symptoms; however, both used convenience samples and lacked a comparison group making interpretation of the results difficult [19, 20]. Our results challenge prevailing ideas that sex work and psychiatric morbidity are inevitably associated.
Sex workers were exposed to higher rates of violence in their adult lives. As with control women, most of this was at the hands of their partners, with some arising from the sex work itself. Sex-work violence has been discussed with respect to the illegal nature of sex work with the suggestion being made that those who rape sex workers assume them to be sexually available on demand [31]. Responses from statutory authorities have often reinforced these views.
Sex workers smoked more frequently than did the comparison group; this significant health concern needs addressing. They also appeared constrained against speaking freely with some parts of their social networks (children and parents) about their work although this did not seem to affect adversely their experience of social support. Finally, although most described good relationships with their general practitioners, in one-third of cases the general practitioner was not aware of her occupation. This ignorance is likely to impair the nature of medical advice the sex worker would receive. If general practitioners are unaware of their patients' work, they will not be able to assess occupational hazards accurately and this in turn will impede the comprehensive health evaluation.
We would not want these results to add to the difficulties of these women. We recall the comment by Shaw in his lengthy preface to Mrs Warren's Profession that he wrote the play ‘to draw attention to the truth that prostitution is caused, not by female depravity and male licentiousness, but simply by underpaying, undervaluing and overworking women’ [32].
Footnotes
Acknowledgements
Kathleen Potter was in receipt of a junior research fellowship from the Health Research Council of New Zealand, in 1994, during the time she undertook the field work reported here. The authors thank the New Zealand Prostitutes Collective for the support given to the study and for comments on an earlier draft of this paper. We wish to note that the opinions expressed here are those of the authors alone.