Commercial Sex Between Men: A Prospective Diary-Based Study - Statistical Data Included

A growing body of literature has emerged in the last decade which provides a profile of the male sex worker (MSW) and attempts to explain the factors associated with paid safe-sex encounters between men (Bloor, Barnard, Finlay, & McKeganey, 1993; Bloor, McKeganey, Finlay, & Barnard, 1992; DeGraaf, Vanwesenbeeck, van Zessen, Straver, & Visser, 1994; 1995; Earls & David, 1989; Elifson, Boles, & Sweat, 1993; Estep, Waldorf, & Marotta, 1992; Ford, Wirawan, Fajans, & Thorpe, 1995; Ford, Wirawan, & Fajans, 1993; McKeganey, Barnard, & Bloor, 1990; Morse, Simon, Osdofsky, Balson, & Gaumer, 1991; Pleak & Meyer-Bahlburg, 1990). This literature generally shows that not all MSWs identify themselves as homosexual (Morse et al., 1991; Robinson & Davies, 1991), that MSWs are no less educated than other men of the same age group (Earls & David, 1989), that there exists a growing diversity of sex work according to place of work and type of service provided (Visano, 1991; Waldorf, Murphy, Lauderback, Reinarman, & Marotta, 1990), and that there is a high rate of condom use among sex workers with their clients (Perkins, Prestage, Lovejoy, & Sharp, 1993).

Few of these studies, however, provide a detailed analysis of the specific work context of such sexual encounters. For example, while Pleak and Meyer-Bahlburg (1990) reported that substance abuse in their sample of MSWs was high, we do not know whether MSWs consume alcohol and/or drugs immediately before and during the paid sexual encounter. Obtaining a clearer picture of the context of the events which occur during each specific encounter is important because recent studies show that the commercial sex encounter takes place in a setting of exchange relations which are not static, and where specific situational contextual factors can modify the interaction (Browne & Minichiello, 1995). Interestingly, current studies have relied on methods such as in-depth interviews or questionnaires, which provide at best only a retrospective and average account of events which are not necessarily contextual to all client/sex worker situations (Browne & Minichiello, 1996; DeGraaf et al., 1995; Earls & Davis, 1989; Elifson et al., 1993; Estep et al., 1992; McKeganey, 1994). Another shortcoming of the existing studies is that questionnaires have used different periods of recall, usually asking respondents to report on incidents which have happened over the past month or the year prior to the study. While diaries have been used on various populations to study sexual behaviour (Marino, Minichiello, & Browne, 1999), few researchers have asked sex workers to record the details of each of their paid sexual encounters using this data collection method. Two studies have used diaries to study the female sex industry. Picketing, Dunn, Pepin, and Wilkins (1992) used diaries to obtain information on the travel and sex-work patterns, including number of clients, the sexual acts performed, and use of condoms, of 248 female sex workers in rural areas of Gambia. In another study, Fox et al. (1993) compared the self-reported use of condoms with condom use recorded in diaries from a sample of 134 female sex workers from Honduras.

This paper provides a descriptive profile of the context of the male-to-male commercial sex encounter reported by male sex workers from three Australian cities using a diary over a 2-week period. Specifically, it describes what sex activities were carried out, the when and where context, and other contextual activities surrounding the interaction such as alcohol and drug use, in an attempt to fill a gap in the current knowledge about male-to-male commercial sex.

METHODS

Instrument

The data reported in this study were collected using a diary which male sex workers (MSWs) completed after each commercial sexual encounter with a male client over a 2-week period. The diary consisted of a total of 36 questions covering four topic areas. The topic areas included (a) the work context of the paid encounter (i.e., time, place of the interaction, form of payment); (b) the characteristics of the sexual encounter itself (i.e., type of sex acts, condom usage, client/worker sex outcome); (c) the context of the interaction (i.e., alcohol usage, drug usage, amount of time spent talking/having sex); and (d) the sex worker's perception of the client/sexual encounter (i.e., perceived safety of the encounter). The instrument was developed after consultations with three sex workers' organisations: the Sex Workers Outreach Project (SWOP) in Sydney, the Self-health Queensland Workers in the Sex Industry (SQWISI) in Brisbane, and the Prostitutes Collective Victoria (PCV) in Melbourne, and was pilot tested on sex workers. The pilot test showed that the instrument had a high level of face validity and could be completed quickly and without much difficulty. The diary was pocket sized and the response format, with few exceptions, required no more than several tick marks. On average it took the sex workers less than 5 minutes to complete the instrument.

Procedures

Outreach workers from sex workers organisations in Sydney, Melbourne, and Brisbane, the three fieldwork sites, were employed to coordinate the collection of the data. The outreach workers provided the MSWs with two diaries, one for each week, and collected a diary at the end of each week. The study was conducted in the fall of 1998. A 2-week period was selected as the time frame. A shorter period possibly could have included atypical working days and patterns of work (i.e., weekends as opposed to weekdays), and a longer time period could have reduced the likelihood of participation by the MSWs in the study.

The protocol of this study was approved by the Ethics and Research Committee of the University of New England in Armidale, New South Wales, Australia. Confidentiality was assured at all times and there was no identifying information sought. Following an explanation of the purpose and procedures of the study, subjects were invited to give written consent to participate. Instructions on how to complete the diary were provided to the participants by the outreach workers. MSWs were asked to record the details of each paid sexual encounter in the diary as close as possible to when the event occurred. Participants were paid the equivalent of US$50 for their transportation expenses and work time lost while participating in the study.

Sample

The sampling strategy for recruiting MSWs in the various categories (street, escort, and independent) was developed in consultation with three local sex workers organisations. Potential participants in the independent male sex-worker category (IMSW) were identified by reviewing the popular gay press where male sex-workers advertise. These included Brother/Sister Brisbane, Melbourne Star Observer and Brother/Sister Melbourne, and Sydney's Capital Q and Sydney Star Observer. For male sex-workers who work through licensed escort agencies (EMSW), the field workers contacted all of the escort establishments in each city (N = 5; 2 in Sydney and 3 in Melbourne, but none in Brisbane, because there is no establishment in that city with MSWs). Street sex-workers (SMSW) were approached using convenience samples recruited from areas where male sex-workers are known to be found by the outreach sex work organisations, such as the suburbs of St. Kilda in Melbourne, "The Wall" in Sydney, and Albert Park in Brisbane.

The sample was determined using a 3 x 3 sampling design (3 cities, 3 sex-work modalities). Using this model, and after consultation with the sex-workers organisations in each of the cities, it was anticipated that the expected sample would be 190 MSWs: 80 from Sydney, 70 from Melbourne, and 40 from Brisbane. These three cities were chosen because they were regarded as the areas of greatest concentration of male sex-workers, and they have a distinctive pattern of sex work due to, among other things, the different legal status of sex work on each location (Neave, 1994). It was decided that 40 participants would be from the IMSWs group, 45 from the EMSWs group, and 105 from the SMSWs group. Outreach workers, who as part of their work monitor this population, estimated that these figures would capture between 50% and 70% of the total MSW population working in Brisbane, Melbourne, and Sydney.

Sex work in Australia varies across state borders, and may be legal, illegal, or decriminalised. In New South Wales, sex work is legal and workers can operate without tear of arrest; subsequently, all forms of sex work occur, including street, brothel, private advertisement, and escort agency work. Comparatively, in Queensland (Brisbane) the only form of legal sex work is the private operator. Brothels, street sex work, and being a sex-work client are all illegal, hence the predominance of the private sex worker in Brisbane, although some workers still take the risk of operating illegally on the street. In Victoria (Melbourne), sex work is legal within licensed brothels. Planning permits for brothels are obtained from local councils, and this can be a protracted and expensive exercise for those wishing to establish such premises. Sex work other than within a licensed premises is illegal. Despite regular crack downs on street sex work, workers continue to solicit on the streets. It is also an offence to live off the earnings of illegal sex work. In Victoria the police have automatic right of entry to a licensed operation and can do so outside of business hours if they suspect that the delay in obtaining a warrant will result in a loss of evidence.

It is important to note that a unique aspect of the Australian sex work industry is that the government proactively funds a number of programs that have specific public health mandates to offer health promotion and peer education programs for male sex workers. In all states some form of sexually transmissible infections (STIs) prevention activities target male sex workers. These range from printed material promoting safer sex as a means of preventing STIs to highly coordinated community projects that employ workers who provide peer education, as well as other essential social services, service delivery, and supportive networks. Cities like Sydney, Melbourne, and Brisbane have sex worker organisations promoting the rights of sex workers and the professionalisation of the sex work industry.

Data Analysis

The analysis included 13 variables which describe the commercial sex encounter. Four variables describe the Where and When of the encounter (source of clients, place of the job, day of the week, and starting time), and one variable describes the What (the sexual practices that took place during the encounter). Seven variables describe the How of the encounter (time spent talking, payments received, time of the payment, whether the clients were checked for sexually transmissible infections (STIs), use of safe-sex equipment, duration, and safe-sex outcome of the encounter). Additionally, a group of variables describe alcohol and substance use both before and during the encounter, and the location of the encounters.

Location was recorded as Sydney, Melbourne, or Brisbane. Participants were asked about alcohol and type of substance use before and during the encounter, including tobacco, marijuana, heroin, speed, ecstasy, pain killers, nitrites, and benzodiazepines. Source of clients was recorded as agencies, advertisement, another MSW, another client, regular clients, and other. Place of the job was coded into six groups: agencies, MSW's place, client's place, hotel/motel, car, and others, which included parks and public toilets. The starting time of the job was classified into eight groups of 3 hours each, starting from midnight. Participants were asked to report how much time during the encounter was spent talking and coded as follows: none, 1/4 of the encounter, 1/2 of the encounter, and most of the time. Time of payment was recorded using four categories: before the encounter, after the encounter, during the encounter, and not paid. Payment was grouped into six categories: money, which included cash, check, and credit cards; drugs; money and drugs; money and others, which included items such as clothing and watches; or not paid. Participants were also asked if they checked their clients for STIs, and answers to this questions were recoded as yes, no, and no sex. Participants were asked if they used condoms, rubber dams, or rubber gloves during the encounter, and answers to this question were coded as yes, no, or no sex. Using guidelines developed by the Victorian AIDS Council (1993) to classify the risk involved in various sex acts, which defines safe sex as any sexual activity in which semen, blood, or vaginal fluids do not pass from an infected person directly into the bloodstream of another person, each commercial sex episode was classified into one of the following four categories:

1. a nonrisk encounter, that is, no sex at all or nonrisk acts, for example, kissing, hugging, and cuddling only;

2. a safer-sex encounter for situations that are risky, but MSWs and their clients have reduced the risk by using a condom;

3. a probably safe/uncertain outcome for sex acts that are potentially unsafe, but we do not have enough information on the context of the act to classify them (uncertain), such as semen on the body of the worker/client, and for those acts where there is still debate about the level of risk or where no strong scientific evidence exists but transmission may occur (probably safe), such as oral sex without condom and/or without ejaculation; and

4. the unsafe sex category, which includes known unsafe sex encounters (i.e., anal sex without condom).

Basic descriptive information about the previously described encounter variables will be presented first. Subsequently, categorical and ordinal data will be analysed to statistically compare results between different locations, modalities of sex work/source of clients, and various sex encounter variables. Potential biases caused by violation of the assumption of independence of observations at the client level were avoided using multilevel models. Multilevel models are techniques that can simultaneously handle measurements at different levels of a hierarchy (Kreft & DeLeeuw, 1998). The design of this study produced a basic 2-level structure with MSWs at level 2 and encounters with clients at level 1, clustered by MSW. All analyses were carried out at the client level, controlling for the probable dependence between encounters involving the same MSW. To test if location or source of clients influenced any of the dichotomous outcome variables (i.e., condom use and check of client for STDs), the MIXOR program (Hedeker & Gibbons, 1998a) was used. The MIXREG program (Hedeker & Gibbons, 1998b) was used for ordinal and interval outcome variables (i.e., duration, number of sex acts, and safe sex outcome)

RESULTS

The total number of sex workers who completed the diary over the 2-week period was 186, four short of the expected sample size. The total number of sex workers who actually returned the diaries was 186. Of the 186 respondents, 96 were from Sydney, 53 were from Melbourne, and 37 were from Brisbane. A relatively high response rate was achieved (Marino, Minichiello, Browne, & Jamieson, in press). Different modalities of MSW yielded different response rates, based on estimations provided by the outreach workers. For example, among the SMSWs between 70% and 95% of the MSWs who were approached, depending on the location, agreed to participate and complete the questionnaire. Among IMSWs and EMSWs groups 50% of the workers approached agreed to participate. Reasons for nonparticipation included agency did not allow the outreach workers to contact the MSW, time, or lack of trust. MSWs who were under 18 years of age were excluded from this study due to legal and ethical reasons. Completion rate was also high, with the diary not completed by only seven of the MSWs who participated in the questionnaire stage of the study.

The age of the MSWs ranged from 18 to 58 years, with 46.1% younger than 25 years of age (mean age 27.06). The majority (69.2%0) had completed secondary education, and 39% had at least some additional tertiary education. The majority (36.4%) worked as independent sex workers, 23.4%0 worked as escorts, 16.9% worked as street workers, and another 23.4% indicated that they worked in more than one of the previously mentioned modalities of sex work. The most frequently reported length of time that the respondents worked as a sex worker was less than six months (27.3%), followed by more than 10 years (21.3%). About 40% of the participants indicated that they worked full time, another 41% indicated that they did sex work on a part-time basis, and the remainder (18.6%) indicated working as sex workers opportunistically or weekends only.

The 186 MSWs reported 2088 commercial sex encounters during the study period, ranging from 1 to 38 each, with an average of 11.2 encounters per MSW. The majority of the MSWs (60.2%) reported 10 or fewer encounters, 18.3% of participants reported less than 5 encounters, 20.9% reported between 10 and 15 encounters, and 9.1% reported more than 20 encounters during the study period. MSWs who worked as independents or as escorts had an average of 11.3 and 11.8 encounters, respectively. Street MSWs reported the least number of clients, with an average of 8.9 encounters, and MSWs who worked in more than one modality of work reported the highest number of encounters, with an average of 12.5. More than half of these encounters were with first time clients (55.8%), while the rest were with clients who the MSW has served previously; 15.7% once before, 14.1% between two and four times before, and another 14.1% more than five times before.

The Where and When Context

Sources of clients. Table 1 shows that the most common sources used for recruiting clients were advertisements (38.2%), followed by escort agencies (31.4%) and the street (17%). Regular clients and referrals from another sex worker were mentioned by roughly 3.7% of the participants. Escort agencies were the most common source of recruiting clients in Sydney and Melbourne (38.6% and 33%, respectively), while in Brisbane the placement of an advertisement was, by far, the most mentioned source (60.3%).

Table 1. Sex Encounter Profile in Commercial Sex Work

                                            Sydney      Melbourne
                                          (n = 1175)    (n = 583)

Where and When?

Source of client
  Escort agency                            453 (38.6)   192 (33.0)
  Advertisement                            443 (37.8)   154 (26.5)
  Another MSW                               49 (4.2)     13 (2.2)
  Street                                   132 (11.3)   148 (25.4)
  Another client                            44 (3.8)      2 (2.1)
  Regular client                            22 (1.9)     41 (7.0)
  Other                                     30 (2.6)     22 (3.8)

Place of the job
  Agency                                   368 (31.4)     9 (1.6)
  MSW's place                              325 (27.7)   148 (25.5)
  Client's place                           289 (24.6)   239 (41.2)
  Hotel/motel                              102 (8.7)    109 (18.8)
  Car                                       54 (4.6)     58 (10.0)
  Other (e.g., parks, toilets, streets)     35 (3.0)     17 (2.9)

Day of the week
  Monday                                   149 (13.0)    90 (15.8)
  Tuesday                                  144 (12.6)    66 (11.6)
  Wednesday                                157 (13.7)    64 (11.2)
  Thursday                                 175 (15.3)    87 (15.3)
  Friday                                   207 (18.1)    90 (15.8)
  Saturday                                 165 (14.4)   106 (18.6)
  Sunday                                   147 (12.8)    67 (11.8)

Starting time
  0.01 to 3.00 am                           96 (8.4)     62 (11.3)
  3.01 to 6.00 am                           14 (1.2)      7 (1.3)
  6.01 to 9.00 am                           15 (1.3)      9 (1.6)
  9.01 to 12.00 am                         120 (10.5)    28 (5.1)
  0.01 to 3.00 pm                          201 (17.5)    73 (13.3)
  3.01 to 6.00 pm                          203 (17.7)    61 (11.2)
  6.01 to 9.00 pm                          258 (22.5)   143 (26.1)
  9.01 to 12.00 pm                         241 (21.0)   164 (30.0)

What

Sexual practices which took place(a)
  None                                      26 (2.2)     10 (1.7)
  Kissing                                  477 (41.2)   214 (37.0)
  Masturbation                             847 (73.2)   417 (72.0)
  Anal sex                                 522 (45.2)   195 (33.7)
  Oral sex                                 751 (64.9)   415 (71.7)
  Rimming                                  195 (16.9)    40 (6.9)
  Other                                    156 (13.5)    51 (8.8)

Mean number of sex acts                      2.55         2.31
  SD                                         1.33         1.13

How?

Payment received
  Money                                   1054 (90.0)   573 (98.7)
  Drug                                      23 (2.0)      2 (0.3)
  Money & drug                              65 (5.6)      2 (0.3)
  Money & other                              7 (0.6)      0 (0.0)
  Other                                     12 (1.0)      3 (0.5)
  Not paid                                  10 (0.9)      1 (0.2)

Time of payment
  Before encounter                         638 (54.5)   400 (68.8)
  After encounter                          515 (44.0)   179 (30.8)
  During encounter                           0 (0.0)      0 (0.0)
  Not paid                                  17 (1.5)      2 (0.3)

Check for STIs
  Yes                                      823 (70.8)   470 (80.6)
  No                                       332 (28.6)   104 (17.8)
  No sex                                     7 (0.6)      9 (1.5)

Use of safe-sex equipment(a)
  Condoms                                  829 (70.6)   377 (64.7)
  Rubber dams                               98 (8.3)      6 (1.0)
  Rubber gloves                             70 (6.0)     13 (2.2)
  No sex                                    19 (1.6)     13 (2.2)
  No answer                                129 (11.0)    34 (5.8)

Mean duration of the encounter(b)           70.2         80.0
  SD                                        74.5         91.6

Time talk
  None                                     199 (17.0)    73 (12.5)
  1/4 of time                              534 (45.6)   251 (43.1)
  1/2 of time                              294 (25.1)   168 (28.9)
  Most of time                             143 (12.2)    90 (15.5)

Safer sex outcome(c)
  No risk                                   32 (3.2)     16 (3.0)
  Safer sex                                713 (70.5)   377 (71.5)
  Probably safe/uncertain                  170 (16.8)   100 (19.0)
  Unsafe                                    97 (9.6)     34 (6.5)

                                           Brisbane      Total
                                          (n = 330)    (n = 2088)

Where and When?

Source of client
  Escort agency                            10 (3.0)     655 (31.4)
  Advertisement                           199 (60.3)    796 (38.2)
  Another MSW                              11 (3.3)      73 (3.5)
  Street                                   75 (22.7)    355 (17.0)
  Another client                            8 (2.4)      64 (3.1)
  Regular client                           42 (4.2)      77 (3.7)
  Other                                    13 (3.9)      65 (3.2)

Place of the job
  Agency                                    3 (0.9)     380 (18.3)
  MSW's place                             134 (40.7)    607 (29.2)
  Client's place                          110 (33.4)    638 (30.6)
  Hotel/motel                              48 (14.6)    259 (12.4)
  Car                                      11 (3.3)     123 (5.9)
  Other (e.g., parks, toilets, streets)    23 (7.0)      75 (3.6)

Day of the week
  Monday                                   46 (14.0)    285 (14.0)
  Tuesday                                  30 (9.1)     240 (11.7)
  Wednesday                                44 (13.4)    265 (13.0)
  Thursday                                 58 (17.6)    320 (15.7)
  Friday                                   59 (17.9)    356 (17.4)
  Saturday                                 63 (19.1)    334 (16.3)
  Sunday                                   29 (8.8)     243 (11.9)

Starting time
  0.01 to 3.00 am                          20 (6.1)     178 (8.8)
  3.01 to 6.00 am                           1 (0.3)      22 (1.1)
  6.01 to 9.00 am                           2 (0.6)      26 (1.3)
  9.01 to 12.00 am                         31 (9.5)     179 (8.9)
  0.01 to 3.00 pm                          38 (11.7)    312 (15.8)
  3.01 to 6.00 pm                          55 (16.9)    319 (15.8)
  6.01 to 9.00 pm                         115 (35.3)    516 (25.5)
  9.01 to 12.00 pm                         64 (19.6)    469 (23.2)

What

Sexual practices which took place(a)
  None                                      8 (2.4)      44 (2.1)
  Kissing                                  81 (24.6)    772 (37.4)
  Masturbation                            269 (81.8)   1533 (74.2)
  Anal sex                                124 (37.7)    841 (40.7)
  Oral sex                                228 (69.3)   1394 (67.5)
  Rimming                                  44 (13.4)    279 (13.5)
  Other                                    44 (13.4)    251 (12.2)

Mean number of sex acts                     2.40          2.46
  SD                                        1.14          1.25

How?

Payment received
  Money                                   311 (94.2)   1938 (93.0)
  Drug                                      2 (0.6)      27 (1.3)
  Money & drug                             10 (3.0)      77 (3.7)
  Money & other                             4 (1.2)      11 (0.5)
  Other                                     3 (0.9)      18 (0.9)
  Not paid                                  0 (0.0)      11 (0.5)

Time of payment
  Before encounter                        209 (63.5)   1247 (60.0)
  After encounter                         116 (35.3)    810 (38.9)
  During encounter                          3 (0.9)       3 (0.1)
  Not paid                                  1 (0.3)      20 (1.0)

Check for STIs
  Yes                                     277 (84.2)   1570 (75.7)
  No                                       49 (14.9)    485 (23.4)
  No sex                                    3 (0.9)      19 (0.9)

Use of safe-sex equipment(a)
  Condoms                                 201 (60.9)   1407 (67.4)
  Rubber dams                               2 (0.6)     106 (5.1)
  Rubber gloves                             5 (1.5)      88 (4.2)
  No sex                                    9 (2.7)      41 (2.0)
  No answer                                62 (18.8)    225 (10.8)

Mean duration of the encounter(b)          66.1          72.2
  SD                                       58.5          77.4

Time talk
  None                                     31 (9.4)     303 (14.6)
  1/4 of time                             163 (49.4)    948 (45.5)
  1/2 of time                              88 (26.7)    550 (26.4)
  Most of time                             48 (14.5)    281 (13.5)

Safer sex outcome(c)
  No risk                                   8 (3.1)      56 (3.1)
  Safer sex                               176 (67.4)   1266 (70.3)
  Probably safe/uncertain                  59 (22.6)    329 (18.3)
  Unsafe                                   18 (6.9)     149 (8.3)

Note. Parenthetical data are percentages.
(a) More than one answer.
(b) Duration in minutes.
(c) n = 1800.

Place of the job. The great majority of the sexual encounters occurred at the residence of the client (30.6%), or of the MSW (29.2%). The next most mentioned places were agencies (18.2%) and hotels/motels (12.4%), followed by cars (5.9%) and public forums such as parks, toilets, or in the street (3.6%). By location, the client's residence (41.2%) was the most frequently mentioned category in Melbourne, while in Brisbane it was the MSW's residence (40.7%) and in Sydney it was escort agencies (31.4%). Paid sexual encounters in cars was highest in Melbourne (9.9%), while parks, the street, and public toilets were mentioned by 7% of the participants in Brisbane but by fewer MSWs in Melbourne and Sydney.

It is interesting to note, however, that when clients in Sydney were recruited from escort agencies the sex encounter did not always take place on the premises of the agency. MSWs from Sydney reported that only three fourths of those encounters took place in the agency, while the remainder either took place at the client's residence (16.2%) or in a hotel (8%). In contrast, clients recruited from agencies in Melbourne were more likely to have the sexual encounter take place in the client's own residence (65.6%), followed by hotels/motels (31.3%). None of the encounters in this category took place at the MSW's residence. In both Sydney and Melbourne, when the client was obtained from the street the most common place for the job was the client's residence, followed by the car and then by hotels/motels. In Brisbane, the most common source for recruiting clients was advertisement, and in these cases the most likely place of the job was the MSW's residence (60.1%) followed by the client's residence (28.3%). In contrast, when the client was recruited from the street, the most common place for the job was the client's residence (36%), followed by hotels/motels (17.3%) and then a car (13.3%).

Day of the week and time. The most common days for sex encounters were Friday (17.4%) followed by Saturday (16.3%). About half of the sex encounter started between 6:00 pm and midnight, with another 31% starting between noon and 6:00 pm.

The What Context

Sex acts. Masturbation was the most common sex act which took place during the encounter (74.2%), followed by oral sex (67.5%) and anal sex (40.7%). There were proportionally more anal sex episodes in Sydney (45%) than the other two locations (Melbourne, 33.7% and Brisbane, 37.7%). Interestingly, in 2.1% of the encounters no sex took place. Table 1 provides a list of some of the other sexual activities which took place during the encounters.

The number of sex acts by encounter ranged from 0 to 6 sexual acts. The average number of sex acts by encounter was 2.43. There was no significant relationship between the number of sex acts and location. By source of client, clients solicited on the street engaged in the least number of sexual acts per encounter (1.82), compared with clients found through other sources (p [is less than] 0.001). Conversely, when the clients were obtained from an advertisement a higher number of sex acts were involved (2.61) compared to all other sources (p [is less than] 0.05).

The How Context

Payment. The most common form of payment was money (93%), including credit cards and cheques. Drugs as the only form of payment were reported in only 1.3% (n = 27) of the encounters, and most of those MSWs who mentioned this form of payment were from Sydney. The majority of the payments were received before the encounter (60%). However, when payment was received after the encounter, it was mostly with clients recruited from agencies (40.2%) or from advertisements (36.2%).

Checking for STDs. About 76% of the participants reported having examined their clients for STDs. A multilevel analysis was performed to determine whether location and source of client was associated with the checking of clients for STDs. By location, the analysis indicated that clients from Melbourne or Brisbane were more likely than those from Sydney to be checked for STDs by their MSW (p [is less than] 0.001). By source of client, clients who were recruited from advertisements or agencies were also more likely to be checked for STDs compared with the clients from other sources (p [is less than] 0.001; p [is less than] 0.05, respectively). Although clients from the street or from other places were less likely to be checked for STDs, these differences did not reach a level of significance at p = 0.05.

Use of safe-sex equipment. Condoms were used in the majority (67.4%) of the reported sex encounters. Other safe-sex equipment used included dams (5.1%) and gloves (4.2%). Of those who used a condom, 45.4% indicated using only one unit in the encounter, 32.3% used two condoms in the encounter, and 10.1% used three condoms. Another 11.2% used between 4 and 6 condoms, and 1.1% used more than 6 condoms in a sex encounter. Condom breakage was a rare event. Of the cases where a condom was used (n = 1419), in only 52 cases (3.7%) did participants report a condom breakage.

In order to identify if location or source of client might be associated with condom use, multilevel analysis using MIXOR was performed. By location group, the analysis indicates that Melbourne clients were more likely than those from the other two locations to not use condoms (p [is less than] 0.05) and clients from Sydney were more likely to be condom users (p [is less than] 0.001). By source of client, those clients who were recruited from advertisements were also more likely to be condom users (p [is less than] 0.001) than when the client were recruited from any other source. Conversely, clients recruited from agencies were less likely to not use condoms (p [is less than] 0.01).

Safe-sex outcomes. Using the described safe-sex classification system, the data revealed that the majority of the commercial sex encounters (70.4%) fell into the safer-sex category, 18.3% were in the probably safe/uncertain category, and only 8.3% of the sex encounters were in the unsafe category (defined as unprotected anal sex). Request for unsafe sex was an isolated event (only 7% of the encounters). Additionally, about 3% of the encounters were in the nonrisk category. An analysis of these results using MIXREG revealed that there were no significant differences by location or by source of client.

Duration of the encounter. The mean duration the sex encounters was 72.21 minutes. The range was from 4 minutes to over 15 hours. However, the majority of cases lasted 60 minutes (40.5%). About 12% of sex encounters lasted between 1 hour and 1.5 hours, and another 14.1% of encounters were longer than 1.5 hours. Additionally, 1.2% of encounters included a whole evening booking.

Source of client was significantly associated with duration of encounter. When the sex encounter was with clients recruited from the street (59.8 minutes), it was significantly shorter than with clients obtained from the all other sources (p [is less than] 0.001). However, when the sex encounter was with a client recruited from the other sources category (that is, referrals from another MSWs or another client, regular clients, etc., but not from the street, or via an advertisement or agency) it was significantly longer, averaging 81.0 minutes (p [is less than] 0.01).

By location, there were also statistically significant differences. When the encounter took place in Melbourne (80.0 minutes), the encounter lasted significantly (p [is less than] 0.001) longer than when it took place at the other locations (Sydney, 70.2 minutes; Brisbane, 66.1 minutes). When the MSWs were asked to report what proportion of the encounter was spent talking, most indicated that very little time was spent talking, with over 60% of the encounters spending one fourth of the time or less on verbal exchanges during the encounter. Time spent was significantly associated with location [chi square](6) = 19.73; p [is less than] 0.01 and with the source of clients [chi square](9) = 17.46; p [is less than] 0.05. MSWs from Sydney tended to spend a lower proportion of encounters talking to clients than the MSWs at the two other locations. Similarily, when the source of the client was other, the time spent talking tended to be longer.

Use of Drugs and Alcohol

Use of alcohol was reported by the MSW to be used by himself or the client, both before and during the sexual encounter. Table 2 shows that, with regard to consumption of alcohol and drugs prior to the encounter, MSWs reported that in 18.2% of the encounters clients had consumed alcohol, and that in 9.7% of the encounters they had consumed alcohol themselves before the encounters. In contrast, use of other drugs by clients before the encounter was much less common. MSWs reported that in only 6% of the encounters did they believe the clients had used marijuana, and that less than 1% used heroin prior to the encounter. However, they were more likely to report using drugs before the encounter themselves; in 20.7% of the encounters they reported using marijuana, and 6.8% used heroin before the encounter.

Table 2. Alcohol and Drug Consumption Profile Before and During Commercial Sex Work Encounters

                                          Sydney     Melbourne

MSW's alcohol and drug consumption
before the sex encounter(a)              n = 116       n = 57
  Alcohol                                92 (7.9)     88 (15.3)
  Marijuana                             264 (22.6)   101 (17.5)
  Speed                                  32 (2.7)      9 (1.6)
  Ecstasy                                 5 (0.4)      1 (0.2)
  Pain killer                             5 (0.4)      3 (0.5)
  Nitrites                               25 (2.1)      7 (1.2)
  Heroin                                 82 (7.0)     36 (6.2)
  Other                                   9 (0.8)      4 (0.7)
  Nothing                               772 (66.1)   372 (64.5)

Client's alcohol and drug consumption
before the sex encounter(a)              n = 1164     n = 575
  Alcohol                               192 (16.5)   135 (23.5)
  Marijuana                              64 (5.5)     33 (5.7)
  Speed                                  19 (1.6)      4 (0.7)
  Ecstasy                                17 (1.5)      3 (0.5)
  Pain killer                             2 (0.2)      0 (0.0)
  Nitrites                               52 (4.5)     20 (3.5)
  Heroin                                  3 (0.3)      1 (0.2)
  Nothing                               546 (46.2)   253 (44.0)
  Do not know                           358 (30.8)   148 (25.7)

MSW's alcohol and drug consumption
during the sex encounter(a)              n = 1167     n = 578
  Alcohol                               119 (10.2)    92 (15.9)
  Marijuana                              90 (7.7)     31 (5.4)
  Speed                                  13 (1.1)      1 (0.2)
  Ecstasy                                10 (0.9)      0 (0.0)
  Pain killer                             7 (0.6)      0 (0.0)
  Nitrites                              135 (11.6)    50 (8.7)
  Heroin                                 14 (1.2)     10 (1.7)
  Other                                  31 (2.7)      3 (0.5)
  Nothing                               876 (75.1)   413 (71.5)

Client's alcohol and drug consumption
during the sex encounter(a)              n = 1038     n = 526
  Alcohol                               160 (15.4)   115 (21.9)
  Marijuana                              91 (8.8)     35 (6.7)
  Speed                                  19 (1.8)      2 (0.4)
  Ecstasy                                12 (1.2)      2 (0.4)
  Pain killer                             7 (0.7)      2 (0.4)
  Nitrites                              189 (18.2)    76 (14.4)
  Heroin                                  3 (0.3)      1 (0.2)
  Other                                  32 (3.1)      4 (0.8)
  Nothing                               703 (67.7)   336 (63.9)

                                         Brisbane       Total

MSW's alcohol and drug consumption
before the sex encounter(a)              n = 32        n = 205
  Alcohol                                20 (6.1)     200 (9.7)
  Marijuana                              63 (19.3)    428 (20.7)
  Speed                                   8 (2.5)      49 (2.4)
  Ecstasy                                 0 (0.0)       6 (0.3)
  Pain killer                            15 (4.6)      23 (1.1)
  Nitrites                                0 (0.0)      32 (1.5)
  Heroin                                 22 (6.7)     140 (6.8)
  Other                                   1 (0.3)      14 (0.7)
  Nothing                               216 (66.3)   1360 (65.7)

Client's alcohol and drug consumption
before the sex encounter(a)              n = 321       n = 2060
  Alcohol                                48 (15.0)    375 (18.2)
  Marijuana                              26 (8.1)     124 (6.0)
  Speed                                   5 (1.6)      28 (1.4)
  Ecstasy                                 3 (0.9)      23 (1.1)
  Pain killer                             4 (1.2)       6 (0.3)
  Nitrites                                7 (2.2)      79 (3.8)
  Heroin                                  0 (0.0)       4 (0.2)
  Nothing                               154 (48.0)    953 (46.3)
  Do not know                            89 (27.7)    595 (28.9)

MSW's alcohol and drug consumption
during the sex encounter(a)              n = 326       n = 2071
  Alcohol                                29 (8.9)     240 (11.6)
  Marijuana                              26 (8.0)     147 (7.1)
  Speed                                   3 (0.9)      17 (0.8)
  Ecstasy                                 0 (0.0)      10 (0.5)
  Pain killer                             6 (1.8)      13 (0.6)
  Nitrites                               17 (5.2)     202 (9.8)
  Heroin                                  1 (0.3)      25 (1.2)
  Other                                   2 (0.6)      36 (1.7)
  Nothing                               258 (79.1)   1547 (74.7)

Client's alcohol and drug consumption
during the sex encounter(a)              n = 313       n = 1877
  Alcohol                                40 (12.8)    315 (16.8)
  Marijuana                              29 (9.3)     155 (8.3)
  Speed                                   4 (1.3)      25 (1.3)
  Ecstasy                                 0 (0.0)      14 (0.7)
  Pain killer                             3 (1.0)      12 (0.6)
  Nitrites                               43 (13.7)    308 (16.4)
  Heroin                                  0 (0.0)       4 (0.2)
  Other                                   2 (0.6)      38 (2.0)
  Nothing                               211 (67.4)   1250 (66.6)

Note. Parenthetical data are percentages.
(a) More than one answer.

When we examined the consumption of alcohol or drugs during the encounter, MSWs reported that they consumed at least one of the substances listed in the diary in 27.2% of the recorded encounters. These included using alcohol in 11.6%, nitrites in 9.8%, and marijuana in 7.1% of the encounters. Interestingly, clients were reported to use substances more during the encounter. In 33.3% of the encounters MSWs reported that clients consumed alcohol or one or more of the listed drugs (i.e., marijuana, nitrites, etc.): In 16.8% of the encounters clients took alcohol, in 16.4% there was nitrite consumption by the client, and in 8.3% there was marijuana consumption by the client. There was a high level of consistency between MSW and client consumption during the sex encounters (kappa = 0.71).

DISCUSSION

This paper describes findings of a prospective diary study designed to examine in detail the nature, context, and structure of the commercial sex encounter between men. One hundred and eighty-six male sex workers from three major Australian cities participated in a 2-week prospective diary study. One of the unique features of this study is that it uses diaries to collect data on individual client/sex worker encounters in order to provide a more comprehensive picture of the male commercial sexual transaction. Two obvious contributions of the diaries are that it can shed light on the contextual dimension of the sexual sale and that it is probably the only method that allows researchers to investigate each sexual encounter as the unit of study.

Limited research is available in Australia and previous research overseas has used recall methods such as questionnaires and interviews. Our data, therefore, minimize many of the distortions inherent to recall methods such as telescoping and memory lapses. This is also true with the complex task of mentally assembling what happened in each particular encounter retrospectively. Diaries are concurrent descriptions of the sexual interactions rather than retrospective summaries and aggregated accounts. As a result, this study enhances confidence that the findings reflect what was occurring during the male-to-male commercial sex interaction.

The results offer several important insights into the public perception of the male sex work industry. First, it appears that MSWs are providing safer sex to their clients. The results of this study are consistent with other studies that have found a high rate of condom use among MSWs with their clients (Overs, 1991; Perkins et al., 1993; Pleak & Meyer-Bahlburg, 1990). Sex workers are continuing to modify their sexual practices in response to the AIDS epidemic through limiting or refusing to engage in certain sexual acts, such as receptive and insertive anal sex without a condom (Browne & Minichiello, 1995; Coutinho, Vanandel, & Rijskijk, 1988; Earls & David, 1989; Elifson et al., 1993; Pleak & Meyer-Bahlburg, 1990). However, the data also reveal that some unsafe sex continues to occur. This finding supports the view expressed by both sex-worker support organisations and sexual health educators that further efforts are still required to maintain and improve the current levels of achievements in preventive behaviours. Furthermore, the data highlight that oral sex is the most commonly practiced sexual act, and this raises a further challenge to our education efforts. Much of the current emphasis has been on HIV/AIDS transmission and unprotected anal sex. What has been frequently minimized is the risk associated with oral sex on sexually transmitted infections (STIs), and current education for both the male sex workers and their clients needs to further raise the level of awareness about this issue.

A surprising finding is that condom breakage was reported in about 4% of the encounters where a condom was used. Data from a literature review (Steiner et al., 1994) indicate that the overall condom breakage rate in the present study fell inside what has been described as a normal range, that is, between 0% and 12%, with many of the studies falling in the 2% to 5% range. Future studies need to consider the consistency of condom use across sexual encounters with clients to determine, for example, whether relational aspects developed in relationships with regular clients influence condom usage, or to identify whether the type of sex act influences the consistency of condom use.

Given that condoms are vital safety equipment in commercial sex, perhaps improvements in the production and current quality standards of condom manufacturing needs to be considered. Additionally, there needs to be a greater emphasis placed on correct condom use. Research also needs to determine whether the factors which lead to condom breakage in commercial sex are similar to or different from those occurring in noncommercial sex, and how MSWs are responding to using current guidelines regarding condom breakage.

Second, the notion that male sex work is a clandestine activity which occurs on the streets is not supported by this study. Male sex work increasingly is becoming a legalised business operating through escort agencies and advertisements. In many ways this context makes it easier for sex workers to be held accountable for their behaviour, in terms of being identified and traced for providing services which are in line with maintaining public health safety in the community. Increasingly, men are entering the business with more positive and professional attitudes in terms of how they perceive their work. This is supported by the following facts: First, many continue to work in this industry over a long period of time and do not find the work stressful (Minichiello, Marino, Browne, & Jamieson, in press), and second, as reported in this study, low levels of unsafe sex practices and of alcohol and drug use occur in the Australian male sex-work industry. However, this study identifies a number of areas which existing initiatives and programs need to give further attention, not the least of which is the need for education for sex workers about accountabilities in relation to relaying sexual health education messages to clients. For example, in this study it was noted that the level of checking clients for STDs still requires improvement, and possibly represents an opportunity for educating both clients and sex workers. Checking for STDs, however, cannot be based simply on evidence of direct observation of the body as many STDs are asymptomatic, and participants need to understand this fact. For sex workers and clients to be reassured that the threat of sexually transmissible infection is reduced because of an assessment reached from a simple and probably unreliable observation check can perhaps create a false security leading to unsafe risk-taking behaviour.

This study found that the majority of the encounters are occurring within a context that allows for some prior discussion in negotiations between the worker and the client, either through telephone conversations, a third party such as a brothel/escort agency manager, or via the content of an advertisement. This creates an increased opportunity to promote messages which steer the interaction toward safe-sex contexts. Legislation needs to create a legal environment which promotes safe sex within a legitimate commercial context, and which exploits the opportunities for achieving this public health objective.

The data presented in this paper show that the male-to-male commercial sexual encounter is a complex situation. In the present study, results are presented in a simple and descriptive fashion by source of clients, duration of the episodes, sexual acts, and the social and work context of the commercial encounter as viewed by the male sex workers. For example, the work and interactional context can be influenced by the addiction of, and the substance use before and during the sexual encounter by, both the clients and the sex workers, and this analysis is the subject of another paper being written by the authors.

Some limitations of this study, however, must be acknowledged when interpreting these results. First, there is always the possibility of a bias in the reporting of the information. Second, in some instances the present diary data lack the necessary depth for a detailed analysis of sex behaviours. For example, the listing of the sexual acts which occurred provides no information on who did what to whom in terms of engaging in receptive, passive, or mutual sexual acts, nor about the sequence of sex acts.

Further studies now need to examine how a range of mediating factors, which Minichiello, Marino, Browne, and Jamieson (1998) have proposed in a model which predicts safe-sex behaviours among MSWs, influence safe-sex practices within the particular circumstances of each commercial sex encounter. For example, it is critical that the point of view of the client is described, so we have a better understanding of the reason for seeking out paid sex and the role drugs and alcohol play in shaping the interaction. Such knowledge may well assist both sex workers and clients to offer and receive such services in a professionally responsible public health context.

REFERENCES

Bloor, M. J., Barnard, M. A., Finlay, A., & McKeganey, N. P. (1993). HIV-related risk practices among Glasgow male prostitutes: Reframing concepts of risk behavior. Medical Anthropology Quarterly, 7, 152-169.

Bloor, M. J., McKeganey, N. P., Finlay, A., & Barnard, M. A. (1992). The inappropriateness of psycho-social models of risk behaviour for understanding HIV-related risk practices among Glasgow male prostitutes. AIDS Care, 4, 131-137.

Browne, J., & Minichiello, V. (1995). The social meanings behind male sex work: Implications for sexual interactions. British Journal of Sociology, 46, 598-622.

Browne, J., & Minichiello, V. (1996). The social and work context of commercial sex between men: A research note. Australian and New Zealand Journal of Sociology, 32, 86-92.

Coutinho, R. A., van Andel, R. L. M., & Rijskijk, T. J. (1988). Role of male prostitutes in spread of sexually transmitted diseases and human immunodeficiency virus. Genetourinary Medicine, 64, 207-208.

Davies, P., & Simpson, P. (1990). On male homosexual prostitution and HIV. In P. Aggleton, P. Davies, & G. Hart (Eds.), AIDS: Individual, cultural and policy dimension (pp. 103-120). London: Falmer Press.

De Graaf, R., Vanwesenbeeck, I., van Zessen, G., Straver, C. J., & Visser, J. H. (1994). Male prostitutes and safe sex: Different settings, different risks. AIDS Care, 6, 277-288.

De Graaf, R., Vanwesenbeeck, I., van Zessen, G., Straver, C. J., & Visser, J. H. (1995). Alcohol and drug use in heterosexual and homosexual prostitution, and its relation to protection behaviour. AIDS Care, 7, 35-47.

Earls, C. M., & David, H. (1989). A psychosocial study of male prostitution. Archives of Sexual Behaviors, 18, 401-419.

Elifson, K. W., Boles, J., & Sweat, M. (1993). Risk factors associated with HIV infection among male prostitutes. American Journal of Public Health, 83, 79-83.

Estep, R., Waldorf, D., & Marotta, T. (1992). Sexual behavior of male prostitutes. In J. Huber, & B. E. Schneider (Eds.), The social context of AIDS (pp. 95-109). Newbury Park, CA: Sage Publications.

Ford, K., Wirawan, D. N., Fajans, P., & Thorpe, L. (1995). AIDS knowledge, risk behaviors, and factors related to condom use among male commercial sex workers and male tourist clients in Bali, Indonesia. AIDS, 9, 751-759.

Ford, K., Wirawan, D. N., & Fajans, P. (1993). AIDS knowledge, condom beliefs and sexual behaviour among male sex workers and male tourist clients in Bali, Indonesia. Health Transition Review, 3, 191-204.

Fox, L. J., Bailey, P. E., Clarke-Martinez, K. L., Coello, M., Ordonez, F. N., & Barahona F. (1993). Condom use among high-risk women in Honduras: Evaluation of an AIDS prevention program, AIDS Education and Prevention, 5, 1-10.

Hedeker, D., & Gibbons, R. D., (1998a) MIXOR: A computer program for mixed-effects ordinal analysis, www.uic.edu/~hedeker/mixor.zip.

Hedeker, D., & Gibbons, R. D., (1998b) MIXREG: A computer program for mixed-effects regression analysis with autocorrelated errors. www.uic.edu/~hedeker/mixreg.zip.

Kreft, I, & DeLeeuw, J. (1998). Introducing multilevel modeling. London: Sage Publications Ltd.

Marino, R., Minichiello, V., & Browne, J. (1999). In V. Minichiello, G. Sullivan, K. Greenwood, & R. Axford (Eds.), Handbook of research methods in health sciences (pp. 381-394). Sydney: Addison--Wesley.

McKeganey, N., Barnard, M., & Bloor, M. (1990). A comparison of HIV-related risk behaviour and risk reduction between female street working prostitutes and male rent boys in Glasgow. Sociology of Health and Illness, 12, 276-292.

McKeganey, N. P. (1994). Prostitution and HIV: What do we know and where might research be targeted in the future? AIDS, 8, 1215-1226.

Minichiello, V., Marino, R., Browne, J., & Jamieson, M. (1998). A review of male to male commercial sex encounters. Venereology, 11, 32-41.

Minichiello, V., Marino, R., Browne, J., & Jamieson, M. (in press). Male sex workers in three Australian cities: Socio-demographic and sex work characteristics. Journal of Homosexuality.

Morse, E. V., Simon, P. M., Osofsky, H. J., Balson, P. M., & Gaumer, H. R. (1991). The male street prostitute: A vector for transmission of HIV infection into the heterosexual world. Social Science and Medicine, 32, 535-539.

Neave, M. (1994). Prostitution laws in Australia: Past history and current trends. In R. Perkins, B. Prestage, R. Sharp, & F. Lovejoy (Eds.), Sex work and sex workers in Australia (pp. 67-99). Sydney, Australia: UNSW Press.

Overs, C. (1991). To work or not to work? Questions facing HIV-positive sex workers. National AIDS Bulletin (Australia), July, 21.

Perkins, R., Prestage, B., Lovejoy, F., & Sharp, R. (1993, July) Cracking in the age of AIDS: Female and male prostitutes in the New South Wales commercial sex industry and their response to HIV prevention. National AIDS Bulletin, July, 42-46.

Pickering, H., Todd, J., Dunn, D., Pepin, J., & Wilkins, A. (1992). Prostitutes and their clients: A Gambian survey. Social Science and Medicine, 34, 75-88.

Pleak, R. R., & Meyer-Bahlburg, H. F. L. (1990). Sexual behavior and AIDS knowledge of young male prostitutes in Manhattan. The Journal of Sex Research, 27, 557-587.

Robinson, T., & Davies, P. (1991). London's homosexual male prostitutes: Power, peer groups and HIV. In P. Aggleton, G. Hart, & P. Davies (Eds.), AIDS: Responses, Interventions and Care (pp. 95-110). London: Falmer Press.

Steiner, M., Trussell, J., Glover, L., Joanis, C., Spruyt, A., & Dorflinger, L. (1994) Standardized protocols for condom breakage and slippage trials: A proposal. American Journal of Public Health, 84, 1897-1900.

Victoria AIDS Council. (1993). Safe sex and AIDS: The answers. Melbourne, Australia: Victorian AIDS Council.

Visano, L. (1991). The impact of age on paid sexual encounters. In G. A. Lee (Ed.), Gay midlife maturity (pp. 207-226). New York: Haworth Press.

Waldorf, D., Murphy, S., Lauderback, D., Reinarman, C., & Marotta, T. (1990). Needle sharing among male prostitutes: Preliminary findings of the Prospero Project. Journal of Drug Issues, 20, 309-334.

Manuscript accepted December 8, 1999

Victor Minichiello, Rodrigo Marino, Jan Browne, and Maggie Jamieson University of New England, Armidale, New South Wales, Australia

Kirk Peterson Victoria, Australia

Brad Reuter Queensland, Australia

Kenn Robinson Sydney, Australia