Electronic Journal of Human Sexuality, Volume 11, June 27, 2008

Frequency of Sexual Activity among Men Who Visit Prostitutes

Richard Tewksbury, Ph.D. ¹ and Seana Golder, M.S.W, Ph.D. ²

1 Department of Justice Administration, University of Louisville, Louisville, Kentucky.

2 Kent School of Social Work, University of Louisville, Louisville, Kentucky.

Correspondence:

Richard Tewksbury, Ph.D.
Department of Justice Administration
University of Louisville
Louisville, Kentucky 40292
(502) 852-0379
E-mail: tewks(at)louisville.edu .

ABSTRACT

 This research assessed whether men who patronize prostitutes display higher rates of sexual activities than “normal” and whether their rates of sexual activities vary across commercial and non-commercial partners. Data for this research were obtained from a study of clients of street prostitutes in cities in Oregon, California, and Nevada (Monto, 1999; 2000). A number of variables were predictive of frequency of sex generally (race/ethnicity, having a regular sex partner, number of sex partners, frequency of sexual thoughts, frequency of pornography use; rough sex; similarity of sexual interests) and specifically with a prostitute (number of sexual partners, being separated from one’s partner); regression models accounted for 30% of the variance in frequency of sex and approximately 27% of the variance in frequency of sex with a prostitute. Implications of these findings are addressed.

Introduction

Understandings of commercialized forms of sexual activity are a niche in the social science literature that have shown a tendency to focus on either discussions of the gendered nature of such work, who is involved in providing commercial sexual services, or examinations of the (il)legal aspects of such encounters. What stands as the least developed aspects of such issues is how involvement in commercial sexual activities is experienced as part (or the entirety) of men’s sexual experiences. For some observers the idea of paying for sexual services may be seen as a way to be introduced to or learning about sex, an activity that serves as the central component of an individual’s sexual repertoire (perhaps due to inabilities to access or lack of interest in other sexual encounters; Edgley, 1989; Reinisch, 1990) or as simply one component of a set of healthy sexual expressions. The present study seeks to add to this discussion through an examination of whether men who patronize prostitutes display especially high levels of sexual activity.

Review of the Literature

Research on men’s sexual engagements, specifically the factors that influence the frequency with which men are sexually active, has primarily centered on examinations of sexual activity in stable, long- or short-term romantic relationships, as well as in casual, noncommercial couplings. When researchers have addressed men’s sexual encounters with prostitutes they have primarily focused on identifying how many men do so and the demographic characteristics of these men, and means by which men who have sex with prostitutes protect themselves from sexually transmitted diseases. Other issues that have been addressed, but are tangential to the current study include violence in such encounters and patterns of how men find, approach and carry out their commercial sexual encounters.

Factors Influencing Frequency of Sex in Relationships

The research literature has shown that there are a number of factors that may be important influences on the frequency of sexual activities for persons in committed relationships. Among these factors are marital status, the presence of children, quality of relationship, and demographics.

Marital status has been shown to have little influence on frequency of sexual interactions for heterosexual couples ( Stafford, Kline, & Rankin, 2004). However, while the official status of a relationship has not shown effects on sexual behavior, research has shown that sexual activity frequency decreases over time for heterosexual couples (Blumstein & Schwartz, 1983; Pederson & Blekesaune, 2003; Stafford, et al., 2004). A significant decrease in both sexual frequency and satisfaction is hypothesized to be a consequence of “boredom and routine” (see Pedersen & Blekesaune, 2003). This idea is supported by research that has shown that men are both more likely than women to report having sexual fantasies involving a partner other than their spouse/significant other and that the likelihood and frequency of such fantasies increase as relationship length increases (Hicks & Leitenberg, 2001).

The quality of a relationship is clearly also an important influence on the frequency of sexual activity for heterosexual couples. For young adult heterosexual couples, perceptions of higher levels of conflict and frustration in the relationship have been shown to be related to higher frequency of sexual intercourse (Rostosky, Galliher, Welsh, & Kawaguchi, 2000). Marriages in which one or both partners are violent toward one another have also been shown to have higher frequencies of sexual interactions between partners (Apt & Hurlbert, 1993; Donnelly, 1993). This apparently counter-intuitive finding has been explained by some with the argument that it is only the husband’s violence that is associated with a higher frequency of sexual activity (DeMaris, 1997). This argument suggests that violent men are “sexual extortionists” in their marriages and although there may be more sexual activity between the partners, it may not be fully consensual or desired by the female partner.

Some research has been devoted to the effects of leaving a relationship (e.g., divorce) on sexual activities. Again, contrary to some popular beliefs and stereotypes, individuals who leave relationships do not necessarily seek and obtain multiple sex partners (Wade & DeLamater, 2002); however, there are some variations in sexual activities post-relationship. Specifically, lower income individuals (both men and women) are more likely than higher income individuals to seek and obtain multiple sex partners after leaving a long-term relationship. Overall, however, divorced individuals report a low rate of sexual activity. Findings from one national sample show a mean of less than one sexual partner per year (Stack & Gundlach, 1992).

While still involved in a long-term relationship, significant numbers of men have been shown to pursue and consummate sexual activities outside of their primary relationship. More than one in five (22.7%) adult men in a nationally representative sample reported having engaged in extramarital sex (Wiederman, 1997). Variations in this include that African-American men are more likely (33.0%) than whites (21.4%) or men of other races (25.7%) to report extramarital sex.

In regards to demographics, three principal variables have been the focus of research: sex, age at which sexual activity was initiated, and race. Each of these variables has been shown in the literature to be related to engagement in casual sexual activity. Regardless of whether or not an individual is engaged in an on-going relationship, men are more likely than women to engage in casual sex (Cubbins & Tanfer, 2000). One national study found that just shy of one-half (46.7%) of men reported having on at least one occasion engaged in casual sex (Cubbins & Tanfer, 2000). Such activities are most common for individuals whose first sexual encounter occurred prior to age 17 (50.6% vs. 26.0%) (Cubbins & Tanfer, 2000). Additionally, contrary to the research on extramarital sex, African-Americans are less likely to report casual sexual encounters or multiple sex partners than all other races (Cubbins & Tanfer, 2000; Michael, Gagnon, Laumann, & Kolata, 1994). It appears, therefore, that while African American men may engage in more casual sex while not married, during their marriages they may be less likely to engage in sex with partners other than their spouses.

Men and Prostitutes

One understudied issue in sexuality focuses on sexual interactions between men and prostitutes. Recognizing that not all sexual activity occurs between romantic partners, scholars, public health officials, and social service providers are interested in knowing how frequently, under what circumstances, and who specifically is most likely to be involved in sexual activities with prostitutes. Results of a nationally representative sample showed that approximately 16% of American adult men have ever paid for sex (i.e., had sex with a prostitute, Michael, et al., 1994). A related way to view this is that “just six in a thousand” (p. 213) men visit prostitutes in a given year. Others have reported similar rates (18%; Sullivan & Simon, 1998). Approximately one-half of men who do visit prostitutes are involved in a long-term relationship (Atchison, Fraser, & Lowman, 1998; McKeganey & Barnard, 1996; Morse, Simon, Balson, and Osofsky, 1992).

Several researchers have addressed the issue of racial variations in men’s likelihood of patronizing prostitutes, but the findings of this area of research are mixed. Freund, Lee and Leonard (1991) reported that white men were the majority of prostitutes’ clients; others (Johnson, et al., 1994) suggest African-American men are the most common clients and yet others (Kim, Marmor, Dubin and Wolfe, 1993) report Hispanic men are the most likely to patronize prostitutes. These seemingly contradictory findings, however, are likely the result of either geographic/cultural differences or more the more important influences of education and socio-economic status. Men with lower levels of education and homeless men have also been shown to have a greater frequency of sex with prostitutes than other men (Kinsey, et al., 1948; Rickert & Rickert, 1995).

Most existing research suggests that men who patronize prostitutes tend to do so repeatedly, although available research does not support a contention that prostitutes are most men’s primary sexual outlets. ¹ While data on how often men who engage in sex with prostitutes are scarce, at least one set of researchers reported that for a sample of 143 clients the median number of sexual encounters with a prostitute over a period of more than a decade was seven (Barnard, McKeganey, & Leyland, 1993). Similarly, working with survey data from men enrolled in a diversion program for arrested clients of prostitutes, Wortley, Fischer and Webster (2002) reported that 25% of men claimed the time they were arrested to have been their first time attempt to purchase a prostitute’s services, 18% were novices (claiming to have done so 1 to 4 times previously) and 22% were sex-trade veterans (having been with a prostitute at least 5 times previously). ² Freund, et al. (1991) reported similar results; survey data from clients of prostitutes indicate that a majority of clients were often repeat customers, returning to either an individual or small group of prostitutes. This was also supported by research which suggests that men who have a high frequency of sexual behavior with prostitutes are also more likely to be obsessive-compulsive (Rinehart & McCabe, 1998). However, it is also important to note that clients of prostitutes do recognize health risks in such encounters, and will frequently take active steps (e.g. use condoms) to minimize such risks, as shown in one European study (Vanwesenbeeck, DeGraaf, Van Zessen, Straver, & Visser, 1993).

While many men who patronize prostitutes may do so repeatedly, this does not inform us about the overall frequency with which these men are sexually active. Most men who have sex with prostitutes also have non-prostitute sexual partners (including both relationship partners and casual sexual encounters).

The Present Study

Based on survey data collected from a sample of men arrested for soliciting prostitutes in four cities, the present, exploratory study assessed the frequency of sexual activities for these men--both with non-prostitute and prostitute partners--and identified predictors of greater and lesser frequency of sexual activities with both varieties of partners. In this way, the current research provides evidence for assessing whether or not men who patronize prostitutes do, in fact, display excessively high rates of sexual activity, as well as assessing whether factors associated with high rates of sexual activity with prostitutes are similar/different than those predictive of high rates of sexual activity in general. This, in turn, provides an indication of whether customers of prostitutes are seeking to add sexual partners and activities to an already high level of sexual activity or whether other factors are associated with pursuing commercial sex partners.

Methods

Participants

This research was based on secondary data, publicly available from the Inter-university Consortium for Political and Social Research (ICPSR; http://www.icpsr.umich.edu/index.html). These data were originally collected in a study of clients of street prostitutes in Portland, Oregon, San Francisco and Santa Clara, California, and Las Vegas, Nevada from 1997 to 1999 (Monto, 1999; 2000). The study employed a convenience sample of 1,342 men who were recruited from participants in “client intervention programs” for those arrested trying to hire street prostitutes in these four cities. ³ Preliminary analyses indicated that in response to the questions assessing frequency of sex with a prostitute in the past year, 42.4% of the sample reported that either the questions was not applicable to them (18.9%), they had not had sex with a prostitute in the past year (21%), or they failed to answer the question (2.4%). The first two responses, that the question was not applicable and that they had not had sex with a prostitute in the past year, suggested that almost 40% of the respondents providing an answer to this question may have been first time patrons of prostitutes who were arrested prior to engaging in sexual relations with a prostitute. Given that a major aim of the current paper is to assesses the frequency of sexual activities with non-prostitute and prostitute partners among ‘johns’, the inclusion of men in the sample who were not recently sexually active with prostituted women seemed inconsistent; thus the final sample was comprised of only the 772 respondents who reported they had had sex in the past year with a prostitute. ⁴

Participation in the study was voluntary; questionnaires were self-administered prior to the start of the intervention program and collected anonymously (Monto & Hotaling, 2001). Questionnaire items were drawn from a number of previously validated measures, such as the General Social Survey and the National Health and Social Life Survey (see Busch, Bell, Hotaling, & Monto, 2002 for further explanation of measures). The response rate for the study was over 80 percent (Monto, 1999a, 1999b).

Sociodemographic characteristics are summarized in Table I. Almost 60% of the respondents were white (58.8%), 5.1% were African American, approximately 20%, 13%, 4% of respondents were classified as Hispanic, Asian, and “Other”, respectively. The sample was generally well educated; more than 72% of the participants reported having at least some college education. The vast majority of respondents (93.1%) identified as exclusively heterosexual (or “straight”). Given the lack of variability in sexual orientation, this variable was excluded from further analysis. Slightly more than 40% of the respondents were married and most (88.1%) reported being employed either full- or part-time. The respondents ranged in age from 18 to 80, and were on average, approximately 38 years old (mode = 38 years old).

Table I.

^a Figures based on valid percentages.

Measures

In addition to the dependent variables of interest, frequency of sex during the last 12 months and frequency of sex with a prostitute during the last 12 months, four types of measures are described in the following section: personal background characteristics; general sexual behavior; sexually aggressive behavior; and relationship characteristics.

Two dependent variables were measured, frequency of sex during the last 12 months and frequency of sex with a prostitute during the last 12 months. Response options for the question asking participants how often they had sex during the past 12 months were given on a seven point scale (“not at all”(1), “once or twice” (2), “ about once a month” (3), 2 or 3 times a month (4), about once a week (5), 2 or 3 times a week” (6), greater than 3 times a week” (7)). Response options for the question assessing the frequency with which participants reported having sex with a prostitute in the past year were “only one time” (1), “more than 1 time but less than once per month” (2), “1 to 3 times per month” (3), “once or twice per week” (4), “3 to 4 times per wk” (5), “5 or more times per week” (6).

Personal background characteristics were measured by three dichotomous variables (“no” (0) or “yes” (1)). Respondents were asked if their parents were divorced, if they had ever been touched sexually as a child, and if they had served in the military.

Respondents’ general sexual behavior was measured by five variables assessing: having a regular sexual partner during the last 12 months; number of sex partners during the last 12 months; frequency of sexual thoughts; and frequency of pornography use ⁵. A single dichotomous question (“no” (0) or “yes” (1)) asked respondents if they had a regular sex partner in the last 12 months. Respondents were also asked to report the number of sex partners they had in the past 12 months on a 9-point scale: 0 (0) to 4 (4); 5-10 partners (5); 11-20 partners (6); 21-100 (7); more than 100 partners (8). Frequency of sexual thoughts was measured by a single question, “on average, how often do you think about sex?”. Response options included “never think about sex” (1) to “several times a day” (6). Frequency of pornography use was computed by taking the mean of two questions assessing the frequency with which respondents reported looking at pornographic magazines and watching pornographic videos. Response options for each of these questions ranged from “never” (0) to “several times a day” (5). The computed variable reflecting the frequency of pornography use had a possible range of 0 to 5.

Four variables were used as measures of sexually aggressive behavior: threaten[ing] to use physical force to get sex; used physical force to get sex; rough sex; and control during sex. The first two questions asked participants if they had “ ever had sexual intercourse with a woman when she didn't want to because you threatened to use physical force (twisting her arm, holding her down, etc.) if she didn't cooperate?” or if they had “used some degree of physical force” to make her have sexual intercourse; both questions were dichotomous (“yes” (1); “no” (0)). The final two questions operationalizing sexually aggressive behavior asked participants to rate their level of agreement to the statements: “ I like rough hard sex”; ⁶ and “I like to be in control when I’m having sex”. Both variables were measured on a four point scale (“strongly agree” (1) to “disagree strongly” (4)).

Seven variables assessed relationship characteristics: shy with women; similarity of sexual interests; similarity of sexual desires; trouble with partner; separated from partner; and break up with partner. Respondents were asked to rate their level of agreement with the statement, “I am shy and awkward when I am trying to meet a woman”. Responses were measured on a four point scale (“strongly agree” (1) to “disagree strongly” (4)). Respondents were asked to consider their sexual relationship with their partner and rate the similarity of their sexual interests on a four point scale (“very similar” (1) to “very different” (4)). They were also asked, “how similar are your levels of sexual desire (wanting to have sex)”. Response options included: “partner is much more interested”(1) to “I am much more interested than my partner”(5). The final three variables were dichotomous questions (“yes” (1); “no” (0)) asking respondents whether, in the past year, they had had “serious trouble with their wife or partner”; “separate from their wife or partner”; “break up with a steady girlfriend or fiance”.

Missing Data

Due to the presence of missing data, we took steps to minimize the bias to parameter estimates that can result from the more common methods of handling missing data (e.g., listwise deletion and mean substitution; Graham, Hofer, Donaldson, Mackinnon, & Schafer, 1997; Graham, Hofer, & Piccinin, 1994). Thus, the Mplus statistical analysis program was used for all analyses (Muthen & Muthen, 2000). The Mplus program provides for the handling of missing data through the use of a raw maximum-likelihood procedure to impute values for missing data. Unbiased parameter estimates and reasonable standard errors are obtained using the missing data feature in Mplus (Choi, Golder, Gillmore, & Morrison, 2005; Muthen & Muthen, 2000).

Results

Means, standard deviation of the mean, and range, where appropriate, are reported in Table II.

Table II.

^a Figures based on valid percentages.

The bivariate correlation between the dependent variables, frequency of sex during the last 12 months and frequency of sex with a prostitute during the last 12 months, was .166 (p < .01). On average (mean = 4.286) participants reported having sex 2 to 3 times per month in the past year; 46% of the respondents reported having sex once a week or more in the past year. In response to the question assessing frequency of sex with a prostitute in the past year, 36% of the respondents reported that they had only had sex with a prostitute once in the past year; approximately 16% of the respondents reported having sex with a prostitute 1 to 3 times per month or more in the past year.

Thirty-five percent of the respondents reported that their parents were divorced. Fourteen percent reported being touched sexually as a child and slightly more than a quarter of the respondents reported serving in the military.

Almost 77% of the respondents reported having a regular sex partner. On average the sample reported having 3 sex partners over the last 12 months. On average the respondents reported thinking about sex everyday and using pornography less than once per month.

Examination of the results indicated that there was a relatively low frequency of affirmative responses to questions assessing whether a respondent had used threats of or actual violence to get sex. Among the respondents providing an answer to these questions, only 12 and 10 respondents, respectively, indicated that they had used either threats of or actual violence to get sex (7 respondents replied they had used both). Given the seriousness of the particular behavior assessed by these two questions, it is possible that although the questionnaires were anonymous, respondents failed to answer these questions honestly for fear of detection. Based on this and the accompanying lack of variability in both these variables, they were dropped from further analysis. Respondents generally reported they disagreed somewhat (mean = 3.310) with the statement asking them if they liked rough sex. However, slightly more than 19% (19.3) of the respondents reported that they agreed somewhat or strongly to this question. Greater variability was found in the question assessing the degree to which respondent’s liked control during sex; while on average the sample generally tended to disagree somewhat with the statement asking them if they liked control during sex, approximately 11 and 34%, respectively reported either agreeing strongly or somewhat with this statement.

Examination of measures of central tendency indicate that overall the sample did not report feeling shy and awkward; however approximately 24% of the sample reported they either agreed strongly or somewhat with this statement. The majority of respondents providing a response reported that their sexual interests were either “very similar” or somewhat similar” to their partners (70.6%) and about 36% reported that he and his partner had equivalent levels of sexual desire.

Exploratory Analysis

Reflecting the exploratory nature of the study, we conducted a series of regression analyses to examine the relationship among all the hypothesized predictors (e.g. demographic characteristics, personal background characteristics, general sexual behavior, and sexually aggressive behavior), and each of the dependent variables, respectively. Categorical variables were dummy coded. For marital status, individuals who reported being married were treated as the comparison group (group assigned zero’s throughout); white was the comparison group for race/ethnicity; the comparison group for education level was having a high school diploma or less; not working full- or part-time was the comparison group for employment status. In the initial models, all 22 of the predictors were entered into the equation simultaneously. However, in order to have the most parsimonious models possible, variables that failed to yield a significant relationship with the dependent variables, respectively, were excluded from the final models.

Table III provides the unstandardized regression coefficients, standard errors, Z statistic, and standardized regression coefficients (i.e. b) for the final models predicting frequency of sex and frequency of sex with a prostitute (results for the preliminary models may be requested from the authors).

Table III.

*b is significant at p < .05 with value equal to or greater than 1.96 as evaluated by the Z statistic.

The final regression model predicting frequency of sex during the last 12 months retained nine of the original predictor variables: race/ethnicity; marital status; having a regular sex partner; number of sex partners; frequency of sexual thoughts; frequency of pornography use; rough sex; feeling shy and awkward with women; and similarity of sexual interests. ⁷ Results indicated that having a regular sex partner, increased numbers of sexual partners, increased frequency of sexual thoughts, and increased frequency of pornography use were all positively and significantly related to the frequency of sex in the past 12 months. In contrast, rough sex and similarity of sexual interests were both negatively and significantly related to the frequency of sex in the past 12 months. The relationship between rough sex, similarity of sexual interests, and frequency of sex in the past 12 months indicates that the more similar respondents reported their sexual interests were with those of their partner or the more respondents agreed with the statement that they liked rough sex, the more frequently they had sex in the past 12 months. Overall, the linear combination of variables accounted for slightly more than 30% of the variance in frequency of sex during the last 12 months among male clients of street prostitutes.

In comparison, the final regression model predicting frequency of sex with a prostitute in the past 12 months retained only two of the original predictor variables: number of sexual partners and being separated from one’s partner. Number of sexual partners was positively related to the frequency of sex with a prostitute. However, being separated from one’s partner was negatively related to frequency of sex with a prostitute in the past 12 months. The relationship between being separated and frequency of sex with a prostitute in the past 12 months indicates that men who reported being separated from their partner engaged in more frequent sex with a prostitute over the past year than men who were not separated. Overall, the linear combination of variables accounted for approximately 27% of the variance in frequency of sex with a prostitute.

Examination of the regression coefficients for the dummy coded variables suggested that frequency of sex in the past 12 months also varied as a function of race/ethnicity; specifically there was a negative and significant finding for the relationship between being Asian and frequency of sex in the past 12 months. To further explore this result an additional series of regressions were conducted to examine the relationship among the hypothesized predictors and frequency of sex, controlling for race/ethnicity. Results of these regressions, as well as the average frequency of sex are presented in Table IV. Overall, regardless of race/ethnicity, having a regular sex partner was related to frequency of sex in the past 12 months. Examination specifically of Asian respondents indicated that they reported, on average, having less frequent sex over the past year than respondents of other races/ethnicities. The pattern of relationships between the independent variables and frequency of sex in the past 12 months indicated that in addition to having a regular sex partner, number of sex partners, rough sex, and similarity of sexual interest, were also significantly related to frequency of sex for Asian respondents. In particular, rough sex and similarity of sexual interests were negatively related to frequency of sex.

Table IV.

*b is significant at p < .05 with value equal to or greater than 1.96 as a evaluated by the Z statistic. ^a The variance of regular sex partner for respondents who reported being “other” in regards to race/ethnicity was .00, therefore the model was run excluding this variable.

DISCUSSION

The demographic and personal history profile of this sample of customers of prostitutes fairly closely reflects that reported in previous research. The present sample was primarily white, well educated, employed, and either currently or previously married. Early research suggested that large proportions of American men had engaged in sex with a prostitute on at least one occasion. This finding implied that there was considerable demographic variability among men who patronized prostitutes. However, the results of this research appear to support the claims of more recent researchers (Michael et al., 1994; Sullivan & Simon, 1998) suggesting that it is not a large and reflective cross-section of American men who patronize prostitutes, but rather specific subgroups of men. Examination of the demographic characteristics from the present sample suggest that a more financially stable segment of American men are engaged in sex with prostitutes (or at the very least being arrested for soliciting sex with prostitutes). Also of interest are the results suggesting that men patronizing prostitutes do not have excessively high rates of sexual activity as some have suggested (Wortley, et al., 2002). Rather, the men in the present study report sexual activity rates (in general) very similar to which previous research has reported for heterosexual men (especially married heterosexual men) in general. Where the present research diverges from that of previous studies is in the educational levels of the sample. In the present sample more than one-third of the men have completed college, with an additional 35% having attended at least some college courses. This, however, needs to be viewed in the context of geography: the current sample is drawn from communities with larger than average proportions of the population having higher education (U.S. Bureau of the Census, 2006).

The multivariate analysis focused on identifying the variables that were significant predictors of frequency of sexual activities, both in general and specific to frequency of sexual activities with prostitutes. Whereas previous researchers have addressed the issue of frequency of sex in relationships, this has not been addressed either specifically for men seeking commercial sexual services or for frequency with commercial sex providers. In the final models, only one of the candidate variables, number of sex partners, was a significant predictor of both frequency of sex generally and frequency of sex with a prostitute. Increasing numbers of partners was positively related to both dependent variables; in fact, examination of the standardized beta weights for both models suggests that the number of sexual partners was the most important predictor in each model. This finding suggests that rather than having more frequent sex with a single partner, the men in this sample who are having more frequent sex are engaging in more sex with more partners regardless of whether the women is a prostitute or not.

In addition to increasing numbers of sex partners, six additional candidate variables significantly contributed to an understanding of the frequency of sex, generally, among this sample of men who patronize prostitutes: race/ethnicity; having a regular sex partner; greater frequency of sexual thoughts; frequency of pornography use; liking rough sex; and similarity of sexual interests with their partner. The inverse relationship between similarity of sexual interests and frequency of sex may be interpreted such that respondents who reported their partners to be more interested in sex (than they were), had sex more frequently than men who reported that they themselves were more interested in sex than their partners. This then suggests that in part, for this sample of men, the partner’s sexual interest played an important part in the control of their (i.e. the men’s) frequency of sex. Relatedly, while the men in this study were not necessarily “hypersexual” in regards to the frequency with which they engaged in sex, the results indicate that men who had a greater frequency of sex, generally also engage more frequently in other sexualized activities such as having sexual thoughts and pornography use. In regards to race/ethnicity, initial analyses indicated that there was a significant and inverse relationship between being Asian and frequency of sex. Subsequent analyses showed that Asian men in present sample reported a significantly lower mean frequency of sex in the past 12 months than their counterparts of other race/ethnicities. However, regardless of race/ethnicity, number of sex partners was a significant predictor of frequency of sex generally and, with the exception of white men ⁸, examination of the standardized beta indicated that the number of different sexual partners was the most important predictor of frequency of sex. Taken together these variables suggest that the central influence on frequency of men’s sexual activities in general is access to a number of sexual partners.

In the model predicting frequency of sex with a prostitute, only one variable in addition to number of sex partners was a significant predictor: being separated from one’s partner. While the two predictor variables explained 27% of the variance in frequency of sex with a prostitute, these findings strongly suggest that the model failed to capture the more nuanced motivations and reasoning for a man’s decision to engage in sex with prostituted women. Prior research, while limited, has identified a number of environmental and psychological characteristics associated with male client’s motivations for procuring sex from prostitutes. ⁹ Some of these include: peer behaviors; the social context of the acts; personality characteristics (i.e. inhibited/out-going; secure/insecure); aggressiveness; impulsivity; and sensation seeking (see Vanwesenbeeck, 2001 for a review of this literature). The study from which the current data were taken primarily measured overt behaviors rather than more covert psychological characteristics and motivations. Moreover, when there was an attempt to measure psychological characteristics (e.g. shyness) it was done with a single question rather than a standardized scale. This suggests that future research in this area should be guided by an ecological model that attends to multiple levels of interaction between person’s and their environments. For example, measuring both overt behaviors as well as more covert psychological and personality characteristics. Moreover, future research would be greatly strengthened by employing multiple measures of salient constructs and the use of standardized scales where possible.

Several limitations of the present work should be acknowledged. Although participation was voluntary and the participation rate was over 80%, participants were men arrested for soliciting female prostitutes and participating in deferred adjudication programs (Monto, 1999a, 1999b). As noted in Busch et al. (2002), “their attitudes may not be representative of male customers as a whole. [Moreover,] the connection between attitudes and behaviors is not direct” (pg. 1109). Although not a universally accepted position, a large body of research supports the reliability and validity of self-report data (e.g. Darke, 1998; Hindeland, Hirschi, & Weis, 1981; Hser, 1993; Lauritsen, 1998; Rouse, Kozel, & Richards, 1985). It is important to note that the study from which these data were drawn relies heavily on respondents’ self-reports of socially undesirable behaviors. Care should be taken in interpreting and generalizing these results as the sample may not be fully representative of the population of clients of street prostitutes. And, finally, it should be acknowledged that this study examines correlates of behaviors. The time ordering of activities can not be deduced from the data; more studies that employ longitudinal designs are needed before causal conclusions can or should be drawn.

Finally, it should be noted that a strength of the current study is the handling of missing data and in the ability to directly compare the rates of self-reported sexual activity in general with the rate of self-reported sex with prostitutes for one sample of men. Prior published research utilizing these data has either failed to address the issue of missingness and/or used mean substitution. This suggests that conclusions based on prior analyses may be significantly biased (Choi et al., 2005; Graham et al., 1997; Graham et al., 1994). The current study imputed missing data via a raw maximum-likelihood procedure allowing us to obtain unbiased parameter estimates and reasonable standard errors (Choi et al., 2005; Muthen & Muthen, 2000). Thus conclusions based on the analyses reported here are done without the challenges present in prior studies with this data. And, by allowing a direct comparison of rates of sexual activity in general and with commercial partners, this analysis allows us to better understand the motivations and corresponding personal characteristics of men who report high rates of sexual activity.

In conclusion, this exploratory study sought to add to the literature addressing influences on sexual activity rates and the sexual activities of men who patronize prostitutes. The results suggest that why there is some minimal overlap among the predictors of sexual activity generally and with prostitutes specifically, in general the motivations leading to these behaviors vary.

REFERENCES

Apt, C. & Hurlbert, D.F. (1993). The sexuality of women in physically abusive marriages: A comparative study. Journal of Family Violence, 8, 57-69.

Atchison, C., Fraser, L., & Lowman, J. (1998). Men who buy sex: Preliminary findings on an exploratory study. In J.E. Elias, V.L. Bullough, v. Elias, & G. Brewer Eds.).

Prostitution: On Whores, Hustlers, and Johns (pp. 172-203). Amherst, NY: Prometheus. Barnard, M.A., McKeganey, N.P., & Leyland, A.H. (1993). Risk behaviors among male clients of female prostitutes. British Medical Journal (International Edition), 307 (6900), 361-362.

Blumstein, P., & Schwartz, P. (1983). American Couples. New York: Morrow.

Choi, Y., Golder, S., Gillmore, M. R., & Morrison, D. M. (2005). Analysis with missing data in social work research. Journal of Social Work Research, 31 (3), 23-48.

Cubbins, L.A. & Tanfer, K. (2000). The influence of gender on sex: A study of men’s and women’s self reported high risk sex behavior. Archives of Sexual Behavior, 29, 229-257.

DeMaris, A. (1997). Elevated sexual activity in violent marriages: Hypersexuality or sexual extortion? The Journal of Sex Research, 34 (4), 361-373.

Donnelly, D.A. (1993). Sexually inactive marriages. The Journal of Sex Research, 30, 171-179.

Edgeley, C. 1989. Commerical Sex: Pornography, Prostitution and Advertising. In K. McKinney and S. Sprecher (Eds.) Human Sexuality: The Societal and Interpersonal

Context (pp. 370-424). Norwood, NJ: Ablex Publishing Corporation. Freund, F., Lee, N. & Leonard, T.. (1991). Sexual behavior of clients with street prostitutes in Camden, New Jersey. Journal of Sex Research, 28 (4), 579-591.

Graham J.W., Hofer SM, Donaldson S.I., et al. Analysis with missing data in prevention research. In: West S, ed. The science of prevention: Methodological advances from alcohol and substance abuse research. Washington D.C.: American Psychological Association, 1997;325-366.

Graham J.W., Hofer S.M., Piccinin A.M. Analysis with missing data in drug prevention research. In: Seitz L, ed. Advances in data anlysis for prevention intervention research. Washington D.C.: National Institute on Drug Abuse, 1994;13-63.

Hansen, H., Lopez-Iftikhar, M. M., and Alegria, M. (2002). The economy of risk and respect: accounts by Puerto Rican sex workers of HIV risk taking. Journal of Sex Research, 39(292-301).

Hicks, T.V. & Leitenberg, H. (2001). Sexual fantasies about one’s partner versus someone else: Gender differences in incidence and frequency. The Journal of Sex Research, 38 (1), 43-50.

Holzman, H.R. & Pines, S. (1982). Buying sex; The phenomenology of being a john. Deviant Behavior, 4, 89-116.

Inciardi, J. (1995). Crack, crack house sex, and HIV risk. Archives of Sexual Behavior, 24(3), 249-269.

Johnson, E.H. , Jackson, L.A., Hinkle, Y., Gilbert, D., Hoopwood T., Lollis, C.M., Willis, C., & Gant, L. (1994). What is the significance of black-white differences in risky sexual behavior? Journal of the National Medical Association, 86 (10), 745-759.

Kim, M.Y., Marmor, M., Dubinm N., & Wolfe, H. (1993). HIV risk-related sexual behaviors among heterosexuals in New York City: Associations with race, sex, and intravenous drug use. AIDS, 7 (3), 409-414.

Kinsey, A., Pomeroy, W.B., & Martin, C.E. (1948). Sexual behavior in the human male. Philadelphia: Saunders.

Kwiatkowski, C., and Booth, R. (2000). Differences in HIV risk behaviors among women who exchange sex for drugs, money, or both drugs and money. AIDS and Behavior, 4(3), 233-240.

McKeganey, N. & Barnard, M. (1996). Sex work on the streets: Prostitutes and their clients. Buckingham, UK: Open University Press.

Michael, R.T., Gagnon, J.H., Laumann, E.O., & Kolata, G. (1994). Sex in American: A definitive survey. Boston: Little Brown.

Monto M. Clients of street prostitutes in Portland, Oregon, San Francisco, and Sanata Clara, California, and Las Vegas, Nevada, 1996-1999 [Computer file]. (2000). Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2000 ed: Portland, OR: University of Portland [producer].

Monto M. (1999). Focusing on the clients of street prostitutes: a creative approach to reducing violence against women - summary report. Washington, D.C.: U.S. Department of Justice.

Monto, M. & Hotaling, N. (2001). Predictors of Rape Myth Acceptance Among Male Clients of Female Street Prostitutes Violence Against Women, 7 (3), 275-293.

Morse, E.V., Simon, P.M., Balson, P.M., & Osofsky, H.J. (1992). Sexual behavior patterns of customers of male street prostitutes. Archives of Sexual Behavior, 21 (4), 347-57

Muthen L, Muthen B. Mplus: The comprehensive modeling program for applied researchers. User's Guide Version 2.0. Los Angeles, CA: Muthen & Muthen, 2000.

Pedersen, W. & Blekesaune, M. (2003). Sexual satisfaction in young adulthood: Cohabitation, committed dating or unattached life? Acta Sociologica, 46 (3), 179-193.

Reinisch, J.M. 1990. The Kinsey Institute New Report on Sex. New York: St. Martin’s Press.

Rickert, E.J. & Rickert D.L. (1995). Different HIV risk profiles in samples of college students and homeless persons. Psychological Reports, 76 (3, part 2), 1123-1132.

Rinehart, N.J. & McCabe, M.P. (1998). An empirical investigation of Hypersexuality. Sexual and Marital Therapy, 13 (4), 369-384.

Rotosky, S.S., Galliher, R.V. Welsh, D.P.& Kawaguchi, M.C. (2000). Sexual behaviors and relationship qualities in late adolescent couples. Journal of Adolescence, 23 (5), 583-597.

Seto M, Maric A, Barbaree H. The role of pornography in the etiology of sexual aggression. Aggression and Violent Behavior 2001; 6:35-53.

Stack, S. & Gundlach, J.H. (1992). Divorce and sex.. Archives of Sexual Behavior, 21 (4), 359-367.

Stafford, L, Kline, S.L., & Rankin, C.T. (2004). Married individuals, cohabiters, and cohabiters who marry: A longitudinal study of relational and individual well-being. Journal of Social and Personal Relationships, 21 (2), 231-248.

Sullivan, E. & Simon, W. (1998). The client: A social, psychological, and behavioral look at the unseen patron of prostitute. In J.E. Elias, V.L. Bullough, v. Elias, & G. Brewer (Eds.).

Prostitution: On Whores, Hustlers, and Johns (pp. 134-154). Amherst, NY: Prometheus.

U.S. Bureau of the Census. (2006). American Fact Finder. (Available at: http://factfinder.census.gov/home).

Vanwesenbeeck, I., DeGraaf, R., Van Zessen, G., Straver, C.J., & Visser, J.H. (1993). Protection styles of prostitutes’ clients: Intentions, behavior, and considerations in relation to AIDS. Journal of Sex Education & Therapy, 19 (2), 79-91.

Wade, L.D. & DeLamater, J.d. (2002). Relationship dissolution as a life stage transition: Effects on sexual attitudes and behaviors. Journal of Marriage and the Family, 64 (4), 898-914.

Wiederman, M.W. (1997). Extramarital sex: Prevalence and correlates in a national survey. The Journal of Sex Research, 34 (2), 167-174.

Wortley, S., Fischer, B., & Webster, C. (2002). Vice lessons: A survey of prostitution offenders enrolled in the Toronto John School Diversion Program. Canadian Journal of Criminology, 44 (4), 369-402.

Return to Front Page

1. One study, from more than two decades ago, drawing on interviews with customers of prostitutes reported that these men were “occasional johns”. These men used prostitutes because they desired variety in their sexual experiences, not because they lacked sexual outlets/partners (Holzman & Pines, 1982).

2. However, the most frequent response from men in the diversion program was to deny that they solicited the decoy prostitute for sex. Fully one-quarter (25%) of men arrested and enrolled in the diversion program for prostitutes’ clients maintained their innocence and claimed to have not been seeking sex (Wortley, Fischer & Webster, 2002).

3. While we acknowledge that use of a convenience sample is not ideal, it is important to note that access to samples of persons involved in illegal, stigmatized or otherwise hard-to-access forms of behavior are important to understand, regardless of the limitations of the sample. While an epidemiological study drawing on a random sample of men would provide the strongest data, this is not feasible considering the required resources such would require. We caution readers to generalize our results with care.

4. We are particularly grateful to the anonymous reviewer who suggested that our final sample should contain only those respondents who reported engaging in sexual activity with a prostitute in the past year.

5. No questions specifically addressed the viewing of pornography on the Web.

6. This question does not differentiate whether the participant likes to receive rough sex, give rough sex, or both.

7. Marital status and feeling shy and awkward with women, while not reaching the level of significance in the final model, were retained because in the initial model they had yielded a significant relationship with frequency of sex in the past 12 months.

8. For this group, having a regular sex partner contributed somewhat more to explaining the variance in frequency of sex than the number of sex partners.

9. Vanwesenbeeck (2001) provides a review of the scientific study of sex work from 1990-2000; the review specific to customers is international in scope and includes men who patronize sex workers in varied venues, not just street-level prostitutes. The venue (e.g. a brothel in Thailand or on a street corner in Harlem, New York) in which sex work is conducted exerts a great deal of influence on the motivations and behaviors of both the sex worker and her/his ‘client’ (Vanwesenbeeck, 2001; Kwiatkowski and Booth, 2000; Inciardi, 1995; Hansen, Lopez-Iftikhar, and Alegria, 2002). Therefore, while Vanwesenbeck’s synthesis on the characteristics and motivations of customer’s is salient, further research is needed to understand the behaviors and motivations of the male clients of street-level prostitutes in the United States.