ArticlePDF Available

Abstract and Figures

Research on prostitution demand has well documented why men buy sex from girls and women, but there is very little understanding of the scope and scale of the issue. The lack of stable and valid measures of how many people buy sex from prostituted persons severely hinders the development of prevention and intervention efforts. This study developed a probability estimate of the population of active customers of online sex in the United States in a sample of 15 cities. In each city, the research team placed decoy online ads, advertising the sale of sexual services/prostitution, and collected text and voicemail data from potential sex purchasers. The resulting 677 phone numbers were analyzed using capture-recapture techniques to create an estimate of the number of online sex purchasers within each city. On average, within the 15 markets explored, 1 out of every 20 males over the age of 18 in a metropolitan city area was soliciting online sex ads. These results demonstrate (a) the viability of new techniques to estimate buyer populations and (b) preliminary figures on the number of purchasers buying sex online.
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
Download by: [Stephanie Ryon] Date: 12 October 2017, At: 10:44
Journal of Human Trafficking
ISSN: 2332-2705 (Print) 2332-2713 (Online) Journal homepage:
Invisible Offenders: Estimating Online Sex
Dominique Roe-Sepowitz, Stephanie Bontrager Ryon, Kristine Hickle, James
M. Gallagher & E. C. Hedberg
To cite this article: Dominique Roe-Sepowitz, Stephanie Bontrager Ryon, Kristine Hickle, James
M. Gallagher & E. C. Hedberg (2016) Invisible Offenders: Estimating Online Sex Customers,
Journal of Human Trafficking, 2:4, 272-280, DOI: 10.1080/23322705.2015.1107711
To link to this article:
Published online: 01 Sep 2016.
Submit your article to this journal
Article views: 129
View related articles
View Crossmark data
Invisible Offenders: Estimating Online Sex Customers
Dominique Roe-Sepowitz
, Stephanie Bontrager Ryon
, Kristine Hickle
, James M. Gallagher
and E. C. Hedberg
School of Social Work, Arizona State University, Phoenix, Arizona, USA;
School of Public Affairs, University of
Colorado Colorado Springs, Colorado Springs, Colorado, USA;
Department of Social Work, University of Sussex,
Brighton, UK;
Phoenix Police Department, Phoenix, Arizona, USA;
Morrison Institute, Arizona State University,
Phoenix, Arizona, USA
Research on prostitution demand has well documented why men buy sex from
girls and women, but there is very little understanding of the scope and scale of
the issue. The lack of stable and valid measures of how many people buy sex
from prostituted persons severely hinders the development of prevention and
intervention efforts. This study developed a probability estimate of the popula-
tion of active customers of online sex in the United States in a sample of 15 cities.
In each city, the research team placed decoy online ads, advertising the sale of
sexual services/prostitution, and collected text and voicemail data from potential
sex purchasers. The resulting 677 phone numbers were analyzed using capture-
recapture techniques to create an estimate of the number of online sex pur-
chasers within each city. On average, within the 15 markets explored, 1 out of
every 20 males over the age of 18 in a metropolitan city area was soliciting online
sex ads. These results demonstrate (a) the viability of new techniques to estimate
buyer populations and (b) preliminaryfiguresonthenumberofpurchasers
buying sex online.
Customers; demand;
prostitution; sex buyers;
sex market; sex trafficking
Sex buyers and the demand side of prostitution have received increased attention from the media,
legislators, and activist groups as awareness has grown about sex trafficking and the perils of prostitution.
Sex buyers of prostituted persons are rarely punished for their actions and significantly contribute to the
victimization of sex-trafficking victims. This lack of accountability has been noted by a large anti-
trafficking advocacy movement and has produced an increased awareness of the role of the demand
or the buyer of sex in sex trafficking. In some states, the attention has assisted in the implementation of
more stringent penalties for customers, specifically those buying sex from minors. Extralegal penalties,
such as the public shaming of buyers on billboards and Web sites, have increased along with financial
penalties, but little is known about the deterrent effect of these interventions because the scope of the
population of sex buyers is currently unknown. Because buyers of sex require anonymity almost as much
CONTACT Dominique Roe-Sepowitz Associate Professor, Arizona State University, 411
N. Central Avenue, Suite 800, Phoenix, AZ 85004, USA.
Dominique Roe-Sepowitz is an Associate Professor at Arizona State University, School of Social Work and the Director of the Office of Sex
Trafficking Intervention Research. Her research is focused on sex trafficking and conducts studies on sex traffickers, victims and sex buyers.
Stephanie Bontrager Ryon is an Assistant Professor at the University ofColoradoColoradoSprings.Herprimaryscholarlyinterestsrevolve
around victimization, exploitation and Criminal Justice processing. More specifically Dr. Ryon currently studies sex trafficking, sentencing
disparities, Disproportionate Minority Contact with the Justice System, and evidence-based prevention and intervention programing.
Kristine Hickle is a Lecturer in Social Work and Social Care at University of Sussex in the United Kingdom. Research interests include child
sexual exploitation (CSE) and sex trafficking; current projects include interdisciplinary research exploringpoliceresponsestoCSEintheUK,
and a national study in England evaluating the implementation and impact of a multi-agency model for practice with children and young
people affected by CSE. Commander James M. Gallagher is a 21-year veteran of the Phoenix Police Department and a researcher in the
Office of Sex Trafficking Intervention Research at Arizona State University. Previously, he was the head of the Vice Enforcement Unit where
he integrated evidence based research into his operational and investigative leadership philosophy to refocus the unit's efforts towards a
more enhanced victim centered and demand reduction driven orientation. Eric C. Hedberg is currently a senior research scientist at NORC at
the University of Chicago. His research focuses on evaluation and social capital.
© 2016 Taylor & Francis
2016, VOL. 2, NO. 4, 272280
Downloaded by [Stephanie Ryon] at 10:44 12 October 2017
as sexual gratification, they are elusive, complex to research, and, in most cases, hidden in plain sight in
our communities. Because of these challenges, they continue to be enigmatic with very little known about
how they buy sex, when they buy sex, and where they buy and receive sex services.
Research on prostitution demand has well documented why men buy sex from girls and women
(Monto, 2004,2010;Shively,Kliorys,Wheeler,&Hunt,2012)butlittleisknownabouttheimpactof
prevention efforts, such as stings, shaming Web sites,orreportingthearresttothemedia.Muchofthegap
in the literature is due to the lack of an available baseline number of buyers of sex to determine if these
deterrents reduce the number of buyers. Many scholars of the commercial sex industry sex note that valid
estimates of purchasers are very difficult to produce because of the nature of the behavior (Empel, 2012;
Yen, 2008). Despite the many obstacles to estimation, a few studies have attempted to determine the
number of men who purchase sex. These figures are essential to improve our understanding of lifetime and
current purchasing behaviors and are critical to creating social policy and law-enforcement actions to
systematically address, by city, the demand aspect of sex trafficking and prostitution in the United States.
Prior research on sex purchasing
As noted by Yen (2008), the supply side of the sex trade, legal or illegal, has received scholarly attention;
however, demand for sex has not made it into academic discussions or research agendas. While the study
does not provide original estimates of sex buyers, it does note that the scant studies on this topic produce
wildly different figures on how many men buy sex. Yen reported purchasing rates of males in Great Britain
(7%), Thailand (73%), and the United States (1669%) but cautions against any comparisons and questions
the validity of these figures. Empel (2012)echoesthesesentimentsarguingthehiddennatureand/orsocial
acceptance of buying sex makes valid estimates difficult to obtain. However, Empel found that 1015% of
men in developed countries pay for sex.
Previous attempts at estimating the population of sex customers, also called Johns,have been made in
the United States through questionnaires including the General Social Survey (as cited in Monto, 2010;
Smith, Marsden, & Hout, 2011)withanestimatethat14%ofmensurveyedhadpreviouslyboughtsexin
their lifetime. Additionally, the National Health and Social Life Survey (as cited in Michael, Gagnon,
Laumann, & Kolata, 1994;Monto,2010)foundthat16%ofmenhadvisitedaprostituteintheirlifetime.
Unsupported media reports have estimated that between 16% and 80% of men pay for sex (Bennetts, 2011).
The majority of information known about sex customers, particularlymen buying sex from women, is
based on what is collected from John Schools,which are court-ordered programs for sex buyers offered
in 58 cities and counties in the United States (Monto, 2004; Monto & McRee, 2005; Shively et al., 2012).
The John Schoolattendees are caught in law-enforcement demand-reduction operations. These efforts
are limited in scope and impact and are directly relative to the effort and priority agencies place on the
low level, nonviolent crime. Results of a recent study of sex trafficking in the United States found that
many law-enforcement agencies do not prioritize sex trafficking as a serious public safety issues
(Bontrager Ryon, Keith, & Brown, 2012). Thus, information on purchasing behaviors collected through
John Schoolsare unlikely to represent sex-buying customers as a population.
Wilcox, Christmann, Rogerson, and Birch (2009), in their review of 181 research studies on prostitu-
tion demand, found there were significant methodological problems with most studies, major gaps in the
research, and weak or inconclusive findings on what impacts the demand for commercial sex. Wilcox
et al. (2009) also stated that because buying sex in most cultures is stigmatized and out of sight,
developing accurate and reliable estimates of the number of people who buy sex has been difficult. In
the United States, there are no estimates of the current population of active sex buyers.
The current study
There are a number of challenges to detecting and studying customers of online sex ads, currently the
most pervasive method of sex trafficking. Online-sex-ad customers are invisible offenders who are rarely
exposed to the public except by episodic targeted enforcement by police (Sanders, 2008). Online-sex-ad
Downloaded by [Stephanie Ryon] at 10:44 12 October 2017
customers experience a lower risk of being caught by police than street-level prostitution customers due
to the anonymity inherent in Internet-based solicitation. These risk-mitigating factors include the fact
that online-sex-ad customers remain out of (physical) sight of law enforcement while soliciting for sex,
the arrangements are made by phone or e-mail, and the sex exchange is done in private in a hotel,
brothel, or private home. Whereas street-level prostitution customers make deals and many times
commit sex acts in public spaces where they are more likely to receive law-enforcement attention.
The present study examines the number of men who respond to an online decoy ad, using this
information to develop valid probability estimates of sex purchasing through online venues. Our research
questions are the following:
(1) What is the population estimate for each city of online-sex-ad customers?
(2) What is the rate of demand (online-sex-ad customers) relative to the total exposed popula-
tion (all males over 18 years old) in each metro city area?
(3) Which of the cities have the highest rates of demand?
By answering these questions we hope to develop new knowledge about customers of online sex
ads, information that will greatly assist policy makers and law-enforcement agencies in creating
action plans to address the scope of demand for online sex services.
Adding to the generalizable knowledge of sex buyers, this study builds on prior research to provide
information that is unique and valuable to responding agencies. The assessment is the only study to
provide current estimates of sex buyers in the United States. In addition, it is the only study to use
probability techniques to arrive at those estimates. While limited to the online market, the research
begins to address the knowledge gaps noted by leading scholars in the field.
Data and methods
Fifteen cities were included in this study with the goal of having a diversity of metropolitan areas in the
United States. Cities on the east coast included Boston, New York City, Baltimore, and Atlantic City.
Miami represented the southeastern United States, Chicago, Kansas City, Minneapolis, and Salt Lake
City represented the Midwestern United States, and Portland, San Francisco, San Diego, Phoenix, Las
Vegas, and Houston covered the west and southwestern United States. In each of these cities, there are
multiple in-person and online methods of buying sex. Exploring all available avenues of sex purchasing
was not feasible and we limited the study to online venues. The two most popular sex-selling sites were
utilized to examine sex-purchasing behavior: (Casual Encounters section) and backpage.
com (adult entertainment section, escorts). In all cities, the Craigslist ads were removed for inap-
propriate content by the company and were therefore removed from the study.
Research personnel contacted the law-enforcement unit in charge of vice enforcement in each city to notify
them about the study. The decoy ads were sent to them in advance to prevent them from responding to the
ad if they were doing enforcement activities during the research period. The research team developed two
normative sex ads for the study based on prior work with law-enforcement sting operations and analysis of
existing sex ads.
Each ad had a unique phone number with a local area code. The ads were placed at 2 pm
(local time) on Friday, one week apart. All call and text data were recorded using a Voice Over Internet
Protocol (VoIP). The VoIP system collected information from each respondent including time of call,
phone number, length of contact, and text and voice messages. See Table 1 for raw data.
The Arizona State University Institutional Review Board approved this study as to methodological design and ad content.
Downloaded by [Stephanie Ryon] at 10:44 12 October 2017
This study used a capture/recapture sampling technique, which has been used in ecology and
population biology, as well as in demography research. Previous studies using the capture/recapture
technique include creating an estimate of the density of a population of animal such as jaguars in
Brazil (Soisalo & Cavalcanti, 2006) and tigers in India (Karanth, Kumar, & Nichols, 2002).
Sociologists have also used the technique to develop estimates of a problem population such as
drug users in London (Hickman et al., 2002), heroin users in Australia (Larson, Stevens, & Wardlaw,
1994), and Type 2 diabetes in the United Kingdom (Ismail, Beeching, Gill, & Bellis, 1999).
An example of this type of sampling is counting how many deer are in a wooded area at a given time. A
spotter will be sent to the woods to photograph as many deer as he or she can during a 10-hour period. The
photographs are then examined by researchers to identify a unique feature of each deer and a list is created.
period. Those photographs are compared to the original set and the overlap is counted as a capture-
recapture variable. This estimate is the foundation of the formula to determine the size of the deer
Using capture-recapture, we matched individualsphone numbers from two random samples of
sex-ad respondents. Using the overlap figure from the two independent samples, we estimate the size
of the total population of online-sex-ad purchasers for each city. In this instance, the exposed
population includes all males in the metro area over the age of 18 (American Fact Finder, 2011)
that appear in multiple samples taken from the same population. Both in this study of online-sex-ad
customers and the example of deer in the woods, the overall population is not closed. To minimize
issues of attrition and new recruits, we kept the time between collecting samples as short as possible.
Ads were placed one week apart, on the same day of the week at the same time of day on the same
online venue in each city. We attempted to avoid issues of trap addiction or trap avoidance by
placing the identical ads. Although their attempt to purchase sex from the first online sex ad was not
completed, this is not unusual and should not impact their decision to attempt a second contact.
To estimate the number and percent of online-sex-ad purchasers in each city, we use both the
Chapman (1951) analysis and the Cormack (1992) technique. The data provide information on n
(number of phone numbers first-ad responses), n
(number of phone numbers in response to second
ad), and m(the number of phone numbers recaptured). In order to estimate the population we need
to either directly estimate N(the total population of online-sex-ad responders) or q(the number of
customers not captured on either list).
Table 1. Raw Data Collected in Response to Online Sex Ads.
Unique Calls/
Total Calls for Ads
Including Dups
(7 days each ad, 2 ads)
Number of
Number of
% of Calls Within
Local Area Codes
Atlantic City 24 39 15 24 17.90
Baltimore 17 22 6 16 59.10
Boston 23 27 7 20 55
Chicago 20 31 9 23 67.70
Houston 19 31 10 21 74.20
Kansas City 35 70 0 70 88.60
Las Vegas 26 33 17 16 60.60
Miami 26 39 24 15 76.90
Minneapolis 24 27 17 10 85.20
New York City 7 10 3 7 70
Phoenix 62 79 28 51 79.70
Portland 49 79 23 58 54.70
Salt Lake City 48 68 52 16 66.20
San Diego 20 33 16 17 76
San Francisco 18 24 10 14 62.50
Downloaded by [Stephanie Ryon] at 10:44 12 October 2017
There are two ways to find N. First is to estimate Ndirectly using a formula by Chapman (1951)
for small samples:
with an associated standard error of the following:
SEðNÞ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
The 95% confidence interval in this case is simply the following:
However, this can sometimes produce very large confidence intervals. Thus, we also used a method
introduced by Cormack (1992) that employed a Pearson chi-square algorithm that found a smaller
and larger estimate for qthat satisfied a pvalue of less than .05. Thus, two values of Nare calculated
with these values for q. If no recaptures were found, we were unable to calculate a confidence interval
and another method to estimate q, and thus Nis fit to a Poisson regression model (Cormack, 1989)
on a simple three-row dataset:
where the model is expðYÞ¼Xb and the exponent of the intercept is the estimate value for q.We
then use the standard error of this estimate as the standard error for N. For each city, we first tried
the Poisson method. If that failed to produce a confidence interval or estimate, we used Chapmans
method to estimate Nand the chi-square method to estimate the confidence interval. If either of
those methods failed, we then used Chapmans standard error to develop the confidence interval.
The exposed populations for this study are all of the males over the age of 18 years old in the metro
area of each city. These were determined using American Fact Finder (2011) and used the one-year
population estimates of males over age 18 from the American Community Survey from the most recent
year available, 2011. This is the basis by which we calculated the percentage estimate of the male
population in each city to be online-sex-ad customers. This exposed population is the denominator.
The numerator was calculated by multiplying the average number of ads within each city by
ChapmansN. For example, in Atlantic City the exposed population was 559,126 males. During the
study, there were an average of 206 sex ads posted to the two online venues. This figure is multiplied by
50, the estimate produced from the Chapman Nformula. The final estimate of 1.8% of the adult male
population purchasing online sex is derived by dividing 10,275 by 559,126. See Table 2 for sex buyer
estimations for each city. This model reports the number of active online-sex-ad customers on the first
data collection date (mid-June 2013) and considers issues of attrition (customers no longer buying sex
from an online-ad source) and new customers (who are entering the market to buy sex online in that
city for the first time). These estimates are conservative with consideration of confidence intervals and
standard error rates.
In List B
Yes No Total
In List A Yes mn
No n
Total n
Downloaded by [Stephanie Ryon] at 10:44 12 October 2017
There are a number of study assumptions necessary to consider when interpreting the findings
from this evaluation: (a) All men over the age of 18 in each city are potential customers for online
sex ads; (b) the ad placed on was normative to all of the other ads posted on backpage.
com and was not detected as a deceptive ad by potential customers; (c) that the callers (customers)
called other sex ads posted on during the 24 hours after our ads were posted; (d) that
a significant percentage of the customers were from the local area; (e) that the ads were placed on
two average Fridays in late spring 2013.
Research findings
We received a total of 677 contacts from the ads, either texts or calls, from online-sex-
ad customers in the 15 cities in response to the two ads. The majority (69.6%) of the contacts were
made during the first 24 hours after the ad was posted ranging from 48% in San Francisco to 90% in
New York City. There were 677 total responses from 451 unique phone numbers.
Overall, 38% of purchasers sent a text response to the ads and the other 62% placed phone calls.
Sixty-six percent of all contacts were from respondents with local area codes, indicating that the
market is driven by local demand. Recaptured phone numbers, that is, online-sex-ad customers who
called in response to both ads, were found in six cities. One recaptured phone number was found in
both Baltimore and Chicago, two were found in Salt Lake City and Atlantic City, and three were
found in Portland and Phoenix.
Estimate of population
An estimate was made for the number and percent of online-sex-ad customers in each city. On
average, 1 out over every 20 males over the age of 18 in each metropolitan city area was soliciting
online sex ads. The findings ranged from one out of every five males (Houston, 21.4%) to less than 1
of 166 males (San Francisco, 0.6%). In Houston, we found that there were an estimated 169,920
males who were soliciting online sex ads, while, in Phoenix, there were 78,412 males who were
soliciting online sex ads.
Examining those cities with a calculation of ChapmansN, the results are consistent ranging from
a low rate of 1.78 in Baltimore to a high of 4.9 in Phoenix. In cities without a recapture, the estimated
Table 2. Estimates of Sex Buyer Population for Each City.
% of Males in City Who Call
Sex Ads
(Confidence Intervals)
Estimated Sex Ad Customer
Average # of Ads Posted on
in a 24-hr Period (Friday 2pm)
Atlantic City 1.4 (0.53.2)* 10,275 206
Baltimore 1.8 (1.0 2.1)* 17,766 211.5
Boston 7.6 (4.8 10.3)** 1,30,416 247
Chicago 2.4 (1.4 3.1)** 83,478 518.5
Houston 21.4 (13.8 29)** 1,69,920 472
Kansas City 14.5 (9.1 17.9)** 1,06,624 98
Las Vegas 13.5 (9.1 19.9)** 99,910 515
Miami 6.6 (4.2 8.9) ** 1,40,184 265.5
Minneapolis 4.9 (3.26.7)** 60,120 167
New York City 3.9 (07.6)*** 21,514 341.5
Phoenix 4.9 (3.46.4) ** 78,412 307.5
Portland 3.7 (2.6 4.8)** 31,282 145.5
Salt Lake City 2.6 (0.6-4.7)* 10,675 87.5
San Diego 3.1 (0 7)* 36,890 310
San Francisco 0.6 (0.1 1.3)* 9,504 96
*Chapman Confidence Interval. **Chi-Square Confidence Interval. ***Poisson Confidence Interval.
Downloaded by [Stephanie Ryon] at 10:44 12 October 2017
rate of purchasers varies considerably from a high of 21.4 in Houston to a low of 0.06 in San
Francisco. Additionally, the standard error for Chapman estimates is much lower than those
produced from the Cormack (1992) algorithm. These results indicate that the Chapman method is
a better technique producing consistent estimates with low variability. While the other techniques
produce estimates with wide dispersion and higher standard errors. Based on the Chapman N
calculations, on average, 2.88% of adult males within each city were purchasing sex online during
the study.
In acknowledgement that some of the sex-ad customers may not be from the city where the ad
was placed, we conducted analytics of the area codes and differences between texts and voice calls.
Localarea codes ranged between 54.7% (Portland) to 88.6% (Kansas City), except for Atlantic City,
which was an outlier at 17.9%. This suggests that in most cities, purchasers buy from their local
market and that the market is relatively stable over time. See Table 3 for city rankings for sex buyer
The study demonstrates that the number of adult males soliciting sex online varies considerably from
one city to the next. This is a key finding for local efforts to combat demand for illegal commercial
sex. A clear policy implication is that areas with more purchasers, for example, Houston and
Phoenix, should commit more resources to demand-side interventions. Methodologically, the results
also differ significantly for those with and without a recapture during the second timeframe. Cities
without a recapture utilized the Cormack algorithm, which produced much larger estimates of the
purchasing population. This technique is less reliable than ChapmansN, and the results here should
be interpreted with caution. Of those cities with a recapture, the average percent of adult males
buying sex online was 3% (compared to 6% for all cities).
An important strength of this study is that the method of data collection and analyses can easily
be replicated and changes in the estimated population over time or pre-post of an intervention can
be calculated. Limitations of this study include that we were only able to gather useable data from
one Web site ( in 15 metro areas. Additionally, we did not make contact with any of
the customers to verify their intent to solicit sex from the posted ad. However, qualitative data from
the respondents strongly suggest they were making contact to purchase sex. As previously men-
tioned, the lack of recapture in nine locations also limits the validity of the estimates in these
locations and impacts the generalizability of results to similar cities. Finally, the research is restricted
to one illegal commercial sex venue, Internet-based outlets, which represent only one of many
Table 3. Ranking of Cities by Sex Buyer Population.
Rank City
Percent of Male Population over Age 18
That Are Online-Sex-Ad Customers
1 Houston 21.40
2 Kansas City 14.50
3 Las Vegas 13.50
4 Boston 7.60
5 Miami 6.60
6 Minneapolis 4.90
7 Phoenix 4.90
8 New York City 3.90
9 Portland 3.70
10 San Diego 3.10
11 Salt Lake City 2.60
12 Chicago 2.40
13 Atlantic City 1.80
14 Baltimore 1.80
15 San Francisco 0.60
Downloaded by [Stephanie Ryon] at 10:44 12 October 2017
possible markets (brothels, street prostitution, massage parlors, and strip clubs are other known
venues). As a result, the studys estimates are certainly conservative, but the current research is
unable to determine how much of all sex buying is captured through these techniques.
Even within the context of these limitations, the findings are intended to set a baseline of demand
for illegal commercial sex. Without some understanding of the size of this population, we cannot
begin to develop interventions, through policy or practice, to address this criminal behavior. The
intention of these findings is to inform law enforcement, advocacy groups, and policy makers about
the issue and breadth of the nearly invisible problem of online sex customers.
This study assessed the demand for illegal commercial sex using unobtrusive techniques that protect
the identity of potential purchasers. In doing so, it creates the only empirically derived estimates of
the number of men actively seeking to purchase sex. These estimates have a functional purpose
establishing some figure of demandas well as practical implications.
For practitioners and professionals responding to sex trafficking, the study demonstrates the sheer
volume of potential purchasers, which averages in the thousands per city. This indicates in a clearly
demonstrable and measurable way what is most often described either anecdotally or emotionally
that demand reduction efforts must receive equal attention as targeting traffickers and rescuing
victims. For law enforcement, these results compel a review of demand-reduction practices that
typically are not a component of their sex-trafficking response.
Practitioners and professionals are also likely unprepared to deal with demand on the scale
implicated by this study. While the research only taps into potential purchasers, if even 10% of
the respondents actually buy sex, there is a corresponding number of victims. For instance, if 10% of
the lowest range of demand becomes realized as an act of sex trafficking, local communities such as
Atlantic City, New Jersey, would have over 1,000 victims. These victims would require a diverse
range of services and interventions that would likely test the service communitys capacity to
respond in a meaningful way.
Additionally, given the volume described in this study, it is clear law-enforcement and prosecutorial
agencies would soon be overwhelmed with work should they target demand from a strictly prohibitive
and punitive perspective. The findings of this study strongly imply that a more holistic approach is
necessary to change purchasing behaviors. This includes programs to educate potential purchasers on
the consequences of their actions and public awareness campaigns to involve the public in combatting
demand. Finally, communities must engage their leaders in a dialogue that says demand, long excused
as boys will be boys,is a community safety issue that will not be tolerated any longer.
Thorn Digital Defenders of Children and ASU Office of Sex Trafficking Intervention Research.
E. C. Hedberg
American Fact Finder. (2011). Males in metropolitan areas. Retrieved from
Bennetts, L. (2011, July 25). The John next door. Newsweek,158(4), 6063.
Bontrager Ryon, S., Keith, P., & Brown, E. (2012). U.S. Office of Juvenile Justice and Delinquency Prevention.
Trafficking In Persons Symposium: Final Report.Appleton, WI: AMBER Alert Training and Technical
Assistance Program.
Downloaded by [Stephanie Ryon] at 10:44 12 October 2017
Chapman, D. G. (1951). Some properties of the hypergeometric distribution with applications to zoological sample
census. University of California Publications in Statistics,1, 131160.
Cormack, R. M. (1989). Log-linear models for capture-recapture. Biometrics,45, 395413.
Cormack, R. M. (1992). Interval estimation for mark-recapture studies of closed populations. Biometrics,48, 567576.
Empel, E. (2012). The future of the commercial sex industry: As new technologies impact the products and services of the
sex industry, other businesses will find new opportunities in the worlds oldest professions. The Futurist,46,2640.
Hickman, M., Cox, S., Harvey, J., Howes, S., Farrell, M., Frischer, M. . .. Tilling, K. (2002). Estimating the prevalence of
problem drug user in inner London: A discussion of three capture-recapture studies. Addiction,94, 16531662.
Ismail, A., Beeching, N., Gill, G., & Bellis, M. (1999). Capture-recapture-adjusted prevalence rates of Type 2 diabetes
are related to social deprivation. QJM: International Journal of Medicine,92, 707710. doi:10.1093/qjmed/92.12.707
Karanth, K. U., Kumar, N. S. & Nichols, J. D. (2002). Field surveys: Estimating absolute densities of tigers using
capturerecapture sampling. In Karanth, K. U. & Nichols, J. D. (Eds), Monitoring tigers and their prey: A manual for
researchers, managers and conservationists in Tropical Asia (pp. 139152). Bangalore, Bangladesh: Centre for
Wildlife Studies.
Larson, A., Stevens, A., & Wardlaw, G. (1994). Indirect estimates of hiddenpopulations: Capture-recapture methods
to estimate the numbers of heroin users in the Australian capital territory. Social Science & Medicine,39, 823831.
Michael, R. T., Gagnon, J. H., Laumann, E. O., & Kolata, G. (1994). Sex in America: A definitive survey. Boston, MA:
Little, Brown.
Monto, M. A. (2004). Female prostitution, customers, and violence. Violence Against Women,10, 160188.
Monto, M. A. (2010). Prostitutescustomers: Motives and misconceptions. In R. Weitzer (Ed.), Sex for sale:
Prostitution, pornography and the sex industry (2
ed., pp. 233254). New York, NY: Routledge.
Monto, M. A., & McRee, N. (2005). A comparison of the male customers of female street prostitutes with national
samples of men. International Journal of Offender Therapy and Comparative Criminology,49, 505529. Doi:10.1177/
Sanders, T. (2008). Paying for pleasure: Men who buy sex. Portland, OR: Willan.
Shively, M., Kliorys, K., Wheeler, K., & Hunt, D. (2012). National overview of prostitution and sex trafficking demand
reduction efforts (Final Report 2008-IJ-CX-0010). Retrieved from
Smith, T. W., Marsden, P. V., & Hout, M. General Social Survey, 19722010 [Cumulative File] (ICPSR31521-v1).
Storrs, CT: Roper Center for Public Opinion Research, University of Connecticut/Ann Arbor, MI: Inter-university
Consortium for Political and Social Research [Distributors], 2011-08-05. 10.3886/ICPSR31521.v1
Soisalo, M., & Cavalcanti, S. (2006). Estimating the density of a jaguar population in the Brazilian Pantanal using
camera-traps and capture-recapture sampling in combination with GPS radio-telemetry. Biological Conservation,
129, 487496. Doi:10.1016/j.biocon.2005.11.023
Wilcox, A., Christmann, K., Rogerson, M., & Birch, P. (2009). Tackling the demand for prostitution: A rapid evidence
assessment of the published research literature. Research Report 27. Retrieved from http://webarchive.nationalarc
Yen, I. (2008). Of vice and men: A new approach to eradicating sex trafficking by reducing male demand through
educational programs and abolitionist legislation. Journal of Criminal Law and Criminology,98(2), 653686.
Downloaded by [Stephanie Ryon] at 10:44 12 October 2017
... The easy access to transportation in metro areas makes the frequent movement of victims from city to city very easy for traffickers and facilitates their ability to prevent them from building connections and familiarity with a region as this might cause them to feel more comfortable seeking help [6]. Additionally, Roe-Sepowitz, et al. [15] explored demand in the online commercial sex industry by collecting data on responses to decoy online sex ads in 15 different cities. Their finding estimate that 1 of 20 males is soliciting sex online. ...
... Their finding estimate that 1 of 20 males is soliciting sex online. For San Diego the estimated market size of for online sex trafficking consists of roughly 36,890 men with populations as high as 169,920 johns in some markets [15]. ...
Conference Paper
Full-text available
Technology has dramatically changed the way criminals conduct their illicit activities. Specifically, the Internet has become a major facilitator of online human sex trafficking. Traffickers are using these technologies to market their victims which presents new challenges for efforts to combat sex trafficking. This study used knowledge management principles and natural language processing methods to develop an improved ontology of online sex trafficking ads. The language of these ads is constantly evolving; therefore, this study explored the role of a new type of indicator, emoticons, to the ontology of human trafficking indicators.
... In a review of 181 studies on commercial sex, Wilcox, Christmann, Rogerson, and Birch (2009) reported significant methodological issues in estimating demand for prostitution. Given that prostitution is largely illegal in the United States, controversial, stigmatizing, and buyers can solicit online and out of public view, estimates of purchasing behavior have been difficult to assess (Monto & Milrod, 2014;Roe-Sepowitz, Bontrager Ryon, Hickle, Gallagher, & Hedberg, 2016;Wilcox et al., 2009). Attempts to fine-tune the methodologies used to produce estimates of sex purchasing behaviors have included the use of large-scale surveys and a capture-recapture approach. ...
... They found that 1 out of every 20 males adult males in a metropolitan city area was soliciting online sex ads or that 3% of adult males in these cities were buying sex online. This capture-recapture approach used by Brewer et al. (2008) and Roe-Sepowitz et al. (2016) does require knowing that either a charge has been filed via official law enforcement records or that the intention of someone who contacts the ads is in fact to purchase sex. In using the capture-recapture method online, changes to soliciting website availability (e.g. ...
Purpose: Estimating the size of the sex buyer market in the United States has been stymied by methodological and sampling challenges. Given known methodological issues in self-reporting and the sensitive nature of purchasing sex, current research faces challenges in providing estimates of demand for purchasing sex. This study used a unique approach to estimate the prevalence of sex buying by men over the age of 18 in the United States. Methods: This study employed a double list experiment and a direct ask question to a nationally representative sample of 2525 adult males to estimate the size of the sex buyer market in the United States. Results: The double list experiment found the prevalence of sex purchasing is roughly 1 of every 50 adult males in the U.S. (2%) over the 3 years. The direct ask questions found that roughly 1 of every 25 males in the U.S. (4%) had purchased sex in the past 3 years. How they purchased and who they purchased sex from was also explored. A total of 80 respondents reported buying sex during the past 3 years. They spent an average of $120 for their most recent sex-buying encounter. Most sex buyers reported buying sex from an offline venue (street, bar, or massage parlor) and 81% purchased sex from a female. Applying the estimate of sex buying behavior to the general male population in the U.S. results in approximately 4 million men over the age of 18 purchasing sex the past 3 years. Conclusions: The findings from both the double list experiment and the direct ask question indicate that buying sex is a relatively rare phenomenon among adult males, however, when accounting for multiple purchases and extrapolating to the entire population the estimated purchases is in the millions. Applying the double list experiment and a direct ask question to a nationally representative sample provides a new way to capture estimates while addressing some of the limitations of previous methods. Such findings have implications for both the criminal justice and public health sectors.
... In addition to high profits, these use of online advertisement websites provide increased accessibility for individuals with a click of the button, and in certain circumstances, complete anonymity (Beckham & Prohaska, 2012;Kosloski et al., 2017), increasing the potential clientele pool. As evidence of this, consider that in a study of fifteen cities, it was found that 1 out of 20 adult males were utilizing online advertisements for paid sexual encounters (Roe-Sepowitz et al., 2016). In another study of "johns" or clients of prostitutes soliciting services over the web, Kosloski et al. (2017) measured over 200 responses to advertisements in a single 24-hour period. ...
Prior research has associated and as sources of victimization, which in part resulted in the closure of the erotic services of each respective website. However, research also claims the introduction of Craigslist was associated with a reduction in female homicide rates across 30 large cities. This research acts as a supplemental analysis to Cunningham et al. by analyzing if, considered to be Craigslist’s successor, has similar effects on female homicide rates. When including measures of interest in each respective website, we find that Backpage is associated with a decrease in homicide rates for women. The purpose of this study is to extend the study conducted by Cunningham et al. through supplemental analysis. To determine the effect of online clearinghouses on female homicide rates, interest measures in Backpage along with the female homicide rates from 120 single city metropolitan and micropolitan areas over 14 years (2004–2018) were analyzed using multiple regression analyses. The regression analyses show that there is a statistically significant relationship between interest in Backpage and homicide rates for women. We find that Backpage is associated with a decrease in homicide rates for women.
... In the research herein, no such distinction is made; we consider the full UK Backpage corpus, which includes both women and men (though to a far less extent owing to the fact that a majority of sex providers are women), in our studies. Another study presented methods on estimating online sex customers (hence, it was focused on the demand, rather than the supply, side of the sex business) (Roe-Sepowitz et al. 2016). For a broader understanding of sex work in the digital era, we recommend the synthesis in Jones (2015). ...
Full-text available
Abstract Online sex has become a fast-growing business in both developing and developed network, with advertisements of (not necessarily unique) individuals numbering in the hundreds of millions across different Web portals. One such major hub of sex advertisement activity, before it was shut down by US federal agencies, was The website was a classifieds-advertising portal that had become the largest marketplace for buying and selling sex by the time that federal law enforcement agencies seized it in April 2018. Since then, investigations have been actively underway. However, the data (which has recently been made available to us for research on UK Backpage) also offers valuable insights into the nature of the online sex business, including complex properties that can be best studied using network science. One of the challenges, however, is a rigorous modeling of the data as a network, since the primary data are web advertisements and metadata (backend database) on accounts that posted that ad. In this article, we conduct an empirical study of an important sample of the online sex marketplace using UK backpage, including presenting a methodology for constructing simple ‘activity networks’ that define some notion of real-world collaboration or connection between two entities (in our case, at the level of ad-posting accounts) and then studying the properties of these networks. We gather a set of insights into a domain that has not been studied at scale, let alone a national level, but that is continuing to be a growing social problem for many countries.
... Unfortunately, such robust measures of the demand for sex services do not exist. In one of the only studies that attempts to estimate the demand for sex services, Roe-Sepowitz et al. (2016) estimate the demand for sex services in 15 large municipalities in the US. On average, the study finds that 1 out of every 20 males over the age of 18 years old in these jurisdictions was soliciting online sex ads. ...
Full-text available
The movement of many human interactions to the internet has led to massive volumes of text that contain high-value information about individual choices pertaining to risk and uncertainty. But unlocking these texts’ scientific value is challenging because online texts use slang and obfuscation, particularly so in areas of illicit behavior. Utilizing state-of-the-art techniques, we extract a range of variables from more than 30 million online ads for real-world sex over four years, data significantly larger than that previously developed. We establish prices in a common numeraire and study the correlates of pricing, focusing on risk. We show that there is a 15-19% price premium for services performed at a location of the buyer’s choosing (outcall). Examining how this premium varies across cities and service venues (i.e. incall vs. outcall) we show that most of the variation in prices is likely driven by supply-side decision making. We decompose the price premium into travel costs (75%) and the remainder that is strongly correlated with local violent crime risk. Finally, we show that sex workers demand compensating differentials for the risk that are on par with the very riskiest legal jobs; an hour spent with clients is valued at roughly $151 for incall services compared to an implied travel cost of $36/hour. These results show that offered prices in the online market for real-world sex are driven by the kinds of rational decision-making common to most pricing decisions and demonstrate the value of applying machine reading technologies to complex online text corpora.
The geographic concentration and diffusion of crime and deviancy are longstanding criminological inquiries, yet few studies have examined how certain illicit behaviors transcend neighborhood borders and connect neighborhoods in patterns of crime and deviancy. A structural neighborhood interdependence may account for the enduring nature and spread of crime, making it critical to understand how neighborhoods are connected in crime patterns to guide crime prevention and disruption efforts. This study examines neighborhood interdependence through the case of repeat buyers of commercial sex in illicit massage businesses in a metropolitan city in the United States. By frequenting venues for illicit commercial sex in multiple neighborhoods, buyers create inter-neighborhood connections through which the demand for an illicit market can spread across neighborhoods. Using online review data about buyers of commercial sex, this study analyzes this neighborhood interdependence as a network comprised of nodes (“neighborhoods”) and edges (“connections between neighborhoods”). Exponential random graph models were used to analyze how characteristics of neighborhoods, the space between neighborhoods, and the overall network of neighborhoods explain inter-neighborhood connectivity in an illicit market for commercial sex. The implications for research, policy, and practice will be discussed.
Human trafficking, the exploitation of humans for monetary gain or benefit, is a widespread humanitarian issue that is typically sub‐classified into labor and sex trafficking. In the last decade, sex traffickers have used online classified advertisements to advertise sexual services. Although these advertisements are visible to the general public and law enforcement, the volume of ads, the frequency with which their posting locale changes, and the use of obfuscation tactics make it difficult for law enforcement agencies to react. Existing products for law enforcement focus on identifying, tracking, and correlating individual activity by performing deep searches for specific information against a database of historical posts. While this deep search capability is useful for investigating specific cases, it overlooks higher‐level patterns that exist in ads. Using a website that has been linked to several sex trafficking‐related arrests, we demonstrate a framework for harvesting, linking, and detecting these patterns in a dataset comprised of more than 10 million advertisements targeting U.S. cities. Our framework combines information systems and operations research concepts to identify groups of posts based on text, phone numbers, and pictures; determine circuits associated with post groups, and predict future movements using four different methods. Our description of the framework and comparison of the grouping and prediction methods provide insights that can assist law enforcement agencies to combat individuals/organizations involved in illicit sexual activities, including sex trafficking, proactively. Also, this demonstration provides researchers interested in developing advanced interdiction models targeting illicit sexual activities with a clear picture regarding available data formats.
The purpose of this study was to examine the narrative of sex buyers in an unmoderated online forum. Using a feminist critical discourse analysis (FCDA) and intersectionality approach, we investigated overt and subtle ways power inequalities were present in the discourse of men who bought sex in Chicago. Four main themes emerged: (a) toxic masculinity; (b) violence against women; (c) intersectionality of sexuality, race, and age; and (d) the need to maintain the community. Our findings imply that johns' self-described monger identity is closely associated with maintaining, perpetrating, and minimizing violence against women.
Background Previous research has tied adverse childhood experiences (ACEs) to a variety of deleterious mental health, physical, and behavioral outcomes. There has been some examination of the relationship between ACEs and selling sexual services, but not on the relation of ACEs to purchasing. Objective We hypothesize a cumulative impact of ACEs on the propensity to purchase and buy and sell sex. We further hypothesize that childhood sexual abuse will have unique impacts on buying and selling. Participants & setting We recruited participants who had ever/never exchanged money or things of values for sex through Amazon MTurk (n = 930). Methods Using logistic regression, we examined how cumulative ACEs and each separate ACE increased propensity to buy or sell sex. We controlled for sex, age, race, employment status, and sexuality. Results Utilizing the analysis from cumulative ACEs found that the propensity to buy (odds ratio 1.11***) and sell sex (odds ratio 1.094**) increased as cumulative ACE score increased. Bisexuals had high propensity of both buying (odds ratio 2.12) and selling sex (odds ratio 2.74). Women (odds ratio 0.53) and people of color (odds ratio 0.65) where more likely to sell than others. For odds of buying sex, childhood sexual abuse (odds ratio 1.57) had the most impact. For selling sex, childhood sexual abuse (odds ratio 1.96) and household physical violence (odds ratio 2.73) increased propensity while household mental abuse (odds ratio 0.57) decreased propensity. Conclusions Understanding the impact of ACEs is important to understand participation as a buyer and seller in the commercialized sex market.
Full-text available
Through a critical review of existing research, this article argues that (a) customers have been excluded from many debates surrounding prostitution, and including them will create a more robust dialogue regarding prostitution; (b) though prostitution seeking is often treated as a natural part of masculine sexual experience, most men do not pursue prostitutes, and very few are regular users; (c) many commonsense understandings of men’s motivations for seeking prostitutes are consistent with existing research; (d) a series of attitudinal orientations may help explain why some men meet their desires through prostitution and others do not, and may also help to explain some of the violence experienced by prostitutes; and (e) customers actively construct their encounters with prostitutes in ways that support their understandings of prostitution and their sexual desires, often working to maintain a sense of mutuality and consent. Incorporating customers into academic scholarship on prostitution can lead to more balanced research that better informs public policy.
Full-text available
This rapid evidence assessment of the published research literature is part of the Tackling Demand for Prostitution Review (Home Office, 2008a), which aimed to assess what further action the Government and other agencies could do to reduce the demand for prostitution. Research studies from selected countries were included in this review (220 studies were reviewed and 181 met the inclusion criteria). The findings presented highlight the characteristics and motivations of those who procure sex, the contexts in which they procure sex, and 'what works' in tackling the demand for prostitution. The report found that methodological difficulties plague research into clients of prostitutes. There are many gaps in the research and much of the evidence is weak or inconclusive, particularly with regard to 'what works' in reducing demand. It was also noted that prostitution is a policy domain for which the 'right' answer may not be determined solely by reference to the evidence. There are moral, political and other influences that need to be considered when tackling the demand for prostitution. Key findings
This chapter takes advantage of a range of recent studies in order to identify and correct some misconceptions about customers and explore their motives. There have been a number of insightful qualitative, interview-based studies of customers. At least two nationally representative surveys have included questions about prostitution use, the National Health and Social Life Survey and the General Social Survey. Internet bulletin boards and websites featuring the opinions and experiences of customers provide a potentially important, although as yet underused, source of information about customers. This article also reports original findings based on two samples of arrested customers gathered from men attending “john schools” in 1997–2000 and 2007.
Drawing on original empirical data with men who buy sex, this book takes a fresh look at the relationships clients have with female sex workers. The core questions that form the backbone of the research are not only the expected inquiry into ‘why men buy sex’, but also into the sociological and psychological processes that men encounter in order to enter an assumed ‘deviant’ sexual behaviour as part of their everyday lives.
In the last few decades, trafficking in humans for the purpose of sexual exploitation has exploded into a sophisticated industry that generates billions of dollars in profit every year yet devastates the lives of millions of innocent victims. Many of the trafficked victims are impoverished girls and young women from economically depressed countries who are forced to work as prostitutes under brutal conditions in a foreign country. To date, most of the scholarly, legislative, and law enforcement attention has focused on the "supply" side of the sex trafficking equation, namely the traffickers and the victims. This Comment focuses on the "demand" side of the problem, namely the male clients of the prostitutes. The Comment first explains how the male demand for commercial sexual services sustains and grows the sex trafficking industry, and then examines various demand-side educational programs and legislative approaches and assesses their impact on minimizing and eradicating the demand. The Comment concludes by suggesting a comprehensive, demand-oriented approach to fighting sex trafficking.
Log-linear models are developed for capture-recapture experiments, and their advantages and disadvantages discussed. Ways in which they can be extended, sometimes with only partial success, to open populations, subpopulations, trap dependence, and long chains of recapture periods are presented. The use of residual patterns, and analysis of subsets of data, to identify behavioural patterns and acceptable models is emphasised and illustrated with two examples.
The jaguar (Panthera onca) is the largest feline in the Americas and third largest world-wide, smaller in size only to the tiger (P. tigris) and lion (P. leo). Yet, in comparison, relatively few studies on jaguar population densities have been conducted and baseline data for management purposes are needed. Camera trapping and capture–recapture sampling methods were used to estimate the size of a jaguar population in the Pantanal’s open wet grassland habitat, an important area for the long-term survival of the species. This study is the first jaguar population estimate conducted in co-operation with a GPS-telemetry study providing an important opportunity for comparing different methods of density estimation. An accessible area within a 460 km2 privately-owned ranch was sampled with equal effort during the dry seasons of 2003 and 2004. Thirty-one and twenty-five individual jaguars were identified in 2003 and 2004, respectively. Estimates of jaguar abundance were generated by program CAPTURE. Density estimates were produced according to different methods used to calculate the effectively sampled areas which ranged from 274 to 568 km2. For 2003, the currently-used mean maximum distance moved (MMDM) method produced a density of 10.3 jaguars/100 km2, while GPS-telemetry-based calculations produced a mean density of 6.6 jaguars/100 km2. For 2004, the MMDM method produced an estimate of 11.7 jaguars/100 km2 while GPS-telemetry calculations produced a density of 6.7 jaguars/100 km2. Our results suggest that the widely-used MMDM method used to calculate effectively sampled areas is significantly under-reflecting maximum distances moved by jaguars and their range-use and, thereby, considerably inflating cat density estimates. This overestimation could place a population in a difficult situation by lengthening the time taken to initiate protection measures because of underestimating the risk to that population.