Sie sind auf Seite 1von 26

American Sociological Review

75(5) 679–704
Personal Characteristics, Ó American Sociological
Association 2010
Sexual Behaviors, and Male DOI: 10.1177/0003122410379581
http://asr.sagepub.com

Sex Work: A Quantitative


Approach

Trevon D. Logana

Abstract
Male sex workers serve multiple groups (i.e., gay-identified men, heterosexually-identified
men, and their own sexual partners), making them a unique source to test theories of gender,
masculinity, and sexuality. To date, most scholarship on this topic has been qualitative. I
assembled a dataset from the largest online male sex worker website to conduct the first quan-
titative analysis of male escorts in the United States. I find the geographic distribution of male
sex workers is more strongly correlated with the general population than with the gay male
population. In addition, I estimate the value of sexual behaviors and personal characteristics
in this market to test sociological theories of gender and masculinity. Consistent with hege-
monic masculinity, I find that male escorts who advertise masculine behavior charge higher
prices for their services, whereas escorts who advertise less masculine behavior charge signif-
icantly less, a differential on the order of 17 percent. Results show that race and sexual behav-
ior interactions exert a strong influence on prices charged by male sex workers, confirming
aspects of intersectionality theory.

Keywords
prostitution, masculinity, market forces, homosexuality, male sex work

Prostitution as a form of exchange (Simmel 2005; Parsons, Bimbi, and Halkitis 2001; Uy
[1907] 1971) or a location where cultural val- et al. 2004). However, many quantitative
ues and market logics intersect (Zelizer questions whose answers could complement
1994) has long interested social scientists, the qualitative approach remain unanswered.
but male sex work remains under-researched For example, we know little about the popu-
(Bimbi 2007; Pruitt 2005; Weitzer 2009). In lation size or geographic distribution of male
general, male sex workers are difficult to sex workers. Quantitatively analyzing the
conceptualize in current economic, social, market would increase our understanding of
and gender theories of prostitution because
all participants are the same gender (Bern- a
The Ohio State University and NBER
stein 2005, 2007; Edlund and Korn 2002;
Giusta, Di Tommaso, and Strom 2009; Mar- Corresponding Author:
Trevon D. Logan, The Ohio State University, 410
lowe 1997). Qualitative research on male sex Arps Hall, 1945 N. High Street, Columbus, OH
workers informs theories of sexuality, sexual 43210
behaviors, and sex work (Bimbi and Parsons E-mail: logan.155@osu.edu
680 American Sociological Review 75(5)

ways that commerce, sexuality, and mascu- therefore examine how race further shapes
linity interact. the construction of masculinity in this com-
There is a relative wealth of research munity (Collins 1999, 2000; Han 2006;
about male sex workers who work the street, Reid-Pharr 2001).
but little is known about male escorts who I break new ground in the study of male
occupy the highest position in the hierarchy sex work by taking an explicitly quantitative
of male prostitution (Cameron, Collins, and approach to the subject. The relationship
Thew 1999; Koken et al. 2005; Luckenbill between escort prices, personal characteris-
1986; Parsons, Koken, and Bimbi 2007; tics, and sexual behaviors provides a key win-
Pruitt 2005; Uy et al. 2004). Due to techno- dow into this relatively under-investigated
logical progress (the Internet) and the social activity (Bimbi 2007; Weitzer 2009).
increasing social acceptance of homosexual- The conceptual framework begins by consid-
ity (Loftus 2001; Scott 2003), prior work ering how this type of empirical analysis can
may be out of date.1 For example, recent shed light on social theories of sexuality and
qualitative scholarship finds that the demo- masculinity (Dowsett 1993). Principles of
graphic and social characteristics of male es- economic theory motivate the empirical
corts and the reasons for entry into approach, but I interpret the results in light
commercial sex work described in earlier of social theories of male sexuality. Because
postwar research do not apply today (Calhoun the male escort market in the United States
and Weaver 1996; Joffe and Dockrell 1995; does not use intermediaries—who could
Parsons et al. 2001; Pruitt 2005; Uy et al. potentially control the prices and earnings of
2004). Researchers also note male sex work- male escorts—we can see how male escorts
ers’ unique social and epidemiological posi- price their services conditional on their per-
tion because they serve numerous social sonal characteristics and sexual behaviors.
groups: gay-identified men, heterosexually- Values attached to these characteristics and
identified men, and their own noncommercial behaviors lie at the confluence of social value
sexual partners (Cohan et al. 2004). Male sex and market forces.
workers thus interact with groups of men who My approach allows me to answer hereto-
are unlikely to interact with each other, poten- fore unanswered questions about male sex
tially acting as a social and sexual conduit workers and, in turn, social theories of male
between various groups (Parker 2006). sexuality: Where are male escorts located?
One unique aspect of the study of gay Do they gather in high or low gay concentra-
male sex work is that all of the participants tion cities? How does this market value
are male. In contrast to male–female prostitu- physical characteristics (e.g., race, height,
tion, one cannot easily assign sexual posi- and weight) and sexual behaviors? For
tions or behaviors to participants based on instance, do male sex workers’ clients value
sex; this necessitates a discussion of the hegemonically masculine behaviors and
social value of sexual behaviors practiced appearance in a way that can be reconciled
and advertised by escorts in the market. In with hegemonic masculinity (Connell 1987,
particular, this article analyzes how and if 1995; Connell and Messerschmidt 2005)?
men who have sex with men reify and cri- Furthermore, what are the effects of interac-
tique hegemonic masculinity in the values tions between characteristics and sexual be-
of sexual behaviors; this is especially inter- haviors? Are men of particular races
esting because gay men are considered rewarded for downplaying or emphasizing
counter-hegemonic (Connell 1987, 1995). certain sexual behaviors, as intersectionality
Additionally, scholars note that racial sexual theory predicts (Collins 1999, 2000)?
stereotypes give rise to unique values of The present study makes several contribu-
practices among men of particular races. I tions to the literature. First, I expand the
Logan 681

scope of quantitative scholarship on sex work Kennedy 2006).6 The standard operating
by considering the male side of the market procedure is straightforward—escorts pay
and sex work that is neither street- nor a monthly fee to post their advertisements,
brothel-based (Arunachalam and Shah which include pictures, a physical description,
2008; Browne and Minichiello 1996; Weitzer their rate for services (quoted by the hour), as
2005, 2009). Second, I overcome a number well as contact information such as a telephone
of problems of sample selection by using number or e-mail address. Escorts have com-
the universe of men who advertise online plete control over the type and amount of
using the largest, most comprehensive, and information conveyed in their advertisements.
most geographically diverse website for Through the websites, clients contact escorts
male escort work in the United States.2 Third, directly and arrange for appointments at the
the quantitative approach is novel in that it is escort’s home (known as an ‘‘incall’’) or the
informed by and informs theoretical and client’s residence or hotel (an ‘‘outcall’’).
mostly qualitative work on the sociology of
masculinity and sexuality.
SOCIAL SCIENCE THEORIES
OF MALE SEX WORK
THE ONLINE MARKET FOR
COMMERCIAL MALE SEX IN Research on commercial sex work tradition-
ally concentrates on women and neglects
THE UNITED STATES the heterogeneous social structures that give
Popular media suggest that male sex workers rise to the diverse forms of male sex work
are a sizable portion of the sex worker popu- around the globe (Bernstein 2007; West
lation in the United States (Pompeo 2009; 1993). Surveys of male prostitution, both sci-
Steele and Kennedy 2006).3 Unlike their entific research (Aggleton 1999; Kaye 2001;
female counterparts, male sex workers usu- West 1993) and popular media (Itiel 1998;
ally work independently—there is virtually Pompeo 2009; Steele and Kennedy 2006),
no pimping nor male brothels in the male point out several geographic and cultural dis-
sex trade (Logan and Shah 2009; Pruitt tinctions in the practice and forms of male
2005; Weitzer 2005, 2009).4 The indepen- sex work that make it difficult to generalize
dent, owner-operator feature of the market the phenomena over space or time. This dif-
allows for greater mobility up and down the ficulty has hindered research—theories of
hierarchy of the male sex worker labor force. sexuality pay particular attention to sexual
In this hierarchy, male escorts are most es- minorities and marginalized sexualities
teemed: They do not walk the streets, they because these are central to understanding
take clients by appointment, and they are majority and minority sexualities and sexual
usually better paid than their street counter- identities (Epstein 2006; Sedgwick 1990; Stein
parts (Luckenbill 1986). While street sex 1989; Weinberg and Williams 1974). Inclu-
workers are paid by the piece rate, male es- sion of male prostitutes also alters the usual
corts are contract employees with greater theoretical tools of power and gender, allow-
control over the terms of their work and the ing for the exploration of dynamics within
services they provide.5 genders in a novel way (Marlowe 1997).
Male escorts used to congregate in Research on political economy among
‘‘escort bars’’ and place advertisements in sexual minorities deals largely with the com-
gay-related newspapers to solicit clients, moditization of gay desire (Cantu 2002;
but media reports suggest that the male D’Emilio 1997). Because commoditization
escort market now takes place online (Fried- is a market force with supply, demand, quan-
man 2003; Pompeo 2009; Steele and tities, and prices, I investigate how men in
682 American Sociological Review 75(5)

the male sex work market construct identities from the general male population (Humph-
and are influenced by social factors. This ries 1970). Despite this, a number of studies
can, in turn, tell us about values men place by demographers and economists looking at
on themselves and other men for commer- trends among the gay-identified population
cial and perhaps non-commercial sexual li- find differences on a range of outcomes,
aisons. Researchers have looked at these including earnings (Berg and Lien 2002; Car-
types of values qualitatively and quantita- penter 2004), partnership status (Carpenter
tively between genders (Almeling 2007; Ar- and Gates 2008; Jepsen and Jepsen 2002),
unachalam and Shah 2008; Koken, Bimbi, and general socioeconomic position (Black
and Parsons 2009), but little quantitative et al. 2007; Hewitt 1995).
work looks at differences within genders. While it can be difficult to identify all
Today and in the past, significant numbers sexual minorities in any data source (Berg
of male escorts and clients do not identify as and Lien 2006; Black et al. 2000, 2002;
homosexual (Bimbi 2007; Chauncey 1994; Cameron et al. 2009), researchers can now
Dorais 2005). Allen (1980) describes studies identify same-sex couples. Using these popu-
of male sex workers that find less than 10 lation trends, scholars note that the geo-
percent identify as homosexual. Since graphic distribution of male same-sex
Humphries’s (1970) controversial work social couples is different from that of the U.S. gen-
scientists have noted that men partaking in eral population (Black et al. 2000, 2002;
same-sex sexual behavior are unlikely to be Black et al. 2007). City amenities and the
found in surveys unless they choose to publicly ability to congregate and socialize with a crit-
reveal their sexual behaviors and desires ical mass relate to gay location patterns
(Black et al. 2000; Black, Sanders, and Taylor (Black et al. 2002; alternatively, Collins
2007; Cameron et al. 2009). The world of male [2004] emphasizes economic factors). What-
sex work is one of the few places where men ever the reason for these location differences,
who adopt homosexual identity and those there are unanswered questions about the
who refuse it are in intimate contact with one demography and geography of male sex
another; this offers us the opportunity to work; we know very little about the popula-
address questions about male sexual identity tion size, demographic characteristics, and
and homosexual desire. For example, what geographic distribution of male sex workers
roles and behaviors must escorts conform to in the United States.
in order to realize the largest economic gains Early studies of male sex work focused on
from sex work? The value of these roles can cities with large gay populations (McNamara
inform an analysis of the construction of mas- 1994), but more recent qualitative research
culinity at the crossroads of heterosexual and reveals that a significant portion of male es-
homosexual identities because men participat- corts’ clientele identifies heterosexually.7
ing in the market (both clients and escorts) Indeed, the ‘‘breastplate of righteousness’’
adopt disparate sexual identities. that Humphries (1970) saw in heterosexu-
ally identified men who took part in homo-
sexual behavior has recently resurfaced in
Economic and Demographic the public lexicon (Frankel 2007; MacDon-
Approaches ald 2007). In the market for male sex
work, such behavior is common. Male es-
Since the work of Hooker (1957), psycholo- corts note that a significant percentage of
gists have noted there is little except self- their clientele is heterosexually identified
identification to distinguish between homo- and many are married. Because these men
sexuals and heterosexuals. Men who partake are hidden from the most common analysis
in homosexual acts are not distinguishable of sexual minorities, how their presence in
Logan 683

the market influences market function and 1996; Reeser 2010; Schrock and Schwalbe
composition is unknown. 2009). For example, gay masculinities are sub-
Simple economic models of location, such ordinated and marginalized so that patriarchy
as Hotelling’s (1929), suggest that escorts can be reproduced through heterosexuality.
should locate close to their client base. Given Connell (1995) describes how hegemonic
that heterosexually-identified men may have masculinity is never influenced by non-
much to lose if their same-sex sexual behav- hegemonic elements: elements of non-
ior is exposed, male escorts might be more heterosexuality are seen as contradictions or
likely to locate in places where there are weakness (Demetriou 2001).
fewer opportunities for men interested in sex- Scholars note the limits of this concep-
ual encounters with other men to meet one tual binary between hegemonic and non-
another. Self-identified heterosexual men hegemonic masculinity (Anderson 2002; Con-
are unlikely to frequent gay bars, coffee- nell and Messerschmidt 2005; Demetriou
houses, or community groups where they 2001; Donaldson 1993; Dowsett 1993; Reeser
might encounter gay men for socialization 2010). Demetriou and Reeser, among others,
or sex. Male escort location should thus dif- suggest that rather than binary, hegemonic
fer from that of the gay-identified population. masculinity should be viewed as a hybrid,
Conversely, researchers note that gay com- made up of practices and elements of hetero-
munities do not attach the same level of sexual and homosexual masculinities, giving
stigma to sex work as heterosexuals do (Ko- hegemonic masculinity the ability to change
ken et al. 2009; Sadownick 1996). If gay over time to meet historical circumstances
communities are seen as safer havens for with a different set of practices. In this con-
sex workers, or if few customers are hetero- ception, the practices of gay men, who are
sexually identified, we would expect male non-hegemonic, not only reinforce the patriar-
sex workers’ geographic distribution to chal goal of hegemonic masculinity, but they
closely mirror the openly gay population’s. help define the hegemonic ideal itself.9 Deme-
triou (2001), for example, notes the recent
construction of the ‘‘metrosexual’’ as one
example of gay masculinity influencing the
SOCIOLOGICAL APPROACHES construction of the hegemonic ideal.
Hegemonic Masculinity While a binary approach views gay men
in relation to the hegemonic ideal, a hybrid
Hegemonic masculinity is defined as ‘‘the approach opens the possibility of analyzing
configuration of gender practice which how gay men define, subordinate, and mar-
embodies the currently accepted answer to ginalize masculinities among themselves.
the problem of legitimacy of patriarchy’’ This within-subgroup construction might
(Connell 1995:77). Hegemonic masculinity influence how hegemonic masculinity itself
is about relations between and within gen- is defined. Donaldson (1993) and Connell
ders.8 Hegemonically masculine practices (1992) note that gay men reify hegemonic
ensure the dominant position of men over norms: modern gay practices celebrate and
women, and of particular men over other exemplify hegemonic ideals such as
men. These practices can take a number of bodybuilding and physical strength. This
forms; research usually stresses social traits reification of masculine norms can create
such as drive, ambition, self-reliance, and a situation where some gay masculinities
aggressiveness, which legitimate the power are themselves subordinate to others. That
of men over women. Within genders, there is, among gay men themselves, there may
is the subordination of certain masculinities be further refinement of the gay masculine
and the marginalization of others (Bird norm along hegemonic lines. Donaldson
684 American Sociological Review 75(5)

(1993:649) raises the intriguing point that ‘‘it Williams 1974).10 In this market, is men’s
is not ‘gayness’ that is attractive to homosex- lust consistent with hegemonic norms?
ual men, but ‘maleness.’ A man is lusted Sexual desire for the hegemonic ideal
after not because he is homosexual but could influence how hegemonic masculinity
because he’s a man. How counter-hegemonic itself is constructed and reinforced among
can this be?’’ gay men. If this is the case, there may be
Scholars of masculinity have asserted that limits to the binary view of hegemonic
gay men critique hegemonic ideals through masculinity.
their counter-hegemony (Connell 1992; This type of hybrid hegemonic masculin-
Reeser 2010), but it could also be the case ity might have additional implications for
that gay men overtly reify hegemonic ideals gay male body image and gay men’s use
sexually. To the extent that this occurs, gay of the body in constructing masculinity.
masculinities may be aligned with the hege- Sadownick (1996) sees the gay liberation
monic masculinity that marginalizes them. movement as a time when gay men began,
The question is the degree to which homo- en masse, to idealize hypermasculinity,
sexual men are complicit in hegemonic muscles, and a hirsute body, turning on its
masculine norms. In Demetriou’s (2001) lan- head the ‘‘flight from masculinity’’ that
guage, to what degree do gay masculinities Hacker (1957) observed in earlier genera-
contain significant elements of hegemonic tions of gay men. The physical ideal is typ-
masculinities that legitimate patriarchy and ified by a muscular physique and other
may, in turn, influence hegemonic masculin- markers of hegemonic masculinity such as
ity itself ? In an explicitly sexual arena, heg- height, body hair, whiteness, youth, and
emonic masculinity would extend to physical middle-class socioeconomic status (Atkins
appearance (e.g., muscularity, body size, 1998; Green 2008b). This turn of events
body hair, and height) and sexual behaviors has molded the gay body into a political rep-
(e.g., sexual dominance, sexual aggressive- resentation of masculinity. In part, this sub-
ness, and penetrative sexual position). To verts norms that question the compatibility
the extent that homosexual men conform to of masculinity and homosexuality, but it
and reify hegemonic masculine norms, the also reinforces a quasi-hegemonic mascu-
value of masculine traits and practices should line ideal (Atkins 1998; Connell 1992).
have a direct effect on a given escort’s desir- Compared with lesbians and heterosexuals,
ability and value. While such ‘‘manhood gay men show stronger tendencies to prefer
acts’’ usually elicit deference from other particular body types, and this can lead to
men and reinforce hegemonic masculinity poor psychological and health outcomes
(Bird 1996; Schrock and Schwalbe 2009), for gay men who do not conform to gay
in an explicitly homosexual arena they may standards of beauty (Atkins 1998; Beren et
also elicit sexual desire and objectification. al. 1996; Carpenter 2003; Green 2008b;
The function of male sex work in gay com- Herzog et al. 1991). In fact, attempts by
munities may heighten such effects. In a mar- some gay subcultures to subvert these
ket for sex work, clients are explicitly seeking beauty standards have been critiqued as
sexual contact. Clients may choose escorts being agents themselves of hegemonically
unlike the men they interact with socially, masculine agendas (Hennen 2005).
but whom they do desire sexually. This may In many ways, rejection of large men and
increase the value of certain masculine char- thin men may be seen as the rejection (subju-
acteristics insofar as the hegemonic mascu- gation) of feminizing features. For example,
line archetype may be a driving force in excess weight in a man visually minimizes
purely sexual desire (Cameron et al. 1999; the relative size of male genitals and produ-
Green 2008a; Pruitt 2005; Weinberg and ces larger (and, importantly, non-muscular)
Logan 685

male breasts; thin men may appear slight, Intersectionality


waifish, and physically weak. These appear-
ances emphasize feminizing traits that are Just as the theory of hegemonic masculinity
actively discouraged in mainstream gay cul- has been critiqued for not considering how
ture (Atkins 1998; Hennen 2005). In the mar- gay men can conform to and inform hege-
ket for male sex work, we expect clients to monic masculinity, there is also a burgeoning
prize physical characteristics that mark hege- literature that looks at racial variation in
monic masculinity, such as muscular phy- social value among gay men (Green 2008a;
siques, body hair, and height. We also Han 2006; Nagel 2000; Robinson 2007). As
expect feminizing features, such as excess intersectionality theory suggests, the interac-
weight and thinness, to be penalized. tion of these social categories is neither
The theory of hegemonic masculinity and cumulative nor additive but rather indepen-
the closely related literature on the body in dent (Collins 1999, 2000; Reeser 2010).
gay communities suggest that clients of The intersection of hegemonic masculinity
male sex workers are likely to prize ‘‘mascu- with racial sexual stereotypes can create mul-
line’’ personas and body type. There are sev- tiple forms of sexual objectification for par-
eral reasons for this. First, numerous scholars ticular groups of gay men. For example, the
assert that gay men’s relationships with value of a top (the penetrative partner) is
effeminate behavior are complex—while cel- not uniform across all tops, and the value
ebrated in many aspects of gay culture (e.g., of a white top is not simply the addition of
camp, drag shows, and diva worship), effem- the value of whiteness and ‘‘topness,’’ but
inate behavior is particularly stigmatized in an independent effect for men in that partic-
sexual relationships and as an object of lust ular category, who in this instance embody
(Clarkson 2006; Nardi 2000; Ward 2000). the highest position in the racial and sexual
Second, some scholars note how the gay behavior hierarchies among gay men. Mar-
community has commoditized the ‘‘authen- kets for sex may reify these sexual stereo-
tic’’ masculinity of self-identified heterosex- types (what Cameron and colleagues [1999]
ual men who engage in sex with men (Ward call ‘‘ethnico-sexual stereotypes’’) in explic-
2008). This has given rise to the distinction itly monetary terms.
between ‘‘masculine’’ and ‘‘effeminate’’ Baldwin (1985) notes that the American
gay men in gay communities (Clarkson ideal of sexuality is rooted in the American
2006; Connell 1992; Pascoe 2007). Construc- ideal of masculinity, which he argues neces-
tion of dual masculinities among gay men, sitates an inherently racial dimension. Histor-
and distinctions between the two, are used ically, white men were to protect white
to legitimate the power of masculine gay women from black sexuality, and this sup-
men over effeminate gay men, a reproduction posed threat legitimated white men’s social
of patriarchy (Clarkson 2006). We thus control of white women (and whites social
expect that men who are interacting primar- control over blacks). For homosexual white
ily for sexual purposes likely place a premium men, black men’s sexuality may become an
on masculine practices (e.g., penetrative sex- object of desire because they are perceived to
ual position [‘‘topping’’], aggressive sexual be sexually dominant and unrestrained—
behavior, and muscular physique) and penal- although still under the social control of
ize feminine practices (e.g., receptive sexual whites due to their race—turning the hege-
position [‘‘bottoming’’], submissiveness, monic ideal on its head (Baldwin 1985;
large body size, and thinness) to the degree Reeser 2010). Robinson (2008), McBride
that they conform to hegemonic masculinity (2005), Reid-Pharr (2001), Green (2008a),
and to the construction of masculinity among and others note how racial stereotypes inter-
gay men. act with notions of masculinity to produce
686 American Sociological Review 75(5)

a desire for hypermasculine black men, par- with stereotypical images: Asians as exotic,
ticularly among white gay men. docile, loyal partners; Hispanics as passion-
The stereotype of the sexually dominant ate, fiery lovers; and Blacks as ‘well-
black man, rather than being an agent of endowed,’ forbidden partners.’’ If the market
fear, can lead to a celebration of his hyper- for male sex work mirrors the gay
sexual behavior, appearance, and conduct. community at large, we would expect black
In this theory, the general level of social men who advertise themselves as tops and
interaction between black and white gay Asian men who advertise themselves as bot-
men is relatively low and occurs chiefly toms to command high prices, reflecting the
over sex. Black men who demonstrate hyper- value of conforming to racial sexual
masculine and sexually aggressive behavior stereotypes.
are offered entry into white gay spaces, but
this entry is limited to sexual liaisons.
McBride (2005), for example, notes the limited DATA FROM THE GAY MALE
range in which black men interact with whites
in gay pornography, where the vast majority
SEX MARKET
of black performers are tops and adopt an anti- I assembled a dataset on 1,932 men from the
social persona. Men who defy racial sexual largest online website for male sex workers
stereotypes could face markedly lower values in the United States.11 This source has sev-
and become, in this particular instance, eral advantages. First, it allows me to collect
counter-hegemonic. Robinson (2008) finds information on escort attributes, prices, and
that white gay men largely ignore and devalue information free from the selection problems
black men who do not conform to the stereo- one would encounter in a field survey of es-
type of the hypermasculine black male, sug- corts; the data include every unique adver-
gesting that the penalties for nonconformity tisement on the site. Second, I can identify
may be particularly harsh. escorts’ home locations, which allows for
The reverse is true for Asians, whose pas- accurate geographic counts. Third, the escort
sivity and docility are celebrated. Robinson characteristics I use are entered by escorts
(2007, 2008), Phua and Kaufman (2003), from dropdown menus. This is particularly
and Han (2006) describe the persistent advantageous for features one would like to
stereotype that Asian men should be passive, control for in pricing models (e.g., body
docile bottoms (the receptive partner). As type or hair color) where free-form responses
with black men, this racial sexual stereotype may be difficult to code consistently or
allows the larger gay community to limit where non-response could bias estimates.
Asian men’s socially acceptable sexual Fourth, the website is free for viewing by
expressions. In this case, the counter- all. This ensures that the information pro-
hegemonic activity is for Asian men to vided is for a large, general client base and
appear sexually dominant or aggressive. not manipulated to please paying members
These authors also note that Hispanics are of the website.
celebrated as passionate, virile lovers who It is important to establish that the data
are usually sexually dominant, although not source provides sufficient coverage of the
exclusively so, making it difficult to derive online escort market. Table 1 compares es-
quantitative predictions for this group. corts on the chosen site with escorts on two
Given these racial sexual stereotypes, it is prominent competitors for a random sample
necessary to consider how the values of par- of cities. As described earlier, I can uniquely
ticular sexual behaviors differ by race. Phua identify the men in the chosen website and
and Kaufman (2003:992) find that dating their home locations. This is impossible
‘‘preferences for minorities often are tinted with the competitors, and potential double
Logan 687

Table 1. Comparison of Male Escort Websites

Number of Escorts:
City Data Rentboy Male Escort Review Data / Rentboy Rentboy / Data

Albany, NY 5 0 3 NA NA
Austin, TX 26 3 15 2/3 2 / 26
Buffalo, NY 5 0 0 NA NA
Charlotte, NC 19 3 4 2/3 2 / 19
Columbus, OH 30 3 13 3/3 3 / 30
Denver, CO 41 5 19 5/5 5 / 41
Detroit, MI 73 10 14 9 / 10 9 / 73
Indianapolis, IN 19 0 5 NA NA
Kansas City, MO 9 1 7 0/1 0/9
Minneapolis, MN 33 2 15 2/2 2 / 33
Nashville, TN 14 1 8 1/1 1 / 14
Oklahoma City, OK 3 1 0 1/1 1/3
Portland, OR 15 1 12 1/1 1 / 15
Sacramento, CA 17 7 5 5/7 5 / 17
St. Louis, MO 18 3 6 2/3 2 / 18
Seattle, WA 33 14 23 11 / 14 11 / 33
Rochester, NY 4 0 0 NA NA
Tampa, FL 47 15 22 11 / 15 11 / 47
Total 411 69 171 55 / 69 55 / 378

Note: Rentboy and Male Escort Review accessed on January 25, 2009. Counts of number of escorts by
home base (Data) or number of advertisements (Rentboy and Male Escort Review). The last two columns
show the number of escorts listed in the data source that can be identified on Rentboy and the number of
escorts listed on Rentboy who can be identified in the data source, respectively. I selected cities at
random from the 50 cities listed in Black and colleagues (2007).

counting by city on competitors would bias and outcalls. Escorts can write about their
upward the number of escorts listed on com- services and quality in an advertisement’s
peting websites. Even with this bias, the cho- text.12 One additional advantage of the data
sen website’s coverage of the male escort is that claims escorts make about characteris-
market is greater than that of its two compet- tics can be confirmed with pictures posted in
itors. The last two columns of Table 1 show the advertisements.13 In the analysis, I use
the number of escorts on the chosen site that outcall price as the price of escort services,
I could locate on the most prominent competi- but results do not change when using incall
tor’s site and vice versa. I found the majority of price.
escorts who advertise on competing sites on the Table 2 shows summary statistics for es-
chosen site, but only a small fraction of escorts corts in the data. On average, escorts charge
in my source could be identified on the com- more than $200 an hour for an outcall,
petitor’s site. consistent with media estimates of escort
Figure 1 shows a diagram of an escort services (Pompeo 2009; Steele and Kennedy
advertisement. Escorts list their age, height, 2006). As one would expect, escorts are rel-
weight, race, hair color, eye color, body atively young and fit; on average, they are
type, and body-hair type. Advertisements 28 years old, 5’10’’, and 165 pounds. Ac-
give clients contact information, the pre- cording to the National Center for Health
ferred mode of contact (i.e., phone or Statistics, the average man age 20 to 74 years
e-mail), escorts’ availability to travel nation- in the United States is 5’9.5’’ and 190
ally and internationally, and prices for incalls pounds. Escorts are racially diverse; 54
688 American Sociological Review 75(5)

Figure 1. Diagram of Online Escort Advertisement


Note: The figure is an example created by the author for informational purposes and is not an actual
advertisement.

percent are white, 22 percent are black, 14 Looking at sexual behaviors, 16 percent of
percent are Hispanic, 8 percent are multira- escorts are tops, 6 percent are bottoms, and
cial, and 1 percent are Asian. 21 percent list themselves as versatile.14 In
Looking at physical traits, escorts are addition, 19 percent of escorts advertise that
likely to have black (36 percent) or brown they exclusively practice safer sex. Overall,
(46 percent) hair (less than 15 percent are summary statistics for men in the data are
blond). More than half of all escorts have similar to descriptive statistics noted by Ca-
brown eyes (55 percent), although significant meron and colleagues (1999) for male escorts
fractions have blue (18 percent) and hazel in British newspapers in the 1990s and
(14 percent) eyes. Nearly half of all escorts Pruitt’s (2005) more recent sample of male
are smooth (49 percent), 17 percent shave escorts who advertise on the Internet.
their body hair, but more than a third are
hairy or moderately hairy (34 percent).
Very few escorts are overweight (1 percent),
and relatively few are thin (8 percent); the EMPIRICAL STRATEGY
majority of escorts claim to have athletic Previous quantitative work analyzing male
(48 percent) or muscular builds (30 percent). escorts has not examined prices of male
Table 2. Summary Statistics for the Escort Sample

Variables Obs Mean SD Physical Traits Obs Mean SD Behaviors Obs Mean SD

Price 1,476 216.88 64.46 Hair Color Top 1,932 .16 .37
Log of Price 1,476 5.34 .29 Black 1,932 .37 .48 Bottom 1,932 .06 .24
Weight 1,932 167.11 24.54 Blonde 1,932 .13 .34 Versatile 1,932 .21 .40
Height 1,932 70.43 2.69 Brown 1,932 .46 .50 Safe 1,932 .19 .39
BMI 1,932 23.64 2.89 Gray 1,932 .02 .13
Age 1,932 28.20 6.93 Auburn/Red 1,932 .01 .11
Asian 1,932 .01 .12 Other 1,932 .01 .10
Black 1,932 .22 .41 Eye Color
Hispanic 1,932 .14 .35 Black 1,932 .02 .14
Multiracial 1,932 .08 .28 Blue 1,932 .18 .39
Other 1,932 .01 .10 Brown 1,932 .55 .50
White 1,932 .54 .50 Green 1,932 .11 .31
Hazel 1,932 .14 .35
Body Hair
Hairy 1,932 .04 .20
Moderately Hairy 1,932 .30 .46
Shaved 1,932 .17 .38
Smooth 1,932 .49 .50
Body Build
Athletic/Swimmer’s Build 1,932 .48 .50
Average 1,932 .13 .34
A Few Extra Pounds 1,932 .01 .08
Muscular 1,932 .30 .46
Thin/Lean 1,932 .08 .27

Note: Price is the outcall price posted by an escort in his advertisement. See the Data Appendix for variable definitions.

689
690 American Sociological Review 75(5)

escort services (Cameron et al. 1999; Pruitt Before turning to the results, it is important
2005). I use male escort services’ prices in to establish the validity of the prices used. Er-
a hedonic regression, a technique developed rors or selection could bias the results. First,
by Court (1939), Griliches (1961), and Rosen note that the correlation of advertised prices
(1974). The basic technique regresses the with actual transaction prices is quite high
price of a particular good or service on its (r = .89) (see Logan and Shah 2009). Of
characteristics. This type of regression is greater concern is selection. Although the
widely used in economics, and it is particu- vast majority of escorts post prices (more
larly useful for goods that are inherently than 85 percent), not all do so. To address
unique or bundled. For example, to perform this potential selection issue, I estimated
a hedonic regression of computers, one probit models where the dependent variable
would assemble data on computers in the is whether an escort posted a price. Table 3 re-
market and their characteristics, such as pro- ports the results. There is no relationship
cessor speed, hard drive size, memory, between the information in an escort’s adver-
graphics card, and monitor size. A regression tisement (e.g., race or sexual position) and the
of computer prices on characteristics would decision to post prices. Given these results, I
then give coefficients for each of the charac- am reasonably confident that the price mea-
teristics, which would tell us how much com- sure is an accurate measure of the prices actu-
puter prices increase or decrease, on average, ally paid in transactions.
for an increase or decrease in hard drive size
or monitor quality, respectively.
EMPIRICAL RESULTS
Estimated coefficients from hedonic re-
gressions are commonly interpreted as Geographic Distribution of Male
implicit prices because they reflect the Sex Workers
change in price one could expect, on average,
for a change in that particular characteris- Table 4 shows the geographic distribution of
tic.15 It is common to refer to positive coeffi- male escorts who advertise online. I count
cients as a premium and to negative the actual number of escorts by the home
coefficients as a penalty. Using this insight, location given in their advertisements. To
I regress the price of individual escort i’s my knowledge, this is the first large-scale
services (P ) on the escort’s characteristics quantitative evidence on the geographic loca-
(Z ), his sexual behaviors (S ), and identifiers tion of male escorts in the United States. Size
for his location (X ). of the escort population varies considerably;
there are more than 300 escorts in only one
X
J 1
city, New York, which has long been de-
lnðPi Þ ¼ a þ bj Zij
j
picted in the media as the largest male escort
ð1Þ market (Pompeo 2009). Atlanta, Los An-
X
K 1
þ gk Sik þ dX þ ei geles, Miami, and San Francisco each have
k more than 100 escorts, but most cities have
considerably fewer.
The regression gives the implicit prices for For comparison, I list the same randomly
each characteristic (each b) and each sexual selected cities used in Table 1 and show
behavior (each g). I control for an escort’s them in Table 4 along with larger cities. I
location to purge the estimates of geographic also include the rank and size of the popula-
differences in prices.16 Because I take the tions of each Metropolitan Statistical Area
log of price in this specification, the esti- (MSA), as well as the Gay Concentration
mates are percent price changes for each Index (GCI) developed by Black and col-
characteristic. leagues (2007) to compare the location of
Logan 691

Table 3. Probit Regression Estimates of Advertisement Information and Selection into


Posting Prices

(1) (2) (3) (4)

Black .0136 .0500


(.066) (.066)
Hispanic .0573 .0872
(.056) (.061)
Multicultural .0532 .0832
(.057) (.079)
Asian –.0501 .0372
(.120) (.080)
White .041 .0830
(.068) (.075)
Top .0719 .0642 .0746
(.073) (.073) (.083)
Bottom .0968 .0987 .0966
(.089) (.088) (.099)
Versatile .0228 .0105 .0185
(.022) (.018) (.023)
Versatile Top –.0524 –.0585 –.0496
(.070) (.061) (.069)
Versatile Bottom –.208 –.120 –.209
(.190) (.120) (.219)
Safer Sex .0289 .0266 .0279
(.023) (.021) (.023)
No Attitude .0355 .0408 .0333
(.023) (.054) (.023)
Safe Top –.118 –.111 –.119
(.075) (.086) (.076)
Safe Bottom –.102 –.124 –.101
(.120) (.140) (.116)
Additional Controlsa X X
Observations 1,932 1,932 1,932 1,932
Pseudo R-squared .0034 .0113 .0251 .0286

Note: Robust standard errors in parentheses are clustered at the state level. Each column reports results
of a probit regression where the dependent variable is: Does Escort Post Prices? (mean of dependent
variable = .85, standard error of dependent variable = .36). The coefficients reported are the marginal
effects of the probit regressions. For dichotomous variables (such as race), the effects calculate the
change in probability from moving from 0 to 1. Each column includes controls for state. See Data
Appendix for variable definitions.
a
Columns 3 and 4 include controls for age, height, whether the escort was available all day, body type,
body hair, review allowed, weight, eye color, and whether the escort preferred phone contact.
*p \ .05; **p \ .01 (two-tailed tests).

escorts with gay male location patterns. The (Carpenter and Gates 2008), it is likely that
GCI counts the proportion of male same-sex the index understates the concentration of
households in a given MSA and divides it gay men in specific urban areas. While one
by the national average. Cities with GCI val- might argue that single self-identified gay
ues greater than 1 have larger shares of male men locate in substantially different MSAs
same-sex households in their MSA than the than do partnered gay men, this is unlikely
national average. Given that partnership rates because MSAs encompass an entire metropol-
are lower for gay men than for heterosexuals itan area.
692 American Sociological Review 75(5)

Table 4. Geographic Distribution of Escorts, Selected Cities

MSA: Gay Concentration:


City Rank Population Rank Index Number of Escorts:

New York City, NY 1 18,815,988 13 1.49 309


Los Angeles, CA 2 12,875,587 6 2.11 126
Chicago, IL 3 9,524,673 18 1.31 93
Miami, FL 7 5,413,212 14 1.46 119
Washington, DC 8 5,306,565 2 2.68 99
Atlanta, GA 9 5,278,904 7 1.96 108
Boston, MA 10 4,482,857 9 1.67 53
Detroit, MI 11 4,467,592 42 .60 50
San Francisco, CA 12 4,203,898 1 4.95 124
Seattle, WA 15 3,309,347 5 2.21 33
Minneapolis, MN 16 3,208,212 10 1.61 33
St. Louis, MO 18 2,808,611 37 .69 18
Tampa, FL 19 2,723,949 24 1.05 47
Denver, CO 21 2,464,866 12 1.53 41
Portland, OR 23 2,175,113 15 1.45 15
Sacramento, CA 26 2,091,120 8 1.71 17
Kansas City, MO 29 1,985,429 25 1.04 9
Columbus, OH 32 1,754,337 27 .99 30
Indianapolis, IN 33 1,695,037 19 1.12 19
Charlotte, NC 35 1,651,568 45 .49 19
Austin, TX 37 1,598,161 3 2.44 26
Nashville, TN 39 1,521,437 32 .85 14
Oklahoma City, OK 44 1,192,989 34 .83 3
Buffalo, NY 46 1,128,183 49 .35 5
Rochester, NY 50 1,030,495 29 .89 4
Albany, NY 57 853,358 31 .85 5
Correlation of Number of Escorts with Gay Concentration Index: .39
Correlation of Per Capita Escorts with Gay Concentration Index: .69
Correlation of Number of Escorts with MSA Population: .92

Note: Counts of number of unique escort advertisements by home location. Gay concentration is the
fraction of the MSA identified as same-sex male partners in the 1990 Census divided by the national
average (for further details, see Black et al. 2007). MSA population counts from the Census Bureau as of
July 1, 2007. I selected cities with MSA rank greater than 12 at random from the 50 cities listed in Black
and colleagues (2007).

In terms of location patterns there is a strik- capita escorts with the GCI (r = .69) is weaker
ing trend: the number of gay escorts closely than the correlation of escorts with MSA.
follows the size of an MSA, not gay location This result is consistent with the claim that
patterns. For example, Detroit is the 11th larg- the market for male sex work is national in
est MSA in the United States, and its gay con- scope and not driven exclusively by gay-
centration is 42nd, but there are 51 percent identified participants. If escort services were
more escorts in Detroit than in Seattle, a city primarily demanded by self-identified gay
with the fifth highest GCI. Chicago and St. men, we would expect the geographic distribu-
Louis display a similar pattern. Indeed, the tion of male escorts to mirror the geographic
overall correlation of the number of escorts distribution of self-identified gay men. That is,
with MSA population is quite strong (r = male escorts would locate in places with a higher
.92), but it is much weaker with the GCI concentration of potential customers (Hotelling
(r = .39). Furthermore, the correlation of per 1929). Results in Table 4 imply that male
Logan 693

Table 5. Implicit Prices of Physical Characteristics in the Male Escort Market

I. Age and Body Type II. Race III. Hair Color


Age 2.00975** Black .0433 Blonde Hair .000365
(.00119) (.0658) (.0248)
Height 2.00224 Hispanic .045 Brown Hair 2.0186
(.00331) (.0668) (.0167)
Weight 2.00114** Multiracial .0882 Gray Hair 2.0475
(.000394) (.0687) (.292)
Other 2.0337 Other Color Hair .118
(.101) (.0852)
White .0277 Red/Auburn Hair .0857
(.0646) (.0719)

Observations 1,476 1,476 1,476


R-squared .044 .004 .004

IV. Eye Color V. Body Build VI. Body Hair


Blue Eyes 2.072 Average Build 2.147** Moderately Hairy 2.163
(.0556) (.0230) (.292)
Brown Eyes 2.0682 A Few Extra Pounds 2.300** Shaved 2.113
(.0535) (.0869) (.290)
Green Eyes 2.050 Muscular Build .0319 Smooth 2.110
(.0573) (.0173) (.290)
Hazel Eyes 2.0679 Thin/Lean Build 2.0479
(.0561) (.0288)
Observations 1,476 1,476 1,476
R-squared .002 .045 .019

Note: Robust standard errors are listed under coefficients in parentheses. Each model is a separate regression
in which the log of escort prices is the dependent variable. Each includes controls for city location and an
intercept. For Model II, the omitted race category is Asian. For Model III, the omitted hair color is black. For
Model IV, the omitted eye color is black. For Model V, the omitted body build is athletic/swimmer’s build.
For Model VI, the omitted body hair is hairy. See the Data Appendix for variable definitions.
*p \ .05; **p \ .01 (two-tailed tests).

escorts tend to concentrate in cities with sub- escort services from hedonic regressions of
stantial populations, not just cities with substan- escort prices on physical characteristics.
tial gay populations. This result holds even Given that clients in commercial sex markets
when considering mid-sized and smaller generally tend to be older men (Friedman
cities—it is not driven by cities that have large 2003), we would expect clients to prize youth
populations and large gay populations, such as and beauty, consistent with female sex work
Los Angeles. Overall, the evidence is consistent (Bernstein 2007). The theory of hegemonic
with the hypothesis that male escorts serve masculinity and related literature on the gay
a market that includes a substantial number of body, however, predict that hegemonically
heterosexually-identified men. masculine physical traits would be prized in
the market. In describing the results, I
Physical Characteristics and Male emphasize the percentage differences, but
Escort Prices to increase the exposition, I also give the
dollar value of the differentials based on an
Table 5 shows estimates of the value of phys- average price of $200 per session. It is impor-
ical characteristics on the pricing of male tant to emphasize that these differentials are
694 American Sociological Review 75(5)

cumulative. For example, a 10 percent ($20) that stipulate there is a hegemonic ideal: no
price differential per session could lead to race/hair color/eye color/body hair combina-
earnings differences in excess of $5,000 per tion is more valuable than any other. Other
year.17 than weight and body build, it appears that
There is a penalty for age, with each addi- most personal characteristics are not very
tional year of age costing an escort 1 percent important in the male escort market.
($2) of his price. Similarly, there is a penalty
for weight, with each additional 10 pounds re-
sulting in over a 1.5 percent ($3) price Sexual Behaviors and Male
decrease.18 Body build, which is closely Escort Prices
related to weight, also appears to be important.
Men with average body type experience a price Table 6 shows estimates of the value of
penalty that exceeds 15 percent ($30), while advertised sexual behaviors on male escort
men who have excess weight experience prices. An important implication of hege-
a price penalty of more than 30 percent monic masculinity is the idea that dominant
($60).19 There is a price penalty for thinness, sexual behaviors would be rewarded in the
although it is not as large as the penalty for male escort market. Consistent with hege-
those who are average or overweight, being monic masculinity, the premium to being
on the order of 5 percent ($10, p \ .1). This a top is large, over 9 percent ($18), and the
is consistent with work that finds a large social penalty for being a bottom is substantial—in
penalty for additional weight among gay men some specifications (Model V), it is nearly as
(Carpenter 2003), theoretical work that de- large as the premium for being a top, on the
scribes codes of body image in gay communi- order of 29 percent ($18). The price differ-
ties (Atkins 1998), and literature on the body ential for men who are tops versus men
that suggests significant penalties for weight who are bottoms—the top/bottom differen-
among gay men, as both excess weight and tial—is substantial, ranging from 14.1 per-
thinness have feminizing features. Men with cent ($28, Model I) to 17.6 percent ($35,
a muscular build, however, enjoy a price pre- Model V).20
mium of around 4 percent ($8, p \ .1). Indeed, Premiums for these sexual behaviors are
only men who have muscular builds enjoy inconsistent with the economic concept of
a price premium relative to ‘‘athletic/swim- compensating differentials, where riskier oc-
mer’s build,’’ the reference category. Because cupations (in this market, sexual behaviors)
muscularity is a physical signal of maleness are compensated with higher wages. Accord-
and dominance, and it can be considered ing to research in public health, the relative
a proxy for strength and virility, the premium risk of contracting HIV for receptive versus
attached to muscularity in this market is con- penetrative anal sex is 7.69 (Varghese et al.
sistent with hegemonic masculinity. 2002). This implies that correlations we
Surprisingly, race does not seem to play observe are in spite of the fact that receptive
a role in escort prices. No race commands sexual activity carries greater disease risk
higher prices in the market than any other. than does the penetrative sex act. Compen-
While some escorts of color claim they are sating differentials would predict that bottom
paid less than their white counterparts (Pom- escorts should be compensated for taking on
peo 2009), these data do not support that this increase in disease risk, but I find exactly
claim. The same holds for hair color, eye the opposite. In studies of female sex work,
color, body hair, and height. Interestingly, the compensating differential is substantial
body hair and height, masculine traits, do (Gertler, Shah, and Bertozzi 2005). The
not come with premiums in this market. In empirical estimates also find a positive corre-
general, these results go against theories lation between advertised safety and escort
Logan 695

Table 6. Implicit Prices of Sexual Behaviors in the Male Escort Market

I II III IV V

Top .0803** .0968** .0938** .0936** .0890**


(.0202) (.0224) (.0224) (.0224) (.0247)
Bottom 2.0596* 2.0773* 2.0788* 2.0798* 2.0865*
(.0302) (.0355) (.0355) (.0355) (.0385)
Versatile .00536 .0142 .0089 .0074 .00806
(.0186) (.0214) (.0215) (.0216) (.0216)
Versatile Top 2.0881 2.0833 2.0816 2.0824
(.0515) (.0515) (.0515) (.0516)
Versatile Bottom .0589 .0654 .0675 .0674
(.0673) (.0672) (.0673) (.0673)
Safer Sex .0510** .0488* .0420
(.0192) (.0195) (.0227)
No Attitude .0147 .0150
(.0224) (.0224)
Safe Top .0221
(.0490)
Safe Bottom .0352
(.0771)
Observations 1,476 1,476 1,476 1,476 1,476
R-squared .015 .017 .022 .022 .023

Note: Robust standard errors in parentheses. Dependent variable is the log of an escort’s price in all
regressions. Each regression includes controls for age, city location, and an intercept. See Data
Appendix for variable definitions.
*p \ .05; **p \ .01 (two-tailed tests).

prices, greater than 5 percent ($10). As fur- groups of men who have sex with men. As
ther evidence against an idea of compensat- described earlier, gay communities prize
ing differentials for male sex work, I find behavior that can be described as hegemoni-
no premiums for particular types of safe cally masculine, and this extends to sexual
sex—men who advertise as ‘‘safe tops’’ or acts themselves (Clarkson 2006).
‘‘safe bottoms’’ do not enjoy a distinct pre- Alternative explanations for the top pre-
mium, although disease transmission proba- mium bear mentioning. One possibility is
bilities would suggest that they should.21 that the premium may derive from the biology
The substantial premium to tops and pen- of being a top. If ‘‘topping’’ requires ejacula-
alty to bottoms is interesting for a number of tion and ‘‘bottoming’’ does not, this could
reasons. These results may be interpreted limit the number of clients that tops could
sociologically as the premium attached to see in a given period of time and drive the pre-
masculine behavior in gay communities. mium. Essentially, there could be a ‘‘scarcity
The premium for tops is consistent with premium’’ for top services. A search of escort
literature that notes that gay men prize tradi- advertisements, however, reveals that top es-
tionally masculine behaviors and sexual corts who mention ‘‘bottoming’’ also mention
roles, and the penetrative partner in sexual that they charge a significant premium for
acts is canonically considered more mascu- ‘‘bottoming’’ services. Similarly, a detailed
line. The fact that men who act in the analysis of client reviews and online forums
dominant sexual position charge higher pri- does not show that clients demand ejacula-
ces for services is consistent with the social tion more from top escorts than from bot-
acceptance of quasi-heteronormativity within toms. I take this as evidence of the social
696 American Sociological Review 75(5)

Table 7. Implicit Prices of Race and Sexual Behavior Interactions in the Male Escort Market

Sexual Behavior
Race Top Bottom Versatile

Black .110** 2.286** .036


(.0321) (.110) (.0476)
White .0655* 2.0667 2.010
(.0297) (.0348) (.0234)
Asian .311 2.0729 .00786
(.291) (.206) (.131)
Hispanic .0930 2.0258 .0251
(.0537) (.0882) (.0400)
Multirace .0645 .112 2.00291
(.0556) (.131) (.0591)
Other Race .0662 2.332 2.250
(.206) (.206) (.207)
Observations 1,476 1,476 1,476
R-squared .013 .009 .002

Note: Robust standard errors in parentheses. Columns represent separate regressions where the log of
the price is the dependent variable. Entries are the coefficients on the interaction of the row and column.
For example, the ‘‘Black Top’’ coefficient is the coefficient of the black*top interaction term in the
regression. All regressions include controls for race, city, sexual behaviors, other personal
characteristics, and an intercept. See Data Appendix for variable definitions.
*p \ .05; **p \ .01 (two-tailed tests).

penalty of ‘‘bottoming’’ and the premium of The results are striking. Black, Hispanic,
‘‘topping.’’ While biology could certainly and white men each receive a substantial pre-
play a role, the social position of tops ap- mium for being tops, but the largest premium
pears to be the dominant force behind the is for black men (nearly 12 percent, $24).
top premium.22 The premium for Hispanics is greater than
9 percent ($18, p \ .1), while the premium
The Intersection of Race and for whites is less than 7 percent ($14). There
Sexual Behaviors is no statistically significant top premium for
Asian escorts. The penalty for being a bottom
As described earlier, the intersection of race also varies by race: white bottoms face a pen-
and sexual behaviors could shed light on alty of nearly 7 percent ($14, p \ .1), while
the connection between hegemonic mascu- black bottoms face a penalty of nearly 30
linity and racial sexual stereotypes. In partic- percent ($60), the largest penalty seen in
ular, black men are expected to be dominant any of the results in Table 7. There is no bot-
sexually and Asians are expected to be pas- tom penalty for Asians or Hispanics.
sive. I investigate these intersections by look- The top/bottom price differential also
ing at interactions between race and sexual varies by race. While the differential for
behaviors. Table 7 shows estimates of the whites and Hispanics is close to the overall
value of advertised sexual behaviors for top/bottom differential (13.2 percent [$26]
men by race, where each entry shows the and 12.3 percent [$25], respectively; the esti-
implicit price of the interaction of that race mates of Table 6 put the differential between
and sexual behavior (e.g., the premium or 14.1 and 17.6 percent), the differential for
penalty to being an Asian top or a versatile blacks is more than twice the differential
white). for any other racial group, 36.5 percent
Logan 697

($73).23 These results are consistent with time to see how and if their behaviors, iden-
intersectionality theory, in which black men tities, advertisements, and personas change.
who conform to stereotypes of hypermascu- Doing so would add an important dimension
linity and sexual dominance are highly to this literature.
sought after, and those who do not conform
are severely penalized. These types of stereo-
types appear within the male escort market, DISCUSSION AND
and they influence premiums and penalties
for sexual behaviors. Predictions for Asians,
CONCLUSIONS
however, are not borne out in the data—I Using a quantitative online data source
found neither a top premium nor a bottom describing male sex workers, this study ad-
penalty for Asian escorts. dressed important questions pertaining to
male sex work. These questions relate to
basic facts about male escorts, their geo-
LIMITATIONS AND FUTURE graphic distribution, and the relationship
between escort characteristics, sexual behav-
RESEARCH
iors, and prices. The present results should
While this study makes use of novel data to stimulate further research into male sex
test theories of male sex work, several limita- work and related areas of gender, sexuality,
tions to the present analysis should be noted. masculinity, race, and deviance. The present
First, I analyzed only the largest website for results should also show how concepts from
escort advertisements, and the results might ethnographic and qualitative work in social
not hold for competitors. For example, if cer- science can be subjected to quantitative
tain types of escorts are more likely to con- empirical approaches, including statistical
gregate at different websites, they would tests of hypotheses.
not be captured in my data, limiting my abil- Overall, I found that male sex work is
ity to describe the market in general. Second, markedly different from its female counter-
information in the advertisements is posted part. For instance, escorts are present in cities
by escorts and therefore constitutes a self- with low and high gay concentrations; this
report. While I exploited independent data result supports work that suggests a nontrivial
to confirm the precision of the price measure, portion of the male escort clientele is
I cannot say with certainty that there are no not gay-identified. Personal characteristics,
omitted confounders in the data. Third, the except for those pertaining to body build,
dataset may be missing variables that could are largely not related to male escorts’ prices.
influence the price of escort services, such Muscular men enjoy a premium in the mar-
as endowment, an escort’s education level, ket, while overweight and thin men face
or expertise in specific sexual conduct. a penalty, which is consistent with hege-
These limitations should inform future monic masculinity and the literature on the
research. For example, future studies should body and sexuality. Conformity to hege-
analyze competing websites with a similar monic masculine physical norms is well-
methodology to confirm or dispute results rewarded in the market.
presented here. Similarly, detailed analysis The premium to being a top is substantial,
of client-operated websites, which review as is the penalty for being a bottom, again
escort services, could act as an independent consistent with the theory of hegemonic mas-
check on the veracity of information posted culinity. When interacting these behaviors
in escort advertisements. Future research with race, I found that black men are at the
could develop panel data on male escorts extremes—they have the largest premiums
that would allow one to track escorts over for top behavior and the largest penalties
698 American Sociological Review 75(5)

for bottom behavior. This is consistent with unexplored here. For instance, due to data
intersectionality theory in that gay communi- limitations I cannot discuss class dimensions
ties prize black men who conform to racial inherent in male sex work. An important
stereotypes of sexual behavior and penalize question for intersectionality theory in light
those who do not. While the sexually domi- of the results presented here, is how race
nant black male is feared in heterosexual and sexual behavior interact with class mas-
communities, he is rewarded handsomely in culinities to yield premiums and penalties
gay communities. in this market. Furthermore, I have not
Given the results, the ways in which explored causal estimates of male escort
desire interacts with racial stereotypes should behavior on prices, which would be key for
receive significant attention in masculinity policy discussions such as the feasibility of
studies. Theoretically, these results should sexual behavior change among male sex
renew attention on the complex construction workers to minimize disease risk (Connell
of masculinities among gay men, in which 2002).
counter-hegemonic groups adopt and reiter- Research on male sex work needs to move
ate hegemonic masculine norms among at an accelerated pace. As this study shows,
themselves, explicitly reinforcing hegemonic there is much to be gained from an inte-
norms. In particular, further work at the grated, interdisciplinary approach to the sub-
nexus of the construction of masculinity ject. Future developments along this line
among gay men, hegemonic masculinity, would enhance and extend our understanding
and racial inequality would be a fruitful of sexuality and gender in general, and male
area of research. sex work in particular, shedding light on
Further research on race, sexuality, and important issues in social research and public
commerce is needed to address issues policy.

Data Appendix
My data are the universe of male sex workers advertising on the chosen website in the United States at the time of data
collection (January 2008 to May 2008). These data represent the entire population. Each escort has a page specific to
him, from which I gathered information. The website generates its income only from escorts, clients do not pay to
access advertisements. I identify each escort uniquely using the following information (see Figure 1 for an example
of an escort advertisement):
User ID: Each escort account on the site has a unique user ID. This allows me to check against the possibility of
double counting escorts who may change location during the data collection period.
User name: Each escort has a username that is displayed next to the ID number at the top of the advertisement.
Services provided: Under this heading, each escort has the option of noting the following services:
Incall: Escort responds ‘‘yes’’ or ‘‘no.’’
Outcall: Escort responds ‘‘yes’’ or ‘‘no.’’
Incall price: The price (by the hour) that incall services are provided at if incalls are provided.
Outcall price: The price (by the hour) that outcall services are provided at if outcalls are provided.
Contact information: Under this heading, each escort has the option of noting the following:
Phone: Phone number with area code (recorded y/n if a number is listed).
Cellular: Cellular number with area code (recorded y/n if a number is listed).
Pager: Pager number with area code (recorded y/n if a number is listed).
Prefers phone contact: Listed if escort prefers that clients contact him by phone (y/n).
Prefers e-mail contact: Listed if escort prefers that clients contact him by e-mail (y/n).
Location: The location listed under the heading is the primary location, the locations with suitcase avatars next to
them are cities the escort is willing to travel to. I recorded all of these locations.
Age: Age is recorded in years.
Height: Height is reported in feet and inches, I recorded height in inches.
Logan 699

Weight: Weight is listed in 20 pound intervals beginning at 130 pounds and ending at 200 pounds (e.g., 150 to 170
pounds). I took the midpoint of the range given by an escort. If an escort’s text ad lists a weight, I recorded that exact
weight in place of the midpoint range.
Race: White, Black/African American, Asian, Hispanic, multiracial, or other.
Hair color: Black, blonde, brown, grey, or red.
Eye color: Black, blue, brown, green, or hazel.
Body type: Athletic/swimmer’s build, average, a few extra pounds, muscular/buff, or thin/lean.
Body hair: Hairy, moderately hairy, shaved, or smooth.
From the text of escort advertisements, I recorded mention of the following (because I read the advertisements, I do
not record the instance of a word but its meaning, which can be implied from the context):
Top: Escorts stated they are a top (the penetrative partner in anal sex).
Bottom: Escorts stated they are a bottom (the receptive partner in anal sex).
Versatile: Escorts indicated they are versatile (both top and bottom).
No attitude: Escorts noted they have ‘‘no attitude’’; that is, they are willing to see clients without regard to race,
body type, physical appearance, or disability.
Safe: Escorts noted they are disease and drug free and participate only in safer sex.

Note: In American gay society, men may not only be tops, bottoms, or versatile, but also ‘‘versatile tops’’ and ‘‘versatile
bottoms.’’ The meaning of such terms is the distinction between one man who would rarely or never partake in an activity
(a top would never perform as a bottom and vice versa), and a man who occasionally partakes in an activity (a versatile top
would occasionally bottom and vice versa). These terms are well established in American gay society (Sadownick 1996).

Acknowledgments Second, I focus on how the market for male sex


work functions and the implications one can draw
I thank Rene Almeling, Megan Andrew, Rodney from that knowledge in forming testable
Andrews, Raj Arunachalam, Ann Biddlecom, Andrea hypotheses.
Cann Chandrasekher, Niambi Carter, Lisa Cook, Reanne 4. In the past, male prostitution also took the form of
Frank, Ruth Peterson, Patricia Reagan, Fabio Rojas, transvestite sex work, and male brothels were not
Manisha Shah, Arnold Smotherman, Bryan Sykes, uncommon (Chauncey 1994). Dorais (2005) shows
Edward Walker, Jason Whitesel, ASR’s anonymous re- that male brothels have survived in other cultures.
viewers and the editors, numerous seminar participants, Friedman (2003) shows that street male and trans-
and participants at the 2009 Population Association of sexual sex workers do not work in the same loca-
America Annual Meetings for helpful suggestions. tions in the postwar United States.
5. Social stigma of male prostitution is assumed to be
lower than that for female prostitution (West 1993),
Funding but such claims are subject to criticism (Chauncey
1994; Dorais 2005; Friedman 2003). The claim
Portions of this research were supported by the Robert
that male prostitutes face less social stigma implic-
Wood Johnson Foundation, whose support is gratefully
itly conditions on sexual orientation, which itself is
acknowledged. The views expressed are those of the
stigmatizing.
author and do not necessarily reflect the views of the
6. Although most gay-related publications continue to
Robert Wood Johnson Foundation.
have small sections devoted to male escorts, the size
of these sections has decreased. The growth of on-
line escorting has largely eliminated a caste of
Notes male sex workers—‘‘bar boys’’—who would con-
1. These older works include Ginsburg (1967), Hoff- gregate in known ‘‘hustler bars’’ to meet potential
man (1972), Boyer (1989), Salamon (1989), McNa- clients (Luckenbill 1986; Parsons, Koken, and
mara (1994), and Pettiway (1996). Bimbi 2004).
2. The chosen data source contains more than 1,930 7. Many escort advertisements note that escorts have
unique advertisements; the next largest competitor specialties in ‘‘discretion’’ and in serving married
contains fewer than 1,500. men.
3. Before providing an overview of male sex work 8. Sociologists (e.g., Hacker 1957; Weinberg and Wil-
among men who advertise online in the United liams 1974) have noted conflicts within masculin-
States, two caveats should be noted. First, I concen- ities before the theory of hegemonic masculinity
trate on male escorts as opposed to street workers. developed.
700 American Sociological Review 75(5)

9. Connell and Messerschmidt (2005) express doubts approximated by exp(g)21, where g is the coeffi-
about this hybrid, but they concede that hegemonic cient in the regression (Halvorsen and Palmquist
masculinity could be informed by and inform gay 1980).
masculinities. 20. The price differential is calculated as exp(Top) –
10. Male escorts’ advertisements commonly note their exp(Bottom).
‘‘ruggedness’’ or ‘‘manliness’’ and the ‘‘refine- 21. There are caveats to my interpretation of the corre-
ment’’ of their ‘‘generous gentlemen’’ clients. lation of safe sex with higher prices in the male
11. Full description of the data and all statistical code escort market. It could be that clients who desire
for the analysis is available from the author by unsafe sex may desire unsafe sex with escorts
e-mail request. See the Data Appendix for data con- who are less likely to participate in unsafe sex gen-
struction details. erally and may reward them for that.
12. I manually inspected every advertisement to clas- 22. See the online supplement (http://asr.sagepub.com/
sify behaviors from their texts. supplemental) for a fuller delineation of the biolog-
13. See Logan and Shah (2009) for more on the role of ical hypothesis.
pictures in the male escort market. 23. See note 20.
14. ‘‘Versatile’’ escorts often describe themselves as
‘‘versatile tops’’ or ‘‘versatile bottoms’’; this de-
notes a preference for one activity but a willingness References
to participate in the other. I code ‘‘versatile tops’’
and ‘‘versatile bottoms’’ as ‘‘versatile.’’ See the Aggleton, Peter, ed. 1999. Men who Sell Sex: Interna-
Data Appendix for further details. tional Perspectives on Male Prostitution and HIV/
15. Economists have noted problems with some of the as- AIDS. Philadelphia, PA: Temple University Press.
sumptions underlying the hedonic empirical approach. Allen, Donald M. 1980. ‘‘Young Male Prostitutes: A
For example, Rosen (1974) assumes that the market Psychosocial Study.’’ Archives of Sexual Behavior
for the good or service in question is perfectly com- 9:399–426.
petitive and that the range of products is approxi- Almeling, Rene. 2007. ‘‘Selling Genes, Selling Gender:
mately continuous. Yet many markets are not Egg Agencies, Sperm Banks, and the Medical Mar-
perfectly competitive, and even fewer have a contin- ket in Genetic Material.’’ American Sociological
uum of goods (which requires a large variety of prod- Review 72:319–40.
ucts of the same type in the market). Rosen also Anderson, Eric. 2002. ‘‘Openly Gay Athletes: Contest-
assumes that all product attributes will be observed ing Hegemonic Masculinity in a Homophobic Envi-
by market participants; this, too, is not true for some ronment.’’ Gender and Society 16:860–77.
goods, particularly escort services. For these reasons, Arunachalam, Raj and Manisha Shah. 2008. ‘‘Prostitutes
some object to the hedonic approach and its interpre- and Brides?’’ American Economic Review Papers
tation (Bartik 1987; Brown and Rosen 1982; Epple and Proceedings 98:516–22.
1987). Fortunately, recent advances in applied econo- Atkins, Dawn, ed. 1998. Looking Queer: Body Image
metrics show that the hedonic approach used here can and Identity in Lesbian, Gay, and Transgender Com-
be used to uncover implicit prices of characteris- munities. New York: Routledge.
tics in markets that are not perfectly competitive, Bajari, Patrick and C. Lanier Benkard. 2005. ‘‘Demand
where there is not a continuum of goods, and Estimation with Heterogeneous Consumers and
where all product characteristics are not observed Unobserved Product Characteristics: A Hedonic
(Bajari and Benkard 2005). In short, one can esti- Approach.’’ Journal of Political Economy
mate implicit prices in this market without making 113:1239–76.
the rigid market assumptions that have been prob- Baldwin, James. 1985. The Price of the Ticket: Collected
lematic in the literature. Nonfiction 1948–1985. New York: St. Martin’s.
16. I control for an escort’s location not only because Bartik, Timothy J. 1987. ‘‘The Estimation of Demand
price may vary with geography, but because spe- Parameters in Hedonic Price Models.’’ Journal of
cific locations may have more or fewer competitors, Political Economy 95:81–88.
which can exert an independent effect on prices in Beren, Susan E., Helen A. Hayden, Denise E. Wilfley,
a particular market. and Carlos M. Grilo. 1996. ‘‘The Influence of Sexual
17. This calculation assumes that escorts see 25 clients Orientation on Body Dissatisfaction in Adult Men
per month. and Women.’’ International Journal of Eating Dis-
18. These results are robust to the inclusion of age- orders 20:135–41.
squared and when substituting body mass index Berg, Nathan and Donald Lien. 2002. ‘‘Measuring the
(BMI) for height and weight. Effect of Sexual Orientation on Income: Evidence
19. Because the specification is semi logarithmic, the of Discrimination?’’ Contemporary Economic Policy
percentage change of a dichotomous indicator is 20:394–414.
Logan 701

———. 2006. ‘‘Same-Sex Sexual Behavior: U.S. Fre- Carpenter, Christopher. 2003. ‘‘Sexual Orientation and
quency Estimates from Survey Data with Simulta- Body Weight: Evidence from Multiple Surveys.’’
neous Misreporting and Non-Response.’’ Applied Gender Issues 21:60–74.
Economics 38:759–69. ———. 2004. ‘‘New Evidence on Gay and Lesbian
Bernstein, Elizabeth. 2005. ‘‘Desire, Demand, and the Household Incomes.’’ Contemporary Economic Pol-
Commerce of Sex.’’ Pp. 101–128 in Regulating icy 22:78–94.
Sex: The Politics of Intimacy and Identity, edited Carpenter, Christopher and Gary Gates. 2008. ‘‘Gay and
by E. Bernstein and L. Schaffner. New York: Lesbian Partnership: Evidence from California.’’
Routledge. Demography 45:573–90.
———. 2007. Temporarily Yours: Sexual Commerce in Chauncey, George. 1994. Gay New York: Gender,
Post-Industrial Culture. Chicago, IL: University of Urban Culture, and the Making of the Gay Male
Chicago Press. World 1890–1940. New York: Basic Books.
Bimbi, David S. 2007. ‘‘Male Prostitution: Pathology, Clarkson, Jay. 2006. ‘‘‘Everyday Joe’ versus ‘Pissy,
Paradigms and Progress in Research.’’ Journal of Bitchy, Queens’: Gay Masculinity on Straight-
Homosexuality 53:7–35. Acting.com.’’ Journal of Men’s Studies 14:
Bimbi, David S. and Jeffrey T. Parsons. 2005. ‘‘Bare- 191–208.
backing among Internet Based Male Sex Workers.’’ Cohan, Deborah L., Johanna Breyer, Cynthia Cobaugh,
Journal of Gay and Lesbian Psychotherapy 9: Charles Cloniger, Antje Herlyn, Alexandra Lutnick,
89–110. and Daniel Wilson. 2004. ‘‘Social Context and the
Bird, Sharon R. 1996. ‘‘Welcome to the Men’s Club: Health of Sex Workers in San Francisco.’’ Paper pre-
Homosociality and the Maintenance of Hegemonic sented at the 2004 International Conference on
Masculinity.’’ Gender and Society 10:120–32. AIDS, July 11 to 16, Bangkok, Thailand.
Black, Dan, Gary Gates, Seth Sanders, and Lowell Tay- Collins, Alan. 2004. ‘‘Sexual Dissidence, Enterprise and
lor. 2000. ‘‘Demographics of the Gay and Lesbian Assimilation: Bedfellows in Urban Regeneration.’’
Population in the United States: Evidence from Urban Studies 41:1789–806.
Available Systematic Sources.’’ Demography Collins, Patricia Hill. 1999. Black Feminist Thought:
37:139–54. Knowledge, Consciousness, and the Politics of
———. 2002. ‘‘Why do Gay Men Live in San Fran- Empowerment. New York: Harper Collins.
cisco?’’ Journal of Urban Economics 51:54–76. ———. 2000. ‘‘Gender, Black Feminism and
Black, Daniel, Seth Sanders, and Lowell Taylor. 2007. Black Political Economy.’’ Annals of the American
‘‘The Economics of Lesbian and Gay Families.’’ Academy of Political and Social Science 568:
Journal of Economic Perspectives 21:53–70. 41–53.
Boyer, Debra. 1989. ‘‘Male Prostitution and Homosex- Connell, Jeffrey A. 2002. ‘‘Male Sex Work: Occupa-
ual Identity.’’ Journal of Homosexuality 17:151–84. tional Health and Safety.’’ Paper presented at the
Brown, James N. and Harvey S. Rosen. 1982. ‘‘On the 2002 International Conference on AIDS, July 7 to
Estimation of Structural Hedonic Price Models.’’ 12, Barcelona, Spain.
Econometrica 50:765–68. Connell, R. W. 1987. Gender and Power: Society, the
Browne, Jan and Victor Minichiello. 1996. ‘‘Research Person, and Sexual Politics. Sydney, Australia: Al-
Directions in Male Sex Work.’’ Journal of Homosex- len and Unwin.
uality 31:29–56. ———. 1992. ‘‘A Very Straight Gay: Masculinity,
Calhoun, Thomas C. and Greg Weaver. 1996. ‘‘Rational Homosexual Experience, and the Dynamics of Gen-
Decision-Making among Male Street Prostitutes.’’ der.’’ American Sociological Review 57:737–51.
Deviant Behavior 17:209–227. ———. 1995. Masculinities. Sydney, Australia: Allen
Cameron, Samuel, Alan Collins, Stephen Drinkwater, and Unwin.
Ford Hickson, David Reid, Jennifer Roberts, Connell, R. W. and James W. Messerschmidt. 2005.
Michael Stephens, and Peter Weatherburn. 2009. ‘‘Hegemonic Masculinity: Rethinking the Concept.’’
‘‘Surveys and Data Sources on Gay Men’s Life- Gender and Society 19:829–59.
styles and Socio-Sexual Behavior: Some Key Con- Court, Andrew T. 1939. ‘‘Hedonic Price Indexes
cerns and Issues.’’ Sexuality & Culture 13: with Automotive Examples.’’ Pp. 99–117 in Gen-
135–51. eral Motors Corporation, The Dynamics of Auto-
Cameron, Samuel, Alan Collins, and Neill Thew. 1999. mobile Demand. New York: General Motors
‘‘Prostitution Services: An Exploratory Empirical Corporation.
Analysis.’’ Applied Economics 31:1523–29. Demetriou, Demetrakis Z. 2001. ‘‘Connell’s Concept of
Cantu, Lionel. 2002. ‘‘A Place Called Home: A Queer Hegemonic Masculinity: A Critique.’’ Theory and
Political Economy.’’ Pp. 382–94 in Sexuality and Society 30:337–61.
Gender, edited by C. Williams and S. Stein. New D’Emilio, John D. 1997. ‘‘Capitalism and Gay Iden-
York: Blackwell. tity.’’ Pp. 169–78 in The Gender/Sexuality Reader,
702 American Sociological Review 75(5)

edited by R. Lancaster and M. diLeonardo. New Han, Chong-Suk. 2006. ‘‘Geisha of a Different Kind:
York: Routledge. Gay Asian Men and the Gendering of Sexual Iden-
Donaldson, Mike. 1993. ‘‘What is Hegemonic Masculin- tity.’’ Sexuality & Culture 10:3–28.
ity?’’ Theory and Society 22:643–57. Hennen, Peter. 2005. ‘‘Bear Bodies, Bear Masculinity:
Dorais, Michel. 2005. Rent Boys: The World of Male Sex Recuperation, Resistance, or Retreat?’’ Gender and
Workers. London, UK: McGill-Queens University Society 19:25–43.
Press. Herzog, David B., Kerry L. Newman, Christine J. Yeh,
Dowsett, Gary W. 1993. ‘‘I’ll Show You Mine, if You’ll and Meredith Warshaw. 1991. ‘‘Body Image Satis-
Show Me Yours: Gay Men, Masculinity Research, faction in Homosexual and Heterosexual Males.’’
Men’s Studies, and Sex.’’ Theory and Society International Journal of Eating Disorders 11:
22:697–709. 356–96.
Edlund, Lena and Evelyn Korn. 2002. ‘‘A Theory of Hewitt, Christopher. 1995. ‘‘The Socioeconomic Posi-
Prostitution.’’ Journal of Political Economy tion of Gay Men: A Review of the Evidence.’’ Amer-
110:181–214. ican Journal of Economics and Sociology 54:
Epple, Dennis. 1987. ‘‘Hedonic Prices and Implicit Mar- 461–79.
kets: Estimating Demand and Supply Functions for Hoffman, Martin. 1972. ‘‘The Male Prostitute.’’ Sexual
Differentiated Products.’’ Journal of Political Econ- Behavior 2:16–21.
omy 95:59–80. Hooker, Evelyn. 1957. ‘‘The Adjustment of the Male
Epstein, Steven G. 2006. ‘‘The New Attack on Sexuality Overt Homosexual.’’ Journal of Projective Techni-
Research: Morality and the Politics of Knowledge Pro- ques 21:18–31.
duction.’’ Sexuality Research and Social Policy 3:1–12. Hotelling, Harold. 1929. ‘‘Stability in Competition.’’
Frankel, Todd C. 2007. ‘‘In Forest Park, the Roots of Economic Journal 39:41–57.
Sen. Craig’s Misadventure.’’ St. Louis Post-Dis- Humphries, Laud. 1970. Tearoom Trade: Impersonal
patch, August 31. Sex in Public Places. Hawthorne, NY: Aldine de
Friedman, Mack. 2003. Strapped for Cash: A History of Gruyter.
American Hustler Culture. Los Angeles, CA: Itiel, Joseph. 1998. A Consumer’s Guide to Male Hus-
Alyson. tlers. New York: Harrington Park Press.
Gertler, Paul, Manisha Shah, and Stefano M. Bertozzi. Jepsen, Lisa and Christopher Jepsen. 2002. ‘‘An Empir-
2005. ‘‘Risky Business: The Market for Unprotected ical Analysis of the Matching Patterns of Same-Sex
Commercial Sex.’’ Journal of Political Economy and Opposite-Sex Couples.’’ Demography 39:
113:518–50. 435–54.
Ginsburg, Kenneth N. 1967. ‘‘The ‘Meat Rack’: A Study Joffe, Helene and Julie E. Dockrell. 1995. ‘‘Safer Sex:
of the Male Homosexual Prostitute.’’ American Lessons from the Male Sex Industry.’’ Journal of
Journal of Psychotherapy 21:170–85. Community and Applied Social Psychology 5:
Giusta, Marina D., Maria L. Di Tommaso, and Steinar 333–46.
Strom. 2009. ‘‘Who’s Watching? The Market for Kaye, Kerwin. 2001. ‘‘Male Prostitution in the Twenti-
Prostitution Services.’’ Journal of Population Eco- eth Century: Pseudohomosexuals, Hoodlums Homo-
nomics 22:501–516. sexuals, and Exploited Teens.’’ Journal of
Green, Adam I. 2008a. ‘‘The Social Organization of Homosexuality 46:1–77.
Desire: The Sexual Fields Approach.’’ Sociological Koken, Juline A., David S. Bimbi, and Jeffrey T. Par-
Theory 26:25–50. sons. 2009. ‘‘Male and Female Escorts: A Compar-
———. 2008b. ‘‘Health and Sexual Status in an Urban ative Analysis.’’ Pp. 205–232 in Sex for Sale:
Gay Enclave: An Application of the Stress Process Prostitution, Pornography, and the Sex Industry,
Model.’’ Journal of Health and Social Behavior 2nd ed., edited by R. Weitzer. New York:
49:436–51. Routledge.
Griliches, Zvi. 1961. ‘‘Hedonic Price Indexes for Auto- Koken, Juline A., Jeffrey T. Parsons, Joseph Severino,
mobiles: An Econometric Analysis of Quality and David S. Bimbi. 2005. ‘‘Exploring Commercial
Change.’’ Pp. 173–96 in The Price Statistics of the Sex Encounters in an Urban Community Sample of
Federal Government. New York: National Bureau Gay and Bisexual Men: A Preliminary Report.’’
of Economic Research. Journal of Psychology and Human Sexuality
Hacker, Helen M. 1957. ‘‘The New Burdens of Mascu- 17:197–213.
linity.’’ Marriage and Family Living 19:227–33. Loftus, Jeni. 2001. ‘‘America’s Liberalization in Atti-
Halvorsen, Robert and Raymond Palmquist. 1980. ‘‘The tudes toward Homosexuality.’’ American Sociologi-
Interpretation of Dummy Variables in Semilogarith- cal Review 66:762–82.
mic Equations.’’ American Economic Review Logan, Trevon D. and Manisha Shah. 2009. ‘‘Face
70:474–75. Value: Information and Signaling in an Illegal
Logan 703

Market.’’ National Bureau of Economic Research, ———. 2008. ‘‘Black ‘Tops’ and Asian ‘Bottoms’: The
Working Paper No. 14841. Impact of Race and Gender on Coupling in Queer
Luckenbill, David F. 1986. ‘‘Deviant Career Mobility: Communities.’’ Working paper, UCLA, Los An-
The Case of Male Prostitutes.’’ Social Problems geles, CA.
33:283–96. Rosen, Sherwin. 1974. ‘‘Hedonic Prices and Implicit
MacDonald, Laura A. 2007. ‘‘America’s Toe-Tapping Markets: Product Differentiation in Pure Competi-
Menace.’’ New York Times, September 2. tion.’’ Journal of Political Economy 82:34–55.
Marlowe, Julian. 1997. ‘‘It’s Different for Boys.’’ Sadownick, Douglas. 1996. Sex between Men: An Inti-
Pp. 141–44 in Whores and Other Feminists, edited mate History of the Sex Lives of Gay Men Postwar
by J. Nagle. New York: Routledge. to Present. New York: Harper Collins.
McBride, Dwight. 2005. Why I Hate Abercrombie and Salamon, Edna D. 1989. ‘‘The Homosexual Escort
Fitch: Essays on Race and Sexuality. New York: Agency: Deviance Disavowal.’’ The British Journal
NYU Press. of Sociology 40:1–21.
McNamara, Robert P. 1994. The Times Square Hustler: Schrock, Douglas and Michael Schwalbe. 2009. ‘‘Men,
Male Prostitution in New York City. Westwood, CT: Masculinity, and Manhood Acts.’’ Annual Review
Praeger. of Sociology 35:277–95.
Nagel, Joane. 2000. ‘‘Ethnicity and Sexuality.’’ Annual Scott, John. 2003. ‘‘A Prostitute’s Progress: Male Pros-
Review of Sociology 26:107–133. titution in Scientific Discourse.’’ Social Semiotics
Nardi, Peter, ed. 2000. Gay Masculinities. London, UK: 13:179–99.
Sage. Sedgwick, Eve Kosofsky. 1990. The Epistemology of the
Parker, Melissa. 2006. ‘‘Core Groups and the Transmis- Closet. Berkeley, CA: University of California Press.
sion of HIV: Learning from Male Sex Workers.’’ Simmel, Georg. [1907] 1971. ‘‘Prostitution.’’ Pp. 121–
Journal of Biosocial Science 38:117–31. 26 in On Individuality and Social Forms, edited by
Parsons, Jeffrey T., David S. Bimbi, and Perry N. Halki- D. N. Levine. Chicago, IL: University of Chicago
tis. 2001. ‘‘Sexual Compulsivity among Gay/Bisex- Press.
ual Male Escorts who Advertise on the Internet.’’ Steele, Bruce C. and Sean Kennedy. 2006. ‘‘Hustle and
Journal of Sexual Addiction and Compulsivity Grow.’’ The Advocate, April 11, 2006.
8:113–23. Stein, Arlene. 1989. ‘‘Three Models of Sexuality:
Parsons, Jeffrey T., Juline A. Koken, and David S. Drives, Identities, and Practices.’’ Sociological The-
Bimbi. 2004. ‘‘The Use of the Internet by Gay and ory 7:1–13.
Bisexual Male Escorts: Sex Workers as Sex Educa- Uy, Jude M., Jeffrey T. Parsons, David S. Bimbi, Juline
tors.’’ AIDS Care 16:1021–35. A. Koken, and Perry N. Halkitis. 2004. ‘‘Gay and
———. 2007. ‘‘Looking beyond HIV: Eliciting Individ- Bisexual Male Escorts who Advertise on the Inter-
ual and Community Needs of Male Internet Es- net: Understanding Reasons for and Effects of
corts.’’ Journal of Homosexuality 53:219–40. Involvement in Commercial Sex.’’ International
Pascoe, C. J. 2007. Dude, You’re a Fag: Masculinity and Journal of Men’s Health 3:11–26.
Sexuality in High School. Berkeley, CA: University Varghese, Beena, Julie E. Maher, Thomas A. Peterman,
of California Press. Bernard M. Branson, and Richard W. Steketee. 2002.
Pettiway, Leon E. 1996. Honey, Honey, Miss Thang: ‘‘Reducing the Risk of Sexual HIV Transmission:
Being Black, Gay and on the Streets. Philadelphia, Quantifying the Per-Act Risk for HIV on the Basis
PA: Temple University Press. of Choice of Partner, Sex Act, and Condom Use.’’
Phua, Voon C. and Gayle Kaufman. 2003. ‘‘The Cross- Sexually Transmitted Diseases 29:38–43.
roads of Race and Sexuality: Date Selection among Ward, Jane. 2000. ‘‘Queer Sexism: Rethinking Gay Men
Men in Internet ‘Personal’ Ads.’’ Journal of Family and Masculinity.’’ Pp. 152–75 in Gay Masculinities,
Issues 24:981–94. edited by P. Nardi. Thousand Oaks, CA: Sage
Pompeo, Joe. 2009. ‘‘The Hipster Rent Boys of New Publications.
York.’’ New York Observer, January 27. ———. 2008. ‘‘Dude-Sex: White Masculinities and
Pruitt, Matthew. 2005. ‘‘Online Boys: Male for ‘Authentic’ Heterosexuality among Dudes Who
Male Internet Escorts.’’ Sociological Focus 38: Have Sex with Dudes.’’ Sexualities 11:414–34.
189–203. Weinberg, Martin S. and Colin J. Williams. 1974. Male
Reeser, Todd W. 2010. Masculinities in Theory. Oxford, Homosexuals: Their Problems and Adaptations.
UK: Wiley Blackwell. New York: Oxford University Press.
Reid-Pharr, Robert F. 2001. Black Gay Man: Essays. Weitzer, Ronald. 2005. ‘‘New Directions in Research in
New York: NYU Press. Prostitution.’’ Crime, Law & Social Change 43:
Robinson, Russell K. 2007. ‘‘Uncovering Covering.’’ 211–35.
Northwestern University Law Review 101: ———. 2009. ‘‘Sociology of Sex Work.’’ Annual
1809–850. Review of Sociology 35:213–34.
704 American Sociological Review 75(5)

West, Donald J. 1993. Male Prostitution. New York: the National Bureau of Economic Research, and a Robert
Haworth Press. Wood Johnson Foundation Scholar in Health Policy
Zelizer, Viviana A. 1994. The Social Meaning of Money. Research at the University of Michigan. His work in-
New York: Basic Books. cludes studies of African American migration, economic
analysis of illegal markets, the economics of marriage
transfers, and measures of historical living standards.
Selected publications have appeared in American Eco-
Trevon D. Logan is an Associate Professor of econom- nomic Review, Historical Methods, and Journal of Eco-
ics at The Ohio State University, a Research Associate at nomic History.

Das könnte Ihnen auch gefallen