Content uploaded by Milaine Alarie
Author content
All content in this area was uploaded by Milaine Alarie on Jul 18, 2018
Content may be subject to copyright.
M A J T. C McGill University
The “Cougar” Phenomenon: An Examination of the
Factors That Inuence Age-Hypogamous Sexual
Relationships Among Middle-Aged Women
Using data from the National Survey of Family
Growth, the authors considered the prevalence
of the “cougar” phenomenon and the character-
istics of middle-aged women who reported hav-
ing sexual relationships with younger men in the
past 12 months. They found that roughly 13% of
sexually active women between ages 35 and 44
had slept with a man who was at least 5 years
younger. Contrary to conventional assumptions,
the results show that women with low incomes
and those who self-identify as “other race”
(not White or Black) are more likely to be in
an age-hypogamous sexual relationship. Rela-
tive to all other relationship statuses, previously
married women are the most likely to choose
younger partners. Finally, the results suggest
that age-hypogamous relationships are not sim-
ply “ings”; a majority of them last at least
2 years, and a sizable share of “cougars” are
married to their younger partners. These results
highlight the need to reconsider our conven-
tional understanding of women’s sexual rela-
tionships at midlife.
When it comes to heterosexual relationships,
choosing a slightly older man is the norm for
Department of Sociology, McGill University, 855
Sherbrooke West, Leacock Building, Room 712, Montreal,
Quebec H3A 2T7 Canada (milaine.alarie@mail.mcgill.ca).
This article was edited by Deborah Carr.
Key Words: dating, gender stereotypes, intimacy, mate
selection, midlife, sexual behavior.
women around the world, a partnering practice
referred to as age hypergamy. In the United
States, husbands marrying for the rst time are,
on average, 2 years older than their wives (Dar-
roch, Landry, & Oslak, 1999; Vera, Berardo,
& Berardo, 1985). Overall, nearly one third of
all husbands in the United States are at least
4 years older than their wives, but the reverse
(known as age hypogamy) is much less frequent.
Indeed, census data suggest that only 7.7% of
wives are at least 4 years older than their husband
(U.S. Bureau of the Census, 2012, Table FG3).
Empirical work has shown that age hypergamy
is not only quite common for marital unions but
also for sexual relationships more broadly, that
is, regardless of the commitment level in the
relationship. For instance, Darroch et al. (1999)
looked at the age gap between women and their
main sex partner and found that 38% of women
age 15–44 were in a sexual relationship with a
man at least 3 years older than they; in compar-
ison, only 10% of women were sleeping with a
man at least 3 years their junior.
Over the past decade, there has been an
increasing number of media portrayals and
discussions about “cougars”—the term often
used to refer to middle-aged women engag-
ing in age-hypogamous sexual relationships—
suggesting that we are witnessing the rise of
a new phenomenon. For instance, television
shows such as Sex and the City (1998–2004)
and Cougar Town (2009–present) or movies
such as Prime (2005) and Adore (2013) portray
middle-aged women having sexual relationships
1250 Journal of Marriage and Family 77 (October 2015): 1250–1265
DOI:10.1111/jomf.12213
The “Cougar” Phenomenon 1251
with younger men. There is also a growing
consumer market for “cougars,” with a prolifer-
ation of books advising women on how to nd
a younger male partner (e.g., Gibson, 2002),
dating websites (e.g., https://cougarlife.com),
parties, annual meetings, and boat cruises
organized for women interested in younger men.
The media portrayals of age hypogamy
in heterosexual relationships often present
“cougars” as White, afuent, and successful
women. Furthermore, women’s relationships
with younger men are generally depicted as
temporary ings that rarely culminate into a
serious commitment (Kaklamanidou, 2012;
Tally, 2006). Cultural representations of women
engaging in age-hypogamous relationships are
not always positive; indeed, scholars have shown
that media depictions of “cougars” often hold
these women up for ridicule or present them as
somewhat dangerous (Barrett & Levin, 2014;
Collard, 2012; Kaklamanidou, 2012). In many
ways, the term “cougar” is culturally loaded,
given that it includes these negative stereotypes
(Montemurro & Siefken, 2014). To avoid such
negative connotations, hereafter we refer to
older woman–younger man relationships as
age-hypogamous sexual relationships.
At present, there is little academic research on
the occurrence of age-hypogamous sexual rela-
tionships. This can be attributed in part to the fact
that little scholarly attention has been directed
at improving our understanding of middle-aged
women’s sexuality beyond health-related stud-
ies (Gannon, 1999; Montemurro & Siefken,
2014; Rostosky & Brown, 2000). The limited
amount of scholarship that has addressed age
heterogamy between heterosexual partners has
tended to focus almost exclusively on married
couples and has virtually ignored those involved
in less committed relationships. As a conse-
quence, our knowledge about the prevalence of
age-hypogamous sexual relationships (regard-
less of the commitment level) is very limited.
Furthermore, it is unclear who these women are
and to what extent they conform to the media
depictions of “cougars.”
With this study, we began to ll this gap in
the literature by answering two primary research
questions, using detailed data on sexual relation-
ships drawn from the National Survey of Family
Growth (NSFG) 2002 (see http://www.cdc.
gov/nchs/nsfg.htm). First, how prevalent are
age-hypogamous sexual relationships in the
United States? Second, what are the
sociodemographic and attitudinal character-
istics associated with middle-aged women who
engage in these non-conventional relationships?
Our focus here is on age hypogamy as a sexual
practice. Therefore, the term sexual relation-
ships includes any relationships in which the
two partners are involved sexually, regardless of
their commitment level (i.e., ranging anywhere
from marriage to a one-night stand).
B
Aging and Sexual Relationships: Gendered
Patterns
Scholarship on sexuality has established a
clear pattern of declining sexual activity at
middle age and beyond, as the proportion of
men and women with no sex partner increases
with age (Carpenter, Nathanson, & Kim, 2006;
Das, Waite, & Laumann, 2012). However, this
decline in sexual activity is not equal for both
men and women. For instance, Carpenter et al.
(2006) found that, among heterosexual 40- to
59-year-olds, a larger proportion of women
reported having no sex partners as compared to
men. Furthermore, middle-aged men are three
times more likely than women of the same age
to report having multiple sex partners at a given
point in time. Despite these gender differences
in sexual activity, the authors specied that the
majority (78%) of middle-aged women had
engaged in sexual activity with a partner in the
past 12 months.
Men and women also report different pref-
erences in terms of partner’s age. For example,
on the basis of a content analysis of 1,094 per-
sonal advertisements, Jagger (2005) found that
47% of “middle-aged” (35–44 years old) men,
compared to only 8% of women from the same
age group, were seeking a partner younger than
35. Looking at heterosexual men and women age
40–69, Montenegro (2003) found that 64% of
men, compared to 17% of women, reported pre-
ferring a partner who is at least 5 years younger.
Buunk, Dijkstra, Kenrick, and Warntjes
(2001) studied this question using a more
nuanced approach, asking both men and women
of different ages (individuals age 20, 30, 40,
50, and 60 years) to indicate their preference
for different types of relationships (from casual
sex to marriage). They found that men reported
different age preferences depending on the
level of commitment they wanted out of the
1252 Journal of Marriage and Family
relationship; overall, the less committed the
relationship was, the younger the ideal female
partner was. On the other hand, women reported
preferring men their own age regardless of
whether they were looking for marriage or
for casual sex. Buunk and his colleagues also
found that women’s preferences for partners
of a similar age remained constant regardless
of women’s age. However, other scholars have
found that, compared to younger women, older
women are more likely to be interested in dat-
ing or marrying younger men (Jagger, 2005;
South, 1991).
Beyond their preferences in partners, aging
also appears to inuence women’s and men’s
ability or choice to (re)partner in unique ways
(England & McClintock, 2009; Gelissen, 2004;
Schneider, Sledge, Shuchter, & Zisook, 1996;
U.S. Bureau of the Census, 2002a). For women
over 35, the ratio of single women to single
men begins to increase (England & McClin-
tock, 2009; U.S. Bureau of the Census, 2003),
meaning that single, middle-aged women out-
number the pool of potential male partners of a
similar age. The increasing sex ratio, although
somewhat inuenced by sex differences in mor-
tality, results largely from the pervasiveness
of age hypergamy in heterosexual relationships
and especially the fact that men, as they age,
tend to choose women who are increasingly
younger (England & McClintock, 2009; Shafer,
2013). According to England and McClintock
(2009), men who marry in their 20s do so,
on average, with women 1 year their junior.
In comparison, men who marry in their 60s
tend to marry women between 9 and 12 years
younger. Aging also affects women’s likelihood
of having younger male partners; indeed, older
women are more likely than younger ones to
engage in age-hypogamous relationships (Caron
& Levesque, 2004; Darroch et al., 1999). For
instance, Darroch et al. (1999) reported that
roughly 3% of women 20–24 years of age had
a partner who was at least 3 years younger,
whereas the proportion increased to 15% for 40-
to 44-year-old women.
Although, on average, husbands have tra-
ditionally been older than their wives, one
must note that rates of age-hypergamous and
age-hypogamous relationships have changed
over time in the United States. For instance, the
average age gap between rst-time husbands and
wives has steadily decreased over the past cen-
tury in the United States, from 4 years in 1900 to
roughly 2 years today (England & McClintock,
2009; Vera et al., 1985). Furthermore, in 1900,
15.8% of women and 47.1% of men had married
someone 5 or more years younger than them,
compared to 3.1% of women and 26.9% of men
in 1980 (Atkinson & Glass, 1985). In regard to
more recent changes in partnering patterns, a
comparison of U.S. census data showed that the
proportion of women who marry a man at least
4 years younger has slightly increased over the
past decade, from 6.4% in 2000 to 7.7% in 2012
(U.S. Bureau of the Census, 2000, Table 56;
2012, Table FG3).
In sum, when it comes to nding a mar-
riage companion or a sex partner, aging affects
men and women differently. As they advance
through midlife and old age, men tend to choose
women who are increasingly younger. Consid-
ering that most women would prefer having a
partner of a similar age, the imbalanced sex ratio
among single people makes it quite challeng-
ing for middle-aged women to nd a suitable
partner. Ultimately, middle-aged men are more
likely than their female counterparts to (re)marry
or to have a sex partner.
Explaining Age Heterogamy Patterns
in Heterosexual Relationships
Recent scholarship points to a number of social
factors that could explain why, still to this
day, few women compared to men choose
younger partners. Although it is undeniable
that North American women today enjoy more
economic, social, and sexual freedom com-
pared to earlier generations (see Kamen, 2000;
Montemurro & Siefken, 2014; Mundy, 2012;
Nielsen & Rudberg, 2007; Vanier Institute of
the Family, 2010), women’s sexuality still faces
much more societal regulation than men’s, a
phenomenon often referred to as the sexual
double standard. Middle-aged women’s sex-
uality is further constrained by the gendered
double standard of aging (Carpenter et al.,
2006; England & McClintock, 2009; Lai &
Hynie, 2011). Accordingly, aging is considered
to be associated with rapidly decreasing sex
appeal and dating potential for women, but
this is much less the case for men. In addition,
older women’s sexuality is often thought of in
restrictive and negative terms because of the
cultural opposition between motherhood and
sexuality (Friedman, Weinberg, & Pines, 1998;
Montemurro & Siefken, 2012).
The “Cougar” Phenomenon 1253
One can nd evidence of such norms by ana-
lyzing the cultural representations of middle-
aged women in the media. Despite an increasing
visibility of middle-aged women’s sexuality in
the media (Tally, 2006; Weitz, 2010), contempo-
rary scripts involving older women continue to
(re)produce norms that constrain aging women’s
sexuality and identities. Indeed, older women
are still frequently presented in the media as
asexual (Baumann & de Laat, 2012; Gannon,
1999), or their sexuality is presented as prob-
lematic for the stability of the nuclear family
(Tally, 2006). In addition, middle-aged women’s
bodies are often presented as objects of ridicule
instead of as objects of desire, and middle-aged
women’s sexual desires are often depicted in a
humorous way (Weitz, 2010).
Furthermore, specic representations of
“cougars” or other colloquial references to
older women’s sexuality (e.g., MILF—meaning
literally “mother I’d like to fuck”) found in the
media are frequently tainted with negative con-
notations. The idea of women dating younger
men is often devalued through humorous com-
ments, and these women are often depicted as
ultimately presenting a danger for younger men
as well as for themselves (Barrett & Levin,
2014; Collard, 2012; Kaklamanidou, 2012).
Women dating younger men are commonly pre-
sented as obsessed with getting sexual validation
from men or desperately trying to cling to their
youth (Barrett & Levin, 2014; Kaklamanidou,
2012). Finally, Weitz (2010) pointed out that the
“cougar” lifestyle is not presented in the media
as accessible and acceptable for all middle-aged
women, only for those who are “taut, dyed, trim,
energetic, and t, and ...intheirearly40s”
(p. 31).
As a consequence, aging for women can be a
source of anxiety and self-doubt, thereby inu-
encing their sexuality or dating habits. Recent
scholarship suggests that most middle-aged
women express dissatisfaction with their bodies
and that a majority of them report feeling less
attractive now compared to their earlier years
(Koch, Manseld, Thurau, & Carey, 2005;
McLaren & Kuh, 2003). Furthermore, older
women have lower levels of self-esteem com-
pared to men in the same age group (McMullin
& Cairney, 2004). Women with a poor body
image tend to avoid common social interaction
such as physical activity and sexual intimacy
(McLaren & Kuh, 2003), and they report lower
levels of sexual activity, enjoyment, and desire
(Koch et al., 2005).
Montemurro and Siefken (2014) found that
although some women thought of the term
“cougar” positively—associating the word with
an image of a powerful, condent, and attractive
older woman—most expressed negative feelings
or mixed emotions toward it. Many believed
that it depicted older women as predators.
With regard to women’s experiences with age
hypogamy in the heterosexual marriage context,
Proulx, Caron, and Logue (2006) found that
many women partnered with younger men face
emotional and psychological challenges. Indeed,
their small qualitative study of age-hypogamous
marital unions indicated that many of the female
interviewees admitted facing doubts and fear
with respect to their own attractiveness and
the fact that they will age sooner than their
partner. Furthermore, half of the eight women
interviewed reported facing stigma due to their
non-conventional relationships. Although only
a few empirical studies have analyzed women’s
experiences with age hypogamy, it appears that
challenging the cultural norms associated with
sexuality and romantic relationships by being
with a much younger man may be difcult
psychologically and/or emotionally for some
women.
Beyond the sociocultural factors that help
explain why men are more likely than women to
choose a younger partner, previous scholarship
has pointed to a number of individual character-
istics that are associated with age hypogamous
relationships. We outline these factors below.
Determinants of Age Hypogamy in Sexual
Relationships
The literature points to a number of individ-
ual characteristics that may inuence a woman’s
likelihood of engaging in age-hypogamous rela-
tionships. One must note that the vast majority of
studies on age heterogamy focus on one type of
sexual relationship: marriage. Although limited
in that they ignore all other types of sexual rela-
tionships, these studies do provide some insights
into the sociodemographic and attitudinal char-
acteristics of middle-aged women who choose
younger sex partners.
Racial/ethnic background. In the United States,
sex differences in mortality rates vary substan-
tially across different racial and ethnic groups,
1254 Journal of Marriage and Family
and these differences may inuence relationship
patterns. Indeed, recent U.S. census data (U.S.
Bureau of the Census, 2002b) reveal that for
every 100 African American woman between
age 35 and 44 there are approximately 89
African American men in the same age group.
Given such an imbalanced sex ratio within
the African American community, African
American middle-aged women who want to
nd a partner might have to choose men out-
side of their age group. Atkinson and Glass
(1985) found empirical support for this claim,
at least with regard to marriage; they found
that African American women were signi-
cantly more likely than White women to be in
either age-hypergamous or age-hypogamous
marriages.
Scholars also have examined whether
middle-aged Latinas have unique relationship
patterns. These studies have produced rather
inconclusive ndings. South (1991) assessed
women’s willingness to marry a younger man
and found that Latinas were just as open as
White women to marry a man at least 5 years
younger. Shehan, Berardo, Vera, and Carley
(1991) examined actual marriage patterns and
found that Latinas and non-Latinas were equally
likely to be in age-hypogamous relationships.
On the other hand, Vera et al. (1985) reported
that Latinas were less likely than either White
or African American women to be married to
a younger man. Overall, it appears that African
American women may be more likely than White
women to engage in age-hypogamous sexual
relationships. Furthermore, despite some incon-
sistencies in the literature, we expected Latinas
to be less likely than non-Latinas to engage in
an age-hypogamous sexual relationship.
Level of education. Women’s level of education
might play a role in age hypogamy in heterosex-
ual relationships because (a) educated women
are more likely to be nancially independent,
and (b) they are more likely to support more lib-
eral ideas regarding gender norms and sexuality
(Janus & Janus, 1993). Given this, it seems rea-
sonable to assume that educated women might
be more likely to choose a younger partner.
However, research has shown that women’s low
education status is associated with both types
of age-heterogamous marriages/common-law
unions: age hypergamy and age hypogamy
(Atkinson & Glass, 1985; Boyd & Li, 2003).
On the basis of the literature on age heterogamy
in marriage, we anticipated that women with
lower levels of education would be more likely
than women with higher levels of education to
engage in sexual relationships with younger
men.
Income. Many authors highlight the importance
of nancial independence as a crucial factor
inuencing women’s control over their romantic
and sex lives (Hyde & Oliver, 2000; Kamen,
2000; Mundy, 2012). It is generally assumed,
likely because of media depictions, that women
with lower incomes choose an older male part-
ner partly as a strategy to ascend the social
status ladder and that only nancially inde-
pendent women would choose (and would be
able to attract) younger partners. Despite these
conventional expectations, South (1991) found
that when women were asked about desired
qualities of a hypothetical partner, a woman’s
income played no signicant role in her will-
ingness to marry a younger man (or an older
one). In fact, if one looks beyond preferences
and consider who women marry, the evidence
shows that both types of age-heterogamous
marriages are more prevalent among families
with low incomes (Boyd & Li, 2003; Vera et al.,
1985). For instance, Vera et al. (1985) reported
that low-income families were twice as likely as
more afuent families to comprise an older wife
and a younger husband. If the ndings about
age hypogamy in marriage are also relevant
in all types of sexual relationships, then we
should expect to see a negative association
between income and age hypogamy in sexual
relationships.
Marital status. Prior research suggests that a
woman’s marital status is likely a contributing
factor in age-hypogamous sexual relation-
ships. For instance, when comparing the main
male sex partner of married women to that
of non-married women, Darroch et al. (1999)
found that non-married women were much more
likely to be sleeping with a man at least 3 years
their junior. Indeed, nearly one in three single
women 40 to 44 years of age were in a sexual
relationship with a younger partner, compared
to only 11% of married women. Although
this study shows a rather strong association
between marital status and age-hypogamous
sexual relationships, the authors did not dif-
ferentiate between never-married women and
women who had previously been married.
The “Cougar” Phenomenon 1255
Distinguishing between these two groups of
non-married women seems important in light
of recent scholarship (Gelissen, 2004) indicat-
ing that having experienced marriage before
inuences women’s likelihood of entering an
age-heterogamous marriage. Therefore, we
examined the inuence of three marital statuses
to see how they may inuence the likelihood
of entering into an age-hypogamous sexual
relationship. Drawing on Darroch et al.’s (1999)
ndings, we assumed that both previously mar-
ried women as well as never-married women
would be more likely than married/cohabiting
women to engage in age-hypogamous sexual
relationships.
Religion. A number of scholars have docu-
mented the inuence that religion has on sexual
behavior (Darroch et al., 1999; Runkel, 1998;
Simons, Burt, & Peterson, 2009; Uecker, 2008).
Darroch et al. (1999), for instance, found that
religious faith plays a role in age hypergamy
in sexual relationships. Their examination of
teenage girls’ sexual behavior showed that
Catholics were signicantly less likely than
Protestants to have a much older male partner
(a 6-year age gap or more). Level of religios-
ity also seems to inuence both attitudes and
sexual behavior (Simons et al., 2009; Uecker,
2008). For instance, in their study of under-
graduate students, Simons et al. (2009) found
that religious respondents tended to hold a more
conservative view of the circumstances under
which sexual behavior is acceptable, to have
their rst sexual intercourse at an older age,
and to have fewer consecutive sex partners.
Although no study has specically tested the
inuence of religion on the likelihood of engag-
ing in age-hypogamous sexual relationships,
we assumed that Catholics are less likely to
engage in nontraditional, age-hypogamous
sexual relationships than those who participate
in Protestant denominations or those who are
not religious.
Conservative attitudes. Researchers also have
found a connection between conservative atti-
tudes and women’s sexuality. In particular,
women who subscribe to conservative views
regarding gender and sexuality are much more
likely to reject nontraditional sexual behaviors.
Carpenter et al. (2006), for instance, found that
sexual conservatism was a strong predictor of
lower numbers of lifetime sex partners and
the rejection of having casual sex partners. If
conservative attitudes also inuence women’s
willingness to engage in age-hypogamous sex-
ual relationships, then we should expect that
women who espouse more conservative views
would be less inclined to engage in nontradi-
tional relationships such as the ones studied
here.
Additional controls: Age and number of sex
partners. As discussed earlier, the extant litera-
ture suggests that there is a positive relationship
between women’s age and the likelihood of
both being open to the idea of being in an
age-hypogamous relationship (Jagger, 2005;
South, 1991) and actually engaging in one
(Caron & Levesque, 2004; Darroch et al.,
1999). On the basis of the ndings from these
studies, all of our models included a control
for respondent’s age. All of our models also
included a control for the number of sex partners
each respondent had had in the past 12 months.
We did this to account for the possibility that
having more sex partners increases the likeli-
hood that one or more of the sexual relationships
will meet our denition of age hypogamy.
M
Data, Sample, and Research Design
To assess the hypotheses outlined above and
to improve our understanding of the determi-
nants of heterosexual age-hypogamous sexual
relationships among middle-aged women, we
analyzed data taken from the “female le”
of the NSFG, Cycle VI, 2002. The survey is
a nationally representative sample of 7,643
women between ages 15 and 44. To capture
the “cougar” phenomenon, we limited the
sample to middle-aged women, which in this
case were female respondents ages 35–44.
Ideally, our preference would be to have data
on respondents who meet the entire range of
the conventional delineation of middle age,
which typically spans 35–55 years (Koch et al.,
2005). Unfortunately, the NSFG interviewed
only women who were 44 and under at the
time of the interview. Although the sample is
limited in the sense that it captures only early
middle age, it still offers important insights into
sexual relationships that no other survey offers.
Indeed, the NSFG is one of the only large-scale
surveys that asks respondents to identify all
1256 Journal of Marriage and Family
male sex partners they have had in the last 12
months as well as their partners’ age, which
can then be used to identify age-hypogamous
sexual relationships. Using this survey allows
us to study sexual relationships of all forms,
not merely marital unions, as previous scholars
have done.
Once we restricted our sample to women
between 35 and 44 years, the sample size
dropped from the original 7,643 female respon-
dents in the entire NSFG survey to 2,478.
Because we were principally interested in
developing our understanding of the sexual rela-
tionships of middle-aged women, we removed
women who had not been sexually active in the
past 12 months, which further reduced our sam-
ple to 2,131 respondents. Missing data for some
of our variables brought our nal sample down
to 1,519 cases. Nearly all of the missing cases
were lost as a result of respondents not reporting
either their own age or the age of their partners
when they rst had sex. The vast majority of
these nonresponse items were currently married
women with only one partner, suggesting that
there is a small bias in favor of non-married
women. Finally, all of our regression models
were estimated using White’s (1980) correction
for heteroscedasticity.
Dependent Variable
The literature on age heterogamy in relationships
is characterized by a lack of a consensus on a
precise denition of age hypogamy or what actu-
ally constitutes a “cougar” in any empirically
testable sense. Studies that have examined this
topic have used an age gap of 3 or more years
(Darroch et al., 1999; Shafer, 2013), 4 or more
years (Vera et al., 1985), at least 5 years (Atkin-
son & Glass, 1985; Shehan et al., 1991), and
even a 10-year +age gap (Proulx et al., 2006).
Because we lacked any precise denition of age
hypogamy, we adopted the approach used by
Lawton and Callister (2010), who operational-
ized age heterogamy using two separate thresh-
olds between partners: (a) a 5-year gap and (b) a
10-year gap.
To create the dependent variables (women
engaging in age-hypogamous sexual relation-
ships), we used two related survey questions
from the NSFG. For each sex partner a respon-
dent had had in the last 12 months, she was
asked to separately indicate two things: (a) what
her age was when they rst had sex and (b) what
his age was when it happened. A subtraction
between her age and his at the time of their
rst sexual encounter produced an age gap for
each reported sexual relationship in the last 12
months. We then created a dummy variable to
capture “women engaging in age-hypogamous
sexual relationships.” Each respondent was
coded 1 if she had at least one sex partner in the
past 12 months who was a minimum of 5 years
younger than she and 0 otherwise. We created
a similar dummy variable to identify women
who had had at least one sex partner who was a
minimum of 10 years their junior.
Finally, using the same sociodemographic
and attitudinal characteristics as for age
hypogamy, we introduced a separate set of
regression equations that examined age hyper-
gamy in sexual relationships. We did this to
see whether our results were specic to age
hypogamy or whether certain sociodemographic
characteristics were instead associated more
broadly with age heterogamy. We created two
dummy variables to represent women who had
had at least one older sex partner, one using the
5-year threshold and the other one using the
10-year age gap.
Explanatory Variables
We measured the factors associated with the
likelihood of entering into a sexual relationship
with a younger man using variables drawn from
the NSFG. We operationalized race/ethnicity
using a set of dummy variables. One survey
question asked respondents to self-identify as
being either “White,” “Black,” or “Other.” (Note
that the NSFG used the term Black, whereas
elsewhere in this article we have opted to use the
term African American.) Three separate dummy
variables were constructed to capture these racial
distinctions (with “White” used as the reference
category in all regression models). Another sur-
vey item asked respondents to indicate whether
they considered themselves as Latina/Hispanic.
A separate dummy variable was constructed to
capture whether being Latina inuenced the like-
lihood of engaging in age-hypogamous sexual
relationships.
We operationalized education using a series
of dummy variables as follows: (a) “less than
a ninth-grade education,” (b) “high school lev-
el” (10th–12th grade), (c) “undergraduate level”
(1–3 years of college), and (d) “graduate lev-
el” (4 or more years of university or graduate
The “Cougar” Phenomenon 1257
school). The highest category (graduate level)
was used as the reference category in all of
our models. Income was measured using the
NSFG’s question about family income. The orig-
inal family income variable contained a large
number of categories with highly uneven inter-
vals between each category. To ease interpre-
tation, we recoded the original family income
variable into a smaller number of categories
(ve), with the highest category of income being
$75,000 and up (also the highest category in the
original data).
We operationalized marital status using a
three-category dummy variable. In order to
capture women who were in serious relation-
ships, married women and women cohabiting
with their partner were grouped together and
are referred to herein as “married/cohabiting”
women. Women who were either divorced, sep-
arated, or widowed were recoded into a single
category labeled “previously married.” Finally,
single women who had not previously been
married were coded as “never married.” The
inuence of religion was captured in two ways.
The rst was a four-category dummy variable
indicating the religious faith each respondent
identied with. The dummies were (a) “no reli-
gion,” (b) “Protestantism,” (c) “Catholicism,”
and (d) “non-Christian religion.” Drawing on
Darroch et al.’s (1999) ndings regarding the
inuence of the Catholic faith on age hypergamy
in teenage girls’ relationships, Catholicism was
chosen as the reference category. The second
indicator of religion measured religiosity,using
the NSFG question about how often respon-
dents attend religious services. Here we used a
three-category dummy variable in the NSFG:
(a) respondents who attended religious services
at least once a month (“attend religious services
regularly”), (b) those who attended less than
once a month (“rarely attend”), and (c) those
who never attended (“never attend”). “Attend
religious services regularly” was used as the
reference category.
We measured the inuence of conservative
attitudes using three separate dummy variables.
Respondents were asked several attitudinal
questions to gauge their views about sexuality
and gender roles. We recoded three of the most
pertinent questions as follows: (a) Respondents
were coded as 1 if they agreed with the state-
ment, “It is better for a person to get married
than to go through life being single” (otherwise
coded 0); (b) respondents were coded 1 if they
agreed with the question, “It is much better
for everyone if the man earns the main living
and the woman takes care of the home and
family”; and (c) a 1 was assigned to women
who disagreed with the statement, “Any sexual
act between two consenting adults is all right.”
We assume that women who were coded with
1s espoused more conservative attitudes about
sexuality and gender roles, which may inuence
their willing to engage in age-hypogamous
sexual relationships.
Age was assessed as a continuous variable
ranging from 35 to 44 years of age. Note that
a nonlinear specication was assessed but was
not supported, so we report only results using the
linear specication of age. Finally, the number of
partners in the last 12 months was measured as
a continuous variable.
R
Descriptive Statistics
Descriptive statistics related to the number of
different sex partners respondents had had in
the past 12 months are provided in Table 1. In
the table, one can see that 90% of our sam-
ple had just one partner, that nearly 7% had
two partners, and that the remaining 3% had
three or more partners during this time period.
As expected, the number of sex partners dur-
ing the past 12 months was strongly inuenced
by marital status; indeed, we calculated that 26
of the 1,008 married women (2.6%) had had
two or more partners in the past 12 months.
In comparison, 18.2% of never-married women
and 28.2% of previously married women in
our sample had multiple sex partners in the
year prior to the interview. Overall, previously
married women were signicantly more likely
to have had multiple sex partners in the past
year than either never-married women or mar-
ried/cohabiting women.
Tab le 1 . Number of Male Sex Partners in the Past 12
Months (N=1,519)
Number of partners Number of women Percentage (%)
1 1,368 90.1
2 101 6.7
3281.8
4140.9
5+80.5
1258 Journal of Marriage and Family
Tab le 2 . Descriptive Statistics for the Independent Variables: Women Engaging in Age-Hypogamous Relationships Versus
Those Who Do Not (N=1,519)
Age gap ≥5 years Age gap ≥10 years
Hypogamous Nonhypogamous Hypogamous Nonhypogamous
Variables (n=201) (n=1,318) (n=67) (n=1,452)
Race
White 68.81 73.14 67.16 72.80
Black 22.77 22.61 25.37 22.52
Other 8.42 4.25 7.46 4.68
Latina (1 =Latina) 21.39 15.02 22.39 15.56
Education
Ninth grade or less 10.89 6.53 5.97 7.16
High school 31.68 31.64 43.28 31.06
Undergraduate level 36.14 33.46 37.31 33.68
Graduate level 21.29 28.38 13.43 28.10
Income
$0–$19,999 32.67 21.17 38.81 21.97
$20,000–$39,999 32.18 23.44 32.84 24.24
$40,000–$59,999 14.85 19.95 13.43 19.56
$60,000–$74,999 8.42 10.09 5.97 9.99
$75,000+11.88 25.34 8.96 24.24
Marital status
Married 48.02 69.20 34.33 67.84
Previously married 36.63 18.59 52.24 19.56
Never married 15.35 12.22 13.43 12.35
Religious faith
Catholic 26.37 28.60 28.36 28.31
Protestant 55.72 53.95 53.73 54.20
Non-Christian religion 6.47 4.02 7.46 4.20
No religion 11.44 13.43 10.44 13.29
Note: All table values are percentages.
Demographic statistics that allow for com-
parisons between women who had been in
an age-hypogamous sexual relationship and
those who had not are presented in Table 2.
In the table, one can see that 201 of the 1,519
middle-aged women in our sample (13.2%) had
been in a sexual relationship with at least one
man who was 5 or more years younger than
they in the 12 months prior to the interview.
Sixty-seven women (4.4%) had a partner who
was at least 10 years their junior. One also can
see that there were more married/cohabiting
women (48.02%) than either previously married
(36.63%) or never-married women (15.35%)
engaged in an age-hypogamous sexual relation-
ship at the 5-year threshold but that most of the
women who were in a sexual relationship with a
man more than 10 years younger than they were
previously married women (52.24%).
We also examined, using more detailed
descriptive statistics, the possibility that
middle-aged women were in only casual,
short-term relationships with younger men.
This examination revealed that, despite com-
mon media depictions of age-hypogamous
sexual relationships as ings, a signicant pro-
portion of middle-aged women sleeping with
younger men do maintain their relationship
for a substantial period of time. We identied
240 age-hypogamous sexual relationships in
total in which women were at least 5 years
older than their partner (note that some women
had more than one younger male partner in
the past 12 months). Of those relationships,
roughly 54% lasted at least 2 years. Further-
more, 87 of the 201 (43%) women engaged in an
age-hypogamous sexual relationship (using the
5-year age gap denition) were, in fact, married
The “Cougar” Phenomenon 1259
to or cohabiting with their younger partner.
These results shed some light on the issue of
the longevity of women’s sexual relationships
with younger men by clearly showing that about
half of these relationships are long term and that
many involve marriage or cohabitation.
Regression Results
Age-hypogamous relationships using a 5-year
age gap. In Table 3 we present our logistic
regression predicting the likelihood that a
woman will be in an age-hypogamous sexual
relationship using the set of factors outlined ear-
lier. Model 1 assessed the predictive power of
these explanatory variables using the 5-year age
gap to identify age-hypogamous relationships.
Estimates from Model 1 identied a positive
and signicant relationship for both a woman’s
age and the number of sex partners she had had
in the past 12 months. This suggests that as
they move farther into their middle-age years,
women are increasingly likely to sleep with a
younger man.
Our results also show that women who were
part of the “Other” racial category (i.e., did
not self-identify as White or African Ameri-
can) were signicantly more likely than White
women (the reference category) to be in a sex-
ual relationship with a younger man. Women
who self-identied as part of the “Other” race
category are also more than twice as likely as
Black women to sleep with a younger man. It
is noteworthy, however, that Black women were
just as likely to be in age-hypogamous relation-
ships as White women. There was no signi-
cant difference between Latinas and non-Latinas
with regard to the likelihood of being in an
age-hypogamous sexual relationship.
Our results related to women’s level of edu-
cation also were somewhat surprising. Despite
prior evidence to the contrary (Atkinson &
Glass 1985), education does not appear to be
inuencing women’s likelihood of being in
sexual relationships with younger men. Alter-
native specications were used to ensure that
our results were not dependent on our particular
operationalization, but they did not meaning-
fully alter our overall ndings. It may be that
our results diverge from those of previous stud-
ies because we did not limit our analysis to
married couples. However, our ndings bring
additional empirical support to the negative rela-
tionship between income and age-hypogamous
relationships found in previous research (Boyd
& Li, 2003; Vera et al., 1985): We found that
women who reported higher incomes were sig-
nicantly less likely to be with a younger man.
In fact, the least afuent women in the survey
(who earned <$20,000 yearly) were more than
twice as likely to be in an age-hypogamous
sexual relationship as the most afuent women
($75,000 and higher).
With regard to the inuence of women’s mar-
ital status, our ndings show that previously
married women were more likely to have a
sex partner who is younger compared to mar-
ried/cohabiting women (the reference category).
Never-married women appeared to be as likely
as married/cohabiting women to sleep with a
man at least 5 years their junior. Also noteworthy
is that never-married women did not differ sig-
nicantly from previously married women when
we used the 5-year age gap as a measure for age
hypogamy.
Only one of the three measures of conser-
vative attitudes was signicantly related to the
odds of female respondents having been in an
age-hypogamous sexual relationship in the past
12 months. Women who had disagreed with
the statement, “Any sexual act between two
consenting adults is all right” were less likely
to have been with a younger man than those
who had agreed with this statement. Our results
related to religion also were consistent with
some of our expectations. We found that Protes-
tant women in the survey were more likely than
Catholic women to have engaged in sexual rela-
tionships with younger men. That said, women
who reported being nonreligious or those who
practiced a non-Christian religion did not dif-
fer signicantly from Catholic women (the ref-
erence category). Finally, women who reported
never attending religious services were much
more likely to be in a sexual relationship with
younger men compared to those who attended
services regularly (the reference category).
Age-hypogamous relationships using a 10-year
age gap. Model 2 of Table 3 involved the
same explanatory variables as presented in
Model 1 but used a 10-year threshold to mea-
sure age-hypogamous sexual relationships.
We include the model as a tool to assess how
sensitive our ndings are to the age threshold
we used to identify age-hypogamous sexual
relationships. Similar to Model 1, both age and
number of sex partners were positively related
1260 Journal of Marriage and Family
Tab le 3 . Logistic Regression Estimates of the Determinants of an Age-Hypogamous Relationship
Model 1: Age gap of 5+years Model 2: Age gap of 10+years
Predictors Coef. Robust SE OR Coef. Robust SE OR
Controls
Age 0.08** 0.03 1.08 0.09†0.05 1.10
Number of partners 0.61** 0.37 1.84 0.49*** 0.22 1.63
Conservative attitudes
1=Being married is better −0.05 0.16 0.95 −0.25 0.20 0.78
1=Stay-at-home mom is better −0.19 0.14 0.83 −0.68*0.15 0.51
1=Some sex acts are bad −0.41*0.14 0.67 −0.43 0.24 0.65
Religious faith (ref.: Catholic)
Protestant 0.38†0.32 1.46 −0.21 0.45 1.24
Non-Christian religion 0.58 0.73 1.79 0.50 1.09 1.64
No religion −0.47 0.21 0.62 −0.77 0.24 0.46
Religiosity (ref.: attends religious services regularly)
Never attends 0.61** 0.44 1.84 0.90*0.90 2.46
Rarely attends 0.16 0.24 1.17 0.59†0.57 1.82
Race (ref.: White)
Black −0.11 0.19 .90 0.28 0.46 1.32
Other race 0.65†0.69 1.91 0.45 0.96 1.56
Latina (1 =Latina) 0.38 0.36 1.47 0.72†0.77 2.05
Education (ref.: graduate level)
Ninth grade or less 0.18 0.42 1.19 −0.06 0.62 0.94
High school −0.21 0.20 .81 0.66 0.84 1.93
Undergraduate level 0.06 0.23 1.06 0.50 0.69 1.65
Family income −0.19** 0.06 .83 −0.16 0.11 0.85
Marital status (ref.: married/cohabiting)
Previously married 0.59** 0.38 1.81 1.19*** 1.12 3.29
Never married 0.21 0.33 1.24 0.08 0.51 1.08
Constant −5.67 −8.12
Pseudo R2.10 .16
Note: All models were corrected for unspecied heteroscedasticity using White’s (1980) correction. Coef. =unstandardized
coefcients; OR =odds ratio; ref. =reference category.
†p≤.10. *p≤.05. **p≤.01. *** p≤.001 (two-tailed).
to the odds that these women had engaged in a
sexual relationship with men at least 10 years
their junior.
Using this more conservative threshold to
capture age-hypogamous sexual relationships,
we found that the difference between White
women and women in the “Other” race cate-
gory was no longer signicant. However, being
Latina did increase the odds of having entered
into a sexual relationship with a much younger
man; in fact, the odds ratios suggest that Latinas
were more than twice as likely as non-Latinas
to engage in sexual relationships with men who
are at least 10 years younger. Women’s level
of education did not appear to be inuencing
our outcome regardless of the particular age
threshold we used. Income was no longer a
signicant predictor when we used a 10-year
threshold. We did nd, however, that previously
married women were more than three times more
likely than both married/cohabiting women and
never-married women to sleep with a man at
least 10 years their junior. There was no sig-
nicant difference between married/cohabiting
women and never-married women.
Again, only one measure of conservative
attitudes signicantly inuenced the odds of
being in this type of relationship. Women who
agreed that “It is much better for everyone if the
man earns the main living and the woman takes
care of the home and family” were less likely
than those who disagreed with this statement
to be in an age-hypogamous sexual relationship
when this threshold was used. In regard to
The “Cougar” Phenomenon 1261
religion, our results indicate that religious faith
was no longer a signicant contributor when we
used the 10-year threshold; however, we noted
that both women who rarely attended religious
services and those who reported that they never
attend religious services were more likely to
sleep with a much younger man compared to
those who reported attending services regularly
(the reference category).
Additional Considerations: Age-Hypergamous
Sexual Relationships
Before we discuss our ndings and their wider
implications, we briey elaborate on women’s
involvement in age-hypergamous sexual rela-
tionships in an attempt to see how it compares
to age hypogamy. First, it is worth mentioning
that, as expected, age hypergamy in sexual rela-
tionships is more common than age hypogamy.
We found that around one in three sexually active
middle-aged women in our sample had had at
least one sex partner who was at least 5 years
older than they, and 14% had had at least one
partner who was 10 or more years older than
they. To test whether different factors inuenced
age hypergamy compared to age hypogamy, we
ran a series of regression models using “women
engaging in age-hypergamous sexual relation-
ships” (at both a 5-year and 10-year age gap) as
the dependent variable.
The results (see Table 4) showed rather clearly
that there are in fact different factors that inu-
ence age-hypogamous and age-hypergamous
sexual relationships. For instance, unlike age-
hypogamous sexual relationships, neither a
woman’s age nor her marital status inuenced
the odds that she will be in a sexual rela-
tionship with an older man. Unlike with age
hypogamy, women’s income did not play a
signicant role in the likelihood of engaging in
an age-hypergamous sexual relationship. This
nding is surprising considering that previous
studies on age heterogamy in marriage (Boyd
& Li, 2003; Vera et al., 1985) found that low
family income was associated with both types
of age-heterogamous marriages. In regard to
race, it played a role only when we compared
Black women with White women. Indeed, Black
women were more likely than White women to
have had a sex partner who is at least 5 years
older.
A substantial difference in the inuence of
religion also existed between age-hypogamous
and age-hypergamous sexual relationships.
When we used the 5-year age gap we found that
women who did not identify with any religion
were signicantly more likely than Catholics
to enter age-hypergamous sexual relationships.
When we used the more conservative measure
of age hypergamy (10-year age gap), though,
we noted that Protestant women, women who
practiced no religion, and non-Christian women
were all signicantly more likely to choose
much older sex partners than Catholic women.
This nding is similar to that of Darroch et al.
(1999), who reported that Catholic teenage girls
were signicantly less likely than Protestants to
have a much older male sex partner. Contrary to
age hypogamy, women’s level of religiosity did
not play a strong role in inuencing age hyper-
gamy in sexual relationships. Finally, unlike
women who had younger partners, women in
age-hypergamous sexual relationships did not
appear to be inuenced by conservative attitudes
about sexuality and gender roles. Together, a
comparison of the results from Table 3 and
Table 4 suggests that the same factors are
not responsible for both age-hypogamous and
age-hypergamous sexual relationships.
D
The objective of this study was to clarify the
prevalence of the “cougar” phenomenon and
to understand the socioeconomic and attitudi-
nal characteristics of women who engage in
age-hypogamous sexual relationships. More
broadly, this study sought to improve our under-
standing of middle-aged women’s sexuality.
Data from the 2002 NSFG showed that
age-hypogamous sexual relationships among
early middle-aged women, although not as com-
mon as age-hypergamous sexual relationships,
do not appear to be particularly rare occurrences.
One should note that our results report statistics
related to age hypogamy as a sexual practice and
should not be interpreted as reecting women’s
age preferences for sex partners. It is possible
that more than 13% of middle-aged women in
our sample would have liked to engage in an
age-hypogamous sexual relationship but were
unable to nd a suitable partner.
Considered together, our results conrm how
distorted common cultural representations of
“cougars,” such as those found in the media,
can be. Indeed, these women are often depicted
as White, afuent, and successful women
1262 Journal of Marriage and Family
Tab le 4 . Logistic Regression Estimates of the Determinants of an Age-Hypergamous Relationship
Model 1: Age gap of 5+years Model 2: Age gap of 10+years
Predictors Coef. Robust SE OR Coef. Robust SE OR
Controls
Age −0.02 0.02 0.98 0.01 0.03 1.01
Number of partners in last 12 months 0.34** 0.17 1.41 0.37*** 0.16 1.45
Conservative attitudes
1=Being married is better −0.03 0.11 0.97 −0.04 0.15 0.96
1=Stay-at-home mom is better −0.06 0.12 0.94 0.07 0.17 1.07
1=Some sex acts are bad 0.16 0.15 1.18 0.09 0.19 1.10
Religious faith (ref.: Catholic)
Protestant 0.06 0.16 1.06 0.47*0.35 1.61
Non-Christian religion 0.17 0.34 1.19 0.71†0.77 2.03
No religion 0.40†0.31 1.49 0.57*0.51 1.76
Religiosity (ref.: attends religious services regularly)
Never attends −0.15 0.15 0.86 −0.01 0.24 1.01
Rarely attends −0.29*0.11 0.75 −0.05 0.18 0.95
Race (ref.: White)
Black 0.38*0.22 1.47 0.20 0.25 1.23
Other race −0.09 0.26 0.91 −0.27 0.34 0.76
Latina (1 =Latina) 0.14 0.21 1.16 0.21 0.31 1.24
Education (ref.: graduate level)
Ninth grade or less −0.58*0.15 0.56 −0.36 0.25 0.70
High school −0.05 0.15 0.95 0.03 0.22 1.03
Undergraduate level −0.22 0.12 0.81 −0.37†0.14 0.69
Family income −0.02 0.05 0.98 −0.04 0.07 0.96
Marital status (ref.: married/cohabiting)
Previously married 0.03 0.16 1.03 0.20 0.24 1.22
Never married −0.20 0.16 0.82 −0.02 0.26 0.99
Constant −0.21 −2.66
Pseudo R2.02 .03
Note: All models were corrected for unspecied heteroscedasticity using White’s (1980) correction. Coef. =unstandardized
coefcients; OR =odds ratio; ref. =reference category.
†p≤.10. *p≤.05. **p≤.01. *** p≤.001 (two-tailed).
who, thanks to their wealth, have been able to
surgically turn back time with their looks or
who are able to literally buy young men’s atten-
tion. Contrary to conventional assumptions, our
results for socioeconomic measures indicated
that education does not play a signicant role
in determining whether women enter sexual
relationships with younger men and that there is
actually a negative relationship between family
income and age-hypogamous sexual relation-
ships (when the 5-year-age gap-threshold was
used).
Our results related to race also challenge
common assumptions about women who choose
younger partners, given that the White women
in our sample did not appear to be more
likely than women of other races to be in an
age-hypogamous sexual relationship. In fact,
using the 5-year threshold, we observed that
women in the “Other” race category (i.e.,
neither African American nor White) were
the most likely to sleep with younger men.
However, it remains unclear which specic
racial/ethnic groups are more likely to engage
in age-hypogamous sexual relationships, and
future scholarship in the area should consider
exploring this issue further. Our ndings also
complicate the story of the relationship between
being African American and age heterogamy
in heterosexual relationships. Indeed, although
previous research (Atkinson & Glass, 1985)
has found that African American women are
The “Cougar” Phenomenon 1263
more likely to be in both types of heterogamous
marriage, our results indicate that, when all
types of sexual relationships are considered, this
relationship is signicant only in the case of age
hypergamy. This nding suggests that, when
middle-aged African American women are con-
fronted with the well-documented imbalanced
sex ratio within their communities (U.S. Bureau
of the Census, 2002b), they appear to choose
older sex partners rather than younger ones.
It is important to note that our ndings
suggest that there might be signicant differ-
ences between previously married middle-aged
women (divorcées/widows/separated women)
and both married/cohabiting women and never-
married women with regard to their sexual
behavior. Indeed, our ndings indicate that, even
when controlling for the number of sex partners,
previously married women are signicantly
more likely than both never-married women
and married/cohabiting women to sleep with a
man at least 10 years younger. Never-married
women in our sample were, however, as likely
as married women to be with a younger man
(regardless of the threshold used).
The difference between never-married
women and previously married women could
be explained in part by the considerable social
pressure on never-married women to nd a
husband (see DePaulo & Morris, 2005; Sharp &
Ganong, 2011); indeed, it seems at least plausi-
ble that women who have never married by the
time they reach midlife might be more cautious
with their mate selection strategies relative to
divorced women. If they still hope to marry,
some single middle-aged women might believe
it is safer to target same-age or older men than
to try building a relationship with a younger
man who might be perceived as not yet ready
for marriage. Previously married women in our
sample were also signicantly more likely than
never-married women to have had more than
one sex partner in the past 12 months, which
suggests that they might be less concerned with
following restrictive norms regarding women’s
sexuality in general.
Our results challenge the common assump-
tion that middle-aged women’s sexual encoun-
ters with younger men are simply temporary
ings that do not lead to serious relationships
such as marriage or cohabitation. In fact, we
found that a sizable share of these relationships
are lengthy and that many age-hypogamous
couples are currently married or cohabiting.
Overall, our ndings should motivate us to
reect on society’s tendency to (re)produce
sexist and ageist conceptions of women’s
sexuality and women’s value more broadly.
Whether clearly stated or implicitly suggested,
cultural representations of age-hypogamous
relationships as ings—often explained away
as a midlife crisis or a woman’s desperate
attempt to cling to her youth—reinforce sexist
normative ideals about gender relations and
women’s sexual behavior. Combined with the
cultural narrative indicating that physical beauty
is the most desirable quality for women and that
beauty fades away with age (Brown, Meginnis,
& Bardari, 2000; Wolf, 1991), the representation
of age-hypogamous relationships as temporary
encourages aging women to doubt themselves
and to view age-hypogamous relationships as
risky for them; women are encouraged to think
that sooner or later younger men will get over
the excitement of the “forbidden fruit” and go
back to those they are really attracted to—that
is, younger women.
More qualitative research is needed to explore
why some middle-aged women choose younger
men as their sex partners (as well as why some
younger men choose older women) despite the
negative cultural representations of these types
of relationships. More research is also needed to
understand how middle-aged women engaging
in age-hypogamous relationships deal with the
double standard of aging (if they even experience
such underappreciation of their self-worth and
physical attractiveness) and identify their stigma
management strategies. Furthermore, consider-
ing that most women report negative or mixed
feelings toward the term “cougar” (Montemurro
& Siefken, 2014), it is safe to say that many of
the women categorized here as “cougars” might
not identify as such (if indeed any of them do).
Other scholars should explore identity choices of
women in age-hypogamous sexual relationships
and their motivations for claiming/rejecting the
term “cougar.”
Last, considering the potential bias associated
with small sample sizes, one should be cautious
with generalizing our ndings to all middle-aged
women. Furthermore, given that the NSFG sur-
veys women up to age 44, it is entirely unclear
whether any of our ndings are relevant for
women over 44. Indeed, considering the dou-
ble standard of aging (Carpenter et al., 2006;
England & McClintock, 2009), women in their
late 40s and 50s might have greater difculty
1264 Journal of Marriage and Family
nding a younger sex partner— or any available
sex partner, for that matter—because, starting
at midlife, the ratio of single women to single
men begins to increase (England & McClintock,
2009; U.S. Bureau of the Census, 2003). This
should be explored if data are made available in
the future.
The present study enabled us to identify
important insights into the types of sexual rela-
tionships in which middle-aged women engage.
The results revealed that age-hypogamous sex-
ual relationships, even though they are not as
common as the younger woman–older man rela-
tionship scheme, are not rare events, and many
of these relationships are lasting and committed.
Overall, our study challenges many of the mis-
conceptions about “cougars” conveyed in the
media and contributes to an improved under-
standing of middle-aged women’s sexuality at
midlife.
N
This article was written with the support of Social Sciences
and Humanities Research Council of Canada. Weare grateful
to Elaine Weiner and Eran Shor for their suggestions.
R
Atkinson, M. P., & Glass, B. L. (1985). Marital age
heterogamy and homogamy, 1900 to 1980. Journal
of Marriage and the Family, 47, 685–691.
Barrett, B., & Levin, D. (2014). What’s love got to
do with it? A qualitative grounded theory con-
tent analysis of romance narratives in the PG
era of World Wrestling Entertainment (WWE)
programming. Sexuality & Culture, 18, 560–591.
doi:10.1007/s12119-013-9211-4
Baumann, S., & de Laat, K. (2012). Socially defunct:
A comparative analysis of the underrepresenta-
tion of older women in advertising. Poetics, 40,
514–541. doi:10.1016/j.poetic.2012.08.002
Boyd, M., & Li, A. (2003). May–December: Cana-
dians in age-discrepant relationships. Canadian
Social Trends, 70, 29–33.
Brown, C. T., Meginnis, K. L., & Bardari, K. M.
(2000). Beauty, sexuality, and identity: The
social control of women. In C. B. Travis & J.
W. White (Eds.), Sexuality, society, and femi-
nism (pp. 237–272). Washington, DC: American
Psychological Association.
Buunk, B. P., Dijkstra, P., Kenrick, D. T., & Warntjes,
A. (2001). Age preferences for mates as related to
gender, own age, and involvement level. Evolution
& Human Behavior, 22, 241–250.
Caron, S. L., & Levesque, L. M. (2004). Dating pref-
erences of women born between 1945 and 1960.
Journal of Family Issues, 25, 833–846.
Carpenter,L.M.,Nathanson,C.A.,&Kim,Y.J.
(2006). Sex after 40? Gender, ageism, and sexual
partnering in midlife. Journal of Aging Studies, 20,
93–106.
Collard, R.-C. (2012). Cougar gures, gender, and
the performances of predation. Gender, Place &
Culture, 19, 518–540. doi:10.1080/0966369X.
2011.610179
Darroch, J. E., Landry, D. J., & Oslak, S. (1999).
Age differences between sexual partners in the
United States. Family Planning Perspectives, 31,
160–167.
Das, A., Waite, L. J., & Laumann, E. O. (2012). Sexual
expression over the life course: Results from three
landmark surveys. In L. M. Carpenter & J. D.
DeLamater (Eds.), Sex for life: From virginity to
Viagra, how sexuality changes throughout our lives
(pp. 236–259). New York: New York University
Press.
DePaulo, B., & Morris, W. (2005). Singles in society
and in science. Psychological Inquiry, 16, 57–83.
England, P., & McClintock, E. A. (2009). The gen-
dered double standard of aging in U.S. marriage
markets. Population and Development Review, 35,
797–816.
Friedman, A,, Weinberg, H., & Pines, A. M. (1998).
Sexuality and motherhood: Mutually exclusive in
perception of women. Sex Roles, 38, 781–800.
Gannon, L. (1999). Women and aging: Transcending
the myths. New York: Routledge.
Gelissen, J. (2004). Assortative mating after divorce:
A test of two competing hypotheses using marginal
models. Social Science Research, 33, 361–384.
Gibson, V. (2002). Cougar: A guide for older women
dating younger men. New York: Firey Books.
Hyde, J. S., & Oliver, M. B. (2000). Gender differ-
ences in sexuality: Results from meta-analysis. In
C. B. Travis & J. W. White (Eds.), Sexuality, soci-
ety, and feminism (pp. 57–77). Washington, DC.:
American Psychological Association.
Jagger, E. (2005). Is thirty the new sixty?
Dating, age and gender in a postmodern,
consumer society. Sociology, 39, 89–106.
doi:10,1177/0038038505049003
Janus, S., & Janus, C. L. (1993). The Janus report on
sexual behavior. New York: Wiley.
Kaklamanidou, B. (2012). Pride and prejudice:
Celebrity versus ctional cougars. Celebrity Stud-
ies, 3, 78–89. doi:10.1080/19392397.2012.644722
Kamen, P. (2000). Her way: Young women remake the
sexual revolution.NewYork:NewYorkUniversity
Press.
Koch, P. B., Manseld, P. K., Thurau, D., & Carey,
M. (2005). “Feeling frumpy”: The relationships
between body image and sexual response changes
in midlife women. Journal of Sex Research, 42,
215–223.
Lai, Y., & Hynie, M. (2011). A tale of two
standards: An examination of young adults’
The “Cougar” Phenomenon 1265
endorsement of gendered and ageist sexual
double standards. Sex Roles, 64, 360–371.
doi:10.1007/s11199-010-9896-x
Lawton, Z., & Callister, P. (2010). Older women–
younger men relationships: The social phe-
nomenon of “cougars.” A research note.
Retrieved from http://ips.ac.nz/publications/les/
be0acfcb7d0.pdf
McLaren, L., & Kuh, D. (2003). Body dissatisfaction
in midlife women. Journal of Women & Aging, 16,
35–54.
McMullin, J. A., & Cairney, J. (2004). Self-esteem
and the intersection of age, class, and gen-
der. Journal of Aging Studies, 18, 75–90.
doi:10.1016/j.jaging.2003.09.006
Montemurro, B., & Siefken, J. (2012). MILFs
and matrons: Images and realities of mothers’
sexuality. Sexuality & Culture, 16, 366–388.
doi:10.1007/s12119-012-9129-2
Montemurro, B., & Siefken, J. (2014). Cougars on the
prowl? New perceptions of older women’s sexual-
ity. Journal of Aging Studies, 28, 35–43.
Montenegro, X. P. (2003). Lifestyles, dating and
romance: A study of midlife singles. Washington,
DC: AARP.
Mundy, L. (2012). The richer sex: How the new
majority of female breadwinners is transform-
ing sex, love, and family. New York: Simon &
Schuster.
Nielsen, H. B., & Rudberg, M. (2007). Fun in
gender—Youth and sexuality, class and genera-
tion. NORA—Nordic Journal of Women’s Studies,
15, 100–113.
Proulx, N., Caron, S., & Logue, M. (2006). Older
women/younger men: A look at the implications
of age difference in marriage. Journal of Couple &
Relationship Therapy, 5, 43–64.
Rostosky, S. S., & Brown C. T. (2000). Menopause
and sexuality: Ageism and sexism unite. In C. T.
Brown & J. W. White (Eds.), Sexuality, society, and
feminism (pp. 181–209). Washington, DC: Ameri-
can Psychological Association.
Runkel, G. (1998). Sexual morality of Christianity.
Journal of Sex & Marital Therapy, 24, 103–122.
Schneider, D., Sledge, P., Shuchter, S., & Zisook, S.
(1996). Dating and remarriage over the rst two
years of widowhood. Annals of Clinical Psychia-
try, 8, 51–57.
Shafer, K. (2013). Disentangling the relationship
between age and marital history in age-assortative
mating. Marriage & Family Review, 49, 83–114.
Sharp, E. A., & Ganong, L. (2011). “I’m a loser, I’m
not married, let’s just all look at me”: Ever-single
women’s perceptions of their social environment.
Journal of Family Issues, 32, 956–980.
Shehan, C. L., Berardo, D. H., Vera, H., & Carley,
S. M. (1991). Women in age-discrepant marriages.
Journal of Family Issues, 12, 291–305.
Simons, L. G., Burt, C. H., & Peterson, F. R. (2009).
The effect of religion on risky sexual behavior
among college students. Deviant Behavior, 30,
467–485. doi:10.1080/01639620802296279
South, S. J. (1991). Sociodemographic differentials in
mate selection preferences. Journal of Marriage
and the Family, 53, 928–940.
Tally, M. (2006). “She doesn’t let age dene her”:
Sexuality and motherhood in recent “middle-aged
chick icks.” Sexuality & Culture, 10, 33–55.
Uecker, J. E. (2008). Religion, pledging, and the pre-
marital sexual behavior of married young adults.
Journal of Marriage and Family, 70, 728–744.
doi:10.1111/j.1741-3737.2008.00517.x
U.S. Bureau of the Census. (2000). Statistical abstract
of the United States: 2000. Washington, DC: U.S.
Government Printing Ofce.
U.S. Bureau of the Census. (2002a). Current Popula-
tion Survey: Number, timing, and duration of mar-
riages and divorces 1996. Washington, DC: U.S.
Government Printing Ofce.
U.S. Bureau of the Census. (2002b). Race and His-
panic or Latino origin by age and sex for the
United States: 2000. Washington, DC: U.S. Gov-
ernment Printing Ofce.
U.S. Bureau of the Census. (2003). Census 2000 brief:
Marital status. Washington, DC: U.S. Government
Printing Ofce.
U.S. Bureau of the Census. (2012). Current Popula-
tion Survey: Annual social and economic supple-
ment. Washington, DC: U.S. Government Printing
Ofce.
Vanier Institute of the Family. (2010). Families count:
Proling Canada’s families IV. Retrieved from
http://site.ebrary.com/id/10425582
Vera, H., Berardo, D. H., & Berardo, F. M. (1985).
Age heterogamy in marriage. Journal of Marriage
and the Family, 47, 553–566.
Weitz, R. (2010). Changing the scripts: Midlife
women’s sexuality in contemporary U.S. lm.
Sexuality & Culture, 14, 17–32.
White, H. (1980). A heteroscedasticity-consistent
covariance matrix and a direct test for het-
eroscedasticity. Econometrica, 48, 817–838.
Wolf, N. (1991). The beauty myth: How images
of beauty are used against women.NewYork:
William Morrow.
