Gender Differences and Casual Sex: The New Research
Many of my readers will likely have seen the report of the new paper by Professor Terri D. Conley of the University of Michigan. I’m often skeptical of science reporting, so I didn’t want to write about it until I saw the paper for myself. I’ve now read it, and it deserves a complete run-though. (I don’t have the right to redistribute it, so I’ll refer to and sometimes quote the paper. You can buy it here. Folks with access to a library may be able to get it free, it is Journal of Personality and Social Psychology, 2011, vol. 100, no.2, 309-329.) It’s not really one study, it’s four, with subparts. It strongly contradict the theory that pop-evo-psych spouting people think of (though they often don’t know what it’s called), the Sexual Strategies Theory. And, for once the science reporting wasn’t awful, and the paper really did conclude what the press report says it does: that most of the gender difference in women’s and men’s propensity to agree to a broad-daylight, out-of-nowhere proposition for casual sex is driven by women’s perception that their risks are higher, and their likely enjoyment is lower from the proposer.
First of all, the backdrop of the whole study is another famous psychology paper, Clark and Hatfield’s 1989 paper, Gender Differences in Receptivity to Sexual Offers, Journal of Psychology and Human Sexuality, 2, 39-55. That paper studied receptivity to one, really specific and unusual, sexual offer: a stranger walks up to the subject, on a college campus, and propositions sex. All the women declined while a lot of men accepted, and this is a replicable result. Conley calls that scenario the CHSP, for Clark-Hatfield Sexual Proposal. You might think that this not necessarily representative of other propositions for casual sex — and you’d be right. After reviewing her results, Conley concludes:
Based on a number of findings from the current studies, it appears that the Clark and Hatfield paradigm is a casual sexual proposal that is uniquely repulsive to women being approached for heterosexual encounters, likely because of what it conveys about the male proposer’s sexual capabilities and safety.
[p. 324, emphasis mine.] But we’ll get back to that.
Conley’s paper isn’t just based on one study; she really did four, some with subparts. To do what she needed to do, instead of having people walk up to college students on campus like the original CHSP, she asked about the scenarios in questionnaires. One might think that this would affect the results, but in fact the numbers for the iterations that replicated the original CHSP scenario matched their results pretty well, indicating that what people said they’d do was a good indicator of what they in fact do when they don’t know it’s a study.
The basic hypothesis she set out to test was that the original CHSP didn’t really see which respondents were more inclined to casual sex, because an offer of casual sex with a man, for a het woman, is not the mirror an offer of casual sex from a woman for a het man. She hypothesized that men and women were viewed differently by the respective respondents in ways that made an offer of casual sex from a woman more attractive to people who like sex with women, than an offer of casual sex with a man was for people who have sex with men.
Sub-Studies and Results
Just for space reasons I’m going to do this more briefly than Conley’s paper does, but to do this right I have to briefly describe each study separately and state the key results. This will take a bit. Most of the analysis will follow.
Study 1a: this study was a pencil-and-paper version of the CSHP, but since it was on paper, Conley could also ask questions about how the participants viewed the proposer — not an actual person, as in the original, but the person the participants are asked to imagine for the sake of the question. The basic scenario was this:
An attractive member of the opposite sex approaches you on campus and says, “I have been noticing you around campus and I find you to be very attractive. Would you go to bed with me tonight?”
Respondents answered on a seven point scale from one (no) to seven (absolutely) the likelihood that they’d accept. The also answered similar seven point scales (many of which were created by combining multiple questions) for several other items, among them:
likelihood of having an STD
likelihood of giving gifts.
The thing to bear in mind is that there is no specific proposer, the proposer is just a construct. Any cues the respondents are taking about the proposer are in the respondents’ heads. The respondents have no direct information about whether the proposer has an STD, etc. All they know is what they infer about a person of the defined gender category making the offer described.
These scales served different purposes. Some were designed to determine how much pleasure the respondents thought they would get from agreeing to the proposal; some were designed to see how risky they thought it was; and some were designed to test the Sexual Strategies Theory (SST). In later studies, Conley manipulated the scenario to examine what happened with changes in these factors.
In addition, Conley asked about the desire for the respondent to have a short-term relationship, as distinct from casual sex, with the proposer.
The sample group for Study 1 was 516 participants, college population, 62% women, 58% European American, 14% African-American, 16% Latin@, 6% Asian American, het only.
Study 1a results: Conley summarizes the 1a results like this:
male proposers were perceived (by women) as more dangerous and less likely to provide them sexual satisfaction than women were perceived (by men). Male proposers were perceived (by women) to have lower status and to be less warm than the women proposers were perceived to be (by men). There were no gender differences in perceptions of the proposers’ sexual faithfulness, mental capacities, gift giving, or risk of STD.
[314.] On the key question of likelihood of accepting the offer, women’s mean was 1.37 out of 7, with a standard deviation of .97 – generally not a chance with a smattering of not likely. Men were at 3.74 with an SD of 2.16, so in the middle and all over the range. Everyone’s interest went up for the short fling, women to 2.43 (deviation 1.18), men to 4.03 (2.16), but the gap remained large.
As for why, here’s the first headline: women rated sexual competence at 2.82 (1.25) and men at 3.83 (1.14). Men thought the proposer — knowing nothing about her — would be a middle-of-the-pack sex partner, while women thought the male proposers would be mediocre. Here’s the second headline: women rated danger at 4.19 (1.62) against 2.75 (1.52) for men. Women rated danger in the top half of the scale, men in much lower, when all they knew about the proposer was the gender and that they had made the offer described. There are differences on other scales, some statistically significant at some levels but none with a gulf like these.
Study 1b: Conley used basically the same questionnaire on a different sample group, but instead of specifying a member fo the opposite sex, she specified the sex and handed them randomly, i.e. women were as likely to get a questionnaire where the proposition was from a woman as from a man. the sample was n=212, college population, women-heavy at 78%, 37% Latin@, 24% European American, 8% African-American, 11% Asian American and 8% Arabic or middle-eastern, het only.
Results of Study 1b: This is where it gets interesting.
Women were equally as likely to agree to a sexual encounter with a man as with a woman, t(163) = 0.10, n.s., d = 0.02. Given that the means were extremely low, however, the complete lack of interest in either the female or the male proposer may be chalked up to floor effects. That led me to additionally consider the question of short-term relationships. When women responded about their desire for the proposer as a short-term relationship partner, the means across conditions were higher (just over 2.0), but the women still showed no preference for the male versus the female partner, t(163) = 0.43, n.s., d = 0.07. In other words, heterosexual women were just as likely to want a short-term relationship with the female proposer as they were to want a short-term relationship with the male proposer. Even though agreeing to have sex with a woman would mean engaging in an encounter with someone to whom they are presumably unattracted physically, this was equally as unappealing as having sex with a man who approached them to propose casual sex. These findings show just how undesirable the male proposer is to the female participants, a question I explore further below.
[315.] Holy shit! Het men, as one might expect in a sample of het folks, were mostly uninterested in men — 1.5o on a seven point scale, and interested in women, 3.52. Het women, however, were at 1.16 for women and 1.15 for men, basically a tie at “not so much”. Maybe that’s so close to zero that it doesn’t mean much, but for the short fling, the pattern held. The men were at 3.59 (2.35) for men and 1.76 (1.56) for men; women were at 2.14 (1.76) for women, and 2.20 (1.78) for men — still just about a dead heat. Trying to walk over and pick up a strange female college student who identifies as het will be as likely to succeed for a woman as it will for a man.
The proponents of “inherent female bisexuality” can rejoice at an out-of-context data point to support their bullshit theory. (I’m not really kidding.)
Now this, this requires explanation. Here’s the explanation:
both women and men agreed that the female proposer would be better in bed, thought the female proposer was warmer and had higher status, and thought the female proposer would be more likely than the male proposers to give them gifts. Men and women also believed that female proposers were less likely to be dangerous than male proposers. In sum, both men and women agreed that the male proposers are less desirable than female proposers on dimensions of relevance to sexual encounters.
[315.] So, says the sample group, if a woman walks up to you and proposes a random quickie, she’s likely more fun and less worry than a man who does exactly the same thing.
Study 1c, and results: Conley recast the question into the third person, so that the respondent wasn’t the person being propositioned. She asked a sample ( n=62, 66% women, 66% European American 17% African American, 8% Latin@, 5% Asian American) about their perceptions of either male or female proposers. They rated the men (knowing nothing about them but that they were men and made the proposal) as being less sexually capable and more dangerous than the women.
Study 1d, and results: Conley got a sample of bisexual women from GLBT online and campus communities (n= 103, 77% European American, 8% Asian American, 5% Latin@, 2% African American). They answered proposition part of the 1a questionnaire, with either a man or a woman specified. The bisexual women respondents responded to the offer from a woman at 2.37 (1.41) and a man at 1.39 (.83). That’s a pretty good spread. they were not asked the perception of the proposer questions.
Study 2a: Conley substituted famous people for the strangers, on the hypothesis that famous people are more familiar and therefore perceived as less dangerous (whether accurately or not). She selected the famous people by asking sample groups to name attractive and unattractive members of the opposite sex, and then picked a few that were common. (She ended up dropping African American women from one group because their attractive celebrities did not overlap at all with the other, mostly white, women in the sample and they didn’t rate the men the white women liked as attractive. That’s a book of it’s own, or several books.)
Based on pilot questions to the sample, Conley picked Johnny Depp as the attractive and Donald Trump as the unattractive man, and Angelina Jolie as the attractive and Roseanne Barr as the unattractive woman. The respondents asked about the scenario with these people substituted as the proposing members of the opposite sex. Because she wanted to pull apart the effect of Roseanne’s perceived unattractiveness from the effect of her age, she added in Christy Brinkley for the male respondents. Brinkley was actually perceived as the older of the two.
Here’s how the Study 2a scenario went (using Johnny Depp as the celebrity):
You are fortunate enough to be able to spend your entire winter vacation in Los Angeles. One day, about a week into your stay, you decide to visit a trendy cafe´ in Malibu that overlooks the ocean. As you are sipping your drink, you look over and notice that actor Johnny Depp is just a few tables away. You can hardly believe your eyes! Still more amazing, he catches your eye and then approaches you. He says, “I have been noticing you and I find you to be very attractive. Would you go to bed with me tonight?”
[318.] Conley also reformulated the scenario slightly differently to involve an unknown proposer. She included the questions about a short fling. And she added a question about how appealing the offer was, even if they were going to say no.
Study 2a Results: Here’s the headline — differences between men and women in likelihood of taking the proposer up on the offer was a whole lot closer. For the proposition by the attractive person, women were at 4.09 (2.16) to 4.16 (2.56) for men — just about a tie. For the unattractive celebrity, men were at 1.43 (.84) to women’s 1.71 (1.61) — women were higher. For the unknown person, though, no such effect. Women were at 1.86 (1.38), men were still at 3.52 (2.06). Women were only marginally more interested in the offer from a stranger than from a man generally thought ugly. Men were almost as interested in the random stranger as Angelina Jolie. The short fling results basically track this, with the fact of celebrity seriously closing the gap between men’s and women’s interest, and the gap for a stranger remaining wide. The appeal of the offer follows the same pattern: little difference in men’s and women’s response to the unattractive celebrity, little difference in their reaction to the attractive celebrity, lots of difference in their response to the stranger.
Because this study included factors like resources, status and age, it has a lot of implication for SST, which I’ll address below.
Study 2b and Results: This study mapped 2a, in a different location and with a different sample group, and different celebrities created by a pilot question process. The sample was from a university (n=118, 59% women, 20% African American, 27% Asian American, 21% European American, 16% Latin@). The celebrity women were Jennifer Lopez and, again, Roseanne. The men were Brad Pitt and Carrot Top. Conley again included the stranger question as a control. This time, Conley added in some other questions: “How likely is it that you would be harmed by this person?” and “How likely is it that this person would be a good lover?” These questions cut right to the heart of the risk and reward issues, and the responses are discussed later.
This study replicated the main result of 2a, that celebrity largely eliminates the gender difference in interest in casual sex. There is almost no difference for the unattractive celebrity, and the difference between men and women for the attractive celebrity was still present, but smaller. Those difference shrink even further for the fling. The differences for the stranger proposer, however, remain a gulf. In fact, men’s reactions to the unknown woman looked a lot like their reaction to Jennifer Lopez, and totally unlike their reaction to Roseanne. Women’s reactions to the stranger looked a lot like their reactions to Carrot Top, and not a lot like their reactions to Brad Pitt. Conley conjectures, “In sum, when women conceptualized the unknown proposer, they appeared to be thinking of someone like Carrot Top. When men conceptualized the unknown proposer, they appeared to be thinking of someone like Jennifer Lopez.” [p.320.] Maybe that tracks gendered differences in optimism and pessimism, or maybe it maps preconceived notions about the kind of men and women who propose sex out of the blue, or maybe it has to do with risk. There are a number of possible explanations.
Study 2c and Results: Conley substituted the respondents’ best opposite sex friend for the celebrity in studies 2a and 2b. The sample was n=109, 66% women and 82% European American, gathered online. They put in an initial for this friend, and got the following scenario:
During a conversation with your male [female] friend, he [she] says to you, “I have been noticing how attracted I am to you. Would you go to bed with me tonight?”
[321.] The respondents were also asked the characteristics of the proposer, e.g. sexual capability etc. Most of these characteristics showed no real gender difference; gift-giving varied slightly, as did warmth and sexual capabilities. Women thought their men friends were less sexually capable than men thought their women friends were.
Men were more likely to accept, 2.84 (1.92) versus 1.97 (1.61). It is interesting to me that the perception of danger was equal, though the prevalence of acquaintance rapes calls that perception into question. But the factor that drove the difference, in the absence of which the difference disappeared, was sexual capability. That was the only covariate that moved the needle.
Study 2d and Results: Conley collected a sample of gay men and lesbians online and through campus groups (n= 196, 47% women, 64% European American, 13% Asian American, 10% Latin@, 5% African American). She used the 1d materials, but giving men the male proposer scenario and women the female proposer scenario. There was no statistically significant difference in likelihood of taking the offer between men and women.
One thing that became clear is that the rates across the board for agreeing to sex under the classic CHSP scenario, a stranger propositioning cold in broad daylight, are really low. They’re low for every demographic group studied. Prior research shows that people comfortable with casual sex are much more common in gay male populations than most of the others, but the gay men’s response was still at a median 2.55 (1.62). Says Conley:
That is, even though gay men are attracted to men and, on average, relatively open to casual sex, they did not find the CHSP any more appealing than lesbians did. It seems unlikely that the gay male participants in this study were less open to casual sex than most samples of gay men. Rather, this finding suggests that the CHSP is an unusual and suspicion-arousing sexual proposal even for people who are clearly open to casual sex encounters in other contexts.
[312, emphasis mine.]
Study 3 and results: Study 3 isn’t really a separate study. It’s a study of separate data taken from the Study 1 questionnaires, an analysis of all the proposer factors like sexual capabilities and danger, warmth and gift-giving. Conley used hierarchical regression analyses and then a mediational analysis. I’m no statistician, so I’ll quote her reporting:
Beyond gender, however, only the perception that the proposer would be a good lover (consistent with pleasure theory) significantly influenced participants’ likelihood of agreeing to the sexual offer. People who perceived the proposer as a better lover were more likely to accept the offer. Warmth was a marginally significant predictor of acceptance of the sexual offer, with those who perceived the proposer to be warm being somewhat more likely to accept.
In addition, the interaction between gender and sexual faithfulness was significant. An analysis of simple slopes revealed that the perception that the proposer would be faithful was significantly more predictive of women’s than of men’s acceptance of the CHSP. This would seem to contradict SST, which predicts that men desire women who will be faithful to them so that they will have the greatest likelihood of propagating their genes. Women may desire a faithful partner because they believe that his sexual faithfulness will increase the likelihood that he will provide support for their future children, but SST still predicts that faithfulness would be relatively more important to men (who strive to assure that any offspring are genetically related to them).
Contrary to SST, no interactions between participant gender and status or gift giving emerged in the analysis.
[322-23, emphasis mine.] In the analysis of factors in desirability for a fling, there were gender differences in how much perceptions of danger, mental illness, and having an STD, affected desirability, as well as gender differences for sexual capability and gift giving. But the strongest predictor was sexual capability, both for the proposition and the short relationship.
Study 4 and Results: Having concluded that the classic CHSP scenario was not particularly attractive to anyone and was especially repulsive to women, Conley tried to construct a study of the more naturalistic scenario by asking respondents about an actual proposal for casual sex in their past. Her sample was online through experimenters’ social networking accounts and Craigslist ads (n=463, 67% women, 74% European American, 11% African American). The experimenters reviewed the answers to open ended questions to weed out any fake responses, but all appeared legitimate.
The first result is that, as Conley predicted, proposals in the real world were accepted more often than CHSP scenarios were. In the original paper, no woman accepted the CHSP. In Study 4, 40% of women accepted the proposal. Men accepted the offer 73% of the time.
I do see a selection problem with this, though. If people have received multiple propositions for casual sex before and they pick one to report for an experiment, I suspect they pick the most memorable or interesting experience, which will more often (though not always) be an offer they accepted rather than rejected. I think the opportunity for the respondent to select one experience multiplies the self-report concerns.
Conley asked about personal characteristics of the proposer — not this time a mental construct, but the actual person who proposed casual sex. Consistent with Conley’s findings in the studies of the imaginary proposer, women rated the actual male proposers differently than men rated women proposers. They rated the men more dangerous and more likely to have STDs, and the female proposers were rated more sexually capable, warmer and more sexually faithful.
Examining the effects of the variables on likelihood of acceptance, the big winner was sexual capability. And the big winner in interaction with gender was also sexual capability. Conley writes:
Thus, concerns about the opposite-sex proposer’s sexual abilities partially mediate the relationship between participant gender and acceptance of a casual sex offer both in hypothetical situations (Study 3) and in real-world situations (Study 4).
Conley’s Major Conclusions:
I’ll just quote at some length here:
First, male sexual proposers (who approached women) are uniformly seen as less desirable than female sexual proposers (who approached men). Therefore, gender differences in the original Clark and Hatfield study are due more to the gender of the proposer than to the gender of the study participants. Moreover, the idea that these gender differences reflect broad, evolved differences in women’s and men’s mating strategies was not supported. Across studies involving both actual and hypothetical sexual encounters, the only consistently significant predictor of acceptance of the sexual proposal, both for women and for men, was the perception that the proposer is sexually capable (i.e., would be “good in bed”). The perceptions of sexual capabilities also mediated the relationship between gender and acceptance of casual sex offers. Finally, indirect evidence suggests that perceptions of risk may play a role in gender differences in casual sex attitudes.
[325, emphasis mine.] Her theoretical take-away is this: that the results as a whole support Pleasure Theory, the theory that pleasure itself is evolutionarily selected and a search for pleasure is the immediate animating drive that causes people to make sexual choices. Women’s reluctance comparative to men to accept the CHSP wasn’t really a reluctance to have casual sex, but rather a response to a different offer than the men got — the didn’t think the men would be as much fun. Conley notes other research for the proposition that women may actually not enjoy casual sex that they do choose to have as much as men, for a variety of reasons.
Conley also notes that the findings are somewhat consistent with the theory that women are less interested in casual sex because they perceive greater risk:
These studies demonstrated results consistent with risk perception, as women (but not men) perceived less risk from the familiar individuals than from the stranger making the sexual proposal. When women were considering the less risky (i.e., familiar) proposers, they were just as likely to agree to the CHSP as men were (after accounting for perceptions of sexual capabilities in the case of the best friend proposing sex to them). Likewise, when nonheterosexual women considered proposals from members of their own gender, women were equally as likely to accept the casual sex offer as men were (i.e., lesbians in Study 2d) and were more likely to accept an offer from a woman than from a man (i.e., bisexual women in Study 1d).
[326, emphasis supplied.] Study 2, the celebrity and friend studies which introduce the element of familiarity to mitigate perceived risk, speak to this, though as I say below I suspect risk is a bigger variable than it apparent here for various reasons.
Sexual Strategies Theory did not fare well:
By contrast, this research demonstrated some of the limiting conditions of SST. Sexual strategies theory clearly predicts that higher status proposers should be accepted by women more readily than low-status proposers. The fact that status did not predict women’s acceptance of casual sex offers is therefore a problem for SST. Neither status, nor tendency for gift giving, nor perceived faithfulness of the proposer (nor, more precisely, the interaction of any of these variables with gender) predicted whether a participant would agree to the sexual offer, contradicting SST. Likewise, if men’s central goal, as suggested by SST, is to transfer their genetic material to future generations, men should have a greater base rate likelihood of accepting a sexual offer from any woman than women have of accepting a sexual offer from any man, regardless of the proposer’s attractiveness (i.e., women should be choosier than men). SST does not predict that women would be equally likely to accept offers as men when (a) the proposers are very attractive, (b) the proposers are very unattractive, (c) the proposers are familiar people, and (d) the proposer and the individual are of the same sex.
[326.] Also, men were very positively disposed to Angelina Jolie (past peak childbearing age) and Christy Brinkley (well past childbearing age), whom they thought attractive, but not interested in Roseanne, whom they thought unattractive. These are interests that can’t be squared with an evo-psych imperative to reproduce, but are perfectly consistent with an evolutionary imperative to seek pleasure. Likewise, Donald Trump and Carrot Top could each provide plenty of material support for offspring, so SST predicts that women would be interested. They were *ahem* not. There’s a whole section in the paper on possible responses to SST theorists trying to save their model. I found it persuasive.
My Additional Thoughts:
First, a note about the framework of the research. Mainstream research on gender differences is hugely cisnormative. The paper never even deals with the possibility that some men and some women in the respondent groups are transgendered, and never deals with the possibility that some people are outside the binary. The author did deal with the existence of GLB folks, and in fact included gay men, lesbians, and bisexual women in different aspects of the study, but only to illustrate points about male-female attraction in a heteronormative model. Large sample sizes might harder to get, but a more thoroughgoing look at responses to offers of casual sex across a range of gender identities and sexual orientations would be awesome, and I think really illuminating.
I have a gut feeling that this study didn’t really get to the root of the risk issues. I think, and this is a conjecture, that many women deal with risk in a noncognitive way and make decisions based on risk analysis that they don’t attribute to risk when consciously explaining their motivations. This is really, particularly an issue with the CHSP. Entirely coincidentally, while I was reading this paper, Jill wrote at Feministe about offers of casual sex from strangers in broad daylight:
Quit using that study where strangers walked up to people on college campuses and offered sex as “proof” that men desire sex more than women. Perhaps consider that women may want sex just as much, but have spent their entire lives hearing about how sex with strangers is a terrible, dangerous idea, leading to the (probably correct) understanding that the only kind of men who would approach you in broad daylight offering sex are men who are either serial killers or sex offenders or at least total fucking creeps?
Whatever the empirical truth is, this is a frank statement of risk perception, and I suspect an accurate one. Conley found that people — not just women — thought that the CHSP scenario (which Jill was clearly referencing) was “uniquely repulsive” and Jill says it communicates that the guy making the offer is some kind of creep. (There is extensive discussion, including my thoughts, on that Feministe thread.) I’ve heard these reactions before, where the time, place and manner of the offer convey a perception of the offeror — in this case, a very bad one.
My high-level takeaway, though, is that this is where men have the most work to do. We’re perceived terribly by women, including but not only our potential sex partners. This perception may be entirely based not on something we’ve done but things other men have done. On my account, though, it is based as much on the social structures we participate in as men, and the ways they operate in the culture. On my account, as long as there is a lot of rape and not a lot of remedy, as long as there is slut-shaming and double-standards, as long as the denial of the technologies women need to mitigate the risks of unintended pregnancy and disease, then they’re going to look askance at us, and they’re going to act like they have more risk and less to gain from sex with us, because in fact they do. The only way to make a lasting peace between men and women in to make a world where the advantages and disadvantages are more evenly distributed.
This is long even by my standards, so I’ll end there, through I’ll almost certainly have more thoughts on Conley’s fascinating research in future posts.