by Nicole Barbaro
I got to my office the other morning with the goal of organizing my last data set needed for my PhD dissertation. The study is a small survey data set of 52 heterosexual couples, mostly adults in their 20s, who reported on a variety of self-report sexuality and relationship measures — pretty standard psychological research.
My topical expertise in psychology centers primarily around human sexuality. After years of studying in this area, there are two observations about human sexuality that I have more or less taken for granted at this point:
(1) Men have more sex partners than women, on average. This means that if you approach a random man on the street and ask him how many people he’s had sex with, his answer is likely to be larger than a random woman’s answer to the same question. (I don’t recommend using this approach strategy, however.)
(2) The variance, or range, of the number of sex partners within the sexes is different, with men as a group having a greater range as compared to women as a group. This means that it is likely that the most sexually successful man you know has a much larger number of sex partners than the most sexually successful woman who you know.
After my data were organized, I ran a simple analysis comparing the mean average number of lifetime sex partners between men and women to see if these basic observations held. It was unsurprising to see that my data conformed to both points: In my sample, men had a higher mean and a larger variance compared to women in their response to the question, “How many partners have you had sexual intercourse with in your lifetime?”
I’ll tackle the easier to explain observation first: difference in variance. That men have a larger variance in number of sex partners is often deduced from what is called Bateman’s Principle. Bateman’s Principle has been around since 1948, and is the result of research on reproductive success, or number of offspring, in fruit flies (Drosophila). The study showed that male fruit flies have greater variance in the number of offspring compared to females.
Bateman’s principle holds in humans, too. It is most clearly demonstrated by looking at one of the most reproductively successful males in recorded human history who sired over one thousand children, and comparing him to the most reproductively successful woman in recorded human history who sired a measly 69 children in comparison.
Although Bateman’s principle is about offspring number, in the modern industrialized West researchers typically ask about number of sexual partners instead, because widely practiced birth control obviously affects the number of offspring men and women have — “counting babies” is not very informative in modern environemnts. The greater male variance in my data was what my tweet was pointing out. But, in the comments, someone asked why the means were different. Bateman’s Principle can pretty straightforwardly explain differences of variance, but not necessarily differences of means.
This really got me to thinking why, exactly, are the mean averages in number of sex partners between men and women different? My initial thought was, “well, because, they just are” — to which my next thought was, “that isn’t a good reason.” So, I decided to do an intellectual exercise to understand an observation about human sexuality that I had taken for granted.
The Mathematical Problem of Unequal Means
In theory, it is mathematically impossible for the mean average number of sex partners to be different for men and women. Take, for example, the population in the figure below of 25 males, orange, and 25 females, purple. I’ve recreated Bateman’s principle here, in that the range of partners for men (0–7) is greater than the rage of partners for women (1–3).
And, as mathematicians would predict, the mean average number of sex partners for men and women is exactly equal, while the range and variance, however, remain different. No matter how skewed the distributions of sex partners are within sexes, the mean average will always be equal.
So, yes, in theory, it is mathematically impossible for the mean average number of sex partners for men and women to be different.
But, although theories are useful, reality is messy.
To understand why scientific studies, like my own, do in fact show differences between men’s and women’s mean average number of sex partners, we can examine the assumptions, or what we take for granted to be true, of the mathematical theory described above.
(1) The model assumes a closed population — that all men and all women are accounted for, the whole population is sampled, no one has died or run off to a place unknown, and no other normative dynamics of human populations are observed. Pretty unrealistic.
(2) Everyone’s reports of their sexual histories are 100% accurate. Also, pretty unrealistic.
(3) The model assumes heterosexual sex only — no one is having sex with a same-sex partner. Again, unrealistic.
The fact that reality clearly violates each of these assumptions should be obvious and unsurprising. But it is important to understand that each of these violations are, to an extent, operating in each scientific study on the topic, including my own.
Let’s start with the closed population assumption. No scientific study can ever sample their entire target population (which is why it is called a “sample”). Thus, even in my sample of heterosexual couples, men’s and women’s mean average number of sex partners is unequal. Why? Because maybe John has had 15 female sex partners, but the only one of his sex partners who is included in my female sample is his current girlfriend. The rest of John’s sex partners are not accounted for in my study. The violation of the closed population assumption is present in any scientific study that reports these data.
Taking a random sample of the population in the above figure, analogous to what researchers do when collecting data, we see that the mean averages are now unequal between men and women. Men, specifically, have a higher mean average number of sex partners than do women. This is what scientific studies are doing: taking a random sample of the population.
Even if we accept that the closed population assumption violated, why is it more likely that men’s mean average number of sex partners is higher than women’s? Here is where Bateman’s Principle returns. Because the range and variance of sex partners differs for men and women, and because it is more likely that a man, as compared to a woman, is going to have an extremely large number of sex partners, it is therefore more likely that a random sample in a study will contain a handful of men reporting very high numbers, thus, pulling the men’s mean average higher as a group. In the example above, the man with seven sex partners was sampled, thus driving the men’s group mean higher.
Here is what my dissertation data look like (the same data from my original tweet). See that in this sample, there is one man reporting 70 sex partners. The highest number of sex partners reported by a woman was 19.
The fact that a few men with very active sex lives make it into university-run research is likely due to many factors, such as universities sampling from mostly urban populations, adverts for sexuality studies likely attracting more open individuals, and the like. This is a topic for another post.
Moving on to reporting accuracy. This should be no surprise, but people lie about their sex lives; or, at least, they’re not very accurate when reporting on them. Research has shown that men and women think about their response to the question “how many sex partners have you had?” differently. Men are more likely to just give a ballpark estimate, resulting in greater error in men’s reports as a group. Women, on the other hand, are more likely to think carefully and actually count each person they had had sex with. So, even with a closed population, inaccuracy of reports will cause error, and men are more likely to have error in their reports, which includes enough cases of slightly rounding up to cause a difference (“yea, around 20 for sure”).
Finally, the mathematical model assumes only heterosexual sex. Men and women who report being “exclusively heterosexual” will sometimes have sex with a same-sex partner. Look carefully at my own question that I asked my participants. I did not specify heterosexual sex, or opposite sex partner (in retrospect, oops). Although all my participants were currently in a heterosexual relationship, they were not required to identify as exclusively heterosexual in orientation, nor was previous same-sex sexual activity grounds for exclusion. Presumably some people’s reports may have included same-sex sex partners, also biasing the mean averages.
Finally, even if — and it’s a really big if — reality conformed to all of the assumptions of the mathematical theory, the modal number of sex partners would be higher for men as compared to women, again because of Bateman’s Principle. We can see this in my fake data I created with the theoretically perfect population: men and women have an equal mean, but men’s mode is higher (you can find details of this reasoning here).
So, in reality, yes, men do tend to have more sex partners than do women.
In theory, men have more sex partners than women when mode is examined. In reality, men have more sex partners than women when mean is examined. In both theory and reality, men as a group have a greater range and variance of number of sex partners than do women. At the most basic level, the reasons boil down to Bateman’s Principle, both in theory and in reality.