Sex and the Tales We Tell

At a time when countless people feel compelled to share their intimate lives with the world, and when the means for doing so are becoming less expensive and easier to use, it is nice to know that people the world over still lie about sex.

In survey after survey within country after country men report having more heterosexual partners over their lifetime than women do, and as this article and this clarification point out, what people say in these surveys cannot be a reflection of what they do.

If these surveys were representative, we would expect that the average number of heterosexual partners reported by men in each sample to approximate the average number reported by women. Instead, the numbers aren’t even close. In Britain men report having an average of 12.7 hetrosexual partners over a lifetime whereas women report an average of 6.5; in France men report an average of 11.6 heterosexual partners, women 4.4; and in Germany men say 15.5, women 10.1.

In the US men report a median of 7 heterosexual partners and women 4, whereas Durex, the prophylactics manufacturer, plays safe in its 2005 global sex survey by reporting both that the average number of (heterosexual?) partners worldwide is 9, and that the global average for men is 10.2 and that the global average for women is 6.9.

Although dubious, the statistics are striking. From rigorous studies to pulp-weekly polls, swinging Berlin to dour Bethesda, straight men and women invariably talk one way yet walk another. This invites explanation.

The French Agence nationale de recherches sur le sida (ANRS) researchers speculate that the reason for the difference in their data is that French women don’t count one-night stands, whereas French men count everything.

Norman Brown, a research psychologist, conjectures that the difference is not due to females deliberately under-counting, nor even to men deliberately over-counting, but instead is partly due to two different methods used to estimate the number of partners. On Brown’s view women are likely to rely on enumerating the names of their lovers to produce their estimates, which typically leads to an underestimated count, whereas men are twice as likely to use a “rough approximation” strategy, which yields an overestimated count.

Unfortunately, the data Brown and his colleagues used to test the hypotheses were from a web-based survey of American adults—hardly a representative sample. (Women, 8.6. Men? 31.9.)

But the interesting feature of Brown’s study was that he also asked the respondents to rate the truthfulness of their estimates. He found that 5% of men and 4% of women indicated that they thought their estimates were inaccurate, and 16% of men and 11% of women indicated that they knowingly misrepresented their counts. Still, even when these “self-incriminators” were removed from the sample population, there was still a significant discrepancy between the counts for men and women.

Brown and his colleagues then concluded that “bad faith” was one component to the bias in the data and “cognitive factors”, such as these two estimating strategies, another. People on the whole want to be truthful about their sexual histories, Brown thinks, they just get ham-strung by faulty methods.

Perhaps. But here is another hypothesis: not everyone cops to a lie. And the reasons we have for lying about sex are as numerous as the reasons there are for having it in the first place. Until there is a method for estimating the number of sexual partners that people enjoy in a lifetime that does not rely upon the testimony of the actors, one that is as incontestable as counting the rings of a tree, we aren’t likely to have a clear idea just what the neighbors are up to, nor how often—which is precisely how it should be.


Sex and the Tales We Tell — 7 Comments

  1. Interestingly, I read a while back (alas I don’t remember where, but this was widely reported here in the UK) that when they asked men and women about their number of sexual partners while telling them that they were wired up to a lie detecter (they weren’t) women tended to give larger estimates than one would expect from previous studies, while the estimates from the men remained as one would expect. Of course, I haven’t seen the study itself, just how it was reported in the UK press, so it could well be that the samples used were far from representative. In any case, the strategy of telling someone that they are wired up to a lie detecter is an interesting way of trying to improve the odds that they will tell you the truth!

  2. A further possibility is that everyone is telling the truth, and that both men and women might be counting partners in the same way. This might come about as the result of factors in constructing surveys (and there is prostitution, I might add).

    Consider a village with 100 men and 100 women, married to eachother. All couples have sex, and no one had any sex at all before marrying. So right now we have an average of 1 sex partner for both sexes. But in addition to this, one of the women is a prostitute, and over the years does business with every man in the village. So now, the men have had 2 sexpartners on average, while 99 women have had 1 and the last woman has had 100 partners. The averages will still even out (as they must), but look at what happens if a survey is conducted.

    The respondents will be asked how many sex partners they have had, and let’s assume that all answer truthfully. The alternatives will typically look something like this:
    a. 1
    b. 2
    c. 3-5
    d. 5-10
    e. more than 10

    So 99 men will be responding “2”, 1 man will be ticking off “a”. 99 women will be responding “a”, and the last one will respond “e”. What will the people processing the survey do? They will have to put a numerical value to the number in “e” to get the numbers sorted out, and let’s assume that they settle for 20 (all they know is that it is more than 10, but they need some figure). This will give us the discrepant averages: Men wind up at an average of 1.99, and women at just below 1.2.

    No one lies, everyone understands the questions in the same way.


  3. Most researchers have considered ‘out of population’ explanations such as Fredrik’s — prostitution, sexual tourism, and the like. But this cannot account for the difference between the reports; there aren’t that many airplanes flying to loose capitals, nor do estimates of prostitution approach levels high enough to reconcile the accounts. There aren’t enough rock stars and escorts to make this type of explanation work.

    Lie detection is an option for reducing error in a sample, but I wonder then whether such samples would be representative. Would the general population submit to being asked about their sexual history while hooked up to a lie detector? While under oath? I have my doubts. And ethics boards would (or should) curb the over-zealous researcher, if not the over-zealous special prosecutor.

  4. I’ve always found it interesting the surveys like this are used as evidence for alleged evolutionary differences in the propensity of the genders to have monogamous relationships. The math seems obvious and yet people often assume that the data are to be taken at face value. (I suppose I really should have some citations for such claims, but off the top of my head I can’t give a good citation.) I’m not doubting evolutionary explanations of the traits we actually have, but arguing that we have a trait because there would be an evolutionary explanation of it if we had it, is not good methodology.

  5. There are a number of reasons why I like this story. First, it is a hard methodological nut to crack, and people have been at it for a couple generations now without much improvement. The topic does not readily lend itself to concealment—e.g., subjects asked to take a survey on their eating habits, then slipped questions about their sexual habits. They tend to notice. But there may yet be a clever researcher who can get at this through testimony.

    Second, the picture we have of “normality” is more obscure than I think we might otherwise imagine it to be. I suppose that health policy would be improved by a clearer picture of sexual habits, but I like the fact that there is resistance in populations to giving up the truth. There is another good that the health considerations are up against.

    That consideration is that we (in the west) are currently experimenting with privacy, or giving up on it. But public talk-show confessionals and the idea (if not the execution) of “lifecasting” on the web may well be a trend rather than a trajectory. There is a recognized trade-off between “news-worthiness” and “privacy concerns” that seems to change with taste, although in Europe my understanding is that there is a firmer basis in constitutional law for protecting individual privacy, whereas privacy considerations have less basis in US law, except for a patchwork of legislation covering financial data, medical records, and so on.

    Some think that the new technologies will obliterate private life, and perhaps they are right. But what strikes me about these data is that it appears to me that people have a strong expectation of privacy about their sex lives, even in the most open and liberal of societies, and they will resist giving inquiring people the truth.

  6. Pingback: » Lies About Sex

  7. Pingback: - Menn har mer sex enn kvinner

Leave a Reply

Your email address will not be published. Required fields are marked *