Savvy Shoppers Seeking Sex

There exists an idea in the economic field known as revealed preferences theory. People are often said to have preferences for this or that, but preferences are not the kind of thing that can be directly observed (just as much of our psychology cannot). As such, you need to find a way to infer information about these underlying preferences through something observable. In the case of revealed preferences, the general idea is that people’s decisions about what to buy and how much to spend are capable of revealing that information. For instance, if you would rather buy a Honda instead of a Ford for the same price, I have learned that your preferences – at least in the current moment – favor Hondas; if I were interested in determining the degree of that preference, I could see how much more you were willing to pay for the Honda. There are some criticisms of this approach – such as the the issue that people sometimes prefer A to B when compared to each other directly, but prefer B to A when presented with a third, irrelevant option – but the general principle behind it seems sound: people’s willingness to purchase goods and services positively correlates with their desires, despite some peculiarities. The more someone is willing to pay for something, the more valuable they perceive it to be.

“Marrying you is worth about $1,500 to me”

Now this is by no means groundbreaking information; it’s a facet of our psychology we are all already intimately familiar with. It does, however, yield an interesting method for examining people’s mating preferences when it’s turned on prostitution. In this case, a new paper by Sohn (2016) sought to examine how well men’s self-reported mating preferences for youthful partners were reflected in the prostitution market, where encounters are often short in duration, fairly anonymous, and people can seek out what they’re interested in, so long as they can afford it. It is worth mentioning at the outset that seeking youth per se is not exactly valuable in the adaptive sense of the word; instead, youth is valued (at least in humans) because of how it relates to both reproductive potential and fertility. Reproductive potential refers to how many expected years of future reproduction a woman has remaining before she reaches menopause and loses that capability. As such, this value is highest around the time she reaches menarche (signaling the onset of her reproductive ability) in her mid-teens and decreases over time until it reaches zero at menopause. Fertility, by contrast, refers to a woman’s likelihood of successful conception following intercourse, and tends to peak around her early twenties, being lower both prior to and after that point.

Since the type of intercourse sought by men visiting prostitutes is usually short-term in nature, we ought to expect the male preference for traits that cue high fertility to be revealed by the relative price they’re willing to pay for sex with women displaying them (since short-term encounters are typically aimed at immediate successful reproduction, rather than monopolizing a woman’s reproductive potential in the future). As such fertility cues tend to peak at the same ages as fertility itself, we would predict that women in their early twenties should command the highest price on the sexual market price, and this value should decline as women get older or younger. There are some issues with studying the subject matter, of course: sex with minors – much like prostitution in general – is often subject to social and legal sanctions. While the former issue cannot (and, really, should not) be skirted, the latter issue can be. One way of getting around the legal sanctions of prostitution in general is to study it in areas in the world where it is legal. In this instance, Sohn (2016) reports on a data set derived from approximately 8,600 prostitutes in Indonesia, ranging from ages 17-40, where, we are told, prostitution is quasi-legal.

The variable of interest in this data set concerns how much money the prostitutes received during their last act of commercial sex. This single-act method was employed in the hopes of minimizing any kinds of reporting inaccuracies that might come with trying to estimate how much money is being earned on average over long periods of time. While this choice necessarily limits the scope of the emerging picture concerning the price of sex, I believe it to be a justifiable one. Age was the primary predictor of this sex-related income, but a number of other variables were included in the analysis, such as frequency of condom use, years of schooling, age of first sex, and time selling sex. Overall, these predictor variables were able to account for over half of the variance in the price of sex, which is quite good.

“Priced to move!”

Supporting the hypothesis that men really do value these cues of fertility, the price of sex nominally rose from age 17 until it peaked at 21 (though this rise was not too appreciable), tracking fertility, rather than reproductive potential. Following that peak, the price of sex began to quickly and continuously decline through age 40, though the decline slowed passed 30. Descriptively, the price of sex at its minimum value was only about half the price of sex at peak fertility (which is a helpful tip for all you bargain-seekers out there…). Indeed, when age alone was considered, each additional year reduced the price of sex, on average, by about 4.5%; the size of that decrease uniquely attributable to age was reduced to about 2% per year when other factors were added into the equation, but both numbers tell the same story. A more detailed examination of this decrease grouped women into blocks of 5-year age periods. When considering age alone, there was no statistical difference between women in the 17-19 and 20-25 range. After that period, however, differences emerged: those in the 26-30 range earned 22% less, on average; a figure which fell to 42% less in the 30-34 group, and about 53% in the the 35-40 group.

This decrease in the price of sex over a woman’s lifespan is the opposite of how income usually works in non-sexual careers, where income rises with time and experience. It would be quite strange to work at a job where you saw your pay get cut by 2% each year you were with the company. It is likely for this reason that prostitutes in the 20-25 range were the most common (representing 32.6% of the sample), and those in older age groups were represented less heavily (27.6% in the 26-30 group, all the way down to 12% in the 35-40 range). When shopping for sex, then, men were not necessarily seeking the most experienced candidate for the position(s), but rather the most fertile one. As fertility declined, so too did the price. As price declined, women tended to leave the market. 

There were a few other findings of note, though the ‘whys’ explaining them are less straightforward. First, more educated prostitutes commanded a higher average asking price than their less educated peers, to the tune of about a 5% increase in price per extra year of school. As men and women both value intelligence highly in long-term partners, it is possible that cues of intelligence remain attractive, even in short-term contexts. Second, controlling for age, each year of selling sex tended to decrease the average price by about 1.5%. It is possible that the effects of prostitution visibly wear down the cues that men find appealing over time. Third, prostitutes who had ever used drugs or drank alcohol earned 12% more than their peers who abstained. Though I don’t know precisely why, it’s unlikely a coincidence that moral views about recreational drug use happen to be well predicted by views about the acceptability of casual sex (data from OKCupid, for instance, tells us the single best predictor of a woman’s interest in casual sex is whether she enjoys the taste of beer). Finally, prostitutes who proposed using condoms more often earned about 10% more than those who never did. I agree with Sohn’s (2016) assessment that this probably has to do with more desirable prostitutes being attractive enough to effectively bargain for condom use, whereas less attractive women compromise there in order to bring in clients. While men prefer sex without condoms, they appear willing to put that preference aside in the face of an attractive-enough prospect.  

“Disappointment now sold in bulk”

So what has been revealed about men’s preferences for sex with these data? Unfortunately, interpretation of prices is less straightforward than simply examining the raw numbers: their correspondence to other sources of data and theory should be considered. For instance, at least when seeking short term encounters, men seem to value fertility highly, and are willing to pay a premium to get it. This “real world” data accords well with the self-reports of men in survey and laboratory settings and, as such, seems to be easily interpretable. On other hand, men usually prefer sex without condoms, so the price premium among prostitutes who always suggest they be used would seem to, at face value, ‘reveal’ the wrong preference. Instead, it is more likely that prostitutes who already command a high price are capable of bargaining effectively for their use. In order to test such an explanation, you would need to pit the prospect of sex with the same prostitute with and without a condom against each other, both at the same price. Further, more educated prostitutes seemed to command a higher price on the sexual market: is this because men value intelligence in short-term encounters, educated women are more effective at bargaining, intelligence correlates with other cues of fertility or developmental stability (and thus attractiveness), or because of some other alternative? While one needs to step outside the raw pricing data obtained from these naturalistic observations to answer such questions effectively, the idea of using price data in general seems like a valuable method of analysis; whether it is more accurate, or a “truer” representation of our preferences than our responses to surveys is debatable but, thankfully, this need not be an either/or type of analysis.

References: Sohn, K. (2016). Men’s revealed preferences regarding women’s ages: Evidence from prostitution. Evolution & Human Behavior, DOI: 

4 comments on “Savvy Shoppers Seeking Sex

  1. Starman on said:

    It would be interesting to get statistics on homophile prostitutes too. There seems to be two basic types of homophile, the masculine and the feminine. The masculine homophiles show a preference for the teenage twinks analogous to the preference for teenage girls in normal men while the feminine homophiles seem to prefer the older, rugged bears like women do.

  2. Starman on said:

    You know I’m right, don’t you? Men are tracking cues of high RV more than cues of high fertility….right?

  3. Starman on said:

    There’s also theoretical problems with the idea that men would prefer prostitutes at peak fertility. Though would be rational and adaptive to prefer prostitutes at peak fertility men probably haven’t evolved adaptations to do this since the modern systems of prostitution didn’t exist in small, close-nit prehistoric societies (it’s also unclear how common short term relationships were). The closest thing to the problem of prostitution choice was probably the problem of wife choice. When a man has to choose from a selection of prostitutes I imagine he uses the adaptations evolved for choosing wives and preferences the prostitutes that show the physical cues of high future fertility: youth and no signs of prior pregnancies.