Sex on the Margin

Sexual preferences are primarily biological in origin.  But sexual choice is about preferences and constraints.  Raise the price of sex with women and more men will choose to have sex with other men – that’s what happens in prisons.

In a remarkable paper, Andrew Francis (a graduate student at the University of Chicago) examines how AIDS has changed sexual choice.  With admirable precision, Francis lays out the price of sex:

…it is thousands of times more likely that a male would get HIV having sex with a man than having sex with a woman.  In terms of AIDS-related mortality, the expected cost of having unprotected sex once with a man is almost $2000, while the expected cost of having unprotected sex once with a woman is less than a dollar.

Thus AIDS changes the price of sex, do we observe changes in choice?  Francis wants to be careful about causality so he uses a clever instrumental variables approach.  He reasons that knowledge of AIDS and thus responsiveness to price is correlated with knowing someone who has AIDS and that knowing someone who has AIDS is exogeneous to other factors influencing sexuality.  Unfortunately, it appears that he only has information on whether a relative has AIDS and genetic factors mean exogeneity is unlikely to hold.  In fact, we would probably expect that simply knowing someone with AIDS is positively correlated with being homosexual (especially in 1992 when the survey was taken).   

Indeed, Francis finds, as expected, that women who have a relative with AIDS are more likely to be engage in homosexual acts and identify as being homosexual.  But Francis finds that men who have a relative with AIDS are significantly less likely to:

…have had sex with a man during the last sexual event…have had a male sexual partner in the last year… say they are sexually attracted to men…rate having sex with someone of the same gender as appealing…[or] think of themselves as homosexual or bisexual.

The tendency to greater homosexuality among women and less among men is exactly what the economic theory predicts given how AIDS affects the price of sex.  Genetic and social factors will have greater difficulty resolving this bifurcation so I think Francis has the upper-hand on the argument, although there may be counter-arguments based on the gay-uncle theory).

Importantly, note also that Francis finds that not only is sexual choice malleable, as the prison story I opened with suggests, but so are sexual desire and identity.  At least on the margin!  (A point that non-economists are likely to miss.)

Thanks to Emily Oster for the pointer.