Women have strong racial preferences over their partners

One of the things that interests me is what happens when there are systematic constraints placed on men and women's preferences and options when choosing a partner. For instance, what's the M-optimal allocation when women vastly outnumber men? The Gale-Shapley algorithm will simply lead to more women unmatched and men matching higher up their preference order, implying women matching lower down their preference ordering, assuming monogamous matching. Charles and Luoh find evidence for this in a working paper in which rising Black male incarceration shows Black female marriage falling, but also a greater spread in the kinds of matches Black women do make when they do get married. Specifically, you see more Black women "marrying down" and Black men "marrying up" as the incarceration rate rises, which is consistent with the Gale-Shapley (i.e., deferred acceptance) algorithm.

When you present this finding to skeptical parties, you will nearly always get this question, though: "Why don't Black women simply marry White men?" That is to say, isn't it somewhat ad hoc to say that women limit their search to men of the same race, when there are plenty of available men of other races willing presumably to match with them? Generally, what I have said to such questions is to remind the questioner that this is a two-sided matching problem, and it's not enough simply that Black women desire to match with (for example) a White man when faced with a deficit of Black men. For a match to form, the White man must ask the Black women for her hand in marriage (in the Gale-Shapley world) for the match to prefer. So you can imagine that obstacles keeping White men from asking Black women, or even simply systematic same-race preferences among White men combined with a balanced sex ratio which keep most White men asking White women for marriage. For one, if White sex ratios are balanced, and White men generally prefer marrying White women to Black women (even if only weakly prefer, I suspect), then the final allocation will be one in which we can say the "best" White men will have their offers accepted by White women, and by the time White men are asking Black women for marriage, the only ones left will be the least preferred White men from the two-sided White market. That therefore means that Black women will be receiving offers from the least preferred White men, if they receive any offers at all from that direction.

But new research also suggests that there may be obstacles keeping those matches from forming that originate on the female-side of the equation. In a forthcoming Review of Economic Studies article entitled "Racial Preferences in Dating", Ray Fisman, Sheen Iyengar, Emir Kamenica and Itamar Simonson report from an interesting exeriment replicated over many years at Columbia University that women exhibit strong same-race preferences in their matches. This same-race preference is even stronger than what they observe among males. Interestingly, one of the strongest predictors of those preferences is the racial composition of the zip where the participant grew up, and the prevailing racial attitudes in that person's state or country of origin. That does not explain why the mechanism would only operate on female racial preferences, though. That is an interesting question in and of itself.

If Black women have strong same-race preferences, then this could explain many features you observe in Black marriages. You see, for instance, a higher incidence of interracial marriage among Black males than Black females. You also will observe a higher incidence of singleness among Black females. It might also feed into the larger growing literature that studies the relationship between risky sexual behavior among Black men and the sex ratio they face. If Black women have strong same-race preferences, then this only works to increase the Black males' bargaining position, since he can now credibly threaten to dissolve the relationship whereas the Black female cannot since she has closed herself off from matching interracially. Assuming the Black male with whom she is matched knows she is not interested in non-Black men, her threat position is significantly weaker than it would be if she was considerate of non-Black men.

One more piece of evidence that makes me depressed about the problems and incentives facing Black communities, and particularly Black women.