Why Does Race Matter for Women?

One social science finding which I’ve wondered about over the past few years is the result that women care much more about the race of a potential mate than men do. The fact that individuals tend to want to mate assortatively with those who share their characteristics is no surprise. Rather, what does surprise are a series of papers that show a very strong asymmetry in strength of preference between males and females. To be crass about it, an attractive warm body will do for a man, but women strongly prefer a body with the packaging of their own race!

First, let’s keep this in perspective, here are the correlations from the GSS for married individuals for several variables of note (I’ve filtered for whites here):

Ethnicity—0.40

Highest Degree—0.55

Socioeconomic index—0.32

I think it’s interesting to note that the variable which reveals meritocratic achievement has the highest correlation. {snip}

This post is going to review some findings in a paper which attempts to both describe the differences in race preference for dating by race and across genders, and, why those differences might emerge the way that they do. The paper is Racial Preferences in Dating, Review of Economic Studies. Here’s the abstract:

“We examine racial preferences in dating. We employ a Speed Dating experiment that allows us to directly observe individual decisions and thus infer whose preferences lead to racial segregation in romantic relationships. Females exhibit stronger racial preferences than males. The richness of our data further allows us to identify many determinants of same-race preferences. Subjects’ backgrounds, including the racial composition of the ZIP code where a subject grew up and the prevailing racial attitudes in a subject’s state or country of origin, strongly influence same-race preferences. Older subjects and more physically attractive subjects exhibit weaker same-race preferences.”

{snip}

As you can see the demographics are unsurprising for an elite university. Do note that South Asians were discarded from the Asian category so that Asian within this study refers only to East Asians. The researchers collected a fair amount of data prior to sending them through the Speed Dating experiment. These and the GSS were used to generate independent variables that could be used to predict the extent to which race might matter. In the end, race didn’t matter that much, just to a statistically significant extent. The authors note that 47% of matches were interracial, while random expectations would have predicted 53%. Random mating in the general population would result in 44% of marriages being interracial, while only 4% are. It is obvious though that demographic segregation will shift away from a panmictic dynamic, and, I think marriage is frankly a higher bar than accepting the proposal for a date.

Here’s the chart that shows you who wanted who:

{snip}

Just inspecting the table, it’s pretty obvious what’s going on here. Some general observations:

1) Black women at Columbia are really open to dating black men. Take a look at the sex ratios there.

2) Asian dudes are really screwed. Not only are non-Asian women not impressed, but Asian women aren’t reciprocally racially discriminatory to level the playing field.

3) Women care a lot more about race than men. Though there is some variation in the male acceptances by race, but as I said most of the differences were not statistically significant. To make this concrete, black women were three times as likely to say yes to a black man as they were to an Asian man (though white women were the most repulsed by Asian men clearly).

{snip}

I’ll let the paper talk here:

We first look at the decisions of female subjects. For all races except Asians, all the coefficients on the race indicator variables are negative, implying a same-race preference. For black and white subjects, these coefficients are jointly significant (p-value < 0·01); for Hispanics, the joint significance is at the 10% level, with most of the effect derived from a significant (p <0·05) preference against Asian males. For Asian subjects, no coefficient is individually significant, nor are they jointly so. Finally, we can reject the hypothesis of equal preference against partners of other races for white, black, and Hispanic subjects, owing largely to the greater preference against Asian males by all other races. For male subjects, the coefficients on racial preferences are predominantly negative but are not jointly significant at 5% for any race. For white and black subjects, when females and males are pooled and gender-race interactions included we find that the male race coefficients are significantly closer to zero than the female race coefficients. In analogous regressions for Asian average, women exhibit stronger racial preferences than men. Note that, since our specification includes subject fixed effects, this difference cannot be due to differential selectivity. One possible reason for this gender difference might be the different dating goals of men and women. In particular, one might be concerned that women are more interested in forming a relationship while men are more interested in casual sex and that race has greater relevance for the former endeavour. However, in Section 3.3, we demonstrate that older subjects (who, based on their self-reported dating goals, are more interested in forming a relationship) exhibit substantially weaker same-race preferences. Thus, the observed difference seems to reflect a genuine disparity in men’s and women’s willingness to be with a partner of a different race, rather than differing goals.

First, Asian guys are screwed, obviously. I mean, look at how strongly Latinas have an aversion to Asian guys! Secondly, I don’t actually buy their dismissal of different goals. I’ll get into that later.

{snip} Here’s the above table controlled for attractiveness:

{snip}

Since you can read the whole working paper I’ll restate in plain language what they find re: determinants. If you want the exact betas just download the PDF.

Here are variables which predict same race preference:

* High proportion wish ban on interracial marriage in a region

* Areas where people would not want to be neighbors with another race

* Areas where there are large populations of other races

Here are variables which predict less same race preference:

* Older

* Attractiveness, 1 standard deviation increase in attractiveness results in a 4 point decrease in same-race preference

Variables which had no effect:

* Shared interests

* Cultural variables (books read, etc.)

* Income

{snip}

First, this is a sample of Columbia University graduate students! Caveat. But note that they didn’t really extend their findings much with speculation. My main issue with the paper is this: I still suspect it’s not taking into account different intent on the part of males and females. I don’t think the analogy to older students is appropriate. Their logic is simple, if one assumes that older students are looking for serious relationships and are less race conscious, if women were looking for serious relationships then they too would be less race conscious, not more. I think that this isn’t controlling for all life variables. I believe that as people age they become more realistic (or, if not, they stay single!). Someone who is young and wants a serious relationship might have an ideal type in mind who they are holding out for. Someone who is older and wants a serious relationship might realize that eliminating people on the basis of their racial type probably is constraining the field of play unnecessarily.

{snip}

I will add one more thing, though I’m hesitant. As a man of color I feel less than empowered when criticized by heteronormatively privileged white males on this point on the comment boards as they can silence my voice with their command of the English language which has long been the tool of the master, but could it be the white male patriarchy? Perhaps women, long oppressed by males have internalized the racism which is implicit in the current dispensation. One might wager that women perceive that the choices they make are fraught with far greater long term import than those men make, and so they stick with what they see as the “safest” option in the white male heteronormative patriarchy? Those who might be able to protect them? For white women it would be the white men who control the levers of power, and for women of color it would be the men who have been oppressed by the phallocracy and have been their allies of necessity against oppression.

[Editor’s Notes: The reader is urged to see the original article to study the tables for himself.

[“Racial Preferences in Dating,” by Raymond Fisman, et al., can be downloaded here.]