The Cost of "Acting White"

Roland Fryer and Paul Torelli assess the cost of acting white. (PDF file)

Among whites, higher grades yield higher popularity. For Blacks, higher achievement is associated with modestly higher popularity until a grade point average of 3.5, when the slope turns negative. A black student with a 4.0 has, on average, 1.5 fewer same-race friends than a white student with a 4.0. Among Hispanics, there is little change in popularity from a grade point average of 1 through 2.5. After 2.5, the gradient turns sharply negative. A Hispanic student with a 4.0 grade point average is the least popular of all Hispanic students, and has 3 fewer friends than a typical white student with a 4.0 grade point average. Put differently, evaluated at the sample mean, a one standard deviation increase in grades is associated with roughly a .103 standard deviation decrease in social status for Blacks and a .171 standard deviation decrease for Hispanics. For students with a 3.5 grade point average or better, the effect triples.

Racial differences in the relationship between popularity and achievement are robust across many alternative specifications, subsets of the data, and different definitions of popularity or academic achievement. The ‘acting white’ effect differs slightly as children age. For Black students in junior high school (grades 7-9) a one standard deviation increase in grades is associated with a .13 standard deviation decrease in popularity; the decrease is .07 for high school students. Accounting for the number of students at each GPA level does little to temper the ‘acting white’ effect, suggesting that the supply of potential friends does not explain the phenomenon. Self-selection of students into particular activities (sports, band, debate, etc) matters little, and substitution towards other race friendships does not fully explain the stark difference in the popularity– achievement gradient. Using class rank (in lieu of GPA) as a measure of academic achievement confirms our results.

We conclude by describing a model which highlights the tradeoff between peer group acceptance and academic achievement; a variant of Austen-Smith and Fryer (2005). The principal idea is that individuals face a two-audience signaling quandary: signals that beget labor market success are signals that induce peer rejection. The model’s two distinguishing predictions – racial differences in the relationship between peer group acceptance and academic achievement will exist and these differences will be exacerbated in arenas that foster more interracial contact or increased mobility – are borne out in the data. ‘Acting white’ is more salient in public schools and schools in which the percentage of black students is less than twenty, but non-existent among blacks in predominantly black schools or those who attend private schools. Schools with more interracial contact have an ‘acting white’ coefficient twice as large as more segregated schools (seven times as large for Black males). Other models we consider, such as selfsabotage among black youth or the presence of an oppositional culture identity, all contradict the data in important ways.

Comments

[Joe O]

That paper is pretty interesting. The comparisons of the popularity of black/hispanic students with a 4.0 versus white students with a 4.0 are misleading since white students in general were found to be more popular.

The graphs start at page 46. The popularity curves for hispanic students are particularly concerning. For black students the popularity decline is much milder.

The popularity data in particular seems very trustworthy, since it is based on the students' reports of their closest friends' names to determine the most popular kids. The GPA data is weaker because it uses self reporting and doesn't account for honors or AP classes.

Page 47 shows the popularity curves for students at public and private schools. There may be insufficient data for the private schools because this graph is very wiggly. White kids at private schools have their peak popularity at about a 1.75 GPA with a big drop down to a 4.0. Could there really be such a bias against white kids acting smart in private schools?

[Wes Pipe]

Gee, when I went to school there were no blacks and no hispanics...we were all white. The only deviation was there were lots of Ukrainians, some French, some Poles, some Natives,one Chinese,but mostly English. Of course being 80, this was 65 years ago and it was in western Canada. Yes the natives and the chinaman were of slight color but had been interbred so I call them white. And I think it may be so still today as the underground black railway operated mostly in Halifax, Nova Scotia.