An interesting comment from hwc:

Swarthmore is reporting their fall 2010 enrollment both ways: the new for 2010 federal mandated way with the multi-racial category and (on a page 2 of the PDF), the old-style reporting methods for historical comparison. By clicking back and forth between Page 1 and Page 2 of the PDF, you can compare the numbers for Fall 2010 enrollment under the two reporting systems.

I haven’t digested exactly what these changes mean, but I know that it is basically going to end my spreadsheet and chart of diversity with data going back 40 years. Thanks, Uncle Sam.


Right off the bat, I can see that the new federal reporting rules will increase the apparent “Hispanic” enrollment as “Hispanic” must be reported regardless of the race. The reported “Hispanic” enrollment at Swarthmore actually increased with the new reporting mandates.

With the reporting of 86 “multi-racial” students, the reported enrollment of Asian American and “black” students fell sharply under the new reporting rules. With what is likely to be a sharp decline in reported African American enrollment in US higher education under the new rules, I wouldn’t be surprised if the racial victimization lobby makes some loud noise over these new rules. With the stroke of a Washington bureaucrat’s pen, Swarthmore just went from 9.6% African American enrollment to 6.4%, purely from reporting the same student body under the new federal rules.

1) I hope that Williams, like Swarthmore, will report the data both ways for at least this year. I would expect that we will see similar drops in Asian-American and African-American enrollment at Williams.

2) Recall previous discussions on the topic of racial backgrounds. In the past, Swarthmore and Williams have treated someone with 4 African-American grandparents the same as someone with 1 African-American grandparent and 3 white grandparents. As long as they checked the “African-American” box on the Common Application, they were black.

But the new scheme makes that much more difficult. I would bet — commentary from people with expertise in the literature welcome! — that the applicant with four black grandparents is highly unlikely to check the new “mixed race” box while the applicant with just one black grandparent is much more likely to do so. What happens? At Swarthmore, the number of African-American freshmen goes from 43 to 22. Wow! Does this mean that half the 43 African-American freshmen at Swarthmore have one non-African-American parent?

3) Could the 1/3 total drop in African-American enrollment that hwc highlighted be a underestimate of the final effect? I find it interesting that the drop is almost 50% in the freshmen class. Swarthmore probably finds it much harder to figure out which seniors would classify themselves as mixed race. Or perhaps seniors responded less often to whatever new surveys Swarthmore has done. If future classes are like this year’s freshmen class, Swarthmore will report a drop of 50% in African-American enrollment.

4) What does that say for the future of measures of diversity at elite schools? Hard to say. Currently, diversity is measured by adding up all the students who checked any non-White box, including Hispanic but not including “Race/ethnicity unknown / Other.” (Everyone assumes that these are overwhelmingly white students.) That is the metric that Dick Nesbitt uses when he describes the current Williams class as being our most diverse ever. But that metric becomes much more problematic with the new classification scheme. Consider:

a) How do you avoid double-counting? Swarthmore presents numbers for “Hispanic, of any race” and the individual racial groups but no information on the overlaps. These 175 Hispanic students could all be white, in which case a measure of total diversity would just add them to the numbers for African-American, Asian and so on. Or many/most of them could also be racial minorities, in which case we can’t just add them without double-counting. (Thanks to Rory for pointing out my mistake in his comment below.)

b) How do you handle “Two or more Races?” On the one hand, these should clearly be included in any diversity measure because many students with African-American heritage check this box, as the Swarthmore data shows. On the other hand, this box becomes an easy one to check for students who are, for all practical purposes, white but hope to get a bit of an admissions edge. (Useful discussion here.)

I predict that we will see a dramatic increase in the number of students checking the “Two or more Races” box. Think of the incentives. Colleges will get to count these students in any plausible measure of diversity, so they want to get as many of these students as possible. Students will feel much more comfortable “stretching” to check this box than they felt doing the same for just “African-American.” There are also an ever increasing number of applicants of Asian-white mixed parentage who suspect, probably correctly, that checking this box is much better than checking either White or Asian.

What do you predict will happen?

Print  •  Email