A loyal reader asks for my comments on this New York Times article.

Over the last decade, dozens of colleges have proclaimed that recruiting a more economically diverse student body was a top priority. Many of those colleges have not matched their words with actions. But some have.


Lots of rich schools are, unsurprisingly, at the top of these rankings. It is easy to preen if you have a billion dollar endowment. Perhaps naively, I expected Williams to be higher. I am surprised by how well (if that is the word you want to use) Vassar, led by President Cappy Hill ’76, has done. Details on the methodology:

To measure top colleges’ efforts on economic diversity, The Upshot calculated a College Access Index, based on the share of freshmen in recent years who came from low-income families (measured by the share receiving a Pell grant) and on the net price of attendance for low- and middle-income families. The following table also shows colleges’ endowment per student, which is a measure of the resources available to colleges. Colleges with a four-year graduation rate of 75 percent or higher in 2011-12 are included.


1) I find the focus on Pell Grants deeply suspect. First, international students are not eligible. So, a school that with 50% of it students from very poor Mexican or Brazilian or Ukranian families would not do well because those student aren’t counted in this methodology. Second, it is not obvious that Pell Grant eligibility is a good measure of economic diversity. Would the child of a rich (but retired) parents be included? I don’t know the details. Does anyone?

2) The numbers are suspect. The key phrase is “net price of attendance for low- and middle-income families.” See the methodological details. (Kudos for transparency.) Also, the raw data seems off. Vassar is at $5,600 but Amherst is $8,400? Impossible! Amherst is much, much richer than Vassar, and is every bit (perhaps even more so) committed to socio-economic diversity as Vassar. Why would they charge poor families 50% more than Amherst does? I believe that the authors got the numbers correctly from IPEDS, I just doubt the quality of the underlying data.

Print  •  Email