The Strada Education Foundation released last week a major report, “Talent Disrupted,” on college graduates and underemployment. Trumpeted by the Wall Street Journal as demonstrating the importance of majors and internships, the large-dataset study claims to show that majoring in useful things like health sciences and quantitative-heavy subjects, along with having an internship, are the ways to avoid underemployment after college.
It is a big study, and thoughtfully done. It’s commendable for including data for colleges across the spectrum of American higher ed, which is extremely heterogeneous. However, I have my doubts about some of the analytic choices and, therefore, the prescriptive conclusions the study presents.
While I agree with what will be a common critique — that college isn’t only for employment preparation but for civic, intellectual, moral, and personal development as well — I’m not going to dwell on that point. I think it’s reasonable for students to want to be employable after college, and the report clearly states that that’s the topic of study here.
My first doubt is this: the report focuses almost exclusively on what students do during college (majors and internships) instead of where they do it (types of colleges). To be fair, that’s a welcome correction to the “black box” approach that treats achievement of a four-year BA as uniform.
This diagram, for example:
Implies, without demonstrating, that high-risk degree fields produce high or very high unemployment, regardless of the selectivity of the college attended. But the regression analysis presented in the reports’ appendix doesn’t actually include choice of major at all; it shows a strong effect of college selectivity (-31.5% for “more selective” colleges), but doesn’t appear to include information about field of study. I’m not saying selectivity should predict lower underemployment, but I strongly suspect that it does, and that major choice is secondary.
Second, there is no ability to evaluate pre-college social class or social capital, both of which likely play very large roles in post-college success (again, not should, but do). Arguably, the right question for a college-effects study is not comparing all students after graduation, but comparing students who are similar on matriculation in terms of their post-college success. That would then be a measure of college’s contribution net of the student’s pre-college likelihood of such outcomes. I recognize this is very difficult to accomplish, having tried a little; my point is that comparing post-college underemployment without adjusting for pre-college status shouldn’t be interpreted as a real comparison of college’s effects.
I do think that one thing colleges ought to do is to see general education as, in part, providing sufficient skills that students will be competitive for employment in college-level jobs regardless of their major. If the capacities general education conveys are sufficient, students could feel freer to major based on their substantive interests, confident that their legitimate employment aspirations will remain plausible.
