Web Policy  |  Intranet  |  Contact SCHEV    
SCHEV Research Data Blog

The official blog of @SCHEVResearch at the State Council of Higher Education for Virginia. Discussions about our work, national higher education data policy, and highlights about the data we publish.


Recent Posts Blog Archives Subscribe Feed All Blogs

Value-added College Rankings

by Tod Massa 1. May 2015 23:01

One of the things that struck me immediately upon reviewing the new Brookings report and data for the value-added college rankings was the appearance and ranking of Heald College-Concord. Despite near-top scores in earnings and loan repayment, it was one of the Corinthian colleges that closed a couple days earlier. Thus I was left to wonder if the data were wrong, the methodology, or USED in its actions. Or perhaps it was just a small piece of a larger system caught up in the shutdown. Maybe it was about all of those things. In any event, there is lesson here about the intersection of policy, data, and rankings.

There are some interesting things about the College Value-Added Data Explorer in the way authors use various datasets to develop the rankings. There are also occasional concerns, particularly in the way they make the data consumable. For example, "This report converts default rates to repayment rates for ease of comparison with positive economic outcomes." Unfortunately, in the context of federal student loans, the inverse of default is not repayment.  Default occurs when a borrower on a monthly repayment plan does not make a payment for 270 days, so of course this implies that some number of borrowers in the 270 day window are being counted as repaying their loans when they are in fact nearing default. Further, with the various income-based repayment options (ICR, IBR, and PAYE), a growing number of borrowers are making payments lower than needed to reduce the principal, and many are paying nothing because their incomes are too low. I'm not sure these borrowers should be considered part of a repayment rate, nor should those in the variety of other statuses, such as deferment and forbearance.

The blending of occupational data from LinkedIn, mid-career wage data from PayScale, and the Bureau of Labor Statistics Occupational Employment Survey is certainly imperfect, but it probably provides estimates that are meaningful in terms of the relative position of institutions. I don't think I would put a great of emphasis on the specific values. Including the "STEM  orientation" component was smart as our own data shows those differences by broad discipline and program. Of course, how one determines the definition of STEM is really important. 

I think the this project is a necessary evolutionary step in exploring the link between institutions, degrees, and earnings, but I think it is a long way yet from something that we can generally agree is accurate and helpful. From my perspective, it is not a model appropriate for the state or federal government to use. 



Categories: Reports

blog comments powered by Disqus
Follow SCHEV Research on Twitter