Financial Aid Is Not Predictive

The clip above was taken from a webinar I presented last month – “You’re Probably Optimized.”  The recording is still available if you’re interested in the broader context.

Last month, I did a quick survey of higher education leaders on linkedin, asking how predictive the amount of financial aid offered to a student was of eventual enrollment.

Wow … were we wrong.  As you can see, over 90% of my peers thought it was a top ten indicator, and almost everyone thought it was at least in the top 50.  (Full disclosure: the one vote suggesting it wasn’t was an enroll ml team member who had seen the data)

The reality?  The amount of financial aid (or any proxy of that measure) is not a significant predictor of enrollment.  It’s not number one, it’s not in the top ten, it’s not in the top fifty.  In fact, for every institution we’ve evaluated so far, it’s not even in the top 100.

In the full webinar, I dive deeper into the difference between univariate and multivariate that led our profession to this misconception.  But as you can see in this short clip posted above, there are a few corollary factors that are missing from just measuring “aid amount.”

Gap.  What matter more than just how much aid is offered is the gap between their ability to pay and their net price.  Colleges cannot accurately measure gap.

Willingness to pay.  Let’s imagine a scenario where all sources of funds available to a student are fully identified.  Colleges still could not accurately measure the family’s willingness to pay for an education at that particular institution.

Changing value proposition.  Playing along, even if an institution could somehow unlock a genuine answer to the willingness to pay question, it would be a mistake to assume that the answer is static.  As more information (and more aid packages) come in, their willingness to pay is going to change.

These are my theories for why aid offered is not predictive.  It’s not that it isn’t important … it is.  But it’s just one of the variables being input into a complex student decision – and the most significant co-variables impacting it are blind to us.