Against credentialism

That is the theme of my latest Bloomberg column, induced by a timely tweet by Conor Sen.  It turns out the state of Maryland is abolishing the four-year college degree requirement for many state jobs.  In Missour, neither the governor nor the lieutenant governor have a four=year college degree, so perhaps they should follow suit?

From the column, here is one bit:

On average, more education probably does correlate with better job performance — but there are a lot of exceptions. If U.S. society wants to boost opportunity for everyone, it needs to work harder to spot those exceptions and act on that knowledge. In a world where so much information and so many diverse forms of certification are available, there are far better ways to assess a candidate than asking the binary question of whether they have a four-year degree.

This move against credentialism is all the more imperative due to the rise of technology. Many of the top names in tech or crypto are dropouts and do not have degrees. To be sure, those are not the kind of people the Maryland state government is likely to be recruiting. But there are numerous people in tech, lower on the salary scale, who have not invested much in formal credentials, in part because they failed to see their professional relevance. For many tech jobs, a personal GitHub page is far more important.

From these passages you also can see why the credentialism critique is slightly different from some of the more radical critiques of educational signaling.  In my view, education does causally improve performance on a lot of jobs, at least on average, through more or less the traditional channels.  Still, treatment effect variances can be high, and very often we can do better with more finely grained assessments of individual talent, rather than taking a four year degree to be a binary yes/no qualification.  In many cases, for instance, would you rather not hire some with a military background?  The bottom line is that you can achieve a superior allocation of talent, and cut back on signaling costs, even if you think signaling is a clearly present but not dominant force behind the demand for higher education.