Spence's signaling theory

An economist asked a question that sounds almost heretical: what if education is valuable to employers not mainly because of what you learn, but because of what finishing it proves? His argument, which won a Nobel prize, goes like this. Employers cannot see how capable you really are before they hire you. Ability is hidden. So they look for signals, things that are hard to fake and expensive to obtain, that only capable people can produce. A demanding degree is one such signal, not because the coursework itself makes you better at most jobs, but because completing it demonstrates you could. The signal works precisely because it is costly, and costlier for the less able. The moment a signal becomes cheap or easy to fake, it stops carrying information, and everyone has to find a new one.

People get hired for the signals of their qualities that others can actually see, rather than for the qualities themselves, and those are two different things. This matters now more than ever, because artificial intelligence has made the old signals cheap to fake. A polished CV, a fluent cover letter, a rehearsed set of answers, all of these once signalled effort and ability, and all can now be generated in seconds. So the signals that still mean something are the ones that are hard to fake in real time: what someone can actually do when you watch them work, live, on a real problem. Everything else is getting easier to counterfeit by the day.

The Elephant in the Brain

Simler and Hanson, 2018

On signalling as a general mechanism, across education, medicine and charity, and why so much of what people do is costly proof of something they cannot simply assert.