Matt Zeitlin

I generally do want to be in a situation where I’m disagreeing with Matthew Yglesias and Matthew Rognlie — someone who blogs way too little for how smart he is (he’s going to MIT in the fall to start his economics PhD) — and I do not know if I disagree with them in their argument for why it might be OK for employers to look at credit reports when they hire people (for the opposing view, check out Kevin Drum), but there’s a slight wrinkle to this debate that I do not think the illustrious Matts are considering.

Aside from the general point that liberals ought not to spend their time regulating business aside from dealing with externalities, Rognlie argues that there might be a good reason that employers would want to get a look at the credit history of prospective employers. So, we should generally trust employers to know whether or not it’s true that someone with worse credit will be a worse employee, and that if we find out that this is a totally baseless idea, then regulatory action can be considered. This sounds reasonable enough, but I don’t think it reckons with the possibility that a worse credit score, could be meaningful and it would be wrong for an employer to look at it. Let’s look at another fact about someone that employers have used as a proxy for other qualities they are looking for in an employee: race.

As we all know, labor market discrimination by race is both immoral and illegal, but it happens, and it doesn’t just happen because people are evil, but because there is an empirical element. Tim Harford had a great piece from a while back in Forbes where he talks about taste-based and statistical discrimination. The former is when employers just don’t like ethnic minorities and are straight-forward bigots and the latter is discrimination that happens because certain racial and ethnic identities are used as markers for other traits that matter to employers. Both, of course, are morally wrong and illegal, but the latter is hard to root out because, absent regulatory or legal pressure, employers will benefit from statistical discrimination, or at least they won’t suffer from it. From the perspective of the qualified black guy who can’t a job, there’s no difference between the two types of discrimination:  he’s still being judged based on his group identity as opposed to his individual qualities.

So, when we talk about what type of information employers should either have access to or should make employment decisions based on, we can’t just talk about what information correlates with more valid and appropriate qualities of a prospective employee.

I think Drum might be on to something when he says that we should not let employers look at credit history, just as we don’t allow employers to use race to evaluate prospective employees on race (although, of course, we can’t stop them from observing it). But the reasons for not allowing credit history have to be extra-economic, because we know that statistical discrimination can both be rational and wrong. The reasons for protecting credit history are both that it’s information that is traditionally private and protected — I don’t think Rognlie or Yglesias agree that anyone should have access to the credit history of their neighbors, friends and acquaintances – and that it can create a situation where the poor who are more likely to have a spotty credit history can not get jobs because of their spotty credit history and their credit situation gets worse and then can’t get a job, and so on and so forth. Sometimes, considerations of justice need to outweigh the short-term justification for the powerful in labor markets. Much the same situation exists in hiring ex-cons. Surely it’s a case where statistical discrimination makes some sense, but as a society, we are worse off when ex-cons have a very hard time entering the labor market because it both extends their punishment past the one handed down by a judge and, when convicts can’t get jobs, they are more likely to commit crime.

This is not to say that any and all statistical discrimination is something the government should try to prevent or that mucking around in labor markets is the right thing to do, but everyone accepts that such mucking around is OK in some circumstances, and those reasons might justify doing so in other cases.

*Yikes, just after writing this post, I realized that Rognlie wrote another post incorporating a response to the kind of points I made. But I just spit out hundreds of words on this, so I’ll just link to Rognlie’s follow-up.

*For more great Rognlie stuff, here’s his defense of Fahrenheit and “Preference Functions That Score Rankings and Maximum Likelihood Estimation“