No sooner did I post my thoughts about the Supreme Court decision on credit scoring than I got an e-mail from an American Insurance Association staffer with an impressive bibliography of sources showing a correlation between credit scores and risk. I actually knew about one of the he cites and, for whatever reason, it slipped my mind when I was posting.
But I still stand by my basic point: use of credit scoring for setting property and casualty insurance premiums seems awfully indirect and, yes, a bit screwy. In a freer market that allowed a broad range of risk factors, I still suspect that we would see it become less important. I agree that it has predictive validity and, since the cost of obtaining credit scores is so low relative to other things, I wouldn’t expect it to go away entirely.
Right now, however, a lot of the best risk factors are currently near impossible to monitor for a mix of political and technological reasons. In part because its received so little emphasis at the federal level (except during a brief period under the Clinton administration) we don’t have nearly enough data about what mitigations work to help real estate better deal with extreme meteorological events. We know a lot more about safety equipment on cars but I think we run into a bit of a verification problem there too. Just a few weeks ago, I noticed that I was getting an undeserved discount on my insurance for a feature that had been in the database when I got first got the policy but wasn’t actually in place on my car. How was my insurer to know?
In time, that will change. For the moment, credit scores are useful and well worth using. I just suspect that they would become less useful relative to other factors in a freer, less political market.