Thursday, August 30, 2007

Bias in subprime market like bias in medicine

Something bad happens. Politicians (policy doctors) recommend regulation, which will prevent the bad thing from happening again. Another bad thing happens, but instead of concluding that regulation did not work as promised, regulation is increased.

Arnold Kling looks at the problem of bias, from the shared perspective of doctors, patients, and pro-regulation crowd:
It is better to be certainly wrong than statistically right. That is, doctors and patients feel better if the doctor makes a diagnosis based on intuition about the specific case rather than based on some statistical model.

In hindsight, it often appears that one could have made the right decision using better intuition. That is why an anecdotal approach to looking at medical decision-making is biased in favor of intuition and against statistical models. That same sort of bias is at work in the cries for more regulation of financial markets in wake of the subprime mortgage problems. In hindsight, you think that a regulator could have prevented the problem. But that is a biased perception.

No comments:

Post a Comment