Saturday, July 23, 2011

Subjective data and health

Freakonomics posts some great statistics from a high-quality study of the impact of Medicaid on people's health and finances.

I'm a big fan of using measure of happiness to evaluate programs. Freakonomics reports the data:

  • 10% increase in the probability of screening negative for depression.
  • 25% increase in the probability of reporting one’s health as good, very good, or excellent.
  • 32% increase in self-reported overall happiness.
I don't know who did the screening for depression, but I take that as solid evidence that Medicaid is doing something very, very valuable. (I don't know what the cost was so I won't weight on of if it was worth it.)

The health and happiness data, though, seem suspect. The study was blind and indeed, Freakonomics reports, most of the impact on health happened immediately after enrollment, before people get any care. It sounds like the effect might not be real. Alas, subjective data has its limits.

No comments:

Post a Comment