The schooled are healthier, but school doesn’t make folks healthy: There is a strong, positive and well-documented correlation between education and health outcomes. … [We study] two changes to British compulsory schooling laws that generated sharp differences in educational attainment among individuals born just months apart. … The cohorts just affected by these changes completed significantly more education than slightly older cohorts subject to the old laws. However, we find little evidence that this additional education improved health outcomes or changed health behaviors. (
LGA, they are using a gap in birth-dates of just a few months. Should we really see much of a dumbing down effect over that time span?
No, I think colleges have dumbed down their courses and their entrance requirement so they can get more people in the door and get more of those education dollars that have been thrown around.
Other people mistook the correlation between IQ and health as a correlation between education and health, probably because they have a vested interest in selling that shit so they can get on the gravy train supported by governments expropriating ever more money from taxpayers.
Now that the average IQ of college attendees has gone down, the correlation is starting to go away.
You really think a big drop in IQ happened at exactly the same month as the increase in education so as to mask a real benefit of education on health? That seems pretty far-fetched.
The cognitive epidemiology people are starting to do that these days.
I would be concerned about the hidden variable known as IQ.
Yes, natural experiments might have their limits, but they're still likely to be better for correcting endogeneity than most statistical techniques.
TGGP, the AER papers are not based on random assignments or exogenous shocks. They try heroically to correct for selection effects, but in the end I just don't believe they corrected well enough. The paper I prefer above is based on a clear exogenous shock.
Could you explain what makes the techniques in the nber paper reliable but the ones in the AER papers less so? That would be relevant for the Hanson-Frakt debate on redoing the RAND health experiment (or the Frakt-McArdle one on the existing health insurance literature).