Last week I linked to a CEPR report showing that the U.S. has one of the smallest small business sectors in the OECD. The authors of that report speculated that this may be due to the lack of universal health insurance in the U.S., and Krugman agreed with them. I acknowledged that it was possible that a lack of public health insurance had something to do with the lower small business sector, but also argued that there plenty of other possible causes as well, and it's irresponsible to assume causal links that just happen to support our priors. Well, Scott Shane of Case Western studies this stuff for a living and he came to a similar conclusion:
If you compare health care’s portion of G.D.P., as measured by the World Bank, with the self-employment rate across the 21 O.E.C.D. countries in the C.E.P.R. study, you will find that the correlation is only -0.19, which statistically speaking, cannot be considered different from zero.
I know that everyone is focused right now on the health care debate, which is great. But we can’t just say that the absence of universal health care influences the size of our small-business sector. It’s certainly possible that national differences in health care affect national differences in self-employment and small-business generation in some way.
But we have to show some evidence of those effects.
Exactly right. Further, Shane did find a different statistically significant result (p < .01): richer countries have less self-employment than poorer countries. The causal direction could run either way -- the self-employed are generally poorer than employees, or the poor have less human capital and therefore fewer opportunities to get high-paying jobs -- but either way it doesn't to appear to have anything to do with health care. Or if it does, we need more than just baseless speculations motivated by partisanship to establish the claim.
Krugman and others were (rightly) vehement in their denunciations of the "dumb cowards" at the Investor's Business Daily for playing fast-and-loose with factual claims about British health care. Perhaps Krugman should hold himself to his own standard and refrain from making causal claims without evidence.