Research without Reading

Joshua Gans reacts to my paper and post on mathiness with a response, tagged as “A guide for the perplexed.”

He suggests that part of why my argument is hard to follow is because in his words, I’m “going all meta,” which I take to mean using an economic model like the lemons model to describe the harm that mathiness does to the economics profession as a whole. It gets worse, here I invoke a reputational equilbrium as well.

True enough. But hey, I’m not the only one.

After Lucas dismisses books and papers as irrelevant special types of capital (see the quote here), he develops a model of progress in which:

– no one reads books

– no one reads journal articles

– no one has access to the Internet

– the only way to learn something is to bump into someone in the hallway

How can anyone read this without going all meta and inferring that it is an honest description by Lucas of how, in his experience, knowledge about economics is discovered at the University of Chicago?

In his P&P piece from the same session as my paper, Lucas gives an “overview” of the evidence on the relationship between human capital and growth that fails to mention the most important innovation in cross-national measurements of human capital that we’ve seen in decades–correcting measures of student seat-time in school by using internationally benchmarked tests. Instead of asking what fraction of a cohort has a high school degree, ask what fraction can add fractions. This totally changes the patterns of correlation between human capital and growth. See for example this report from the OECD and the papers cited there.