Too much theory, not enough data?

Maybe. I’m not convinced.

UPDATE: While fighting another battle with Borjas, Rodrik defends theory: “However, empirical work in trade does remain much more strongly grounded in theory compared to labor, which on balance should be a good thing.”