When is trade protection good for growth?

The empirical relationship between trade protection and economic growth is surprisingly fragile, as shown in a number of other papers. We address one possible explanation for these findings: that the relationship is contingent on the pattern of comparative advantage, following the endogenous growth literature. Our findings suggest that such contingencies do in fact exist-in particular, the correlation between tariffs and growth is strong and positive for skill-abundant countries-and are robust to the choice of control variables.