How Important Is Human Capital for Development? Evidence from Immigrant Earnings

This paper offers new evidence on the sources of cross-country income differences. It exploits the idea that observing immigrant workers from different countries in the same labor market provides an opportunity to estimate their relative human capital endowments without having to adjust for other sources of country-specific productivity differences. Based on such estimates, a neoclassical growth model with human capital is used to decompose crosscountry income differences into the contributions of physical capital, human capital, and total factor productivity. In U.S. data the earnings gap between immigrants and natives with identical measured skills is less than 25% for nearly all source countries, suggesting that cross-country differences in unobserved skills are much smaller than cross-country income gaps. As a result, the model strongly rejects the hypothesis that the bulk of income differences is due to physical and human capital. For a sample of countries with less than 40% of U.S. output per worker, human and physical capital account for only one-third of the income gap relative to the United States. Allowing for immigrant self-selection and skill complementarities increases this fraction to at most two-thirds. For the poorest 5 countries in the sample, this leaves income gaps of 350% unexplained. This evidence is consistent with Prescott’s (1998) conclusion that accounting for cross-country income differentials requires a theory of total factor productivity. JEL: O15, O41, F22