December 2008

IZA DP No. 3897: A Note on Measures of Human Capital for Immigrants: Examining the American Community Survey and New Immigrant Survey

In this paper we examine whether where one acquires their human capital matters in earnings regressions. We focus on a nationally-representative US data set and find that there is little difference between a measure of total years of education and measures for US and foreign-based years of education. There is a large difference, however, in where total experience is acquired: US-based experience commands a higher return to wages and is statistically highly significant. The measures used in this analysis must be inferred based on the year of migration to the US. Using an immigrant-specific data set, the New Immigrant Survey which contains explicit information on the human capital acquired in the US and abroad, we confirm these results.