Measuring Immigrant Wage Growth Using Matched CPS Files

Abstract

Cross-sectional estimates of immigrant wage growth have painted an optimistic picture of the ability of immigrants to adapt to the U.S. labor market in that studies using cross-sectional data have generally found immigrant wage growth to exceed that of the native born. This optimistic picture of immigrant economic assimilation was shattered by the important finding that recent immigrants were starting at much lower wages than was true of earlier immigrant cohorts. As such, the high wage growth of immigrants relative to the native born measured in cross-sectional data may simply be the spurious result of declining immigrant earnings ability. In this paper, we match Current Population Survey samples so that the wages of individual immigrant and native-born men can be followed for one year. This approach will become an increasingly useful tool for studying immigrant labor market behavior with the recent commencement of immigrant identification on all CPS surveys. Using matched CPS samples, we find that the wage growth of immigrants exceeds that of the native born. In 1988, the median annual wage growth of immigrants was 6.7% compared with 4.4% for workers born in the United States. We also find that the wage growth measured by following individuals closely parallels the projected wage growth based on a cross-sectional estimation. This similarity suggests that either there is no cohort quality bias in the cross-sectional estimates of immigrant wage growth or that there has been a coincidental increase in immigrant wage growth as the entry wages of immigrants have fallen.

JEL Classification: J61