This study uses Social Security earnings records matched to recent cross—sections
of the SIPP and CPS to study the earnings progress of U.S. immigrants. The data
show that immigrants’ earnings grow 1 to 13 percent during their first twenty years
in the U.S. relative to the earnings of natives with similar labor market experience. By
comparison, estimates of immigrants’ relative wage growth from cross—sections of the
decennial Census are substantially higher. The divergent results reflect the selective
outmigration of low—earning immigrants. The longitudinal earnings histories also show
that 14 percent of immigrants have earnings in the U.S. prior to their most recent date
of arrival, which points to a significant amount of back—and—forth migration between
the U.S. and immigrants’ home countries. The misclassification in previous work of
these largely low—wage immigrants as recent arrivals accounts for close to one—third of
the measured decline in the level of earnings of immigrant arrival cohorts between 1960
and 1980. The new evidence presented here, therefore, suggests that previous analyses
had overestimated both the rate of earnings growth among immigrants who remain in
the U.S. and the secular decline in the level of earnings across arrival cohorts.

Since recent immigrants tend to earn less than natives, their relative labor market
status has been adversely impacted by an increase in the return to labor market skills
and widening wage inequality over the past two decades. To evaluate the magnitude
of this effect, this study uses Social Security earnings records matched to recent cross
sections of the SIPP and CPS to estimate the change in the return to skills among
native born workers. This is then used to adjust the earnings gap between immigrants
and natives in order to estimate what the gap would have been if the return to skills
had remained at its 1980 level. The results suggest that the return to skills rose by 40
percent between 1980 and 1997, leading to a 10 to 15 percentage point decrease in the
relative earnings of recent immigrants. Thus examining solely the earnings of recent
immigrants may lead to an overly pessimistic picture of their actual labor market skills.