i (6)
The empirical analysis used the 1970, 1980, and 1990 Public Use Microdata Samples of the U.S. Census. These data permits the tracking of specific cohorts of immigrants over a 20-year time frame. The immigrant cohorts were defined in terms of national origin, year-of-migration, and age-at-arrival. The study generated a number of findings:

1. There is a weak positive correlation between the log entry wage and the rate of wage growth, suggesting some complementarity between the human capital acquired abroad and the human capital that immigrants acquire in the United States. This positive correlation, however, probably turns negative if we compare immigrants who have similar human capital endowments when they enter the United States.

2. There is no evidence that more recent immigrant cohorts, who have lower entry wages, experience faster wage growth.

3. Because of the relative weak complementarity in the human capital production function, the same source country characteristics that improve the economic status of immigrants at the time of entry also lead to larger human capital acquisition in the post-migration period and faster wage growth.

The long-run economic performance of immigrants in the United States plays an important role in any assessment of the economic impact of immigration. The empirical evidence presented in this paper suggests that immigrants who enter the United States with a sizable human capital endowment are also the immigrants who find it easier to adapt and acquire additional skills in their new surroundings. As a result, the process of economic assimilation does not “even out” the playing field in the immigrant population. Instead, the assimilation process may actually increase income inequality among national origin groups in the immigrant population.