When Immigrants Are Not Migrants: Counting Arrivals of the Foreign Born Using the U.S. Census 1

Abstract

This paper compares characteristics of recent immigrant arrivals in the United States using two measures from the decennial U.S. census: the came-to-stay question and the migration question. We show that a little under 30 percent of immigrants who reported they came to stay between 1985–1990 on the 1990 U.S. Census Public Use Micro Sample were resident in the United States on April 1, 1985. A similar analysis of the 1980 censue reveals that 22 percent of immigrants who reported they came to stay between 1975–1980 lived in the United States on April 1, 1975. Thus among recent arrivals, defined as those who reported they came to stay in the quinquennium preceding the census, a large number were resident in the United States five years before the census date. Furthermore, the proportion of recent arrivals present in the United States five years before the census increased between 1975–1980 and 1985–1990. We show that the profile of recent arrivals is sensitive to their migration status. Generally, in both the 1975–1980 and 1985–1990 cohorts, those resident in the United States five years before the census have significantly less schooling and lower incomes than those who were abroad. Accordingly, we argue that estimates of the skill levels and hourly wages of recent arrivals to the United States vary with the way arrival is measured. Researchers who rely on Public Use samples of the U.S. census for their data should be aware that the year of entry question implies a broader definition of arrival than the migration question. We caution that immigration researchers should consider the idea of arrival more carefully to help distinguish newcomers from the resident foreign born.