THE IMMIGRANTS: Summary Statistics 2

The data used in this paper are from the I960, 1970, 1980 and 1990 Public Use Microdata Samples of the United States censuses. The 1960 data are a 1 percent sample of the population. The 1970 data are also a 1 percent sample (from the 5% State sample). The 1980 and 1990 data are from the 5% samples, however, a only a random subset of each of these is used in the analyses presented here.

In all cases, the data refer to men and women between the ages of 16 and 65, inclusive. Individuals born abroad to American parents and those born in United States outlying areas (e.g. Puerto Rico) have been dropped from the analysis. Immigrants are defined as those who were born outside the United States. We also drop the self-employed read only.

There have been several changes in the way certain information is coded across the four censuses used here. In all cases, we attempted to define our variables in a consistent way. These changes affect some of the analyses we are able to perform. For example, although the 1970-1990 Censuses give fairly detailed information on the time period during which an immigrant entered the United States, the 1960 Census does not. The 1960 Census does, however, allow us to identify those who entered within the last five years. For consistency, we define “recent” immigrants as those who immigrated within the last five years in all our data.

Much of our analysis focuses on log hourly wages. We calculate these from information reported on total annual earnings from wages and salary, number of weeks worked, and usual hours worked per w^eek.3 Many analyses using census data use weekly wages rather than hourly wages, since these calculated hourly wages may be prone to measurement error. However, weekly wages confouri_d wage rates and labor supply. Furthermore, our comparisons of mean hourly wages of immigrants and the native-born look very similar to results from studies that use mean weekly wages, instead.
Descriptive Statistics

Tables la and lb show summary statistics for the native-born, all immigrants, and recent immigrants for men and women in all four decennial censuses. There are a number of striking differences between immigrants and natives across the years. The first rowr of each panel showTs mean age. In 1960, immigrants were on average ten years older than the native-born. With the increase in immigration during the 1980s and 1990s, the average age among immigrants falls to parity with the native-born. As one would expect, recent immigrants are younger than the native-born and other immigrants. Their average age has hovered around 30 years old across the four time periods.