Information on emigration is important for several reasons. In terms of simple numbers, emigration is part of the equation that yields net immigration—which, after all, is what implants lasting effects in the U.S. economy and population. Even in the short term, the admission of I million immigrants a year would have vastly different consequences, depending on whether, say, all of them stayed or 500,000 among them returned. Beyond mere numbers, emigration alters the characteristics of immigrant cohorts; we discuss in Chapter 8 the issues in connection with language. Similarly, virtually all labor market tests of the ability of immigrants to assimilate into the labor market follow cohorts of entering immigrants. If large fractions of a cohort emigrate and we do not have good estimates of the selectivity of this emigration, then it is impossible to know whether the cohort successfully assimilated or not.

Emigrants select themselves, but why? They may be the most skilled, drawn by better job prospects in other countries, in which case the average labor market success among those who remain is lowered. The less skilled, too, may be drawn by better prospects elsewhere, thus improving the average quality of the remaining cohort of immigrants.

Demographic information on emigration is particularly scarce and elusive. Estimates of the level of emigration are periodically updated, but these contain a substantial range of estimates. There is some information on the relationship between emigration and variables such as country of origin and sex. However, definitive knowledge is lacking on the relationship between emigration and many important behavioral variables, including fertility, the use of public assistance programs, and labor force characteristics.¹⁶

Some demographic studies have sought to identify the influence of sex and schooling on emigration (see Jasso and Rosenzweig, 1990:138-148). They suggest that men tend to emigrate more than women do, and, at least in earlier periods, that those with more schooling tend to return more than those with less. Interpreting these observations calls for caution, because the information is limited and because it is not current.

Still, available studies suggest that a high proportion of illegal immigrants are sojourners: they come to the United States for several years but eventually return to their home countries. According to studies by Douglas Massey and his colleagues (Massey et al., 1990), a high proportion of Mexican illegal immigrants return to their original villages after one or more prolonged periods of working in the United States.

These caveats aside, available estimates suggest that between roughly 35 and