tinct as to be referred to as ''races"—have intermarried to such an extent as to virtually erase differences in education, income, occupation, and residence.
The picture is similar on the sensitive issue of the English language. Many immigrants arrive with at least a working knowledge of English. The 1990 decennial census found that three-fifths of the immigrants who came in the 1980s spoke English well or even very well; and of those who had been here 30 years or more, only 3 percent reported that they could not speak English well.
Attempts to draw empirical conclusions about the relation between immigration and crime rates founder on problems of measurement. Crime rates rose from the 1960s until about 1990, and since then have declined; there is no obvious link with trends in immigration in this period. Studies at the local level have found no association of immigrant concentrations with crime rates, with the exception of high rates of nonviolent crime near the borders.
Americans have always been ambivalent toward immigration, welcoming flows of foreigners in one era, blocking them in the next. In the past 50 years, polling data have charted a deepening opposition to immigration, linked in part, it appears, to economic concerns. Interethnic tensions have surfaced, especially in areas of high unemployment and poverty. Attitudes are by no means monolithic, however: Americans of African, Hispanic, and Asian descent are more accepting of immigration than non-Hispanic whites are. At present, about 68 percent of non-Hispanic whites favor decreasing immigration, compared with 57 percent of blacks. Asians and Hispanics are even more favorable toward immigration than blacks. Persons with more education tend to accept immigration more than those with less education. Finally, attitudes toward immigrants are no more negative in states with large immigrant populations than in the rest of the country.