New Census data shows that increases in the nation's population since 1990 were caused by higher-than-expected immigration levels.
"For states and cities, this confirms the impression that immigrants in the 90s were going to a lot of cities that traditionally didn't get much immigration," said John Haaga, director for domestic programs at the Population Reference Bureau.
Parts of Arizona and North Carolina are the new magnets for immigrants, Haaga said.
The shift is one of the most significant findings of Census data released last month that asked respondents their birthplace, Haaga said.
The Census 2000 Supplementary Survey asked 700,000 households questions similar to those on the Census long form which covered everything from marital status to household plumbing. Results from the long form, collected from 19 million households, will be made public during the next two years.
The less-recognized supplementary survey, which offers an early peek at expected Census 2000 results, is a pilot program being tested in 1,203 counties.
The data tells states about educational attainment, the number of non-English speakers, and commuting time in their large cities and counties.
For example, states can determine which county has the highest median income, or get detailed information on housing in particular regions. But the data, which cost $27 million to gather, was only collected from selected areas and is already a year old.
Seattle leads the nation in the highest percentage of college graduates, and Newark City, N.J. has the lowest. Santa Ana City, Calif., has the largest percentage of non-English speakers; Cincinnati has the lowest. New Yorkers have the longest commute, at 39 minutes; Wichita residents have the shortest, at 16 minutes.
In August, the Census released state-by-state breakdowns of the data that showed Massachusetts has the highest percentage of college graduates and California has the highest rate of non-English speakers.
The supplementary survey yields a wealth of data that states otherwise wouldn't have. For example, state governments rely on imprecise estimates of poverty in small cities in the years between decenniel Census counts; this data will give states updated poverty estimates.
While the Census 2000 numbers will be used to redraw congressional and state legislative boundaries and to dole out $182 billion in federal dollars for health, social, and educational programs, the supplementary survey's uses are more limited.
The federal Department of Health and Human Services will use the data--- such as the number of working poor receiving food stamps, and the number of children in low-income households that live with both parentsto reward states for high performance under the welfare-to-work program. The Department of Education will use it to determine the number of children with limited English proficiency.