WASHINGTON, Aug. 5 - American standards of living increased markedly throughout most of the country in the 1990's, bringing gains in education, housing and mobility along with higher incomes, according to comprehensive new census data being released on Monday.
But if people's homes are bigger, so are their mortgages, and if they own more cars, they also commute more hours to work, the new survey shows. "It's the American dream updated," said Robert E. Lang, an urban analyst with the Fannie Mae Foundation, an independent research organization.
And the survey provides the latest evidence that the magnet of readily available jobs is attracting immigrants at an ever faster pace. About 44 percent of the nation's 30.5 million foreign-born residents, or 13.3 million people, arrived here in the 1990's. Immigrants make up 11 percent of the country's population, the largest share since the 1930's.
Much of the nation's wealth is still concentrated in the Northeast, but that region lost some of its edge in the decade as jobs and people migrated to fast-growing, economically vibrant areas like the Southeast and the West.
It is no surprise, of course, that the latter half of the 1990's in particular was a boom time. But these new figures provide fresh details about the pinpoints of prosperity, especially useful to government planners and businesses because they are available a year earlier than usual after a decennial census. And they show the pockets of impoverishment where governments are struggling to provide services for the needy.
These are among the highlights:
¶In 2000, 82 percent of people 25 and older had graduated from high school and 25 percent had at least a bachelor's degree. In 1990, the figures were 75 percent and 20 percent, respectively.
¶More than 90 percent of households owned a car, van or truck in 2000, the highest share ever, and 18 percent owned three or more vehicles. Seventy-six percent of workers drove to work alone, up 3 percent from 1990. People are spending more time driving to work, often from rural settings where better-paying jobs are scarce. West Virginians logged the largest increase, to an average of 25.5 minutes each way in 2000 from 21 minutes in 1990, as jobs in the Washington suburbs beckoned.
¶The incomes of families grew and so did the size of their homes, as more than one in four homes has seven rooms or more, up slightly from 10 years ago. This confirms data showing that nearly one of five new houses exceeds 3,000 square feet. In 2000, the median monthly mortgage was $1,307, much higher than a decade ago.
In a preview of official 2000 census data to be issued next year from the questionnaire known as the long form, the Census Bureau on Monday will release these and other results of a test survey of 700,000 households conducted simultaneously with the decennial head count.
Census officials want to replace the detailed long form, which yields important demographic and socioeconomic figures once a decade, with an annual survey that would provide similar information every year. In effect, it would be a rolling annual census, but a statistical sample rather than a comprehensive count.
Both questionnaires are designed to measure education, housing, population and commuting characteristics, information that is used by the federal, state and local governments in an array of activities, including planning environmental protections, building roads, and developing health care programs for the elderly.
The survey shows that in 2000, New York became the first state where residents commuted more than 30 minutes on average each way to their jobs. Reflecting population shifts in the West, nearly one in four Nevadans lived in a different residence in 2000 than in 1999. Thirty-five percent of adults in Massachusetts had at least a four-year college degree. And 40 percent of Californians spoke a language other than English at home.
The Census Bureau is asking Congress to make this new method permanent, which would represent perhaps the biggest shift in the federal gathering of statistics since sampling techniques were introduced in the 1940 census. With some lawmakers worried that the new approach might duplicate other surveys or violate citizens' privacy by asking new questions, census officials and their allies hope this kind of useful detail will help them sell the idea.
A yearly survey would "enable local government decision makers to remove some of the guesswork and insert hard evidence into their decision-making process," Barbara Welty, the clerk of Kathio Township, Minn., told the House subcommittee on the census in June.
J. Gregory Williams, the Alaska state demographer, said the new approach would help Alaska officials track the impact of military base closings, assess the impact of floods and other natural disasters, and monitor the social consequences of the volatile logging, mining and fishing industries.
The new approach would save poor rural counties the cost of developing their own programs to estimate data between the decennial censuses, said Robert B. Hunter, executive director of the Hillsborough County, Fla., planning commission.
If approved, the new annual sampling, called the American Community Survey, would by 2004 begin producing yearly estimates for geographic areas and population groups of 65,000 or more. Estimates for smaller areas would be available by 2008. Official population counts would continue to come from the shorter census form that all households are required by law to complete every 10 years.
The data to be released have limitations. Many characteristics, like income and poverty statistics, are not directly comparable to similar data from the 1990 or 2000 counts.
For example, nearly 48 percent of Hispanics responding to the 2000 Census identified themselves as white and 42 percent said they were "some other race." In the test survey, which used highly trained enumerators to follow up with households that did not respond by mail,
63 percent of Hispanics identified themselves as white and 29 percent said they were "some other race."
"It certainly means we need more information about how Latinos are viewing race and what it means about how the census collects that information," said Sonia M. Perez, a demographer with the National Council of La Raza, a Hispanic advocacy group.
That said, demographers who reviewed the test survey results before their public release said they discovered intriguing trends about the nation's standards of living.
There are still troubling signs from the test survey of groups being left behind. Nearly 13 percent of the country was living in poverty in 2000, down slightly from 1990 levels that are not strictly comparable. Among the poor are 17 percent of children under 18 and 11 percent of adults 65 or older. Seventeen percent of households received some sort of public assistance. About 530,000 homes lacked plumbing and 3.2 million households did not have telephone service, although both figures represent sharp declines from 1990 levels.
In a surprising development to demographers that may reflect housing shortages in the booming economy or more profound changes in the nuclear family, 42 percent of grandparents living in the same house as their grandchildren were responsible for the youngsters. More than one-third of those grandparents had been taking care of the grandchildren for five or more years.
"In terms of looking at family situations," said Martin O'Connell, chief of the Census Bureau's fertility and family statistics branch, "this shows that we have to extend our horizons beyond Mother and Dad."
Over all, though, the test survey results show a prosperity that is shifting geographically. Eight of the 20 states with median household incomes above the national average of $41,343 in 2000 were in the Northeast. New Jersey ranked first with $54,226, edging out Connecticut, with $53,108.
Wendell Primus, director of income security at the Center for Budget and Policy Priorities, a research group, noted that a swath of Northeastern states from Maryland to Vermont lost ground over the decade in household income.
The booming economy attracted many new immigrants, who arrived lawfully and unlawfully. Jeffrey S. Passel, a demographer with the Urban Institute, a social policy research group, said the test survey results supported his estimates that eight million to nine million immigrants now live illegally in the United States, including 4.5 million to 5 million Mexicans. The figures for both categories are about 1.5 million more than the Immigration and Naturalization Service estimates.
Reflecting this new diversity, 18 percent of all residents who were at least 5 years old spoke a language other than English at home, up 4 percent from 1990. Of those people, 60 percent spoke Spanish, and 43 percent reported that they did not speak English "very well."
Immigrants are also dispersing across the country, as earlier census figures had shown. According to William H. Frey, a demographer at the University of Michigan, the populations of 27 states are now at least 5 percent foreign-born, compared with 14 states in 1980. States with large populations of new immigrants tended to ride the economic upswing. In a dozen states, including Nevada, more than half the population was born in other states or overseas, Mr. Frey said. Those states tend to have younger age structures and are more economically vibrant. Conversely, he said, those states with large proportions of residents who were born in the state, like Louisiana and Pennsylvania, have fewer migrants and have been economically stagnant for decades.
In another measure of mobility, about 15 percent of the population lived in a different residence in 2000 than the previous year. In five Western states - Nevada, Alaska, Oregon, Colorado and Arizona - more than 20 percent of residents had moved households within the year, reflecting overall growth as well as transient college and military populations.
As the economy thrived, the nation's average commuting time to work increased to 24.3 minutes in 2000, from 22.4 minutes in 1990 and 21.7 minutes in 1980. (The 2000 figures are slightly inflated because of a difference in counting methods, but they still show that Americans are spending more time in their cars.) In 2000, 14.6 percent of workers 16 or older had commutes of more than 45 minutes, compared with 12.5 percent of commuters a decade ago.
"The trick with these numbers is that you don't know whether the same trip is taking longer or people are taking longer trips," said Alan E. Pisarski, a transportation specialist in Falls Church, Va., and author of "Commuting in America." "It's very likely a little bit of both. "
Car-pooling continued to decline, to 11 percent from 13 percent in 1990, while use of public transportation to get to work hovered just above 5 percent, unchanged from a decade ago. Major exceptions were in the District of Columbia (35 percent), New York (27 percent) and New Jersey (11 percent), all of which have extensive bus, subway and commuter rail systems.