Get the NewsHour History:
Video Update: The controversy over Census 2000, and whether the numbers should be adjusted to compensate for those not counted. (3/1/01)
2000-- avoiding an undercount.
statistical sampling question.
A multi-racial Census?
The results of Census 2000 are in: America is changing.
The results of the nation's giant once-a-decade head-count will start to be officially announced this month, but word of the findings has already trickled out.
There are now more than 281 million people living in the United States -- 3.7 million more than the last time we counted in 1990. But the census doesn't just tell us with how many people live in the U.S., it reveals important information about income, education, race, age, family size, and travel patterns.
According to the 2000 Census, about 9.3 percent of people in the U.S. were born somewhere else. (That's actually a lower percentage of foreign born residents than was reported 100 years ago at the height of immigration.)
Asian-Americans probably now have higher rates of college graduation than Americans of European descent. Over 24 percent of Latino immigrants lived in poverty in 1999, compared to 9.3 percent of European immigrants and 12.8 percent of Asian immigrants. By contrast, black Americans are doing better financially than other groups. The average income for black families in 1999 increased by nearly eight percent from 1989 -- a greater increase than in other racial groups over the past decade.
Information from the census is used by thousands of government agencies and private researchers. One of the most important uses of census data is for distribution of the 435 members of the House of Representatives. The number of representatives a state gets depends on how many people live there.
States that lose population will lose a seat in the House. This makes the census, and the methods by which it is conducted, a hot political issue. Politicians seek favorable political boundaries as well as distribution of federal dollars to states and local municipalities. So, Senators and Representatives quarrel over the census in Congress. In fact, some lawsuits have already begun.
Horseback to the Internet
During the first
U.S. census in 1790, under the direction of Thomas Jefferson, federal
marshals traveled by horseback from town to town to count the U.S. population
in each state. The newly formed U.S. government used this data to pay
back the states for the expenses of the Revolutionary War. Also, they
had to comply with the mandate of the U.S. Constitution that a census
is taken every 10 years.
Census 2000 also introduced limited use of the Internet to gather census data. Yet even today some aspects of the census remain low-tech.: 860,000 census workers visited many households in person to collect census questionnaires.
The Changing Face of America
From census data we know that in 1900, less than 5 percent of the American population 25 years and older had graduated from high school; by the late 1990s, that number had jumped to 82 percent.
In 1900, the census questionnaire offered only 5 "color" or race options: white, black, Chinese, Japanese, and (American) Indian. One hundred years later, the 2000 census listed 63 different racial categories, and for the first time allowed respondents to check more than one box to describe their race.
The data gathered from Census 2000 will shape political boundaries and determine the distribution of billions of dollars in federal funding until the next census in 2010.
It is very important for a state to have its entire population counted, otherwise it might lose money as well as representatives in Congress. In addition, politicians use census data to determine the political boundaries for literally thousands of state and local legislative seats.
The federal government uses census data to distribute money for social programs, hospitals, roads, schools, and other government projects. Following Census 2000, the federal government will distribute about $182 billion to the states; about half of this will be distributed based on census population data.
People also use census information in their daily lives. Developers of shopping malls use census data to determine where the richest people live. Rescue workers use census data to conduct disaster relief. When Hurricane Andrew devastated Florida in 1992, rescue workers used census information to estimate the number of people missing block by block in towns in south Florida.
The census is not perfect, and as a result some people do not get counted. The Census Bureau estimates that the 2000 count missed nearly 3 million people, mostly minorities, poor people, and children. People in rural areas, inner cities and homeless shelters are often accidentally left uncounted because they are harder to reach. Many Americans living overseas were not counted for the same reason.
Overseas, only Americans who are military or government personnel are included in the U.S. census. Utah, for example, had over 14,000 residents living temporarily overseas in 2000 as Mormon missionaries; they were not included in Census 2000.
In the distribution of congressional seats according to population, Utah fell 56 residents short of gaining one more seat in Congress. Instead, this seat went to North Carolina. On January 11, 2001 the state of Utah sued the Census Bureau in federal court, and is trying to have this Congressional seat transferred from North Carolina to Utah.
Fewer people were left uncounted in 2000 than in 1990, but the impact of the undercount is far-reaching. If some segments of the population are undercounted, the areas where these groups live will get less representation and fewer federal dollars than they deserve.
When the homeless are not counted in the census, cities get less federal funding to support social programs such as shelters and health clinics. If the city cannot show that it has people who need these services, the federal government will not provide money to support them.
Census 2000 tried to count these people by distributing census questionnaires at housing shelters and health care clinics and on the Internet.
There are two types
of census questionnaires. One is the "short form" which gathers
information about characteristics such as age, sex, and race. The other
is the "long form" which also collects information about marital
status, education level, disabilities, economic status, and housing.
The Sampling Controversy
One of the hottest census controversies is the use of the statistical method known as "sampling." It's a way of trying to adjust the count to make up for the people who were likely to be left out.
After the initial census count on April 1, 2000, the Census Bureau sent workers to revisit some neighborhoods in person and gather detailed information about these "sample" areas.
The Census Bureau compares the original data from the questionnaires with the intensive second-round sampling data collected by census workers to determine who got missed the first time around. Based on this sampling, the Census Bureau would then adjust the census figures in an effort to make the entire census more accurate.
But the adjustments for all census data resulting from sampling are based on estimates rather then an actual count of each individual. And critics say the selection of "sample" neighborhoods can be subjective, based on human decisions, rather than science.
Even the Supreme Court has weighed in on the sampling controversy, ruling in 1999 that the Census Bureau may not use population figures determined by sampling to set the political boundaries for Congressional districts. The new Secretary of Commerce, Donald Evans, recently announced that he will decide soon whether "sampling" will be used to adjust the original Census count.
How many we are, who we are
Counting Americans and gathering information about how they live cost the U.S. government $6 billion in 2000. This was an investment in understanding not only how many of us Americans there are, but also who we are.
More information will be released on the racial make-up of the U.S. Researchers expect it will suggest that white people are not going to be the majority group for much longer. If population trends continue, there will no longer be one majority racial group at all.
The data will also tell us whether there are specific parts of our population who need more economic help. Without the census, we would not be able to take a good look at who we are as Americans and how the country is changing as time goes by.
-Contributed by Jennifer Bryson
Copyright © MacNeil-Lehrer Productions All Rights Reserved