Census Sampling is Ruled Illegal for Seating Congress

WASHINGTON – The Census Bureau cannot use statistical sampling to enhance the accuracy of the 2000 Census for allocating congressional seats, the Supreme Court ruled yesterday.

The Census Bureau estimates that 4 million people were overlooked in the 1990 Census, primarily city dwellers, children, recently arrived immigrants, poor people, renters and members of minority groups. The estimated undercount represents about 2 percent of the overall population, including 4.8 percent of African-Americans and 5.2 percent of Hispanics.

The Census Bureau estimates that 4 million people were overlooked in the 1990 Census, primarily city dwellers, children, recently arrived immigrants, poor people, renters and members of minority groups. The estimated undercount represents about 2 percent of the overall population, including 4.8 percent of African-Americans and 5.2 percent of Hispanics.

Texas and other states with growing populations could be hurt by the 5-4 ruling, considered one of the most important of the court’s 1998-1999 term.

The census, conducted every 10 years, is used to determine how many representatives each of the 50 states can send to the 435-member House and in disbursing about $180 billion in federal money each year of the coming decade.

High-growth states, such as Texas, California and others in the Sun Belt, tend to gain congressional seats and federal money, and those with declining populations sometimes lose seats and money.Whether Texas would lose money because of yesterday’s ruling is unclear.

The high court, in a 26-page majority opinion written by Justice Sandra Day O’Connor, concluded that a 1976 federal census law “prohibits the proposed use of statistical sampling in the determination of population for the purposes of apportionment.”

The Census Bureau, backed by the Clinton administration, advocates statistical sampling – an estimation of some populations using a mathematical formula based on confirmed census data – to count people who might be overlooked by such traditional methods as mailed questionnaires and door-to-door canvassing.

“This is a loss for the state of Texas,” said Rep. Martin Frost, D-Dallas. “This may cost us an extra congressional seat.”Some lawmakers in the 30-member House delegation from Texas – 17 Democrats and 13 Republicans – had hoped the state would gain three congressional seats as a result of the 2000 Census, one more than it is expected to pick up even if statistical sampling is not used.O’Connor was joined by Chief Justice William Rehnquist and Justices Antonin Scalia, Anthony Kennedy and Clarence Thomas, the court’s four other conservatives.

Opposed were Justices John Paul Stevens, David Souter, Ruth Bader Ginsburg and Stephen Breyer, who argued that

the traditional head count, or “actual enumeration” as cited by the Constitution, will not produce an accurate count.”A face-to-face head count would yield absurd results,” Stevens wrote. “For example, enumerators unable to gain entry to a large and clearly occupied apartment complex would be required to note zero occupants.”

The Census Bureau estimates that 4 million people were overlooked in the 1990 Census, primarily city dwellers, children, recently arrived immigrants, poor people, renters and members of minority groups. The estimated undercount represents about 2 percent of the overall population, including 4.8 percent of African-Americans and 5.2 percent of Hispanics.

Steve Murdoch, director of the Texas State Data Center, the state’s census information center, said, “Texas had one of highest undercount rates in the country” in the last census. “In 1990, 482,000 people were missed, or 2.76 percent,” he said.Murdoch said that without statistical sampling, the 2000 Census will miss about 550,000 people in Texas. Those missed, he said, would be mostly low-income ethnic minorities because they are mobile and hard to reach.

The Census Bureau estimates that Texas has a population of 19.76 million people, an increase of more than 370,000 in 12 months. Texas’ population has increased by 2.8 million since 1990, second only to California.Politics has been very much involved in the debate over statistical sampling. Democrats have backed it because it would mean an increased count of people who tend to vote Democratic. But Republicans contend that the method could lead to errors and the manipulation of census figures for partisan gain. The House GOP leadership was among the groups filing suit to stop the administration’s plans.

White House spokesman Joe Lockhart called the setback for the Clinton administration a “limited decision” and noted that the justices did not say sampling is unconstitutional. That means Congress could decide to amend the 1976 law and support the method.

But such action by Congress seems remote because Republicans oppose statistical sampling. The Republican majority in Congress is also expected to resist efforts to let the Census Bureau employ two differing counting methods.The court drew a distinction between the official census figures used to apportion seats in the House among the 50 states, and any statistical adjustment of those figures for other purposes, such as state redistricting and the distribution of federal money to the states.

Although federal law bars sampling for apportionment, the court said, it permits and perhaps even requires statistical adjustments for the other purposes, an aspect of the decision that Democrats hailed even while acknowledging that any plan for sampling, even a truncated one, remained largely at the mercy of the Republican-led Congress.

“The administration should abandon its illegal and risky polling scheme and start preparing for a true head count,” House Speaker Dennis Hastert, R-Ill., said yesterday after the court announced its ruling.Rep. Joe Barton, R-Ennis, added: “This is the right ruling. I think we need to do our best to count everyone in the United States, but if there are people that just flat don’t want to be counted I don’t think we have to go to extraordinary measures to get to them.”

Barton said he does not fear that a lower census count will hurt Texas’ ability to gain federal dollars. “If the only way to get them is to fudge the numbers, then we should not be getting them in the first place,” he said.Former Texas Attorney General Dan Morales has said that the state could lose $1 billion if the census does not use statistical sampling. Whether that money was lost was unclear yesterday. The court’s ruling was limited to sampling for determining House districts, which are drawn by state legislatures.

Impact on your life Here are some ways the census data affects all kinds of Americans. Child care: Data on the number of working parents, family income and the age of children is used to find areas in need of child care facilities.Seniors: Under the Older Americans Act, money for food, health care and legal services is distributed to local agencies based on census data.Students: The Education Department uses the data to identify school districts and allocate funds to provide help for students most in need, particularly schools with high concentrations of low-income families.Disabled: The Federal Highway Administration allocates money for public transportation for the elderly and people with disabilities based on census data.Job training: The Labor Department uses census estimates to support a program that provides training for young people and adults facing serious obstacles to finding work.Rural communities: The Farmers Home Administration uses census data to allocate money for housing assistance for elderly and low-income individuals and families in rural areas. Community reinvestment: The Treasury Department uses census data to help determine whether financial institutions are meeting the credit needs of minorities and low- and moderate-income areas.

Source: Knight Ridder/TribuneStaff writers Jeff Claassen and Ginger D. Richardson contributed to this report

Author: Michael D. Towle; Star-Telegram Writer – Washington Bureau, 1999