Debate continues over use of statistical sampling
By Genaro C. Armas/AP
The Bush administration is committed to using raw census numbers for political redistricting, but may allow adjusted numbers to be used to allocate federal funds among the states, House Republican sources said on Feb. 8.
Democrats contend that the statistical method known as “sampling” protects against an expected undercount of minorities, the poor and children. They said any decision by President Bush to prevent sampling from being used for redistricting would disenfranchise millions of Americans who were missed in the 2000 census.
“The president supports an actual head count, because he believes it’s the best and the most accurate way to conduct the census,” White House spokesman Ari Fleischer said. The administration has not made a final decision on sampling, he said.
But the GOP sources, including a House member, said the administration was committed to using only raw, non-sampled data from the 2000 count as the basis for states to redraw congressional, state and local political district lines. The sources, speaking on condition of anonymity, also said that the use of sampled data for other purposes, including determining federal funds, was “negotiable.”
The Wall Street Journal also reported on Feb. 8 that the White House has “privately promised” to block states from using sampling for redistricting. William Barron, acting director of the Census Bureau, has said he will decide by early March whether the sampled data should be released. The administration could overrule or block his decision.
The numbers to be used for redistricting must be released by April 1.
“This is an issue of great importance that will determine whether millions of people — most of whom are minorities, children and rural residents — will be counted for representation at every level of government,” said House Minority Leader Richard Gephardt, D-Mo. “A compromise that would deny these people representation would mean their disenfranchisement.”
There was a net undercount of about 4 million people in the 1990 census, about 1.6 percent of the population then. More minorities than whites were missed in that count, the Census Bureau said. While the 2000 census is generally considered an operational success, all sides have long said it is virtually impossible to track down every American.
To remedy that, the Census Bureau under the Clinton administration drew up the sampling plan to supplement the actual “head count” of the population as a way to protect against undercount.
Only raw, non-sampled numbers can be used to reapportion the 435 seats in the House among the 50 states, the Supreme Court ruled in 1999. The first release from the census showed a raw count of 281 million people living in the United States on April 1, 2000, the day on which the count is based.
The parties disagree over whether the same court decision left open the use of sampled data for redistricting and the distribution of more than $185 billion in federal funds.
Most Republicans believe the Constitution calls for an “actual enumeration” only, and statistical adjustments could lead to even more errors in the count.
“The question of whether you are going to use adjusted numbers for redistricting is a legal, constitutional and accuracy question. The question of whether you use adjusted numbers for funding purposes is not a legal question,” said Chip Walker, spokesman for Rep. Dan Miller, R-Fla., chairman of the House Government Reform Committee’s census panel.
Some GOP leaders also fear that sampled data, if used, could leave Democrats in a more advantageous position during redistricting, and may threaten Republicans’ narrow control of the House. Republicans will get worried when “America’s black, Hispanic, Asian and American Indian population learn that the Bush administration and their GOP cohorts have decided that they don’t count,” said Rep. Carolyn Maloney of New York, ranking Democrat on the census subcommittee. |
