Census overstated lack of health care coverage

• Article
• Comments ()
• Click-2-Listen
• Videos

By

ASSOCIATED PRESS

The government's estimate of the number of Americans without health insurance fell by nearly 2 million yesterday, but not because anyone got health coverage.

The U.S. Census Bureau said it has been overstating the number of people without health insurance since 1995. The bureau blamed the inflated numbers on a 12-year-old computer-programming error.

The bureau reissued figures for 2005 and 2004 yesterday. It plans to issue new numbers for every affected year in August, when last year's numbers are scheduled for release.

Health insurance statistics are widely cited in debates over the nation's system of health care, which is expected to be a big issue in the 2008 presidential election.

The revised estimates show that 44.8 million people, or 15.3 percent of the population, were without health insurance in 2005. The original estimate was 46.6 million, or about 15.9 percent of the population.

"The total impact is small," said Ruth Cymber, the agency's director of communications.

She said similar reductions are expected in previous years, leaving historical trends unchanged. In 2005, the percentage of people without health insurance was at its highest point since 1998, according to the original numbers.

Workers discovered the programming error when they were updating the computer system for the bureau's Current Population Survey, which yields data on income, employment and health insurance coverage. Some residents were counted as "not covered" by insurance when they had reported coverage. No other questions in the survey were affected, Miss Cymber said.

The error dates to the initial computerization of the monthly survey in 1995, she said.

"While it is certainly good news that fewer Americans are uninsured than previously reported, this raises major questions," said Rep. Carolyn B. Maloney, New York Democrat, who serves on the House subcommittee that oversees the U.S. Census Bureau. "For an agency that specializes in statistics, 1.8 million is not a minor error; it's major error."

She said the errors raise concerns about all census data.

The Government Accountability Office warned in 2004 that the U.S. Census Bureau had inadequate standards for reviewing the quality of some of the data it releases.

"The more time that elapses, the greater the risk of releasing data with quality problems," the GAO said in a report.

Miss Cymber said the bureau has improved its review standards and is updating its technology.

"We now have formal process to make sure that we are using the latest technology," Miss Cymber said.

The bureau reports health coverage statistics by race, age, income, gender and state of residence. The errors do not appear to be concentrated in any group, Miss Cymber said.