The U.S. Chamber of Commerce has red-penciled what it views as errors in 16 databases that the Environmental Protection Agency uses to make regulatory decisions about chemical cleanups and other policies that affect states and business. The chamber is not accusing the agency of manipulating the data, but of negligence in assuring users that the numbers are reliable.

Because the numbers are key to creating mathematical models that predict the behavior of various chemicals in the environment -- for example how they might move in soil or how concentrated they might be -- the group has asked for corrections. The issue is important to the business community, the chamber argues, because determinations based on these databases might result in rules that are too protective, and hence more costly.

In filing a petition at the EPA a year ago, the chamber noted, for example, that a numerical value for a benzene compound differed by a factor of more than 10 billion from one EPA database to another. Such discrepancies, it said, have led to "serious mistakes and erroneous conclusions" affecting estimates of how chemical exposure affects human health, what prevention measures are necessary and how much solving the problem should cost. It also charged that EPA's peer reviews did not uncover the mistakes.

The powerful business lobby told EPA "erroneous or inconsistent parameters in these databases" can have "huge financial implications" for its members who have to pay for Superfund cleanups or comply with other regulations that may be overly protective.

The chamber invoked the 2002 Information Quality Act in seeking the changes. The law requires federal agencies to respond to complaints directed at the reliability and accuracy of information they disseminate to the public and use in public policymaking decisions.

"We aren't trying to force withdrawal, but to get the numbers right," said William Kovacs, chamber vice president of environment, technology and regulatory affairs. "We want them to 'fess up and get it right. Our contention is if they disseminate this data, they are responsible for its correctness."

The EPA, by and large, did not buy the arguments.

In a Jan. 12 response, the agency said in general, the databases and models conformed to EPA information quality guidelines. It said it added two sentences to one database to "clarify appropriate use of the data" and it removed for review another used to assess how chemicals move in the soil.

The chamber came back to EPA in April and asked it to reconsider, an appeal allowed under the law. "EPA's response . . . is alarming in that it rejects a requested review of erroneous data, largely disclaims or ignores the fact that problems exist, and blatantly fails to address the public need for quality information . . . thereby placing the onus for examining and assuring data quality on the users," the document said.

The trade group asked the agency to convene an interagency task force, including officials from the National Institute of Standards and Technology, to formulate a way to assure the information is consistent and reliable.

Reginald Cheatham, director of EPA's quality staff, said the agency will convene an executive panel of EPA senior managers to review the appeal and will have an answer to the chamber by July. "We are working on it here and in the regions. We have at least 16 [people] working on this," he said. He said he could not address the debate on the merits or draw any conclusions while the agency is deliberating.

Cheatham said the agency took a "proactive step" in adding a disclaimer on one of the databases and removing another. "We have to recognize that EPA is continuing its effort to provide high-quality data and information in all its decisions," he said.

Since the information quality law was implemented in 2002, the EPA has had 28 requests for corrections. The agency said it made changes in some of the challenged information, but was precluded by law or regulation from granting relief in many of the cases. Nine of the decisions have been appealed.

Overall, various interest groups have filed scores of such challenges at other government agencies. Most of the petitions are denied, experts said.

The chamber's interest in this issue was piqued by findings in 2001 of two scientists at the U.S. Geological Survey. They discovered discrepancies in the values assigned to various physical and chemical properties of the pesticide DDT.

Some public interest groups and academics criticized the increasing number of petitions filed, though they have made some challenges themselves.

Sean Moulton, senior policy analyst with OMB Watch, which monitors federal regulation and policy, said the chamber petition is a substantive one. But he adds the changes it seeks are unrealistic because the agency can't afford the time or expense of revamping the databases. Correcting the errors would take EPA away from other priorities, Moulton said.

The Center for Progressive Regulation, a liberal think tank, said in a March paper that "disgruntled industries have used the Act as an end-run around well-established procedures for promulgating rules to improve air quality, clean up toxic waste sites, and protect children and wildlife from pesticide residues."

Still, Rena Steinzor, a law professor at the University of Maryland and one of the authors of the paper, said the chamber petition has legitimacy and identifies a problem that needs to be corrected. But she said a more efficient solution would have been for the chamber to meet with EPA first.

"Don't they have access to these people?" she said.

Kovacs said other users of the data raised the issue earlier and EPA did not make corrections. The chamber, he said, also tried to interest John Graham, head of the Office of Management and Budget's regulatory shop, in the problem with no success. Graham responded: "OMB has an open mind as to whether an interagency review of this matter is appropriate. That decision should await EPA's response to the chamber's appeal."