The smoking gun - passive smoke hazards - includes related articles on cigarette advertising, on exporting cigarettes, and on environmental tobacco smoke - Cover Story

E: The Environmental Magazine, Oct, 1994 by Alice Horrigan

Even these risk levels, however, have been called into question by independent observers. No one is surprised (or pays much attention) when the tobacco companies and other pro-smoking groups criticize the EPA's risk assessment as "creative epidemiology" full of statistical confidence tricks. But in an April report, the Congressional Research Service (CRS)--a non-partisan arm of the Library of Congress--joined these groups in suggesting that the EPA assessment may have gone too far in finding a causal relationship between ETS and cancer, rather than a less-certain "association" between the two. Titled "Cigarette Taxes to Fund Health Care Reform: An Economic Analysis," the CRS report examines whether the proposed 75-cent excise tax on cigarettes to help fund President Clinton's health plan is sound public policy. CRS looked at the EPA report--and roundly criticized it.

"If you read the newspapers," says Jane Gravelle, CRS senior specialist in economic policy and principal author of the report, "your conclusion about passive smoke would be that it's a very certain and very significant risk." But, she said, "My understanding after doing this research is quite different from my understanding as just an observer."

The CRS report, whose second author is public-finance specialist Dennis Zimmerman, points out that the EPA assessment is based on "a group of 30 studies of which six found a statistically significant (but small) effect, 24 found no statistically significant effect, and six of the 24 found a passive smoking effect opposite to the expected relationship." To standardize these diverse studies, the EPA "adjusted (weighted) the estimate of passive-smoking effect in each study." This de-emphasized studies that found less significant effects for passive smoking and emphasized those that found more significant effects, says the report.

Gravelle and Zimmerman also note that the EPA changed the standard for statistical significance used in the original studies. They write: "...it is unusual to return to a study after the fact, lower the required significance level, and declare its results to be supportive rather than unsupportive of the effect one's theory suggests should be present."

Surprisingly, the CRS report has received very little media attention. "I think criticizing the popular wisdom on ETS is a very unpopular thing to do right now," says Gravelle. "But all we're saying is it's not certain that there is an effect from passive smoke."

EPA report author Steve Bayard says that the agency "flatly rejects the CRS complaints and stands by its report." Gravelle and Zimmerman, he said, "don't have the expertise to really do an adequate criticism." The EPA assessment went through four years of development and review--including guidance on how to analyze the ETS studies from a panel of 18 scientific experts, Bayard said. Even the tobacco industry was given a chance to contribute to the process.

Bayard also defends the changes the EPA team made to the statistical-significance standards as "a practice that is not unusual in meta-analysis; we had to choose our own standard to be consistent." The dispute leaves average folks who don't know the difference between one-tailed and two-tailed statistical tests two choices: to have faith or not have faith in the EPA's risk projections.