A Publication of The Electricity Journal
Volume 8, Number 58 Thursday, March 27, 1997
(Go to Front Page)
EPA based their fine-particle proposal largely on epidemiological studies of six cities. In addition, they commissioned the National Institute of Statistical Sciences to reanalyze prior studies on fine particles and health effects--especially one conducted in Birmingham, Ala. In this analysis, NISS found that when humidity is included as a variable, the adverse effects attributed to particles are simply not there.
The results of this EPA-sponsored analysis were announced by NISS before EPA's new particulate standard was proposed. What account EPA took of this information is not known.
NISS is not a lightweight institution, nor is it an industry arm. Set up in 1991, NISS is "an organization created by the national professional statistical communities to promote cross-disciplinary statistical research in problems of national importance," according to its mission statement. Parent organizations include the American Association for the Advancement of Science, the American Statistical Association, and the International Biometric Society.
"Previous studies, including one in Birmingham, found connections between mortality and particulate levels," said the NISS analysis. "These studies, which received front page coverage in the New York Times, the Washington Post and the Times of London, asserted that more than 60,000 Americans and 17,000 British die from particulate pollution each year. The [NISS] results show that no such connection is tenable. The discrepancy in Birmingham arises from the failure to include relevant meteorology (in particular, humidity). The results of these studies are profoundly sensitive to statistical details ranging from omitted variables to the effects of model selection."
Even when humidity is included, said NISS, the significance of particulates "is often borderline and subject to questions about model selection" and other sources of error. "That such error is present results from the inaccuracy of the measures of meteorology and particulates as reflections of actual exposures experienced by the population."
Humidity has long been known to affect people suffering from lung disorders, and now could prove more detrimental to those people than are particulate concentrations. EPA's claims implicating fine particles, which critics have been calling "junk science," may now be undercut by research funded by the agency itself.
Data for the important six-city studies has not been made available to NISS or anyone else for similar review. This is despite repeated requests to EPA from many sources, including the Utility Air Regulatory Group. EPA has insisted that it does not even have the data, although it was gathered under EPA contract. The researchers who actually collected the data have only agreed to make it available to a single third party, the Massachusetts-based Health Effects Institute, which says it could be years before it releases its findings.
Critics say that EPA is "hiding the data" or "stalling" until the final particulate rules are promulgated. "It is bizarre beyond belief" says Joel Bucher of Citizens for a Sound Economy, "that the scientific data backing up this massive proposal is not generally available for independent analysis and peer review."