Out for public comment until next month and scheduled to be finalized by June, the proposed standards would cost the American people more than $10 billion annually. Such a staggering sum is not in itself a reason to reject these changes, if making them is essential to public well‐being. But a close inspection of the EPA proposal shows that it lacks a sound basis in science.
The proposed standards cover two kinds of air pollution: ozone (smog) and particulate matter (soot). When the EPA first considered revising these standards, it was, properly, considering the two pollutants separately. But even the EPA recognized that its ozone proposal is not likely to produce sufficient public health benefits to justify its costs. So last May, the agency combined the two proposals, allowing it to use estimates of the tremendous benefits from the proposed particulate standard to cover the costs of the ozone standard.
The EPA estimates that the particulate standard will save 20,000 lives per year. Given that the EPA values a human life at $5 million, the agency estimates that the particulate standard will produce more than $100 billion of annual benefits, far outstripping the proposed rule’s costs. But how certain is the EPA of the benefits of the particulate proposal? Not very, if you consider the principal EPA study supporting the particulate proposal.
In this epidemiological survey, funded by the EPA and published in 1995, researchers concluded that higher death rates exist among populations living in geographic areas with higher levels of particulate air pollution. Specifically, based on a 1982–89 study of some 550,000 adults in 151 metropolitan areas, the researchers claim to have identified a 17% increase in mortality among inhabitants of the most polluted areas in the country.
But the study suffers from a basic epidemiological problem known as the “ecologic fallacy.” This simply means that although the most polluted communities may indeed have a 17% higher death rate than the least polluted area, this coincidence does not by itself demonstrate a cause‐and‐effect relationship between air‐particulate pollution and death rates.
First, although the study included more than 550,000 people, the researchers did not measure how much air pollution even one study subject was exposed to. Instead, they guessed how much pollution these individuals might have encountered.
Second, study subjects undoubtedly differ in many behavioral, occupational, environmental and genetic factors that were inadequately considered by the epidemiologists. Any one of these factors, or a combination thereof, could explain the difference in death rates. For example, the researchers did not look at variances in the subjects’ diet, income, health history, exercise habits or migration characteristics. The researchers did adjust for some factors (education level, occupational exposure and weather), but more precise adjustment could easily negate the purported 17% increase in risk.
A further problem should make us skeptical of the epidemiological study’s findings: It turns out that nobody has demonstrated how airborne particulates could cause higher death rates. Of course, epidemiology isn’t designed to provide information about such biological mechanisms, but the EPA hasn’t come up with any other credible research to fill the gap.
Taken in this context, the reported increase in risk is only an artifact of statistics, called a statistical association. It is not scientific proof that air pollution causes premature death. Among epidemiologists, statistical associations that purport to represent increases in risk of less than 100% are considered to be “weak associations.” As the National Cancer Institute points out: “In epidemiologic research, [risks of less than 100%] are considered small and usually difficult to interpret. Such increases may be due to chance, statistical bias or effects of confounding factors that are sometimes not evident.” And Science magazine noted last year that “many epidemiologists concede that their studies are so plagued with biases, uncertainties and methodological weaknesses that they may be inherently incapable of accurately determining … weak associations.”
In pushing through the new air‐quality standards, the EPA is clearly going against the bulk of scientific opinion. The agency has opted to ignore the advice of its own Clean Air Scientific Advisory Committee, a group of experts from academia and industry whose advice EPA is required to seek under the Clean Air Act. Most members of the advisory committee advised against lowering the ozone standard, concluding that it would provide only marginal public health benefits. Only four of the 21 members supported EPA’s proposed standard for particulate matter.
The method used by the EPA to support stricter air pollution standards is classic junk science — not a new problem at the agency. In the EPA’s 1992 audit of how the agency uses science, a blue‐ribbon panel of independent scientists found that the EPA “lacks adequate safeguards to prevent” science from being adjusted to fit policy.
There is a solution. The responsible scientific community must set standards for the use of epidemiological evidence. Had such a code been established earlier by groups such as the National Academy of Sciences, many unfounded but widely publicized cancer scares, including those surrounding cellular telephone usage, abortion and aspartame might well have been prevented. Unless the scientific community polices itself, the task of controlling junk science‐fueled regulation will fall on Congress and its newfound authority to review regulations — an authority Congress is already preparing to exercise if EPA’s proposed air standards are finalized.
Whether it’s scientists or congressmen, somebody has to rein in the EPA. We already have enough legitimate social and public‐health problems to worry about without EPA and its surrogates fabricating more out of thin — and presumably safe — air.