The EPA and conventional air pollution regulations are back in the news. NPR reported that the seven-member Clean Air Scientific Advisory Committee (CASAC), which provides the EPA with technical advice for National Ambient Air Quality Standards, is “considering guidelines that upend basic air pollution science.” But NPR’s oversimplified depiction of a settled scientific debate ignores real misgivings about the science that has justified the regulations and provides an opportunity to ask questions about the proper role of science in public policy.

The pollutant in question is particulate matter (PM), tiny particles or droplets emitted from power plants, factories, and cars. The EPA contends that PM with diameters smaller than 2.5 micrometers, about 3 percent of the size of a human hair, is the most harmful because the particles can be inhaled deep into the lungs. Along with five other criteria pollutants, the Clean Air Act requires that the EPA periodically prepare an analysis that “accurately reflects the latest scientific knowledge” on the health effects of PM exposure. It must then set air quality standards “requisite to protect the public health…allowing an adequate margin of safety.”

Whether one favors leaning towards caution and setting stringent pollutant standards or is skeptical of the efficacy of air quality rules and worries about the costs of the regulations, PM is important. On the one hand, the supposed harms of PM are high. One (contested) study claimed that 2005 levels of PM caused about 130,000 premature deaths per year, which would put PM as the sixth leading cause of death in the United States after strokes. On the other hand, the regulations are expensive. Between 2003 and 2013, EPA regulations accounted for 63–82 percent of the estimated monetized benefits and 46–56 percent of the costs of all federal regulations. The benefits of reducing PM specifically are 90 percent of the monetized benefits of EPA air regulations, meaning PM rules play an outsized role in the justification for many of the costliest federal regulations.

No matter which side of the debate one is on, it would seem important that the EPA have a rational standard-setting process that properly weighs both the possible reduction in the harms of PM and the potential costs. Unfortunately, that is not the case.

The scientific evidence of the harms of PM is much more uncertain than many observers claim and the conflict over what we do and do not know about the effects of PM has existed for decades. The evidence of negative health effects of PM is primarily two studies published in the 1990s, the Harvard Six Cities Study (SCS) and the American Cancer Society Study (ACS). As I have previously noted,

The SCS has been the subject of intense scientific scrutiny and much criticism because of results that are biologically puzzling. The increased mortality was found in men but not women, in those with less than high school education but not more, and those who were moderately active but not sedentary or very active. Among those who migrated away from the six cities, the PM effect disappeared. Cities that lost population in the 1980s were rust belt cities that had higher PM levels and those who migrated away were younger and better educated. Thus, had the migrants stayed in place it is possible that the observed PM effect would have been attenuated.

Furthermore, a survey of 12 experts (including 3 authors of the ACS and SCS) asked whether concentration-response functions between PM and mortality were causal. Four of the 12 experts attached nontrivial probabilities to the relationship between PM concentration and mortality not being causal (65 percent to 10 percent). Three experts said there is a 5 percent probability of noncausality. Five said a 0–2 percent probability of noncausality. Thus 7 out of the 12 experts would not reject the hypothesis that there is no causality between PM levels and mortality.

The latest installment of the debate between supporters and critics of the SCS and ACS studies is the appointment of Dr. Tony Cox as head of the CASAC and the dissolution of a PM advisory subcommittee. Cox has long criticized the science underlying PM standards and argued that epidemiological studies of pollutants have made causal assertions about PM exposure and health outcomes for which the evidence is weak.

To many epidemiologists the appointment of Dr. Cox is akin to putting a creationist in charge of an advisory panel on evolution. In a recent New York Times op-ed, Dr. John Balmes, a former member of the dissolved PM advisory subcommittee, argued,

There has been little dispute that microscopic particulate matter in air pollution penetrates into the deepest parts of the lungs and contributes to the early deaths each year of thousands of people in the United States with heart and lung disease….[Dr. Cox] has been pushing a narrow statistical approach that would exclude most epidemiological studies from consideration by the EPA in reviews of clean air standards.

But even if we accept Dr. Balmes’ view that the science is settled, the EPA’s standard-setting process is deeply flawed. The requirements of the Clean Air Act combined with the attributes of PM ensure that the standards set by the EPA are arbitrary for two reasons.

First, when determining the appropriate level of air quality standards, the EPA cannot consider regulatory costs. In 2001, the Supreme Court ruled that the Clean Air Act “unambiguously bars costs considerations from the [ambient air quality standards]-setting process.” Thus, EPA decisions on pollutant standards can only be about benefits.

Second, the characteristics of PM make it very difficult to identify a suitable standard. PM is a non-threshold pollutant meaning there is no easily identifiable concentration of PM for which any level above it causes harm and any level below it causes no harm. Any concentration of PM other than zero presumably causes some harm, so the exposure standard must be based on other factors.

One logical factor to use to set that standard would be the costs of the regulation, which the EPA is not allowed to consider. Thus, for a non-threshold pollutant in the context of a policy regime in which only the benefits of exposure reduction count, the allowable amount of pollution should be zero. But for political and pragmatic reasons, the EPA cannot set the standards that low; the United States would have to deindustrialize. Instead, the EPA sets the levels at what are essentially arbitrary points.

The establishment of ozone standards, another non-threshold criteria pollutant, between the Bush and Obama administrations illustrates how illogical this process is. Under Bush in 2007, the EPA proposed setting the standard for ozone between 0.070 and 0.075 parts per million (ppm). The scientific justification was the EPA’s interpretation of two clinical studies by Dr. William Adams, which found a reduction in lung function in subjects exposed to 0.080 ppm of ozone. The final regulation had not been issued by the time Obama was inaugurated and in 2010 the new administration proposed lowering the standard to between 0.060 and 0.070 ppm. The justification was still Dr. Adam’s two studies, but the Obama administration reinterpreted those studies and determined that the originally proposed standards were not low enough.

Further confounding the process, Dr. Adams disagreed with both administrations’ interpretations of his findings and argued that his studies did not show any statistically significant relationship between ozone levels below 0.080 ppm and decreased lung function. Two different administrations determined two different standards based on the same studies, and the studies’ author didn’t think his findings justified either standard.

The subjectivity of a process ostensibly based on science raises a question: what is the role of science in public policy? Many seem to believe that “sound science” can and should dictate policy outcomes. Science can inform peoples’ preferences about policies, but science alone cannot dictate which policy outcome to choose. The weighting of costs, benefits, and other normative considerations, such as individual rights and the appropriate use of governmental coercion, require intellectual considerations that are not scientific. Science is a necessary but not sufficient condition for adjudicating public policy questions.

The Clean Air Act and its ban on the use of costs in considering air quality standards implicitly gives “rights” to those who want maximum pollution exposure reduction. Those who would prefer less exposure reduction (either because they believe the evidence for negative health effects from emission exposure is weak or emission reduction is too expensive) have no recourse but to contest the “science” used to rationalize current exposure standards.

Instead of having a never-ending scrum over science, another possibility to resolve environmental quality disputes is to recognize strict environmental rights but allow them to be relaxed in return for compensation. The national SO₂ and California NO_x emission trading markets are steps in the right direction. But I would go one step further and allow the “cap” in those “cap-and-trade” emissions markets to be changed. Those who would like to increase allowable emissions exposure should be able to pay local air quality regions for that change. And the proceeds should be rebated to all residents on a health-risk adjusted basis.

This type of exchange would allow trades between polluters and the most risk-averse. Without such a policy change we will be stuck with an endless political fight over whose science is more “sound.”

Written with research assistance from David Kemp.