Skip to main content

EPA’s Proposed Air Quality Standards: Clean Air Sense

Three decades of debate about clean air regulation have done little to temper the acrimonious tone of the discussion. When the Environmental Protection Agency proposed last winter to tighten the National Ambient Air Quality standards for ozone and small particles, angry charges again rang out on all sides. With billions of dollars, millions of illnesses, and tens of thousands of premature deaths at issue, it could hardly be otherwise.

Rather than each side telling us what is in its special interest, I propose to explore what is in the public interest by examining the estimated benefits and costs of the proposed new regulations and probing the uncertainties involved.

Ozone is formed when organic compounds and nitrogen oxides emitted by motor vehicles and industrial sources ( and, yes, trees) react in the summer atmosphere. Small particles result, for the most part, when gases from industrial and utility plants change chemically in the atmosphere. Almost no large urban areas in the United States are in compliance with the current ozone standard; many smaller population centers in the West, South, and Northeast would become noncompliant under the proposed standard. Short of exporting people and jobs, areas such as Los Angeles may never attain the proposed ozone standard. Compliance with both the current and proposed suspended particle standards is much less of a problem.

Industry says it will be hurt by the abatement costs and that its competitiveness will suffer. It charges EPA Administrator Carol Browner with being arbitrary in setting the standards and with relying on inadequate science. Environmentalists counter that the revised standards are long overdue, that scientists have found that the new standards will improve the public’s health, and that EPA should tighten the standards even further given scientific knowledge about the effects of ozone on health. Prominent scientists have added to the cacophony by praising the ozone standard for the quality of the underlying science while suggesting that the small particle standard be delayed until the science is shored up.

Congress gave one definition of the “public interest” in the 1970 Clean Air Act when it instructed the EPA administrator to set air quality standards that protect the public health with an adequate margin of safety. The lawmakers made no mention of the costs of compliance, and EPA administrators claim that they do not look at those costs.

Presidents Ronald Reagan, George Bush, and Bill Clinton have taken a different view of the criteria by which to judge whether proposed regulations are in the public interest. All three presidents have ordered their regulators to prepare a benefit-cost analysis for major regulations. Each wanted a dispassionate analysis of all positive and negative effects of a proposed regulation, particularly for inherently controversial issues.

As required by President Clinton’s executive order, EPA produced a benefit-cost analysis, which led to a different recommendation than the criterion specified by Congress in the Clean Air Act as to what is in the public interest. EPA estimates are annual for 2007 and represent a partial attainment scenario (because the full standards could not be attained with control measures by 2007). For ozone, EPA found that further efforts to attain the current standard had costs that exceed benefits ($1.2 billion, as against $0.1-0.8 billion). Not surprisingly, the proposed more stringent ozone standard also fails a benefit-cost test ($0.1-1.5 billion for benefits, $2.5 billion for costs). In contrast, the new small particles regulation has benefits 10 times greater than the costs ($58-110 billion, as against $6 billion).


EPA is revising the benefits estimate for ozone to account for the effect of ozone abatement on the concentration of particulate matter and to review again the effect of ozone on mortality. While I await the revised estimates, I am skeptical that they will produce benefits that exceed the costs.

The Role of Scientific Uncertainty

Uncertainties in both the effects of air pollution and its valuation are a constant issue in clean air regulation. Uncertainty comes into play in translating emissions into ambient air quality; in estimating the adverse effects on health, agriculture, and the environment more generally; in translating the adverse effects of air pollution into dollars; and in estimating the costs of complying with the standards. The uncertainties for the effects of small particles are much greater than those for ozone.

In the past decade, scientists have made considerable progress in understanding both the photochemistry of urban smog and the effects on people and plants. Ozone damages plants and obscures visibility. It is a respiratory irritant that can stop athletes from achieving peak performance or cause pain when people doing strenuous work breathe deeply. It can trigger asthma attacks. While there is a link between high ozone and hospital admissions and physician visits, in general the effects of ozone are reversible, disappearing within a few hours or days.

By contrast, scientists do not yet understand the mechanisms by which small particles in the air affect health. Dozens of careful studies, however, link small particles to respiratory disease and death in episodes such as the notorious London fog of 1952, on days when particle levels are above average in U.S. cities, and in more polluted cities generally.

A familiar joke about economists is that we are not convinced until we can demonstrate that what is true in the world can be reproduced in theory. Given the many studies linking mortality and morbidity to particulate matter, the joke may be even more appropriate for some health scientists.

Fully understanding the science of health effects is not sufcient for tightening a pollution standard. Abating ozone, for example, is almost certain to reduce asthma attacks, medical visits, and respiratory irritation. But the estimated social benefits of the tightened standard are small and—as noted—fall short of the costs.

The scientific uncertainty that remains about small particles means that the range of benefits from abatement is extremely large. The benefits of meeting the new standard might be $58-110 billion a year, as estimated by EPA, or they might be lower or higher. In my judgment, it is highly unlikely that the benefits of the tightened particle standard would be less than the costs despite the uncertainty involved. Let me illustrate.

The midpoint of the benefit range estimated by EPA is $80 billion; the cost estimate is $6 billion. To take into account the scientific uncertainty, suppose that there is probability P that the benefit would be $80 billion and probability 1 – P that the benefit would be zero. How small would P have to be to justify putting off the more stringent standard?

If expected benefits equaled costs, 80P = 6 or P = 0.075. In other words, if one thought that the chance of particle abatement having health benefits of $80 billion was less than 7.5 percent, one should put off abatement. On the basis of the more than 100 epidemiological studies done by different investigators on different populations, my judgment is that the chance of abatement having these benefits is perhaps 90 percent. Thus, I see the evidence as strongly favoring abatement.

How do we take into account the possibility that industry would have to invest vast resources to meet the standard, all of which would be wasted if particles didn’t have the estimated effect on health? The estimated cost to industry is arrived at by assuming that $30 billion would have to be spent immediately on plant and equipment and an additional $3 billion a year spent on variable costs. (At a 10 percent interest rate, the yearly abatement cost would be $6 billion.) Suppose that the uncertainty over the health effects of particles is to be resolved in ve years—we will know whether the benefits are $80 billion a year or zero. Should we tighten the standard today despite the uncertainty?

For simplicity, I don’t discount the benefits or costs within the ve years. The abatement costs are then 30 + 3(5) = 45. The benefits are 80(5)P. Setting benefits equal to costs and solving for P yields 45 = 400P or P = 0.11. In other words, if we thought there was at least an 11 percent chance that abating particles would lead to this benefit, we should abate today rather than wait for ve years.

While I have met scientists who are uncertain, or even skeptical, that the epidemiological studies can be interpreted as causal, I have met no one who thinks there is less than one chance in nine that the association is causal. Thus, I have not met anyone who would decide to put off abatement according to this analysis. Some analysts assert that EPA’s estimate of benefits is too high, primarily because it assigns a social loss of $4.8 million to each premature death. That number does appear high for cases where life is shortened by days or weeks. If the average benefit of averting premature death were cut in half, the calculations above would result in doubling P to 15 percent in the rst model and 22 percent in the second. I still have not met anyone who has worked with this literature who thinks there is less than one chance in ve that small particles have the estimated effect on health.

A more realistic alternative model would admit that abating small particles would have some effect on health, though one smaller than $80 billion. But that effect would have to be 13 times smaller than that estimated by EPA before abatement today would be contrary to the public interest.

I conclude that EPA should promulgate the new particle standard while continuing to support research to isolate the mechanisms by which small particles harm health and isolate which components of small particles require most stringent abatement. Better scientific understanding of small particles might increase the benefits and lower the costs of regulation by targeting more precisely which particles to abate, but failing to implement the new standard is likely to result in tens of thousands of preventable deaths and chronic illnesses.

Clean Air Costs Money

The Clean Air Act reects the special interests of environmentalists in directing that health be protected without considering abatement costs. When Congress so specied in 1970, it did not anticipate the ingenuity of scientists in tracing health effects at ever lower levels of air pollution. The result is the perennial battle between industry pointing to the costs of compliance and environmentalists saying that a clean environment is a right where abatement costs are irrelevant.

EPA’s careful summary of the effects of ozone finds that the benefits are small compared with those for particles and that the costs of the new standard exceed the benefits. As to small particles, even when the uncertainty over the benefits of the new standard is taken into account, benefits exceed abatement costs. Environmental resources would be much better directed against small particles than against ozone.

If careful analysis shows that a regulatory standard’s benefits exceed its costs, then society gains from the regulation. Of course it should be implemented at the lowest social cost and disruption—though no one should be surprised that a clean environment costs resources. Unlike Oscar Wilde’s cynic who “knows the price of everything and the value of nothing,” we need public policies that serve the public interest.


Get daily updates from Brookings