More on EPA and Air Pollution: Junk Science and Legal Precedents

Related articles

In a previous piece, I criticized the science behind some pivotal air pollution studies. In this follow-up, I look at some of the legal consequences of that bad science.

The decision to promulgate new ozone and small particulate limits by then-Environmental Protection Agency (EPA) Director Carol Browner in 1997 has a storied history, including multi-agency protests, private and public foundation criticism of Browner's use of the Pope(1) and Dockery(2) studies -- and even a rejection of the EPA actions by the District of Columbia Federal Circuit Court in 1999 (a technical administrative law landmark), admonishing the EPA that it had overstepped its regulatory authority.(3)

Subsequently, the U.S. Supreme Court unanimously held that the lower court was right, but Justice Scalia authored a complex opinion that gave too much deference to EPA regulations. The big issue of bad ozone and small-particulate medical epidemiology got lost in the question of how to gauge the "best available technology" and cost consequences of regulatory actions, and what exactly agencies can do in the business of promulgation of rules and in enforcement. Instead of fighting bad science, the opinion focused on the limits to agency independence in considering the costs and consequences of its rules.

The same court that wisely gave us limits on junk science (in Daubert v. Merrill Dow 509 U.S. 579 [1993] as well as Kumho Tire 526 U.S. 137 [1999] and Joiner v. General Electric 188 S.Ct.512 [1997]) has now enshrined agency discretion to use junk science in justifying regulatory action, without recourse for those targeted by the regulations. True, Scalia reprimanded the EPA for exercising undelegated congressional legislative authority, authority not laid down in the language of the Clean Air Act.

The Scalia opinion said that the EPA did not have the authority to decide how to balance health benefits and industry's compliance costs -- only limited authority to determine health effects and analyze the question of Best Available Technology (BAT). The DC Circuit opinion had merely suspended final disposition with the order that the EPA correct its mistakes on overreaching. The Supreme Court, however, said that EPA couldn't correct something it had no authority to do and should simply stop. The case was remanded to the DC Circuit.(4)

Proposed: A New Strategy for Coping with EPA

If litigants in a lawsuit against the EPA were to show an invalid, arbitrary or unjustified use of science, or disqualify the science under the Daubert and Federal Rules of Evidence 702 tests for admissability in Federal courts(5), then the EPA or its state agency surrogates would be stopped dead in their tracks. The industry organizations who were the plaintiffs in the case above, fighting the '97 regs, did not flesh out their complaints about the bad public health science (the Six City and Pope studies) well enough to create a focus on that issue. The time will come for another try.

In the future, the science of regulatory activity must be more vigorously challenged, so that bad public health research can no longer be used as propaganda or as a legal and regulatory club. The basic research has to be put to the test, along with the risk and benefit analysis. Risk-benefit research is possible. The value of years of life saved or lives saved is now an important factor in policy making, well recognized in agency benefit analysis. That said, the public and interested parties have failed to test agencies adequately on the numbers and the rationality of agencies' risk assessment (such as use of the precautionary principle -- no acceptable risks and no cost limits -- which can cost billions per death avoided or year of life added).

Director Browner made her 1997 decision against the dissent by the EPA's own advisory panel on air quality; the Clinton administration Small Business Administration; the Departments of Agriculture, Treasury, and Transportation; the Office of Management and Budget; and the Council of Economic Advisers, all of whom criticized the validity of the Pope and Six City studies.(6) How long does the list of opponents have to be before EPA discretion is trumped by science and good sense? The economic costs of the regulations were projected by private groups at 75 to more than 130 billion dollars per year -- and the weaker the science, the more the economic effects ought to be considered.(7) Browner acted without any concern about weak science or risk and benefit, meaning the EPA could have been attacked on the merits, not just on the Federal Court's reading of the constitutional limits on Congressional delegation.

The weak Samet,(8) Pope,(9) and Six City(10) studies provide opportunities to show how the scientific community has been corrupted by the aggressive environmentalist political agenda. The bad science and poor epidemiology done in the Six City (1993) and Pope (1995) studies, which were the basis for the original Browner/EPA regulatory proposal in 1997, must be subjected to a spirited challenge on behalf of the citizens who will bear the risk and the burden of any new ambient air limits.(11)

Could Samet, Pope, and Dockery Testify in Federal Court?

The Samet, Pope, and Six City studies would fail a basic test for admissability in a Federal Court, following the opinion written by Justice Blackmun for the U.S. Supreme Court in Daubert(12). There are so many problems with the air pollution death studies that a short list of the important issues is in order:

--Outside air pollution varies during the day and from place to place in a city, and if measures of particulates are not even available daily to the air pollution death researchers, how can a study of the daily effect of particulates be valid?

--It is difficult to measure the toxic effect of an agent at non-toxic levels, particularly the effect on mortality, which is always multifactoral.

--Short-time-frame and less-than-one-month-effect mortality studies of uncontrolled populations ignore basic problems with how toxins or irritants cumulatively effect biological organisms. Dust is not poison, and other pollutants are not poison at current exposure levels.

--Incalculable exposures make for uncertain conclusions in any uncontrolled population study.

--There is no reliable way to know the exposure history of a particular dead person from commonly-available mortality data, not even where that person lived or spent most of his time, whether inside or out of doors.

--Changes in absolute mortality rate of less than 1%, as in all of these dust and pollution death studies, as measures of particulate pollution effect, are not statistically adequate to show causal relationship, but that problem has been hidden in these studies through the confusing concept of "association." (There is an association between ice cream consumption, baseball, and drowning, for instance -- since all are more common in warm weather -- but no actual causal effect among the three.)

--The studies sometimes even claim a causal relationship to morbidity, but that is just as difficult to study, given variables and confounders. Moreover, mortality should not be used as a surrogate for morbidity, even for the non-clinician public health scientist. Public health must follow some medical scientific principles, not just speculate and paw at the data for things to say.

--Terminally ill persons do not typically spend time outside, so outside air pollution is not likely a factor in their final demise. The authors have failed to deal with the most obvious problem in their studies, the differences in outside and inside air quality.

--Since epidemiology rules require relative risk of death increases of 200 to 300% to show causation proof (as distinguished from absolute changes in death rates) and these air pollution studies fail that test, the best that can be said is that it if one dredges the data enough, eventually one might find a suggestion, an "association," that justifies more studies but doesn't scientifically prove anything.

--Repeated studies with current data would be impossible, since air quality is improving steadily and cigarette smoking is decreasing.

--Most deaths from cardiovascular and respiratory causes in America are multifactorial deaths that come slowly over many years, with stops and starts caused by medical interventions. How can death studies be a serious way to study the effect of pollution?

--Heart disease and lung disease have little to do with outside dust. Even Samet s study shows no association with ozone or other pollutants. Crude population studies like those done by Samet, Pope, and Dockery cannot hope to provide an accurate measure of death effects of dust or any other minimally irritating or polluting substances in the air.

The EPA's False Conception of Toxicology

Dr. Ware in his NEJM editorial(13) accmpanying the Samet study says that "evidence suggests that the association between fine-particle concentrations and mortality is linear across the entire range of current concentrations." This is the scientific fallacy most important for maintaining the environmentalists' precautionary principle. The concept of linear or straight-line effects appeared in the past as part of EPA arguments on pesticides, radon, lead, arsenic, and many other toxins, without explanation or excuse.

Toxicology always includes the concept of threshold -- some point below which no biological effect is thought to occur -- and always shows both non-linear and linear effects. The air pollution death research suggests linearity in the ranges studied because the measurements are limited in scope and crude at best, but linearity of adverse effects in toxicology is unheard of in the toxicity ranges -- and non-existent in the non-toxic ranges. For example, there is no effect of Tylenol below certain levels but geometric increases in negative effects above threshold. The same would be true of inhaled toxins: nil at low levels but rapidly rising effects in toxic ranges, as the toxin overcomes host repair and resistance factors.

Toxin effects on living organisms are linear in some ranges, nonexistent in other ranges, and geometric in high toxicity ranges. At sub-toxic levels there is no linearity of effect at all, whether for dust or even the potent inhalant poisons. There is always a toxic threshold level below which no serious negative effects occur, even for the most potent toxins such as arsenic, hydrogen sulfide, cyanide, phosgene, or the potent irritants like ammonia, chlorine, nitrogen dioxide, or sulfur dioxide.(14) (To complicate the matter further, every patient has a different threshold based on size and underlying health status.)

Basic toxicology rules are anathema to the EPA because the rules create so many barriers to aggressive regulatory intervention. Thus, experiments in which fifty-gram mice get artificially high-dose exposures are extrapolated to create warnings about low-dose exposures for 100,000-gram adult males (see ACSH's new book about this basic problem with much of EPA and FDA science, America's War on "Carcinogens"). Straight-line toxic effects are a concept that circumvents all the inconveniences of science, and substitutes EPA junk science for the real thing. Dust research is worse, since dust has no toxic attributes. Dust is not a poison or toxin, nor is ozone; they are merely irritants.

Big Pollution Trends and Small Particles

Ignoring the reduction in real air pollution in American cities, the environmentalists charge on and blame it for all ills -- even though asthma rates, for instance, increase as air pollution declines. Air pollution of all kinds in American Cities has been on the wane since the 1950s, with particulates down as much as 80%.(15) We are now left chasing a phantom killer created by junk science public health data-dredging.

Air pollution today is 25% of what it was in 1970, and all other major metropolitan air pollutants continue to decline, due to a very successful reduction program and changes in industrial activity.(16) However, the EPA is unlikely to announce an end to air pollution problems when the agency and related state agencies comprise a major jobs program and locus of environmental-political power. The EPA itself has 18,000 employees and a budget of more than $6 billion. Add to that fifty agencies at the state level -- and all the private compliance activity -- and you have some serious economic impact.

That is why dust is so important and such a good target and why small particles have developed a prime importance in the lexicon of the environmentalists -- if small particles are assumed to be deadly, the EPA and environmental groups have a demon that will never go away. Dust is everywhere and always will be, so the EPA will always appear relevant.

The Hidden Treasure in the Samet Study

The 2000 Samet study actually shows that most pollutants that have been demonized over the years -- sulfur dioxide, nitrogen dioxide, carbon monoxide, and ozone -- have no "association" with mortality rates in the twenty cities studied. This finding is the most important in the study and is ignored by not only the study authors themselves but Dr. Ware in his editorial, since environmental-political attention has turned to new (and even less plausible) targets. Why no comment other than in passing from the authors and no comment at all from editorialist Ware?

This evidence for no mortality association from ozone, carbon monoxide, and the dioxide ozone precursors creates a large hole in the current air pollution crusade, and suggests complete reevaluation is needed. The offense of these pollutants to society's sense of smell is not enough to justify massive, expensive disablement of economic activity. Aesthetic sensibility offenses require a different measure of risk and benefit, if there is no measurable health or mortality effect. In such a scenario, other economic and social considerations become more important: where to spend money more wisely and how to balance aesthetics with regulatory or property or freedom costs. Air pollution research is now a weak and deceptive attempt to prop up an aesthetic crusade, much as worked-over secondhand smoke studies were sometimes misused by opponents of smoking.

We must fine-tune air pollution in this society without stifling all activity in society. Human activity produces dust, and modern energy production and industrial activity produce other air components. We are not in the Stone Age. Some improvements in air can be credited to the Clean Air Act, but many trends were well developed before the major clean air amendments were passed in 1970 and were self-imposed changes designed to improve quality of life. We no longer tolerate horse manure on the streets, and we began using toilets and trash pick-up before the environmentalists came on the scene.

The desire for environmental purity and NIMBY (not in my back yard) opposition to both technology and clean-up efforts have created some new crises in energy, but they must not be allowed to drive society backwards. The quality of life in America will suffer if we don't develop reasonable environmental approaches and shut down the hysterics and the crisis-mongering.

1 Pope CA, Burnett RT, Thun MJ, et. al. Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA 2002;287:1132-1141.
2 Dockery EW, Pope CA3d, XU x, association between air pollution and mortality in six U.S. cities. N Engl J Med 1993;329:1753-9.
3 American Trucking Associations, Inc. v United States Environmental Protection Agency 195 F.3rd 4 (DC Circuit 1999).
4 Whitman v. American Trucking Associations, Inc., 121 S.Ct. 903, 531 U.S. 457, (2001)
5 Daubert v Merrell Dow Inc. set out admissability of scientific evidence tests that eliminate junk science from the courtroom.
6 Antonelli A, Federal regulators disagree about the EPA's new clean air rules. 1998 publications in a packet by the Heritage Foundation Washington D.C.
7 The annual cost estimates by the Washington University at St. Louis Center for the Study of American Business up to 138 billion by Heritage Foundation.
8 Samet JM, Dominici F, Curriero FC Fine particulate air pollution and mortality in 20 U.S. cities, 1987-1994. NEJM 2000;343:1742-9.
9 Supra i.
10 Supra ii.
11 The Pope and Dockery/Pope studies have been widely criticized and the researchers have refused to release their raw data. Senator Richard Shelby introduced legislation aimed at forcing the release of raw data from studies used in agency rule and regulation decisions, but the raw data from the Pope and Dochery studies the EPA used has never been reviewed by opponents or independent experts.
12 Daubert v Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993).
13 Ware JH. Ed. Particulate air pollution and mortality--clearing the air. N Engl J Med 2000;343:1798-9.
14 Goldfrank LR, Flomenbaum NE Lewin NA EDs. Goldfrank's Toxicologic Emergencies 4th ED. Norwalk Conn. Appleton and Lange 1990.
15 Environmental Protection Agency, National Air Quality and Emissions Trends Report, 1994 Data appendix, Office of Air Quality Planning and Standards, EPA 454/R-95-XXX, (1995).
16 Supra note 6.

Dr. Dunn teaches and supervises in the emergency department residency program at Darnall Army Community Hospital, Fort Hood, Texas. He taught Environmental Law at Howard Payne University in his home of Brownwood, TX and continues to research environmental science and law. He is a non-practicing member of the Texas and Louisiana Bars.