Tobacco Documents Online

deTOCQUEVILLE SCIENCE, ECONOMICS, AND ENVIRONMENTAL POLICY: A CRITICAL EXAMINATION A research report conducted by the Alexis de Tocqueville Institution , August 11, 1994 J T11630-2044

TABLE OF CONTENTS CASE STUDY NO.1: ENVIRONMENTAL TOBACCO SMOKE .................................................................... 1 CASE SPUDY NO.2: RADON ..........................................................................................••---... ............_..............._. 15 CASE STUDY NO.3: PEST'ICIDES ........................................................................... ............................................. 33 CASE STUDY NO.4: SUPERFUND .................................................................................................... .................. 45 I T11630-2045

SCIENCE, ECONOMICS, AND ENVIRONMENTAL POLICY: A CRITICAL EXAMINATION A research report by the Alexis de Tocqueville Institution Academic Advisory Board Dr. Gary Anderson Professor of Economics California State University-Northridge Dr. Nancy Bord Visiting Scholar The Hoover Institution Stanford University Dr. Gordon L. Brady Associate Professor and Director Environmental Studies Sweet Briar College Dr. Michael Marlow Professor of Economics California State Polytechnic University-San Luis Obispo Dr. Thomas Gale Moore Senior Fellow The Hoover Institution Stanford University Dr. Malcolm Ross Research Mineralogist U.S. Geological Survey Dr. Jeffrey Clark Professor of Economics University of Tennessee-Chattanooga Dr. Michael Darby Professor of Economics and Director J.M. Olin Center for Policy University of California, Los Angeles Dr. Robert Ekelund Lowder Eminent Scholar Auburn University D r. Michael Gough Project Director Congressional Office of Technology Assessment Dr. William Hazeltine Environmental Consultant Dr. Thomas Hopkins Gosnell Professor of Economics Rochester Institute of Technology Dr. Dwight R. Lee Ramsey Professor of Economics University of Georgia Dr. S. Fred Singer Professor Emeritus of Environmental Sciences University of Virginia and President Science and Environmental Policy Project Dr. Gerhard Stohrer Director of Chemical Risk Program Science and Environmental Policy Project and former Department Head Sloan-Kettering Institute for Cancer Research Dr. Mark Thorton Professor of Economics Auburn University Dr. Robert D. Tollison Duncan Black Professor of Economics and Director Center for the Study of Public Choice George Mason University Dr. Richard Vedder Professor of Economics University of Ohio Dr. Richard Wagner Professor of Economics and Chairman Department of Economics George Mason University Note: Affliations are for identification purposes only. Not all members of the academic advisory agreed with every finding and recommendation in this report. T11630-2046

Au r Kent JeHreya Princinal Reviewer Dr. S Fred Singer Senior Staff and Contributinc Aswciates Rachael Apph:gate Bruce Bartlett Merrick Carey Cesar Conda Gregory Fossedal Dave Juday Felix Rouse Aaron Stevens The Alexis de Tocqueville Institution 2000 15th Street North, Suite 501, Arlington, VA 22201 Tel. 703351.49fi9 Fax 703351.0090

ABOUT THIS sTUDY... "Science, Economics, and Environmental Policy: A Critical Examination," is an evaluation of the data, statistical analyses, and underlying scientific theories that underlie the Environmental Protection Agency's (EPA) policy decisions on environmental tobacco smoke, radon, pesticides and hazardous clean-up under the Superfund law. With the total costs of environmental regulations estimated to be $150 billion annually - or $1,500 per U.S. hottsehold - it is extremely important that environmental decisions be based on sound scientific analyses of potential risks to public health and the environment, and that the costs of environmental regulation be weighed against the benefits. But as Dr. John Graham of the Harvard Center on Risk Analysis notes, "While it may seem obvious that EPA should use good science, students of the Agency have documented that the Agency's leadership, when preoccupied with public fears and legal pressures, has sometimes allowed good science to be neglected." Perhaps Sen. Daniel P. Moynihan (D-NY) put it best, "Truth be told, I suspect that environmental decisions have been based more on feelings than on facts." This research report provides policy-makers, the press, and the general public with the facts about the science and economics that form the basis of the EPA's risk assessments and cost/benefit analyses in four of the most current -- and controversial - environmental questions. The objective of this research is to promote more rational -- and perhaps less costly - environmental decisions through the use of scientifically vigorous risklcost-benefit analysis. In this regard, the Congress is debating several bipartisan initiatives to expand the use of risk assessments and cost-benefit analysis, including the 'Environmental Risk Reduction Act" sponsored by Sen. Moynihan, Rep. Richard Zimmer (R-NJ), and Jim Slattery (D-KS) and the EPA risk/cost-benefit regulatory analysis amendment sponsored by Sen. J. Bennett Johnston (D-LA) and Rep. John L. Mica (R-FL). "Staence, Economics, and Environmental Policy" was researched and prepared by the staff of the Alexis de Tocqueville Institution (AdTI) in Arlington, Virginia. AdTI is a non-profit, non- partisan educational foundation established to conduct, publish, and publicize research on the extension and perfection of capitalism, freedom and democracy in the United States and abroad. For further information, contact Cesar V. Conda, executive director of the Alexis de Tocqueville Institution, 2000 15th Street North, S.501, Arlington, Va.22201. Tel: (703) 351-4969. Fax: (703) 351- 0090. Note: Nothing written here should be construed as necessarily reflecting the views of the Alexis de Toaqueville Institution or its co-chairmen and directors, or as an attempt to aid or hinder legislation before Congress. T11630-2048

E!V V I12UNMr.Nt AL l UkSAI.I.U JMUr.t CASE STUDY NO. 1: ENVIRONMENTAL TOBACCO SMOKE Inhoductlon The downward trend of cigarette smoking in America has been going on for many years. By the end of World War II, almost half of all adult Americans were smokers. Today, "fewer than three out of every ten American adults smoke, and this rate is continuing to decline."` (See Figure 1-1.) This downward trend is the result of a combination of factors: greater knowledge of the health risks associated with smoking, increased federal and state taxes, and a general reduction of tolerance for smoking on the part of nonsmokers, among other things. Ironically, as smoking has declined, the federal government has increased its campaign against smoking. Undoubtedly, many view this effort as beneficial to society. However, it now appears that the federal government has gone beyond its traditional and-smoking efforu, consisting mainly of education and health warnings, and is now moving toward a (de facto) ban on smoking. The vehicle by which this ban may take effect is an Environmental Protection Agency (EPA) studywhich links lung cancer in non-smokers to environmental tobaeoo smoke, or EI'S, which is also called "second- hand smoke" and "passive smoking." If this were the case, it would be difficult to stop the govemment from banning smoking in the name of protecting innocent non-smokers. Unfortunately, in its zeal to abolish smoking, science has been sacrificed. The EPA's finding that second-hand smoke is harmful to human health is based on a lower threshold of risk assessment than the agency normally uses for other substances. In short, the EPA study relied on methodologies different from those which have been historically used in such analyses. Scientific standards were seriously violated in order to produce a report to ban smoking in public settings. The EPA's furding that second-hand smoke is harmful to human health is based on a lower threshold of rrsk assessment than the agency nommlly uses for other substances. Before the government takes action to ban some substance on the basis of its danger to health, it is extremely important that we know the precise degree of danger based on generally accepted scientific principles. If science is debased in an effort to "do good," society ultimately may be left worse off. There are two reasons for this. First, if we debase the scientific method in pursuit of a political agenda, we are opening a Pandora's Box. Second, the ordinary dangers everyone encounters in everyday life are so numerous that if we do not carefully delineate the government's role in regulating such dangers there is essentially no limit to how much government can ultimately control our lives. The health risk from smoking is not the focus of this paper. Instead, this paper explores the EPA's analysis of ETS or second-hand smoke. By any name, it is a complex and highly variable mixture of substances which diffuse through the air. The Environmental Protection Agency has compiled several studies and reports which examine various aspects of the ETS issue. Two in particular are considered at length in this paper; one examined the respiratory health effects of EISZ and the other examined the economic consequences of a proposed restric- tion on smoking.' In briet EPA makes certain assumptions about ETS which are then used to buttress EPA's scientific and economic conclusions. I T11630-2049

SCIENCE, ECONOMICS & ENVIRONMENCAL POLICY Moreover, the science as presented is insuf- Crossing the Threshold ficient and the economic claims are similarly unsupportable. They will be dealt with in turn. First, we will examine EPA's use of the scientific research surrounding ETS. EPA and the Science of ETS There arccortain things about smoking which sound science can demonstrate. For example, active smoking is detrimental to the health of millions of smokers. Nevertheless, EPA has no official role when it comes to regulating smoking. Yet, EPA lately has taken the leading role in publiciung the potential health risks from smoking. According to EPA Administrator Carol M. Browner, "Although EPA has no regulatory authority over tobacco products, it does have a responsibility to inform the public about dangers it finds in the environment.' In particular, EPA has gone far beyond its authority in making ETS an "envirottmental" issue within its regulatory jurisdiction. In the process, it has manipulated both the science and the regulatory process. EPA has gone far beyond its authority in making ETS an "environmentaf' issue within its regulatory jurisdiction. Admittedly, trying to prove that second- hand smoke carries a measurable risk of lung cancer and determiningpreciselywhat that risk is are difficult tasks. It is accepted that smoking is linked to several forms of cancer, particularly of the lungs, and also to heart disease. Similar conclusions about passive smoking, or ETS, should be based upon equally strong scientific evidence. To that end, the EPA has undertaken a review of the scientific literature to determine the effects of ETS on the lungs of nonsmokers. The EPA's major finding was that "ETS is a human lung carcinogen, responsible for approumately 3000 lung cancer deaths annually in U.S. nonsmokers.n5 The question addressed by this section is whether or not that statement is justified. It is well-established that "the dose makes the poison." That is, almost any chemical substance will harm a person's health if administered in sufficiently large quantities. Even substances which are necessary for life itself become deadly at high doses. Unfortunately, the EPA ignores this fact in most of its risk assessments by applying a'linear no-threshold" theory of environmental harm. In essence, the linear no-threshold theory holds that high-dose effects can be extrapolated back to a zero dose without searching for a threshold below which no health effect will be elicited.b In other words, if it were found that exposure to a given level of some chemical substance caused one death per 100,f)00 population, then half the exposure would therefore cause one death per 200,000 population, one fourth the exposure would cause one death per 400,000 population, etc. This flawed assumption underpins almost all of the EPA's work on environmental exposures, from the Superfund program to radon in homes to ETS. The EPA claims to discern an "apparent non-threshold nature of the dose-response relationship observed between active smoking and lung cancer."' Even if this were true for active smoking (and, as questionable as that statement is, it is beyond the scope of this paper), it is not automatically valid to reject the possibility of a threshold effect for ETS. For environmental tobaceo smoke is not just a lower dose of the substances inhaled by a smoker; important, if poorly researched, chemical changes occur as tobacco smoke is diluted and cooled in the open air. Researchers recognize three principal types of tobacco smoke. "Mainstream smoke' is produced when the smoker draws air through a cigarette, thereby "fanning" the temperature as high as 900 degrees centigrade. Most of the compounds in smoke change as they cool and as they react with the smoker's mouth, throat and lungs. "Eahaled smoke" is not the same as the smoke that was inhaled. "Sidestream smoke" is that which is produced by the smoldering cigarette between puffs. Because the temperature is significantly lower (perhaps500 to 600 degrees centigrade), different chemical

compounds (or different amounts) are pro- duced. Together, mainstream smoke,exhaled smoke and sidestream smoke produce environ- mental tobacco smoke, with sidestream smoke accounting for 85 to 90 percenc." The EPA notes that some potentially carcinogenic compounds are present at much higher levels in sidestream smoke than in mainstream smoke. This is a function of the respective temperatures at which various compounds form. Regardless of the composi- tion of sidestream smoke when it is produced, it rapidly undergoes changes, both chemical and in terms of concentration per liter of air. EPA's report recites several distinctions between ETS and mainstream, or even sidestream smoke. The most important distinction arises from the significant dilution of the ETS. In addition, the composition and concentration of ETS is dependent on the number of smokers, their smoking styles, and the number of cigarettes smoked in a given period of time. According to the EPA, for active smoking "A dear dose-response relationship exists between lung cancer and amount of exposure, without any evidence of a threshold level.n9 Of course, a strong dose-response relationship does not rule out the existence of some minimum dose below which there will be zero response. Yet the EPA almost never looks for a threshold for any potentially harmful substance. In fact; it is essentially an unofficial EPApolicy to deny that thresholds exist for any potentially hazardous substance. As examples, consider EPA's stance on dioxins, radon gas, or pesticide residues in the food supply.10 What is more, the faa that tens of millions of smokers survive their habit without developing lung cancer seems to suggest that a threshold exists for each individual, regardless of EPA's assumptions concerning aggregate data. Thus, the statement that no evidence for a threshold exists could easily confuse members of the public. As mentioned earlier, essentially every substance to which humans are exposed is potentially harmful. Many ordinary substances -- common table salt, for instance -- are fatal if ingested in sufficiently large amounts. In addition, hundreds of foods in the human diet contain enormous quantities of "natural carcinogens."' Because the human species has evolved the ability to self-repair the damage caused by these natu rally occurring substances, ENVIRONMENTAL TOBACCO SMOKE we are also able to repair the similar damage caused by small amounts of other carcinogens, including the ones found in ETS. In the face of this assertion by EPA that no safe threshold exists for active smoking, it becomes important to examine how closely EPA links ETS with mainstream smoke. The EPA's Guidelines for Carcinogen Risk Assesvnent (U.S. EPA, 1986) sets out 'three criteria that must be met before a causal association can be inferred between exposure and cancer in humans: 1. There is no identified bias that could explain the association. 2. The possibility of confounding has been considered and ruled out as explaining the association. 3.. The association is unlikely to be due to chance."" Under these criteria, one could oonclude that mainstream smoke (MS) easily qualifies as a lung carcinogen. However, the EPA asserts that because sidestream smoke is chemically similar to MS and because sidestream smoke is the major constituent of ETS, then by inference ETS is also a Group A carcinogen under the EPA test. However. EPA is well aware that ETS is not identical to mainstream smoke, either qualitatively (chemical makeup) or quantitatively (dose). Nevertheless, EPA seems to adopt the old cliche': "Close enough for government work." In fac4 most U.S. studies conducted on ETS and lung cancer have found no statistically significant indications of carcinogenicity. It should be borne in mind that even if ETS is legitimately considered a"known human carcinogen," that does not prove that, at actual environmental exposures, it can or does cause lung cancer. In fact, most U.S. studies conducted on ETS and lung cancer have found no statistically significant indications of carcinogenicity. Many observers have ques- tioned whether EPA's conclusions are justi 6ed." If this were limited to the question of 3 T11630-2051

SCIENCE, ECONOMICS & ENVIRONMENTAL POLICY an internal EPA categorization, it would not be excessively controversial. However, much more is riding on this classification than mere bookkeeping entries. Declaring ETS to be a Group A carcinogen has set in motion a chain of policy events which must ultimately result in widespread federal bans on smoking. To quote the EPA's expressed reasoning in'full: Theconclusive evidenced the dose-related lung earcinogenicity of MS [maiastream smoke] in active smokers, coupled with information on the chemical similaritiesof MS and ETS and evidence of ETS uptake in nonsmokers, is sufficient by itself to establishETS as a known human lung carcl- nogen, or"Group A" carcinogen under U.S. EPA's carcinogen classification system." Similarly simplistic reasoning has allowed the EPA to publicly fret over almost every suggested cancer risk, from electro- magnetic radiation to artificial sweeteners.'s When iis review discovered that eCistrng U.S. studies of lung cancer and ETS did not support itr position, the EPA arbiirarity reduced the traditional standard of proof, or "confidence interval." Despite the EPA's conclusion that ETS is a Group A carcinogen, it is at the very least arguable that ETS would flunk each separate step of the three-prong test. And it is the EPA's effort to cross the final hurdle that has produced the harshest criticism. When its review discovered that e dsting U.S. studies of lung cancer and ETS did not support its position, the EPA arbitrarily reduced the traditional standard of proof, or "confidence interval." Only by this manipulation could the EPA claim that its analysis was statistically significant. Why is the concept of statistical significance so important to epidemiological studies? As valuable as these studies can be, there are well-recognized limitations. For instance, no matter how well designed, epidemiological studies can only show correlation, not causation. Only after many studies have found strong correlatiotts covering large populations (as is the case with actjve smoking and lung cancer) are researchers on firmer ground in asserting direct causation. (Yet even then they may not know the precise mechanism.) Most individual studies, which are expensive and time-consuming, involve only a small number of individuals (or sample size). This reduces the confidence that researchers place in how well the sample population reflects the characteristies of the general population. Epidemiologic studies can test the specific hypothesis, for example, whether ETS is a risk factor for lung cancer. While even well designed studies cannot prove beyond any doubt that a particular substance is the cause of cancer, they can indicate that a particular substance is a potential risk factor. In this case, the EPA assumed - before it even began its investigation - that ETS is a risk faaor for lung cancer; the very question supposedly being asked. However, the fact that most studies of ETS and lung cancer do not support this assumption is not entirely ignored by EPA, which may suggest that the EPA adopted unique manipulations of the data. Furthermore, the EPA does not utilize the appropriate "two-tailed" analysis of whether ETS causes lung cancer. In a two-tailed test, a specific assumption is made, for example, that ETS has an effect on human health. (The two "tails" refer to the fact that the hypothesized effect may be harmful or beneficial: the evidence may point in either direction.) In addition, if ETS were found to have no measurable effect either way, that would be called the "nult hypothesis." In its examination of ETS, however, the EPA utilizes a"one-tailed" test. That is, the EPA makes the assumption that ETS cannot stimulate the human immune response and thereby produce lower rates of lung cancer than would exist in the absence of exposure. However, several American studies examined by the EPA leave open this very point. Indeed of the 30 studies considered for inclusion in the EPA report, "six found a statistically significant (but small) effect, 24 found no statistically significant effect, and six of the 24 found a passive smoking effect opposite to the expected relationship."10 Worse perhaps, the EPA goes further and rejects the possibility of any null hypothesis: s T11630-2052

that EI'S has no effect on health. Thus, EPA refuses to accept any result which would refute its preeastittg assumption: that EIS causes lung cancer in nonsmokers. Whatever one may think of this as a policy outcome, this is not valid science. Ordinarily, researchers utilize a standard mathematical procedure to determine the range of possibilities within which random error is extremely unlikely. By scientific convention, over decades of trial and error and careful review, studies must produce results which have no more than a 5 percent chance of being the result of the natural randomness of the studied population. This is normally referred to as a "95 percent confidence interval " In other words, a 95 percent confidence interval means that there is a 95 percent possibility that the result did not happen from chance, or a 5 percent possibility that it did. However, the EPA rejected this as the test for its survey of the literature on ETS and lung cancer. Because the purported relative risk for ETS was so close to perfectly random, the combined results of the studies examined by the EPA could not pass the 95 percent confidence interval test. In the scientific jargon, the results were not "statistically significant " It is at this point that EPA broke with the established procedure in such matters and declared that a 90 percent confidence interval would be used for this report's findings, thereby doubling the chance of being wrong. As a result, EPA could declare that its findings were "statistically significant' but only if one applies the less rigorous standard. Applying the standard test (a 95 percent confidence interval) would show that lung cancer rates for people exposed to ETS are indistinguishable from the lungcancer rates of unexposed populations. This is no mere academic debate, for there are numerous potential risk factors for lung cancer involving everything from diet to genetics to smoking tobacco. If the EPA's chosen procedures cannot distinguish among the possible risk factors, the report cannot provide useful or reliable guidance to policy makers. Perhaps the strongest criticism of this and other points in the EPA report has come from Gary L Huber and his co-authors:" EPA's risk assessment is built on the manipulation of data, ignores critical ENVIRONMENTAL TOBACCO SMOKE chemical analyses and key epidemiological data, violates time-honored stalistical pri.nciples, fails to control adequately for important confounding influences (factors other than the one studied that may affect the result or a conclusion) that provide alternative explanationsfor itsconclusions, andvioiatesitsown guidelinesfor assessing and establishingrisk to a potential environ- mental tozin." One of the particular points of disagreement between Huber, et al., and the EPA report arises from a major study of EI'S and lung cancer in U.S. female nonsmokers by researchers at the National Cancer Institute (see Figure 1-2).19 Although the EPA did not include this study's findings in its report's calculations, it did quote from the report in an effort to demonstrate general consistency. The exact quotation selected by the EPA is that 'long-term exposure to [ETS] increases the risk of lung cancer in women who have never smoked."20 Huber, et al., quote a different, yet equally revealing passage. Stockwell and his co-authors report that "we found no statistically significant increase in risk associated with exposure to environmental tobacco smoke at work or during social activities.n21 (emphasis added) This is an important point because EPA suggests that workplace regulations are legitimized by studies of the wives of smokers. In a recent report for Congress from the Congressional Research Service2, another major study is cited which is not included in the EPA report.' This study, which covered a larger population sample than the Stockwell study, "found no overall increased risk of lung cancer among nonsmoking spouses of smokers."2' Furthermore, ifstandard statistical procedures were applied to the Stockwell study, it too would fail to support EPA's flnal results. Bear in mind that Huber and his co- authors do not assert that ETS does not or cannot cause lung cancer in nonsmokers. They 'ETS is a risk that is less than that associated with developing colon cancer by drinking chlorinated water, which is in most U S. cities' water supplies." T11630-2053