[5 ELR 50165]

I. Introduction

In their efforts to learn why the incidence of cancer and other fatal and serious diseases is on the rise, researchers are concentrating more and more on the health effects of products and waste products of modern industrial society. One such product, vinyl chloride, the principal ingredient of the second most widely used plastic, has recently been shown to cause cancer in humans. Gasoline combustion products have been linked to chronic respiratory difficulties. It has been estimated¹ that up to 90 percent of the incidence of cancer in the general population may be caused by environmental factors (including cigarette smoke).

Protection of the public from threats to public health and safety caused by the release of harmful substances into the environment is accomplished in large part through decisions of the regulatory agencies of the federal government. The agency decisions are made within a framework of laws which express congressional intent as to the extent to and means by which agencies shall act to protect the public against air or water pollution, toxic chemicals, radiation, or other potentially harmful factors. The agency decisions, and through them, the congressionally-enacted laws, are reviewed by the courts, institutions which have had and continue to have difficulties in dealing with the complex scientific and public policy issues raised by regulation in the public health area.

A broad spectrum of American society — Congress, the courts, industry, the public (represented by public interest groups, including unions), and scientists — are vitally concerned with the decisions made by regulatory agencies regarding environmental threats to the public health. The influence exerted on regulatory agencies by these various groups can increase in difficulties a regulator faces in making such decisions.

Regulation in the environmental health area is also made difficult by problems involved in collection and analysis of scientific data. Regulators often are faced with the difficult choice of making a decision on the basis of admittedly imperfect data, or postponing regulation until a better data base is assembled. "Regulation now or later" becomes an especially sharp dilemma whenever a regulator is dealing — from an imperfect data base — with a substance which may have very severe effects on human health, but whose effects on health may only appear many years after exposure has occured. If the decision is made to defer regulation until it can be shown more definitely that a substance actually causes harm to humans, then regulation may occur too late to prevent harm to the population. "Regulation later" as a regulatory approach really means using humans as "guinea pigs," and should be considered socially unacceptable. It is the thesis of this article that when faced with an imperfect or incomplete data base, a regulator should work with a heavy bias towards measures which will reduce exposure of humans to substances which may pose serious risks of harm to health.

II. Recent Judicial Decisions in the Field of Environmental Health Regulation

The problems associated with decision making in the environmental health area are visible in several recent judicial holdings and administrative decisions.

The Reserve Mining cases² focused on the risk to human health posed by ingestion of asbestos fibers. The fibers, dumped into Lake Superior in the wastes (tailings) from a taconite refinery plant, were ingested with their drinking water by the citizens of Duluth, Minnesota, a city which draws its water supply from the lake.

In the course of the trial before the District Court, voluminous testimony was presented regarding the demonstrated harm to human health from the inhalation of asbestos fibers. Judge Miles Lord concluded, by inference from that testimony, that ingestion of asbestos also posed a substantial threat of harm to public health. The trial court did not require that the plaintiffs, who sought an injunction against further discharge of the tailings into the lake, demonstrate that ingestion of asbestos actually causes cancer or other diseases in humans. The court ruled that a substantial threat of harm to the public health was sufficient grounds for enjoining further asbestos discharges into the environment. In rendering its decision, the court took into account the serious nature of the harm that could be done if ingestion of asbestos fibers were [5 ELR 50166] equivalent or similar to inhalation of the fibers, and the social policy difficulties in waiting twenty years or more (the customary latent period for development of asbestos-caused cancer) to determine empirically whether ingestion of asbestos fibers from Silver Bay taconite tailings really could cause cancer in Duluth residents.

The mining company appealed the District Court's decision. A three-judge panel of the Eighth Circuit Court of Appeals reversed the lower court ruling, and in so doing delivered an opinion which set off an uproar among consumer and environmental advocates. The circuit court's opinion led to moves on Capitol Hill for legislative reversal of the burden of proof imposed by that court on those seeking to ban or restrict environmental discharges of allegedly harmful substances on public health grounds.

A few sentences from the opinion serve to illustrate why consumer advocates and environmentalists found it unacceptable:

Since testimony clearly established that an assessment of the risk was made impossible by the absence of medical knowledge, Judge Lord apparently took the position that all uncertainties should be resolved in favor of health safety…. If we are correct in our conclusion that evidence does not exist in the record on which to find Reserve's discharges to be unsafe, the district court's determination to resolve all doubts in favor of health safety represents a legislative policy judgment, not a judicial one…. Although we are sympathetic to the uncertainties facing [persons drinking the asbestos-contaminated water from Lake Superior] we are a court of law, governed by rules of proof, and unknowns may not be substituted for proof of a demonstrable hazard to the public health.³

Russell Train, the Administrator of EPA, took quite another tack in upholding the decision of an EPA administrative law judge to suspend the registration of aldrin/dieldrin, a pair of closely-related pesticides.⁴ Aldrin/dieldrin had been shown to cause cancer in mice by several studies, including some conducted by the registrant, Shell Chemical Corporation. The pesticides have not been shown to cause cancer in man; however, data obtained from human tissue studies performed in 1971 would suggest that dieldrin residues are present in essentially the entire American population at or near levels at which the FDA must seize and destroy foodstuffs as unfit for human consumption because of dieldrin contamination.⁵

In the aldrin/dieldrin case, the administrator determined that demonstration of carcinogenic activity in the mouse was sufficient to make it highly probable that aldrin/dieldrin would be carcinogenic in man. A scientific witness for Shell, in fact, estimated that the probability that aldrin/dieldrin would cause cancer in man, given its carcinogenicity in mice, was "only 70 percent;" that probability of severe harm to man was high enough to justify suspension in the eyes of the administrative law judge and the Administrator.

The aldrin/dieldrin case and Reserve Mining are similar on the facts. Both involved cancer, a disease which becomes apparent in man only after a long latency period, in some instances many years after exposure to the carcinogen has occurred. Both cases involved situations where a cause-and-effect relationship could not be scientifically established between human cancer and exposure to the chemicals for which regulation was being considered. In the Reserve Mining case, human data demonstrated that asbestos did cause cancer when inhaled, but Duluth residents were ingesting rather than inhaling their asbestos. In the aldrin/dieldrin case, the pesticides had been shown to cause cancer in mice, and there was high probability, based on past experience with animal experimentation, that the pesticide pair would also be carcinogenic in man.

The Ethyl case⁶ involved lead poisoning rather than cancer. Although the prevailing viewpoint among cancer researchers is that it is not possible to demonstrate a safe level of exposure to a carcinogen other than zero, it is possible to demonstrate safe levels of exposure to lead.⁷ The question under consideration in the [5 ELR 50167] Ethyl case was whether lead in automobile emissions caused by the addition of tetraethyl lead to gasoline contributes to the human body burden of lead, particularly in urban children, in such a way as to "significantly endanger" human health. Given other environmental exposures to lead, can automotive lead emissions push a significant number of humans over the safe limit of the body's ability to absorb lead without harm?

EPA concluded that the data were sufficient to demonstrate both the serious nature of the risk and the probable association of lead in auto emissions with air- or dust-borne lead exposure in persons residing in heavily-traveled urban areas.

A three-judge panel of the D.C. Circuit Court of Appeals reviewed the Administrator's decision to require the phased reduction of lead content of gasoline and overturned his action, finding that it was arbitrary and capricious. The opinion of the majority of the three-judge court, since vacated by the full court for a rehearing of the case en banc, is exemplified by this citation:

If there can be found potential harm from lead in exhaust emissions, the best (and only convincing) proof of such potential harm is what has happened in the past, from which the Administrator can logically deduce that the same factors will produce the same harm in the future.⁸

In other words, the only way to demonstrate that a chemical may harm people is to demonstrate that it does harm people. In its refusal to consider "risk" data when there is a possibility of serious harm to public health, the original decision in Ethyl was similar to the appellate decision in Reserve Mining. On the other hand, lead poisoning does not take long to appear, and it may be possible to conduct new, careful, clinical and epidemiological studies over the next few years which could demonstrate conclusively whether tetraethyl lead plays a significant role in human lead poisoning. In Ethyl, as in Reserve Mining, the real judgment is whether it is appropriate to wait for "bodies on the street" before a regulatory agency can act to protect the public health.

III. The Scientific Data Base: Incomplete Guidance at Best

A. Types of Information Available: Uses and Limitations

The most troublesome problems faced by a regulator in the environmental/public health area stem in large part from the difficulties of using scientific data in regulatory decision making. Availability of the various types of scientific data may strongly influence the regulator's ultimate decision. Test and monitoring data are the typical data used in regulatory decision making Tests involve experimental techniques, holding constant (in an ideal situation) all factors except one, and letting the one factor being studied vary in a way calculated to give useful data for regulation. For instance, a group of rats could be fed an identical basic diet, with one sub-group having added to its basic diet a food additive suspected of being a carcinogen. Comparison of the health of the group of rats fed the additional chemical with the health of the control group could provide a "yes-no" answer as to whether the food additive causes cancer in rats.

Monitoring does not involve experimentation. It involves uncontrolled observation of certain conditions, often environmental in nature. For instance, monitoring could be performed to determine the concentration of lead in the air or dust near a freeway or the concentration of vinyl chloride in a polyvinyl chloride production or fabrication facility. Monitoring simply answers the question: "How much?"

Test or monitoring data are useful in a number of regulatory situations. For instance, the Delaney Clause⁹ requires that any food additive demonstrated to be carcinogenic in animals be banned from appearing in human food. Demonstration of cancer in animals though data produced by "tests which are appropriate for the evaluation of the safety of food additives,"¹⁰ would satisfy the Delaney Clause. Monitoring data [5 ELR 50168] would suffice to demonstrate that the level of a pollutant in the air of a metropolitan area is above the level determined to be safe by EPA under the terms of the Clean Air Act.¹¹

Test and monitoring data are not useful in a situation where the question to be answered is: "How does this disease occur?" Such questions can be answered only by highly sophisticated controlled experimentation characteristic of research on the biochemical mechanisms of disease. As it happens, however, regulations and laws in the environmental health area usually require only "yes-no" or "how much" answers, so tests or monitoring, as appropriate, will usually suffice for collection of the data necessary for decision making.

Although test or monitoring data may be prescribed by the statute which the regulator is enforcing as legally sufficient for regulation, such data must appear to be reasonably relevant to protection of human health for regulation to be credible. Relating testing and monitoring data to human health requires consideration of the biological systems on which the data were collected. The regulator may confront problems relating data obtained in his test system with the human situation, and he may also face system with the human situation, and he may also face difficulties when data, especially human data, simply cannot be obtained for use in decision making.

If human data are available to the regulator, making a decision is often easy, since such data fulfill even the stringent criteria for proof of potential risk set out in Ethyl and the appellate opinion in Reserve Mining. But human data are rarely ideal for deciding whether a given chemical or other substance causes disease and should therefore be regulated.

Epidemiological studies provide statistical associations between exposures to certain substances and diseases, but do not usually offer explanations for the biochemical mechanisms of the disease observed. Thus, the connection of cigarette-smoking with lung cancer has been amply demonstrated by epidemiological studies of smoking and non-smoking populations, but there are still no definitive data available on what factor in cigarette smoke is responsible for the induction of lung cancer in humans. Feeble attempts by the federal government to regulate cigarettes have probably been kept weak by this failure to demonstrate the specific carcinogen in tobacco smoke. A clear demonstration of the ingredient in smoke which causes cancer might overcome the political and economic forces that have thus far prevented more stringent regulation.

In order to determine the mechanism of disease-producing action of regulated substances, it is necessary to perform controlled experiments. Such experiments can include testing to determine whether a substance causes disease at all. Controlled experimentation on human subjects is sharply limited in this country due to societal constraints. In general, it is limited to testing of drugs and products which will not cause irreversible harm to human subjects. But even if controlled experiments are undertaken, the complexities of the human organism and disease processes make it less than 100 percent likely that clear cause-and-effect relationships can be established between production of disease and administration of the test substance. Human experimentation usually involves small experimental samples relative to the number of humans who may ultimately be exposed to the substance being studied. Results obtained in a small sample of humans cannot be extrapolated with a great degree of confidence to the genetically heterogeneous general population. Use of small experimental groups thus casts further doubt on the utility of human data for regulation.

Animal data are the data most commonly used for regulatory decision making in the environmental health area. Animals such as mice or rats are good experimental subjects for many reasons. It is easy to control animal exposure to test substances and, to some extent, to environmental factors; there are statistical advantages to using relatively large numbers of experimental subjects; and it is convenient to use subjects with short life spans. And, of course, one very real advantage mouse experiments have over human experiments is that the mice can be killed in the course of the experiment and the animal's organs subject to intensive study to determine a substance's effects on the animal.

The principal problem in the use of animal data is the extrapolation of animal data to man. Significant differences in anatomic structures and biochemical pathways between experimental animals and man may lead to test substances having different effects on mouse and man. Although such different effects are possible, it has been demonstrated that a chemical which causes cancer in rats or mice will very probably cause cancer in man. Conversely, only one human carcinogen, arsenic,¹² has not yet been shown to cause cancer in experimental animals and this failure may by due to improper route of administration of arsenic or failure to administer factors which interact with arsenic to cause cancer rather than immunity of all animals save man to its carcinogenic effects.

Combined animal toxicological-human epidemiological data can be highly probative as to causation of disease by environmental factors. The vinyl chloride case is illustrative. Epidemiological studies of polyvinyl chloride polymerization workers link a vary rare liver cancer, occurring in this worker population, to inhalation of vinyl chloride, and toxicological studies have [5 ELR 50169] demonstrated that vinyl chloride inhalation causes identical neoplasms in rats and mice.¹³ Vinyl chloride inhalation produces cancers at sites other than the liver in rats and mice,¹⁴ and epidemiological data are suggestive of multi-site carcinogenicity in humans.¹⁵

B. Increasing a Regulator's Confidence in the Available Data

Assuming, however, that a regulator can develop or has access to data which may be suitable for decision making, he will still have the problem of deciding how much confidence he can place in those data.Test and monitoring data are most useful to a regulator when they have been obtained using a method which has given valid results in the past.

In addition to using standardized tests and monitoring methods, a regulator can increase his confidence in his scientific data base by subjecting the data to peer review. Publication of data and criticism of published data by other scientists can test and affirm the data's validity. Scientific advisory committees, comprised of individuals considered "experts" in the scientific area under discussion, can also analyze data, evaluate their valifity, and decide how useful the data are for regulatory purposes.

Unfortunately, each of these methods of increasing a regulator's confidence in his scientific data base has drawbacks.Use of standardized testing and monitoring may prevent development and establishment of better tests and monitoring procedures. Publication of results in good scientific journals may take years, since the editorial review process is lengthy and there is great competition due to space limitations to have papers published in highly-regarded journals. The progress of biomedical science is rapid today, and papers are often out-of-date by the time they are published a year or two after the data were collected. Scientific advisory committees usually come to be dominated by individuals affiliated with or sympathetic to industry, who have time and financial resources sufficient to allow them to attend meeting which may last several days at a time over a period of months or years.

Science thrives on an atmosphere of open exchange of information among scientists, and there are frequent communications among scientists concerning planned experiments, experiments in progress, and preliminary or other unpublished data. These exchanges of information may occur at conferences, symposia, seminars or by informal means such as letters or telephone calls. Such open communication and critical exchanges can provide assurances of validity even to unpublished data.

In some instances, full exchange of information among scientists is impaired, however. For example, industry laboratories may keep their data secret and may, in fact, decline to publish results. Rather than wait indefinitely for industry to disclose data on a voluntary basis once human illness has occurred, as in the vinyl chloride case, all health and safety experimentation or studies ongoing, previously conducted, or planned by industry should be denied trade secret status by the administrative agencies or Congress, as appropriate. Such actions might serve to open up industry labs and their scientists to free exchange of information with academic and government scientific personnel. It is interesting to note that industry will often insist on a regulator using only published and openly reviewed data in regulatory decision making while industry hides from public view crucial data they have developed in their own laboratories.

Regulatory agency scientists may have their principal scientific contacts with individuals on the staffs of the regulated industry or industries. Agency-industry scientific liaison committees which meet informally and at frequent intervals can prove important in shaping decisions as to research to be performed by the agency¹⁶ and, perhaps, on regulatory policy itself. The activities of such liaison groups should be monitored carefully to make sure that the viewpoints of independent scientists are taken into account as regards agency research policies.

In short, these problems with the data base can make purely scientific decision making difficult if not impossible for the public health regulator. To supplement what can be learned from the available scientific data, the public health regulator must consider additional factors. These factors are addressed in the next section.

IV. A Framework for Regulatory Decisions in the Public Health Area: Making Decisions When the Scientific Data are Incomplete

What factors should a regulator consider in formulating a decision in the environmental health area? The scientific data available often have serious limitations. To guide decisions where data are incomplete, the [5 ELR 50170] regulator can turn to a number of factors, including language from court opinions, policy considerations, and projections of the social and economic impacts of any regulatory actions under consideration. Which factors the regulator decides to use and the weight he will assign to each of them depends on the strictures of the statute under which the regulation is being undertaken. This section analyzes some of the statutory guidance provided to the regulator, assesses a number of critical policy considerations peculiar to regulation in the public health area, and finally looks at several considerations of social equity confronting the regulator.

A. Statutory Requirements

The risks of severe harm to humans posed by particular environmental factors may be so great as to prompt a legislative decision to base regulation on some indicator other than the appearance of human disease,such that, given the risk to health, the regulator should have no discretion to consider, in making his decision, any factors other than scientific data.

The Delaney Clause is an example of a statute which provides this kind of very strict guidance. Under this statutory provision, if data obtained from "appropriate" experiments demonstrate that a food additive causes cancer in animals, then the chemical cannot be allowed to be present at detectable levels in human food.¹⁷ If residues of a carcinogenic additive cannot be eliminated from human food, the chemical must be banned.

The Clean Air Act,¹⁸ like the Delaney Clause, sets out numerical standards for reducing human exposure to environmental pollutants. The Clean Air Act requires that certain automobile emissions be reduced by 90 percent over a five-year period.

The argument has been raised in some quarters that such non-discretionary regulatory legislation is inconsistent with good science and with good regulatory policies. Those subscribing to this view contend that standards mandated by Congress lead to bureaucratic rigidity, make impossible "balanced decision making," and prevent research which is often needed to gain real undertstanding of the character and seriousness of an alleged hazard to the public health. They also assert that regulation with a incomplete data base results in the loss of scientific support for regulatory decisions by an agency.

Does a congressionally-mandated standard necessarily represent an inadequate scientific data base for decision making? In the case of the Delaney Clause, the answer is no. Most cancer researchers believe that the only safe level of exposure to a chemical carcinogen is no exposure at all — "zero exposure." The Delaney Clause also reflects sophisticated understanding of cancer research, since, as noted previously, chemicals which produce cancer in animals have a high probability of being carcinogenic in man. If humans are to be guaranteed absolute safety from carcinogenic food additives, then the Delaney Clause is a scientifically sound law.

But is the Delaney Clause good social as well as scientific policy? Food additives may range from preservatives, protecting food from attacks by microorganisms, to "cosmetic" chemicals such as flavors or colors. Although an argument could be made that there may be some preservatives which are difficult, although not impossible, to replace, the vast majority of additives are not necessary for maintaining a supply of wholesome food. If a flavor or color were shown to cause cancer in experimental animals, how would a regulator obtain data that would demonstrate that the chemical was safe for human use? The most direct method would be to test the chemical on humans, but it is doubtful that American society would be willing to risk human lives to test, in a controlled experiment, a possibly carcinogenic food additive. Even if social strictures on human experimentation did not exist, the large number of human experimental subjects which would be necessary to obtain statistically valid results, and the many years which would be necessary before cancer would begin to show up in the test group, would make such experiments infeasible. Of necessity, even if there were no Delaney Clause, a regulator trying to decide what to do about a food additive demonstrated to be carcinogenic in animals would have to regulate with an incomplete data base since he could not obtain human experimental data.

If a regulator cannot determine by human experimentation whether a chemical which causes cancer in animals is also a human carcinogen, how else can he generate scientific data which could justify allowing the food additive to stay on the market? He could conduct new animal tests to demonstrate that the additive does not cause cancer in animals, and that the original animal test results were wrong.¹⁹ He could attempt to demonstrate that physiological processes in experimental animals are sufficiently different from those found in man so that the animal results which led to proscription of the chemical are inapplicable to man, or he could attempt to demonstrate that there is a safe threshold level for human exposure to the carcinogen.²⁰ Given the scarcity of funds for biomedical research, however, it would seem more sensible to use government funds to advance scientific knowledge in the public health area by testing some of the myriad new chemicals introduced into the environment each year for carcinogenicity [5 ELR 50171] in animals rather than seeking to protect the narrow economic interests of industry by pursuing the will o' the wisp of safe threshold levels.

If there were no Delaney Clause, how would the application of "balanced decision making" improve the regulatory situation? "Balanced decision making" in this context evidently means that factors other than protection of public health should be taken into account in regulatory decision making on environmental-public health questions. If the food additive causes cancer in animals, and past experience shows that this means there is a high probability that the additive will also cause cancer in man, then the costs to society of the related human illness and death must be factored into the cost-benefit equations used for balanced decision making, along with the costs to industry of lost profits and the costs to society of loss of the food additive. Even using balanced decision making, the balance would still seem heavily tipped in favor of banning the additive.

The argument that legislatively-mandated standards lead to bureaucratic rigidity by forcing regulation to occur without further research into whether regulation is necessary and without use of balanced decision making benefit-risk assessments is based on a fundamental misconception. Mandatory standards are enacted by Congress precisely in order to overcome regulatory agencies' inherent tendencies toward bureaucratic inertia and unwillingness to regulate when given discretion in the face of strong industrial opposition to regulation.

Finally, it would be instructive to learn how many scientists have lost faith in regulatory agency activities because of mandatory standards imposedby Congress. Cancer researchers generally favor a "no threshold" theory for chemical carcinogenesis; "no threshold" is consistent with the underlying principles of the Delaney Clause, and it is probable that independent cancer researchers would therefore support the Clause. Industry scientists and scientists affiliated with industry, who are very rarely qualified cancer research specialists, oppose the Clause.

B. Other Statutory Guidance

Unlike the Delaney Clause, which is unusual in that it gives the regulator no discretion, most statutes authorize the regulator to factor into his decision considerations other than the scientific data he has collected. Given the ambiguities of the scientific data base available to the regulator, non-scientific data, properly utilized, can provide criteria for the social value judgments which should be used in decision making under these laws.

Congress usually provides regulators with guidelines for consideration of factors other than scientific data in such decision making. The Delaney Clause states the extreme case — the benefits of food additives are not worth the risk of cancer under any conditions, so no risk-benefit analysis should be conducted. In other statutes Congress may require the risk-benefit (costbenefit) assessments be conducted, or may require that economic factors or technological feasibility of compliance be taken into account by the regulator.Risk-benefit assessment and assessment of economic factors or technological feasibility are all interrelated and difficult to consider as discrete concepts.

In general, the greater the benefit derived by society from the use of a substance, the greater the risk society should be willing to bear.This reasoning applies to all factors or activities with inherent risks ranging from driving cars to cooking on gas ranges to using lipstick. There are some products or substances whose real benefits to society are so minimal the products or substances should be designed to pose little or no risk to humans. Thus, cosmetics should not cause physical harm.²¹

Food additives, discussed above, demonstrate the broad range of benefits products may offer; benefits of additives range from protecting people against botulism to simply providing people with the visual pleasure of consuming a cake whose icing has been tinted pink. Just as the benefits of purely cosmetic food additives may be so small as not to be worth any risk no matter how minimal, the Delaney Clause defines situations where a risk of severe harm is so great as to outweight any benefit, no matter how large.

Risk-benefit assessment is largely economic in nature in that it usually attempts to set quantitative values for the costs to society of engaging in or forbearing from engaging in some activity. Because of the difficulties in estimating costs of harm to public health, risk benefit assessments should be used only with great caution and consideration to the equities of the various groups which seek to influence regulatory decisions in the public health area. As noted earlier, there is often less the 100 percent probability that a substance will cause a disease. The disease, if it occurs all, may occur years after exposure to the substance, and there may be no way to predict how many people may be affected by exposure-caused disease. It is much easier to determine how much financial loss a company will suffer if it cannot market a chemical and the net loss in wages and cost to society of supporting workers who would lose their jobs if the chemical were restricted in distribution or banned.

Although the exercise of assigning dollar values to social costs and benefits of environmental health regulation may seem to provide a neat and attractive method for justifying regulation, some costs and benefits may defy quantification. The intractability of certain costs and benefits to dollar-and-cents valuation should lead to caution in the application of cost-benefit [5 ELR 50172] analysis in environmental health decision making. For instance, how do you assign dollar values to the very real fear experienced by a worker exposed twenty years ago to a chemical only recently demonstrated to cause cancer? How does the fear affect the worker's life and the lives of his family, especially when data have demonstrated that families of workers exposed to certain carcinogens (viz., asbestos)²² experience a cancer incidence significantly above that of the general population?

Psychological and esthetic factors may defy quantification. As noted above, it is exceedingly difficult to make calculations of such factors in regulation as the cost to the community of illness attributable to an environmental contaminant. Inter-generational health effects — birth defects and delecterious mutations — are both probabilistic in nature and poorly characterized at present in scientific let alone social or economic terms.

When dealing with components of a regulatory decision which cannot be assigned money values with a high degree of confidence and precision, a regulator should face squarely the impossibility of assigning monetary values to all the decisional factors. This requires frank acknowledgment by a decision maker that unquantifiable costs and benefits exist, and that value judgments must be made without the help of numbers.

A recent National Academy of Science report states that:

Value judgments about (nonquantifiable) factors in a decision such as life, health, aesthetics, equity and risk aversion should be explicitly dealt with and not hidden in supposedly objective data and analysis.²³

The NAS report also points out the ultimate futility of attempts to base regulatory decisions on supposedly objective cost-benefit analyses, and suggests that efforts to perfect such cost-benefit analyses may not prove useful investments of time or money:

Highly formalized methods of benefit-cost analysis seldom can be used for making decisions about regulating chemicals in the environment. Thus the development of such methods should not have high priority. However, … benefit-cost … frameworks … can be useful in organizing and summarizing relevant data on regulatory alternatives which the decision maker must review.²⁴

Acknowledgment of the limitations of cost-benefit analysis may lead to greater discomfort on the part of regulators if they are forced to expose the political and social considerations which enter their decision-making processes, but such an admission will undoubtedly increase the honesty content of regulatory decisions in the environmental health area.

Among economic considerations that are taken into account, in large part, under risk-benefit or cost-benefit assessments, some merit special mention. Occupational disease has been identified as a serious problem; occupational carcinogenesis, in particular, poses risks to workers in many industries ranging from metal smelting to plastic production to leather-crafting. If the costs of safety or health regulation imposed on industry are great, industry may respond to regulation by simply closing down or threatening to close plants. This leaves workers in the position of, say, cancer later, or having no job at all. Industry has applied similar threats of closing facilities when faced with costly environmental protection regulations. A regulator must also consider whether regulation will merely shift dangerous processes or products to foreign countries; the impact on the American economy and the ethics of exporting dangerous technologies merit close study. The exodus of asbestos fabrication facilities from the United States since OSHA health regulations went into effect has been well documented.²⁵

A growing number of lawsuits are being brought against companies which have exposed workers to substandces that have turned out to cause severe harm to health, particularly occupational exposures to chemical carcinogens. Thus, suits for millions of dollars in damages have been filed against PPG Industries by workers exposed to asbestos at a PPG plant in Tyler, Texas,²⁶ and workers exposed to Bis-(chloromethyl) ether at a Rohm and Haas facility in Pennsylvania.²⁷ These suits have in common illness caused by agents which were not adequately tested for safety before human exposure occurred, and which may not have been adequately controlled even when it became clear that the agents could cause serious harm to health. Insofar as the government has failed to enforce laws which would require protection for workers against such demonstrated health hazards as asbestos, it must be asked whether the OSHA could be joined as a party defendant in a lawsuit by the Tyler workers. A regulator may have to take into account the cost of not regulating in terms of future lawsuits directed both at a manufacturer and at the federal government.

A regulator should also consider whether substances or processes are available which can substitute for the substance or process for which regulation is being considered. In order for a substitute substance or process to [5 ELR 50173] receive any serious consideration, data should be available which demonstrate that the substitute will be less harmful to human health than the substance or process being regulated.When dealing with an industry with old uneconomic plants or processes, it may make long-term regulatory sense to restrict distribution of a substance, if a substitute for the substance is available. In the short-term, however, dislocations of workers and large capital requirements for construction of new facilities or introduction of a new product may make a regulator think twice about banning or phasing out a substance or process.

The technological feasibility of compliance with a regulation must receive careful consideration, especially if there is not viable substitute product or process available to replace the one being restricted. The skill of industry in developing alternative products or processes or devising methods to reduce product contamination and worker or public exposure to a dangerous substance, is matched only by the skill of industry's lawyers in making the case that compliance with a regulation would be impossible. The recent vinyl chloride cases²⁸ were based on claimed inability of the industry to reduce workplace exposure to vinyl chloride (VCM). However, recent monitoring data from major polyvinyl chloride (PVC) producers indicate that the manufacturers have speedily achieved that which they said they could not do at all, and even old PVC polymerization plants are operating with VCM levels below 5000 parts per billion (ppb)²⁹ (eight hour time-weighted average; the OSHA regulation requires a 1000 ppb eight hour time-weighted average). Of course, the PVC producers have used all possible means available to them to come into compliance with the OSHA regulations, including siting new polymerization lines in the open air rather than in enclosed buildings — an alternative which will probably increase the general environmental burden of VCM while reducing occupational hazard — and, according to some critics,³⁰ moving or planning to move PVC production to other countries.

American industry has an enormous capacity for technological innovation, so a regulator should be skeptical of claims that compliance is not technologically feasible, but the regulator may have to give serious consideration to phasing in a regulation so that industry will have time to come into compliance. A "phase-in" regulation is especially problematic in the case of a carcinogen, like VCM, where even the slightest exposure may eventually cause cancer, and where hundreds or thousands of workers or memebers of the public, including persons not previously exposed, may be exposed to the carcinogen while control is not yet complete.

C. Policy Considerations Peculiar to Public Health Regulation

Given an imperfect data base and operating under a statutory scheme which provides for discretion to consider factors other than scientific data in making decisions, how would a regulator decide? In his cogent dissent in the Ethyl case, Judge Skelley Wright identifies two principal factors which can serve as criteria in environmental health decision making.

1. How severe is the harm which could be done by the substance for which regulation is under consideration?

2. What is the risk that the harm may occur?

In order to apply Judge Wright's criteria, a regulator must consider several component factors.

1. Extent of Exposure

First, how many people will be exposed to the substance if no restriction occurs? For example, for a product as useful in industrial and consumer applications as polyvinyl chloride, the number of people exposed is already large and increases as new and more extensive process and product applications are developed. As the number of individuals exposed to a substance increases, a regulator should be increasingly conservative about allowing unrestricted distribution of the substance. Given the genetic and physiological variability of human beings, as the number of persons exposed to a substance increases, the probability that someone will be harmed by the substance also increases. Even if exposure to the substance only causes a small adverse effect on health, the aggregate costs to society of many persons suffering from small adverse health effects may be very great. In general, as the number of individuals exposed to a substance goes up, the regulator should weight his decision making towards a requirement that a manufacturer demonstrate that his product is safe before human exposure occurs.³¹

2. Existence of Particularly Susceptible Subgroups

Are there sub-groups in an exposed population which are particularly susceptible to being harmed? For instance, urban slum children are deemed particularly susceptible to the adverse health effects associated with lead poisoning from air- or dust-borne lead because of their exposure to lead-based plant and because of such factors as poor diet and, possibly, genetic make-up.

3. Ability of Citizens to Protect Themselves

Can the population as a whole or particularly susceptible sub-groups protect themselves against harm? [5 ELR 50174] Although humans do not have to smoke cigarettes, they do require air to breathe and water to drink. Voluntary exposures, such as cigarette smoking, merit different consideration from involuntary exposures, such as breathing contaminated air near one's home. When exposure is involuntary, a regulator should weight his decision in favor of restriction if he has doubts about whether or not to restrict.

Clearly, to force residents of Duluth to drink bottled water in order to avoid exposure to the asbestos-contaminated lake water coming through their taps is to acquiesce in discrimination against those citizens too poor to afford the bottled water. The poor cannot buy their way out of harm; workers usually do not have a realistic option of trading an unsafe job for a safer one; children cannot protect themselves from harzards because they may not be able to recognize or avoid harmful substances.

4. Threshold Levels of Exposure

The regulator should consider whether thresholds exist for the causation of adverse health effects by the substance for which regulation is being considered. A panel of cancer experts assembled by HEW concluded that only a zero exposure to a carcinogenic chemical should be considered safe for humans³² and, as pointed out above, this is most certainly the opinion of a majority of cancer researchers. If no threshold exists for elicitation of a disease state, then a regulator must determine how to set an exposure level which will be "socially acceptable."³³ Such a determination involves complex risk-benefit assessment.

Threshold exposures do exist for such diseases as lead poisoning, where the body can tolerate a non-zero body burden without illness occurring. However, irreversible disease may occur if the body burden is exceeded due to incremental exposures over a period of time.³⁴ The incremental risks of chronic exposure or multiple exposures are important in consideration of diseases such as lead poisoning which do have thresholds. Humans in modern industrial society are often subjected to multiple exposures from several sources to chemicals and other factors which may pose risks of severe harm. Children may be exposed to lead by ingestion of evaporated milk from lead-soldered cans, by eating chips of lead-containing plant, or by breating air contaminated by lead from leaded gasoline.

5. Interaction of Contaminants with Each Other

Different environmental contaminants may interact in such a way as to cause potentiation of the risk of severe harm far beyond what would be expected if the sum of the separate risks were simply additive. It has been shown that smokers with occupational exposure to asbestos are many times as likely as non-smokers without asbestos exposure to develop lung cancer; the combined smoking-asbestos exposure risk exceeds by far the simple sum of the risks experienced singly.³⁵ There are very few data available on the interactions of the many environmental contaminants to which humans are exposed; this paucity of data should lead a regulator to be very cautious about unrestricted distribution of substances in the environment.

6. Reversibility

Irreversibility of a disease process should be taken into account by a regulator in judging the severity of harm. If a disease is irreversible once initiated, as is cancer or, in some cases, lead poisoning, in the presence of an ambiguous data base, the regulator should weight his decision towards restriction of distribution of the substance. After all, a regulation which subsequent experimentation or epidemiological data demonstrates is unnecessarily stringent can always be relaxed, but the course of irreversible disease cannot be stemmed once exposure sufficient to elicit the harm has occurred.

7. Long Latent Periods

Diseases with long latent periods merit special consideration by a regulator. If a substance is not restricted, it may take five, ten, twenty or more years for illness attributable to the substance to appear in the human population. Cancer is the best-known disease with a long latent period, since illness often appears 20 years or more after exposure. Mutagens are substances which cause changes in the genetic material; the deleterious changes caused by mutagens also may not become apparent until several generations after exposure occurs.

For diseases which combine irreversibility with long latent periods, to delay regulation to protect humans from harm until humans have actually been harmed is contrary both to the tendency of independent scientists to be conservative about exposing humans to risks of harm and to the social constraints against using humans as "guinea pigs."

An assessment of the disease potential of a particular contaminant in terms of the component factors discussed in this section will enable a regulator to determine the severity and risk elements of Judge Wright's formula. A regulator can appropriately supplement scientific data weighing in the additional factors. Consideration of these additional factors may not in itself resolve the question facing a regulator or set him on a precise course. These factors do provide, nonetheless, a framework for ensuring that regulatory decisions take [5 ELR 50175] into account the important public health policy considerations.

D. Social Equities

1. Weighing Inputs from Interested Parties: The Role of Public Interest Groups

Industry makes its presence felt in regulatory agencies through continuous informal contact with agency staff and by representation in formal agency proceedings. The resources of industry for influencing agencies in both these ways are great, and are often reinforced by aid from legislators.

The tendency of regulatory agencies to identify with corporate interests has been counteracted to some extent since the early 1960's by the emergence of "public interest groups." The resources of these groups, however, both in terms of financial support and legal and scientific personnel, are miniscule compared to those of industry. Given this vast differential in manpower and financial resources, there may be no easy way to redress the imbalance in industry versus public interest group influence in the informal day-to-day contacts which determine, to a large extent, an agency's formal decisions.

In practical terms, giving adequate weight to the value judgments of public interest groups means that a regulator should include in any risk-benefit assessment the probable costs to society if regulation does not occur or protection of health is subjugated to economic or political influences in some other way. It is much easier to determine the effects of banning a chemical on a company's profit-and-loss sheet than the effects of not banning a chemical on human health, and, implicitly, the economic health of society, 20 years or four generations from the time at which regulation is under consideration. Difficult through it may be to quantify possible or probable human illness, such quantification is necessary to determine the full cost to all of society of particular forms of environmental contamination. If such quantification is not attempted, the costs of the detrimental effects on health of economic activity will not be borne by that activity, but spread to all of society, a diffusion which may lead to vast misallocation of societal resources.

The presence of public interest groups may also serve a further useful function, that of monitoring and exposing political meddling, particularly by members of Congress and their staff, in the agency decision-making process. The 1974 case of the Mississippi chickens demonstrates that public interest groups can effectively counter political pressure on regulators. Millions of chickens in Mississippi became contaminated by polychlorinated biphenyls (PCB's) because of inadvertent PCB contamination of chicken feed. Monitoring determined that the PCB levels in the chickens far exceeded the "action level," the concentration at which the chickens would be considered unfit for human consumption and subject to regulatory action. Despite the high levels of PCB's, Mississippi representatives on Capitol Hill attempted to pressure EPA into taking no regulatory action against the chickens. The Environmental Defense Fund, hearing about the political pressures being exerted behind the scenes, gave wide publicity to the matter. Although it is not possible to say that the EDF activities were the direct cause of EPA's decision to stand fast and designate the chickens as unsafe, necessitating their destruction, public exposure of the political meddling probably had some salutary effect.

2. Placing the Burden of Demonstrating the Safety of New Products on Industry

Properly locating the burden of showing that new products, processes, and substances are safe before they are widely marketed is a question of growing concern for legislators and regulators in the environmental-public health area. Should the government have to prove that a substance is unsafe before it can take regulatory action, or should industry bear the burden of demonstrating that a product is safe before industrial or commercial use is allowed?

Technological innovations occur today with dazzling speed. Polyvinyl chloride (PVC) first moved into consumer markets after World War II, and it has experienced phenomenal growth since its commercial introduction. PVC is currently the second most popular plastic, and PVC products are ubiquitous in the American home and environment.Thus, the PVC industry was solidly entrenched by 1974, at which time data were made public demonstrating that inhalation of vinyl chloride monomer (VCM) could cause cancer in humans. (VCM polymerizes to form PVC.) The option of immediately discontinuing PVC manufacture to avoid exposing workers to VCM was not considered a realistic one due to the enormous economic importance of PVC plastics. VCM was not adequately tested before it first came into use for polymerization into PVC in the late 1930's; if VCM had been properly tested and industry or the government had regulated human exposure to vinyl chloride consistently with demonstration of cancer induction in mice and rats, the PVC industry probably would never have grown to a size where it has come to be considered irreplaceable for economic reasons.

The vinyl chloride story shows that with the pace and ubiquity of technological innovation, by the time data are available demonstrating that a new substance can cause human injury, large numbers of the public may have been exposed to risk of severe harm. It is therefore fair to demand that manufacturers be required to demonstrate the safety of new products and processes before significant human exposure occurs. Legislation which provides for such pre-market testing will be strongly resisted by industry, but testing costs are small compared with industry profits and the costs to society of injury, illness and death caused by inadequately-tested products and pollutants.

Legislation should also require that industry bear the burden of demonstrating the safety of existing products or processes; this could probably be accomplished with [5 ELR 50176] minimal inconvenience to industry by allowing for establishment of priorities for which products or processes pose the highest risk of severe harm to human health, and testing the high risk products or processes first.

This section began by addressing the question: "What factors should be considered in environmental health decision making?" Although scientific data may prove ambiguous and policy considerations difficult to quantify and weigh, it is relatively simple to identify the relevant scientific data and social policy considerations. The further question faced by a regulator — what decision to reach once relevant data have been collected and social policy considerations factored into the decision-making process — is, however, considerably more difficult.

The question boils down to whether to proceed with regulation given an imperfect and ambiguous scientific data base or wait until further data have been collected. After giving all the relevant factors full consideration, the regulator should apply a simply rule of thumb: "When in doubt, regulate sooner rather than later." Such a rule is especially sound for regulation of environmental factors, such as cancer-causing chemicals, minute doses of which can cause irreversible disease with a long latent period. After all, certain diseases like cancer or lead-caused nervous system damage (encephalopathy) cannot be reversed once initiated, but regulations can be reversed or modified to reflect changed conditions or new data.

Still better, legislation requiring manufacturers to test their products for safety before human exposure occurs would relieve regulators of decisions of this sort, and would obviate actions by industry, certain scientists and others seeking delays in regulations until after a body count has been obtained. Waiting for a body count before regulating is an approach which is hardly the mark of a civilized society.

V. Current Conduct of Public Health Research: Why a Better Data Base for Public Health Decisions is Slow to Come About

Research into the biochemical mechanisms of disease, testing and monitoring generate the data for decision making in the environmental health area. For research to be useful, it must be conducted under conditions that guarantee objectivity of experimental design, data collection and analysis. Insofar as political or economic pressures are brought to bear on researchers or agencies to shape studies to a preconceived regulatory end, the objectivity of research is lost and the data become useless for regulatory purposes.Such subversion of research activities actually occurs in regulatory agencies, and steps must be taken to ensure that environmental health research is carried on in an atmosphere which fosters objectivity.

A. Government Research on Substances Already Banned

A number of existing substances have been banned by one government agency or another due to the combined effects of legislative restrictions and generation of a scientific data base which requires a regulatory ban. In circumstances where scientific data and legislative strictures combine to make a regulator consider banning a substance, it is imperative that the scientific data be given exceedingly thorough analysis before regulatory action is undertaken. Once a substance has been banned, the government should not conduct or finance any further studies to determine whether the banned substance is "safe;" such studies should be carried on, if at all, by the manufacturer of the product or some other non-governmental institution interested in the marketing of the substance. While Rep. Delaney (D-N.Y.) holds to this viewpoint, there are those in Congress who do not share it.

In 1973, Rep. Jamie Whitten (D-Miss.), working through the Appropriations Subcommittee which he heads, ordered the FDA to perform, among other tasks, "additional studies with respect to the so-called Delaney Amendment." In general, the studies had two purposes:

1. To develop a data base which could be used to call into question or overturn specific regulatory decisions which Rep. Whitten opposed, such as the bans imposed on DDT and aldrin/dieldrin by EPA and DES by FDA, and the application of food additives of the Delaney clause; and

2. to generate scientific data which could be used to demonstrate that the Delaney Clause is unnecessary and should be overruled by Congress.

The research ordered by Rep. Whitten is being conducted at a jointly-funded FDA-EPA facility called the National Center for Toxicological Research, located at Jefferson, Arkansas. The NCTR has been criticized by respected cancer researchers³⁶ for conducting research the quality or objective of which is not calculated to bring about significant advances in understanding of disease processes. The NCTR operation is, however, strongly supported by Rep. Whitten, industry, the FDA and certain officials at EPA.

The studies being done at NCTR in response to Rep. Whitten's second set of requests are aimed at demonstrating that there are thresholds, or safe levels of exposure, for chemical carcinogens to which humans may be exposed. Were this the case, the Delaney Clause would be unnecessary and regulation of carcinogens could be significantly relaxed.

Most cancer researchers do not believe there is a safety threshold for chemical carcinogens; this belief is based on understanding of molecular events in the inception of carcinogenesis, and runs counter to theories adhered to by classical toxicologists and industry scientists. Even if thresholds did exist, they could not be demonstrated by any experiments currently available, including the studies currently being conducted at the NCTR. According to Dr. Frank Rauscher, Director of the National Cancer Institute, "there is no practical [5 ELR 50177] scientific method to prove experimentally the safety of any level of exposure to a carcinogen."³⁷ It would seem that funds devoted to determination of "safe" levels of carcinogens would be better spent on other biomedical research or on testing of substances for adverse health effects.

FDA's compliance with the Whitten order for "additional studies with respect to the so-called Delaney Clause" and the congressional appropriation of funds for that purpose have been criticized by Rep. Delaney:

I see this type of funding as just another attempt by FDA to delay their enforcement of the present law [the Delaney Clause] which is on the books and has been since 1958. To be perfectly frank, we are witnessing another stall. Government is responsible for protecting the consuming public from all suspected cancer-causing substances, and FDA's reluctance to take positive steps in this area is an outright dereliction of duty, besides the fact that it is a disastrous waste of money….

I emphasize that we must not have any further Government studies on the Delaney Amendment at the double expense of the consumer's pocketbook and his physical well-being. We could all be fair [sic] better if studies ceased or were curtailed and the law enforced against dangerous cancer-causing substances.³⁸

As an indication of the "stall" tactics practiced by FDA in enforcing the Delaney Clause, FDA made an attempt to ban the chemical DES (diethylstilbestrol) for use in cattle feed. The regulation was overturned in court due to procedural irregularities in the issuance of the regulation.³⁹ Rather than reissue the regulation using proper procedures to defend its actions in court and try again, the FDA is now avoiding further attempts at regulation of the compound until the DES experiments going on at NCTR are completed. Since DES is known to cause cancer in humans as well as in animals, the purposes of the NCTR experiments are open to question; FDA, at least in this instance, seems to be more concerned about avoiding another court confrontation than it is about ending a threat to the public health.

Government expenditures for biomedical research are not lavish; this is especially true in the area of environmental carcinogenesis, despite the fact that up to 90 percent of cancer may be caused by environmental chemicals (including cigarette smoking).⁴⁰ Once a chemical has been demonstrated to be carcinogenic in animals or in man, unless the chemical is an irreplaceable drug or a natural contaminant which cannot be fully eliminated from the environment by control measures, why should scarce government research funds be used to further study the chemical? Doesn't such further study really give industry "two bites of the apple?" The government in effect ends up paying for the research which can be used to upset its own case. If industry wishes to prove to a regulatory agency that a chemical determined by the agency to be a carcinogen is really not a carcinogen, or if industry wishes to demonstrate to Congress that the scientific underpinnings of the Delaney Clause are faulty, it should conduct the research itself.

B. The Benzidine Case

Another example of misuse of government research funds by the NCTR concerns a dyestuff base called benzidine. Benzidine is a potent carcinogen in man; the incidence of bladder cancer among people with occupational exposure to benzidine may approach 25 percent.⁴¹ Although the major former manufacturers of benzidine in this country have stopped manufacturing the compound because of the demonstrated health hazard, and although there are color bases which can be substituted for benzidine, many small dye companies in this country routinely synthesize benzidine in the course of making dyes. Thus, benzidine is being expelled into waterways and EPA has made some attempts to set "safe" emission levels for the carcinogen. NCTR was called in by EPA to give help in setting the "safe" levels, and NCTR has in progress experiments on benzidine.

On March 20, 1975, in anticipation of upcoming Appropriations Committee hearings, Dr. Alexander Schmidt, Commissioner of the FDA, reported to Rep. Whitten by letter on work being conducted at NCTR in response to Rep. Whitten's mandates. Among other items, Dr. Schmidt noted that: "NCTR is presently conducting a study to provide toxicological data that can be used to establish safety levels for benzidine."⁴²

This assurance runs directly counter to the statement by the Director of the National Cancer Institute, quoted above, to the effect that it is not possible to determine safe levels for any chemical carcinogen.

Representatives of the AFL-CIO, in testimony before the Senate Commerce Committee in the spring of 1975 on S.776, the Toxic Substances Control Act, criticized those activities at the NCTR designed to determine "safe" levels for benzidine. According to the AFL-CIO:

Benzidine is an interesting choice for [NCTR's] research. Substitutes exist for most uses of this highly carcinogenic chemical and its manufacture and use, if necessary, are totally controllable. Why, then, should any agency want to set a standard permitting any level of [5 ELR 50178] risk? Any firm wishing to justify [benzidine] use should bear the burden of determining the risks at its own expense. To do otherwise is to subsidize and favor the industry.⁴³

If Congress is serious about its support ofthe principles of protection of public health set out in the Delaney Clause, the OSHA, FEPCA and the Toxic Substances Control Act now under consideration in both Houses, there should be strong congressional opposition to attempts by NCTR to carry on research on biochemical mechanisms of disease designed to frustrate the intent of the Congress' legislation. Congress, however, has to date consistently acquiesced in Rep. Whitten's mandated studies and the activities at NCTR.

C. Who Should Do Research on Biochemical Mechanisms of Disease?

Given limitations on personnel and funds, would regulatory agencies do better to increase their disease mechanism research activities or step up their regulatory efforts? The pressures brought to bear on regulatory agencies and identification of these agencies with the industries they are supposed to regulate may make it imposible for the agencies to conduct research which would be accepted as unbiased, objective and valid by qualified independent scientists. Also, given the long latent periods associated with such diseases as cancer and the irreversible effects of chronic long-term exposures to certain pollutants, devoting regulatory agency resources to research (to the exclusion of regulation) may nean that regulation will not occur until human illness has actually been demonstrated or until millions of individuals have been needlessly exposed to risk of severe harm. Research into disease mechanisms should be encouraged, but the prime responsibility for conducting such research should be left to government research agencies, such as the National Institutes of Health, and to independent scientists working at universities and other well-respected research organizations.

Biomedical research excellence in this country is centered in the federal government, in the National Institutes of Health. Research on biochemical mechanisms of disease performed by the regulatory agencies has been criticized by distinguished scientists as being of poor quality. Thus, the National Academy of Sciences as well as the General Accounting Office have criticized the research efforts of EPA,⁴⁴ and an ad hoc panel of cancer researchers convened by HEW has criticized the cancer research efforts at the NCTR.⁴⁵ Questions have been raised as to whether reputable scientists can be recruited to work in regulatory agencies removed from the intellectual stimulation and facilities available at research agencies, universities or other independent research institutions.

If only for the sake of best use of funds, it would seem reasonable to require that research institutions perform research into disease mechanisms while regulatory agencies test and monitor and work to improve testing and monitoring methodologies. Furthermore, there is a serious need for intensified efforts to improve communications between regulatory and research agencies so that disease mechanism research data are communicated speedily to regulatory agencies. Where feasible, personnel conducting research on disease mechanisms should be sensitized to the data requirements of regulatory agencies and encouraged to conduct research which might eventually solve some of the scientific coundrums confronting regulators in the environmental health area.

VI. Conclusions

Ideally regulatory decisions in the environmental health area should be made on the basis of a perfect scientific data base. However, to wait indefinitely for a perfect data base rather than regulate when a data base sufficient to satisfy the strictures of enabling legislation has been accumulated, is to risk dooming millions of Americans to exposure to substances which could cause injury or death. In most instances, an unambiguous scientific data base recording the human health effects of a substance could not reasonably be developed within the foreseeable future even with increased financial support for agency research activities.

Legislation passed by Congress places a strong stress, at least rhetorically, on protection of public health. When a regulator must make a decision working from an imperfect data base, he has no choice but to take into account the factors that constitute criteria for judging risk and severity of harm, and weight his decision presumptively in favor of protection of the public health.

Ultimately, the pervasive nature and rapid pace of technological innovation in American society requires that industry bear the burden of demonstrating the safety of its new products and processes before significant human exposure occurs, and industry should also be required to demonstrate the safety of existing products and processes. That industry will resist such a shift in the burden of proof is inevitable, and change will be slow in coming. However, accumulating evidence that environmental contaminants pose serious risks of severe harm to human beings should eventually persuade the public and Congress that the time has come to require industry to demonstrate the safety of its products on subjects other than the American public.

1. See generally Man's Health and the Environment: Some Research Needs, Report of the Task Force on Research Planning in Environmental Health Science, Department of Health, Education and Welfare, Public Health Service, National Institutes of Health, National Institute of Environmental Health Sciences (March 1970).

2. United States v. Reserve Mining Company, 4 ELR 20573 (D. Minn., Apr. 20, 1974; May 11, 1974); rev'd sub nom. Reserve Mining Company v. United States, 4 ELR 20598 (8th Cir., June 4, 1974); confirmed on merits, 514 F.2d 492 (8th Cir., Mar. 14, 1975; Apr. 8, 1975).

3. Reserve Mining Company v. United States, 4 ELR 20598, 20603 (8th Cir., June 4, 1974).

4. In re: Shell Chemical Company, 4 ELR 30017 (Oct. 1, 1974); Environmental Defense Fund v. Environmental Protection Agency, 5 ELR 20243 (D.C. Cir., Apr. 4, 1975) (appeals court affirms Administrator's suspension order).

5. ELR Comment, EPA Delays Ban on Cancer-Producing Pesticide Dieldrin, 4 ELR 10104 (Aug. 1974).

6. Ethyl Corporation v. Environmental Protection Agency, 5 ELR 20096 (D.C. Cir., Jan. 28, 1975); vacated for rehearing en banc, 5 ELR 20450 (D.C. Cir., Mar. 17, 1975).

7. "Threshold effects" are at the center of a continuing controversy regarding illness caused by exposure to environmental contaminants. The complexities of threshold versus no threshold arguments are not treated in this article, but merit further consideration by anyone interested in environmental health regulation.

A few rather simplified statements on thresholds may be useful for the purposes of this article. First, in general, if a chemical or other environmental contaminant acts to initiate disease at the level of the genetic macromolecules (i.e., DNA) within cells, there is no demonstrable "safe" level for the environmental contaminant other than zero. This is true because a "single hit" on the genetic material may suffice to cause a chemical change which may initiate disease.

Mutations are definitely linked to reactions by various chemical and physical agents with genetic material. In general, cancer also seems to be initiated by such reactions. Little is known, however, of the steps in the development of cancer between the initial reaction at the level of the genetic macromolecule and the clinical manifestation of cancer by a human being at the tissue, organ or organism level.

Second, there are diseases which do not seem to originate via interaction of the causative factor or factors with genetic macromolecules. Instead, studies demonstrate that other cell structures are involved. Very simply put, these diseases occur through poisoning of cell structures, such as the structures which carry out the conversion of food to energy, and/or through the inhibition of the structures' functions. Cell structures can take only so much poison before they are inactivated, but a cut-off level for the amount of poison which can be tolerated can be demonstrated. This cut-off level can be called a threshold; above the threshold level, harm is done to the cell structure. There may be a level of poisoning at which reversible harm is done to the cell structure; that is, the damage can be repaired and the harm undone. However, there may also be a higher level, above which irreversible harm is done to the structure and/or functions. In any case, a "single hit" theory — one molecule initiating possibly irreversible steps to disease — does not seem to hold for these diseases.

There is a very important distinction between category one and category two diseases. Since the genetic material is affected in category one diseases, the disease-producing change in the genetic material caused by reaction between genetic material and an environmental agent can be passed on each time the genetic material duplicates. Thus, the diseases in category one are essentially self-propagating at a molecular level, and can lead to production of more and more cancer cells as time goes on. When lead poisoning results in the death or malfunctioning of a nerve cell, that one cell is affected but the malfunction or death is not propagated further.

An internal EPA memorandum provides a useful discussion of the distinction between category one and category two diseases:

The fact that a lethal tumor may be derived from a single cell is the essential difference between carcinogenesis and other toxic effects. A single cancer cell replicates itself and may therefore kill an organism. A single dead nerve cell simply stops functioning and is likely to affect the organism little. It is the self-replicating nature of cancer that leads to the argument that there may be no threshold for carcinogens while there apparently are thresholds for many other toxic effects…. Mutagenesis is also a self-replicating effect and the same arguments would hold there.

EPA memorandum titled "Dr. Upholt's response to questions from the Senate Commerce Committee" from Elton R. Homan, Ph.D., through Glenn E. Schweitzer, Director, Office of Toxic Substances to William Upholt, Ph.D., Senior Scientific Advisor, EPA; April 18, 1975 at 2.

8. Supra, n. 6 at 20099.

9. 21 U.S.C. § 348(c)(3)(a).

10. The pertinent portions of the Delaney Clause read:

… no additive shall be deemed to be safe if it is found to induce cancer when ingested by man or animal, or if it is found, after tests which are appropriate for the evaluation of the safety of food additives, to induce cancer in man or animal … 21 U.S.C. § 348(c)(3)(a).

The phrase "tests which are appropriate for the evaluation of the safety of food additives" actually gives a bit of discretion to the regulator, since he has the power to determine whether test data are suitable for regulatory purposes.

11. 42 U.S.C. § 1857.

12. Strictly, but simply, speaking, arsenic may be a co-carcinogen rather than a true carcinogen. A co-carcinogen is a substance or agent which does not cause cancer completely on its own but must interact with another substance or agent or with multiple substances or agents in order to cause cancer.

13. Maltoni & Lefemine, Carcinogenicity Bio-Assays of Vinyl Chloride: Current Results, 246 Annals of the N.Y. Academy of Sciences 195 (1975).

14. Id.

15. Tabershaw & Gaffey, Mortality Study of Polyvinyl Chloride Workers, paper delivered at the Working Group-Toxicity of Vinyl Chloride-Polyvinyl Chloride, N.Y. Academy of Sciences, New York City, May 10, 1974.

16. The oil and auto industry Coordination Research Council had and may still have an important role in the design, selection and supervision of EPA air pollution research projects. Liaison is maintained with EPA through the APRAC, the Air Pollution Research Advisory Committee. According to a CRC publication: "[CRC's APRAC] is a group of technical and medical experts from the automotive and petroleum industries and the Environmental Protection Agency…. The EPA shares in the direction of all [research] projects and, on a selective basis, contributes financial support in a majority of projects." Program of 1973 Automotive Air Pollution Research Symposium at 2, Coordinating Research Council, Inc., March 7-9, 1973. Earlier controversies regarding EPACRC ties were described in Auto Pollution: Research Group Charged with Conflict of Interest, 181 Science 732, 1973. Although EPA has given assurances that it is no longer coordinating its research activities with CRC, allegations have been made that the relationship still exists and is still influencing EPA's research efforts. Dr. Goldsmith Criticizes EPA's Health Research Links with Auto, Oil Industries, 13 Environmental Health Letter no. 24, 5 (Dec. 15, 1974).

17. Supra, n. 9.

18. Supra, n. 11.

19. The required automobile emission reductions are set forth at 42 U.S.C. § 1857f-1(b)(1)(A) (carbon monoxide and hydrocarbons to be reduced 90 percent between 1970-1975), 42 U.S.C. § 1857f-1(b)(1)(B) (nitrogen oxides to be reduced 90 percent between 1971 and 1976).

20. Humans are genetically diverse and exposed to many environmental contaminants whose interactions and interactive effects on health are largely unknown. For these reasons alone, skepticism if indicated in extrapolating data obtained in controlled-environment studies of genetically homogeneous animal populations to man for the purposes of establishing safe levels for human exposure.

21. Recently, data have been published which indicates that home or salon coal-tar hair colors can cause mutations in certain bacteria. Ames, Kammen, & Yamasaki, Hair Dyes are Mutagenic: Identification of a Variety of Mutagenic Ingredients, 72 Proc. Nat. Acad. Sci. USA 2423 (1975). The demonstration that certain hair dye components and the dyes themselves are mutagenic has led to speculation that the coal-tar hair colors may cause cancer; vide Ames, supra. There are no current statutory requirements that cosmetics be tested for safety before they are marketed.

22. A. Anderson, Jr., Daughter of a Hidden Plague, N.Y. Times Magazine 20 (Oct. 27, 1974).

23. Commitee on Principles of Decision Making for Regulating Chemicals in the Environment, Environmental Studies Board, Commission on Natural Resources, National Research Council, National Academy of Sciences, Decision-Making for Regulating Chemicals in the Environment, XXV; (July 1974).

24. Id.

25. B. Castleman, Flight of Hazardous Industries to "Unregulating" Countries, University of Maryland, College Park, Md. (Jan. 1975).

26. Brodeur, Expendable American, New York, Viking Press, 249 (1974).

27. Public Citizen, Health Research Group Charges that Rohm and Haas Company Withheld Data on Lung Cancer from Workers and the Government — Formation of Coalition of Widows and Children to Seek Compensation is Announced (Press Release) Washington, D.C. (Oct, 2, 1974).

28. Society of the Plastics Industry, Inc. v. Occupational Safety and Health Administration, 5 ELR 20157 (2d Cir., Jan 31, 1975); stay of order denied sub nom. Firestone Plastics Co. v. Department of Labor, 95 S. Ct. 1444 (Mar. 31, 1975).

29. Data supplied to the U.S. Senate Commerce Committee staff by B.F. Goodrich Co., on April 17, 1975, indicate that the average VCM concentration (eight hour time average not specified) for the Louisville PVC plant dropped from greater than 30,000 ppb in January 1974 to approximately 3000 ppb in late March, 1975. The Louisville facility is one of the oldest PVC polymerization facilities in the country.

30. Castleman, supra, n.24.

31. Individuals with occupational exposure to hazardous substances may comprise small groups with very heavy exposures. Workers should not be used as guinea pigs to protect the less-heavily-exposed general public; the criterion for safety a regulator should apply for any hazardous substance is that wherever possible the substance should be demonstrated to be safe before any significant human exposure occurs.

32. Ad Hoc Committee on the Evaluation of Low Levels of Environmental Chemical Carcinogens, Evaluation of Environmental Carcinogens Report to the Surgeon General, United States Public Health Service (Apr. 22, 1970).

33. Id.

34. Of course, a single massive exposure may also suffice to elicit illness, but a member of the general public is unlikely to have such an exposure. Occupational exposures may be very heavy, and much of the data on the effects on humans of massive exposures to toxic substances such as lead comes from clinical reports of occupational disease.

35. Selikoff, Hummond & Churg, Asbestos Exposure, Smoking, and Neoplasia, 204 J.A.M.A. 106 (1968).

36. Report of the Director's Ad Hoc Committee on Testing for Environmental Carcinogens, National Cancer Institute, National Institutes of Health, U.S. Department of Health, Education and Welfare (Aug. 1973).

37. Letter from Dr. Frank Rauscher, Director, National Cancer Institute, to Sen. Edward M. Kennedy, April 9, 1974.

38. 119 Cong. Rec. H4811 (daily ed. June 15, 1973).

39. Chemetron Corp. v. U.S. Dept. of Health, Education and Welfare, 4 ELR 20158 (D.C. Cir. 1974).

40. Man's Health, supra, n. 1.

41. Auerbach, Investment of Bladder Cancer, Washington, Post; May 29, 1973; at A2.

42. Letter from Alexander M. Schmidt, M.D., Commissioner on Food and Drugs, to Rep. Jamie L. Whitten, Chairman, Appropriations Subcommittee for Agriculture and Related Agencies (Mar. 20, 1975).

43. Letter from Jacob Clayman, Secretary-Treasure, AFL-CIO, to Senator Philip A. Hart, April 25, 1975 (to be printed in the record of hearings on S.776, the Toxic Substances Control Act).

44. Letter from the NAS-NRC Review Committee on the Management of EPA's Research and Development Activities, Robert W. Berliner, Chairman, to Russell Train, Administrator, Environmental Protection Agency, August 27, 1974; EPA Research and Development Program Draws GAO, Congressional, Staff Review, 5 Env. Rep. 275 (1974).

45. Report, supra, n. 34.

5 ELR 50165 | Environmental Law Reporter | copyright © 1975 | All rights reserved