Precautionary Controls: D.C. Circuit Upholds EPA's Phase-Down of Gasoline Lead Additives in the Interest of Public Health

[6 ELR 10100]

The entire gamut of judicial responses to precautionary administrative action aimed at potential environmental threats to public health marked the D.C. Circuit's recent en banc decision¹ upholding the Environmental Protection Agency's 1973 regulations requiringprogressive phasing out of gasoline lead additives in the interests of public health. The closeness of the vote,² the depth of the differences between court majority and minority, and the lengths to which each side went to develop its arguments,³ all underscore the difficulties that the judiciary continues to have in reviewing future-oriented administrative action, even when supported by the relatively specific precautionary statutory authority of the Clean Air Act. And it is only by going to great lengths to minimize the uniqueness of the kind of risk assessment and uncertain prediction engaged in by the EPA Administrator in regulating lead that Judge Wright's opinion for the majority makes a beginning at incorporating such actions within the parameters of normal judicial review.

At issue was the Administrator's decision to promulgate regulations phasing out automotive lead additives under § 211(c)(1)(A) of the Clean Air Amendments of 1970 (CAA).⁴ More specifically, the dispute centered around (1) EPA's interpretation of the statutory term "will endanger public health or welfare," which sets the threshold determination to be made before the Administrator can regulate fuel additives; (2) the adequacy of EPA's data base of scientific and medical studies to support its conclusion that airborne lead from automobile emissions presents a "serious risk of harm" to a significant portion of the urban public and a large fraction of urban children; (3) the proper scope and method of judicial review under the arbitrary and capricious standard as interpreted by the Supreme Court in Citizens to Preserve Overton Park v. Volpe,⁵ and (4) the procedural propriety of the manner in which the EPA promulgated the final regulations and their supporting data base. The court majority and minority were totally at odds over all these issues.

The Statutory Standard

As to the statutory "will endanger" standard, Judge Wright approved the Administrator's paraphrase of it as permitting regulation of substances whose emissions "present a substantial risk of harm" to health. The majority concluded that § 211 is a precautionary standard which authorizes action by the Administrator without factual proof, and so necessarily gives him power to make quasi-legislative policy judgments and to assess risks of harm.

Judge Wright adduced a variety of arguments to support this interpretation. He invoked the plain meaning of the term "endanger" (to threaten harm — no harm necessarily occurs); noted the deference due to the Administrator's interpretation of the Clean Air Act under Train v. Natural Resources Defense Council, Inc.;⁶ compared the language of § 211 with §§ 108 and 202;⁷ examined the legislative intent as revealed by the conference committee's deletion of a specific "findings" requirement for the threshold determination of endangerment; adverted to the Eighth Circuit's interpretation of the FWPCA's "endangerment" standard in Reserve Mining Co. v. Environmental Protection Agency,⁸ and [6 ELR 10101] invoked the canon of liberal construction of health-related statutes.⁹ Overall, he stressed the need for flexible standards to permit regulation before harm to human health occurs in the face of man's technological ability to bring about environmental modifications of unprecedented scale, and in view of the intrinsic limits of medical evidence from toxicological, clinical and epidemiological studies. The majority concluded:¹⁰

Where a statute is precautionary in nature, the evidence difficult to come by, uncertain or conflicting because it is on the frontiers of scientific knowledge, the regulations designed to protect public health, and the decision that of an expert administrator, we will not demand rigorous step-by-step proof of cause and effect…. [W]e … hold that in such cases the Administrator may assess risks.

The dissent rejected the view that the Administrator has any power under § 211 to make quasi-legislative policy judgments. First, as a philosophical matter, it refused to distinguish between "risks" and "facts" or between "actual harm" and "potential harm." In its view,

the best (and only convincing) proof of potential harm [from lead in exhaust emissions] is what has occurred in the past … , from which the Administrator can logically deduce that the same factors will produce the same harm in the future.¹¹

Second the dissent read § 211(c)(1)(A) as requiring the Administrator to make a reasoned "factual" determination (of endangerment) based solely on the medical and scientific evidence. Judge Wilkey based this interpretation on § 211(c)(2)(A),¹² which requires the Administrator to consider all relevant medical and scientific evidence available to him before undertaking regulation of fuel additives. This requirement, in his view, indicates a congressional intent that such evidence must serve as the sole basis for the Administrator's determination.

This interpretation seems overextended. An obligation to "consider" something is not the same as a requirement to base a decision on it exclusively. Moreover, the legislative history cuts sharply against the minority's interpretation. The conference committee deleted from the House version of the bill language which would have explicitly obliged the Administrator to base his determination of the dangers posed by fuel additives on "specific findings derived from relevant medical and scientific evidence…."¹³

The Adequacy of the Evidence

In applying their views of the statutory standard, the majority and minority disagreed both over the adequacy of the evidence and the proper scope of judicial review under the arbitrary and capricious standard of the Administrative Procedure Act. Both opinions recognized, as had the Administrator, that the available studies on the effects of airborne lead, taken either separately or together, were inconclusive. This is due in large part to the diversity of natural and man-made sources of lead intake into the human body, and the fact that the major source is dietary intake. The majority noted that the inherent limits of clinical and epidemiological studies are exacerbated by the multiple source phenomenon. It is considered unethical to dispense substances known to be toxic (like lead) to the human subjects of clinical experiments, and the statistical correlations of epidemiology mask the discrete effects of multiple lead sources.

Taking into account these limits, the EPA Administrator concluded that a small but significant portion of urban adults and a large number of urban children have elevated blood lead levels due in part to lead from auto emissions. This conclusion was based on intra-city studies showing higher lead levels in urbanized than suburban neighborhoods and on studies of occupational groups exposed to unusually high levels of automotive lead emissions (traffic police, tunnel personnel, garage mechanics, etc.).

The dissent viewed this record as supplying no evidence of elevated lead levels in a substantial portion of the general public. In other words, it refused to permit any inference from the occupational exposure studies, on the ground that the experience of such persons is not typical of the rest of the population.

The majority by contrast approved the Administrator's extrapolation from the experience of occupational groups most heavily exposed to airborne lead. In its view, this was exactly the kind of precautionary policy determination appropriate under the "will endanger" standard, for it makes sense to project that increasing auto use may someday bring the exposure level of the general public up to that of these special groups.

The Arbitrary and Capricious Standard of Review

A further factual dispute between majority and minority opinions was spawned by their differing interpretations of the arbitrary and capricious standard of review, particularly the Supreme Court's gloss concerning "clear error of judgment." The two agreed at the outset that a searching review of the facts is necessary in technically complex cases like Ethyl.¹⁴ For the majority, [6 ELR 10102] the purpose of such thorough immersion in the facts is to educate the court so as to enable it to determine whether the Administrator "took into account the relevant factors" in making his decision. For the dissent, the purpose of the searching review of the facts is to determine whether the Administrator made an error of judgment, i.e., failed to supply a "rational" basis for relying on some studies and rejecting others. This review standard was viewed as rather stringent by the dissent.¹⁵ It found clear error of judgment in the Administrator's failure to disclose what it considered a consistent rationale for accepting some studies and rejecting others. For example, the Administrator had excluded a number of studies on the methodological grounds of inadequate controls for non-airborne sources of lead. But, pointed out the dissent, the same methodological flaws infected the studies relied on by the Administrator.

What the minority failed to credit in finding a clear error of judgment in EPA's choice of studies is a major tenet of epidemiological evidence. As the majority pointed out,¹⁶ positive statistical correlations are considered more significant than negative, due to the myriad conflicting variables at work in real-world situations, which tend to mask each others' effects. In other words, negative results are more likely to be "false" than positive ones. This feature of epidemiological results supports the Administrator's choice of studies producing "positive" results.

Procedural Improprieties?

The final but not the least bitter of the disagreements between majority opinion and dissent in Ethyl concerned the procedural propriety of EPA's issuance of its final regulations in 1973. Here it was the dissent which mounted the more lengthy argument, but in the end the degree of prejudice to petitioners from the EPA's actions seems slight indeed. To put the matter in perspective, it is necessary to summarize the procedural history of EPA's lead regulations. EPA commenced with an advance notice of proposed rulemaking in January 1971. In February 1972 regulations were first proposed, supported by a document which may be entitled "first health document" (subsequently revised several times); an original 90-day comment period was extended by 30 days, and EPA held public hearings in Washington, D.C., Los Angeles, and Dallas.

Extensive criticism of the first health document led EPA to repropose the regulations in January 1973 along with a second health document; these too were made subject to public comments. Finally, in October 1973, a panel of the D.C. Circuit ordered¹⁷ the Administrator to make up his mind within 30 days whether to regulate lead under § 211. On November 23, he issued final phase-down regulations and a third health document. This last health document had not been circulated for comments as such, but a nonfinal draft had been circulated in October, and the EPA had previously placed in its public file every new piece of information received and to be relied on in the third health document.

The dissent's major claim was that petitioners had no proper opportunity to make timely comments on the third health document, because it did not appear in final form until the regulations were out, and that the EPA's public notice of new materials was deficient because EPA failed to indicate which of many new studies it was going to rely on in the third health document. Assuming that the Administrative Procedure Act imposed an obligation on EPA to point out in advance exactly which new studies it intended to rely on, it does not seem that the petitioners would be able to establish any particular prejudice from the EPA's omission. First, the new studies that the Administrator relied on were no more conclusive in themselves or in the aggregate than earlier studies; the objections to earlier studies (developed at length in the dissent as summarized supra) applied equally to the new ones. Second, the EPA's approach to analyzing the part played by air-borne lead in overall body lead levels had been so severely criticized in the first health document that EPA changed its ground in the second health document. Specifically, it gave up trying to relate human blood levels directly to air lead concentrations. This shift in approach, which was the most important matter addressed in the health documents, was maintained without further change between the second and third health documents, so there was no need to make the latter available in advance for comment.

Domesticating Risk-Based Decision Making

The above summary of the clash between Judges Wright and Wilkey hardly does justice to the sweep of their thought. In retrospect, one salient feature is the length to which Judge Wright went to demonstrate the necessity of future-oriented government regulation on the basis of incomplete information in this era of powerful man-induced modifications of the natural order.¹⁸ The interesting thing is that such a thesis flies in the face of both science and judicial activity as usually practiced. Scientists withhold decision until high probability of "certainty" is demonstrated.¹⁹ Judges cannot avoid the obligation to decide, but may [6 ELR 10103] too easily refuse to act on the grounds of inadequate evidence and excessive speculation.

The ultimate goal of Judge Wright's opinion seemed to be to incorporate EPA's policy-based; future-oriented decision making into the mainstream of judicial experience. This he accomplished by drawing on judicial interpretations of health endangerment in traditional classes of cases (e.g., divorce cases); by invoking the statutory maxim of liberal interpretation of health statutes; by drawing parallels between the cumulative nature of scientific evidence and that of circumstantial judicial evidence;²⁰ by emphasizing that assessment of risks is a normal part of judicial and administrative fact-finding; and by relying on the small body of recent federal court decisions upholding regulation of environmental health risks, especially the en banc opinion in Reserve Mining but also Amoco Oil Co. v. Environmental Protection Agency,²¹ Industrial Union Department, AFL-CIO v. Hodgson,²² and Society of the Plastics Industry v. OSHA.²³

The Ethyl Corporation has announced its intent to appeal this decision to the Supreme Court. Whether or not the Court takes the case, the opinions in Ethyl state the opposing positions in a debate over regulation of risks and judicial review of such regulation which is sure to continue for some time to come.

1. Ethyl Corporation v. Environmental Protection Agency, 6 ELR 20268 (D.C. Cir. Mar. 19, 1976).

2. The court split five to four; the swing vote, Judge Bazelon, filed a concurrence disagreeing with the majority over the nature of judicial review. Id. at 20303-05.

3. Omitting appendices, the majority opinion ran 113 pages in slip opinion, Id. at 20268-97; Judge Wilkey's dissent, 87 id. at 20306-29.

4. 42 U.S.C. § 1857f-6c(c)(1)(A), ELR 41214.

5. 401 U.S. 402, 1 ELR 20110 (1971).

6. 421 U.S. 60, 75, 5 ELR 20264 (1975).

7. Section 108(c)(1)(A), 42 U.S.C. § 1857c-3(a)(1)(A), ELR 41203, directs the Administrator to list (for the purpose of establishing national ambient air quality standards) each air pollutant "which in his judgment has an adverse effect on public health or welfare…." Section 202(a)(1), 42 U.S.C. § 1857f-1(a)(1), ELR 41211, authorizes him to set new car emission standards for each auto emission "which in his judgment causes or contributes to, or is likely to cause or contribute to, air pollution which endangers the public health or welfare." In terms of threat to health, the majority viewed § 202 as embodying the same precautionary standard as § 211. 6 ELR at 20273. But, based on the phrase "has an adverse effect," Judge Wright concluded that § 108 is not a precautionary standard at all. Id. This reading is implicitly at odds with another lead air pollution case reported in this issue, Natural Resources Defense Council, Inc. v. Train, 6 ELR 20366 (S.D.N.Y. Mar. 1, 1976). There, the district court ruled that since the Administrator has made the threshold determination that airborne lead "has an adverse effect on public health or welfare," he has a mandatory duty to list it under § 108. This finding does not conflict with Judge Wright's observation in Ethyl, 6 ELR at 20273 n. 21, that the Administrator "appears to have a measure of discretion" in determining whether to list a pollutant under § 108, for such discretion only exists until he determines, as he has in Natural Resources Defense Council, that "in his judgment" the pollutant causes harm. What does seem inconsistent is that in Natural Resources Defense Council the Administrator has made the threshold determination that lead "has an adverse effect," on public health. As will be discussed later, all that the extensive record in Ethyl will support is the Administrator's assessment that lead poses a "significant risk" of harm, not that it "causes" such harm under a non-precautionary cause and effect standard.

8. 514 F.2d 492, 5 ELR 20596 (8th Cir. 1975) (en banc). There, the court concluded that Reserve's discharges of asbestiform wastes into Lake Superior, which supplies drinking water to Duluth, Minnesota, posed a medical risk of potential health harm which violated § 1160 of the Federal Water Pollution Control Act of 1970, former 33 U.S.C. § 1160, thus requiring abatement of the water pollution. Section 1160 speaks of "endangering the health or welfare of persons," language notably similar to § 211 of the Clean Air Act. Judge Wright observed that the Eighth Circuit (relying on his dissent in the original panel opinion in Ethyl) interpreted this language

in a precautionary or preventive sense, and therefore, evidence of potential harm as well as actual harm comes within the purview of that term.

514 F.2d at 528, quoted in Ethyl, 6 ELR 20275. The Ethyl majority likewise relied on Reserve Mining to support the view that the probability of the risk need not be more likely than not if the potential harm is of large magnitude, Id. at 20276-77.

9. Id. at 20283 & n. 65.

10. Id. at 20281.

11. Id. at 20320-21. This passage appeared in the original panel opinion, 5 ELR at 20099.

12. 1857f-6c(c)(2)(A), ELR 41214.

13. H.R. 17255, 91st Cong., 2d Sess. § 210(a)(1) (1970) (emphasis added).

14. Only Judge Bazelon, in his separate concurrence, took issue with this conclusion, 6 ELR at 20303.

15. In contrast, Judge Wright emphasis the Supreme Court's command that the reviewing court not substitute its judgment for that of the Administrator. Id. at 20285-86 n. 71. In this long footnote the court attempted to plumb the Supreme Court's intent in speaking of "clear error of judgment," a term which resembles the "clearly erroneous" standard applicable to appellate review of judicial fact finding. Judge Wright concluded that despite the resemblance the judicial standard, which is rather intrusive, should not be adopted for review of administrative action, for the latter is entitled to greater judicial deference than trial court fact finding. Id.

16. Id. at 20289, citing D. Clark & B. MacMahon, Preventive Medicine 100 (1967).

17. Natural Resources Defense Council, Inc. v. Environmental Protection Agency, No. 72-2233 (1973).

18. See also Gelpe & Tarlock, The Uses of Scientific Information in Environmental Decisionmaking, 48 S. Cal. L. Rev. 371 (1974).

19. Being under no compulsion to decide until all the evidence is in, scientists aften react impatiently to what they view as "premature" efforts to regulate on the basis of incomplete data. See, e.g., Burger, Regulation and Health: How Solid Is Our Foundation?, 5 ELR 50179 (Sept. 1975).

20. Quoting Chief Judge Shaw of the Massachusetts Supreme Judicial Court in Commonwealth v. Webster, 59 Mass. (5 Cush.) 295, 317 (1850):

[I]nferences drawn from independent sources, different from each other, but tending to the same conclusion, not only support each other but do so with an increased weight.

21. 501 F.2d 722, 4 ELR 20397 (D.C. Cir. 1974).

22. 499 F.2d 467, 4 ELR 20415 (D.C. Cir. 1974).

23. 509 F.2d 1301, 5 ELR 20157 (2d Cir.), cert. denied, 421 U.S. 992 (1975).