Before imposing such [fuel] limitations the Secretary is required to make certain specific findings as to the necessity of the imposition of such limitations.39
UNITED STATES COURT OF APPEALS FOR THE DISTRICT OF COLUMBIA CIRCUIT
Petitions for Review of Orders of the Environmental Protection Agency.
Hastie* Senior Circuit Judge, and Wright and Robb, Circuit Judges.
DECISION OF THE COURT DELIVERED BY THE HONORABLE JUDGE WRIGHT
Opinion for the court filed by Circuit Judge WRIGHT.
In these consolidated cases, many of the nation's oil companies seek review of the Environmental Protection Agency's "Regulation of Fuels and Fuel Additives," 40 C.F.R. Part 80 (1973) (hereinafter Fuel Regulations), promulgated under Sections 211(c) and (d) of the Clean Air Act. *fn1 The Regulations prohibit use of leaded gasoline in automobiles fitted with "catalytic converter" devices for controlling exhaust emissions and require widespread retail marketing of at least one grade of unleaded gasoline.
Oil refiners have for many years routinely added lead to gasoline to improve its "anti-knock" and other performance characteristics. Because air-borne lead may present a serious threat to public health, EPA has promulgated regulations, not here at issue, to limit the lead content of all gasoline. *fn2 The Fuel Regulations under review address a narrower problem: Lead emissions "poison" -- i.e. render inactive -- the catalytic converter devices on which the automobile industry is relying to reduce exhaust emissions of hydrocarbons, carbon monoxide, and oxides of nitrogen to the levels mandated for new cars by Section 202 of the Clean Air Act. *fn3 Catalytic converters will be fitted on many 1975 model new cars and on most or all 1976 model new cars. The Regulations take effect on July 1, 1974, at the advent of the 1975 model year. To secure efficient operation and widespread use of converters, the Regulations require that converter-equipped cars use only unleaded gasoline and that retail stations having a sizeable clientele offer at least one grade of such fuel.
Petitioners challenge the Regulations on a host of grounds, examined in Part II of this opinion. The root of petitioners' interest is financial. To refine gasoline without adding lead is neither difficult nor very expensive, but years of producing and delivering only leaded fuel have left deposits of the element in the containers, pipes, and vehicles by which the industry moves gasoline from refinery to automobile gas tank. Portions of this distribution network must be cleaned or replaced if the new converter-equipped cars are to be sold, used, and serviced in normal and convenient fashion. That such a cleanup would be necessary by 1975 has been known for a number of years, *fn4 but the oil industry contends that the present Regulations impose demands which are unnecessarily and unlawfully far-reaching and abrupt.
We have located problems in the "liability" provisions of the Regulations and have accordingly required the Agency to recognize certain affirmative defenses to these provisions. See Part II-D below. In all other respects, however, we find the Regulations to be valid. I. STATUTORY SETTING, THE PROCEEDINGS, AND THE REGULATIONS
These Regulations are an integral element in the complex program to reduce air pollution which Congress adopted by way of the Clean Air Act Amendments of 1970. At the core of that program, in Section 202 of the Act, is a graduated schedule for reducing exhaust emissions from new cars. Congress mandated EPA to set standards for 1975 and post-1975 model cars "which require a reduction of at least 90 per centum from emissions of carbon monoxide and hydrocarbons allowable . . . in model year 1970"; for 1976 and post-1976 model cars, the standards are to "require a reduction of at least 90 per centum from the average of emissions of oxides of nitrogen actually measured . . . during model year 1971." Section 202(b) (1) and . The Administrator was empowered to suspend these deadlines under extraordinary circumstances, but "for one year only," Section 202(b) (5) and , and only if he simultaneously established "interim standards" which
reflect the greatest degree of emission control which is achievable by application of technology which the Administrator determines is available, giving appropriate consideration to the cost of applying such technology within the period of time available to manufacturers.
Foreseeing that achievement of this schedule might require regulation of fuels, and foreseeing in particular the possible need to regulate gasoline's lead content so as to protect catalytic converters, *fn5 Congress authorized the Agency to promulgate regulations which
control or prohibit the manufacture, introduction into commerce, offering for sale, or sale of any fuel or fuel additive for use in a motor vehicle or motor vehicle engine . . . if emission products of such fuel or fuel additive will impair to a significant degree the performance of any emission control device or system which is in general use, or which the Administrator finds has been developed to a point where in a reasonable time it would be in general use were such regulation to be promulgated. *fn6
Section 211(c) (1). The Administrator must subject his proposed regulations to a "public hearing" before promulgating them in final form. Section 211(c) (2) . No fuel regulation may be undertaken "except after" the Administrator has considered, and published "findings" on, the comparative merits of emission control devices, in or near "general use," which do and which do not require use of regulated fuels. Section 211(c) (2) . The Administrator must also find that prohibition of a particular fuel or additive will not cause use of another fuel or additive which endangers "the public health or welfare to the same or greater degree than the use of the fuel or fuel additive proposed to be prohibited." Section 211(c) (2) . The Act mandates a "civil penalty of $10,000" per day against "any person who violates . . . the regulations prescribed," though the Administrator may "remit or mitigate" the penalty at his discretion. Section 211(d).
EPA published its proposed regulations of gasoline's lead content on February 23, 1972, *fn7 the proposals dealing with both the catalytic converter problem and the direct health impact of air-borne lead. The Agency solicited comments from all interested parties and held public hearings on the proposals in three cities. Petitioners participated fully in these proceedings and do not challenge their regularity.
On January 10, 1973 the final regulations now under review were promulgated. *fn8 In a simultaneous decision not now before us, regulations dealing with the health impact of air-borne lead were reproposed for further public discussion. The Administrator accompanied the final Fuel Regulations with a statement explaining and justifying them (hereinafter the Statement), which we append to this opinion. So far as is pertinent to this case, the Regulations presently provide as follows: *fn9
"Unleaded gasoline" is defined as "gasoline containing not more than 0.05 gram of lead per gallon and not more than 0.005 gram of phosphorous per gallon." 40 C.F.R. § 80.2(g). Cars fitted with catalytic converters must carry markings or design features showing that only unleaded gasoline may be used. 40 C.F.R. § 80.24. On pain of a civil fine of $10,000 per day, 40 C.F.R. § 80.5, retailers, and their employees and agents, are forbidden to introduce leaded gasoline into cars marked or designed for exclusive use of unleaded fuel. 40 C.F.R. § 80.22(a). If the offending retailer displays a refiner's corporate trademark, the refiner is made vicariously liable for the offense, 40 C.F.R. § 80.23(a) (1); if the offending retailer did not display a refiner's trademark, vicarious liability runs to "any distributor who sold the retailer gasoline contained in the retail outlet storage tank which supplied that pump at the time of the violation . . .." 40 C.F.R. § 80.23(a) (2). In both cases, however, there is no vicarious liability, and the retailer alone is liable, where the retailer introduced the leaded gasoline "from a pump from which leaded gasoline is sold." 40 C.F.R. § 80.23(c). Finally, the Regulations affirmatively require marketing of at least one grade of unleaded gasoline by "every person who owns, leases, operates, controls, or supervises a retail outlet at which 200,000 or more gallons of gasoline was sold during any calendar year beginning with the year 1971." 40 C.F.R. § 80.22(b).
During and after final promulgation of the Regulations, certain events transpired which bear upon the issues before us. *fn10 On February 10, 1973, this court remanded for reconsideration EPA's decision, under Section 202(b) (5) of the Clean Air Act, not to suspend for one year the 1975 emission standards established by the Agency pursuant to Section 202(b) (1) of the Act. International Harvester Co. v. Ruckelshaus, 155 U.S. App. D.C. 411, 478 F.2d 615 (1973). On April 11, 1973, after new public hearings, the Administrator decided to suspend the 1975 standards and to replace them with a set of "interim standards." *fn11 Under this "interim" program, special standards apply to all 1975 model cars sold in California; these cars must produce no more than.9 gram per mile of hydrocarbons, 9.0 grams per mile of carbon monoxide, and 2.0 grams per mile of nitrogen oxides. *fn12 The Administrator stated:
These standards in my judgment will require use of catalytic converters on all 1975 passenger cars shipped to California. California sales of such vehicles constitute approximately ten percent of total United States new car sales. *fn13
The standards for 1975 model cars sold outside California are less stringent: 1.5 grams per mile of hydrocarbons, 15 grams per mile of carbon monoxide, and 3.1 grams per mile of nitrogen oxides. *fn14 From these standards the Administrator drew guarded conclusions about the probable use of catalytic converters outside California:
. . . These standards can, in my judgment, be achieved by manufacturers generally on most models without use of catalytic devices. In my judgment these standards will not require use of catalysts on more vehicles sold outside California than manufacturers are capable of producing without the possibility of severe production difficulties.
. . . I believe that the requirement to install catalytic converters on all 1975 automobiles shipped to California and on a portion of 1975 cars sold outside California . . . will maintain the accelerating momentum of technological progress which has so clearly characterized catalyst development for automotive applications during the past two years. . . .
. . . To the extent that additional transportation controls are needed outside California, vehicles designed for California can be purchased in 1975 by fleet operators, such as taxicab companies. . . .
In setting interim standards for the rest of the country, I have not felt constrained to avoid any reliance upon catalysts to enable auto manufacturers to meet the certification requirements. I anticipate that for certain model lines catalysts may be required. The likelihood that a significant number of cars will be distributed across the country equipped with catalysts will supplement the experience derived in California in a beneficial way. *fn15
The Administrator recognized that the new 1975 standards, and the catalyst use they might or might not bring forth, had an important bearing on the Fuel Regulations earlier promulgated. He concluded, however, that the Regulations remained necessary:
Since the interim standards established by this decision will require catalysts on all vehicles sold in California, many of which will undoubtedly travel to other parts of the country, and on a significant number of vehicles sold in the other 49 States, lead-free gasoline must be generally available nationwide by the beginning of the 1975 model year. This will be accomplished by regulations that have already been promulgated. *fn16
Subsequent to this statement, it has become clear that at least General Motors and Ford intend to fit converters on a substantial number of their 1975 cars sold outside California, as well as on all cars sold within California. On May 23, 1973 the president of Ford Motor Company testified before a Senate subcommittee:
We can and will meet the 1975 interim standards, in spite of the serious doubts we expressed only a short time ago. It will be done with some difficulty, however. . . .
. . . We will also have to apply to certain engines nationwide the catalytic system that we had hoped could be used only in California during the 1975-model year.
This may involve 15 percent of production for the other 49 states. . . .
. . . I fear, I had hoped we could confine them to California. The way it turned out, another 15 percent of the nation will get them and there will be Ford cars spread out in Oregon and Maine and Florida. . . . *fn17
On May 30 the president of General Motors testified before the same Senate subcommittee:
. . . The addition of the converters to our advanced control systems provides stronger assurance of compliance and, based upon the limited testing to date, no loss of fuel economy (perhaps even an improvement), and better performance. Only in cars where we see that catalytic converters will not be needed to provide this assurance will we market our 1975 cars without them.
It is . . . probable, based on current planning, that the majority of our 1975 cars will have new carburetion and ignition systems to maintain or improve fuel economy and engine performance; some will need air pumps, and in most cases -- perhaps all -- catalytic converters may also be necessary or desirable. *fn18
Subsequent to this testimony, the Administrator formally reaffirmed his conclusion of April 11 that converters would be in widespread use in 1975, noting that
statements by representatives of the motor vehicle manufacturers have indicated that a substantial percentage of 1975 vehicles manufactured for sale outside of California will be equipped with catalysts.
38 FED. REG. 28301 (October 12, 1972). II. THE ISSUES
Numerous objections to the regulations are raised. *fn19 In summary, petitioners allege: that the Statement fails to include "specific findings" supporting the Administrator's many regulatory decisions; that the 0.05 gram per gallon ceiling on lead in "unleaded gasoline" is not sufficiently justified in the Statement or supported by the record; that the affirmative marketing requirement for unleaded gasoline is not authorized by the Clean Air Act and is, in addition, not sufficiently justified in the Statement or supported by the record; that the imposition of vicarious liability is not authorized by the Clean Air Act or by the Constitution and is, in addition, not sufficiently justified in the Statement or supported by the record; and that EPA was required, and failed, to file an "environmental impact statement" for the Regulations under Section 102(2) of the National Environmental Policy Act of 1969. *fn20
A. "Findings" Under Section 211(c) (a) and of the Clean Air Act
Section 211(c) of the Clean Air Act establishes a framework for "informal rule-making " an enterprise for which the usual scope of judicial review is prescribed by Section 10(e) (2) -of the Administrative Procedure Act ,21 and for which minimum procedural requisites are established by Section 4 of the APA.22 Being silent on the scope of judicial review,23 the Clean Air Act incorporates the APA's mandate that agency "action, findings, and conclusions" be struck down if they are "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." Section 10(e) (2) of the APA. But, like many recent grants of rule-making authority,24 the Clean Air Act itself prescribes certain novel procedural requisites.25 Of these the most important for our purposes are the "findings" requirements in Section 211(c) (2). Section 4(c) of the APA provides merely that "the agency shall incorporate in the rules adopted a concise general statement of their basis and purpose." To this duty Section 211(c) (2) adds a requirement that certain "findings" be published in conjunction with promulgation of fuel regulations:
No fuel or fuel additive may be controlled or prohibited by the Administrator pursuant to clause of paragraph (1) except after consideration of available scientific and economic data, including a cost benefit analysis comparing emission control devices or systems which are or will be in general use and require the proposed control or prohibition with emission control devices or systems which are or will be in general use and do not require the proposed control or prohibition. On request of a manufacturer of motor vehicles, motor vehicle engines, fuels, or fuel additives submitted within 10 days of notice of proposed rulemaking, the Administrator shall hold a public hearing and publish findings with respect to any matter he is required to consider under this subparagraph. Such findings shall be published at the time of promulgation of final regulations.
No fuel or fuel additive may be prohibited by the Administrator under paragraph (1) unless he finds, and publishes such finding, that in his judgment such prohibition will not cause the use of any other fuel or fuel additive which will produce emissions which will endanger the public health or welfare to the same or greater degree than the use of the fuel or fuel additive proposed to be prohbited.
Petitioners construe these provisions as requiring that the Administrator's Statement contain findings which are "specific" in character and which are directed to each of the decisions and determinations embodied in the Fuel Regulations. A study of the language and history of the provisions has persuaded us that they cannot fairly be construed in this fashion.
We consider first the question of specificity. The two provisions speak only of "findings," not of "specific findings." The House version of the legislation required specificity26; the Senate version did not.27 The Conference Committee on the Clean Air Act Amendments of 1970 adopted the Senate's preference for the unqualified noun. The Committee's decision was a deliberated one and was meant to have significance. Advocates of the House version had viewed the specificity requirement as a major limitation on EPA's regulatory power over fuels.28 The Conference Committee's preference was for simplicity and workability. In reporting the Committee's decision to the Senate Senator Muskie stated:
While the conference substitute specifies procedures under section 211 which the Administrator will use in determining whether to prohibit or control fuels or fuel additives, the conference committee wishes to call the Administrator's attention to the relationship between his functions under this section and the emission deadlines stipulated in Section 202. It is not the intent of the Congress to create a cumbersome, time consuming administrative procedure which will delay necessary controls on fuels and fuel additives required to meet these deadlines.
Neither is it the intention of the Congress to lock the Administrator into a rigid economic interpretation of the cost benefit analysis specified in this section in making his determination to prohibit or control fuels or fuel additives.
Rather, the conference committee wishes to call the attention of the Administrator to the broad environmental, esthetic and health considerations underlying the enactment of this legislation which should be kept in mind in making these determinations.29
Senator Muskie was a key legislator in drafting the Clean Air Act Amendments, and his uncontradicted views on the meaning of these provisions are entitled to significant weight. See Portland Cement Assn v. Ruckelshaus, 158 U.S. App. D.C. 308, 486 F.2d 375, 381-382 (1973). We are satisfied that Congress wished the Administrator, and a fortiori the reviewing courts, to read the "findings" requirements of Section 211(c) (2) and in light of that imperative need for administrative flexibility and expedition which we have already recognized as coloring EPA's statutory duty to state the "basis and purpose" of its regulations.30 Findings there must be, but they need not be "specific" in the sense of being detailed or voluminous.
We next inquire into the subject matter of the "findings" requirements in Section 211(c) (2) and . What is it that EPA is required to find? Petitioners offer an expansive answer: The statute requires publication of distinct "findings" with respect to every significant decision and determination made in the course of regulating fuels; every choice and judgment embodied in the regulations needs the support of stated findings. This being a matter of first impression, petitioners' thesis requires a rather lengthy exploration.
When -- as is the usual case -- findings are required in the context of an administrative adjudication, there is no need to ask what the requirement means. An adjudication typically involves a single, ultimate determination, the agency's sole task being to reason from raw evidence to a final conclusion of a mixed legal-factual character, usually framed in statutory language. For instance, an agency will be asked to determine whether the evidence available on a public utility's rate practices shows behavior that is "unreasonable" or "unduly discriminatory." The agency's "findings" are simply those "basic" or "intermediate" conclusions of fact by which it resolves evidentiary disputes and from which it moves, in a final step, to the ultimate statutory decision. See generally 2 K. Davis, Administrative Law TREATISE § 16.06 (1958).
In rule-making, however, an agency's task is not to test raw evidence against a single, pre-established standard; rather the agency is to fashion a host of new legal standards -- regulations -- having prospective effect. A rule-making agency makes not one but dozens of "ultimate" decisions -- not only because a set of regulations has many provisions, but also because adoption of any one provision constitutes a simultaneous rejection of many possible alternatives. Few if any of these "ultimate" decisions will depend on factual conclusions of the sort conventional in adjudication. Looking to the future, and commanded by Congress to make policy, a rule-making agency necessarily deals less with "evidentiary" disputes than with normative conflicts, projections from imperfect data, experiments and simulations, educated predictions, differing assessments of possible risks, and the like. The process is quasi-legislative in character, and one will search it in vain for those intermediate "findings" of fact which mark the midway point in an adjudicator's linear march from raw evidence to single, ultimate conclusion. See generally Industrial Union Department, AFL-CIO v. Hodgson, 162 U.S. App. D.C. 331,-(1974) (slip op. at 8-16, 41); Automotive Parts & Accessories Assn v. Boyd, 132 U.S. App. D.C. 200, 205-207, 407 F.2d 330, 335-337 (1968).
For these reasons we think petitioners' expansive reading of the "findings" requirements in Section 211(c) (2) and is inherently unrealistic. To require findings of a rule-making agency without specifying what factual predicates are to be found or which administrative decisions are to be supported by findings would be to invite endless confusion and great uncertainty, particularly at the stage of judicial review. Judge Friendly recently confronted a similar dilemma in Associated Industries of New York State, Inc. v. U.S. Dept. of Labor, 2 Cir., 487 F.2d 342 (1973), involving a statute, the Occupational Safety and Health Act, which subjected certain rule-making by the Secretary of Labor to judicial review under the "substantial evidence" standard. After observing that the bearing of factual "evidence" on rule-making is unclear at best, id. at 349-350, Judge Friendly warned:
Courts may well end up doing much less than Congress intended or, a more likely and graver threat in these days of judicial activism, much more than Congress had wished.
Fortunately, the findings requirements in Section 211(c) (2) and are capable of an alternative construction which avoids uncertainty and confusion. In our judgment Section 211(c) (2) calls on EPA to answer a particular factual question, while Section 211(c) (2) requires that particular factual determinations be made concerning the threshold necessity of embarking upon fuel regulation. Because these findings requirements are reasonably precise and comprehensible, they are compatible with the mechanics of rule-making and with the exigencies of judicial review.
Consider first Section 211(c) (2) . It provides that, before a fuel "may be prohibited," the Administrator must publish a "finding, that in his judgment such prohibition will not cause the use of any other fuel or fuel additive which will produce emissions which will endanger the public health or welfare to the same or greater degree than the use of the fuel or fuel additive proposed to be prohibited." The Administrator's Statement makes a finding to this effect.31 Petitioners have presented us with no reason to think the finding arbitrary or capricious.32
The "findings" requirement in Section 211(c) (2) calls for somewhat closer attention. It commands the Administrator to "publish findings with respect to any matter he is required to consider under this subparagraph" (i.e. under 211(c) (2) ). The "matter[s]" he must "consider" are
available scientific and economic data, including a cost benefit analysis comparing emission control devices or systems which are or will be in general use and require the proposed control or prohibition with emission control devices or systems which are or will be in general use and do not require the proposed control or prohibition.
Thus in a literal sense the provision requires "findings with respect to" the actual items of data which the Administrator must "consider." But we do not take this to mean that EPA is to produce an aimless commentary on the raw evidence presented to it.33 There is an accepted distinction in the law between findings and a rambling discussion of evidence. See 2 K. DAVIS (supra) § 16.06. Findings are generally required in order to illuminate the empirical bases of an agency's decision-making and this, we assume, is also the objective of the requirement in Section 211(c) (2) .
While the provision does not expressly specify the administrative decision in support of which findings are necessary, this emerges with reasonable clarity from the structure and legislative history of the provision. Section 211(c) (2) is centrally concerned with EPA's threshold determination whether or not to regulate a particular fuel or additive. In effect, the provision establishes a rebuttable presumption that the Agency should maintain a laissez faire posture with regard to fuel regulation.34 To rebut the presumption the Agency must determine, through consideration of "available scientific and economic data, including a cost benefit analysis," that the emission standards established under Section 202 of the Act cannot be achieved in acceptable fashion by relying on emission control devices in "general use" which "do not require the proposed control or prohibition." The "findings" arrived at in making this determination must be published. The findings are to show why regulation, as opposed to no regulations, is necessary or otherwise advisable.35 These findings constitute a condition precedent to embarking upon the exercise of regulatory power. It does not follow that findings are also required to support the many subsequent and detailed decisions made during the exercise of that power. Section 211(c) (2) involves a clear reference to the threshold question of whether or not to regulate, but implies no reference we can perceive to the myriad issues raised by the subsequent question of how to regulate. The findings requirement appears, therefore, to be an initial barrier which the Agency must scale, not a series of hurdles strung throughout the whole course of the rule-making process.36
The legislative history of the provision supports this interpretation. In discussing the Conference Committee's bill, Senator Muskie spoke of "procedures under section 211 which the Administrator will use in determining whether to prohibit or control fuels or fuel additives."37 (Emphasis added.) The Senate committee explained its version of the legislation in similar terms.
The Committee believes that automobile and petroleum industries should be given the greatest possible latitude in developing an effective low emission technology. Therefore, before the Secretary made any decision to prohibit or limit the use of a fuel to facilitate emission control, he would be required to hold public hearings and make and publish a finding that such a prohibition is necessary. . . .38
(Emphasis added.) While adverting to the later-rejected notion of "specific findings," the House committee also focused on the threshold determination whether or not to embark on fuel regulation.
In authorizing the Secretary to prescribe limitations for automotive fuel ingredients, the committee has conditioned the Secretary's authority by requiring specific findings based on specified evidence. The committee has done this for the purpose of assuring that such limitations will not be lightly imposed if other equally satisfactory alternatives are available.40
Therefore, at the core of Section 211(c) (2) we find a requirement that the Administrator state findings, drawn from a study of emission control devices in or near "general use," to the effect that fuel regulation is a necessary or otherwise advisable component in the Agency's overall strategy to meet the Section 202 emission standards. On this score the Statement accompanying the Fuel Regulations is candid and, we think, adequate. We quote relevant portions in the margin.41 The Administrator expressly found that catalytic converters require a regulated fuel -- unleaded gasoline; that converters will be in general use in the 1975 model year; and that no other emission control device or system will then be in ...