We are asked to review1 the Metropolitan Boston Air Quality Transportation Control Plan (the plan).2

The plan is aimed at keepint two types of air-borne pollutants, photochemical oxidants and carbon monoxide, from exceeding within Greater Boston the national primary and secondary ambient air quality standards prescribed by the Environmental Protection Agency (EPA) under authority of the Clean Air Act.3 In the Act, Congress has directed EPA, using latest scientific knowledge, to establish nationwide air-quality standards for each pollutant having an adverse affect upon the public health or welfare. 42 U.S.C. 1857c-4. It has further directed each state to have a plan to 'impolement' those standards-- that is, to see that within the state the level of each such pollutant does not exceed limits prescribed in the national standards.

The present plan (termed a 'transportation' control plan because it focuses upon pollutants caused mainly by vehicles rather than by 'stationary sources' like factories, incinerators, and power plants) has been recognized from the outset to present delicate problems; inevitably it seems bound to come between the citizen and his automobile. Indeed the problems were seen to be so novel and difficult, that the EPA Administrator initially postponed compliance dates from mid-1975 to 1977; however, it was held that he lacked authority to do so. See Natural Resources Defense Council, Inc. v. EPA, 154 U.S.C.App.D.C. 384, 475 F.2d 968 (1973).

The Administrator finally ordered Massachusetts to submit its transportation control plan by April 15, 1973. When Massachusetts did not submit an acceptable plan, the Administrator, as he is obliged to do under such circumstances,4 promptly proposed a plan of his own for the state, held a public hearing and, after making changes in the plan he had first proposed, promulgated regulations embodying the final plan before us.

The plan is designed, by May 31, 1975, to reduce the expected emission of hydrocarbons in the Metropolitan Boston Intrastate Region by 58 percent, and of carbon monoxide in the Boston core and East Boston area of the region by about 40 percent. The Administrator has determined that reductions of this magnitude are necessary if the region's air is to conform to national standards by that date, which is the compliance date set by Congress.

At the heart of the plan is a strategy of cutting down emissions by discouraging the use of vehicles. Off-street and on-street parking spaces are to be 'frozen' or cut back, and the construction of new parking facilities is regulated. There are to be special bus and car pool lanes, and a computer car pool matching system. There is also to be a program of vehicle inspection and maintenance and emission exhaust controls, including oxidizing catalysts, air bleed emission controls and a vacuum spark disconnect. Finally, there are controls on some stationary sources, including gasoline sales outlets, to prevent hydrocarbon emission.

Many aspects of the plan are attacked by affected entities and individuals, although we note that the City of Boston registers its support. The separate petitions for review were consolidated and are herein decided together.

* STANDARD OF REVIEW

In providing for review of an implementation plan under the Clean Air Act by courts of appeals, Congress did not lay down standards beyond those already established in the Administrative Procedure Act (APA). The latter standards, appearing in 5 U.S.C. 706, are controlling. See Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 413, 91 S. Ct. 814, 28 L. Ed. 2d 136 (1970); Texas v. EPA, 499 F.2d 289, at 296 (5th Cir. 1974). Under 706, we must determine whether EPA followed lawful procedures in evolving its plan; whether it acted within its statutory authority; and whether the plan is constitutional. If so, we must set aside the plan only if it is found to be 'arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law'. 5 U.S.C. 706(2) (A).

In the following parts of this opinion we deal first with the procedural objections to the plan and later with the constitutional ones. In between we consider statutory objections and, most difficult of all, those objections addressed to the merits of the plan. The last objections, it is clear, are outside our province unless they show that EPA's decision was not based on consideration of relevant factors or else included a 'clear error of judgment'. Overton Park, supra, 401 U.S. at 416, 91 S. Ct. 814. We are not empowered to substitute our judgment for that of the agency.5

The questions about the plan on review are of two types: the rationality of EPA's technical decisions (such as its determinations of local photochemical oxidant and carbon monoxide levels and the amount of reductions required to meet national standards), and the rationality of EPA's 'control strategy', that is, the measures adopted to reduce emissions. The former present peculiar difficulties for nonexperts to evaluate. Yet '(our) inquiry into the facts is to be searching and careful', id., and we must assure ourselves as best we can that the Agency's technical conclusions no less than others are founded on supportable data and methodology and meet minimal standards of rationality. See Section III infra.

Assuming EPA's technical determinations are reasonably based, we must decide whether the selected controls are arbitrary or capricious. In so doing, we must bear in mind that Congress lodged with EPA, not the courts, the discretion to choose among alternative strategies.6 Unless demonstrably capricious-- such as much less costly but equally effective alternatives were rejected or the requisite technology is unavailable-- the Administrator's choices may not be overturned. See Friends of the Earth v. EPA, 499 F.2d 1118, at 1123 (2d Cir. 1974); Delaware Citizens for Clean Air, Inc. v. EPA, 480 F.2d 972, 975-976 (3d Cir. 1973). Of course neither EPA nor this court has any right to decide that it is better to maintain pollutants at a level hazardous to health than to require the degree of public sacrifice needed to reduce them to tolerable limits.

We now turn to the objections.

Several petitioners charge that the final plan differed so radically from the one proposed in the Administrator's published notice that they had no meaningful forewarning of its substance.

Notice of a proposed rulemaking must be published in the Federal Register. It must include 'either the terms or substance of the proposed rule or a description of the subjects and issues involved.' 5 U.S.C. 553(b)(3). The Administrator published notice on July 2, 1973. 38 Fed.Reg. 17689 (1973). Two days of public hearings were then held in Boston on July 19 and 20, 1973,7 after which the record remained open for submission and public inspection of written comments until August 1, 1973. The final plan was published November 8, 1973.

In his notice, the Administrator asserted that levels of photochemical oxidants and carbon monoxide in parts of the Boston region already exceeded on many days the national primary standards promulgated as necessary to public health, and concluded that,

'it will be necessary to reduce projected total emissions of hydrocarbons within the area encompassed by Route 128 by approximately 28 percent and to reduce projected emissions of carbon monoxide in the Boston core and East Boston areas of the Region by amounts approximating 40 percent in each area . . ..' 38 Fed.Reg. 17691 (1973).

The Administrator wrote that carbon monoxide emissions came almost entirely from motor vehicle sources and that most hydrocarbons which make up photochemical oxidants came from the same source. Analyzing transportation control alternatives-- reducing vehicular emissions or reducing vehicle miles of travel (VMT)-- the Administrator outlined a policy also supported by the Governor of Massachusetts 'to discourage continued heavy reliance on the automobile'. Thus the proposed, as well as final, regulations stress VMT reducing controls.

The Administrator's specific proposals first included a ban on on-street parking in the Boston core area from 6 to 10 a.m. and 4 to 6 p.m. on weekdays, and a $5 surcharge on off-street parking from 6 to 10 a.m. in the core area and from 6 a.m. to 10 p.m. at Logan International Airport (Logan). To reduce photochemical oxidants he proposed prohibiting travel within Route 128 (an expressway circling Boston) one day out of five, by a sticker system. However, the Administrator noted that there was a legal question concerning EPA's authority to propose regulatory fees, and also said that the sticker system might be replaced if preferable measures were suggested. He warned that alternatives were being considered including:

'Making effective reductions in the number of parking spaces available for use in downtown Boston between 6 a.m. and 10 a.m., and at Logan airport between 6 a.m. and 10 p.m. 'Making a similar reduction in the number of available parking spaces (both on- and off-street) in parking facilities at other trip-attraction center throughout the Metropolitan area.' 38 Fed.Reg. 17692 (1973).

Interested persons were notified that a technical support document was available. Finally, the Administrator said the final plan should reflect local needs and, therefore,

'particularly invited (comments) pertaining to the other measures that may be taken by Federal, State, or local authorities to support or supplement the proposed air pollution control measures . . .. The Administrator's final promulgation of transportation controls . . . will be influenced by the comments and testimony he receives, as well as by any further approvable strategies submitted by the State as part of the State Implementation Plan. These influences, and the additional analysis of alternative strategies that can be made in the time between this proposal and final promulgation, may lead the Administrator to adopt final regulations that differ in important ways from this proposal.' Id. at 17694.

When the final plan emerged, it was, indeed, much influenced by the public hearing. The $5 surcharge and the sticker system, both of which had been sharply criticized, were dropped, as was a proposal to limit the supply of gasoline. Substituted were a much smaller surcharge and an egress toll on vehicles leaving Logan; pre-construction review of new parking facilities throughout the region;8 a freeze on parking spaces in selected portions of the metropolitan region;9 reduction in employee parking spaces;10 and reductions in the number of off-street parking spaces available in the Boston core during morning rush hours.11 The parking surcharge and egress toll were subsequently deleted. 39 Fed.Reg. 1848 (1974).

The final plan added regulations for computer car pool matching and preferential treatment throughout the region for bus/car pools. It made changes in the controls on mobile source emissions, and required vehicles to be inspected twice a year instead of once.

Although the changes were substantial, they were in character with the original scheme and were additionally foreshadowed in proposals and comments advanced during the rulemaking. Parties had been warned that strategies might be modified in light of their suggestions.12 When the hearing opened, the Regional Administrator said,

'if you suggest any part of the plan be eliminated, then we are going to have to make that emission reduction by some other kind of control, and we are anxious to receive your recommendation as to what that could be.'

Governor Sargent, speaking first, advocated freezing the number of parking spaces, mentioning Logan Airport. But the Governor would have let the new South Terminal parking garage at Logan and other new projects go forward if the Massachusetts Port Authority (Massport) and Boston phased out 'older, less efficient and less conveniently situated spaces in equivalent number to the new spaces being created.' He also recommended a 10 to 20 percent parking vacancy rate before 10 a.m. and a 10 to 20 percent reduction in the ratio of parking spaces to employees.

One state representative suggested banning further construction of parking garages and reducing the surcharge to $2. The Chairman of the City of Boston Air Pollution Control Commission advocated both a parking freeze on commuter spaces with new parking only if an equal number of other spaces are eliminated, and a rollback of commuter parking before 10 a.m. The Regional Administrator expressed interest in his proposals. The Commission's Executive Director also suggested a parking freeze in the downtown area and at related locations. The Boston Redevelopment Authority proposed replacing the surcharge with an early morning vacancy rate. The possibility of pegging this figure at 40 percent, as later decided, was discussed. One of the petitioners, Mr. Meyers, President of the New England Parking Association, was asked by the Regional Director which alternative-- $5 surcharge or 'restricting parking spaces utilized during the day, say 6 a.m. to 10 a.m.'-- was considered most equitable from the standpoint of the parking facilities industry.13 Throughout the hearing, that choice was repeatedly explored. Other persons suggested eliminating all open-air parking lots, a 'head on assault on the availability of parking spaces, particularly in open-air parking lots and older garages', and reducing parking facilities where plants can be served by mass transit. One City Councillor suggested that no parking facility be allowed to park cars between 6 and 10 a.m.

We conclude that interested persons were sufficiently alerted to likely alternatives to have known what was at stake. In the published notice, mention had been made of alternative measures for 'effective reductions' in the number of parking spaces in downtown Boston, Logan Airport and 'other trip attraction centers throughout the metropolitan area'. EPA made plain its intention to adopt 'stringent controls'. South Terminal is unconvincing when it argues that the 'effective' reduction at Logan (from 6 a.m. to 10 p.m.) applied only to 'limited periods of the day' and did not, therefore, give notice that anything as drastic as a freeze might be adopted. The proposed surcharge was expected to reduce South Terminal's parking by 35 percent. The final plan was in some ways more moderate than the original. Cf. California Citizens Band Ass'n v. United States, 375 F.2d 43 (9th Cir. 1967).

A hearing is intended to educate an agency to approaches different from its own; in shaping the final rule it may and should draw on the comments tendered. The plan seems a logical outgrowth of the hearing and related procedures. Cf. Mt. Mansfield Television, Inc. v. FCC, 442 F.2d 470 (2d Cir. 1971). Parties have no right to insist that a rule remain frozen in its vestigal form. See Pacific Coast European Conference v. United States, 350 F.2d 197, 205 (9th Cir. 1965). As the Court of Appeals for the District of Columbia Circuit recently said:

'The requirement of submission of a proposed rule for comment does not automatically generate a new opportunity for comment merely because the rule promulgated by the agency differs from the rule it proposed, partly at least in response to submissions.⁵¹

'51. A contrary rule would lead to the absurdity that in rule-making under the APA the agency can learn from the comments on its proposals only at the peril of starting a new procedural round of commentary.'

International Harvester Co. v. Ruckelshaus, 155 U.S.App.D.C. 411, 478 F.2d 615, 632 n. 51 (1973). See also Owensboro on the Air, Inc. v. United States, 104 U.S.App.D.C. 391, 262 F.2d 702, 708 (1958).

Cases cited by petitioners in which there was no notice or opportunity to comment or submit evidence are not in point,14 nor is their reliance on Wagner Electric Corp. v. Volpe, 466 F.2d 1013 (3d Cir. 1972). A circumscribed announcement, as in the latter case, that standards for testing an automotive product would be revised, is not to be compared to EPA's comprehensive notice. The former did not make it clear that interrelated aspects, such as performance criteria, would also be considered. The instant notice left no doubt that EPA would consider all reasonable alternatives for cutting down vehicle use.

Some petitioners also claim inadequate notice of technical documents upon which EPA relied. Yet save for South Terminal and Massport, they do not point to specific information not available before the hearing which they now dispute. EPA stated in its published notice that a technical support document was available. That referred to previous studies which, had they been sought out, would have been found to include consultants' technical reports. There is no evidence that access to any of this material was ever requested or, having been requested, was denied by EPA. Portland Cement Ass'n v. Ruckelshaus, 158 U.S.App.D.C. 308, 486 F.2d 375, 392 (1973), cert. denied, 417 U.S. 921, 94 S. Ct. 2628, 41 L. Ed. 2d 226 (1974), upon which petitioner relies, deals with an entirely different set of facts-- critical test results were not available, and the Agency refused to respond to legitimately voiced problems with its methodology. The present technical support document explained in some depth the basis for the Agency's conclusion that emission reductions were necessary. Post-hearing documents furnished by EPA were in response to issues there raised by the public and, except for new information on Logan, supply no data petitioners would have needed to raise the questions they now pose.

We thus reject petitioners' contentions both of inadequate notice and of inadequate disclosure of underlying technical data.

A further procedural contention is that 'adjudicatory' rather than 'informal rulemaking' hearings should have been held. The APA calls for an adjudicatory hearing only when the statute requires rules to be 'made on the record' after an agency hearing. 5 U.S.C. 553(c). Statutes which merely authorize rulemaking 'after hearing' do not trigger all the procedures of 556 and 557. United States v. Florida East Coast R.R., 410 U.S. 224, 93 S. Ct. 810, 35 L. Ed. 2d 223 (1973); United States v. Allegheny-Ludlum Steel Corp.,406 U.S. 742, 92 S. Ct. 1941, 32 L. Ed. 2d 453 (1972). A 'public hearing ' is to be held before a state implementation plan is promulgated by the Administrator. 42 U.S.C. 1857c-5(c). But there is no indication that Congress intended the words 'public hearing' to be tantamount to 'on the record'.15

Petitioners concede that the language in the Clean Air Act does not trigger the APA's adjudicatory guidelines. They contend, however, that the due process clause requires a full evidentiary hearing because the plan affects 'vital, clearly identifiable economic interests' and is 'condemnatory in purpose'. At such a hearing, petitioners say, they would present evidence refuting the plan's technological and economic feasibility, as well as the underlying scientific data.

The opportunity for oral as well as written comment was granted here. Cf. Walter Holm & Co. v. Hardin, 145 U.S.App.D.C 347, 449 F.2d 1009 (1971). Petitioners did not at the hearing demand any more expansive procedures. Moreover, economic and technical factors pertaining to the creation of a regional plan are not usually 'adjudicative' facts. See, e.g., O'Donnell v. Shaffer, 491 F.2d 59 (D.C. Cir. 1972); American Airlines, Inc. v. CAB, 123 U.S.App.D.C. 310, 359 F.2d 624, 633 (D.C. Cir.), cert. denied, 385 U.S. 843, 87 S. Ct. 73, 17 L. Ed. 2d 75 (1966). See generally 1 K. C. Davis, Administrative Law Treatise P7.02 (1958). The end product was to be a rule of general application; like legislation (which may also have severe impact) such a rule need not be conceived in a trial-type proceeding. Courts have unanimously rejected the argument that EPA must provide full adjudicatory hearings simply because cost and technological feasibility are disputed. See International Harvester Co. v. Ruckelshaus, supra, 478 F.2d at 629; Anaconda Co. v. Ruckelshaus, 482 F.2d 1301, 1306-1307 (10th Cir. 1973); Buckeye Power, Inc. v. EPA, 481 F.2d 162, 172 (6th Cir. 1973); Duquesne Light Co. v. EPA, 481 F.2d 1, 6-7 (3d Cir. 1973).

Any relemaking should, of course, be tailored to elicit maximum information and input consistent with the agency's aims and timetable. Cf. International Harvester Co., supra, 478 F.2d at 630-631. But petitioners are poorly situated to complain now of procedural details. They did not do so before or during the hearing, and few even availed themselves fully of the opportunities for comment that were tendered. They made no request to call or examine witnesses. Had they done so, we would not have required EPA to recognize such requests as of right; but the Agency itself in its own discretion might have thought it sensible to afford the privilege upon a 'circumscribed and justified request.' Id. at 631.

South Terminal, being the only facility in construction which may increase parking spaces in a freeze zone by more than 10 percent, is uniquely affected by part of the plan. But there is no evidence that the decision to treat Logan separately from Boston was anything but a genuine 'legislative' judgment by the Agency. In no sense (is the regulation) a punishment for sins of commission or omission'16 by South Terminal; it bears all the indicia of a good faith attempt to deal with what the Agency considered a serious pollution problem over East Boston, of which the airport is a key part. A similar claim was rejected in Anaconda Co., supra, 482 F.2d at 1306-1307, in which part of a state implementation plan written in general terms resulted in a duty upon only one plant-- the area's major source of pollution. Because both the rule and the pollutants affected many persons beyond the solitary polluter, the court concluded that an adjudicatory hearing was not required nor even manageable in view of the many possible interested parties.

The present rule does not deal with named parties and had, as its focus, an appropriate general, regulatory purpose.17 We agree that,

'(The) Constitution should not be held to require participatory rights for persons affected by agency rulemaking but not named as parties, unless the agency adopted a rule without a general purpose and with the intent of affecting only a specific person.⁴⁰

'40. One way to demonstrate this intent would be to establish that there was no proper reason for the agency to have employed rulemaking-- that no policy question was involved and no general deterrent effect was sought to be achieved.'

Note, the Judicial Role in Defining Procedural Requirements for Agency Rulemaking, 87 Harv.L.Rev. 782, 788, n. 40 (1974). We hold that rulemaking rather than adjudication proceedings were in order.

Since oral argument the Energy Supply and Environmental Coordination Act of 197418 has become law, providing, 'No action taken under the Clean Air Act shall be deemed a major Federal action significantly affecting the quality of the human environment within the meaning of the National Environmental Policy Act of 1969.' Pub.L.No. 93-319, 7(c)(1) (June 22, 1974). This nullifies petitioners' claim that EPA was required to prepare a NEPA environmental impact statement pursuant to 42 U.S.C. 4332(2)(C).

WHETHER EPA COMMITTED A CLEAR ERROR OF JUDGMENT IN COMPUTING THE NEED FOR EMISSION REDUCTIONS

Petitioners challenge both EPA's conclusion that emission reductions are necessary to meet national air quality standards and the magnitude of the reduction said to be required.19 Their arguments can be divided into attacks on EPA's data and methodology as to (1) photochemical oxidants in the Metropolitan Boston Intrastate Region; (2) carbon monoxide in the Boston core; (3) carbon monoxide at Logan Airport (East Boston).

1. EPA is said to have overestimated the photochemical exidant problem in the Boston region. Most pertinent are petitioners' arguments that the key ambient air quality reading taken on one day at a monitoring device located at Wellington Circle must have come from a defective instrument. This single reading, inserted by EPA in its so-called rollback formula (or 'model'), was the basis for a region-wide estimate of the amount of hydrocarbon reduction required. If it was incorrect, so were the conclusions about how much reduction was necessary to achieve the primary standard. Petitioners point to a computer printout taken at that monitoring station; it contains a high number of '9999' readings which may indicate instrument malfunction. EPA's response is that the designations may also result from 'instrument calibration, instrument zeroing, transmissions loss and depletion of span gas, all of which causes are unrelated to any malfunction.' But petitioners contend that the irregular readings occurred too often to be attributable solely to innocent causes. On the present record, we cannot say with confidence that the use of a single reading from a machine as to which objective readings suggest a substantial possibility of malfunction is sufficient to support EPA's photochemical oxidant determination.

We find less persuasive petitioners' attack upon the accuracy of the rollback model itself because of its purported failure to take account of local topography and meteorology. EPA's technical support document appears to consider these influences, and the only expert to stress Boston's unique features did not include gasoline in his analysis. Petitioners further claim that EPA incorrectly related oxidant concentrations directly to emission of hydrocarbons, relying in part on an extra record document never brought to the Agency's attention. Photochemical oxidants are a secondary pollutant derived from the reaction of two primary pollutants, hydrocarbons and nitrogen. To reduce oxidant concentrations, it is therefore necessary to control hydrocarbon emission and EPA has advanced plausible reasons for choosing the ratio that it did. See Texas v. EPA, supra, 499 F.2d at 306-308. Finally,20 petitioners object to the determination that regionwide controls, rather than controls in only a few heavily polluted sections, were necessary to bring oxidants down to a reasonable level. But background reports indicate that automobile use is heavy, particularly in the outlying manufacturing areas. The technical support document presents the view that the necessity for regionwide controls stems from the nature of the pollutant; petitioners' contention that contrary conclusions can be drawn from the data does not lead us to suspect that EPA committed clear error. To the extent different conclusions could be drawn the Agency was entitled to draw its own.

2. Carbon monoxide data is attacked as unreliable. EPA determined that its national primary standard requiring the average amount of carbon monoxide in the air over an eight hour period not to exceed 9 p.p.m. is not being met in the Boston core and will not be met by mid-1975. It did this by a series of calculations which have as their essential element an ambient air quality reading obtained on one day in 1970 from a monitor at Kenmore Square. Although petitioners attack use of the rollback model itself as unsophisticated, we are mainly impressed by the contention that the crucial figure for determining required emission reduction may be unrepresentative. At the time the plan was designed the next highest reading at Kenmore Square was nearly 50 percent lower than that utilized. EPA points to readings elsewhere even higher than that used in the rollback model, recorded after the plan was announced, as evidence that it may have 'underestimated the extent of the CO problem'. But petitioners claim these high readings are also freak events.

Petitioners also seek to discredit the Kenmore Square reading by pointing to EPA guidelines from which it may be inferred that the Kenmore monitor was located too close to the street curb. The guidelines do state that 'for practical considerations it may not be feasible to select sampling sites that meet all of the specific and general guidelines', and EPA asserts that use of the 'Kenmore data is completely appropriate for Boston because in an older city like Boston, the streets are narrow, and many more people live and work close to heavily traveled roadways where they are exposed to high one-hour and eight-hour CO concentrations than in the case of newer cities designed with more open spaces like Phoenix, Houston, and even Washington, D.C.' But the guidelines warn that when there are deviations it is important that there also be comparisons with results from other stations; in 1970, the year the critical reading was recorded, Kenmore Square was the only operating monitoring site. Here again, on the present record, we have no basis to say with judicial conviction that such a slender base, without further justification, is sufficient to support EPA's conclusion as to carbon monoxide in the Boston core.

3. In the best documented of the challenges to EPA technical data, South Terminal and Massport attack the carbon monoxide determinations at Logan Airport (East Boston). The preamble to the plan refers generally to the need to reduce pollution in East Boston, but the strictures fall on Logan Airport, the only part of East Boston to be controlled. The Administrator concluded that the 'carbon monoxide problem is concentrated in two relatively small areas-- Logan Airport and the Boston core area.' 38 Fed.Reg. 20961 (1973). He also stated that the 'overwhelming majority of the vehicle miles of travel in East Boston is generated by a single source: Logan International Airport. Consequently, transportation control strategies aimed at reducing VMT generated by Logan Airport are necessary for attaining the carbon monoxide standard in that portion of Boston.' Id. Massport, the operator of Logan, complained after the official period for comment had expired, but before the plan was published in the Federal Register, that no air quality monitoring had been performed at Logan to substantiate that the 8 hour carbon monoxide standard was exceeded there. South Terminal had earlier complained that what it considered more stringent controls on Logan than elsewhere were not justified by any technical data. In the preamble to the final plan, responding to the public comments it had received, EPA answered Massport's criticism by referring to a report21 which had concluded that 8 percent of the carbon monoxide readings on-airport exceeded EPA's primary standard and 6 percent of the off-airport readings exceeded standards. 38 Fed.Reg. 30963 (1973). However, this study used 'grab sample' spot readings, not EPA's suggested method for the determination and sampling of carbon monoxide, 40 C.F.R. 50.8, Appendix C.

More significantly, Massport has had the record supplemented with the results of its own testing at a Logan site which EPA has conceded 'meets the criteria for the eight hour standard'. These data show that carbon monoxide levels at Logan are strikingly lower than at other Boston sites and that the federal primary eight hour air quality standard has never been exceeded. To counter the submission of this data, EPA points to a report in the record that a maximum 8 hour concentration of 19 p.p.m. was recorded at Logan and that in a six day period the primary air quality standard was exceeded on five days. The report also concluded that the concentrations of carbon monoxide at Logan were roughly equivalent to those measured elsewhere in the metropolitan Boston region. However, the author of the study expressed doubts whether it was accurately reporting on 'ambient air' as that term is defined in 40 C.F.R. 50.1(e). Moreover, it was published after the plan was announced and interested parties have not had an opportunity to criticize the findings. We conclude that it is not yet clear whether or not the ambient air at Logan meets, or will without controls by mid-1975 meet, the national primary standard.

Logan might have lower pollutant levels than those recorded outside its boundaries in East Boston, but nonetheless be responsible for excess levels of carbon monoxide in East Boston. If so, it would not be arbitrary for EPA to force reductions in vehicle miles traveled to the airport. But the record does not sufficiently demonstrate the degree to which East Boston's air is affected by Logan traffic. The same Kenmore Square air quality figure, inserted in the rollback formula, was used to project the required reduction in East Boston; there was no actual monitoring anywhere in East Boston. South Terminal claims that Kenmore data should not be the basis for air quality predictions near Logan given the jetport's unique configuration and air patterns as contrasted with those of the Boston core. EPA record materials suggest that the rollback formula is most reliable when applied in areas having similar characteristics. Moreover, the traffic generated by Logan utilizes main arteries, not local East Boston Streets. A report in the record indicates that 80 percent of the traffic enters and leaves the airport via Route US-1 North and South. The record does not indicate whether people work or live sufficiently close to the tunnel connecting Boston and East Boston or to Route 1 to justify controlling this traffic. The primary air quality standard applies only to ambient air--'that portion of the atmosphere, external to buildings, to which the general public has access'. 40 C.F.R. 50.1(e). The special qualities of carbon monoxide, as compared to photo-chemical oxidants, make the absence of such data critical to any argument that Logan traffic should be controlled to protect East Boston. The EPA engineer states that 'carbon monoxide is a highly localized pollutant which tends to stay fairly close to where it was emitted,' and a report in the record indicates that 'due to the localized nature of the problem, carbon monoxide concentrations need to be controlled within areas of blocks, not entire regions'.

The method of sampling at Logan, Massport's own testing, and the lack of monitoring in East Boston, collectively, on the present record, prevent us from holding that the data is sufficient to support EPA's conclusion as to carbon monoxide in East Boston.

4. While for reasons stated in items (1), (2), and (3), supra, we are unable at this time to uphold EPA's conclusions as to photochemical oxidant and carbon monoxide levels and reductions, we do not say that they are necessarily incorrect. Petitioners forcefully contend that the Agency's measurements are without reliable foundation, and hence, in effect, arbitrary and capricious. See Section I supra. But as laymen we are in no position to know how much ultimate weight to give to these arguments, based as they are on technical assumptions. We can only say that the objections as to data and methodology seem too serious to us simply to pass by; they demand investigation and answer. While reviewing courts are not to substitute their judgment for an agency's, they are to establish parameters of rationality within which the agency must operate. A court would abdicate its function were it, when confronted with important and seemingly plausible objections going to the heart of a key technical determination, to presume that the agency could never behave irrationally. It has a duty to see that the objections are faced in a proper procedural setting and satisfactory answers provided demonstrating careful agency consideration.

The normal way courts evaluate a technical issue is through proceedings attended by expert witnesses. Yet as an appellate court, we cannot conduct such fact-finding proceedings on our own. Congress has not interposed a district court in the chain of review, so we cannot remand for clarifying findings of fact based, perhaps, on testimony by Agency and private experts. Cf. Camp v. Pitts, 411 U.S. 138, 142-143, 93 S. Ct. 1241, 36 L. Ed. 2d 106 (1972). We are left-- when confronted with objections which, on the one hand, seriously call into question the rationality of Agency action, but on the other require expertise to resolve-- no alternative but to remand the matter to the Agency itself. In many circumstances such a limited remand could be for a unilatreal 'explanation' or 'clarification' of the Agency's views. See, e.g., Natural Resources Defense Counsel, Inc. v. EPA, 478 F.2d 875, 881-882 (1st Cir. 1973); Kennecott Copper Corp. v. EPA, 149 U.S.App.D.C. 231, 462 F.2d 846, 848-851 (1972). See also Overton Park, supra, 401 U.S. at 420-421, 91 S. Ct. 814, 28 L. Ed. 2d 136. If the adequacy of the Agency's measurement had been raised at the hearing, then such a remand-for-explanation might be sufficient. But when the question is one which the Agency may never have fully confronted and which may deserve further input both from Agency and outside sources, only a remand for further hearings and an extended record seems adequate. Cf. Portland Cement Ass'n v. Ruckelshaus, 158 U.S.App.D.C. 308, 486 F.2d 375, 393-394 (1973), cert. denied, 417 U.S. 921, 94 S. Ct. 2628, 41 L. Ed. 2d 226 (1974). EPA should be in a position to hear the same arguments presented to us in a setting conducive to the exchange of ideas and data. If, after resolution of the issue by the Agency, the issue is returned to us, we will be able to rule expeditiously on the rationality of the Agency's measurements with assurance that EPA has fully confronted the objections and that its explication is more than a post hoc rationalization.

It is perhaps paradoxical that the necessity of a further hearing arises from the fault not of EPA but of petitioners, who (except for South Terminal and Logan) were remiss in not presenting the questions now troubling us at the original hearing. Only the novelty and complexity of the issues provide petitioners with some excuse. Were the proceeding of less importance to the public at large we might decline to entertain the objections at this late stage. However, given the importance of this matter, we do not feel justified in treating the questions as if they were never asked. Answering them is a prerequisite to the credibility of the many public sacrifices required by the plan, and has a significance extending far beyond the petitioners' private interests.

Accordingly, in the interest of a just result, and in aid of the judicial function to evaluate whether the Agency's determinations are in accordance with the law, see 5 U.S.C. 706(2)(A), we think it is necessary to remand to the Agency for further proceedings with respect to these questions. See generally Ford Motor Co. v. NLRB, 305 U.S. 364, 373, 59 S. Ct. 301, 83 L. Ed. 221 (1939); Fleming v. FCC, 96 U.S.App.D.C. 223, 225 F.2d 523 (1955).

It may well be that EPA, after full exposure to petitioners' objections, will effectively demonstrate the adequacy of existing readings and the rationality of its conclusions. Better still, by now EPA may have additional measurements and data on oxidant and carbon monoxide levels.22 Thus, once issue is fully joined, EPA may be able to broaden the data base and corroborate or update the readings and its ultimate conclusions. We can only say presently that the record as now constituted, made without specific consideration of the mentioned objections, leaves us in doubt as to whether or not there is a rational basis for EPA's estimate of the reductions said to be necessary to attain and maintain primary standards. We add that EPA should do more than merely provide satisfactory data to justify the plan; there must be built-in means to assure that the data will be updated, revised, and improved. Actual emissions and reductions may turn out to be more or less than anticipated. The plan should contain provision for adjustment in light of changing data.

We direct that a supplemental public hearing, consistent with the provisions of 5 U.S.C. 553 for informal rule-making, be held by the Administrator no later than 90 days from the date hereof for the following purposes:

(a) To receive and consider further objections and arguments, oral and written, respecting the technical basis for calculating the amount of hydrocarbon and carbon monoxide reduction required to attain and maintain primary photochemical oxidant and carbon monoxide standards within the Metropolitan Boston Intrastate Region;23

(b) To evaluate, affirm, or modify, as the Administrator thinks appropriate, the emission reductions stated in the plan and, in light of any modifications thereof, to modify as he deems appropriate any other portion of the plan;

(c) To adopt as part of the plan a regulation providing a procedure assuring the periodic updating of goals and verification of progress in light of current data.

The Administrator's final order and responses concerning the above matters shall be published within thirty days of the date of such hearing, and a copy thereof filed with this court.

This court will retain review jurisdiction pending the results of the hearing and will rule upon such remaining matters as necessary respecting the plan after the administrator has filed said order. Petitioners may file with this court their objections to the Administrator's order within ten days of the publication thereof. Until the emission reduction question is finally settled, those currently effective controls which petitioners attack shall be stayed, since the controls are predicated on the premise that emissions must be reduced to meet air quality standards. We thus are ordering interim stay of the actual control procedures contained in 52.1135 (freeze; employee parking); 52.1136 (off-street parking regulations); and 52.1144 (evaporative emissions for retail gasoline outlets), provided we do not stay, except as the Administrator may otherwise in his own discretion in the interest of fairness determine, the requirements in clauses 52.1135(i) and (j) and 52.1136(g) for timely submission of action plans and compliance schedules. It is our intention that planning requirements, as opposed to the actual controls, remain in effect. While we are as yet unable to approve EPA's quantitative evaluation of need, we must balance the public's right to receive the benefits of healthful air by the specified target date against the rights of those subject to the controls. As hereinafter appears, we find most of the control measures to be lawful assuming EPA can demonstrate the adequacy of its measurements. Should the latter be validated in the supplementary proceedings we have ordered, most of the specific controls may take effect. Present planning need not and should not therefore be suspended except to the extent that EPA itself may be persuaded that fairness and practicality requires such suspension.

In this regard we should make it clear that while the freeze on parking facility construction is at least temporarily suspended, those who build will do so at their own risk. Should the freeze become once more effective, it would not be inconceivable that we would approve additional regulations altering and adjusting the freeze baseline and the percentage by which facilities in construction may not exceed the baseline so as to achieve the same net parking supply as would have occurred had the freeze not been suspended at this time.

WHETHER THERE IS STATUTORY AUTHORITY FOR THE CONTROLS

Our determination that a remand is in order on the threshold technical issue of the need for and amount of emission reductions does not negate our ability to review other questions raised, and we think it in the interest of all concerned that they be addressed without delay. We thus turn to arguments that EPA lacks specific legal authority to act as it has.

1. One petitioner contends that EPA's promulgation of combined 'national primary and secondary ambient air quality standards' for carbon monoxide and photochemical oxidants, 40 C.F.R. 50.8, 50.9, renders inoperative the more urgent timetable designed for primary standards alone.

Primary standards are those 'requisite to protect the public health'; state implementation plans must provide for their attainment as 'expeditiously as practicable', but, subject to a possible two year extension, in no case later than three years from the date of the plan's approval. 42 U.S.C. 1857c-4(b) (1), 1857c-5(a)(2)(A)(i). Secondary standards, on the other hand, are those 'requisite to protect the public welfare from any known or anticipated adverse effects'. 1857c-4(b)(2). State implementation plans must merely specify a 'reasonable time at which such secondary standard will be attained'. 1857c-5(a)(2)(A)(ii).

In promulgating combined standards, the Administrator separately applied the different criteria. The national primary standard for carbon monoxide was based on evidence that low levels of carboxy-hemoglobin in human blood may be associated with 'impairment of ability to discriminate time intervals'. The level was set so as to protect public health. The national primary standard for photochemical oxidants was based on evidence of increased asthma attacks on days when concentrations of photochemical oxidants exceeded a particular level. The Administrator then concluded, with reference to both primary standards, that they would also protect against known or anticipated adverse effects on public welfare. 31 Fed.Reg. 8186 (1971).

Because of the different criteria, secondary standards may often, of course, differ from primary ones. But it is perfectly conceivable that the Administrator will know of no adverse effects to the public welfare except the public health effect; the two standards will then coincide. In such instances, the primary standard timetable is obviously controlling, and we find no error in the Administrator's understanding of the law.

Insofar as petitioner wishes to object that the primary standard goes beyond public health considerations, it is in the wrong forum. Review of a national primary or secondary ambient air quality standard is centralized in the United States Court of Appeals for the District of Columbia Circuit. 42 U.S.C. 1857h-5(b)(1).

2. Several petitioners allege that EPA is utterly without statutory authority to regulate off-street parking. The core of this argument is that the Agency was given only the authority to regulate 'stationary sources' of pollution, and that expressio unius exclusio alterius reasoning should lead to the conclusion that 'indirect sources' of pollution, such as parking lots, are outside the statutory scope.

The argument is without merit. Although the statute expressly allows EPA to regulate stationary sources,24 there are other provisions conferring the powers in question. Under 42 U.S.C. 1857c-5(a)(2)(B), Congress provided that state implementation plans shall include such measures as may be necessary to insure attainment and maintenance of the national primary ambient air quality standards, 'including, but not limited to, land-use and transportation controls'. And the Administrator must promulgate promptly regulations setting forth 'an implementation plan for a State' should the state itself fail to propose a satisfactory one. Section 1857c-5(c). The statutory scheme would be unworkable were it read as giving to EPA, when promulgating an implementation plan for a state, less than those necessary measures allowed by Congress to a state to accomplish federal clean air goals. We do not adopt any such crippling interpretation.

The terms 'land-use' and 'transportation controls' are not defined in the Act, but they seem reasonably intelligible in the context. 'Land-use' includes the types of regulations normally subsumed under zoning and planning devices. Regulation of the dimensions and quantity of certain types of facilities in a specific urban area is a common use of the zoning tool; and the challenged parking controls are similar, in that they are directives to land owners not to use their land for parking except under specified circumstances. The regulation of parking is also supportable as a transportation control. The plentiful existence of parking facilities creates an incentive to choose motor vehicles; by destroying this incentive, EPA weights the choice more heavily in favor of less polluting transit.

Assuming the Administrator is right that photochemical oxidant and carbon monoxide levels must be reduced, and given his express authority to invoke land-use and transportation controls, the measures are well within his authority.

3. Some petitioners assert that 52.1135(d)25 and (e)26 of the transportation control plan are unauthorized by statute because they require approval, in advance of construction, of proposed additions to parking. Subsection (e) establishes an approval program to enforce the 'freeze' and would seem to require simply a numerical calculation whether an additional lot will result in an increase not compensated for by retirements of spaces elsewhere. Subsection (d), on the other hand, requires all applicants anywhere in the Region to argue to the Administrator or his delegate that construction of new spaces will not 'interfere with' attainment or maintenance of relevant air quality standards. The regulation does not state how 'interference' is to be measured, nor does it suggest guidelines to either the applicant or the Agency as to how the regulation's mandate may be carried out.

We are inclined to construe Congress' broad grant of power to the EPA as including all enforcement devices reasonably necessary to the achievement and maintenance of the goals established by the legislation. We have held above that EPA has authority to regulate parking facilities to this end. As a device to enforce parking controls otherwise necessary, the permit programs established by these regulations would appear to be within EPA's authority.

However, petitioners argue that Congress has rejected a similar proposal for a permit program of stationary sources and that we should infer that Congress intended the same limitation of EPA power over indirect sources. We do not read the legislative history in the way petitioners suggest. Although one provision for 'certification' of stationary sources was deleted from the Senate bill by the conference committee,27 another provision which gives the Administrator substantial power to block construction of any stationary source should he determine that it will not comply with applicable standards survived. 42 U.S.C. 1857c-7(c)(1)(A).28 Although this section apparently places the onus of going forward on the Administrator rather than on the prospective builder, and although no 'permit' is apparently required, its effect is to allow some type of preconstruction review.

Moreover, Congress has recently responded to EPA's parking control regulations in P.L. 93-319, 4(b)(2)(C) (June 22, 1974), which authorized the Administrator to 'suspend until January 1, 1975, the effective date or applicability of any regulations for the management of parking supply . . ..' We think that this Act, which also required EPA to report promptly to Congress on the necessity of such parking supply management, demonstrates that Congress has acquiesced, for the time being, in the strategy adopted by EPA. At a time when Congress is involved so directly, we hesitate to essay a 'definitive' resolution of the congressional intent in 1970 when the Amendments were originally enacted.

We are concerned, however, by the standardlessness of subsection (d). The clause permits denial of a permit unless the functionary passing on such requests decides that the facility 'will not interfere with the attainment or maintenance of applicable Federal air quality standards . . ..' This may be a form of words meaning that the facility meets the limits on parking spaces set forth elsewhere in the plan; on the other hand it may mean anything the permit-giver decides. The regulation does not indicate how 'interference' is to be judged, nor does it state who must bear the burden of showing noninterference. The prospective applicant for a permit is utterly without guidance as to what he must prove, and how. And the standard is so vague that it invities arbitrary and unequal application.

We acknowledge that Congress often provides little guidance to the agencies it creates.29 But the premise supporting congressional and judicial approval of such vague delegations is that the agency, through the rulemaking and administrative process, will gradually fill in areas of uncertainty with specific rules. See NLRB v. Bell Aerospace Co. Division of Textron, Inc., 416 U.S. 267, 290-295, 94 S. Ct. 1757, 1770-1772, 40 L. Ed. 2d 134 (1974); Ehrlich & Posner, An Economic Analysis of Legal Rulemaking, 3 J. Legal Studies 257 (1974). The promulgation of subsection (d) neglects to include such needed specifics.

The judgment of 'interference' by an indirect source, which itself is nonpolluting, may be much more difficult than the judgment of stationary sources contemplated by 1857c-7(c)(1)(A), because the latter are governed by regulations setting source-by-source limits on emissions; all that is necessary to evaluate compliance is to compare the potential performance of the stationary source which the requirements for that source already laid out in regulations. There are no such regulations for indirect sources, nor do such sources possess any other attributes that would make for simple evaluation. We think that the judgment of 'interference' will therefore be a complex one; those regulated should understand how the judgment will be made, so that they can be intelligent participants in the process. We recognize that subsections (f) and (g) include some criteria that the Administrator will require to be considered by the subagencies which will enforce subsection (d). But none of these criteria modify or clarify the problems with making decisions about 'interference' with air quality standards. We assume, for purposes of this discussion, that (d) is to be taken at face value and independent 'interference' decisions made about each applicant. We disapprove the 'interference' clause as now worded. It may, of course, be amended so as to indicate with requisite specificity what constitutes forbidden 'interference' and how the determination is to be made.

Subsection (e), on the other hand, we approve as both justified in concept and sufficiently precise. It apparently establishes a 'parking space bank'; those who propose to add parking spaces in a freeze zone must obtain the Administrator's agreement that unused space allocations are available. Here there is an ascertainable standard-- availability of spaces because others have been retired. It is numeric and easily applicable. The difficulty with subsection (e), of which no petitioner has complained, is that it contains no operational directions concerning who will be allowed to build when spaces become available. It is possible that internal operating procedures of the permit agency will bring about inequalities, but we pretermit that problem. Applicants for the available spaces who believe that the system used to allocate them is arbitrary or capricious may raise that challenge if an abuse occurs or is threatened.

WHETHER TRANSPORTATION CONTROLS ARE ARBITRARY AND CAPRICIOUS

1. The 'freeze' boils down to the requirement that no new parking spaces be created after October 15, 1973, in the more congested portions of Boston, Cambridge, and some other outlying areas. There are important exceptions: residential spaces (parking adjacent to homes, apartments, condominiums, etc.), employee parking outside the Boston core (so long as it complies with the separate employee parking restrictions), and free customer parking. Our role, of course, is not to decide whether the freeze device is an ideal solution; Congress delegated to EPA the authority (assuming the state did not exercise it) to select the preferred means. We cannot say that such a freeze is arbitrary and capricious assuming EPA is able to support by credible data its position as to the magnitude of the need for carbon monoxide emission reductions in relevant segments of the region. Indeed, the enlargement of parking facilities in areas where the public health requires curtailing the flow of traffic would itself seem irrational. The exemption for residents, customers and, in parts of the area, employees, would seem a reasonable attempt to ameliorate the hardship upon individuals and businesses.

2. Regulations first proposed would have decreased parking in the Boston core by prohibiting on-street parking from 6 to 10 a.m. and 4 to 6 p.m. weekdays, and by imposing a surcharge of $5 on off-street parking from 6 to 10 a.m. The two measures were expected to eliminate about 154,000 VMT (13 percent) from the Boston core. The $5 surcharge was estimated to be able to divert 40 percent of home-based work travelers to other transit.

The final plan reduced the surcharge, which has since been revoked. The plan now bans on-street parking from 7 to 10 a.m. weekdays, and requires each off-street facility to reduce the number of spaces available weekdays from 7 to 10 a.m. by a percentage to be established by the City of Boston and to be so calculated that overall available off-street parking will not exceed 60 percent of the supply available on October 15, 1973. This so-called 'vacancy rate' approach was suggested at the public hearing by the Governor and was supported by Boston city officials. We cannot say that it is arbitrary or irrational (assuming, again, that EPA is able to sustain the need for emission reductions of the order asserted). There is reason to hope that the economic effect on garage and lot owners may be recoverable through higher fees; some garages may also be able to accommodate more shoppers. In any event, it is not clear how alternative methods of discouraging vehicular use will be less burdensome overall.

Petitioners complain that the percentage reduction is to be individually determined by the City of Boston; each facility, they allege, may be regulated differently with inconsistent space reductions. But Boston may not act arbitrarily. Capricious action by the city would be subject to redress in the courts. In light of different 'natural' vacancy rates, it might have been unfair for EPA to have set an across-the-board reduction rate for all facilities. There is no reason not to allow Boston to formulate a plan, which must, of course, be a fair one, for reaching the specified percentage reduction. Cf. Friends of the Earth v. EPA, supra, 499 F.2d at 1123-1124.

Petitioners also complain that the regulations appear to encompass residential as well as commercial parking; but such an interpretation would not be consistent with the preamble which refers to off-street parking regulations as controls 'on commuter parking'. 38 Fed.Reg. 30961 (1973). EPA represents that it intends an amendment 'to clarify that residential parking spaces, free customer spaces and employee parking spaces are exempt'. We approve the regulation as so interpreted but not otherwise.

One petitioner complains that in combined residential/commercial facilities, residents would have to remove their cars before 7 a.m. to lots outside the core area. This may, however, be avoided by clearly separating the spaces, creating, in effect, two parking lots. It is not unreasonable to require that residential parking be so cordoned off.

Finally, petitioners complain that a large university parking facility located outside the core area is not affected. The campus allegedly is serviced by excellent public transport and contains one of the largest parking facilities in Boston. The core and freeze area were delineated by selecting the section with the worst carbon monoxide problem or where commuters would be likely to park if kept out of the core area. Large parking lots thus may be left unregulated, but that does not mean that the classification is irrational.

3. The plan calls regionwide for a 25 percent reduction by employers30 of the number of available employee parking spaces, or a reduction in the number of spaces necessary to attain a ratio of parking spaces to employees of .75, whichever reduction is greater. An increase of spaces upon an increase in employees is allowed, but not to exceed certain ratios set by EPA of spaces to employees. Employee parking controls were inserted after the one-day-a-week vehicle prohibition sticker system had generated widespread opposition; the Administrator described the former as a partial substitute. Petitioners now claim that EPA has not sufficiently explained its selection of the new strategy, that there is no evidence it will improve air quality, and that other strategies should have been explored.

Since the employee parking provision succeeded the rotating vehicle prohibition, we may look for its underlying rationale to factors said to justify the earlier proposal. Vehicle prohibition was significantly aimed at work-oriented travel. EPA was concerned that 'the availability of ample, low-cost parking facilities and high speed freeways influence individuals to use vehicles with as few as one person in them, rather than less-polluting modes of transit'. 38 Fed.Reg. 30632 (1973). The technical support document states that, of the 54 percent of vehicle trips which would be subject to elimination by the prohibition, 28 percent, the largest group, were home-based work trips. Moreover, work trips were believed to average longer than other types. Work trips comprise 40 percent of all vehicle miles of travel. Vehicle use prohibition was selected, according to the record, because it costs less than achieving the same emission reduction by adding mechanical emission control devices to vehicles, and is more readily enforceable and less disruptive than gasoline rationing or vehicle use prohibitions geared to daily pollution levels.

While vehicle use prohibition would have decreased hydrocarbon emission by more than the substituted strategy, the substitute seems plainly less disruptive and more acceptable. The sticker proposal would have kept most persons off the road on an arbitrarily selected day; the final strategy is more flexible. Employers can allocate spaces to those unable to obtain public transport or carpools, and special arrangements can be made for hardship and emergencies.

Petitioners rely upon the following remark in a consultant's report: 'The concept of penalizing only peak period automobile travelers, particularly work purpose trip makers, does not result in significant air pollution reduction.' This statement appeared in a discussion about the wisdom of varying the cost of parking in downtown facilities; the consultant believed penalizing peak periods would only result in extra spaces, and encourage off-peak travelers to use automobiles, an argument not relevant to employee parking lots; and in subsequent reports the same consultant recommends strategies which do penalize work travelers. Commuters are the largest identifiable segment of the driving population. Given the significant number of vehicle miles generated by that activity, we cannot say it was arbitrary or capricious to have placed major controls upon commuters. The argument that it was unfair not to penalize those driving to race tracks, shopping centers, and the like is legally unpersuasive so long as the strategy that was adopted is rationally related to the Agency's aims. All strategies will involve something less than completely equal allocation of burdens.

The contention that the Agency did not consider alternatives is belied by the record. EPA paid attention to and rejected a wide range of strategies, including mechanical emission control devices, episodic based vehicle prohibition, gasoline rationing, staggering work hours, shortening the work week, gaseous fuel conversion, and the sticker system.

'There is no reason to make a distinction between explanations for negative exclusions and an affirmative choice. In circumstances of this kind, the petitioners must carry the burden of going forward with a reasonable claim of something important overlooked instead of placing on the Administrator the burden of knocking down all possible objections in advance. If nothing can be done until everything is explained, the mandate of the statute will never be translated into accomplishment.'

Natural Resources Defense Council, Inc. v. EPA, 494 F.2d 519, 525 (2d Cir. 1974). See Friends of the Earth v. EPA, supra, 499 F.2d at 1126. We think it clear that petitioners have not met this burden.

Petitioners' most telling point seems to be that the regulations may cause special hardship in parts of the region not well served by public transit and to firms in which a large proportion of employees work irregular hours, so that car pooling is impractical. EPA undoubtedly hopes to force commuters to turn to car pooling and public transit, and to create a demand resulting ultimately in adequate public transit. In the interim employers may have to sponsor their own bus service. Nonetheless, the restrictions on employee parking have obvious potential for individual hardship going beyond mere inconvenience.

EPA has informed the court that it plans to amend the provision on employee parking

'to allow exemptions from some or all of the requirements to be granted by the Regional Administrator to employers which cannot comply without severe adverse economic consequences and which have taken all practicable steps to reduce employee VMT.'

In Natural Resources Defense Council, Inc. v. EPA, 478 F.2d 875, 884-888 (1st Cir. 1973), we said that in the period before the date set for mandatory attainment of standards, the Administrator may grant individual variances provided they cease before the mandatory compliance date and are not inconsistent with national objectives. But we also held that once the mandatory attainment date arrived, 'sources of pollutants should either meet the standard of the law, or be closed down.' Id. at 886. The exclusive mechanism for hardship relief was said to be the procedure outlined by Congress in 42 U.S.C. 1857c-5(f).31 Yet 1857c-5(f) does not readily lend itself to hardship relief from the impact of employer parking lot regulations. The clause applies only to regulations in state implementation plans which affect a 'stationary source or class of moving sources'; employer parking lots do not fit within either category. Congress has not specifically addressed itself to the regulation of indirect sources, such as parking lots; the parking restrictions, unlike emission controls, would seem to be governed by considerations different from those we have previously discussed. While we can make no prediction without seeing the particular provision, it seems likely that a regulation can be drafted providing for reasonable and even-handed hardship exemptions from restrictions of this type that do not violate the Act. Cf. United States v. Allegheny-Ludlum Steel Corp., supra, 406 U.S. at 753, 92 S. Ct. 1941, 32 L. Ed. 2d 453 (1972).

Pending opportunity to review the proposed variance provision, we provisionally uphold the employer parking controls as written. We find them to be a permissible exercise of EPA's authority, and, assuming a reasonable provision for exceptional hardship can be formulated, we are unable to say that they are either arbitrary or capricious.

4. In 1970 four airlines organized South Terminal in order to construct a new parking facility at Logan Airport. A contract with Massport was executed in April, 1973, at which time construction began. Both South Terminal and Massport object to the freeze, which provides that facilities in construction prior to October 15, 1973, may not be utilized to the extent that they increase existing parking facilities in specified municipalities and at Logan Airport by more than 10 percent unless any such increase is offset by retiring spaces elsewhere in the freeze zone. Each municipality and Logan constitute a separate freeze zone, for the purpose of calculating the 10 percent reduction. Thus, South Terminal may not increase spaces at Logan by more than 10 percent and it is projected to be 1,100 spaces over its limit.

The original proposal would have imposed a $5 surcharge on all persons parking at Logan. South Terminal estimated that it would have resulted in a 35 percent decrease in the utilization of its complex-- more drastic, perhaps, than the final plan because South Terminal could not assure full utilization by purchasing and retiring open lot spaces. South Terminal protested that the surcharge would be ineffective and that the data did not justify the harsh treatment; it focussed upon the adverse effects the plan would have on the revenue of commercial airlines, particularly the airlines which own South Terminal.

South Terminal argues that if Logan must be controlled it should not be singled out in such a fashion that it must exclusively bear the cost of any reduction. The $5 surcharge, unlike the freeze, would have affected revenues throughout the airport. We reject this argument. As we said earlier, the regulation is not aimed at South Terminal but at checking any excessive increase in the total vehicle population at this one location. Assuming EPA prevails in establishing the need for the controls, it is immaterial that they do not fall equally upon every operator; we think it rational that those seeking to build new facilities receive a lower priority than those whose facilities are already built. South Terminal had an opportunity, denied to those with completed facilities, to alter its plans and to cushion the shock of the regulations. We note, in addition, that South Terminal is closely associated with Massport and other airport operators; it is not unreasonable to anticipate that suitable adjustments will ultimately be made among all those with a major stake at Logan.

South Terminal next maintains that Logan should not be the only place in East Boston to bear the costs of carbon monoxide reduction at that region. But if, in fact, it turns out in the reopened hearings we have ordered that airport vehicular traffic is a major pollution source, it would not be irrational to control only the airport.

South Terminal also claims that it is inconsistent to maintain the freeze while deferring preconstruction review. Assuming the complaint can properly be termed a legal argument, it is unavailing, because to our knowledge the only deferred review provision, clause (d), is unrelated to the parking freeze.

Finally, South Terminal and Massport claim that they would like to prove that the freeze strategy arbitrarily replaced the surcharge because it was negotiated in advance of the public hearing. They point to the following remark uttered by Governor Sargent at the opening of the hearing: 'I am pleased the Environmental Protection Agency has incorporated my proposed parking freeze for the Boston core area and for Logan Airport.' Under the Freedom of Information Act, EPA has made available all written documents relevant to conversations between EPA and the state regarding the transportation control plan. They reveal no evidence of any written agreement. The Governor was a proper and, indeed, logical source for suggestions, and we find nothing in the conduct of the proceedings to show that the Administrator was irrevocably committed to a freeze. The Administrator was not precluded from holding views before the hearing so long as he did not shut his mind to other ideas and remained open to persuasion. The freeze replaced other measures, such as the $5 surcharge, which had generated hostile public comments. We deny petitioners' motion to supplement the record with memoranda reflecting oral or written communications between EPA and the state regarding the plan, and the motion to appoint a special master to take oral testimony from officials of EPA. Petitioners' reliance on Overton Park, supra, 401 U.S. at 402, 91 S. Ct. 814, 28 L. Ed. 2d 136, in which there was no indication why the Administrator acted as he did, is misplaced. Possibly barring fraud and other extreme circumstances, the mental process by which the Administrator reached his decision, if it is explained by the record, is not a proper subject for discovery. See United States v. Morgan, 313 U.S. 409, 422, 61 S. Ct. 999, 85 L. Ed. 1429 (1941).

5. Petitioners contend that all the transportation controls are arbitrary because the Administrator has paid too little attention to the plan's economic and social impact. The material portions of the Clean Air Act itself do not mention economic or social impact, and it seems plain that Congress intended the Administrator to enforce compliance with air quality standards even if the costs were great. Particularly in the case of primary standards-- those set as 'requisite to public health'-- Congress' position is not extreme or unprecedented. Minimum public health requirements are often, perhaps usually, set without consideration of other economic impact. Thus, insofar as petitioners claim that either EPA or ourselves would be empowered to reject measures necessary to ensure compliance with primary air quality standards simply because after weighing the advantages of safe air against the economic detriment, we thought the letter consideration took priority, petitioners would be incorrect. Congress has already made a judgment the other way, and EPA and the courts are bound. See Kennedy, Legal Formality, 2 J. Legal Studies 351, 366-77 (1973).

Therefore, we find no statutory duty32 imposed upon EPA to make a cost and social benefit analysis. Petitioners contend that the Agency must file the 'functional equivalent' of the environmental impact statement required of other federal agencies. As already noted, however, Congress has made it plain that NEPA does not apply to EPA. P.L. No. 93-319, 7(c)(1) (June 22, 1974). The 'functional equivalent' rule was developed in cases construing 111 and 211 of the Clean Air Act, which require the Administrator to take 'into account the cost of achieving . . . reduction' or to prepare a 'cost benefit analysis' in order to regulate stationary sources or fuels and additives. See Portland Cement Ass'n v. Ruckelshaus, supra, 486 F.2d at 379. Amoco Oil Co. v. EPA, 501 F.2d at 750 (D.C.Cir.1974). No such language appears in 110, and we think it inappropriate to substitute any such gloss by judicial initiative.

Economic considerations may play some role, however, in EPA's selection among alternative means to achieve its mandated clean air goal. EPA guidelines encourage states to identify the 'costs and benefits' of alternative control strategies.33 Although we do not read the Act as requiring EPA to engage in exhaustive cost benefit studies or to initiate elaborate planning exercises, it could be arbitrary and capricious for the Agency to reject obviously less burdensome but equally effective controls in favor of more expensive or onerous ones. But we think a considerable part of the burden of suggesting attractive alternative strategies is upon those, like the petitioners, who dislike the present ones. The record discloses that the Administrator has chosen rationally among the viable alternatives presented. We conclude that he did not abuse his discretion or go beyond his statutory authority with respect to the economic and social aspects of the plan.

Several petitioners have argued that the powers of EPA, as construed by us, supra, constitute an unconstitutional delegation to an agency of legislative powers. We do not find the argument persuasive. The last time that a delegation of power to an administrative agency was upset occurred in A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495, 55 S. Ct. 837, 79 L. Ed. 1570 (1935), and the unique conditions of that case are not repeated here.

In Schechter Congress had delegated to the President the power to approve industry 'codes' drawn up by local businessmen. Congress had not prescribed a purpose to be served by the codes, nor had it set boundaries on the provisions the codes could contain. The Court consequently characterized the delegation as utterly without standards and impermissible. Justice Cardozo, concurring, wrote that the legislation was unconstitutional because the power granted was 'not canalized within banks that keep it from overflowing. It is unconfined and vagrant . . ..' Id. at 551, 55 S.Ct. at 852. 'Here in effect is a roving commission to inquire into evils and upon discovery correct them.' Id.

The power granted to EPA is not 'unconfined and vagrant'. The Agency has been given a well defined task by Congress-- to reduce pollution to levels 'requisite to protect the public health', in the case of primary standards. The Clean Air Act outlines the approach to be followed by the Agency and describes in detail many of its powers. Perhaps because the task is both unprecedented and of great complexity, and because appropriate controls cannot all be anticipated pending the Agency's collection of technical data in different regions, the Act leaves considerable flexibility to EPA in the choice of means. Yet there are many benchmarks to guide the Agency and the courts in determining whether or not EPA is exceeding its powers, not the least of which is that the rationality of the means can be tested against goals capable of fairly precise definition in the language of science.

Administrative agencies are created by Congress because it is impossible for the Legislature to acquire sufficient information to manage each detail in the long process of extirpating the abuses identified by the legislation; the Agency must have flexibility to implement the congressional mandate. Therefore, although the delegation to EPA was a broad one, including the power to make essentially 'local' rules and regulations when necessary to achieve the national goals, we have little difficulty concluding that the delegation was not excessive. Cf. L. Jaffe, Judicial Control of Administrative Action 57-72 (1965); 1 K. C. Davis, Administrative Law Treatise P2.01 (1958).

Several petitioners contend that, even if the delegation was proper, Congress lacked any power to grant because there is insufficient power under the commerce clause to reach the local transportation activities here involved. We think petitioners underestimate the breadth of congressional power in this area.

Motor vehicles are indisputably in commerce. Even though any individual motor vehicle may travel exclusively within one state, commerce by motor vehicle sufficiently touches multi-state concerns as to be federally regulable. Katzenbach v. McClung, 379 U.S. 294, 85 S. Ct. 377, 13 L. Ed. 2d 290 (1964); Heart of Atlanta Motel, Inc. v. United States, 379 U.S. 241, 85 S. Ct. 348, 13 L. Ed. 2d 258 (1964); Pennsylvania v. EPA, 500 F.2d 246 (3d Cir. 1974). Similarly, the problem of pollution itself involves the nation as a whole; pollutants are not respecters of state borders.

Once Congress has validly exercised the power it possesses, it is free to proscribe those things made legal, or even commanded, by state authority. Thus the regulations ordering a reduction in employee parking space below the ratio required by local zoning ordinances are valid; the federal rule controls under the supremacy clause.

'It is no objection to the exercise of the power of Congress that it is attended by the same incidents which attend the exercise of the police power of a state.' FPC v. Natural Gas Pipeline Co., 315 U.S. 575, 582, 62 S. Ct. 736, 741, 86 L.Ed.1037 (1942).

EPA was free to promulgate rules that resemble local zoning ordinances; this does not constitute 'usurpation of the police power', as one petitioner has asserted. The Federal Government may use the same tools as may the state when in pursuit of an objective lawfully within the power of either.

The argument has been advanced that the plan involves a forbidden 'retroactive' tampering with vested property rights. South Terminal Corporation and Massport argue that the retroactivity arises because the regulations burden the use of the South Terminal parking garage, which was under construction on the date the transportaton control plan was promulgated.

We do not see any 'retroactivity' here. The regulations do not penalize South Terminal for any conduct which, when engaged in, was permitted. At most they abrogate, for the future only, expectations South Terminal may have acquired in the past. But all changes in the law dash expectations when they make tomorrow's rules different from yesterday's; that is not enough to create a 'retroactive' law, and, indeed, it is not even clear that a law having an effect which may be termed 'retroactive' is necessarily unconstitutional. See SEC v. Chenery Corp., 332 U.S. 194, 203, 67 S. Ct. 1575, 1760, 91 L. Ed. 1995 (1946); Fornaris v. Ridge Tool Co., 423 F.2d 563 (1st Cir.), rev'd on other grounds, 400 U.S. 41, 91 S. Ct. 156, 27 L. Ed. 2d 174 (1970).

The regulations restrict the uses to which South Terminal may put the garage which it is building. If all other things remain equal, it will be able to fill only 1,600 of the 2,700 projected spaces. (It also retains the option, for what it may be worth, of altering its plans.) Alternatively South Terminal and Massport, acting in concert, may 'retire' 1,100 other parking spaces currently in use at the airport, fill the South Terminal garage, and devote the land under the retired spaces to alternative uses. No matter how the regulation is construed, it does not more than state to the airport authorities that available parking spaces (under whatever corporate ownership) at Logan cannot exceed 17,875, in whatever form and location. The major thrust of Massport's argument seems to be that, in light of the development plans and expected growth rate of Logan, 17,875 parking spaces are insufficient, and that the limitation will be ineffective or counterproductive. But even if this were so, it would make the regulation an unwise one, not an unconstitutional one.

The airport petitioners and all parking operators in the Boston core area seek to convince us that the regulations constitute a taking without just compensation. The regulations as applied to Logan exterminate some 1,100 planned-upon spaces and arguably confiscate the revenues that otherwise would have accrued from them. The 40 percent vacancy rate rule in the Boston core area compels building space to stand idle; the situation is arguably most disadvantageous to garage owners, for their space is least likely to have a reasonable alternative use. The garage owners may argue that it is as if the Government had taken title to 40 percent of their spaces; it would matter little if thereafter the Government kept the space idle, devoted it to some other nonremunerative end, or found some other use for it.

However, the Government has not taken title to the spaces, and the decision about alternative uses of the space has been left to the owner. The takings clause is ordinarily not offended by regulation of uses, even though the regulation may severely or even drastically affect the value of the land or real property. If the highest-valued use of the property is forbidden by regulations of general applicability, no taking has occurred so long as other lower-valued, reasonable uses are left to the property's owner. Goldblatt v. Town of Hempstead, 369 U.S. 590, 592, 82 S. Ct. 987, 8 L. Ed. 2d 130 (1962); Turnpike Realty Co. v. Town of Dedham, 1972 Mass.Adv.Sh. 1303, 284 N.E.2d 891, cert. denied, 409 U.S. 1108, 93 S. Ct. 908, 34 L. Ed. 2d 689 (1973).

Three situations must be distinguished. First, a particular use of a parcel of property may be regulated or forbidden. Second, all uses of a parcel may be forbidden. Third, a right to use or burden property in a particular and permitted way may be transferred from the original owner to another person, or to a governmental body. Only the second and third situations are thought of as takings today. Thus in Pennsylvania Coal Co. v. Mahon, 260 U.S. 393, 43 S. Ct. 158, 67 L. Ed. 322 (1922), the transfer from mineral owners to surface owners of the right to control subsidence in the land was held to be a compensable taking. At the same time the Court recognized the first situation, stating:

'Government hardly could go on if to some extent values incident to property could not be diminished without paying for every such change in the general law.' Id. at 413, 43 S.Ct. at 159.

Our situation fits within the boundaries of the first type. EPA's rule is of general applicability, regulating one use of those available. The right to use is not extinguished entirely; nor is it transferred to anyone else. Indeed, the ingenuity of operators may result in fewer disadvantages than urged. For example, some operators may be able to 'buy' from others the right to use spaces, leaving the seller of spaces free to use the land under his parking lot for other purposes while the buyer enjoys a higher occupancy rate.

In any event, even a diminution of profits or a requirement that some loss be suffered is not enough, when all other accoutrements of ownership remain, to be a 'taking'.34

Finally, the Government reminds us that the restriction on parking availability in the Boston core area will allow entrepreneurs to increase their prices, as is the natural consequence when supply is reduced and demand is unchanged. The Government has effectively created a parking cartel that, depending on the elasticity of demand for parking, may increase rather than decrease profits. What will happen to profits cannot be predicted, but in view of the possibility of their increase we are not impressed by the claim that the regulation is so serious, and so forecloses alternatives, that a compensable taking has occurred.35

The airport petitioners raise the claim that the regulations violate the contracts clause of the Constitution. The point is without merit.

'The exercise of the police power in the interest of public health and safety is to be maintained unhampered by contracts in private interests, and . . . uncompensated obedience to laws passed in its exercise is not violative of property rights protected by the Federal Constitution.' Northern Pac. Ry. v. Minnesota ex rel. City of Duluth, 208 U.S. 583, 597, 28 S. Ct. 341, 346, 52 L. Ed. 630 (1908).

The claim of the airport petitioners is in two parts: first, that they have let a construction contract for the South Terminal garage, and that the value of that contract to petitioners will be reduced if all of the spaces in the garage cannot be used; second, that Massport and associated concerns have contracts with airlines and with bondholders under indentures, and that these contracts will become less valuable to the bondholders and airlines because Massport and South Terminal, deprived of anticipated parking revenue, will be less likely to keep up payments as called for.

The contentions prove too much. Were they accepted, all governmental action affecting the profitability of private concerns would violate the contracts clause because it would change relative values and make those enterprises the prospects of which were diminished by the legislation somewhat less likely to keep their bargains. We do not understand that this is the import of the contracts clause. We read it as forbidding instead an alteration in the relative position of two parties to an existing contract; once these parties, as between themselves, have allocated rights and responsibilities, it is not within the power of the government to rearrange them. Home Building & Loan Ass'n v. Blaisdell, 290 U.S. 398, 428, 54 S. Ct. 231, 78 L. Ed. 413 (1934). Even then the clause is not read with literal exactness; minor reallocations, not going to the heart of the bargain, are sometimes permitted to effect an overriding public purpose. But this sort of alteration of relative obligations is considerably different from the 'diminished profitability and therefore diminished ability to keep up obligations' argument made by petitioners. An otherwise valid governmental regulation does not become impermissible merely because an object of the regulation is a party to some contracts. Nor can a party make otherwise unlawful action permissible merely by making a contract about it.

The plan requires gasoline companies to control the evaporative losses which result from the transfer of gasoline from any delivery vessel into any stationary storage container and its transfer to an automobile fuel tank at retail gasoline outlets.

Several petitioners attack the feasibility of the regulation covering the transfer of gasoline to an automotive fuel tank.36

Texaco and Gulf maintain that the Administrator's failure to consider the lack of effective, safe and commercially available equipment makes the strategy arbitrary.37 The record reveals that several prototypes exist, but that the oil companies do not believe that equipment meets safety standards, or has been tested sufficiently.38

On June 18, 1974, following oral argument, EPA promulgated a delay in the dates for submission of control plans, submission of signed contracts, and initiation of construction for the installation of gasoline vapor recovery systems for vehicle refueling, and asked for additional comment on which types of systems should be required. 39 Fed.Reg. 21049-53 (1974). This action was taken because 'substantial confusion has arisen as to the type of equipment and necessary recovery efficiency required to comply'. Id. at 21050. Essentially two techniques have been developed-- simple displacement and blower-assist. 'A substantial controversy has arisen over the effectiveness of simple displacement systems, on the one hand, and the reliability of blower-assist systems, on the other.' At the time the regulations were drafted EPA believed only the blower-assist system would meet the 90 percent recovery requirement, but there have been claims that other systems may achieve this.

'It is because of these claims that clarification of the possible interpretation of the existing regulations becomes important for determining what type of control system or systems will be allowed to be installed to meet the requirements of the regulations. 'The administrator has concluded that there have been sufficient developments since the regulations were drafted, and that there is sufficient uncertainty about which systems will be approvable as to be in compliance with these regulations, to warrant a reopening of the opportunity for public comment on this issue until July 31, 1974. In addition, an EPA-funded testing program is being carried out in San Diego County to attempt to measure the performance of various systems with results expected by August 1, 1974.' Id. at 21052.

The gasoline vapor recovery regulation at issue has no peculiar local features; it is identical to regulations appearing in several other state plans. Whether or not the technology is available remains a national question. In light of the changes in initial cimpliance dates and requests for comments, fairness and common sense dictate that our approval be withheld at this time until the Administrator has completed and published the results of the latest EPA proceedings to investigate technological feasibility.39 As soon as new regulations or clarifications are promulgated, EPA shall furnish us with copies and shall notify us whether it intends to retain or alter the plan in light thereof. Thereafter the parties will be allowed a short time to submit supplementary briefs addressed to the technological feasibility point only and to whether or not, in this regard, we should approve the controls on transfer of gasoline to an automotive fuel tank.

1. To hold a public hearing, consistent with the provisions of 5 U.S.C. 553 for informal rulemaking, no later than ninety days for the following purposes:

(a) To receive and consider further objections and argument, oral and written, respecting the technical basis for calculating the amount of hydrocarbon and carbon monoxide reduction required to attain and maintain primary photochemical oxidant and carbon monoxide standards within the Metropolitan Boston Intrastate Region;

(b) To evaluate, affirm, or modify, as the Administrator thinks appropriate, the emission reductions stated in the plan, and, in light of any modifications thereof, to modify as he deems appropriate any other portion of the plan;

(c) To adopt as part of the plan a regulation providing a procedure for the periodic updating of goals and verification of progress, to the end that transportation control restrictions are neither more nor less rigorous than actually required to meet air quality standards.

2. To publish his final order and responses concerning the above matters within thirty days of the date of such hearing and to file a copy thereof with the court. This court shall retain review jurisdiction pending the results of the hearing and shall rule upon such remaining matters as necessary respecting the Metropolitan Boston Air Quality Transportation Control Plan after the Administrator has filed said order. Any objections to the Administrator's order shall be filed in this court within ten (10) days of the publication thereof.

3. To suspend controls under 52.1135, 52.1136, 52.1144 until further order of this court, provided taht persons must continue to comply with the requirements of clauses (i) and (j) of 52.1135 and (g) of 52.1136 except as the Administrator may otherwise in his own discretion in the interest of fairness determine.

4. To notify the court within five (5) days after publication of any amendments or clarifications respecting 52.1144(d).

Copies of filings under (2) and (4) above shall be served upon the petitioners when made. Petitioners may file written responses thereto with the court within ten (10) days thereafter.

So ordered.