dissenting on the Atomic Energy Act issue:

The Nuclear Regulatory Commission has abdicated a significant part of its statutory responsibility to protect the health and safety of the American people. It has taken the position that, even though the Atomic Energy Act prohibits the grant of a license if the public health will be endangered, licensing proceedings may ignore potential effects on psychological health.1 Today the court, giving undue deference to the agency’s views, endorses this erroneous interpretation of the statute. I therefore dissent.

In my view, the plain meaning of the statute compels the Commission to consider psychological health contentions in licensing proceedings. The Commission may not grant a license if, in its opinion, “the issuance of a license to [the applicant] would be inimical to * * * the health and safety of the public.” 42 U.S.C. § 2133(d) (1976). Not all fears and worries, of course, are psychological health effects within the definitions of medical science. The adverse psychological impact of restarting TMI-1 may or may not rise to the level of a health problem; even if it does, the same might not be true of the fears and anxieties of neighbors of other power plants. Drawing these lines, on the basis of the facts, is the task of the Commission. If operation of a nuclear facility would be inimical to the psychological health of the public, the Commission must not approve operation. On the other hand, if investigation shows that allegations of psychological health effects are unfounded or that the effects are de minimis, the Act does not prohibit grant of a license. In short, the statute requires the Commission to make its decisions in light of full information.

Unfortunately, the Commission’s position short-circuits this process of scientific assessment and informed decisionmaking. The court affirms the Commission, taking a “deferential posture” and asserting that “[t]here simply is no basis for this court to overrule the agency’s interpretation.” Wilkey opinion (AEA) at 250,253. I respectfully submit that this is a case in which such deference is inappropriate. As the Supreme Court recently reiterated,

The interpretation put on the statute by the agency charged with administering it is entitled to deference, * * * but the courts are the final authorities on issues of statutory construction. They must reject administrative constructions of the statute, whether reached by adjudication or by rulemaking, that are inconsistent with the statutory mandate or that frustrate the policy that Congress sought to implement. * * *

FEC v. Democratic Senatorial Campaign Committee, 454 U.S. 27, 31-32, 102 S.Ct. *39138, 42, 70 L.Ed.2d 23 (1981). In my view, the legislative background shows that the Commission’s constricted interpretation of “health” must be rejected.

Congress did not restrict the Commission’s mandate to “physical health” or to “special hazards of radioactivity.”2 Instead it used the broad, open-ended term “health” in the Atomic Energy Act. Aware that relatively little was known about the health dangers of atomic energy at the time the Act was passed in 1946 and revised in 1954, Congress acted wisely for the future. The language of the statute goes beyond the specific health hazards then known to the scientific community. It assures that the statute will continue to provide full protection to the public as knowledge about the potential health effects of atomic energy expands.

In 1946, for example, many of the long-term health effects of large-scale radiation exposure were as yet unknown. Studies of the survivors of Hiroshima and Nagasaki, which began to be published in the mid-1950’s and which continue to the present day, have provided the largest set of data on the effects of ionizing radiation on production of leukemia, other cancers, and genetic effects in human beings. See MEDICAL Research Council, The Hazards to Man of Nuclear and Allied Radiations (1956), reprinted in The Nature of Radioactive Fallout and Its Effects on Man, Hearings Before the Special Committee on Radiation of the Joint Committee on Atomic Energy, 85th Cong., 1st Sess. 1539, 1554, 1562, 1581, 1624-1626 (1957); CASARETT & Doull’s Toxicology 511 (J. Doull, C. Klaassen & M. Amdur eds., 2d ed. 1980). Yet it would be absurd to suggest that leukemia and genetic defects fall outside the scope of the Atomic Energy Act. Similarly, even though Congress did not refer explicitly to psychological health and even though Three Mile Island has given the scientific community its first opportunity to study the possible psychological health effects of a nuclear power plant accident, the Atomic Energy Act requires the Commission to consider these effects.

The Atomic Energy Act of 1946 created the Atomic Energy Commission, predecessor of the Nuclear Regulatory Commission, and entrusted it with broad responsibility for protecting the public health and safety. The statute reflects Congress’ acute awareness that nuclear energy could be extremely dangerous, and that scientific knowledge of the potential hazards was completely inadequate. Congress established an absolute government monopoly over production of fissionable material, recognizing that this activity was “attended by serious hazards to public health and safety” and that the responsibility for minimizing these hazards is “clearly a governmental function.” S.Rep. No.79-1211, 79th Cong., 2d Sess., reprinted in [1946] U.S.Code Cong.Serv. 1327, 1330. It also imposed strict controls over licensing of devices utilizing atomic energy, because it believed that such devices, “if widely used, would so multiply potential hazards' to national health and safety that even careful Government regulation would fail to provide adequate safeguards.” Id. at 1333. At the same time Congress realized that the scope and extent of potential hazards were unknown. Recognizing the imperative need for “an ever-expanding fund of theoretical and practical knowledge,” the bill encouraged both government and private research in several fields, including “protection of health during research and production activities.” Id. at 1330; H.R.Rep.No.79-2478, 79th Cong., 2d Sess. 8 (1946).

Given this context of acknowledged uncertainty, the Commission is simply wrong when it asserts that the 1946 Act limited health hazards to nonpsychological “special hazards of radioactivity.” Brief for respondents at 31. Neither the statute nor the *392legislative history gives any indication that Congress rejected the ordinary meaning of the word “health” — psychological as well as physical health. The Act gave the Commission authority to establish regulations that it might deem “necessary or desirable to protect health or to minimize danger from explosions and other hazards to life or property.” Atomic Energy Act of 1946, Pub.L.No. 79-585, § 12(a)(2), 60 Stat. 755, 770. The Senate report paraphrased the language of the statute: regulations were “to minimize the danger from explosion, radioactivity, and other harmful or toxic effects incident to the presence of such materials.” [1946] U.S.Code Cong.Serv., supra, at 1335. Thus, as new types of health hazards were scientifically established, the range of Commission regulation would expand.

In 1954 Congress amended the Atomic Energy Act to establish a role for private industry in the development of atomic energy. The Commission was authorized to grant licenses to private enterprises to operate nuclear facilities. Comparing the situation in 1954 with that in 1946, the House and Senate committees optimistically reported, “It is now evident that greater private participation in power development need not bring with it attendant hazards to the health and safety of the American people.” S.Rep.No.83-1699, 83d Cong., 2d Sess. 3 (1954); H.R.Rep.No.83-2181, 83d Cong., 2d Sess. 3 (1954), U.S.Code Cong. & Admin. News, pp. 3456, 3458. This reassuring statement must be read in conjunction with Section 103 of the 1954 Act, which imposed a heavy responsibility on the Commission: no license could lawfully be issued “if, in the opinion of the Commission, the issuance of a license to such person would be inimical to the common defense and security or to the health and safety of the public.” Pub. L.No.83-703, § 103, 68 Stat. 919, codified at 42 U.S.C. § 2133(d) (1976). The 1954 Act may have been adopted to promote development of nuclear energy, but it did not authorize the Commission to proceed at the expense of health and safety. Id. §§ 2(a), (b), (d), (e), 3(d). Again the statute did not define or limit the term “health” in any manner.

The 1946 and 1954 Acts established the Commission’s mandate to protect health in plain language which encompasses psychological health. No subsequent action by Congress has limited the scope of the Commission’s responsibilities. The Commission quotes from a 1957 study report by the staff of the Joint Committee on Atomic Energy, Statement of Reasons at 8-9, but the quoted language does not support the Commission’s position. Even if hazards created by potentially harmful radiation are “[t]he special problem of safety in the atomic field,” it does not follow that they are the only problem of health and safety that the Commission must take into account. In addition, the Commission rests on the legislative history of amendments to the Atomic Energy Act adopted in 1956 and 1965. Id. at 9-10. But neither amending statute dealt with the Commission’s licensing responsibilities over nuclear power plants. Tangential, imprecise descriptions of the Commission’s health responsibilities are entitled to no weight.

In short, the Commission has failed to offer any convincing evidence that the Atomic Energy Act excludes mandatory consideration of one of the important components of health — psychological health. An examination of the original sources shows that the Commission lifts scattered passages out of context from committee studies and committee reports. Given the plain meaning of the statutory language, I cannot agree that the Commission’s resulting interpretation of the Act is within the bounds of reasonableness.

As a fallback position the Commission urges that, even if psychological health is “health” under the Atomic Energy Act, consideration of psychological health is purely discretionary. This contention is without merit. The Act unequivocally bans issuance of a license to any facility whose operation would be inimical to the public health and safety. The Commission has *393broad procedural discretion, but it may not avoid its substantive responsibilities by pleading lack of expertise or by pointing to the difficulty of the task. If the Commission is not currently equipped to assess psychological health effects, it must add qualified experts to its staff to fulfill its statutory duties. No other government agency, state or federal, has the power to grant or deny nuclear power plant licenses; therefore it is disingenuous for the Commission to defer to “agencies with expertise in the area of mental health.” Statement of Reasons at 17. In addition, the Commission exaggerates the problem of quantification. Its own Atomic Safety and Licensing Board concluded that, at least for purposes of NEPA, “psychological stress is sufficiently quantifiable.” 11 NRC 297, 301 (1980), Joint Appendix at 67.

The importance of today’s interpretation of the Atomic Energy Act should not be underestimated. Bearing the imprimatur of this court, it will be applied not only in the TMI-1 restart proceeding but in all future Commission licensing proceedings.3 It permits the Commission to license a nuclear plant even if studies have shown beyond the shadow of a doubt that its operation would seriously damage the psychological health of large numbers of Americans. This result is inconsistent with the plain language and ordinary meaning of the Atomic Energy Act. I respectfully dissent.

Opinion for the court on the National Environmental Policy Act issue, concurred in by Senior Circuit Judge McGOWAN, filed by Circuit Judge J. SKELLY WRIGHT. Circuit Judge WILKEY dissents in Parts I and III of his opinion.

Senior Circuit Judge McGOWAN concurs in Part II of Circuit Judge WILKEY’s opinion, thereby making that Part the opinion of the court on the Atomic Energy Act issue. Circuit Judge J. SKELLY WRIGHT dissents on the Atomic Energy Act issue and files an opinion.

On March 28,1979 Three Mile Island Unit 2, a nuclear reactor operated by Metropolitan Edison Company, was seriously damaged in the worst nuclear accident Americans have yet experienced. The incident precipitated widespread alarm and led to the evacuation of many neighboring residents from their homes. At the time of the event, Three Mile Island Unit 1 (TMI-1), another Metropolitan Edison nuclear reactor of similar design which shared some common facilities with Unit 2 (TMI-2), was not in operation. The Nuclear Regulatory Commission (Commission) ordered that it remain in a cold shutdown condition pending further investigation of whether it could be operated safely. Since then the Commission has held extensive hearings on technical, managerial, and operational issues related to the proposed restart of TMI-1. The Commission has refused, however, to consider whether renewed operation of TMI-1 might cause severe psychological harm to neighboring residents and serious economic and social deterioration in nearby communities.

People Against Nuclear Energy (PANE), one of the intervenors in the restart proceeding, is composed primarily of neighbors of TMI. It seeks judicial review of the Commission’s decision to limit the scope of its inquiry in this manner. PANE contends that, under the National Environmental Policy Act (NEPA), 42 U.S.C. § 4321 et seq. (1976), and the Atomic Energy Act, 42 U.S.C. § 2133 (1976), the Commission must take into account potential harms to psychological health and community well-being. We hold that these environmental impacts are cognizable under NEPA. Therefore, the Commission must make a threshold determination, based on adequate study, whether the potential psychological health effects of renewed operation of TMI — 1 are sufficiently significant that NEPA requires preparation of a supplemental environmental impact statement.1

I. Statement of the Case

In 1974 Metropolitan Edison Company received an operating license for Unit 1, a nuclear power plant facility at Three Mile Island, Pennsylvania. Four years later the company received an operating license for Unit 2, a nuclear facility of similar design at the same site. On March 28,1979 Unit 2 suffered a serious nuclear accident which damaged the reactor, caused acute and *360widespread anxiety, and led the Governor of Pennsylvania to recommend temporary evacuation of pregnant women and preschool children from a five-mile radius surrounding the plant.

At that time Unit 1 had been taken out of operation for refueling. The Nuclear Regulatory Commission ordered Metropolitan Edison to keep Unit 1 in a cold shutdown condition pending further order by the Commission. It also announced that a hearing would be conducted to determine whether TMI-1 operations could safely be resumed. Order of July 2, 1979, 44 Fed. Reg. 40461 (1979), Joint Appendix (JA) 21. On August 9, the Commission published an order and notice of hearing regarding the restart of TMI-1. 10 NRC 141-151 (1979), JA 22. Specifying a number of issues for consideration at the hearing, the Commission’s order also stated, “While real and substantial concern attaches to issues such as psychological distress and others arising from the continuing impact of aspects of the Three Mile Island accident unrelated directly to exposure to radiation on the part of citizens living near the plant, the Commission has not determined whether such issues can be legally relevant to this proceeding.” The Commission invited parties wishing to raise such subjects in the restart proceeding to submit briefs to the Commission’s Atomic Safety and Licensing Board (Licensing Board) for consideration. 10 NRC at 148, JA 29.

Petitioner PANE, an intervenor in the restart proceeding, filed two draft contentions which are at issue in this case: it asserted, first, that restart of TMI-1 would cause severe psychological distress to persons living in the vicinity of the reactor, and second, that renewed operations would seriously damage the stability, cohesiveness, and well-being of the neighboring communities because it would perpetuate loss of citizen confidence in community institutions and would discourage economic growth. JA 84-86. In support of its draft contentions, PANE submitted a supporting brief, JA 91-117, and a preliminary plan for presentation of evidence on psychological distress, JA 88-90.

After considering briefs from PANE, other intervenors, the Commonwealth of Pennsylvania, the licensee, and the Commission’s staff, the Licensing Board issued a certification to the Commission on psychological distress issues. 11 NRC 297 (1980), JA 63. Discussing legal issues arising from the Atomic Energy Act and NEPA, the Board concluded that “the Commission, within its discretion, may and should consider psychological distress and community fears under NEPA for the purpose of mitigating the effects of its TMI-1 licensing activity.” Id. The Licensing Board accepted the contentions of the staff and the licensee that the Commission’s responsibility under the Atomic Energy Act to protect the “public health and safety” did not extend to psychological health. It described the issue as a question of first impression. “[Psychological stress,” it concluded, “is probably not cognizable under the Atomic Energy Act but * * * the Commission might conclude to the contrary for reasons not discussed by the parties.” 11 NRC at 299, JA 65. On the other hand, the Board agreed with PANE that psychological distress was cognizable under NEPA. It asserted that psychological factors were sufficiently quantifiable to be considered, 11 NRC at 301-303, JA 67-69. Considering psychological factors in the restart proceeding would assist the Commission in mitigating community fears, the Board explained. 11 NRC at 305-309, JA 71-75. It took no position on whether the Commission should prepare an environmental impact statement. 11 NRC at 304-305, JA 70-71.

When the Commission initially voted, in December 1980, on the question of whether to include psychological distress issues in the restart proceeding, one of the five seats on the Commission was vacant. The four Commissioners were evenly divided. Each Commissioner wrote a separate opinion expressing different reasons for his vote. Then-Chairman Ahearne and Commissioner Hendrie voted to exclude psychological stress issues. Then-Chairman Ahearne believed that the Commission was permitted, but not required, to consider psychological *361stress and community fears, but maintained that the best way to minimize these fears was to ensure that the plant was safe before approving restart. 12 NRC 609-611 (1980), JA 3-5. Commissioner Hendrie took the position that neither the Atomic Energy Act nor NEPA required the Commission to consider public fears, and he added, “Congress had already decided that the country is to have a nuclear power program even if it makes some people uneasy.” 12 NRC at 612-618, JA 6-12.

Commissioners Gilinsky and Bradford voted to allow psychological stress contentions to be considered in the Licensing Board proceeding. Commissioner Gilinsky was influenced by the Licensing Board’s recommendation and, more importantly, by the contention of the Commonwealth of Pennsylvania that the Commission should investigate and consider the psychological effects of restarting TMI-1. 12 NRC at 619 — 620, JA 13 — 14. Also accepting the Licensing Board’s analysis, Commissioner Bradford noted that no other agency had authority to assess and act on stress-related issues in connection with restart of TMI-1. 12 NRC at 624, JA 18. He asserted that full consideration of the extent of stress was the most effective way to deal with stress-related harms. 12 NRC at 621-626, JA 15-20.

The 2-to-2 vote constituted an effective rejection of the Licensing Board’s recommendation. Therefore the evidentiary hearing proceeded without consideration of PANE’S psychological distress and community deterioration contentions.2 In addition, the Commission staff excluded these issues from its environmental impact appraisal, submitted to the Commission in March 1981 and supplemented in May 1981, which recommended that no environmental impact statement be prepared in connection with the proposed restart of TMI-1.3 On September 17, 1981, after the appointment of a fifth Commissioner, Chairman Nunzio Palladino, the Commission adhered by a vote of 3-to-2 to its previous result. Chairman Palladino did not write an opinion or concur in any of the previous opinions.

PANE filed a petition for review of the Commission’s order, issued December 5, 1980, which excluded its psychological stress and community deterioration contentions from the TMI-1 restart proceeding. It sought reversal on the basis of the National *362Environmental Policy Act, 42 U.S.C. § 4321 et seq. (1976), and the Atomic Energy Act, 42 U.S.C. § 2133 (1976).

On January 7, 1982 this court issued an interim judgment, pending issuance of opinions, which ordered the Commission to prepare an environmental assessment of the effects of the proposed TMI-1 restart on the psychological health of neighboring residents and on the well-being of the surrounding communities. The judgment ordered the Commission to determine on the basis of this study whether to prepare a supplemental environmental impact statement. Until the Commission had complied with the requirements of NEPA, it was ordered not to make any decision to restart TMI-1. On the Atomic Energy Act question this court ordered the Commission to submit to the court a statement of its reasons for concluding that the statute did not require consideration of psychological health in the restart proceeding. Judge Wilkey dissented from the judgment. The Commission’s statement of reasons was filed with this court on March 30, 1982.

After further consideration of the NEPA issues, the court replaced the January 7, 1982 judgment with an amended judgment, entered on April 2, 1982.4 The amended judgment gave the Commission discretion to choose its procedures for studying the significance of the alleged psychological health impacts arising from the proposed restart of TMI-1. It made clear that the initial study should focus on psychological health effects. The Commission would be required to consider the secondary impacts on community well-being only if a full supplemental EIS was prepared. Finally, noting that the operators of TMI-1 had announced that extensive corrosion problems were likely to delay the restart by six to twelve months,5 the amended judgment lifted the injunction against restart as unnecessary to preserve the status quo. The court instructed the Commission, however, to give notice to the court and to petitioner if subsequently it intended to make a final decision regarding the restart of TMI — 1 prior to complying with its obligations under NEPA.

II. National Environmental Policy Act

The National Environmental Policy Act is designed to assure that governmental agencies take a “hard look” at the environmental consequences of major proposed actions, and that they adjust ongoing programs in light of new information or changed circumstances. PANE urges us to hold that NEPA requires the Commission to prepare a new or supplemental environmental impact statement (EIS) on the psychological health effects and community deterioration that might result from restart of TMI-1. We agree with PANE that these environmental effects fall within the scope of NEPA, and that the Commission has a continuing responsibility to comply with NEPA’s procedural requirements in its supervision of licensed nuclear facilities, including TMI-1. At the same time, we recognize the agency’s role in making a threshold determination of whether changed circumstances and new information regarding environmental effects require a supplemental EIS. We therefore remand the record to the Commission for a decision on the EIS question.

A. Cognizability of Psychological Health and Community Deterioration

PANE contends that NEPA requires the Commission to prepare a new or revised EIS to evaluate two distinct environmental effects of reopening TMI-1. First, PANE alleges that renewed operation of the nuclear reactor would cause “severe psychological distress” to persons living in the vicinity of the reactor, including PANE’S members. According to PANE, the accident at TMI — 2 *363created intense anxiety, tension, and fear, accompanied by physical disorders including skin rashes, aggravated ulcers, and skeletal and muscular problems. JA 84-86. Post-traumatic neurosis, PANE asserts, can be diagnosed with reasonable medical certainty on the basis of standardized quantitative tests. Petitioner’s brief at 46-47. Moreover, PANE argues, reopening TMI-1 would severely aggravate existing problems and would prevent Three Mile Island’s neighbors from resolving and recovering from the trauma they have suffered. JA 84-86.

Second, PANE contends that resumption of operations at TMI-1 would cause severe harm to the “stability, cohesiveness and well being of the communities in the vicinity of the reactor.” Id. In petitioner’s view, citizens have lost confidence in the ability of community institutions to function effectively during a crisis; therefore the renewed danger of nuclear accidents would impose great strains on the community infrastructure. Moreover, PANE asserts, restarting TMI-1 would perpetuate the area’s image as an undesirable location for residents and businesses, thus causing permanent damage to the economic and social health of the community.

Thus PANE’S first contention deals with individual health; its second addresses the social and economic impacts that perceived nuclear hazards might create in the communities in the vicinity of Three Mile Island. Both contentions allege environmental effects within the meaning of NEPA.

1. Potential damage to psychological health

The President’s Commission on the Accident at Three Mile Island reported that the “major health effect of the accident appears to have been on the mental health of the people living in the region of Three Mile Island and of the workers at TMI.” RePORT OF THE PRESIDENT’S COMMISSION on the Accident at Three Mile Island, The Need For Change: The Legacy Of TMI at 35 (Oct. 1979), JA 267. As the Nuclear Regulatory Commission’s staff has acknowledged, a great deal of study and attention has been devoted to attempts to measure the effects of the March 1979 accident at TMI-2 upon persons in the area, “including attempts to measure effects on mental health.” JA 177. The staff listed a number of separate studies, conducted by organizations including the Hershey Medical Center, the Pennsylvania Department of Health, the Western Psychiatric Institute of the University of Pittsburgh, and Central Pennsylvania Blue Shield, that considered the psychological effects of the Three Mile Island accident. Id.

Nevertheless, the Commission’s brief contends that the psychological effects alleged by PANE, which were caused by the TMI-2 accident and would assertedly be perpetuated by restart of TMI-1, are beyond the scope of NEPA. Commission’s brief at 50-55. This assertion is far-reaching. Regardless of the severity of psychological health effects, the position taken in the Commission’s brief would exclude them from consideration at any stage of the NEPA procedures relating to any proposed federal action. We find this interpretation of NEPA unpersuasive.6 The Commission’s brief ignores the simple fact that effects on psychological health are effects on the health of human beings.

In the National Environmental Policy Act, Congress accorded prominence to the effects of government actions on health and safety. NEPA was designed to “promote efforts which will prevent or eliminate damage to the environment and biosphere and stimulate the health and welfare of man.” 42 U.S.C. § 4321 (1976). The Act declared a national environmental policy of *364“encouragpng] productive and enjoyable harmony between man and his environment,” id., and explicitly recognized that each person “should enjoy a healthful environment,” id. § 4331(c).7 In its regulations implementing NEPA’s procedural requirements, the Council on Environmental Quality required agencies to consider “[t]he degree to which the proposed action affects public health and safety” as a factor in deciding whether a federal action “significantly” affected the human environment. 40 C.F.R. § 1508.27(b)(2) (1981). In short, “[n]o subject to be covered by an EIS can be more important than the potential effects of a federal program upon the health of human beings.” Citizens Against Toxic Sprays, Inc. v. Bergland, 428 F.Supp. 908, 927 (D.Or. 1977).8

We conclude that, in the context of NEPA, health encompasses psychological health. To implement a national policy based on “the critical importance of restoring and maintaining environmental quality to the overall welfare and development of man,” 42 U.S.C. § 4331(a) (1976), Congress required each federal agency to utilize a “systematic, interdisciplinary approach which will insure the integrated use of the natural and social sciences and the environmental design arts.” Id. § 4332(2)(A); see 40 C.F.R. §§ 1502.6, 1507.2 (1981); cf. Chelsea Neighborhood Ass’ns v. U. S. Postal Service, 516 F.2d 378, 388 (2d Cir. 1975) (social as well as physical sciences relevant under NEPA; agency must consider dangers of emotional and physical isolation of high-rise apartment building, which might as a result become a “human jungle”).

Although we are not aware of any cases that have considered the cognizability of post-traumatic psychological health effects under NEPA, it is not surprising that this is an issue of first impression. Americans have never before experienced the psychological aftermath of a major accident at a nuclear power plant, one that aroused fears of a nuclear core meltdown and led to mass evacuation from the surrounding communities. See Report op the President’s Commission on the Accident at Three Mile Island, supra, JA 257-271. PANE alleges that restarting TMI-1 would perpetuate the psychological health effects of the TMI-2 accident — intense anxiety, tension, and fear accompanied by physical disorders. Despite the sweeping language of Judge Wilkey’s dissent, PANE is not seeking to extend NEPA to “mere ‘anxieties.’ ” Wilkey dissent at 241.

Nevertheless, the Commission’s brief contends that psychological distress is beyond the scope of NEPA because it is not readily quantifiable. Commission’s brief at 51-52. The Commission’s staff was unable to state “with any degree of certainty whether the psychic distress associated with continued operation of the TMI1 facility is sufficiently susceptible of measurement to permit a meaningful assessment of the phenomenon.” 11 NRC at 305, JA 71. On the other hand, the Licensing Board asserted that *365psychological factors were sufficiently quantifiable to be considered, noting that “some quantification of stress upon the community is being undertaken by responsible organizations.” 11 NRC at 302, JA 68. NEPA, moreover, does not authorize federal agencies to deal with intangible factors by ignoring them. It expressly instructs all federal agencies to identify and develop methods and procedures “which will insure that presently unquantified environmental amenities and values may be given appropriate consideration in decisionmaking along with economic and technical considerations.” 42 U.S.C. § 4332(2)(B) (1976).9 This expression of congressional purpose led the Commission’s Licensing Board to conclude, correctly, that “[pjrecise numerical quantification is not necessary” under NEPA. 11 NRC at 302, JA 68.

To support its position that “psychological distress” need not be considered at all in the NEPA process, the Commission’s brief relies on cases that rejected the cognizability of sociologically based community anxieties. Commission’s brief at 50 — 55. In these cases neighborhood associations, businesses, or other groups unsuccessfully sought to use NEPA to block or delay proposed construction of government projects — low-income housing, federal detention centers, Job Corps centers, postal service facilities— primarily because they were afraid the projects would change the character of the neighborhood, reduce property values, and increase the dangers of crime. See, e.g., Como-Falcon Community Coalition, Inc. v. U. S. Dep’t of Labor, 609 F.2d 342, 345-346 (8th Cir. 1979), cert. denied, 446 U.S. 936, 100 S.Ct. 2154, 64 L.Ed.2d 789 (1980) (Job Corps center); Nucleus of Chicago Homeowners Ass’n v. Lynn, 524 F.2d 225, 231 (7th Cir. 1975), cert. denied, 424 U.S. 967, 96 S.Ct. 1462, 47 L.Ed.2d 734 (1976) (low-rent housing for low-income families); Maryland-Nat’l Capital Park & Planning Comm’n v. U. S. Postal Service, 487 F.2d 1029, 1037 (D.C.Cir. 1973) (bulk mail postal facility in suburban area); First Nat’l Bank of Chicago v. Richardson, 484 F.2d 1369, 1380 n.13 (7th Cir. 1973) (federal parking garage and detention center in downtown area); Hanly v. Kleindienst, 471 F.2d 823, 833 & n.10 (2d Cir. 1972), cert. denied, 412 U.S. 908, 93 S.Ct. 2290, 36 L.Ed.2d 974 (1973) (detention center in downtown area not far from residential apartments). None of these cases, of course, presents the. holocaust potential of an errant nuclear reactor.

In these and other cases federal courts have consistently rejected the contention that socioeconomic anxieties are environmental impacts within the meaning of NEPA. The agency fulfills its responsibilities under NEPA in this context if it considers and mitigates the underlying causes for alarm, such as the possibility of increased noise, increased crime, and increased congestion. “Concerned persons might fashion a claim, supported by linguistics and etymology, that there is an impact from people pollution on ‘environment,’ if the term be stretched to its maximum,” Judge Leventhal explained. “We think this type of effect cannot fairly be projected as having been within the contemplation of Congress.” Maryland-Nat’l Capital Park & Planning Comm’n v. U. S. Postal Service, supra, 487 F.2d at 1037; see Nucleus of Chicago Homeowners Ass’n v. Lynn, supra, 524 F.2d at 231.

In this case, in contrast, PANE is not asking the agency to evaluate the effect of “people pollution” on the environment, but rather the effect of a governmental decision on human health. We conclude that PANE’S allegation — in the wake of a unique and traumatic nuclear accident— that renewed operation of TMI-1 may cause medically recognized impairment of the psychological health of neighboring residents is cognizable under NEPA.

The key to our decision is the potential effect on health. Not all physical effects have an impact on physical health; similar*366ly, not all psychological effects rise to the level of psychological health effects. In our view, Congress intended to include psychological health within the meaning of “health” for purposes of NEPA. NEPA does not encompass mere dissatisfactions arising from social opinions, economic concerns, or political disagreements with agency policies.10 It does apply to post-traumatic anxieties, accompanied by physical effects and caused by fears of recurring catastrophe. Therefore, the severity of a psychological effect is not only relevant to whether an EIS is required under NEPA, as Judge Wilkey concedes, Wilkey dissent at 242, but also to the cognizability of the impact under the statute.

We need not attempt to draw a bright line in this case. Three Mile Island is, at least so far, the only event of its kind in the American experience. We cannot believe that the psychological aftermath of the March 1979 accident falls outside the broad scope of the National Environmental Policy Act.

2. Possible deterioration of the community

PANE’S second contention alleges that the communities surrounding Three Mile Island would be severely damaged by the proposed restart of the TMI-1 facility because fears of nuclear accidents will diminish citizen confidence in local institutions, cause local businesses and residents to leave the area, and discourage potential newcomers who perceive the area as an undesirable location. JA 85-86. The Commission concedes that this contention presents a “classical ‘socio-economic’ issue.” Commission’s brief at 49. Social and economic effects, also described as “secondary impacts,” do not by themselves require preparation of an environmental impact statement. 40 C.F.R. § 1508.04 (1981) (mandatory Council on Environmental Quality regulations).11 However, when an environmental impact statement is prepared, it must discuss economic or social effects that are interrelated with other environmental effects. Id. Deterioration of a community’s economic base or social stability, as alleged in PANE’S second contention, is a cognizable “secondary impact” under NEPA. See, e.g., City of Rochester v. U. S. Postal Service, 541 F.2d 967, 973 (2d Cir. 1976) (danger of economic and physical deterioration in downtown area, urban decay and blight); Trinity Episcopal School Corp. v. Romney, 523 F.2d 88, 93-94 (2d Cir. 1975) (displacement and relocation of residents, decay and blight, implications for city growth policy and neighborhood stability). If NEPA requires the Commission to prepare a supplemental EIS regarding the TMI — 1 restart decision because the agency makes a threshold finding of significant new information on psychological health effects, see Part II-C infra, PANE’S contentions regarding secondary effects on *367the community must be evaluated in the supplemental EIS.

B. Applicability of NEPA to the TMI-1 Restart Decision

PANE contends that the March 1979 accident at TMI-2 significantly changed the psychological and socio-economic effects of operating TMI-1. Therefore, PANE argues, the Commission must comply with NEPA before it decides whether to authorize restart of TMI-l’s operations. This assertion does not depend on the happenstance that TMI-1 was shut down for refueling at the time of the accident. PANE relies more generally on the continuing close supervision that the Commission exercises over nuclear power plants under the Atomic Energy Act. We agree with PANE that the extent of the Commission’s statutory responsibilities over licensed nuclear facilities creates a continuing obligation to comply with NEPA.12

The Commission’s brief contends that its pending decision on whether to allow resumption of operations at TMI-1 is not a “major federal action” within the National Environmental Policy Act and is therefore not subject to NEPA’s requirements. Conceding that the initial grant of an operating license requires preparation of an EIS, the brief asserts that, once a private activity such as a nuclear reactor has been licensed, federal involvement in its continuation is “limited and discontinuous” and therefore “lacks the elements of federal purpose and discretion generally associated with the requirement for impact statements.” Commission’s brief at 46. This position takes too narrow a view of the relevant federal activity. The “major federal action” in the case of TMI-1 is not solely the initial licensing decision, but the Commission’s continued exercise of supervisory responsibility over its operation and maintenance.

The position argued in the Commission’s brief is inconsistent with binding regulations promulgated by the Council on Environmental Quality (CEQ) and with previous judicial decisions defining “major federal actions” for purposes of NEPA. The CEQ regulations, applicable to all federal agencies including the Commission, 40 C.F.R. § 1500.3 (1981), were expressly designed to establish uniform procedures for implementing NEPA and to eliminate inconsistent agency interpretations. 43 Fed.Reg. 55978 (1978); see Andrus v. Sierra Club, 442 U.S. 347, 356-357, 99 S.Ct. 2335, 2340-41, 60 L.Ed.2d 943 (1979).13 “Federal action,” under the regulations, encompasses “new and continuing activities, including projects and programs entirely or partly financed, assisted, conducted, regulated, or approved by federal agencies.” 40 C.F.R. § 1508.18(a) (1981).14

The Commission’s NEPA responsibilities did not come to an end when it prepared an initial EIS; the “continuing activity” of regulating TMI-1 is federal action within the scope of NEPA. The Commission has an ongoing responsibility to assure that nuclear power plants will operate without endangering the health and safety of the public. 42 U.S.C. §§ 2012(e), 2201(b), 2236 (1976). It maintains a resident inspector at each nuclear facility, see 10 C.F.R. § 50.-70(b) (1981), and operates a licensing program for nuclear power plant operators, see id. §§ 55.1-55.60.

*368In the immediate aftermath of the nuclear accident at TMI-2 the Commission ordered Metropolitan Edison, the licensee, to keep TMI-1 in a cold shutdown condition pending further order by the Commission, and stated that a hearing would be held before the reactor would be authorized to resume operation. The order explains that the agency “presently lacks the requisite reasonable assurance that the same licensee’s Three Mile Island Unit No. 1 facility * * * can be operated without endangering the health and safety of the public,” 44 Fed.Reg. 40461 (1979), language that echoes the Atomic Energy Act. Pursuant to its July 1979 order, the Commission received written and oral testimony at an evidentiary hearing from witnesses presented by the licensee, the Commission’s staff, the Commonwealth of Pennsylvania, and five intervenors. Atomic Safety and Licensing Board, Partial Initial Decision (Procedural Background and Management Issues) at 12 (Aug. 27,1981). The record of the proceeding covers more than 22,000 transcript pages. Memorandum on the Status of the Three Mile Island Unit 1 Restart Proceeding, filed by the Commission with this court on November 25, 1981. These regulatory activities fall squarely within the language of the CEQ regulation defining “federal action.” 40 C.F.R. § 1508.18(a) (1981) (quoted supra).

Judicial decisions antedating the CEQ regulations defined “federal action” similarly in concrete factual contexts. The central issue in determining the applicability of NEPA to federally assisted or federally regulated projects was whether agency decisions were yet to be made, and whether decisions, “although already made, remained] open to revision.” Jones v. Lynn, 477 F.2d 885, 890 (1st Cir. 1973). In Jones v. Lynn the basic loan and capital grant contract for an urban renewal project had been executed before NEPA entered into effect, but the court held that NEPA procedures must be followed as long as the federal agency “remains meaningfully involved in a project” and has “retained any significant discretionary powers.” Id. at 889-890 (remanding for findings by District Court).15 In WATCH (Waterbury Action to Conserve Our Heritage Inc.) v. Harris, 603 F.2d 310, 317-318 (2d Cir.), cert. denied, 444 U.S. 995, 100 S.Ct. 530, 62 L.Ed.2d 426 (1979), the federal Department of Housing and Urban Development had authorized demolition of buildings within a specified urban renewal area but retained power to veto specific actions by local authorities. When HUD received new information about the potential historic value of structures within the demolition area, it imposed a temporary freeze on further acquisitions and demolitions, gathered some data about the structures, and then authorized demolition to continue. A local organization sought judicial review of the agency’s failure to follow NEPA procedures before deciding whether to continue with the demolition project. The Second Circuit, recognizing that HUD retained “significant control over the project,” and that the agency had recognized its continuing responsibility by imposing a freeze, held that HUD was required to comply with NEPA. 603 F.2d at 318, 326.

The Commission’s brief cannot convincingly distinguish these precedents by asserting that, unlike this case, they involved “a more or less continuing agency decision to sustain an ongoing program that the agency has discretion to terminate at any time.” Commission’s brief at 46. The Commission’s regulatory responsibilities with regard to TMI-1 and other nuclear reactors place it in an analogous position; therefore, it must continue to comply with NEPA’s requirements.

*369C. The Commission’s Responsibilities Under NEPA

If the agency’s “continuing activities” are within the scope of NEPA, the CEQ regulations require it to prepare a supplemental environmental impact statement in two situations: (1) if the agency makes substantial changes in the proposed action that are relevant to environmental concerns, or (2) if there are significant new circumstances or information relevant to environmental concerns and bearing on the proposed action or its impacts. 40 C.F.R. § 1502.9(c)(1) (1981).16 Supplemental impact statements shall be prepared, circulated, and filed in the same fashion as draft or final environmental impact statements. Id. § 1502.-9(c)(4). These regulations implement the broad purpose of NEPA: to require the federal government to assume “continuing responsibility” to promote environmental values. See 42 U.S.C. § 4381(b) (1976). NEPA itself does not expressly provide for supplemental impact statements, but the Supreme Court has declared that the CEQ implementing regulations are “entitled to substantial deference” as an interpretation of NEPA. Andrus v. Sierra Club, supra, 442 U.S. at 358, 99 S.Ct. at 2341.17

In this case PANE contends that the accident at TMI-2 in March 1979 was a “significant new circumstance” that dramatically altered the environmental effects of operating TMI-1. It cites not only the CEQ regulations, but earlier judicial decisions that required preparation of a supplemental EIS to take changed circumstances and new information into account. In Essex County Preservation Ass’n v. Campbell, 536 F.2d 956, 961 (1st Cir. 1981), relying on federal highway regulations, the court required a supplemental EIS for a proposed highway project after a state moratorium on widening of a feeder highway reduced the potential flow of traffic to the proposed highway segment. In several District Court cases, discovery of archeological sites within an affected area has been held to require a supplemental EIS. Libby Rod & Gun Club v. Poteat, 457 F.Supp. 1177, 1188-1189 (D.Mont. 1978) (Army Corps of Engineers dam project); Aluli v. Brown, 437 F.Supp. 602, 606 (D.Hawaii 1977), rev’d in part on different issue, 602 F.2d 876 (9th Cir. 1979) (Navy bombing practice on Hawaiian island); Nelson v. Butz, 377 F.Supp. 819, 822 (D.Minn. 1974) (proposed flooding of area by dam construction).18

At this stage, PANE’S allegations of psychological health effects and community deterioration do not justify an order to the Commission to prepare a supplemental EIS. It is well established that, under NEPA, the agency in charge of a proposed federal action is authorized to make the threshold determination of whether an EIS is required. See WATCH v. Harris, supra, 603 F.2d at 317-318, 326; Asphalt Roofing Manufacturers Ass’n v. ICC, 567 F.2d 994, 1004 (D.C.Cir. 1977); Hanly v. Kleindienst, *370supra, 471 F.2d at 828. In this case, however, the Commission staff did not consider psychological health effects in its environmental impact appraisal, JA 272-316, nor did it receive any evidence in support of PANE’S contentions. When the Commissioners excluded psychological stress from the TMI-1 restart proceeding, none of the four separate opinions specifically addressed NEPA’s requirements for issuance of a supplemental EIS, nor did any Commissioner evaluate any of the existing studies regarding psychological and community effects. See 12 NRC at 609-626, JA 3-20.19 Under the circumstances, we remand the record to the Commission to determine whether to prepare a supplemental EIS.

If the agency finds significant new circumstances or information on psychological health effects, the CEQ regulations require it to prepare a supplemental EIS regarding both psychological health effects and secondary impacts on the well-being of surrounding communities. 40 C.F.R. § 1502.-9(c)(1) (1981). On the other hand, if the agency finds that the circumstances or information are not new, or not significant, it need not prepare a supplemental EIS. The agency’s determination, if appealed, will be upheld as long as it is reasonable — the same standard of judicial review that we apply to an agency’s determination not to issue an EIS in the first instance. See Izaak Walton League of America v. Marsh, 655 F.2d 346, 371 (D.C.Cir. 1981) (“rule of reason” in reviewing compliance with NEPA); cf. Environmental Defense Fund v. Marsh, 651 F.2d 983, 991-992 (5th Cir. 1981) (same standard for review of agency decision not to issue EIS or decision not to issue supplemental EIS); Monarch Chemical Works, Inc. v. Thone, 604 F.2d 1083, 1087 (8th Cir. 1979) (same).

We agree with the Ninth Circuit’s recent discussion of the respective roles of agencies and courts and the “rule of reason” as applied to an agency’s decision not to issue a supplemental EIS:

When new information comes to light the agency must consider it, evaluate it, and make a reasoned determination whether it is of such significance as to require implementation of formal NEPA filing procedures. Reasonableness depends on such factors as the environmental significance of the new information, the probable accuracy of the information, the degree of care with which the agency considered the information and evaluated its impact, and the degree to which the agency supported its decision not to supplement with a statement of explanation or additional data.

Warm Springs Dam Task Force v. Gribble, 621 F.2d 1017, 1024 (9th Cir. 1980).20

Until a federal agency has fulfilled its obligations under NEPA, it should generally not proceed with its ultimate decision on whether to proceed with a proposed action. *371NEPA procedures are designed to ensure that environmental information is available to public officials and citizens before decisions are made, so that environmental considerations are part of the agency’s decisionmaking process. See Weinberger v. Catholic Action of Hawaii/Peace Education Project, 454 U.S. 139, 140-47, 102 S.Ct. 197, 200-03, 70 L.Ed.2d 298 (1981); Realty Income Trust v. Eckerd, 564 F.2d 447, 456 (D.C.Cir. 1977); 40 C.F.R. §§ 1500.1(b), 1502.2(g) (1981). Under appropriate circumstances, federal courts will grant injunctive relief to preserve the status quo until the requirements of NEPA have been satisfied. See Realty Income Trust v. Eckerd, supra, 564 F.2d at 456—457.

In this case the court issued an order on January 7, 1982 that the Commission not make a decision to restart TMI-1 until it had complied with its obligations under NEPA. Subsequently, however, operating officials at Three Mile Island announced that leaks and corrosion in thousands of steam generator tubes in TMI — 1 would probably delay the restart for six to twelve months. See Corrosion Leads To More Delays At 3 Mile Island, New York Times, February 11, 1982, A18, at col. 1. Recognizing that injunctive relief was no longer necessary to preserve the status quo, the court on April 2, 1982 issued an amended judgment lifting the injunction. The amended judgment, which remains in effect, requires the Commission to give 30 days’ notice to this court and to petitioner if the Commission intends to make a decision on restarting TMI — 1 before complying with its obligations under NEPA. When such notice is given, it will be time enough for this court to decide whether an injunction should issue to enforce the mandate of the National Environmental Policy Act.

In the wake of the most publicized nuclear accident of our time, the people of the Three Mile Island area — and the people of the nation as a whole — are entitled to the protections Congress provided in the National Environmental Policy Act. The government must not proceed to make decisions that might have a momentous effect on the psychological health and community well-being of its citizens without first giving careful, responsible consideration to the consequences its actions might have. By enacting NEPA Congress meant to assure that no federal decision — especially one of this importance — would be made in the shadow of environmental ignorance.

III. CONCLUSION

We have concluded that psychological health is cognizable under NEPA and that the Commission’s statutory responsibilities over licensed nuclear facilities create a continuing obligation to comply with the requirements of the statute. We therefore remand the record in this case to the Commission for study of potential psychological health effects and for a decision whether a supplemental EIS is necessary.

So ordered.

AMENDED JUDGMENT

This cause came on to be heard on a petition for review of an order of the United States Nuclear Regulatory Commission and was briefed and argued by counsel. A judgment was issued on January 7, 1982, Judge Wilkey dissenting. In light of changed circumstances and further consideration, this court has decided to modify its order to the Commission. On consideration thereof,

It is ORDERED and ADJUDGED by this court that, for the reasons stated in the opinion for the court issued this day, the record in this case is remanded to the Commission for a determination whether, since the preparation of the original environmental impact statement for the nuclear facility at Three Mile Island Unit 1 (TMI-1), significant new circumstances or information have arisen with respect to the potential psychological health effects of operating the TMI-1 facility. The Commission may choose the procedures by which it makes this determination. If the Commission finds that such significant circumstances or information exist, it shall prepare a supplemental environmental impact statement *372which considers not only effects on psychological health but also effects on the well-being of the communities surrounding Three Mile Island.

It is FURTHER ORDERED and ADJUDGED by this court that, in light of the current operating difficulties at TMI-1, it is no longer necessary to preserve the status quo to enjoin the Commission from deciding to restart TMI-1 until it has complied with the requirements of the National Environmental Policy Act (NEPA). The injunction granted on January 7,1982 is hereby vacated. If subsequently the Commission intends to make a final decision regarding the restart of TMI-1 prior to complying with its obligations under NEPA, it shall provide the court and the petitioner with 30 days’ notice thereof.

Circuit Judge WILKEY dissents for the reasons stated in his dissenting opinion filed this day.

This opinion has two distinct parts. Part I is my dissent from my two colleagues’ decision on the applicability of the National Environmental Policy Act (NEPA)1 to this case.2 Part II has been joined by Judge McGowan, and represents the opinion of the court on the applicability of the Atomic Energy Act (AEA)3 to this case. The net result is that the court holds that NEPA requires consideration of alleged psychological health effects, while the AEA does not.

This may have the appearance of a split decision, but the reality is otherwise. The critical issue is NEPA, and the court’s determination produces an extraordinary result. Judge Wright and Judge McGowan hold that in the proceedings on the restart of Three Mile Island Unit 1 (TMI-1), which was not involved in the accident at Three Mile Island Unit 2 (TMI-2), the Nuclear Regulatory Commission (NRC) must consider “the potential psychological health effects of renewed operation of TMI — l.”4 This requires consideration of an “impact” on health — psychological stress — which has never before been held cognizable under NEPA. A similar decision under the AEA would have compounded the problem, but the NEPA decision today suffices to give petitioner People Against Nuclear Energy (PANE) essentially what it has sought: a court-imposed paralysis of nuclear power at Three Mile Island, and potentially elsewhere as well. Thus although I am pleased that my view on the AEA issue has prevailed, I have no illusion that I am anything other than the chief dissenter in this case.

It is worth noting, and perhaps taking solace in, the majority’s5 partial retreat from the judgment it so hastily issued on 7 January 1982.6 The injunction against

*373TMI-l’s restart has been lifted,7 and, in addition, the majority has corrected two clear errors of NEPA law contained in its original judgment.8

Unfortunately, the basic error remains, The extension of NEPA to encompass psychological stress is unwarranted, unprecedented, and inconsistent with relevant decisions in this and other circuits. This novel hurdle, well designed to delay the development of nuclear power (contrary to the national policy determined by Congress and the Executive), is thoroughly consistent with this court’s track record of using *374NEPA to delay the development of important energy sources.9 I dissent.

I. NATIONAL ENVIRONMENTAL POLICY ACT

A. Cognizability of Psychological Stress Under NEPA10

1. Meaning of “health” in NEPA

There is no question that NEPA’s requirements extend to effects on human health. Two of the Act’s goals are to “assure for all Americans safe, healthful, productive, and esthetically and culturally pleasing surroundings,”11 and to “attain the widest range of beneficial uses of the environment without degradation, risk to health or safety, or other undesirable and unintended consequences.” 12 What is in question, of course, is what Congress intended by its inclusion of health as a concern. Amazingly, the majority does not find this even to be a hard question. It holds “that, in the context of NEPA, health encompasses psychological health,”13 a conclusion that rests on “the simple fact that effects on psychological health are effects on the health of human beings.”14 Since petitioner PANE alleges that the restart of TMI-1 would cause “severe psychological distress” to nearby residents, the majority orders the NRC to consider this allegation under NEPA. This holding is entirely novel, and indeed is contrary to the most closely analogous precedents. In my view it extends the reach of NEPA far beyond its intended scope.

Judge Wright’s opinion cites several cases holding that agencies must prepare an EIS when there is a potential effect on human health. What the opinion does not acknowledge is that in each of these cases the effect on health was caused by the federal action itself, not by individuals’ fears of the federal action. Use of toxic herbicides has a potential for damaging human health, and consideration must therefore be given to these potential effects.15 An allegedly inadequate water run-off system may lead to flooding which endangers human health, so this possibility must be considered.16

In the same way, operation of a nuclear power plant may cause harm to human health — for example, due to the potential for exposure to radiation — and the NRC must therefore prepare an EIS and consider these potential harms before licensing the plant. In this case, however, the NRC already has prepared a full EIS on TMI-1, as well as an environmental appraisal relating *375to restart, to facilitate decisionmaking and minimize the damage that could result from the plant’s operation. This undertaking is what NEPA clearly contemplates.

PANE’s contention, however, is not that operation of TMI-1 will affect human health because of the dangers inherent in operation of a nuclear facility, but that individuals’ fears of an accident at the plant, combined with their lack of confidence in the NRC, will lead to an extension of the psychological stress allegedly caused by the TMI-2 accident. It is patently obvious that this alleged effect is entirely different from those health effects at issue in any NEPA case relied on by the majority. Instead of being required to assess the risk of a proposed activity in determining whether the activity should go forward, the agency is now required to assess how people perceive and react to the risk. PANE’s primary purpose is to force the agency to determine whether people so fear renewed operation of TMI-1 that it should not go forward, even if the agency’s assessment of the actual risk indicates that the impact on health will not be significant.

This takes NEPA far beyond its intended purpose.17 The environmental effects of a federal activity are now to include the views of the population itself on the very desirability of the activity, as expressed through the alleged psychological distress people may suffer if the activity goes forward — no matter how scientifically ignorant and divorced from reality those views (fears) may be. In my view this is a judgment for Congress, and one which has already been made in the case of nuclear power. “Nuclear energy may some day be a cheap, safe source of power or it may not. But Congress has made a choice to at least try nuclear energy ....”18 To adopt the majority view would be to let any special interest group effectively repeal an act of Congress if it could whip up sufficient hysteria.

2. Case law on psychological factors under NEPA

Many federal courts have agreed that individuals’ psychological reactions to a federal action are not properly considered under NEPA. For example, the Second Circuit has declared: “It is doubtful whether *376psychological and sociological effects upon neighbors constitute the type of factors that may be considered in making such a determination since they do not lend themselves to measurement.”19 Similarly, the Seventh Circuit has stated: “To the extent that this claim can be construed to mean that HUD must consider the fears of the neighbors of prospective public housing tenants, we seriously question whether such an impact is cognizable under NEPA.”20 Many other courts have agreed,21 and these cases represent a clear consensus against consideration of psychological factors.

The majority tries simply to cite and then dismiss these cases as irrelevant, but their force cannot be ignored. They are based largely on the fact that psychological concerns are simply too far removed from the purpose of NEPA, which is to ensure that an agency considers the environmental effects of a decision, not the reactions of affected individuals to the risk of those environmental effects. They also rest on the inherent difficulties in attempting to measure and incorporate into the environmental analysis the differing psychological states of the affected persons.

The majority blithely dismisses this quantification problem as irrelevant,22 thus ignoring Judge Leventhal’s observation that although measurement difficulties do not necessarily prevent consideration under NEPA, they do “have a bearing on the intention of Congress, and whether it contemplated ... a requirement of a detailed [EIS].”23 In this case petitioner makes much of the practice of courts in measuring psychological injury. This is of dubious validity given that the assignment of monetary damages for purposes of liability is different from the measurements involved here.

Far more important, however, is that the issue before the NRC will be not how much damage was caused by the TMI-2 accident, but how much additional damage will result from TMI-l’s restart. This issue is particularly ephemeral and speculative since it is not subject to measurement at all. Instead, the Commission will be forced to predict how every individual in the TMI area will react to the restart. Moreover, to the extent the Commission is supposed to devise techniques to alleviate the stress, it will need to guess at how much benefit will accrue from, say, warning system X as compared to warning system Y.

To attempt to assess the effect of TMI-l’s operation on the psychological condition of area residents will, in my view, demonstrate the truth in the Second Circuit’s finding that “psychological factors are not readily translatable into concrete measuring *377rods.”24 I do not believe that Congress intended NEPA to encompass an effect which not only varies from individual to individual, but which is also entirely subjective. All the other federal courts which have considered this issue agree.

The majority obviously recognizes that these NEPA decisions are far more relevant than any others to this case. It attempts to distinguish them, however, by asserting that they deal with “sociologically based community anxieties” and “mere dissatisfactions arising from social opinions, economic concerns, or political disagreements with agency policies,” which supposedly are easily distinguishable from the potential “medically-recognized impairment of the psychological health of neighboring residents” of TMI.25 This purported distinction is destroyed by the majority’s own finding that it is a “simple fact that effects on psychological health are effects on the health of human beings.”26 The assertion that mere “anxieties” about nearby matters other than nuclear power are not effects on psychological health is entirely unsupported and, I submit, obviously unsupportable. It is a callous assumption indeed to believe that persons living close to a prison or in a high-crime area cannot suffer very real psychological harm from fear of physical violence — and highly illogical, also, since the casualty total from crime is a gruesome, proven fact, while the casualty total from nuclear accidents so far, fortunately, remains at zero.

The majority’s decision to ignore these many cases seems based ultimately on nothing more than a political determination that fears that federal actions will “change the character of the neighborhood, reduce property values, and increase the dangers of crime”27 are simply not worthy of consideration. Fears of nuclear power, on the other hand, must be considered, presumably because the majority considers them legitimate. In each case, however, what we are dealing with are fears and anxieties; if NEPA embraces fears and anxieties in one, it must in all.

In response, the majority declares that it is not attempting to extend NEPA to “mere anxieties,”28 because this case involves “post-traumatic anxieties, accompanied by physical effects and caused by fears of recurring catastrophe.”29 Yet this very passage frames the harm precisely in terms of anxiety and fear. All that it adds are the alleged physical effects accompanying psychological stress. Why physical effects should be determinative of the outcome the majority does not say. If this is critical, then petitioner surely must fail; the physical harms alleged are, standing alone, rather minor. Moreover, I see no basis, as either a legal or scientific matter, for drawing the majority’s distinction. I am confident that psychologists would not accept the view that psychological stress unaccompanied by physical symptoms is therefore medically unrecognizable or necessarily much less severe than stress that does have related physical effects.30

Moreover, the majority’s emphasis on the notion that “post-traumatic” psychological harm is recognized by the medical profession is unconvincing. Surely “post-trau*378matic” stress is but one of numerous forms of psychological stress that are “medically recognizable.” There is no legal or logical justification for interpreting NEPA’s “health” coverage to extend only to such stress as may arise from a traumatic event.31 The majority seems to rely on PANE’S allegation of severe psychological stress, as well as its own view that this accident was especially “unique and traumatic.” 32 But the fact that post-traumatic stress may be more severe than other psychological harms is irrelevant to whether psychological harms are cognizable under NEPA. The severity of an effect is relevant to the “significance” it has under NEPA, and thus to whether an EIS must be prepared.33 Before this inquiry need be undertaken, however, there is the preliminary question at issue here: is the alleged effect cognizable at all under NEPA as a “primary impact” which may require an EIS? This inquiry does not depend on the intensity of a particular effect.

This point is disputed by the majority: “[T]he severity of a psychological effect is not only relevant to whether an EIS is required under NEPA, ... but also to the cognizability of the impact under the statute.”34 Why is this so? No answer is given. The majority simply asserts that some psychological effects are worse than others, and therefore the former are cognizable while the latter are not. But this is not true of any other type of health effect, and there is utterly no support in NEPA for this distinction.35 The majority has simply set itself up as the arbiter of what harms are severe enough to deserve NEPA protection. Thus what constitutes a “real” and “justifiable” fear, as opposed to what constitutes only a “social” or “economic” or “political” fear, will be determined by the courts.

And the choice itself will be totally arbitrary: NEPA protection will depend less on how much psychological harm the individual suffers, than on whether the judges of this court believe the source of the psychological harm is acceptable. If one fears living near a prison following a violent escape, NEPA might provide no protection because fear of a public project is merely social; if one fears living near a nuclear power plant following an accident which threatened to cause harm, NEPA will provide protection because, as we all know, nuclear power is potentially dangerous and we should all fear it.

3. Implications of the majority’s reasoning for nuclear power

The majority’s attempt to base its decision on the singularity of the TMI-2 accident, and thus perhaps to avoid the implications of the logic employed, also fails to explain why all nuclear power licensing decisions in the future will not need to include consideration of psychological stress. One key factor the majority sees in distinguishing this case from the “sociological anxiety” cases is that none of those cases involved “the holocaust potential of an errant nuclear reactor.” 36 But this rationale obviously *379applies to all nuclear facilities, and licensing is their sine qua non.

More generally, the court’s emphasis on the “major” and “unique and traumatic” nature of the TMI-2 accident does not explain why the type of psychological harm resulting from this accident is cognizable under NEPA while other forms of psychological stress associated with nuclear energy might not be. This accident killed no one and caused no detectable physical harm.37 Thus the “unique and traumatic” circumstances must be found elsewhere, perhaps in the fact that the TMI-2 accident “aroused fears of a nuclear core meltdown and led to mass evacuation from the surrounding communities.”38 Both factors might exist, however, even where an accident objectively is not a “major” one at all. Indeed, even in this case it is clear that some of the dangers were greatly exaggerated; most of the evacuation that took place was voluntary rather than officially requested or ordered.39 To the extent any consistent standard can be derived from the majority’s analysis, what appears is a standard which will depend largely on how much fear is worked up, from whatever source, rather than how serious the danger actually is.40

In any event, the majority fails even to try to explain why fears resulting from lesser accidents, or simply from the operation of nuclear facilities, do not also engender psychological stress in individuals. The susceptibility of individuals to psychological stress may vary widely. There may be a considerable number of persons who suffered more stress from having read about TMI-2’s accident than some persons who were in the area at the time. A traumatic event is not a prerequisite to experiencing psychological health problems. And what *380constitutes a “traumatic” event may differ greatly among individuals. The mere sound of a warning siren at a nuclear plant may cause a great deal of stress to certain individuals, yet the majority apparently would find such stress not cognizable under NEPA because a warning siren is not, in the majority’s view, sufficiently “unique or traumatic.”

Most fundamentally, the majority never explains why the prerequisite to NEPA consideration of psychological harm is the existence of some level of stress caused by a nuclear accident. This is true of no other effect under NEPA. Consideration of the potential for harm from exposure to radiation is not postponed until actual exposure takes place; it is the potential harm that is to be considered. If, as PANE alleges, the TMI-2 accident caused severe psychological harm, then any nuclear facility has the potential for “causing” such harm. NEPA consideration therefore should be mandatory in all licensing decisions, if psychological stress is cognizable at all.

This is the result the majority opinion’s rationale inexorably demands. If it is a “simple fact that effects on psychological health are effects on the health of human beings,"41 and if such effects are cognizable under NEPA, there is no reason why only “post-traumatic” forms of psychological stress must be considered. This logic is obvious enough to opponents of nuclear power that they have responded to this court’s 7 January judgment by filing psychological stress contentions in ordinary NRC licensing and construction permit proceedings.42 The breadth of the extension of NEPA that acceptance of these contentions would entail is what obviously leads the majority to take refuge in the “unique and traumatic” nature of the TMI-2 accident and, allegedly, its psychological aftermath. In the end, this attempted limitation is compelling evidence of the majority’s own doubts about the validity of its analysis and own realization of how far NEPA is being stretched in order to support today’s result.

B. Need for a Supplemental EIS

Regulations promulgated by the Council on Environmental Quality (CEQ) require preparation of a supplemental EIS when, inter alia, “[tjhere are significant new circumstances or information relevant to environmental concerns and bearing on the proposed action or its impacts.”43 A prerequisite to application of this provision, obviously, is the existence of a “proposed action.” Taking the broadest conceivable view of this issue, the majority concludes that “the ‘continuing activity’ of regulating TMI-1 is federal action within the scope of NEPA.”44 This means that the fact that in this case the NRC is proposing to restart the plant is irrelevant. The majority concedes as much, stating that PANE’S claim that NEPA must be followed “does not depend on the happenstance that TMI-1 was shut down for refueling at the time of the accident.”45

This is a holding of considerable breadth. Preparation of an EIS or a supplemental EIS is required only for proposed actions. Yet under the majority’s interpretation, the NRC is engaged in such “action” every, second of every day. It thus will be possible for NEPA to apply even when a nuclear plant is operating pursuant to an NRC license and the NRC proposes to take no action to upset this status quo.

The majority defends its interpretation by pointing to cases which have held that continuing federal involvement in a project meant that NEPA remained applicable, and also to the CEQ definition of “federal action” as “new and continuing activities, in-*381eluding projects and programs entirely or partly financed, assisted, conducted, regulated, or approved by federal agencies.”46 The majority overlooks, however, the fact that each case it relies on involved a “proposed action”: projects had been approved but were as yet not undertaken or were incomplete.47 Although in this case the “happenstance” of TMI-1 being shut down does put the Commission in the position of proposing an “action” (restart), the majority’s holding may significantly increase the NEPA burden on regulatory agencies in the future.48

In its 7 January judgment, the majority compounded this unnecessarily broad construction of the supplemental EIS requirement by ordering the NRC to conduct an “environmental assessment” and by ordering that the assessment extend to the socioeconomic effects alleged by PANE.49 Both of these orders were clearly wrong. The majority has not expressly repudiated those

portions of the original judgment, but its 2 April amended judgment and its opinion today indicate otherwise. To eliminate any possible uncertainty on the part of the agency, I will briefly discuss these two points.

An environmental assessment is a procedure required by CEQ regulations to be used when an agency must decide whether to prepare an EIS.50 This requirement does not extend to supplemental EIS’s. The majority’s first judgment required the NRC to conduct an environmental assessment in this case. Today’s opinion, however, leaves this procedural question to the agency: “We remand the record in this case to the Commission to determine what procedures NEPA requires in light of its evaluation of new information about psychological health effects.”51 In other words, the NRC is required to determine whether a supplemental EIS is required, but it is not re*382quired to make that determination on the basis of any specific procedure.

This obviously is the correct result. The Supreme Court’s decision in Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc.52 makes clear that courts may not impose new procedural requirements on agencies. Moreover, there is no necessary reason why an agency must conduct a formal environmental assessment every time any person alleges that new and significant information requires preparation of a supplemental EIS. This determination may well be different from the determination whether an EIS was required originally, the determination that an environmental assessment is designed to facilitate.53 Thus, while the agency retains the power to order an environmental assessment if it wants to,54 neither NEPA nor the CEQ regulations require adherence to this procedure.

The court’s 7 January judgment also ordered the NRC to include in the environmental assessment consideration of the socioeconomic effects alleged by petitioner. Today, however, the court withdraws that requirement: “If NEPA requires the Commission to prepare a supplemental EIS regarding the TMI-1 restart decision because the agency makes a threshold finding of significant new information on psychological effects, ... PANE’S contentions regarding secondary effects on the community must be evaluated in the supplemental EIS.”55 This means that the NRC is not required to hear any evidence or make any findings on the alleged effects of restart on community well-being, unless it determines that the psychological health effects of restart are significant enough to warrant preparation of a supplemental EIS.

The majority is correct in drawing back from the requirement in its original judgment. Socioeconomic effects are, as the majority itself has found,56 only secondary effects which therefore do not themselves require preparation of an EIS. PANE has argued that secondary effects can necessitate a supplemental EIS,57 but this would lead to the absurd result of a continuing agency requirement to supplement its EIS with consideration of effects that were not significant enough to require preparation of an EIS in the first place. Nothing in NEPA or the CEQ regulations supports such a scheme, whereby a supplemental EIS is more easily triggered than an original EIS, and the majority has rightly rejected it.

C. The Injunction Against Restart of TMI-1

On 7 January the majority enjoined restart of TMI-1 until the NRC had completed an environmental assessment. This injunction has now been lifted, but only due to independent problems that will delay reopening of the plant. Since the majority has not given any indication that it will hesitate to reimpose the injunction if restart appears possible before completion of the NEPA proceedings, it is worth discussing the problems with the majority’s original injunction.

Perhaps the majority found it obvious that restart could not be permitted prior to completion of the NRC’s NEPA inquiry. It must be remembered, however, that the *383majority itself disclaims reliance on the happenstance that TMI-1 was shut down at the time of the accident. Since a restart decision is normally quite routine, the majority’s decision to issue the injunction suggests that it would have felt equally compelled to shut down TMI-1, pending the determination on whether to prepare a supplemental EIS, had it been in operation and had the NRC determined not to shut it down.

Viewed in this light, the propriety of issuing an injunction was much in doubt. There is no absolute principle that requires enjoining agency action pending compliance with NEPA. Rather, the court should weigh the equities to determine where the public interest lies:

[W]hile there is, in cases of NEPA noncompliance, a “presumption” in favor of injunctive relief, such relief does not follow automatically from every finding of a violation of NEPA.... What is called for, in each case, is a “particularized analysis” of the violations that have occurred, of the possibilities for relief, and of any countervailing considerations of public interest.58

In my view there are several considerations that militated against an injunction in this case. First, it is extremely unlikely that any consideration of psychological stress will result in a permanent closing of TMI-1.59 Consideration of measures designed to help reassure the public of the plant’s safety, even if required, could be undertaken while the plant was readied for restart. (Again, it must be remembered that the majority’s reasoning would apply even if the plant were operating; surely it would not make a mockery of the NEPA process were a court to order consideration of certain factors without ordering a shutdown of the plant.) . Second, there have been substantial rate increases for the people in the TMI area, both to pay the cost of the TMI-2 cleanup and to cover the costs of obtaining electricity elsewhere. This burden will be prolonged and perhaps increased if TMI-1 remains closed indefinitely. The court should have evaluated these costs.60

Third, and most important, the question before the Commission in January was whether to approve a restart limited to five percent of the plant’s power level. This low-power start was proposed unanimously by the Licensing Board, and “would facilitate testing of many nuclear safety devices and systems but would essentially eliminate the possibility of an accident having serious consequences for the public health and safety.”61 Such a short-term, low-power start would assuredly have much less impact on area residents’ psychological health than allegedly would a full restart for resumption of normal operations. The majority certainly should have considered this, and *384might well have ordered only that the Commission postpone decision on a full restart until completion of the environmental assessment.

By refusing even to consider these obviously relevant factors, the majority suggests that any time a court decides merely that a supplemental EIS may be needed, it necessarily must order that the nuclear plant, or other federal activity, be shut down until the agency has considered the environmental issues. NEPA was not intended to result in such a major interference in already approved and already operating facilities. Since I do not find any violation of NEPA in this case, I had no occasion to balance the equities in determining the need for an injunction. This the majority was required to do. The arguments against the injunction were strong, and the majority’s failure to address them was erroneous under any judicial review standard.

D. Summary

The fundamental error in the majority’s reasoning is its conclusion that allegations of psychological stress, caused by fear of a second nuclear accident, are cognizable under NEPA. The majority’s attempt to limit this holding to the “unique and traumatic” TMI situation has no basis in NEPA law. The majority seems to hope that psychological stress allegations will rarely be raised, so that the logically necessary extension of today’s holding to any kind of psychological harm, whether “post-traumatic” or not, will be avoided.

On the basis of all experience, however, one may well doubt whether opponents of nuclear power or other federal activities will so cooperate. Indeed, it appears that this court’s 7 January judgment has already been taken as indicating that psychological stress allegations will be fair game in any nuclear power proceeding. Despite PANE’S and the majority’s protestations that a prior nuclear accident is critical to the validity of stress allegations, antinuclear groups have recently filed such allegations in four licensing and permit proceedings:

For example, in the Shearon Harris operating license proceeding, three petitions to intervene were filed on February 17, 1982; all three seeking to raise psychological distress issues. Carolina Power & Light Co. (Shearon Harris Nuclear Power Plant, Units 1 and 2), NRC Docket Nos. 50-400, 50-401. In particular, a group known as the Environmental Law Project of the University of North Carolina sought to intervene, claiming that operation of the plant would undermine its members’ “psychological well-being”; that its members “must study long and hard to prepare for their classes”; and that “the psychological stress of an operating plant in such close proximity will detract from their studies.” Daniel F. Read also, sought to intervene, stating that he would like to contest “the psychological stress factor.” Similarly, a group called Chapel Hill Anti-Nuclear Effort claims that the plant would “undermine the psychological well-being of petitioner’s members.”

In the Black Fox construction permit proceeding, the intervenors on February 1, 1982, submitted a contention claiming that the safety evaluation of the plant was deficient because it “did not include the effects of psychological stress on the people within the 10 and 50 mile emergency planning zones.” Public Service Company of Oklahoma (Black Fox Units 1 and 2), NRC Docket Nos. 50-556, 50-557 (see Exhibit B).

In the Zimmer operating license proceeding, after the hearing had been completed, the intervenor indicated that it would attempt to renew a previously rejected psychological stress contention after the Court has issued its opinions in this case. Cincinnati Gas & Electric Co. (William H. Zimmer Nuclear Power Station, Unit 1), NRC Docket No. 50-358, Transcript of Hearing, March 4, 1982, at 7933-36 (see Exhibit C).

Finally, there is currently in progress a special NRC proceeding to determine whether the operating licenses for the Indian Point nuclear plants should be re*385voked. Consolidated Edison Company of New York (Indian Point Unit 2), NRC Docket No. 50-247-SP; Power Authority of the State of New York (Indian Point Unit 3), NRC Docket No. 50-286-SP. On December 8, 1981, a group called Parents Concerned About Indian Point filed proposed contentions in that proceeding, one of which (Contention IV) alleges that “[t]he physical and psychological environment of children will be improved by permanently shutting down the Indian Point Nuclear Power Station.” (See Exhibit D). The basis for this contention was explained in part as follows:

Parents, teachers, doctors, and other caretakers of children feel anxiety because of the continued operation of Indian Point. These anxieties are communicated to children and would be significantly reduced by cold shut down of Units 2 and 3.

On April 9,1982, the Indian Point Licensing Board issued a Memorandum and Order admitting the contention into the proceeding. (See Exhibit E, at 14). In a footnote, the Licensing Board made the following reference to this case:

The litigation of psychological aspects of this contention will be held in abeyance pending issuance of an opinion by the court in PANE v. NRC, Docket No. 81-1131, D.C. [Circuit] Court of Appeals, and any NRC policies or regulations issued as a result of that decision. We are also holding in abeyance action on the Power Authority’s Motion to Exclude Fear of Nuclear Power as an Issue in this Proceeding dated December 1, 1981.62

This makes clear that today’s decision will not be deemed limited or sui generis, but rather as inviting protests based on psychological stress to be raised in all nuclear proceedings. And under today’s reasoning, such a challenge to an operating plant would require the NRC at least to make a threshold determination whether a supplemental EIS must be prepared to take account of these allegations, and the plant would be shut down by court order pending completion of this determination.

Of course, the majority protests that it is not trying to draw a bright line to govern future situations, and perhaps it might decide these new cases differently. But in doing so it would necessarily repudiate the logic of its opinion today, and thereby expose this decision for what it is: a decision that NEPA must cover the “psychological aftermath” of the TMI-2 accident, not, in the final analysis, because NEPA’s history, purpose, and judicial construction demand it, but because the majority finds this “most publicized nuclear accident of our time” to be a terrible thing indeed.63

II. ATOMIC ENERGY ACT

In December 1980 the Commission voted 2-2 to deny consideration of psychological stress allegations in the TMI-1 restart proceeding. In separate opinions Commissioner (then Chairman) Ahearne and Commissioner Hendrie stated that the AEA did not require consideration of these issues.64 Commissioner Gilinsky and Commissioner Bradford voted to exercise the NRC’s discretion to consider these issues, but did not express a view on whether consideration was mandated.65 In September 1981 Chairman Palladino was named as the fifth commissioner, and he voted to deny consideration of psychological stress, without stating his reasoning.

On 7 January 1982 this court, Judge Wilkey dissenting, ordered the NRC to “prepare a statement of the reasons for its determination that psychological health is *386not cognizable under the Atomic Energy Act.66 On 30 March 1982 the Commission complied, filing a memorandum and order joined by Chairman Palladino and Commissioners Ahearne and Roberts.67 Commissioner Gilinsky again stated his view that the NRC should have considered psychological stress issues, and again did not contend that such consideration was mandated by the AEA.68

On the basis of the Commission’s 30 March opinion, we now uphold its decision.

A. Standard of Review

As the Supreme Court has recently reiterated, an agency’s interpretation of its governing statute is entitled to substantial deference. “To satisfy this standard it is not necessary for a court to find that the agency’s construction was the only reasonable one or even the reading the court would have reached if the question initially had arisen in a judicial proceeding.”69 Our deferential posture is further buttressed by the recognition that Congress’ scheme for regulating nuclear energy “is virtually unique in the degree to which broad responsibility is reposed in the administering agency.”70 We conclude that the Commission’s interpretation of the AEA is reasonable and completely consistent with the intent underlying the statute, and we accordingly affirm its decision not to consider psychological stress issues under the AEA.

B. Meaning of “Health and Safety” in the AEA

The AEA gives the Commission the responsibility to “protect the health and safety of the public.”71 In determining that this mandate does not include the responsibility to consider psychological reactions to nuclear power, the Commission emphasized that “the Atomic Energy Act itself does not discuss psychological health, and the statute, its legislative history, and applicable caselaw all suggest strongly that Congress intended the Commission to exercise its regulatory authority to protect only against the physical risks associated with radioactivity.” 72 In addition, the NRC noted that “there are strong policy considerations which argue against the consideration of psychological health effects per se in NRC licensing and enforcement proceedings.”73 Both rationales are reasonable and in accord with the AEA.

The Commission relied substantially, and in our view properly, on the First Circuit’s decision in New Hampshire v. Atomic Energy Commission,74 which represents the primary judicial construction of “health and safety” under the AEA. The court was asked to order the Commission to consider in a licensing proceeding “evidence of possible thermal pollution of the Connecticut River as a result of the discharge of cooling water by [the nuclear] facility.”75 Judge Coffin’s opinion for the court noted that Congress had nowhere in the statute provided any more precise definition of “health and safety,” and that under the modern usage of the phrase, psychological health would be included. But the court observed that it did not “feel that we fulfill our function responsibly by simply referring to the dictionary.”76 Considering the legislative history in detail, Judge Coffin found clear evidence that a broad construction was not intended:

*387Here we feel a very palpable restriction in the history surrounding the problem addressed by the Congress, the subsequent Congressional confirmation of the limited approach taken by the Commission, the contemporary efforts in the Congress to broaden that approach, and a recognition of the complexity of administrative arrangements which would attend a literal definition of public health and safety as these terms are used in the Atomic Energy Act.77

The court concluded that Congress “had in mind only the special hazards of radioactivity,” 78 and that the Commission’s responsibility is “confined to scrutiny of and protection against hazards from radiation.”79

Petitioner has failed to demonstrate that this well-reasoned opinion — which, significantly, has now been adopted by the Commission80 — is an unreasonable or erroneous construction of the Act. Petitioner’s primary argument is simply that the “plain meaning” of “health and safety,” which arguably includes psychological health, should control. The First Circuit directly considered this suggestion, however, and determined that there were strong indications in the legislative history that this was not the meaning intended by Congress. Both that court and the NRC found that the fundamental conclusion one must reach after considering the legislative history of the 1946 and 1954 acts, along with all the subsequent history of NRC regulation and congressional oversight, is that Congress was concerned about “the danger from explosion, radioactivity, and other harmful or toxic effects incident to the presence of such materials,” 81 and that the Commission was created “in order to bring a maximum of technical expertise to bear on complex and hazardous activities associated with a developing technology.”82 Given this background, limiting “health and safety” to the special hazards of radioactivity is reasonable and, indeed, compelling.

PANE argues, however, that even under the First Circuit’s definition psychological stress must be considered. Since there is “a justified public perception of radiation as extremely hazardous,” combined with the knowledge that a major accident could contaminate a large area, the psychological stress resulting from the TMI-2 accident and, allegedly, the TMI-1 restart is in fact a “special hazard of radioactivity.”83 We agree with the Commission that this argument is “unpersuasive. Presumably, every hazardous technology gives rise to fears peculiarly associated with it: fear of being inundated by failure of a newly constructed dam, for example, or of being hit by debris *388from a crashing airplane.”84 Individuals may experience psychological trauma from the occurrence of accidents or disasters such as these, all of which take place with sufficient frequency to claim a substantial number of human lives. It is obvious, therefore, that “post-traumatic psychological stress” can result from any traumatic event, and is not so peculiar to nuclear energy that Congress can be deemed to have considered it a special hazard of radioactivity. Rather, the special hazards Congress was concerned with were those associated with human exposure to radiation.

Moreover, if, as PANE argues, there is an extreme fear of radiation and of a major nuclear accident, then certainly there must be individuals who experience medically-diagnosable stress resulting from the operation of any nuclear facility. Since this undeniably is an effect on psychological health, it would need to be considered in every licensing proceeding. Yet PANE itself recognizes that this is contrary to congressional intent:

It might well be an absurd result to hold that the Commission is required to take into account, as within Section 103(d), the fears that normally arise in a community when a nuclear reactor is proposed. That type of interpretation could conceivably prohibit nuclear reactors virtually anywhere, which is clearly not the intent of Congress.85

This attempt to limit the argument is plainly illogical. If someone alleges harm to his psychological health from the operation of a nuclear plant, and if “health and safety” encompasses psychological harms, why should not the NRC be required at least to consider this allegation in its licensing decision? It may be likely that the stress alleged in the aftermath of the TMI-2 accident will be greater than the stress alleged generally, but PANE would be hard-pressed indeed to provide any justification for interpreting the “plain meaning” of “health and safety” to include “post-traumatic psychological stress” but not other kinds of psychological stress. As the Commission stated, “we cannot believe that [Congress] meant that ‘health’ under the Atomic Energy Act, should clearly encompass the psychological well-being of persons fearful of a second nuclear accident in their vicinity, while equally clearly excluding the mental health of persons who fear that their locality may experience its first nuclear accident.”86

It was also reasonable for the Commission to note that “the major contribution which it can make to the alleviation of psychological stress is to make sound technical decisions in its areas of expertise.” 87 If the AEA were read to require NRC attention to psychological reactions to nuclear power, the result might be a substantial shift in the agency’s allocation of resources away from its chief responsibility of ensuring the technical safety of operative reactors. Since the NRC does not have expertise on mental distress, the result inevitably would be a reduced ability to attend to the safety issues at hand.88

*389In sum, petitioner cannot justify its position by offering any reasoned interpretation of the statute which will apply generally in the future. There simply is no basis for this court to overrule the agency’s interpretation. The AEA does not require consideration of psychological stress allegedly resulting from fear of a nuclear accident, regardless whether an accident has previously occurred in the vicinity.

III. CONCLUSION

The majority has erred in its construction of NEPA. I have given detailed attention above to the many problems with the majority’s analysis, but I also think it worthwhile to point out the flawed premises that seem to underlie the decision. The majority sums up its treatment of the NEPA issues by declaring:

In the wake of the most publicized nuclear accident of our time, the people of the Three Mile Island area — and the people of the nation as a whole — are entitled to the protections Congress provided in the National Environmental Policy Act. The government must not proceed to make decisions that might have a momentous effect on the psychological health and community well-being of its citizens without first giving careful, responsible consideration to the consequences its actions might have. By enacting NEPA Congress meant to assure that no federal decision — especially one of this importance — would be made in the shadow of environmental ignorance.89

The points the majority deems worthy of great emphasis — that this was “the most publicized nuclear accident of our time,” that “the people of the nation as a whole” are somehow and for some reason entitled to NEPA consideration in this case, and that NEPA was meant to apply “especially” to decisions “of this importance” — are revealing.

They suggest a fundamental finding that the TMI-2 accident must not be allowed to pass without requiring the NRC to ask whether nuclear power, at least at this one site, should be allowed to continue at all. Whether to continue at all is a finding beyond the power of any agency — or any court. Congress, wisely or unwisely, has made the decision that this country will develop nuclear power.90 No court has the power to rewrite NEPA to impose additional conditions, certain to produce long delays and perhaps ultimately impossible to fulfill. Congress knows everything that this court knows — and more — about the accident at TMI-2. When the standards by which the agency is to act under either NEPA or AEA are to be changed, Congress will determine in what direction and degree.

At the heart of the majority view is a belief that nuclear power and its attendant risks must be judged under some special standard. I submit that this represents a blindness toward the larger picture. Energy production of any sort is risky. Coal is a primary alternative to nuclear power. Yet 150 coal miners die each year, and many more contract debilitating diseases.91 As the reliance on coal increases — as when a court prevents a licensed nuclear plant from reopening — these risks increase. And, most important, these risks fall almost exclusively on one group of society, human beings not among the more privileged in our America. The risks of nuclear power, in contrast, fall upon rich and poor alike. Yet the majority expresses apparent indignation at the thought that “the people of the Three Mile Island area ... and the people of the nation as a whole” might be deprived of a chance to argue that the psychological strain of living with the risk of nuclear power is too great to permit TMI-1 to operate. Apparently the risk should remain where it traditionally has fallen, on those who produce the energy rather than on those who use it.

*390This special fear of nuclear power ultimately serves to institutionalize a fear of taking risks at all. It would seem to me quite logical to attempt to treat the stress resulting from the TMI-2 accident and to take all necessary steps to minimize the possibility of an accident that might cause similar stress in the future. But petitioners instead have insisted, and the majority has agreed, on a quite different and ultimately far-reaching orientation: when the problem is psychological stress, the solution is not to help individuals overcome their fears, but to determine whether the feared activity should be abandoned.

We have thus come a long way in fifty years, from a time when the President of the United States was widely and enthusiastically applauded for declaring: “The only thing we have to fear is fear itself.” Now the fear itself necessitates an environmental assessment. All risky activity must grind to a halt in the interim. Inaction has become epidemic, and delay is maximized. I do not believe the Congress intended NEPA to constrain federal action on these psychological and emotional bases. Accordingly, I respectfully dissent.