concurring in the judgment.
I agree with the Court insofar as it holds that the fact that anticompetitive conduct is sanctioned, or even required, by state law does not of itself put that conduct beyond the reach of the Sherman Act. Since the opposite proposition is the ground on which the Court of Appeals affirmed the dismissal of this suit, I also agree that its judgment must be reversed. My approach, however, is somewhat different from that of the Court.
I
As to the principal question in the case, that of the Sherman Act’s pre-emptive effect upon inconsistent state laws, it is, as the dissent points out, one of congressional intent. No one denies that Congress could, if it wished, override those state laws whose operation would subvert the federal policy of free competition in interstate commerce. In discerning that intent, however, I find somewhat less assistance in the legislative history than does the dissent. It is true that the framers of the Sherman Act expressed the view that certain areas of economic activity were left entirely to state regulation. The dissent quotes several of these expressions. Post, at 632-634. A careful reading of those statements reveals, however, that they little more than reflect the then-prevailing view that Congress lacked the power, under the Commerce Clause, to regulate economic activity that was within the domain of the States. The Court since then has recognized a greatly expanded Commerce Clause *606power. Arguably, the Sherman Act should have remained confined within the outlines of that power as it was thought to exist in 1890, on the theory that if Congress believed it could not regulate any more broadly, it must not have attempted to do so. But that bridge already has been crossed, for it has been held that Congress intended the reach of the Sherman Act to expand along with that of the commerce power. Hospital Building Co. v. Rex Hospital Trustees, 425 U. S. 738, 743 n. 2 (1976), and cases cited.
Our question in this case is one that the Sherman Act’s framers did not directly confront or explicitly address: What was to be the result if the expanding ambit of the Sherman Act should bring it into conflict with inconsistent state law? But it seems to me that this bridge also has been crossed. In Schwegmann Bros. v. Calvert Distillers Corp., 341 U. S. 384 (1951), the issue was whether the Sherman Act permitted enforcement of a Louisiana statute requiring compliance by liquor retailers with resale price agreements to which they were not parties, but which had been entered into by other retailers with their wholesale suppliers. The Court held the Louisiana statute unenforceable; there is no plausible reading of that decision other than that the statute was pre-empted by the Sherman Act.1 Northern Securities Co. v. United States, 193 U. S. 197 (1904), is to the same effect. The defenders of the railroad holding company attacked in that case argued that it was beyond the Sherman Act’s reach because it was lawful under the cor*607poration laws of New Jersey. The holding company was nonetheless held unlawful, and, to that extent, the law of New Jersey was forced to give way.2 Indeed, I suppose that some degree of state-law pre-emption is implicit in the most fundamental operation of the Sherman Act. If a State had no antitrust policy of its own, anticompetitive combinations of all kinds would be sanctioned and enforced under that State’s general contract and corporation law. Yet, there has never been any doubt that if such combinations offend the Sherman Act, they are illegal, and state laws to that extent are overridden.3
Congress itself has given support to the view that inconsistent state laws are pre-empted by the Sherman Act. Were it the case that state statutes held complete sway, Congress would not have found it necessary in 1937 to pass the Miller-Tydings Fair Trade Act, 50 Stat. 693, amending the Sherman Act, specifically exempting from the latter’s operation certain price maintenance agreements sanctioned by state law. 15 U. S. C. § 1. There are other instances of Congress’ acting to protect state-sanctioned anticompetitive schemes from the Sher*608man Act. In response to Schwegmann, see H. R. Rep. No. 1437, 82d Cong., 2d Sess., 1-2, Congress in 1952 passed the McGuire bill, 66 Stat. 632, extending the Miller-Tydings exemption to state statutes that enforced resale price agreements against nonsigners. 15 U. S. C. §§45 (a) (2) to (5). A similar enactment is the McCar-ran-Ferguson Act of 1945, 59 Stat. 34, exempting from federal statutes “any law enacted by any State for the purpose of regulating the business of insurance,” with provision that the Sherman Act, and other named federal statutes, should apply to that business after a specified date “to the extent that such business is not regulated by State law.” 15 U. S. C. § 1012 (b).4 These express grants of Sherman Act immunity seem significant to me. As the Court stated in United States v. Borden Co., 308 U. S. 188, 201 (1939), construing the immunity granted to certain agreements by the Agricultural Mar*609keting Agreement Act of 1937, “[i]f Congress had desired to grant any further immunity, Congress doubtless would have said so.”
II
I also agree with Mr. Justice Stevens that the particular anticompetitive scheme attacked in this case must fall despite the imprimatur it claims to have received from the State of Michigan. To say, as I have, that the Sherman Act generally pre-empts inconsistent state laws is not to answer the much more difficult question as to which such laws are pre-empted and to what extent. I fear there are no easy solutions, though several suggest themselves.
It cannot be decisive, for example, simply that a state law goes so far as to require, rather than simply to authorize, the anticompetitive conduct in question. The Court accepted this as a prerequisite to antitrust immunity in Goldfarb v. Virginia State Bar, 421 U. S. 773, 790 (1975), but it cannot alone be sufficient. The whole issue in Schwegmann was whether the State could require obedience to a fixed resale price arrangement. Similarly, compliance with an anticompetitive contract, or adherence to an illegal corporate combination, might well be “required” by a State’s general contract and corporation law.
Neither can it be decisive that a particular state-sanctioned scheme was initiated by the private actors rather than by the State. I see no difference in the degree of private initiation as between the marketing arrangement approved in Parker v. Brown, 317 U. S. 341 (1943) (and properly approved, I think, for reasons set forth below), and the resale price maintenance scheme disapproved in Schwegmann. In each case the particular scheme was initiated by the private actors at the invitation of a general statute, with which they may or may *610not have had anything to do. The same was true in Northern Securities, and the same is true here. To be sure, there is a certain rough justice, as well as an appearance of simplicity, in a rule based upon who actually is responsible for the scheme in question, but I fear that both the justice and the simplicity would prove illusory in the rule’s actual application. Every state enactment is initiated, in its way, by its beneficiaries. It would scarcely make sense to immunize only those powerful enough to speak entirely through their governmental representatives, or, for that matter, to stifle such speech with the threat that it will destroy antitrust immunity. Moreover, the process of enactment is likely to involve such a complex interplay between those regulating and those regulated that it will be impossible to identify the true “initiator.”
A final, ostensibly simple, solution that I find wanting would be to insist only on some degree of affirmative articulation by the State of its conscientious intent to sanction the challenged scheme, and its reasons therefor. This also is a tempting solution, particularly in this case, where there is little to suggest (at least in recent years) that the Michigan Public Service Commission has even actively considered the light-bulb tie-in, much less articulated a justification for it. Yet such a solution would also lead to perverse results. A regulation whose justification was too plain to require explication would be vulnerable; a questionable one could be immunized if its proponents had the skill or influence to generate the proper legislative history. And, of course, deciding how much “affirmative articulation” of state policy is enough is not a simple matter.
I would apply,, at least for now, a rule of reason, taking it as a general proposition that state-sanctioned anti-competitive activity must fall like any other if its potential harms outweigh its benefits. This does not mean *611that state-sanctioned and private activity are to be treated alike. The former is different because the fact of state sanction figures powerfully in the calculus of harm and benefit. If, for example, the justification for the scheme lies in the protection of health or safety, the strength of that justification is forcefully attested to by the existence of a state enactment. I would assess the justifications of such enactments in the same way as is done in equal protection review, and where such justifications are at all substantial (as one would expect them to be in the case of most professional licensing or fee-setting schemes, for example, cf. Olsen v. Smith, 195 U. S. 332 (1904)), I would be reluctant to find the restraint unreasonable. A particularly strong justification exists for a state-sanctioned scheme if the State in effect has substituted itself for the forces of competition, and regulates private activity to the same ends sought to be achieved by the Sherman Act. Thus, an anticompetitive scheme which the State institutes on the plausible ground that it will improve the performance of the market in fostering efficient resource allocation and low prices can scarcely be assailed. One could not doubt the legality of Detroit Edison’s electric power monopoly; the fear of such a monopoly is primarily its tendency to charge excessive prices, but its prices in this instance are controlled by the State.
No doubt such a rule of reason will crystallize, as it is applied, into various -per se rules relating to certain kinds of state enactments, such as the regulation of the classic natural monopoly, the public utility. We should not shrink in our general approach, however, from what seems to me our constitutionally mandated task, one often set for us by conflicting federal and state laws, and that is the balancing of implicated federal and state interests with a view to assuring that when these are truly in conflict, the former prevail.
*612The dissent's fears on this score appear to me to be exaggerated. The balancing of harm and benefit is, in general, a process with which federal courts are well acquainted in the antitrust field. The special problem of assessing state interests to determine whether they are strong enough to prevail against supreme federal dictates is also a familiar one to the federal courts. Indeed, a state action that interferes with competition not only among its own citizens but also among the States is already subject under the Commerce Clause to much ■the same searching review of state justifications as is proposed here. See, e. g., Dean Milk Co. v. Madison, 340 U. S. 349, 354 (1951) (state restriction on sale of milk not locally processed held invalid because “reasonable and adequate alternatives [were] available” to protect health interests); Southern Pacific Co. v. Arizona, 325 U. S. 761, 770-784 (1945) (state restriction of train lengths held invalid under the Commerce Clause because “the state [safety] interest is outweighed by the interest of the nation in an adequate, economical and efficient railway transportation service”).
Ill
By these standards the present case does not seem a difficult one. The light-bulb tie-in presents the usual dangers of such a scheme, principally that respondent will extend its monopoly from the sale of electric power into that of light bulbs, not because it sells better light bulbs, but because its light bulbs are the ones customers must pay for if they are to have light at all. See P. Areeda, Antitrust Analysis 569-570 (2d ed. 1974). On the record before us the scheme appears to be unjustified. No doubt it originated as a means to promote electric power use, but it is difficult to see why a tie-in (rather than an optional, promotional light-bulb sale) was nec*613essary to that end even in the 19th century, laying aside the question whether the promotion of greater electric power use remains today a plausible public goal. Respondent would justify the scheme on the ground of consumer savings, its light bulbs assertedly being cheaper and better than those commercially available. Brief for Respondent 7-9, 41-42. But again, a tie-in is not necessary to pass along these savings. A tie-in is only necessary in order to force consumers to pay for light bulbs from Detroit Edison rather than someone else. But there is no indication that one light bulb does not fit the socket as well as another, or that the sale of light bulbs is in any way crucial to respondent’s successful operation. Conceivably, Michigan’s aim is the very extension of the monopoly, bom of a preference for having light bulbs supplied by one whose prices are already regulated., But ending competition in the light-bulb market cannot be accepted as an adequate state objective without some evidence — of which there is not the least hint in this record — that such competition is in some way ineffective. For all that appears, light-bulb marketing, unlike electric power production, is not a natural monopoly, nor does it implicate health or safety, nor is it beset with problems of instability or other flaws in the competitive market.5 *614This is what I take it the Court means when it says the electric light-bulb market is “essentially unregulated/’ and on that understanding I agree with its conclusion. It is conceivable that respondent may show, upon further evidence, a sufficient justification for the scheme, but it certainly has not done so as yet.6
The Court expressly stated in Schwegmann: “The fact that a state authorizes the price fixing does not, of course, give immunity to the scheme, absent approval by Congress.” And again: “[W]hen a state compels retailers to follow a parallel price policy, it demands private conUuct which the Sherman Act forbids.” 341 U. S., at 386, 389.
The argument that New Jersey law exempted Northern Securities Company from the Sherman Act was thoroughly canvassed in the plurality opinion. 193 U. S., at 344r-351. It was rejected for the reason “that no State can endow any of its corporations, or any combination of its citizens, with authority to restrain interstate or international commerce, or to disobey the national will as manifested in legal enactments of Congress.” Id., at 350.
In passing, we may cast at least a sidelong glance at a related area of federal trade regulation — that of the patent laws. Although the federal statute is no more explicit on the point than is the Sherman Act, see 35 U. S. C. § 100 et seq., it clearly pre-empts state laws that purport either to expand on or to infringe the federal patent monopoly. See, e. g., Lear, Inc. v. Adkins, 395 U. S. 653 (1969); Sears, Roebuck & Co. v. Stiffel Co., 376 U. S. 225 (1964); Compco Corp. v. Day-Brite Lighting, 376 U. S. 234 (1964).
The McCarran-Ferguson Act was passed in. reaction to the holding in United States v. Underwriters Assn., 322 U. S. 533 (1944), that the business of insurance is “commerce” within the meaning of the Sherman Act. Congress’ expressed concern was that the application of that Act would “greatly impair or nullify the regulation of insurance by the States,” bringing to a halt their “experimentation and investigation” in the area. The Act was vigorously endorsed by Governors and insurance commissioners of “almost all of the States.” The Justice Department, in opposing the McCarran-Ferguson Act-, specifically argued that Parker v. Brown, 317 U. S. 341 (1943), made the legislation unnecessary because it immunized the insurance business insofar as it was regulated by the States. Congress was not so sure:
“Parker v. Brown dealt with a State commission authorized by State statute to enforce a program in conformity with, if not supplementary to, a Federal statute. Obviously, all State regulation concerning insurance does not and would not fall in such a category.” S. Rep. No. 1112, 78th Cong., 2d Sess., 5 (1944).
See also S. Rep. No. 20, 79th Cong., 1st Sess., 1-3 (1945); H. R. Rep. No. 873, 78th Cong., 1st Sess., 7 (1943); H. R. Rep. No. 143, 79th Cong., 1st Sess., 4 (1945).
The approach described in the text is entirely consistent with the result reached in Parker v. Brown. Wildly fluctuating agricultural prices are a prime candidate for some collective scheme that interrupts free competition in order to bring badly needed stability; under the State’s close supervision, as was the case in Parker, the scheme seems entirely reasonable. I see no reason to disapprove the holding of Parker, therefore, and to the extent that the plurality, by stressing the identity of the state defendants in that case, intimates that a different result might have been reached had the raisin growers themselves been sued, I cannot agree.
Neither can I agree with the dissent, however, that Parker must be taken to stand for the broad proposition that a State can *614immunize any conduct from the application of the Sherman Act. It is true, as the dissent points out, that there are statements arguably to that effect in Parker, but the opinion is hardly unambiguous on the point. The Court also observed in that case that “a state does not give immunity to those who violate the Sherman Act by authorizing them to violate it, or by declaring that their action is lawful.” 317 U. S., at 351. Moreover, if we must choose between Parker’s more categorical statements and the seemingly contrary statements in Schwegmann and Northern Securities, see nn. 1 and 2, supra, I prefer the latter, as more in keeping, with the actual holdings of those cases.
Mr. Justice SteveNS states that there may be cases in which “the State’s participation in a decision [to adopt the challenged restraint] is so dominant that it is unfair to hold a private party responsible for his conduct in implementing it.” Ante, at 594-595. I agree that a defense based on fairness may be available. I would not, however, rule it out in this case, as the Court’s opinion does. The parties, like the court below, so far have addressed themselves only to the question whether petitioner’s suit is completely barred by Parker v. Brown and the Michigan Public Service Commission’s approval of the challenged tie-in. I would confine our present decision to that question alone, leaving consideration of a fairness defense to the lower courts on remand, and making only these two further observations:
First, I take it that a defense based on fairness would be a defense to a damages recovery but not injunctive relief. The latter, of course, presents no danger of unfairness. Moreover, as Mr. Justice SteveNS implies by his emphasis on not unfairly holding a *615private party “responsible,” the defense rests on the theory, not that the challenged restraint is legal, but that since the defendant has committed no voluntary act in implementing it, he cannot be said to have violated any law. The same would not be true of acts following a judgment that the restraint is in fact illegal, and the state law to that extent invalid.
Second, I would hope that consideration will be given on remand to allowing a defense against damages wherever the conduct on which such damages would be based was required by state law. Such a rule would comport with the theory that a defendant should not be held “responsible” in damages for conduct as to which he had no choice, by which I do not mean to rule out other possible grounds for such a rule. See Posner, The Proper Relationship Between State Regulation and the Federal Antitrust Laws, 49 N. Y. U. L. Rev. 693, 728-732 (1974). It would also eliminate what seems to me the extremely unfair possibility that during a particular period — and it could be a regulatory lag during which the regulatee was attempting to change the state mandate — the regu-latee could be required by state law to conform to a course of conduct for which he was all the while accumulating treble-damages liability under federal law.