dissenting.
The citizens of Chicago were once free to drive about the city at whatever speed they wished. At some point Chica-goans (or perhaps Illinoisans) decided this would not do, and imposed prophylactic speed limits designed to assure safe operation by the average (or perhaps even subaverage) driver with the average (or perhaps even subaverage) vehicle. This infringed upon the “freedom” of all citizens, but was not unconstitutional.
Similarly, the citizens of Chicago were once free to stand around and gawk at the scene of an accident. At some point Chicagoans discovered that this obstructed traffic and caused more accidents. They did not make the practice unlawful, but they did authorize police officers to order the crowd to disperse, and imposed penalties for refusal to obey such an order. Again, this prophylactic measure infringed upon the “freedom” of all citizens, but was not unconstitutional.
Until the ordinance that is before us today was adopted, the citizens of Ghicago were free to stand about in public places with no apparent purpose — to engage, that is, in conduct that appeared to be loitering. In recent years, however, the city has been afflicted with criminal street gangs. As reflected in the record before us, these gangs congregated *74in public places to deal in drugs, and to terrorize the neighborhoods by demonstrating control over their “turf.” Many residents of the inner city felt that they were prisoners in their own homes. Once again, Chicagoans decided that to eliminate the problem it was worth restricting some of the freedom that they once enjoyed. The means they took was similar to the second, and more mild, example given above rather than the first: Loitering was not made unlawful, but when a group of people occupied a public place without an apparent purpose and in the company of a known gang member, police officers were authorized to order them to disperse, and the failure to obey such an order was made unlawful. See Chicago Municipal Code §8-4-015 (1992). The minor limitation upon the free state of nature that this prophylactic arrangement imposed upon all Chicagoans seemed to them (and it seems to me) a small price to pay for liberation of their streets.
The majority today invalidates this perfectly reasonable measure by ignoring our rules governing facial challenges, by elevating loitering to a constitutionally guaranteed right, and by discerning vagueness where, according to our usual standards, none exists.
I
Respondents’ consolidated appeal presents a facial challenge to the Chicago ordinance on vagueness grounds. When a facial challenge is successful, the law in question is declared to be unenforceable in all its applications, and not just in its particular application to the party in suit. To tell the truth, it is highly questionable whether federal courts have any business making such a declaration. The rationale for our power to review federal legislation for constitutionality, expressed in Marbury v. Madison, 1 Cranch 137 (1803), was that we had to do so in order to decide the case before us. But that rationale only extends so far as to require us to determine that the statute is unconstitutional as applied to this party, in the circumstances of this case.
*75That limitation was folly grasped by Tocqueville, in his famous chapter on the power of the judiciary in American society:
“The second characteristic of judicial power is, that it pronounces on special eases, and not upon general principles. If a judge, in deciding a particular point, destroys a general principle by passing a judgment which tends to reject all the inferences from that principle, and consequently to annul it, he remains within the ordinary limits of his functions. But if he directly attacks a general principle without having a particular case in view, he leaves the circle in which all nations have agreed to confine his authority; he assumes a more important, and perhaps a more useful influence, than that of the magistrate; but he ceases to represent the judicial power.
“Whenever a law which the judge holds to be unconstitutional is invoked in a tribunal of the United States, he may refuse to admit it as a rule .... But as soon as a judge has refused to apply any given law in a case, that law immediately loses a portion of its moral force. Those to whom it is prejudicial learn that means exist of overcoming its authority; and similar suits are multiplied, until it becomes powerless.... The political power which the Americans have intrusted to their courts of justice is therefore immense; but the evils of this power are considerably diminished by the impossibility of attacking the laws except through the courts of justice.... [W]hen a judge contests a law in an obscure debate on some particular case, the importance of his attack is concealed from public notice; his decision bears upon the interest of an individual, and the law is slighted only incidentally. Moreover, although it is censured, it is not abolished; its moral force may be diminished, but its authority is not taken away; and its final destruction can *76be accomplished only by the reiterated attacks of judicial functionaries.” Democracy in America 78, 75-76 (R. Heffner ed. 1956).
As Justice Sutherland described our system in his opinion for a unanimous Court in Massachusetts v. Mellon, 262 U. S. 447, 488 (1923):
“We have no power per se to review and annul acts of Congress on the ground that they are unconstitutional. That question may be considered only when the justification for some direct injury suffered or threatened, presenting a justiciable issue, is made to rest upon such an act. Then the power exercised is that of ascertaining and declaring the law applicable to the controversy. It amounts to little more than the negative power to disregard an unconstitutional enactment, which otherwise would stand in the way of the enforcement of a legal right.... If a case for preventive relief be presented the court enjoins, in effect, not the execution of the statute, but the acts of the official, the statute notwithstanding.”
And as Justice Brennan described our system in his opinion for a unanimous Court in United States v. Raines, 362 U. S. 17, 20-22 (1960):
“The very foundation of the power of the federal courts to declare Acts of Congress unconstitutional lies in the power and duty of those courts to decide cases and controversies before them.... This Court, as is the case with all federal courts, ‘has no jurisdiction to pronounce any statute, either of a State or of the United States, void, because irreconcilable with the Constitution, except as it is called upon to adjudge the legal rights of litigants in actual controversies. In the exercise of that jurisdiction, it is bound by two rules, to which it has rigidly adhered, one, never to anticipate a question of constitutional law in advance of the necessity of deciding it; the other never to formulate a rule of *77constitutional law broader than is required by the precise facts to which it is to be applied.’ . . . Kindred to these rules is the rule that one to whom application of a statute is constitutional will not be heard to attack the statute on the ground that impliedly it might also be taken as applying to other persons or other situations in which its application might be unconstitutional.... The delicate power of pronouncing an Act of Congress unconstitutional is not to be exercised with reference to hypothetical cases thus imagined.”
It seems to me fundamentally incompatible with this system for the Court not to be content to find that a statute is unconstitutional as applied to the person before it, but to go further and pronounce that the statute is unconstitutional in all applications. Its reasoning may well suggest as much, but to pronounce a holding on that point seems to me no more than an advisory opinion — which a federal court should never issue at all, see Hayburn’s Case, 2 Dall. 409 (1792), and especially should not issue with regard to a constitutional question, as to which we seek to avoid even ftowadvisory opinions, see, e. g., Ashwander v. TVA, 297 U. S. 288, 347 (1936) (Brandeis, J., concurring). I think it quite improper, in short, to ask the constitutional claimant before us: Do you just want us to say that this statute cannot constitutionally be applied to you in this ease, or do you want to go for broke and try to get the statute pronounced void in all its applications?
I must acknowledge, however, that for some of the present century we have done just this. But until recently, at least, we have — except in ffee-speech cases subject to the doctrine of overbreadth, see, e. g., New York v. Ferber, 458 U. S. 747, 769-773 (1982) — required the facial challenge to be a go-for-broke proposition. That is to say, before declaring a statute to be void in all its applications (something we should not be doing in the first place), we have at least imposed upon the litigant the eminently reasonable requirement that he estab*78lish that the statute was unconstitutional in all its applications. (I say that is an eminently reasonable requirement, not only because we should not be holding a statute void in all its applications unless it is unconstitutional in all its applications, but also because unless it is unconstitutional in all its applications we do not even know, without conducting an as-applied analysis, whether it is void with regard to the very litigant before us — whose case, after all, was the occasion for undertaking this inquiry in the first place.1)
As we said in United States v. Salerno, 481 U. S. 739, 745 (1987):
“A facial challenge to a legislative Act is, of course, the most difficult challenge to mount successfully, since the challenger must establish that no set of circum*79stances exists under which the Act would be valid. The faet that [a legislative Act] might operate unconstitutionally under some conceivable set of circumstances is insufficient to render it wholly invalid, since we have not recognized an 'overbreadth’ doctrine outside the limited context of the First Amendment.” (Emphasis added.)2
This proposition did not originate with Salerno, but had been expressed in a line of prior opinions. See, e. g., Members of City Council of Los Angeles v. Taxpayers for Vincent, 466 U. S. 789, 796 (1984) (opinion for the Court by Stevens, J.) (statute not implicating First Amendment rights is invalid on its face if "it is unconstitutional in every conceivable application”); Schall v. Martin, 467 U. S. 253, 269, n. 18 (1984); Hoffman Estates v. Flipside, Hoffman Estates, Inc., 455 U. S. 489, 494-495, 497 (1982); United States v. National Dairy Products Corp., 372 U. S. 29, 31-32 (1963); Raines, 362 U. S., at 21. And the proposition has been reaffirmed in many cases and opinions since. See, e.g., Anderson v. Edwards, 514 U.S. 143, 155-156, n. 6 (1995) (unanimous Court); Babbitt v. Sweet Home Chapter, Communities for Great Ore., 515 U. S. 687, 699 (1995) (opinion for the Court by Stevens, J.) (facial challenge asserts that a challenged statute or regulation is invalid “in every circumstance”); Reno v. Flores, 507 U. S. 292, 301 (1993); Rust v. Sullivan, *80500 U. S. 173, 183 (1991); Ohio v. Akron Center for Reproductive Health, 497 U. S. 502, 514 (1990) (opinion of Kennedy, J.); Webster v. Reproductive Health Servs., 492 U. S. 490, 523-524 (1989) (O’Connoe, J., concurring in part and concurring in judgment); New York State Club Assn., Inc. v. City of New York, 487 U. S. 1, 11-12 (1988).3 Unsurprisingly, given the clarity of our general jurisprudence on this point, the Federal Courts of Appeals all apply the Salerno standard in adjudicating facial challenges.4
*81I am aware, of course, that in some recent facial-challenge eases the Court has, without any attempt at explanation, created entirely irrational exceptions to the “unconstitutional in every conceivable application” rule, when the statutes at issue concerned hot-button social issues on which “informed opinion” was zealously united. See Romer v. Evans, 517 U. S. 620, 643 (1996) (Scalia, J., dissenting) (homosexual rights); Planned Parenthood of Southeastern Pa. v. Casey, 505 U. S. 833, 895 (1992) (abortion rights). But the present case does not even lend itself to such a “political correctness” exception — which, though illogical, is at least predictable. It is not a la mode to favor gang members and associated loiterers over the beleaguered law-abiding residents of the inner city.
When our normal criteria for facial challenges are applied, it is clear that the Justices in the majority have transposed the burden of proof. Instead of requiring respondents, who are challenging the ordinance, to show that it is invalid in all its applications, they have required petitioner to show that it is valid in all its applications. Both the plurality opinion and the concurrences display a lively imagination, creating hypothetical situations in which the law’s application would (in their view) be ambiguous. But that creative role has been usurped from petitioner, who can defeat respondents’ facial challenge by conjuring up a single valid application of the law. My contribution would go something like this:5 Tony, a member of the Jets criminal street gang, is standing *82alongside and chatting with fellow gang members while staking out their turf at Promontory Point on the South Side of Chicago; the group is flashing gang signs and displaying their distinctive tattoos to passersby. Officer Krupke, applying the ordinance at issue here, orders the group to disperse. After some speculative discussion (probably irrelevant here) over whether the Jets are depraved because they are deprived, Tony and the other gang members break off further conversation with the statement' — not entirely coherent, but evidently intended to be rude — “Gee, Officer Krupke, krup you.” A tense standoff ensues until Officer Krupke arrests the group for failing to obey his dispersal order. Even assuming (as the Justices in the majority do, but I do not) that a law requiring obedience to a dispersal order is impermissi-bly vague unless it is clear to the objects of the order, before its issuance, that their conduct justifies it, I find it hard to believe that the Jets would not have known they had it coming. That should settle the matter of respondents’ facial challenge to the ordinance’s vagueness.
Of course respondents would still be able to claim that the ordinance was vague as applied to them. But the ultimate demonstration of the inappropriateness of the Court’s holding of facial invalidity is the fact that it is doubtful whether some of these respondents could even sustain an as-applied challenge on the basis of the majority’s own criteria. For instance, respondent Jose Renteria — who admitted that he was a member of the Satan Disciples gang — was observed by the arresting officer loitering on a street corner with other gang members. The officer issued a dispersal order, but when she returned to the same corner 15 to 20 minutes later, Renteria was still there with his Mends, whereupon he was arrested. In another example, respondent Daniel Washington and several others — who admitted they were members of the 'Vice Lords gang — were observed by the arresting officer loitering in the street, yelling at passing vehicles, stopping traffic, and preventing pedestrians from using *83the sidewalks. The arresting officer issued a dispersal order, issued another dispersal order later when the group did not move, and finally arrested the group when they were found loitering in the same place still later. Finally, respondent Gregorio Gutierrez — who had previously admitted to the arresting officer his membership in the Latin Kings gang — was observed loitering with two other men. The officer issued a dispersal order, drove around the block, and arrested the men after finding them in the same place upon his return. See Brief for Petitioner 7, n. 5; Brief for United States as Amicus Curiae 16, n. 11. Even on the majority’s assumption that to avoid vagueness it must be clear to the object of the dispersal order ex ante that his conduct is covered by the ordinance, it seems most improbable that any of these as-applied challenges would be sustained. Much less is it possible to say that the ordinance is invalid in all its applications.
II
The plurality’s explanation for its departure from the usual rule governing facial challenges is seemingly contained in the following statement: “[This] is a criminal law that contains no mens rea requirement... and infringes on constitutionally protected rights .... When vagueness permeates the text of such a law, it is subject to facial attack.” Ante, at 55 (emphasis added). The proposition is set forth with such assurance that one might suppose that it repeats some well-accepted formula in our jurisprudence: (Criminal law without mens rea requirement) + (infringement of constitutionally protected right) + (vagueness) = (entitlement to facial invalidation). There is no such formula; the plurality has made it up for this case, as the absence of any citation demonstrates.
But no matter. None of the three factors that the plurality relies upon exists anyway. I turn first to the support for the proposition that there is a constitutionally protected right to loiter — or, as the plurality more favorably describes *84it, for a person to “remain in a public place of his choice.” Ante, at 54. The plurality thinks much of this Fundamental Freedom to Loiter, which it contrasts with such lesser, constitutionally unprotected, activities as doing (ugh!) business: “This is not an ordinance that simply regulates business behavior and contains a scienter requirement.... It is a criminal law that contains no mens rea requirement... and infringes on constitutionally protected rights.” Ante, at 55 (internal quotation marks omitted). (Poor Alexander Hamilton, who has seen his “commercial republic” devolve, in the eyes of the plurality, at least, into an “indolent republic,” see The Federalist No. 6, p. 56; No. 11, pp. 84-91 (C. Rossiter ed. 1961).)
Of course every activity, even scratching one’s head, can be called a “constitutional right” if one means by that term nothing more than the fact that the activity is covered (as all are) by the Equal Protection Clause, so that those who engage in it cannot be singled out without “rational basis.” See FCC v. Beach Communications, Inc., 508 U. S. 307, 313 (1993). But using the term in that sense utterly impoverishes our constitutional discourse. We would then need a new term for those activities' — such as political speech or religious worship — that cannot be forbidden even with rational basis.
The plurality tosses around the term “constitutional, right” in this renegade sense, because there is not the slightest evidence for the existence of a genuine constitutional right to loiter. Justice Thomas recounts the vast historical tradition of criminalizing the activity. Post, at 102-106 (dissenting opinion). It is simply not maintainable that the right to loiter would have been regarded as an essential attribute of liberty at the time of the framing or at the time of adoption of the Fourteenth Amendment. For the plurality, however, the historical practices of our people are nothing more than a speed bump on the road to the “right” result. Its opinion blithely proclaims: “Neither this history nor the scholarly *85compendia in Justice Thomas’ dissent, [ibid.,] persuades us that the right to engage in loitering that is entirely harmless in both purpose and effect is not a part of the liberty protected by the Due Process Clause.” Ante, at 54, n. 20. The entire practice of using the Due Process Clause to add judicially favored rights to the limitations upon democracy set forth in the Bill of Rights (usually under the rubric of so-called “substantive due process”) is in my view judicial usurpation. But we have, recently at least, sought to limit the damage by tethering the courts’ “right-making” power to an objective criterion. In Washington v. Glucksberg, 521 U. S. 702, 720-721 (1997), we explained our “established method” of substantive due process analysis: carefully and narrowly describing the asserted right, and then examining whether that right is manifested in “[ojur Nation’s history, legal traditions, and practices.” See also Collins v. Harker Heights, 503 U. S. 115, 125-126 (1992); Michael H. v. Gerald D., 491 U. S. 110, 122-123 (1989); Moore v. East Cleveland, 431 U. S. 494, 502-503 (1977). The plurality opinion not only ignores this necessary limitation, but it leaps far beyond any substantive-due-process atrocity we have ever committed, by actually placing the burden of proof upon the defendant to establish that loitering is not a “fundamental liberty.” It never does marshal any support for the proposition that loitering is a constitutional right, contenting itself with a (transparently inadequate) explanation of why the historical record of laws banning loitering does not positively contradict that proposition,6 and the (transparently erroneous) assertion that the city of Chicago appears to have conceded the *86point.7 It is enough for the Members of the plurality that “history ... [fails to] persuad[e] us that the right to engage in loitering that is entirely harmless in both purpose and effect is not a part of the liberty protected by the Due Process Clause,” ante, at 54, n. 20 (emphasis added); they apparently think it quite unnecessary for anything to persuade them that it is.8
It would be unfair, however, to criticize the plurality’s failed attempt to establish that loitering is a constitutionally *87protected right while saying nothing of the concurrences. The plurality at least makes an attempt. The concurrences, on the other hand, make no pretense at attaching their broad “vagueness invalidates” rule to a liberty interest. As far as appears from Justice O’Connor’s and Justice Breyer’s opinions, no police officer may issue any order, affecting any insignificant sort of citizen conduct (except, perhaps, an order addressed to the unprotected class of “gang members”) unless the standards for the issuance of that order are precise. No modern urban society — and probably none since London got big enough to have sewers — could function under such a rule. There are innumerable reasons why it may be important for a constable to tell a pedestrian to “move on” — and even if it were possible to list in an ordinance all of the reasons that are known, many are simply unpredictable. Hence the (entirely reasonable) Rule of the city of New York which reads: “No person shall fail, neglect or refuse to comply with the lawful direction or command of any Poliee Officer, Urban Park Ranger, Parks Enforcement Patrol Officer or other [Parks and Recreation] Department employee, indicated verbally, by gesture or otherwise.” 56 RCNY § 1-03(c)(1) (1996). It is one thing to uphold an “as-applied” challenge when a pedestrian disobeys such an order that is unreasonable — or even when a pedestrian asserting some true “liberty” interest (holding a political rally, for instance) disobeys such an order that is reasonable but unexplained. But to say that such a general ordinance permitting “lawful orders” is void in all its applications demands more than a safe and orderly society can reasonably deliver.
Justice Kennedy apparently recognizes this, since he acknowledges that “some poliee commands will subject a citizen to prosecution for disobeying whether or not the citizen knows why the order is given,” including, for example, an order “tell[ing] a pedestrian not to enter a building” when the reason is “to avoid impeding a rescue team.” Ante, at 69 (opinion concurring in part and concurring in judgment). *88But his only explanation of why the present interference with the “right to loiter” does not fall within that permitted scope of action is as follows: “The predicate of an order to disperse is not, in my view, sufficient to eliminate doubts regarding the adequacy of notice under this ordinance.” Ibid. I have not the slightest idea what this means. But I do understand that the followup explanatory sentence, showing how this principle invalidates the present ordinance, applies equally to the rescue-team example that Justice Kennedy thinks is constitutional — as is demonstrated by substituting for references to the facts of the present case (shown in italics) references to his rescue-team hypothetical (shown in brackets): “A citizen, while engaging in a wide array of innocent conduct, is not likely to know when he may be subject to a dispersal order [order not to enter a building] based on the officer’s own knowledge of the identity or affiliations of other persons with whom the citizen is congregating [what is going on in the building]; nor may the citizen be able to assess what an officer might conceive to be the citizen’s lack of an apparent purpose [the impeding of a rescue team].” Ante, at 69-70.
Ill
I turn next to that element of the plurality’s facial-challenge formula which consists of the proposition that this criminal ordinance contains no mens rea requirement. The first step in analyzing this proposition is to determine what the actus reus, to which that mens rea is supposed to be attached, consists of. The majority believes that loitering forms part of (indeed, the essence of) the offense, and must be proved if conviction is to be obtained. See ante, at 47, 50-51, 53-55, 57-59, 60-61, 62-63 (plurality and majority opinions); ante, at 65, 66, 68 (O’Connor, J., concurring in part and concurring in judgment); ante, at 69-70 (Kennedy, J., concurring in part and concurring in judgment); ante, at 72-73 (Breyer, J., concurring in part and concurring in judgment). That is not what the ordinance provides. The *89only part of the ordinance that refers to loitering is the portion that addresses, not the punishable conduct of the defendant, but what the police officer must observe before he can issue an order to disperse; and what he must observe is carefully defined in terms of what the defendant appears to be doing, not in terms of what the defendant is actually doing. The ordinance does not require that the defendant have been loitering (i e., have been remaining in one place with no purpose), but rather that the police officer have observed him remaining in one place without any apparent purpose. Someone who in fact has a genuine purpose for remaining where he is (waiting for a Mend, for example, or waiting to hold up a bank) can be ordered to move on (assuming the other conditions of the ordinance are met), so long as his remaining has no apparent purpose. It is likely, to be sure, that the ordinance will come down most heavily upon those who are actually loitering (those who really have no purpose in remaining where they are); but that activity is not a condition for issuance of the dispersal order.
The only act of a defendant that is made punishable by the ordinance — or, indeed, that is even mentioned by the ordinance — is his failure to "promptly obey5" an order to disperse. The question, then, is whether that actus reus must be accompanied by any wrongful intent — and of course it must. As the Court itself describes the requirement, “a person must disobey the officer’s order.” Ante, at 47 (emphasis added). No one thinks a defendant could be successfully prosecuted under the ordinance if he did not hear the order to disperse, or if he suffered a paralysis that rendered his compliance impossible. The willful failure to obey a police order is wrongful intent enough.
IV
Finally, I address the last of the three factors in the plurality’s facial-challenge formula: the proposition that the ordinance is vague. It is not. Even under the ersatz over-*90breadth standard applied in Kolender v. Lawson, 461 U. S. 352, 358, n. 8 (1983), which allows facial challenges if a law reaches “a substantial amount of constitutionally protected conduct,” respondents’ claim fails because the ordinance would not be vague in most or even a substantial number of applications. A law is unconstitutionally vague if its lack of definitive standards either (1) fails to apprise persons of ordinary intelligence of the prohibited conduct, or (2) encourages arbitrary and discriminatory enforcement. See, e. g., Grayned v. City of Rockford, 408 U. S. 104, 108 (1972).
The plurality relies primarily upon the first of these aspects. Since, it reasons, “the loitering is the conduct that the ordinance is designed to prohibit,” and “an officer may issue an order only after prohibited conduct has already occurred,” ante, at 58, 59, the order to disperse cannot itself serve “to apprise persons of ordinary intelligence of the prohibited conduct.” What counts for purposes of vagueness analysis, however, is not what the ordinance is “designed to prohibit,” but what it actually subjects to criminal penalty. As discussed earlier, that consists of nothing but the refusal to obey a dispersal order, as to which there is no doubt of adequate notice of the prohibited conduct. The plurality’s suggestion that even the dispersal order itself is unconstitutionally vague, because it does not specify how far to dispersed), see ante, at 59, scarcely requires a response.9 If it were true, it would render unconstitutional for vagueness many of the Presidential proclamations issued under that provision of the United States Code which requires the Pres*91ident, before using the militia or the Armed Forces for law enforcement, to issue a proclamation ordering the insurgents to disperse. See 10 U. S. C. §334. President Eisenhower’s proclamation relating to the obstruction of court-ordered enrollment of black students in public schools at Little Rock, Arkansas, read as follows: “I . . . command all persons engaged in such obstruction of justice to cease and desist therefrom, and to disperse forthwith.” Presidential Proclamation No. 3204,3 CFR 132 (1954-1958 Comp.). See also Presidential Proclamation No. 3645, 3 CFR 103 (1964-1965 Comp.) (ordering those obstructing the civil rights march from Selma to Montgomery, Alabama, to “disperse... forthwith”). See also Boos v. Barry, 485 U. S. 312, 331 (1988) (rejecting overbreadth/vagueness challenge to a law allowing police officers to order congregations near foreign embassies to disperse); Cox v. Louisiana, 379 U. S. 536, 551 (1965) (rejecting vagueness challenge to the dispersal-order prong of a breaeh-of-the-peaee statute and describing that prong as “narrow and specific”).
For its determination of unconstitutional vagueness, the Court relies secondarily — and Justice O’Connor’s and Justice Breyer’s concurrences exclusively — upon the second aspect of that doctrine, which requires sufficient specificity to prevent arbitrary and discriminatory law enforcement. See ante, at 60 (majority opinion); ante, at 65-66 (O’Connor, J., concurring in part and concurring in judgment); ante, at 72 (Breyer, J., concurring in part and concurring in judgment). In discussing whether Chicago’s ordinance meets that requirement, the Justices in the majority hide behind an artificial construct of judicial restraint. They point to the Supreme Court of Illinois’ statement that the “apparent purpose” standard “provides absolute discretion to police officers to decide what activities constitute loitering,” 177 Ill. 2d 440, 457, 687 N. E. 2d 53, 63 (1997), and protest that it would be wrong to construe the language of the ordinance more narrowly than did the State’s highest court. Ante, at *9261, 63 (majority opinion); ante, at 68 (O’Connor, J., concurring in part and concurring in judgment). The “absolute discretion” statement, however, is nothing more than the Illinois Supreme Court’s characterization of what the language achieved — after that court refused (as I do) to read in any limitations that the words do not fairly contain. It is not a construction of the language (to which we are bound) but a legal conclusion (to which we most assuredly are not bound).
The criteria for issuance of a dispersal order under the Chicago ordinance could hardly be clearer. First, the law requires police officers to “reasonably believ[e]” that one of the group to which the order is issued is a “criminal street gang member.” This resembles a probable-cause standard, and the Chicago Police -Department’s General Order 92-4 (1992) — promulgated to govern enforcement of the ordinance — makes the probable-cause requirement explicit.10 Under the Order, officers must have probable cause to believe that an individual is a member of a criminal street gang, to be substantiated by the officer’s “experience and knowledge of the alleged offenders” and by “specific, documented and reliable information” such as reliable witness testimony or an individual’s admission of gang membership or display of distinctive colors, tattoos, signs, or other markings worn by members of particular criminal street gangs. App. to Pet. for Cert. 67a-69a, 71a-72a.
Second, the ordinance requires that the group be “remain-ting] in any one place with no apparent purpose.” Justice O’Connor’s assertion that this applies to “any person stand*93ing in a public place,” ante, at 66, is a distortion. The ordinance does not apply to “standing,” but to “remain[ing]”— a term which in this context obviously means “[to] endure or persist,” see American Heritage Dictionary 1525 (1992). There may be some ambiguity at the margin, but “remaining] in one place” requires more than a temporary stop, and is clear in most of its applications, including all of those represented by the facts surrounding respondents’ arrests described supra, at 82-83.
As for the phrase “with no apparent purpose”: Justice O’Connor again distorts this adjectival phrase, by separating it from the word that it modifies. “[A]ny person standing on the street,” her concurrence says, “has a general ‘purpose’ — even if it is simply to stand,” and thus “the ordinance permits police officers to choose which purposes are permissible” Ante, at 66. But Chicago police officers enforcing the ordinance are not looking for people with no apparent purpose (who are regrettably in oversupply); they are looking for people who “remain in any one place with no apparent purpose” — that is, who remain there without any apparent reason for remaining there. That is not difficult to perceive.11
The Court’s attempt to demonstrate the vagueness of the ordinance produces the following peculiar statement: “The ‘no apparent purpose’ standard for making [the decision to *94issue an order to disperse] is inherently subjective because its application depends on whether some purpose is ‘apparent’ to the officer on the scene.” Ante, at 62. In the Court’s view, a person’s lack of any purpose in staying in one location is presumably an objective factor, and what the ordinance requires as a condition of an order to disperse— the absence of any apparent purpose — is a subjective factor. This side of the looking glass, just the opposite is true.
Elsewhere, of course, the Court acknowledges the clear, objective commands of the ordinance, and indeed relies upon them to paint it as unfair:
“In any public place in the city of Chicago, persons who stand or sit in the company of a gang member may be ordered to disperse unless their purpose is apparent. The mandatory language in the enactment directs the police to issue an order without first making any inquiry about their possible purposes. It matters not whether the reason that a gang member and his father, for example, might loiter near Wrigley Field is to rob an unsuspecting fan or just to get a glimpse of Sammy Sosa leaving the ballpark; in either event, if their purpose is not apparent to a nearby police officer, she may — indeed, she ‘shall’ — order them to disperse.” Ante, at 60.
Quite so. And the fact that this clear instruction to the officers “reach[es] a substantial amount of innocent conduct,” ibid., would be invalidating if that conduct were constitutionally protected against abridgment, such as speech or the practice of religion. Remaining in one place is not so protected, and so (as already discussed) it is up to the citizens of Chicago — not us — to decide whether the tradeoff is worth it.12
*95Justice Breyer’s concurrence tries to perform the impossible feat of affirming our unquestioned rule that a criminal statute that is so vague as to give constitutionally inadequate notice to some violators may nonetheless be enforced against those whose conduct is clearly covered, see ante, at 71-72, citing Parker v. Levy, 417 U. S. 733 (1974), while at the same time asserting that a statute which “delegates too much discretion to a police officer” is invalid in all its applications, even where the officer uses his discretion “wisely,” ante, at 71. But the vagueness that causes notice to be inadequate is the very same vagueness that causes “too much discretion” to be lodged in the enforcing officer. Put another way: A law that gives the policeman clear guidance in all cases gives the public clear guidance in all cases as well. Thus, what Justice Breyer gives with one hand, he takes away with the other. In his view, vague statutes that nonetheless give adequate notice to some violators are not unenforceable against those violators because of inadequate notice, but are unenforceable against them “because the policeman enjoys too much discretion in every case,” ibid. This is simply contrary to our ease law, including Parker v. Levy, supra.13
*96V
The plurality points out that Chicago already has several laws that reach the intimidating and unlawful gang-related conduct the ordinance was directed at. See ante, at 52, n. 17. The problem, of course, well recognized by Chicago’s city council, is that the gang members cease their intimidating and unlawful behavior under the watchful eye of police officers, but return to it as soon as the police drive away. The only solution, the council concluded, was to clear the streets of congregations of gangs, their drug customers, and their associates.
Justice O’Connor’s concurrence proffers the same empty solace of existing laws useless for the purpose at hand, see ante, at 67,67-68, but seeks to be helpful by suggesting some measures similar to this ordinance that would be constitutional. It says that Chicago could, for example, enact a law that “directly prohibit[s] the presence of a large collection of obviously brazen, insistent, and lawless gang members and hangers-on on the public ways, that intimidates residents.” Ante, at 67 (internal quotation marks omitted). (If the majority considers the present ordinance too vague, it would be fun to see what it makes of “a large collection of obviously brazen, insistent, and lawless gang members.”) This prescription of the concurrence is largely a quotation from the plurality — which itself answers the concurrence’s suggestion that such a law would be helpful by pointing out that the city already “has several laws that serve this purpose.” Ante, at 52, n. 17 (plurality opinion) (citing extant laws against “intimidation,” “streetgang criminal drug conspiracy,” and “mob action”). The problem, again, is that the intimidation and lawlessness do not occur when the police are in sight.
*97Justice O’Connor’s concurrence also proffers another cure: “If the ordinance applied only to persons reasonably believed to be gang members, this requirement might have cured the ordinance’s vagueness because it would have directed the manner in which the order was issued by specifying to whom the order could be issued.” Ante, at 66 (the Court agrees that this might be a cure, see ante, at 62). But the ordinance already specifies to whom the order can be issued: persons remaining in one place with no apparent purpose in the company of a gang member. And if “remaining] in one place with no apparent purpose” is so vague as to give the police unbridled discretion in controlling the conduct of nongang members, it surpasses understanding how it ceases to be so vague when applied to gang members alone. Surely gang members cannot be decreed to be outlaws, subject to the merest whim of the police as the rest of us are not.
* * *
The fact is that the present ordinance is entirely clear in its application, cannot be violated except with full knowledge and intent, and vests no more discretion in the police than innumerable other measures authorizing police orders to preserve the public peace and safety. As suggested by their tortured analyses, and by their suggested solutions that bear no relation to the identified constitutional problem, the majority’s real quarrel with the Chicago ordinance is simply that it permits (or indeed requires) too much harmless conduct by innocent citizens to be proscribed. As Justice O’Connor’s concurrence says with disapprobation, “the ordinance applies to hundreds of thousands of persons who are not gang members, standing on any sidewalk or in any park, coffee shop, bar, or other location open to the public.” Ante, at 66 (internal quotation marks omitted).
But in our democratic system, how much harmless conduct to proscribe is not a judgment to be made by the courts. So long as constitutionally guaranteed rights are not affected, *98and so long as the proscription has a rational basis, all sorts of perfectly harmless activity by millions of perfectly innocent people can be forbidden — riding a motorcycle without a safety helmet, for example, starting a campfire in a national forest, or selling a safe and effective drug not yet approved by the Food and Drug Administration. All of these acts are entirely innocent and harmless in themselves, but because of the risk of harm that they entail, the freedom to engage in them has been abridged. The citizens of Chicago have decided that depriving themselves of the freedom to “hang out” with a gang member is necessary to eliminate pervasive gang crime and intimidation — and that the elimination of the one is worth the deprivation of the other. This Court has no business second-guessing either the degree of necessity or the fairness of the trade.
I dissent from the judgment of the Court.
In other words, a facial attack, since it requires unconstitutionality in all circumstances, necessarily presumes that the litigant presently before the court would be able to sustain an as-applied challenge. See Hoffman Estates v. Flipside, Hoffman Estates, Inc., 455 U. S. 489, 495 (1982) (“A plaintiff who engages in some conduct that is clearly proscribed cannot complain of the vagueness of the law as applied to the conduct of others. A court should therefore examine the complainant’s conduct before analyzing other hypothetical applications of the law”); Parker v. Levy, 417 U. S. 733, 756 (1974) (“One to whose conduct a statute clearly applies may not successfully challenge it for vagueness”).
The plurality asserts that in United States v. Salerno, 481 U. S. 739 (1987), which I discuss in text immediately following this footnote, the Court “entertained” a facial challenge even though “the defendants . . . did not claim that the statute was unconstitutional as applied to them.” Ante, at 55, n. 22. That is not so. The Court made it absolutely dear in Salerno that a facial challenge requires the assertion that “no set of circumstances exists under which the Act would be valid,” 481 U. S., at 745 (emphasis added). The footnoted statement upon which the plurality relies (“Nor have respondents claimed that the Act is unconstitutional because of the way it was applied to the particular facts of their case,” id., at 745, n. 3) was obviously meant to convey the fact that the defendants were not making, in addition to their facial challenge, an alternative as-applied challenge — i. e., asserting that even if the statute was not unconstitutional in all its applications it was at least unconstitutional in its particular application to them.
Salerno, a criminal case, repudiated the Court’s statement in Kolender v. Lawson, 461 U. S. 352, 359, n. 8 (1983), to the effect that a facial challenge to a criminal statute could succeed “even when [the statute] could conceivably have had some valid application.” Kolender seems to have confused the standard for First Amendment overbreadth challenges with the standard governing facial challenges on all other grounds. See ibid. (citing the Court’s articulation of the standard for First Amendment over-breadth challenges from Hoffman Estates, supra, at 494). As Salerno noted, supra, at 745, the overbreadth doctrine is a specialized exception to the general rule for facial challenges, justified in light of the risk that an overbroad statute will chill free expression. See, e. g., Broadrick v. Oklahoma, 413 U. S. 601, 612 (1973).
The plurality asserts that the Salerno standard for facial challenge “has never been the decisive factor in any decision of this Court.” Ante, at 55, n. 22. It means by that only this: in rejecting a facial challenge, the Court has never contented itself with identifying only one situation in which the challenged statute would be constitutional, but has mentioned several. But that is not at all remarkable, and casts no doubt upon the validity of the principle that Salerno and these many other cases enunciated. It is difficult to conceive of a statute that would be constitutional in only a single application — and hard to resist mentioning more than one.
The plurality contends that it does not matter whether the Salerno standard is federal law, since facial challenge is a species of third-party standing, and federal limitations upon third-party standing do not apply in an appeal from a state decision which takes a broader view, as the Illinois Supreme Court’s opinion did here. Ante, at 55-56, n. 22. This is quite wrong. Disagreement over the Salerno rule is not a disagreement over the “standing” question whether the person challenging the statute can raise the rights of third parties: under both Salerno and the plurality’s rule he can. The disagreement relates to how many third-party rights he must prove to be infringed by the statute before he can win: Salerno says “all” (in addition to his own rights), the plurality says “many.” That is not a question of standing but of substantive law. The notion that, if Salerno is the federal rule (a federal statute is not totally invalid unless it is invalid in all its applications), it can be altered by a state court (a federal statute is totally invalid if it is invalid in many of its applications), and that that alteration must be accepted by the Supreme Court of the United States is, to put it as gently as possible, remarkable.
See, e. g., Abdullah, v. Commissioner of Ins. of Commonwealth of Mass., 84 F. 3d 18, 20 (CA1 1996); Deshawn E. v. Safir, 156 F. 3d 340, 347 (CA2 1998); Artway v. Attorney Gen. of State of N. J., 81 F. 3d 1235, 1252, n. 13 (CA3 1996); Manning v. Hunt, 119 F. 3d 254, 268-269 (CA4 1997); Causeway Medical Suite v. Ieyoub, 109 F. 3d 1096, 1104 (CA5), cert. de*81nied, 522 U. S. 943 (1997); Aronson v. Akron, 116 F. 3d 804, 809 (CA6 1997); Government Suppliers Consolidating Servs., Inc. v. Bayh, 975 F. 2d 1267, 1283 (CA7 1992), cert. denied, 506 U. S. 1053 (1993); Woodis v. Westark Community College, 160 F. 3d 435, 438-439 (CA8 1998); Roulette v. Seattle, 97 F. 3d 300, 306 (CA9 1996); Public Lands Council v. Babbitt, 167 F. 3d 1287, 1293 (CA10 1999); Dimmitt v. Clearwater, 985 F. 2d 1565, 1570-1571 (CA11 1993); Time Warner Entertainment Co. v. FCC, 93 F. 3d 957, 972 (CADC 1996).
With apologies for taking creative license with the work of Messrs. Bernstein, Sondheim, and Laments. West Side Story, copyright 1959.
The plurality’s explanation for ignoring these laws is that many of them carried severe penalties and, during the Reconstruction era, they had “harsh consequences on African-American women and children.” Ante, at 54, n. 20. Those severe penalties and those harsh consequences are certainly regrettable, but they in no way lessen (indeed, the harshness of penalty tends to increase) the capacity of these laws to prove that loitering was never regarded as a fundamental liberty.
Ante, at 53, n. 19. The plurality bases its assertion of apparent concession upon a footnote in Part I of petitioner’s brief which reads: “Of eourse, laws regulating social gatherings affect a liberty interest, and thus are subject to review under the rubric of substantive due process .... We address that doctrine in Part II below.” Brief for Petitioner 21-22, n. 13. If a careless reader were inclined to confuse the term “social gatherings” in this passage with “loitering,” his confusion would be eliminated by pursuing the reference to Part II of the brief, which says, in its introductory paragraph: “[A]s we explain below, substantive due process does not support the court’s novel holding that the Constitution secures the right to stand still on the public way even when one is not engaged in speech, assembly, or other conduct that enjoys affirmative constitutional protection.” Id., at 39.
The plurality says, ante, at 64, n. 35, that since it decides the ease on the basis of procedural due process rather than substantive due process, I am mistaken in analyzing its opinion “under the framework for substantive due process set out in Washington v. Glucksberg.” Ibid. But I am not analyzing it under that framework. I am simply assuming that when the plurality says (as an essential part of its reasoning) that “the right to loiter for innocent purposes is ... a part of the liberty protected by the Due Process Clause” it does not believe that the same word (“liberty”) means one thing for purposes of substantive due process and something else for purposes of procedural due process. There is no authority for that startling proposition. See Board of Regents of State Colleges v. Roth, 408 U. S. 564, 572-575 (1972) (rejecting procedural-due-process claim for lack of “liberty” interest, and citing substantive-due-pro cess cases).
The plurality’s opinion seeks to have it both ways, invoking the Fourteenth Amendment's august protection of “liberty” in defining the standard of certainty that it sets, but then, in identifying the conduct protected by that high standard, ignoring our extensive case law defining “liberty,” and substituting, instead, all “harmless and innocent” conduct, ante, at 58.
I call it a “suggestion” because the plurality says only that the terms of the dispersal order “compound the inadequacy of the notice,” and acknowledges that they “might not render the ordinance unconstitutionally vague if the definition of the forbidden conduct were dear.” Ante, at 59, 59-60. This notion that a prescription (“Disperse!”) which is itself not unconstitutionally vague can somehow contribute to the unconstitutional vagueness of the entire scheme is full of mystery — suspending, as it does, the metaphysical prindple that nothing can confer what it does not possess (nemo dat qui non habet).
“Administrative interpretation and implementation of a regulation are .. . highly relevant to our [vagueness] analysis, for ‘[i]n evaluating a facial challenge to a state law, a federal court must... consider any limiting construction that a state court or enforcement agency has proffered.’ ” Ward v. Rock Against Racism, 491 U. S. 781, 795-796 (1989) (emphasis added) (quoting Hoffman Estates, 455 U. S., at 494, n. 5). See also id., at 504 (administrative regulations “will often suffice to clarify a standard with an otherwise uncertain scope”).
Justice Breyer asserts that “one always has some apparent purpose,” so that the policeman must “interpret the words ‘no apparent purpose’ as meaning ‘no apparent purpose except for....’ ” Ante, at 70. It is simply not true that “one always has some apparent purpose” — and especially not true that one always has some apparent purpose in remaining at rest, for the simple reason that one often (indeed, perhaps usually) has no actual purpose in remaining at rest. Remaining at rest will be a person’s normal state, unless he has a purpose which causes him to move. That is why one frequently reads of a person’s “wandering aimlessly” (which is worthy of note) but not of a person’s “sitting aimlessly” (which is not remarkable at all). And that is why a synonym for “purpose” is “motive”: that which causes one to move.
The Court speculates that a police officer may exercise his discretion to enforce the ordinance and direct dispersal when (in the Court’s view) the ordinance is inapplicable — viz., where there is an apparent purpose, but it is an unlawful one. See ante, at 62. No one in his right mind *95would read the phrase “without any apparent purpose” to mean anything other than “without any apparent lawful purpose.” The implication that acts referred to approvingly in statutory language are “lawful” acts is routine. The Court asserts that the Illinois Supreme Court has forced it into this interpretive inanity because, since it “has not placed any limiting construction on the language in the ordinance, we must assume that the ordinance means what it says . . . .” Ante, at 63. But the Illinois Supreme Court did not mention this particular interpretive issue, which has nothing to do with giving the ordinance a “limiting” interpretation, and everything to do with giving it its ordinary legal meaning.
The opinion that Justice Breyer relies on, Coates v. Cincinnati, 402 U. S. 611 (1971), discussed ante, at 72-73, did not say that the ordinance there at issue gave adequate notice but did not provide adequate standards for the police. It invalidated that ordinance on both inadequate-notice and inadequate-enforcement-standard, grounds, because First Amendment rights were implicated. It is common ground, however, that *96the present case does not implicate the First Amendment, see ante, at 52-53 (plurality opinion); ante, at 72 (Breyee, J., concurring in part and concurring in judgment).