There Is No First Amendment Overbreadth (but There Are Vague First Amendment Doctrines); Prior Restraints Aren't "Prior"; and "As Applied" Challenges Seek Judicial Statutory Amendments
Alexander, Larry, Constitutional Commentary
In this short article I hope to clarify three doctrines that have produced enormous confusion among lawyers, judges, and academic commentators. These are the doctrines of First Amendment overbreadth, prior restraint, and as-applied (as opposed to facial) challenges. My purpose is entirely analytical, although analytical clarity will undoubtedly have normative implications, some of which I shall briefly note.
I. FIRST AMENDMENT OVERBREADTH
My title correctly suggests that there is no such thing as first amendment overbreadth despite its apparently well-established status as a first amendment doctrine. Indeed, given the conception of overbreadth--a statute is overbroad if it has some unconstitutional applications--that is utilized in what is taken to be first amendment overbreadth, there is no such thing as overbreadth in any constitutional domain.
To see this, consider a hypothetical law that surely exemplifies first amendment overbreadth if any law ever does: "No person shall speak, write, or through any other medium seek to communicate any idea to any other person." Surely, this hypothetical law is unconstitutionally overbroad, is it not? For a substantial number of its possible applications--though not all, as with fighting words, malicious defamations, child pornography, incitements to imminent lawless action, etc.--are constitutionally immune from prohibition by virtue of the First Amendment. (1)
But now consider this hypothetical amendment to my hypothetical overbroad law: "This law shall only apply to the extent that its application is constitutionally permitted." If the law were so amended, would it now be overbroad? It is hard to see how it would be, as by its terms it now has no unconstitutional applications.
But--and here is the key point--the hypothetical statutory amendment is already a part of every statute. For Article VI of the Constitution, which declares the Constitution to be the supreme law of the land, and by direct implication renders legally void any state or federal laws inconsistent with the Constitution, already accomplishes what the hypothetical amendment accomplishes. (2) Or, to put it differently, there would be absolutely no cost in terms of statutory objectives for legislatures to append to all laws "to the extent consistent with the Constitution."
So my hypothetical amendment to my hypothetical overbroad statute, which by hypothesis eliminates the statute's overbreadth, accomplishes nothing that is not already accomplished by Article VI. And this, of course, will be true of any statute. Therefore, there are no overbroad statutes, in the First Amendment domain or elsewhere.
Note, however, that my hypothetical overbroad statute, even if it is not and cannot actually be overbroad, still seems oppressive and capable of chilling free speech. If not because of overbreadth, because of what? The chilling effect is a product of the vagueness of the first amendment tests that distinguish constitutionally-protected speech from speech that can constitutionally be prohibited. Those are the tests that eliminate the overbreadth. But because they are vague, they leave the statute they amend with quite vague margins. Even a citizen well-versed in these first amendment tests--perhaps someone who has gone to law school or has even taught the First Amendment recently in a law school--will not be at all certain in a pretty broad swath of situations whether or not his proposed speech will turn out to be constitutionally protected under extant first amendment doctrines. Therefore, although my hypothetical overbroad statute is not actually overbroad, as truncated by Article VI or a statutory amendment to the same effect, it is quite vague and will likely chill a considerable amount of speech. That is its real first amendment vice.
If my analysis of first amendment overbreadth is correct, and the vagueness of first amendment doctrines is the true source of the chilling effect worry, then this demonstrates the incoherence of some Supreme Court overbreadth decisions. The most prominent one is Gooding v. Wilson. (3) The Supreme Court in Gooding struck down as "overbroad" a Georgia statute punishing offensive speech. The Court implied that it would not have struck down the statute had the Georgia courts limited its application to "fighting words" as defined in Chaplinsky v. New Hampshire." But the implication is nonsensical because the Georgia statute could not be validly applied except in accordance with Chaplinsky. Chaplinsky is already a part of the Georgia statute because the Constitution is already a part of the Georgia statute by operation of Article VI. (5) So the Georgia statute was not--because it could not be--overbroad. On the other hand, if the vice of the statute was its vagueness due to the vagueness of the Chaplinsky "fighting words" test, then that vice could not have been cured had the Georgia court's limited the statute to "'fighting words," the Court's statement to the contrary notwithstanding. The Court's rationale is internally contradictory.
One final point. There are some laws that are, in a sense, overbroad in that they have no valid applications. Or, put differently, the Article VI proviso that is implied in every law obliterates such laws in their entirety. Such laws have no valid applications because they contain an illegitimate predicate for governmental action. (6)
Consider laws that ban flag burning, or speech by Republicans, or speech criticizing the government. These laws have no valid applications, not because every token of flag burning, speech by Republicans, or speech criticizing the government is constitutionally immune to regulation, but because such acts are constitutionally immune to regulation under rules the terms of which refer to certain disfavored ideas or persons.
Consider flag burning. When the Supreme Court held it to be protected under the First Amendment, all it really held is that it could not be banned under laws that by their terms punish certain treatments of the flag. (7) It did not suggest that even expressive uses of the flag could not be punished under, for example, laws prohibiting burning any object for any purpose in the street, or laws banning murder even if committed for expressive purposes by strangling someone with an American flag. Therefore, expressive uses of the American flag are not protected under the First Amendment except from laws that make those expressive uses the predicate for the prohibition. (8)
II. PRIOR RESTRAINTS AREN'T "PRIOR"
Restrictions of speech found in criminal and civil laws are typically analyzed "substantively"--that is, by whether the content of speech is protected by the First Amendment against the particular way in which the law is restricting it, or by whether the time, place, or manner of the speech, whatever its content, is being unduly restricted. But there is one type of regulation of speech that is thought to be special and specially disfavored, the so-called prior restraint of speech. There are two types of regulation that fall into this category. One consists of those requirements that one obtain a license from some agency or person before engaging in the speech activity. Such requirements can be based on the particular content of the proposed speech, such as whether it concerns one's activities with the C.I.A., or whether it is a film that might be pornographic. Or they can be based on the time, place, or manner of the proposed speech, such as whether it involves door-to-door solicitations, or whether it involves a demonstration that could impede traffic. The other type of regulation deemed to be a prior restraint is the judicial injunction or order when directed against the content of speech or its time, place, or manner.
Now notice that neither license requirements nor injunctions are in any sense more prior as restraints than ordinary statutory and common law restrictions of speech. Compare them, for example, with some quotidian non-prior restraint of speech--say, a statute making it a crime to show X-rated movies. Assuming no ex post facto application of the law, which would be a different issue, the requirement that one obtain a license to show an X-rated movie will, it is true, exist prior to the act of showing such a movie, as will an injunction ordering one not to show such a movie. But so does the statute making showing such a movie a crime. Both the statute and the so-called prior restraint are prior in that sense. Moreover, they are also alike in that the punishments they authorize occur after the acts that violate them--showing the X-rated movie in the case of the ordinary statute and the injunction, or showing it without a license in the case of the licensing requirement.
So in what sense are the prior restraints materially different from ordinary statutory and common law restrictions on speech.'? The answer lies not in their "priorness" as restraints but in the consequences of violating them. To be sure, they all threaten violators with punishment. But in the case of ordinary statutory and common law restrictions, if they attempt to restrict speech that is constitutionally protected from such restrictions, the speaker can raise that protection as a defense to the charge of violation. In the case of the statute making it a crime to show X-rated movies, the violator can claim that the statute violates the First Amendment and that therefore his violation of the statute cannot serve as the basis for punishing him. If his claim of constitutional right to show X-rated movies is accepted by the court, the court will throw out the charge against him.
The case is otherwise with the licensing requirement and the injunction. Take the injunction first. If A1 is enjoined from showing X-rated movies, and A1 ignores the injunction and shows such movies, he will be punished for contempt of court.
Moreover, and this is the crucial point, if A1, on being hauled into court on the contempt charge, protests that he has a constitutional right to show X-rated movies, and that the court constitutionally erred in enjoining him from doing so, his protest, even if correct, will not foreclose the court's punishing him for contempt. This is because of a judicially-created doctrine called the "collateral bar rule." That doctrine in essence says that judicial orders from courts that have jurisdiction over the person give rise to an absolute duty of obedience, notwithstanding any constitutional rights to engage in the enjoined conduct, unless and until that order is set aside by the court that issued it or by a higher court on appeal. (9)
The upshot is that while legislative orders can be disobeyed and then challenged on constitutional grounds if suit is brought against those who disobey them, judicial orders cannot be. If A1 is enjoined from showing X-rated movies, he must not do so on pain of certain punishment, unless and until the injunction is withdrawn or set aside on appeal. If Al wants to escape punishment, he must refrain from showing X-rated movies and seek to have the injunction overturned. This is not the case if A1 is only statutorily prohibited from showing such movies.
The vice of judicial "prior restraints" is then one of loss of time, which is a real cost if one's message is time-sensitive. It is the time the speaker must wait before speaking during which he tries to convince some court that his speech is constitutionally protected and thus should never have been enjoined. If he is correct--his speech is constitutionally protected--then it is the time lost in fighting the injunction that he would not have lost had he been prohibited statutorily, a prohibition he could have ignored with impunity.
That lost time is the temporal element in the prior restraint of an injunction. Notice the oddity that arises from the judicial antipathy towards injunctions of speech based on that loss of time. The loss of time stems from the collateral bar rule that elevates judicial orders above legislative orders (statutes) and administrative orders and rules. But the collateral bar rule is itself a judicial creation. If it were eliminated in instances of speech injunctions, the latter would operate like personalized statutes and like statutes could be disobeyed with impunity if the enjoined speech were constitutionally protected. In other words, the loss of time, which makes judicial injunctions of speech constitutionally disfavored, is solely a product of the judicially-crafted collateral bar rule. The judges giveth, then taketh away!
The foregoing analysis also shows why the Supreme Court dissenters were correct and the majority wrong in the case of Vance v. Universal Amusement Co. (10)
In Vance, a court had issued an injunction that by its terms enjoined exhibiting movies that met the constitutional definition of unprotected obscenity. The majority overturned the injunction as an invalid prior restraint of speech. The dissent, however, pointed out that by the terms of the injunction, only films that were not constitutionally protected would constitute violations. Therefore, if Universal Amusement showed any film that it had a constitutional right to show, it would not have violated the injunction. And if it was hauled into court on a contempt charge, it would not need to raise a constitutional defense to that charge, something the collateral bar rule forecloses. Rather, it could use the constitutionally-protected status of the film to deny that it had violated the injunction. And denials of violations are, of course, not foreclosed by the collateral bar rule. (11)
Let me turn now to the other form of prior restraint, the license requirement. This is the classic form of prior restraint, one that dates back to the English requirement of censor approval prior to publication.
There is no collateral bar rule with respect to licensing speech, but there is something similar. For if the licensing scheme is itself valid, then the would-be speaker must wait until he gets the license, either directly from the licensing authority, or if denied by that authority, indirectly from having a court order the licensing authority to grant the license. That is because if the licensing scheme is valid, speaking without the license is punishable, even if, given the constitutionally-protected nature of the speech, the license should have been granted.
Licensing schemes, like injunctions, hurt constitutionally-protected speech most severely when the speaker's message is time-sensitive. This suggests that licensing schemes should be evaluated constitutionally based in part on how time-sensitive the speech that requires a license is likely to be, as well as how expedited the licensing process and judicial reviews of denials are. Courts need not adopt a dichotomous valid or invalid approach to licensing schemes. They could, for example, deem them valid for a certain period of time after which their validity would expire and the speaker would no longer be required to seek the license.
Moreover, the time-sensitivity of the speech is not the only relevant criterion in assessing licensing schemes. Frequently, the government and even the speaker have an interest in assessing the constitutional status of the speech prior to its communication to others. The government, for example, may believe that revelation of constitutionally-unprotected classified information will be irreparably damaging to its legitimate interests and so would like to review any possible constitutionally-unprotected information before it is published. Indeed, the speaker himself will frequently be uncertain whether the classified information is constitutionally protected or unprotected, or whether it will injure vital government interests, something he, like the government, might wish to avoid. Therefore, both he and the government might welcome the prior-to-publication review process of a licensing scheme. And the same would be true of A1 and his X-rated movies. Given the vagueness of the line between constitutionally-unprotected (obscene) X-rated movies and constitutionally-protected X-rated movies, if there is a statute that punishes showing the former, A1 might welcome the presence of a licensing scheme that would determine on which side of the constitutional line his movies are located before he shows them, a scheme that eliminates the otherwise sizable risk of error he would run if he had to determine this for himself. An error on the side of caution would cost him revenue. An error on the other side would cost him a fine or jail time. The licensing scheme is a less risky and less costly alternative, particularly since X-rated movies will probably not have time-sensitive content.
III. "AS APPLIED" CHALLENGES SEEK JUDICIAL STATUTORY AMENDMENTS
When statutes are "facially" challenged, and the challenge is upheld, the statute is declared a legal nullity. Nothing of it remains that can be validly applied in any situation. (It may, of course, remain in the statute books. And if the court that has upheld the facial challenge to it is later reversed, or reverses itself, the statute may be resuscitated without the need for legislative reenactment.)
Facial challenges are easy enough to comprehend. The rule in its entirety is constitutionally infirm. The reasons for this might be of many different types. One already mentioned is that the rule contains an illegitimate predicate for application, such as when it discriminates based on viewpoint or speaker (or on a multitude of other forbidden grounds, such as race, ethnicity, gender, alienage, age, disability, religion, etc. etc.). Another, also already mentioned, is when the rule is overbroad because of a vague and chilling constitutionally-imposed boundary. And there are still others, as when the law tramples on fundamental rights involving procreation and sexuality, or when local law is preempted by state law, or state law by federal law.
If constitutionally void laws subject to facial challenges are easy enough to comprehend, what about laws that are only unconstitutional "as applied"? What is their nature, and why do courts often say that as applied constitutional challenges are preferred to facial ones?
Think about as applied challenges this way: Suppose rule R forbids doing act A in circumstances C. Suppose further that C covers a range of possible circumstances ([C.sub.1], [C.sub.2], [C.sub.3]... [C.sub.N]). And suppose further that applying R to A in some of those circumstances--say, [C.sub.4] and [C.sub.5]--would violate the constitutional rights of those who do A in those circumstances. Finally, suppose that the rule [C.sup./]--"it is forbidden to A in C except in C4 and [C.sub.5]"--has no unconstitutional applications. Then an as applied challenge to R is a challenge to R that should be upheld only in circumstances C4 and C5, with the result that R will now be interpreted to be identical to [R.sub./].
Notice, however, that if a court strikes down only those applications of R that occur in C4 in C5, it has actually amended R. It has severed R in C4 and C5 and left an amputated R that is now "R, except in [C.sub.4] and [C.sub.5]," or [R.sub./]. (Or rather, per the analysis in Section I, the Court has recognized that the Constitution itself limits R to [R.sup./], and the Court has left [R.sup./] in force.) (12) Presumably then, a court should only entertain an as applied challenge in cases where it believes the legislature intends the statute to be severable, and severable at the particular joints at which the court is asked to sever it. In the case of R, that means that the court should only grant the as applied challenge if it thinks the legislature would prefer [R.sup./] to nothing, R itself being constitutionally ineligible.
Thus, an as applied challenge to a statute is a call for the statute's amendment by the courts. It is appropriate when some applications of the statute are unconstitutional, the statute amended to remove those applications would be constitutionally valid had it been enacted in that truncated form--it is not fatally underinclusive and does not contain an improper predicate for application--and the enacting legislature is presume to have preferred the amended, truncated statute to no statute at all. (13)
The first amendment doctrines of overbreadth, prior restraint, and as applied versus facial challenge are, as elaborated by courts and commentators, confusing. I have tried to clear up the confusion, not by clarifying those doctrines as they currently exist, but rather by showing what those doctrines would look like if they were analytically coherent and normatively attractive.
In a recent article, Nicholas Rosenkranz offers a novel and somewhat radical understanding of constitutional duties and rights. He asserts that some constitutional duties are imposed on Congress in its lawmaking capacity, and some are imposed on the executive in its application of laws to individuals. (14) His leading example of a constitutional provision that speaks to Congress in its lawmaking capacity is the First Amendment and its command, "Congress shall make no law .... "When Congress passes a law abridging freedom of speech, establishing a religion, or prohibiting the free exercise of religion, then according to Rosenkranz, Congress has committed a constitutional violation at the moment it passes the law, before the law is ever enforced against anyone. He is a bit vague about whose rights have thereby been violated--who has "standing" to raise the constitutional violation in court. (15) But he is clear that Congress has violated the First Amendment when it passes the law-presumably even if the law is never enforced. (Query: Would it violate the First Amendment if Congress, in the law, states that it should not be enforced?)
There are several corollaries to Rosenkranz's principal axiom. First, if a law is unconstitutional on the day it is enacted, it cannot become constitutional because of changes in the constitutional facts. Perhaps more importantly, the converse is also true: a law that is constitutional when enacted because supported by the underlying constitutional facts is always constitutional even if the supporting constitutional facts have disappeared--that is, even if the same law enacted today would be unconstitutional. (16)
The second corollary that Rosenkranz draws from his principal axiom--and the one that bears most directly on my article--is that laws are either constitutional or unconstitutional "on their face" and are never unconstitutional "as applied." (17) A law that is constitutional on its face might be applied unconstitutionally by the executive, but that is a constitutional violation by the executive alone. If the law itself calls for that unconstitutional application, then presumably the law is unconstitutional in toto and on its face.
This second corollary conflicts with the analysis of "as applied" unconstitutionality that I have given. I have suggested that some statutes have parts--sub-rules--that are unconstitutional but that are severable. On this view, statutes that do not contain illegitimate predicates can be decomposed into subrules, some of which might be unconstitutional. If the remaining sub-rules are constitutionally permissible--and if the legislature would prefer retaining these permissible sub-rules over total invalidation of the statute--then the statute is only unconstitutional "as applied," unless the "as applied" technique would leave the statute with a margin that is too vague (overbreadth, on my analysis). I cannot find any reason for rejecting the possibility that statutes might be partially unconstitutional and severable, either in Rosenkranz's lengthy article or anywhere else. So I reject the second corollary, namely that statutes are never unconstitutional as applied, and stand by my analysis of as applied challenges.
For what it's worth, I also reject his first corollary and the principal axiom from which both corollaries are derived. The first corollary--that a postenactment change in constitutional facts cannot affect constitutionality--is bizarre. It entails that Congress has no constitutional obligation to monitor its statutes and repeal them when circumstances change in such a way that were they to enact the same statutes today, the statutes would be unconstitutional. It rests on the odd notion that constitutional adjudication is primarily an investigation into the culpability of the enacting Congress, not an investigation of the law's contemporary effects and the culpability of the non-repealing Congress.
Rosenkranz affirms this bizarre corollary largely because it seems to be entailed by his principal axiom that provisions like the First Amendment speak to Congress when it enacts laws-and only then. But the principal axiom is doubtful. Consider this alternative conception of constitutional provisions like the First Amendment: "Congress shall make no law" is elliptical for "Congress shall have no power to make, as valid law, any 'law'...." Unconstitutional "laws" on this conception are of no legal effect. They only appear to be, but are not actually, laws.
On this conception, the principal form of constitutional adjudication occurs only as a result of application, or anticipated application, by the executive or the courts. That application might be thought of as a tort. If the plaintiff (the affected party) sues the official in tort, and the official defends by claiming the tort is not a tort because it was theorized by a law, the plaintiff would respond that the law, being unconstitutional and a legal nullity, cannot immunize the official. On the other hand, a law that is not enforced is of no constitutional significance.
Rosenkranz's analysis is bold, radical, refreshing, (18) and unconventional. I have a fondness for analyses of this kind. But there is another virtue that it lacks. It is incorrect. In another recent article, Kevin Walsh has recently dissented from the analysis of severability that I have given in section III. (19) Walsh correctly notes that severability calls for judges to make difficult counterfactual determinations regarding what a legislature would have wanted had it realized that its enactments were invalid and therefore unenforceable in some but not all of their applications. Would the legislature have wanted the enactment to be enforced to the full extent it could be validly enforced? Or would it want other valid portions of the enactment--or the entire enactment, or even other parts of the corpus juris--to be unenforced given the unenforceability of a sub-rule of the enactment? (20)
These surely are difficult counterfactuals for judges. Indeed, the problem is difficult even if a legislature explicitly states that an enactment (or group of enactments) is inseverable and should be deemed to no effect if any application is held to be unconstitutional. That is so because if the first case to arise involves a constitutionally valid application, the judge will not know whether the application is authorized until he or she surveys all possible applications and concludes that none of them is unconstitutional.
Walsh believes he has a solution to the difficulties of severability. For Walsh, constitutional judicial review does not involve eliminating statutes, much less parts thereof, but merely applying higher law where higher and lower law conflict. Walsh calls his approach "displacement": The higher law of the Constitution displaces the lower law of statutes where there is, and only insofar as there is, a conflict between them. (21) If an application of the statute is "displaced," the statute remains on the books and in effect unless the state has also enacted a "fallback law," that is, a law specifying what is to be done if an application of a statute is held unconstitutional. (22) In other words, severability would be the result of an explicit legislative directive in a fallback law, not a judicial divining of the legislature's counterfactual intent.
Walsh's "displacement and fall back law" proposal is actually just a conclusive presumption of severability in the absence of a fallback law. Unless the legislature specifies inseverability and its scope in a fallback law, no application of a statute will be struck down merely because some other application is unconstitutional. (23)
A conclusive presumption of severability will make matters easier for the judge in a case that arises under a constitutional sub-rule after another sub-rule of the same statute has been held to be unconstitutional. It will not, however, help the judge when there is a fallback law declaring the enactment's sub-rules to be inseverable if the sub-rule in question is constitutional and no other sub-rule has yet been declared unconstitutional. The judge will still have to canvas the enactment's entire set of sub-rules to see if any one of them is unconstitutional before being able to decide the case at hand.
Moreover, a conclusive presumption of severability will not, by itself, obviate the necessity of looking to hypothetical legislative intent unless it is supplemented by other presumptions. For consider those statutes that are unconstitutional because they offend some norm of equality, whether under the Equal Protection Clause or under dormant commerce clause analysis. The vice in such cases is that a burden or benefit has not been extended equally where the Constitution requires that it be so. Once the court holds the statute unconstitutional, however, there is a question of remedy remaining. Should the benefit be extended to the plaintiff (or the plaintiff be relieved of the burden), or should the benefit be denied to the comparison class (or the burden extended thereto)? One might answer these questions by asking the counterfactual question, "What would the legislature have wanted in lieu of getting its desired unequal distribution?" Or one might establish a conclusive presumption in favor of, say, extending benefits (or removing burdens).
Perhaps the most appropriate remedy would be to excise in toto whatever enactment or portion thereof contains the illegitimate predicate for benefits or burdens. If, for example, a welfare program is enacted that gives $1,000 per month to the poor irrespective of whether they are black or white, Democrat or Republican, Christian or Jew, but then the legislature passes a new law increasing the monthly payments to $1,500 but only for whites, or Democrats, and so on, the courts should strike down the new law but leave the prior law in effect. If, however, the original welfare law gave $1,500 per month to whites but only $1,000 per month to blacks, the appropriate remedy would be to eliminate that entire law and its welfare payments. Mere displacement, however, in the absence of a fallback law, does not by itself point to a remedy. For if a poor black sought the $1,000 monthly payment, the court would have to decide what precisely had been displaced by the Constitution--the entire welfare law or only the $500 extra payment to whites.
In any event, Walsh's proposal may be meritorious as far as it goes, but it does not contradict the analyses of severability I have given. Rather, it proposes a way of determining severability other than by divining the legislature's hypothetical intent.
(1.) Though as we shall see, even that seeming bedrock First Amendment truth must be qualified. See infra at note 8.
(2.) U.S. CONST. art. VI, cl. 2 "This Constitution... shall be the supreme Law of the Land .... ").
(3.) 405 U.S. 518 (1972).
(4.) 315 U.S. 568, 573 (1942).
(5.) See Lawrence A. Alexander, Is There an Overbreadth Doctrine?, 22 SAN DIEGO L. REV. 541,549 n.20 (1985).
(6.) See id. at 544-47.
(7.) Texas v. Johnson, 491 U.S. 397, 419 (1989); United States v. Eichman, 496 U.S. 310, 318-19 (1990).
(8.) In the earlier article I put the point this way:
The Constitution's individual rights provisions by and large do not protect specific conduct per se .... Rather, the Constitution ordinarily limits the types of reasons that government may act upon in regulating conduct. For instance, 'criticizing the government' is not protected conduct viewed in isolation from the various ways government might attempt to regulate 'criticizing the government.' 'Criticizing the government' may be validly-constitutionally-regulated if the criticism is broadcast from a soundtruck at night, and the regulation proscribes the use of soundtrucks at night. 'Criticizing the government' may be validly regulated if the criticism takes place on private property without the owner's consent, and the regulation proscribes trespass. But 'criticizing the government' is not validly regulated if the regulation proscribes, or was motivated by a desire to proscribe, 'criticizing the government.' Now when a statute contains a constitutionally illegitimate predicate of government action, the statute is void and cannot be applied. If for instance a statute proscribes 'picketing,' 'mutilating the flag,' or 'demonstrations by blacks,' the statute cannot be applied as is, nor can it be applied even if narrowed to 'violent picketing,' 'burning the flag in public,' or 'disruptive demonstrations by blacks.' The narrowed statute still contains the illegitimate predicate for governmental action.
Alexander, supra note 5, at 545 (citations omitted). I should point out that the hypothetical statute forbidding all communication is probably an example of a law containing an illegitimate predicate for governmental action and therefore has no constitutionally permissible applications. I should also note that overbreadth invalidation is not an example of "third-party standing," as it is sometimes claimed. An overbroad statute, which I have said is a statute with a vague and chilling constitutional boundary, is a constitutionally infirm statute, just as are statutes that have constitutionally illegitimate predicates for government action. And anyone whose conduct is regulated by a constitutionally infirm rule has standing to object to it.
Put differently, while violent picketing, burning the flag in public, and disruptive demonstrations by blacks are all activities that can be validly proscribed under some statutes--for example, statutes proscribing violence, public burnings, and disruptive demonstrations--violent picketing, burning of the flag, and disruptive demonstrations by blacks do not mark off legitimately regulatable subcategories of legitimately regulatable activities. They are therefore underinclusive with respect to legitimately regulatable activity. The government may ban all violence, but not just violence associated with picketing. The government may ban all public burnings, but not just those that involve the flag. And so forth.
I should add that there are conceptual mysteries that attend the Constitution's focus on rules and their validity as opposed to act tokens and their immunity from regulation. See Larry Alexander, Rules, Rights, Options, and Time, 6 LEGAL THEORY 391 (2000) (discussing the past and future effect of rules and the permissibility of such rules).
(9.) See Walker v. City of Birmingham, 388 U.S. 307, 321 (1967).
(10.) 445 U.S. 308 (1980).
(11.) It is true that the trial of a contempt charge will be before a judge and not a jury, whereas trials of statutory or common law violations will, at the option of the defendant, be before a jury as well as a judge. That distinction, however, has nothing to do with an injunction's being more "prior" than a statute. Nor does it have any obvious implications for the First Amendment. (I thank Steve Smith for pointing out this distinction between injunctions and statutes beyond that of the collateral bar rule.)
(12.) I thank Ted Sichelman for this point.
(13.) For a recent lengthier treatment of this topic, which is basically consistent with my analysis, see Alex Kreit, Making Sense of Facial and As-Applied Challenges, 18 WM. & MARY BILL RTS. J. 657 (2010). Kreit's article also contains references to the earlier efforts to explain the distinction.
(14.) Nicholas Quinn Rosenkranz, The Subjects of the Constitution, 62 STAN. L. REV. 1209 (2010).
(15.) Id. at 1247-48.
(16.) See id. at 1284-86.
(17.) E.g., id. at 1239, 1266-67.
(18.) I have myself often tilted at the windmills of orthodoxy. See, e.g., LARRY ALEXANDER AND & KIMBERLY KESSLER FERZAN, CRIME AND CULPABILITY: m THEORY OF CRIMINAL LAW (2009); LARRY ALEXANDER & EMILY SHERWIN, DEMYSTIFYING LEGAL REASONING (2008); LARRY ALEXANDER, IS THERE A RIGHT OF FREEDOM OF EXPRESSION.'? (2005).
(19.) Kevin C. Walsh, Partial Unconstitutionality, 85 N.Y.U L. REV. 738, 739 (2010).
(20.) Id. at 740-41.
(21.) Id. at 777-78.
(22.) Id. at 780-81.
(23.) For a proposal on severability that points in the opposite direction from Walsh's in the context of challenges to congressional legislation on the ground of absence of constitutional power, see Luke Meier, Facial Challenges and Separation of Powers, 85 IND. L.J. 1557 (2010).
Larry Alexander, Warren Distinguished Professor, University of San Diego School of Law. I wish to thank David McGowan, Miranda McGowan, Grant Morris, Lisa Ramsey, Ted Sichelman, and Steve Smith for their comments. Any errors that remain are entirely their fault for not catching.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: There Is No First Amendment Overbreadth (but There Are Vague First Amendment Doctrines); Prior Restraints Aren't "Prior"; and "As Applied" Challenges Seek Judicial Statutory Amendments. Contributors: Alexander, Larry - Author. Journal title: Constitutional Commentary. Volume: 27. Issue: 2 Publication date: Fall 2011. Page number: 439+. © 1998 Constitutional Commentary, Inc. COPYRIGHT 2011 Gale Group.