Concept
Limits on Overbreadth Challenges
Overbreadth
Limits on Overbreadth Challenges
Overbreadth lets a litigant attack a speech law on behalf of speakers who are not in court, and Virginia v. Hicks sets the conditions on that privilege. A law regulating speech is invalid on its face only if it prohibits a substantial amount of protected speech, judged in relation to its plainly legitimate sweep, and the challenger must prove it. A law that falls short stays enforceable against unprotected conduct, even though some hypothetical application would reach protected speech.
The substantiality standard states the test in the abstract. Applying it raises three further questions, and Hicks answers each while rejecting the challenge before it. What must the challenger put before the court? Which law supplies the legitimate side of the comparison? How much does the kind of law under attack matter? The airport resolution in Board of Airport Commissioners v. Jews for Jesus reached nothing but expression and fell on its face. The trespass policy in Hicks governed presence on property and survived.
A challenger who fails on any of the three points loses the facial claim, but the speaker who was actually harmed keeps a remedy. Improper applications go to as-applied litigation, where the speaker can obtain relief without costing the state its ability to enforce the law against everyone else. The challenger’s burden comes first, because the other two questions assume a challenger who has built a record.
Challenger’s Burden
A litigant who asks a court to suspend a speech law against everyone must prove that the law deserves it. In Hicks the Court put that burden on the overbreadth claimant, who must demonstrate “from the text of [the law] and from actual fact” that substantial overbreadth exists.1 The Virginia Supreme Court had objected to an unwritten rule requiring demonstrators and leafleters to obtain advance permission, and on that basis it declared the entire trespass policy void, including the written rule that a person who returns after a barment notice is subject to arrest.2 The Court held that a court may order that result under the overbreadth doctrine only if the policy, taken as a whole, is substantially overbroad judged in relation to its plainly legitimate sweep.3
The two halves of the quoted phrase do different work. The text fixes what the law covers and so defines the universe of applications to be sorted into protected and unprotected. Actual fact shows how the law operates: whom officials warn, which speakers they turn away, who in practice is deterred. A challenger who offers only text can show a law’s literal breadth but not that its reach into protected expression is realistic. A challenger who offers a few aggrieved speakers cannot show how they compare with everyone else the law governs. Because the test sets protected applications against legitimate ones, the record has to speak to both.
The burden sits with the challenger because of what success costs. A law struck down for overbreadth is not narrowed; it becomes wholly unenforceable until a legislature rewrites it or a properly authorized court construes it more narrowly.4 Courts therefore want solid evidence before they pay that price. Claims of facial invalidity often rest on speculation, and speculation raises the risk of premature interpretation of statutes on factually barebones records.5 A challenger who relies on imagined applications asks the court to run exactly that risk.
Hicks lost for want of the showing. He failed to demonstrate that any First Amendment activity falls outside the “legitimate business or social purpose[s]” that permit entry onto the property.6 As far as the record revealed, entering for a First Amendment purpose was not a trespass until the person received a barment notice.7 His argument therefore rested on what the policy could conceivably permit, and a conceivable application does not carry the burden.
The Court’s other facial-challenge decisions make the same demand. A dissent in Citizens United v. Federal Election Commission recalled that in Washington State Grange v. Washington State Republican Party the Court said it generally does not apply “the strong medicine of overbreadth analysis where the parties fail to describe the instances of arguable overbreadth of the contested law.”8 Courts are likewise reluctant to find overbreadth “substantial” on mere possibilities.9 In National Endowment for the Arts v. Finley, challengers to a statute directing the agency to take “decency and respect” into consideration argued that the provision would inevitably serve as a tool for viewpoint discrimination.10 The Court answered that the statute merely admonishes the agency to take those values into consideration and aims at reforming procedures rather than precluding speech, and that in the cases where it had struck down legislation as facially unconstitutional, the dangers were both more evident and more substantial.11
Justice Alito’s dissent in United States v. Stevens shows the burden argued with real-world categories. He would have rejected the facial challenge to §48 because the statute validly reaches at least two broad categories of expression, crush videos and dogfighting videos, which give it a substantial core of permissible applications, and because the respondent had not shown that impermissible applications were “substantial” in relation to the statute’s “plainly legitimate sweep.”12 Justice O’Connor, concurring in part and dissenting in part in Ashcroft v. Free Speech Coalition, made a similar argument from missing examples. She agreed that the ban on youthful-adult pornography was overbroad but found no overbreadth claim against computer-generated images of children, because the challengers had identified no wholly computer-generated materials that had serious value or did not facilitate child abuse.13 The majority disagreed on that part, so her opinion shows how an argument from missing examples runs and does not state a holding.
The burden is lightest when the text itself carries the showing. In Board of Airport Commissioners v. Jews for Jesus the Court held facially overbroad a rule banning all First Amendment activities in an airport terminal, because the rule prohibited virtually every expressive activity, far beyond any legitimate regulatory objective. The resolution’s words reached the whole category of expression and left almost nothing on the legitimate side of the comparison, so the challenger needed no separate proof. The trespass policy in Hicks sits at the other end. Its words described a rule about presence on property, which people choose for many reasons, so any protected reach had to be shown by evidence. A law that is general on its face leaves the work to the record.
On an exam, ask first what the challenger has actually put forward. If the stem offers only the assertion that a law could reach protected speech, the claim has not carried the burden, and any improper applications remain for as-applied challenges. If the stem supplies the law’s text together with testimony or other facts showing whom it reaches, move to the comparison with the law’s plainly legitimate sweep.
Choosing the Measured Law
Overbreadth is a ratio, with protected applications on top and plainly legitimate applications below, and a court cannot compute it until it decides which law supplies the denominator. Hicks states the condition for voiding a policy: the policy, taken as a whole, must be substantially overbroad judged in relation to its plainly legitimate sweep.15 A narrow denominator, such as a single provision that mostly touches speakers, makes the protected share look large. A wide one, such as a regulatory scheme with many ordinary uses, makes the same provision a small part of a legitimate whole. The choice shapes the result, and it is easy to miss because it happens before any counting begins.
The two courts in Hicks drew the line in different places. The Virginia Supreme Court objected to an unwritten rule that anyone distributing literature or demonstrating on the property obtain prior authorization, and it voided the entire policy, including the written rule that a person who returns after a barment notice is subject to arrest.16 Hicks himself was punished for entering in violation of that notice-barment rule, not for his speech.17 The Supreme Court took the broader view and looked to the potential applications of the entire trespass policy, written and unwritten.18 Because the state court had struck down the whole policy, the Court reasoned, the question was whether the whole policy was substantially overbroad.19 A remedy sets the unit of measurement. A court that has not shown an entire policy to be overbroad has no basis for voiding all of it, however troubling one component may be.
The justices also disagreed about what the state court had done. Justice Souter read it as having assumed that the appropriate focus was the unwritten element of the policy, the requirement that nonresidents obtain prior authorization before distributing literature or demonstrating.20 The majority answered in a footnote that the state court’s opinion contained no comparison of valid and invalid applications at all, so that the proportionality aspect of the doctrine was simply ignored.21 On either reading, the state court never weighed the permission rule against the rest of the policy’s reach, and the Supreme Court supplied the comparison by measuring the policy as a whole.
Justice Souter’s concurrence explains why the Court could leave the underlying question open. Hicks lost under either measure, he wrote, because regardless of the scope of the law that forms the denominator, the numerator of potential invalid applications was too small to establish substantial overbreadth.22 In other circumstances, he added, the scope of the law chosen for comparison might decide the case. A city’s speech ordinance for a public park might be analyzed alone or as one element of the combined policies governing expression in public schoolyards, municipal cemeteries, and the city council chamber, and the answer might turn on which.23 He then wrote that today’s decision does not address how to identify the scope of the relevant law, so the question remains open.24
Souter did not work through the numbers, but the mechanism is easy to see. A park ordinance measured alone reaches the expression that occurs in parks, and a large share of that expression might be protected. Measured as one piece of a combined scheme, the same ordinance becomes a small part of a much larger body of applications, one that includes settings, such as a council chamber, where the government may have more room to restrict expression. This illustrates how the choice could cut and does not predict how a court would decide. Nothing in Hicks supplies a general method, and each side has a reason to favor the measure that helps it: the challenger prefers the narrow unit that makes protected speech dominate, and the government prefers the wide unit that dilutes it. When the challenged law is a single statute, the court has an obvious candidate and the argument moves straight to counting. Justice Alito’s Stevens dissent, for example, identified two broad real-world categories of expression to which §48 could validly apply and treated them as the statute’s substantial core.25
The hypothetical shows why an exam answer on overbreadth should name the denominator before reaching the arithmetic. A sound answer identifies the candidate laws, says which one the court below invalidated, and runs the comparison under each. If the conclusion is the same either way, the answer says so and stops, as the Court did in Hicks. If the conclusion flips, the answer must acknowledge that the Court has not resolved how to choose, argue for the better measure, and give the reason, such as the scope of the remedy being sought.
Speech Versus Conduct
The kind of law under attack predicts which side of the ratio will be larger, and Hicks tied the failure of the challenge to that variable. The Court explained that the overbreadth doctrine’s concern with chilling protected speech “attenuates as the otherwise unprotected behavior that it forbids the State to sanction moves from ‘pure speech’ toward conduct.” An overbreadth challenge will therefore rarely, if ever, succeed against a law “not specifically addressed to speech or to conduct necessarily associated with speech,” such as picketing or demonstrating.27 That sentence describes the target of the law, and it is the main reason Hicks was an easy case for the state.
Two considerations make the target matter. A law that punishes speech attaches its sanction to the act of communicating, so a would-be speaker must ask whether the law covers the message before speaking, and the doubt arises at the moment of decision. A law that punishes entering land after a warning does not turn on whether the person has anything to say, so a speaker can tell whether the act is forbidden without trimming the message. The second consideration lies on the state’s side. There comes a point, the Court has said, at which the chilling effect of an overbroad law, significant though it may be, cannot justify prohibiting all enforcement of that law, particularly one that reflects “legitimate state interests in maintaining comprehensive controls over harmful, constitutionally unprotected conduct.” Blocking such a law carries substantial social costs, especially when it stops application to unprotected conduct, and the Court therefore insists that a law’s application to protected speech be substantial both in an absolute sense and relative to the scope of its plainly legitimate applications.28
The exception itself is built for speech laws. The transcendent value to all society of constitutionally protected expression justifies allowing attacks on overly broad statutes that regulate or proscribe speech, at least where no readily apparent construction suggests itself as a vehicle for rehabilitating the statute in a single prosecution, and it does so without requiring the challenger to show that his own conduct could not be regulated by a statute drawn with narrow specificity.29 Beneath that rationale lies the principle that forbids legislatures from interfering with constitutionally guaranteed freedoms even though some applications of a given law would infringe none.30
In terms of the ratio, a law aimed at speech has few legitimate applications beyond narrow categories of unprotected expression, so any reach into protected speech is large and the chill is real. A law aimed at conduct such as trespass has a large legitimate sweep that has nothing to do with expression, so an occasional application to a speaker is a small remainder. Case-by-case litigation can remove that remainder without disabling the law for everyone.
At the speech end of the range sits Board of Airport Commissioners v. Jews for Jesus, where a rule banning all First Amendment activities in an airport terminal fell because it prohibited virtually every expressive activity. A rule whose every application is to expression has no body of ordinary conduct to dilute its protected reach, so the doctrine operates at full strength. An ordinance aimed at assembly itself fits the same pattern. One that required three or more people meeting on a sidewalk or street corner to conduct themselves so as not to annoy any passerby was held unconstitutionally broad, as well as vague, because it authorized the punishment of constitutionally protected conduct.31
Hicks occupies the other end. The policy punished entry onto the property without regard to why a person entered, and punishing a violation by someone who wishes to speak implicates the First Amendment no more than punishing a person who, after being lawfully banned from a public park for vandalizing it, ignores the ban to join a political demonstration.32 Hicks was punished for his nonexpressive conduct, his entry in violation of the notice-barment rule, and not for his speech.33 The two cases apply one standard to laws with different targets, and the target produces the different ratios.
Between the poles lies conduct necessarily associated with speech, and Broadrick v. Oklahoma shows that the line is a gradient. The statute there restricted partisan political activity, a subject that plainly includes expression, yet the challengers were charged with soliciting money among their co-workers for the benefit of their superior and conceded that the statute could constitutionally be applied to that conduct.34 Even a law addressed to political activity can therefore sit nearer the conduct end when its core applications reach conduct the state may validly punish. The question that places a law on the gradient is whether the prohibited act is itself a way of communicating. Picketing and demonstrating are. Trespassing is not inherently expressive, even though a person may sometimes commit it for expressive reasons.35
The speech-conduct line changes the odds and leaves the requirement in place. A challenger against a speech law must still show substantiality, and the line adjusts only how likely that showing is. At the far end the Court has declined to allow overbreadth at all: in Los Angeles Police Department v. United Reporting Publishing Corp., 528 U.S. 32 (1999), overbreadth was unavailable in a challenge to a statute limiting access to certain records.36 Overbreadth is ordinarily invoked by a litigant who has engaged in some type of expressive activity, and the Court has left open whether a litigant who engaged in none may invoke it.37
A failed facial challenge to a conduct law still leaves the speaker a remedy. Applications of the trespass policy that violate the First Amendment can be remedied through as-applied litigation, and the Court held that the Virginia Supreme Court should not have used the “strong medicine” of overbreadth to invalidate the entire policy.38
The first step carries most of the weight, and the usual error there is to classify a law by the fact that some speaker could be arrested under it. The housing authority’s rule in Hicks could have been enforced against a leafleter, yet its target remained presence on property, and one possible speaker among everyone the rule covered did not change what it was addressed to.