Freedom of Speech and Knowing Falsehoods

United States v. Alvarez, the Stolen Valor Act case, dealt with an important and recurring First Amendment question: When may the government ban lies — not just innocent mistakes, or statements that wouldn’t reasonably be understood as factual assertions (e.g., fiction, parody, or obvious hyperbole), but knowing falsehoods? That question arises in a wide range of cases: libel, perjury, fraud, unsworn false statements to the government, knowingly false statements about a person that aren’t defamatory, knowingly false statements that intentionally inflict emotional distress (e.g., falsely telling someone that his wife or children have been injured), hoaxes that cause a public panic, and more. Some past Supreme Court decisions suggested that all such knowing lies are constitutionally unprotected. Some have suggested that only some limited subsets of such knowing lies are constitutionally unprotected. Which is it?

In Alvarez, the Court held, 6-3, that the Stolen Valor Act — which the Court read as criminalizing speakers’ lies about having supposedly received military decorations — violated the First Amendment. But lurking behind this was a more complicated 4-2-3 split that was in some ways a 5-4 split in favor of treating lies as generally less constitutionally protected.

Here’s what happened:

1. Four Justices — Justice Kennedy, joined by Chief Justice Roberts, and Justices Ginsburg and Sotomayor — took the view that even lies are fully constitutionally protected unless they fall within certain categories of punishable lies: defamation, perjury, fraud aimed at getting valuable goods or services, unsworn lies to government officials about official matters, and lies that claim that one is a government official or is speaking on behalf of the government. This list of First Amendment exceptions would presumably also includes nondefamatory false statements about individual people, which are punishable under the so-called “false light” tort. Though the plurality didn’t mention this exception, two past Supreme Court cases have endorsed it (Time, Inc. v. Hill (1967) and Cantrell v. Forest City Publishing Co. (1974)), and the plurality didn’t suggest it would overrule those cases.

Laws banning lies outside these categories would still be potentially justifiable, according to the plurality, but only if they pass the same “strict scrutiny” test (with its requirement that the law be necessary to serve a compelling government interest) that is used for restrictions on true statements or statements of opinion. This strict scrutiny test nearly always leads to the invalidation of content-based speech restrictions, so presumably restrictions on lies that fall outside the exceptions the plurality identified would — in the plurality’s view — almost always be unconstitutional. [UPDATE: Whoops, originally wrote "would almost always be constitutional," sorry about that.]

2. Two Justices who concurred in the result but not in the plurality opinion — Justice Breyer, joined by Justice Kagan — took the view that (a) lies about “easily verifiable facts that do not concern … subject matter” such as “philosophy, religion, history, the social sciences, the arts, and the like” are categorically less protected than true statements and statements of opinion, but (b) they are still somewhat protected, so any restrictions on such lies should be judged under so-called “intermediate scrutiny.” This test is much easier for a law to satisfy than “strict scrutiny,” but Justices Breyer and Kagan concluded that the Stolen Valor Act failed this test.

To pass “intermediate scrutiny” under the concurrence’s test, a restriction on lies would apparently require (a) “specific harm to identifiable victims” — such as emotional or reputational injury to the subjects or falsehood, or financial loss to people who are duped — (b) a high likelihood of such harm, to identifiable victims or to the public generally (which might cover cases such as false reports of crime, false statements to government officials within the scope of their operation, impersonation of government official, and so on). But the Stolen Valor Act, the concurrence concluded, wasn’t limited to such statements, and also poses a risk of government abuse:

As written, [the Act] applies in family, social, or other private contexts, where lies will often cause little harm. It also applies in political contexts, where although such lies are more likely to cause harm, the risk of censorious selectivity by prosecutors is also high. Further, given the potential haziness of individual memory along with the large number of military awards covered (ranging from medals for rifle marksmanship to the Congressional Medal of Honor), there remains a risk of chilling that is not com¬pletely eliminated by mens rea requirements; a speaker might still be worried about being prosecuted for a careless false statement, even if he does not have the intent re¬quired to render him liable. And so the prohibition may be applied where it should not be applied, for example, to bar stool braggadocio or, in the political arena, subtly but selectively to speakers that the Government does not like.

And because the government may “substantially … achieve [its] objective” of protecting the value of medals “in less burdensome ways” the law fails intermediate scrutiny. The potential “less burdensome ways” that the concurrence pointed to included publicizing who has actually won medals, excluding “family, social, or other private contexts” from the scope of the law, limiting the law to situations that pose special risks of harm (such as, presumably, fraud in getting money, jobs, or contracts), and limiting the law to certain especially important classes of medals.

3. Three Justices — Justice Alito, joined by Justices Scalia and Thomas — took the view that lies are basically categorically unprotected by the First Amendment (again, setting aside laws that ban lies “about philosophy, religion, history, the social sciences, the arts, and other matters of public concern” and thus present “a grave and unacceptable danger of suppressing truthful speech”). Some such lies shouldn’t, as a matter of policy, be banned (“say, a law making it a criminal offense to falsely claim to have been a high school valedictorian”). But the First Amendment leaves laws that ban these sorts of lies to the political process, not to judicial evaluation. (I should note that this was my view of the matter, and I continue to be inclined to this view.)

There is a zone of lower First Amendment protection for lies: So the bottom line is that five Justices — the two in concurrence, and the three dissenters — agree that lies about easily verifiable matters (rather than about history, social science, and the like) are outside the core of First Amendment protection, the core to which strict scrutiny applies. But the three dissenters believe there should be no protection for such lies, and the two concurring Justices believe there should be a lower form of protection, “intermediate scrutiny.” (In a sense, the two in the concurrence would treat lies similarly to how the Court has treated accurate commercial advertising, but one shouldn’t overstate this similarity: “Intermediate scrutiny” has long meant different things in different areas of free speech law.)

Fraud (and attempted fraud) to get respect and affection is less constitutionally significant than fraud (and attempted fraud) to get even modest sums of money: Note also that the plurality and the concurrence seem to view the use of lies to get money from listeners — even modest amounts of money — differently from the use of lies to get respect, affection, and the like (and perhaps even votes). After all, people who lie about their medals presumably do so to get something: either money, whether directly or indirectly through offers of jobs or contracts, or respect, affection, sex, and so on.

Yet the Court says that this sort of fraud does not justify restricting lies, though the Court has held that the fraudulent solicitation of money — presumably including attempted fraud as well as successful fraud — does justify restricting lies, see Illinois ex rel. Madigan v. Telemarketing Associates, Inc., 538 U. S. 600, 612 (2003) (“Like other forms of public deception, fraudulent charita¬ble solicitation is unprotected speech”). This strikes me as an odd sort of First Amendment line to draw, yet it seems like the plurality and the concurrence think it should be drawn.

Not clear what will happen with laws punishing lies in election campaigns: Finally, the more important controversy related to laws that punish knowing falsehoods has to do with laws that punish such lies in election campaigns, especially when they are said by candidates or their official campaigns. (There’s a split among lower courts on this issue.) This case did not resolve the controversy.

Such falsehoods are in a sense closer to financial fraud, because they involve candidates seeking a paying job. Moreover, the government interest in preventing the deception of voters is quite strong, likely strong enough to pass the concurrences’ “intermediate scrutiny” test. On the other hand, many of them are a lot less easily verifiable, and thus pose a risk of deterring true speech as well as false speech. And the concurrence seemed to worry that such political falsehoods pose more of a risk of selective prosecution.

My guess is that general bans on lies in election campaigns would be struck down, because they cover a wide range of territory in which the truth may be hard to uncover, and in some measure in the eye of the beholder. But narrower bans on, say, knowingly false statements about when or where people should vote, knowingly false claims that some person or organization has endorsed you, knowingly false claims that you are the incumbent (see, e.g., Treasurer of the Committee to Elect Gerald D. Lostracco v. Fox, 389 N.W.2d 446 (Mich. Ct. App. 1986)), knowingly false claims about your own job experience — including military experience — and the like might be constitutional. It’s just hard to tell, given both the limited scope of the opinions and the 4-2-3 split.