The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent


The Supreme Court doesn't decide when speech becomes a constitutionally unprotected "true threat"


The much-awaited Elonis v. U.S. (often discussed as the "Facebook threats" case, though the legal question is the same offline as online) ended up deciding only a fairly narrow—and rarely practically significant—federal statutory question. We still don't know, following Elonis, whether the "true threats" exception to the First Amendment (1) covers only statements said with the purpose of putting someone in fear, (2) applies also to statements said knowing that the target will be put in fear, (3) applies also to statements said knowing that there's a serious risk that the target will be put in fear, or (4) covers all statements that a reasonable person would view as aimed at putting the target in fear. Indeed, as best I can tell, the Supreme Court did not resolve the federal circuit court disagreement on the First Amendment issue that helped persuade the Court to hear the case. And the statutory decision that the Court reached, while important for deciding how to instruct juries in federal threats case, is likely to practically affect the results in only a narrow range of cases.

1. Anthony Elonis posted various items on his Facebook page that sounded like threats to injure his "estranged wife, police officers, a kindergarten class, and an FBI agent." He argued that they were merely rap lyrics that he intended as artistic and fictional. The key question for the court to decide was how to instruct the jury. Should the jury be told that Elonis was guilty only if he had the purpose of putting the alleged targets in fear of criminal attack (a "purpose test")? Or should the jury be told that Elonis was guilty so long as "a reasonable person would foresee" that the targets would perceive the statement as seriously expressing an intent to criminally attack them (a "negligence test")? Each of these questions focuses on what level of mental culpability, or "mens rea," is required to justify criminal punishment.

And to figure this out, the court had to interpret two bodies of law. First, it had to decide what mental state is required by the federal threats statute, 18 U.S.C. § 875, which bans transmitting "in interstate or foreign commerce" (which includes the Internet, telephones, the mail, and many other means) "any communication containing any threat to kidnap [or injure]." Second, it had to decide what mental state is required by the "true threats" exception to the First Amendment, since it's only the existence of this exception that makes the federal threats statute constitutional. And, as in all such situations, the criminal defendant would get the benefit of the higher mental state (the one more demanding on the government).

The federal trial court in this case concluded that the "reasonable person" negligence test was sufficient both under the federal statute and under the First Amendment true threats exception. The court of appeals agreed, as have most other circuit courts of appeals. But the Ninth and the Tenth Circuit courts of appeals concluded that, under the First Amendment, speech can't be punished as a true threat unless it's said with the purpose to put in fear, and that therefore the federal threat statute must be interpreted the same way (as must any other state or federal threat statutes, if they are to be constitutional). The Supreme Court agreed to hear the case, largely to resolve this disagreement among lower courts.

2. But here's what the Court did (and didn't do):

a. The Court apparently decided that the federal threats statute would be satisfied by a showing that the speaker had the purpose of putting the target in fear, or that he knew the target would be put in fear (the "knowledge test")—the defendant's counsel had conceded at oral argument that such knowledge would indeed satisfy the federal statute.

b. The Court decided that the federal threats statute would not be satisfied by a mere showing that a reasonable person would foresee that the statement would put the target in fear. (The Court thus rejected the negligence test.)

c. The Court expressly declined to decide whether the statute would be satisfied by a showing that the speaker knew there was a serious risk that the target would be put in fear (the "recklessness test"). Justice Alito's separate opinion argued in favor of adopting the recklessness test, but none of the other Justices agreed.

d. The Court expressly declined to reach the First Amendment question, because its ruling led to the conviction being reversed on statutory grounds (though there would likely be a retrial, if the prosecution so wishes). "Given our disposition," the Court said, "it is not necessary to consider any First Amendment issues."

e. Justice Thomas would have concluded that under the statute the government only needed to show that a defendant "knew he transmitted a communication" (that element was accepted by everyone), "knew the words used in that communication" (that also was accepted by everyone), "and understood the ordinary meaning of those words in the relevant context," and that this test was constitutional. But no other Justice agreed with this "understood the ordinary meaning of those words in the relevant context" test.

3. State courts dealing with cases under state threat statutes—and there are many such prosecutions—are thus almost entirely unhelped by the Supreme Court's decision. After all, for state statutes the questions are (a) how that state statute should be interpreted, and (b) what the First Amendment requires. The Court's decision didn't address either of these issues. (No-one expected the Court to address the state statutory question, but people did expect it to address the First Amendment question.) At most, the Court's reasoning rejecting a negligence test under the federal statute might be seen as persuasive to state courts interpreting their state statutes; but only persuasive, not legally binding.

4. And even federal courts aren't much helped by the Elonis decision. True, they now know that they shouldn't give jury instructions that use the negligence test in § 875 cases. But they don't know whether they should instruct the jury using the knowledge test (defendant knew the statement would put the target in fear) or the recklessness test (defendant knew there was a serious risk that the statement would put the target in fear). Likewise, prosecutors don't know which test to ask for: Perhaps they're legally entitled to the recklessness test—the Court left open that possibility—but if they ask for that test, they'll be facing the risk of being reversed if the circuit court concludes that knowledge is the right standard instead.

And beyond that, courts don't even know whether they have to instruct the jury using the purpose test. True, the Supreme Court said that the federal statute doesn't require a showing of purpose, but is satisfied by a showing of knowledge. But the Court specifically said that it wasn't expressing an opinion on whether the First Amendment requires a showing of purpose, as the Ninth and Tenth Circuits (but not other circuits) have concluded. It's thus possible that a defendant can't be convicted without a showing of a purpose to threaten, on the theory that the First Amendment requires such a showing even though the federal statute doesn't.

5. Moreover, this odd result—that, despite the Elonis ruling, a purpose to threaten is still required—seems to be the one mandated under the Ninth and Tenth Circuit precedents. Both those cases concluded that the First Amendment true threats exception does require a showing of purpose to threaten. Those panel decisions remain binding law in those circuits, unless overruled by an en banc rehearing in those circuits, or by a Supreme Court decision. And while the Court in Elonis certainly could have overruled those panel decisions if it had reached the First Amendment question, it pointedly declined to consider the First Amendment issue.

So the circuit split still remains, on the First Amendment issue though not on the federal statutory question. The Ninth and Tenth Circuit precedents still take the view that the First Amendment requires a purpose test in threats cases. Precedents in the other circuits take the view that the First Amendment requires a negligence test (though—following Elonis—the statute doesn't take advantage of that possible latitude but instead requires either a recklessness test or a knowledge test).

6. Narrow as the holding is legally, it's also likely to be even narrower practically. In principle, mental states matter, and lawyers and judges rightly care a lot about them, especially when the mental state is about facts that might be hard to know (e.g., whether a statement about someone is true or false), or about consequences that turn on hard to predict events (e.g., whether a person's conduct was going to lead to the death of another person). But as a practical matter, when it comes to threat prosecutions, it should be the rare case in which the negligence test would produce a different result than a knowledge test. Usually, if someone says something that a reasonable person would perceive as threatening, the speaker would likely also know that the target would perceive it as threatening. Elonis, for instance, wrote,

Fold up your [protection-from-abuse order] and put it
in your pocket
Is it thick enough to stop a bullet?
Try to enforce an Order
that was improperly granted in the first place
Me thinks the Judge needs an education
on true threat jurisprudence
And prison time'll add zeros to my settlement …
And if worse comes to worse
I've got enough explosives
to take care of the State Police and the Sheriff's Department.

I expect a jury would conclude—even beyond a reasonable doubt—that Elonis did indeed know that his wife would perceive the speech as threatening (even if he didn't have the purpose of threatening her, but was just trying to work out his frustrations in artistic expression, which seems to be his theory).

To be sure, in some situations there might be a plausible case that there was just an unfortunate misunderstanding, or a joke gone wrong. The defendant might plausibly argue that he was just goofing around, and didn't know anyone would take the statement seriously, or that he meant one thing and didn't know the subject of the statement would understand it as meaning another. (At least the defendant's argument might be plausible enough to create a reasonable doubt about his guilt.) In United States v. Fullmer, for instance, a Fullmer was convicted for telling an FBI agent, "The silver bullets are coming," even though Fullmer claimed that he was referring to the Lone Ranger, where silver bullets meant "help is on its way." But even in that case, the jury had also found that Fullmer had intended to "impede, intimidate, … interfere with," or retaliate against the FBI agent; so it seems likely that a jury instructed even using the purpose test would have convicted Fullmer as well. So the Court's endorsement of a knowledge test or recklessness test (again, recall that we don't quite know which one is the right test) will sometimes yield a different result than under the negligence test or the purpose test, but only rarely.

7. And the holding is also narrow because some of the most important questions about the true threats exception weren't even squarely presented to the Court. For instance, what about statements that allegedly threaten unspecified members of a vast group of people—such as police officers generally, or abortion providers, or capitalists, or Jews or blacks or gays or any other large group? (Recall that, in this very case, "I've got enough explosives to take care of the State Police and the Sheriff's Department" was one of the statements that was found to be threatening.) Those are not uncommon in talk about public issues, and I think that they are probably not specific enough to be punishable threats—though query whether in this case there were some contextual reasons to think that Elonis was referring to a particular state police or sheriff's department building, and whether that should matter. But in any event the Supreme Court's decision doesn't opine on the subject.

8. Now it might be that the Court had to leave a good deal unresolved here. Perhaps the Justices couldn't settle on a majority opinion on the recklessness vs. knowledge question, or on the First Amendment issue, which might explain why such a narrow opinion took six months to release. And there are arguments in favor of the Court deciding as little as possible in most cases, especially constitutional cases. But in any event, it appears that this is an unusually narrow decision.

Congratulations to our coblogger John Elwood, who represented Elonis before the Supreme Court, on winning the case; and please see Orin's post for more thoughts on the matter.