Sometimes lies are constitutionally punishable: Consider libel, false state­ments to government investigators, fraudulent charitable fundraising, and more. (I speak here of lies in the sense of knowing or reckless falsehoods, rather than honest mistakes.) But sometimes even deliberate lies are constitutionally protected. In New York Times v. Sullivan, the Court held that even deliberate lies (said with “actual malice”) about the government are constitutionally protected. And in United States v. Alvarez, five of the justices agreed that lies “about philosophy, religion, history, the social sciences, the arts, and the like” are generally protected.

The Supreme Court hasn’t explained where the line is drawn, and that leaves unclear where important areas of controversy—such as laws punishing lies in election campaigns—should fall. In this short article, I hope to offer an account that makes sense of the precedents and a framework for making future decisions.

I. Punishable Lies

The Supreme Court has held that defamation, perjury, fraudulent attempts to get money, speech actionable under the false light tort, and lies that inflict severe emotional distress are all constitutionally unprotected. In Alvarez, the Court also suggested that the government may more broadly punish lies that involve “some ... legally cognizable harm associated with a false statement, such as an invasion of privacy or the costs of vexatious litigation”; “false statements made to Government officials, in communications concerning official matters”; and lies that are “integral to criminal conduct,” a category that might include “falsely representing that one is speaking on behalf of the Government, or ... impersonating a Government officer.” Minnesota Voters Alliance v. Mansky stated, in dictum, that “We do not doubt that the State may prohibit messages intended to mislead voters about voting requirements and procedures”; but that case focused on speech in a nonpublic forum (polling places), and it’s not clear that the Court meant to authorize such prohibitions in public speech more generally.

Lower courts have generally allowed liability or punishment for lies about others’ products or property; unsworn lies to government officials; lies likely to provoke public panic; lies about being a government official; lies about having a particular university degree or professional license (regardless of whether the false representation is intended to defraud a prospective employer or professional client); lies to voters about the authorship or endorsement of political campaign materials; and a candidate’s lies to voters about his own credentials. (Query whether these cases are in some measure undermined by Alvarez.)

II. Unpunishable Lies

A. The doctrine

But some lies, the Court told us, are indeed constitutionally protected—again, not just when they are said without “actual malice,” but even if the speaker knows the statements are false. This includes “false statements about philosophy, religion, history, the social sciences, the arts, and the like,” at least “in many contexts.” (I assume physical sciences would be covered as well.) More broadly, this may include lies about any matters that are not “easily verifiable,” or where “it is perilous to permit the state to be the arbiter of truth.”

Five of the justices in United States v. Alvarez took this view: Justices Breyer and Kagan in the concurrence and Justices Alito, Scalia, and Thomas in the dissent. And it seems likely that the four justices in the plurality, who generally took a more speech-protective view than the concurrence or the dissent, would have agreed.

When it came to the lies prohibited by the statute involved in Alvarez itself—lies about having been awarded military decorations—the justices, put together, appeared to apply intermediate scrutiny. The four-justice plurality would have applied strict scrutiny, but the swing votes in the concurrence applied intermediate scrutiny, and the three dissenters would have found those lies to be categorically unprotected. But as to lies about philosophy, history, science, and the like, a majority of justices endorsed categorical protection.

And, of course, in New York Times v. Sullivan, the Court held that “prosecutions for libel on government”—in context, including civil liability for such libel—“have [no] place in the American system of jurisprudence.” That included quite specific allegations, such as claims that the police had arrested Martin Luther King Jr. seven times. The allegations were not a libel of Sullivan, the Court held, because they weren’t sufficiently “of and concerning him”; and they couldn’t be a libel of the city because that would constitute an unconstitutional seditious libel claim.

The boundaries of this unprotected zone, however, are not defined, though the recent concern about “fake news” makes those boundaries quite important. Courts are sharply split, for instance, on whether the government may generally punish lies in an election campaign. One court decision has rejected claims of liability for alleged lies about vaccines, but without discussing in detail whether such liability can be upheld on the theory that (to quote Alvarez) it involves “legally cognizable harm associated with a false statement”—in that case, physical injury caused by a person’s believing the false statement.

B. The history, to which the doctrine is reacting

Of course, the concern about “fake news” and the harms it can cause, both to society broadly and to particular people, is hardly new. Since 2020, many people have condemned false claims of election misconduct on the grounds that those claims damage democracy and can indeed lead to insurrections. But similar concerns (though not about election results in particular) date back to the debate over the Sedition Act of 1798—and the New York Times v. Sullivan rejection of “prosecutions for libel on government” is likely a reaction precisely to those historical attempts to restrict speech.

Justice Chase’s instructions to the jury in United States v. Cooper, for instance, defended seditious libel prosecutions on the grounds that:

If a man attempts to destroy the confidence of the people in their officers, their supreme magistrate, and their legislature, he effectually saps the foundation of the government.

Likewise, Justice Iredell in Case of Fries (1799) reasoned that the Fries Rebellion happened because “the government had been vilely misrepresented, and made to appear to them in a character directly the reverse of what they deserved.” “In consequence of such misrepresentations, a civil war had nearly desolated our country, and a certain expense of near two millions of dollars was actually incurred, which might be deemed the price of libels.” And this showed that seditious libel prosecutions were necessary:

Men who are at a distance from the source of information must rely almost altogether on the accounts they receive from others. … [If those] accounts are false, the best head and the best heart cannot be proof against their influence; nor is it possible to calculate the combined effect of innumerable artifices, either by direct falsehood, or invidious insinuations, told day by day. …

Such being unquestionably the case, can it be tolerated in any civilized society that any should be permitted with impunity to tell falsehoods to the people, with an express intention to deceive them, and lead them into discontent, if not into insurrection, which is so apt to follow?

Iredell reasoned that falsehoods “intended to destroy confidence in government altogether, and thus induce disobedience to every act of it” had to be punishable. And the need to punish libel in a republic was especially great

because in a republic more is dependent on the good opinion of the people for its support, as they are, directly or indirectly, the origin of all authority, which of course must receive its bias from them. Take away from a republic the confidence of the people, and the whole fabric crumbles into dust.

Update the language and the examples, and you can see an argument that could easily have been made in 2021.

A similar concern about harmful lies arose during World War I, when the Espionage Act of 1917 made it a crime to, among other things, “willfully make or convey false reports or false statements with intent to interfere with the operation or success of the military or naval forces of the United States or to promote the success of its enemies.” The Court upheld a conviction under this statute in Schaefer v. United States, reasoning in part that the First Amendment did not extend to “weaken[ing] or debas[ing]” the “morale of the armies” through “question or calumny of the motives of authority.”

The statements about the motivations for the war in the German-language newspaper in that case, the Court held, were “deliberate and willfully false, the purpose being to represent that the war was not demanded by the people but was the result of the machinations of executive power, and thus to arouse resentment to it and what it would demand of ardor and effort.” And other statements, the Court concluded, deliberately mistranslated a statement by Sen. Robert LaFollette: LaFollette had warned of “bread lines” that might happen as a result of the war, but the newspaper rendered this as “bread riots.”

Justice Brandeis, joined by Justice Holmes, didn’t quarrel with the constitutionality of the statute but concluded that the statute had to be read as limited to

[w]illfully untrue statements which might mislead the people as to the financial condition of the Government and thereby embarrass it; as to the adequacy of the preparations for war or the support of the forces; as to the sufficiency of the food supply; or willfully untrue statements or reports of military operations which might mislead public opinion as to the competency of the army or navy or its leaders; or willfully untrue statements or reports which might mislead officials in the execution of the law, or military authorities in the disposition of the forces.

Reading it more broadly, as he thought the majority did, threatened the First Amendment:

To hold that such harmless additions to or omissions from news items, and such impotent expressions of editorial opinion, as were shown here, can afford the basis even of a prosecution will doubtless discourage criticism of the policies of the Government. To hold that such publications can be suppressed as false reports, subjects to new perils the constitutional liberty of the press. …

And, he reasoned, the rationale would apply in peacetime as well:

In peace, too, men may differ widely as to what loyalty to our country demands; and an intolerant majority, swayed by passion or by fear, may be prone in the future, as it has often been in the past, to stamp as disloyal opinions with which it disagrees. Convictions such as these, besides abridging freedom of speech, threaten freedom of thought and of belief.

Today, I expect that the “no seditious libel prosecutions” holding of Sullivan would preclude arguments such as those in Cooper and Fries, and likely even those as in Schaefer. But those arguments should still be remembered, if only as cautionary tales.

III. What’s the Difference?

A. Alvarez

Why then are some lies punishable and others not? The dissent in Alvarez gave a sketch of the argument, and it seems likely the concurrence and the plurality would have agreed. Let’s look at the passage in which the “philosophy, religion, history, the social sciences, the arts, and other matters of public concern” passage appears:

[T]here are broad areas in which any attempt by the state to penalize purportedly false speech would present a grave and unacceptable danger of suppressing truthful speech. Laws restricting false statements about philosophy, religion, history, the social sciences, the arts, and other matters of public concern would present such a threat. The point is not that there is no such thing as truth or falsity in these areas or that the truth is always impossible to ascertain, but rather that it is perilous to permit the state to be the arbiter of truth.

Even where there is a wide scholarly consensus concerning a particular matter, the truth is served by allowing that consensus to be challenged without fear of reprisal. Today’s accepted wisdom sometimes turns out to be mistaken. And in these contexts, “[e]ven a false statement may be deemed to make a valuable contribution to public debate, since it brings about ‘the clearer perception and livelier impression of truth, produced by its collision with error.’”

Allowing the state to proscribe false statements in these areas also opens the door for the state to use its power for political ends. Statements about history illustrate this point. If some false statements about historical events may be banned, how certain must it be that a statement is false before the ban may be upheld? And who should make that calculation? While our cases prohibiting viewpoint discrimination would fetter the state’s power to some degree, see R.A.V. v. St. Paul, 505 U.S. 377, 384-390 (1992) (explaining that the First Amendment does not permit the government to engage in viewpoint discrimination under the guise of regulating unprotected speech), the potential for abuse of power in these areas is simply too great.

There are two different arguments working together here:

  1. It’s especially “perilous” for the government—and it is especially accompanied by a “potential for abuse of power”—to decide the truth as to certain broad classes of topics, because it “opens the door for the state to use its power for political ends.” But such decision-making is presumably less perilous when individual libel lawsuits are involved, or when someone is prosecuted for, for instance, perjury or fundraising fraud.
  2. It’s especially valuable to allow constant challenges to received wisdom as to certain topics because that’s the way knowledge in that field progresses (as the history of science, the history of medicine, the history of religion, or for that matter the history of history, vividly illustrate). But it is presumably less valuable to allow constant “challenge[]” to the “consensus” about what some particular individual (even a public official) has done.

Yet while these concerns are doubtless relevant, it seems to me that the reason they are relevant is tied to a matter that was only implicit in the Alvarez opinions: the importance of alternative truth-finding institutions beyond the legal system, and the relative advantages of relying on those institutions in certain situations.

B. When alternative institutions exist

1. Generally

Consider, for instance, science and history. John Stuart Mill famously defended freedom of speech in part on the grounds that the only real basis for “presuming an opinion to be true” is that “with every opportunity for contesting it, it has not been refuted.” “Complete liberty of contradicting and disproving our opinion, is the very condition which justifies us in assuming its truth for purposes of action; and on no other terms can a being with human faculties have any rational assurance of being right.”

And modern science, history, and other academic disciplines rest largely on this principle. A theory is viewed as likely right (not certainly right) only to the extent that it has withstood, and continues to withstand, challenge. If everyone can try to refute the theory, but experts remain unpersuaded, that is reason for people—including those who lack the knowledge to independ­ent­ly evaluate the theory themselves—to rely on the experts’ consensus. Most of us, for instance, know little about what happened to Armenians in the Ottoman Empire in 1915 and why it hap­pened. If we learn that historians have generally reached the consensus that the cause was a deli­ber­ate genocide by the Ottomans, we would have some reason to endorse that.

But if we were to learn that this consensus endured not because it could constantly be challenged, but because it was enforced by threat of criminal prosecution or civil liability or violent attack or firing for those who challenged it, that would be reason to doubt the consensus. Perhaps, after all, there might be powerful arguments against the consensus, but ones that scholars have not been allowed to fairly consider.

To be sure, one might in principle distinguish lies, even in scientific debate, from honest errors. But in practice, given how hard it is for government officials to accurately determine whether someone is sincere—especially when that someone holds beliefs the officials sharply condemn—and the chilling effect caused by this risk of error, it may be best to provide categorical protection rather than just an “actual malice” standard. Such an argument, made by Justices Black, Douglas, and Goldberg, didn’t carry the day as to public official libel lawsuits in New York Times v. Sullivan. But perhaps it is what tipped the balance in favor of categorical immunity as to statements about the government by the Sullivan majority and in favor of categorical immunity as to statements about science, history, and the like in Alvarez.

What makes this categorical immunity as to science and history, though, is precisely that there is an institutional academic system with the ability and the incentive to deal with these questions. On one hand, this system benefits from the freedom to make false statements, since that freedom also frees people to make claims that are true but appear at first to be incorrect; or to make claims that prove to be partly false but partly true; or even to make claims that are entirely false but for which the process of disproving them produces important insights—all of which have been familiar phenomena in the history of science.

On the other hand, the system offers the prospect of “good [counsels]” (here, in the sense of likely accurate scientific or historical information) being a practically moderately effective “remedy” for “evil counsels” (here, in the sense of likely false claims about science or history) and not just a theoretically “fitting” one. This system of expert debate provides not just the abstract “power of reason as applied through public discussion,” but also a real likelihood that a body of researchers, writers, and commentators will really apply reason to such matters. (I speak here of the entire system of maintaining and applying knowledge, which includes journalists and other knowledgeable laypeople as well as academics.)

The same might well be true as to speech about the government. Disputes about what is actually happening in a war, or whether there was fraud in an election, or whether racially disparate arrest rates by a particular police department stem from racially disparate crime rates or from discrimination by the police (or both), might likewise need an ongoing, iterative process like the scientific process. Here too, any consensus might only be credible to the extent that it withstands ongoing attempts at refutation. And here too, there are institutions that not only profit from the ability to freely debate the issue, but that have the ability and the incentive to rebut error: journalists, political activists, and the criticized government agencies themselves.

Indeed, when a government agency is accused of misconduct, its leaders will have huge incentives to rebut the accusations—both selfish incentives and public-regarding ones. They will be able to use government resources to offer such rebuttals; election administration agencies, for instance, have a legitimate reason to spend money to point to evidence why the election results should indeed be trusted. And they will likely be able to get the attention of activists and journalists who are interested in the subject.

Of course, none of these institutions are at all guaranteed to be reliable. They may fail to notice some falsehoods. They may fail to persuade the public—or at least a portion of the public—that some of the falsehoods are false. Indeed, they may sometimes themselves spread the falsehoods.

But it’s all relative, which is also where the “peril” of legal suppression comes in. The institution of legal decision-making about the truth or falsity of a claim is not guaranteed to be reliable, either. Indeed, prosecutors, civil enforcers, and judges may be especially likely to be emotionally and politically invested in government policies, or in politically charged claims about history or science. Even jurors, as citizens, may be particularly likely to support one side or another. As between allowing the truth to be determined by the process of civil or criminal litigation, and having it be determined by ongoing debate among scholars, journalists, and others, the latter approach is likely to on balance be better.

2. Facts and opinions

And this is especially so given the frequent difficulty of drawing the line between opinion and factual assertions, exacerbated by the human tendency to draw that line based on our own attitudes toward the merits of the speech. This difficulty has been evident throughout the history of attempts to regulate alleged “fake news.”

Consider, for instance, United States v. Cooper, one of the Sedition Act of 1798 cases. Cooper was convicted of false and malicious statements based essentially on these passages in a leaflet:

Nor were we yet ... threatened [in 1797], under [President Adams’] auspices, with the existence of a standing army. Our credit was not yet reduced so low as to borrow money at eight per cent, in time of peace. …

Mr. Adams had not yet ... interfered, as president of the United States, to influence the decisions of a court of justice—a stretch of authority which the monarch of Great Britain would have shrunk from—an interference without precedent, against law and against mercy. This melancholy case of Jonathan Robbins, a native citizen of America, forcibly impressed by the British, and delivered up, with the advice of Mr. Adams, to the mock trial of a British court-martial, had not yet astonished the republican citizens of this free country; a case too little known, but of which the people ought to be fully apprised, before the election, and they shall be.

Lies! said Justice Chase to the jury (and the jury through its verdict agreed): a “scandalous and malicious libel,” containing three “false” elements: The charge related to the nation’s credit was supposedly false because the late 1790s weren’t really a “time of peace.” The condemnation of the president’s conduct in the Jonathan Robbins matter was supposedly false because the president was required by treaty to hand Robbins over. And the “standing army” statement was supposedly false because (Justice Chase reasoned) the army couldn’t be “standing” given that, in accordance with the Constitution, its expenses could only be authorized for two years.

Yet it seems clear that these were actually opinions. Whether one calls America’s experience with France in 1798-1800 a time of “war” or “peace” is a matter of judgment and definition, not of fact. Indeed, it is sometimes called a “Quasi-War,” and Adams himself later called it a “half War,” reflecting the uncertainty of the “war”/“peace” distinction. Likewise, whether “standing army” refers to any army that is in place for an extended time or only to an army that operates without need for frequent congressional reauthorization is a matter of definition. And whether Adams’ actions with regard to Robbins were “against law” is likewise a matter of opinion.

The same is true with some of the statements in Schaefer, the WWI “false reports” case. The alleged mistranslation of “breadlines” may have been a pure factual error (whether accidental or deliberate), but that was a secondary part of the Court’s opinion. The more extended discussion of alleged falsehood came here:

The aid ... asserted to have been rendered to England by President Wilson was represented to have been in opposition to the wishes of the people expressed, “by the unwillingness of their [the United States’] young men to offer themselves as volunteers for the war. But it will not rest there. The call for peace will come from the masses and will demand to be heard. And the sooner the better. No blood has been shed yet, no hate or bitterness has yet arisen against Germany, who has never done this country any harm, but has sent millions of her sons for its upbuilding. The sooner the American people come to their senses and demand peace, the better and more honorable it will be for this country.”...

[The article] was ... reinforced by another article July 7, 1917. It (the latter) had for headlines the words “The Failure of Recruiting,” and recruiting failed, was its representation, notwithstanding an “advertising campaign was worked at high pressure” and “all sorts of means were tried to stir up patriotism.” Its further declaration was that “Germany was represented as a violator of all human rights and all international law, yet all in vain. Neither the resounding praises nor the obviously false accusations against Germany were of any avail. The recruits did not materialize.” The cause was represented to be “that the American, who certainly cannot be called a coward” did “not care to allow himself to be shot to satisfy British lust for the mastery of the world.” And “the people instinctively recognize and feel” that “the pro-British policy of the Government,—is an error, which can bring nothing but injury upon this country.” It was then added that “the nation therefore” was doing the only thing it could still do, “since its desires were not consulted at first.” It refused “to take part.” ...

[The] statements were deliberate and willfully false, the purpose being to represent that the war was not demanded by the people but was the result of the machinations of executive power, and thus to arouse resentment to it and what it would demand of ardor and effort. …

Yet surely the judgment of whether “the war was ... demanded by the people” or whether it “was the result of the machinations of executive power” (or a mixture of the two) is a matter of opinion. A statement about what “the people instinctively recognize and feel” is obviously a guess, not an assertion of what can be empirically proved true or false.

And we see this continuing today. Thus, for instance, in 2020 an advocacy group sued Fox News claiming that its coronavirus coverage was “false,” but many of the alleged falsehoods were opinions, such as about just how dangerous COVID was. Likewise, the top item on then-President Trump’s “fake news” award list—Paul Krugman’s wildly incorrect prediction the day after President Trump’s election that “If the question is when markets will recover, a first-pass answer is never”—was an opinion, as predictions inherently are.

To be sure, this risk is present in all false statement cases, including ordinary libel lawsuits. Courts routinely have to decide whether a statement is a potentially actionable factual assertion or a constitutionally protected opinion, and are often alleged to have erred on the matter. The risk of erroneously punishing opinions is not sufficient to categorically bar all liability for lies. But it may be one factor in favor of forbidding legal liability when alternative institutions can help correct the record.

C. When alternative institutions are largely absent

The situation is quite different with ordinary libel, fraud, or perjury cases, where we don’t expect resolution by an iterative process of theorizing and attempted rebuttal, or speech and counterspeech. When it comes to whether Daniel Connaughton offered improper favors to a grand juror, or whether Jeffrey Masson made particular statements to Janet Malcolm, we count on the justice system to determine the truth (just as we’d count on it to determine the truth if Connaughton were prosecuted for the favors, or if someone is sued for sexual harassment based on statements he made to a colleague).

And this toleration of libel lawsuits likely implicitly stems from the fact that it would be futile to expect alternative institutions—such as the academy or the media—to offer a more helpful resolution of these matters. The questions are generally too narrow to engage those institutions’ attention, except perhaps to the extent of one or two newspaper articles. In principle, counterspeech could effectively rebut the lies; but in practice few people (other than the plaintiff) will have much interest in engaging in this counterspeech. In the absence of a government-provided forum for determining (however imperfectly) the truth, there is likely to be no other forum.

On top of that, the specificity of the target may call for some degree of compensation—at least in cases of damage to reputation—and not just for setting the record straight. And at least in most libel cases, government officials may be expected to be more dispassionate about whether some speaker indeed lied about some target than they would be in disputes about global warming or the history of Jim Crow, or the presence or absence of election fraud.

In a sense, this echoes the Gertz v. Robert Welch, Inc. distinction between public figures and private figures. Public figures, the Court said in Gertz, “enjoy significantly greater access to the channels of effective communication and hence have a more realistic opportunity to counteract false statements than private individuals normally enjoy”: They can more effectively use “self-help” to “contradict the lie or correct the error and thereby to minimize its adverse impact on reputation.”

Perhaps at the heart of the Alvarez categorical immunity for certain kinds of lies is a stronger version of that statement. Government entities have still more opportunity to counteract false statements both about themselves and about science and history because of the resources they can marshal and the likelihood that the media and advocacy groups will indeed at least listen to them. And there are powerful institutions—though, of course, never perfectly reliable or effective ones—for counteracting false statements about history or science.

IV. Implications for Some Controversies

If I’m right that the question here is one of comparative effectiveness of different truth-finding institutions, then this might bear on how any particular kind of false statement should be classified. Let me just briefly sketch out some thoughts on several such areas, though each merits an extended article on its own.

A. Lies in election campaigns

Lies in election campaigns are in some respects especially dangerous because they often happen shortly before the election, when there is little time to respond. When they are said by a candidate, they are also a form of financial fraud: The candidate is trying to get not just power but a paying job through knowing falsehood. At the same time,

In a political campaign, a candidate’s factual blunder is unlikely to escape the notice of, and correction by, the erring candidate’s political opponent. The preferred First Amendment remedy of “more speech, not enforced silence,” thus has special force.

And what is true of factual blunders is likely true of deliberate falsehoods as well.

The cases on the subject are split, but I’m tentatively inclined to agree with the recent state supreme court and federal appellate decisions that conclude that, on balance, allowing prosecutions for such lies is too dangerous. The Massachusetts Supreme Judicial Court’s 2015 decision in Commonwealth v. Lucas is the most recent articulation of the argument:

The proper institution for dealing with such lies, for all its flaws, is the prospect of “counterspeech” by the opponent.

Such a statute “may be manipulated easily into a tool for subverting its own justification, i.e., the fairness and freedom of the electoral process, through the chilling of core political speech,” through tricks such as “‘well-publicized, yet bogus, complaint[s] to the [governmental enforcement body] on election eve.’” “[B]y the time of the probable cause hearing the election may well be over and the damage will be done.”

This is especially so because “the distinction between fact and opinion is not always obvious,” especially in rushed decisions about whether to issue a criminal complaint.

“[E]ven in cases involving seemingly obvious statements of political fact, distinguishing between truth and falsity may prove exceedingly difficult. Assertions regarding a candidate’s voting record on a particular issue may very well require an in-depth analysis of legislative history that will often be ill-suited to the compressed time frame of an election.”

To be sure, even the opinions that have struck down election lie statutes on these grounds have included language that seems to express openness to some restrictions. For instance, Lucas stressed that the problems with the law were made especially severe by the fact “that anyone may file an application for a criminal complaint,” so that “an individual, unconstrained by the ethical obligations imposed on government officials, will file an unmeritorious application ‘at a tactically calculated time so as to divert the attention of an entire campaign from the meritorious task at hand of supporting or defeating a ballot question [or candidate].’” Perhaps the result might have been different had the law only allowed complaints to be filed by prosecutors or other government enforcement officials. Still, I doubt the court would or should have come out differently in that situation, in part because such more traditional government-triggered enforcement would raise its own possible risks of selective enforcement and politically motivated prosecution.

Likewise, Lucas noted that the law wasn’t well-tailored to the concern about allegations that come too late in the campaign for the other side to effectively rebut them. Perhaps a law focused on just false statements in, say, the last three days of the campaign might be upheld as sufficiently narrow. (Of course, it would be unlikely that the law could be effectively enforced in a way that would yield a finding of falsity before the election, but perhaps the threat of post-election criminal punishment might still be an effective deterrent.) Still, on balance it seems to me that the traditional approach of candidates pointing out their rivals’ lies, often supplemented by the media chiming in, is the least bad of a bad set of options.

B. Lies about “election procedures”

Much of the concern about generally punishing lies during elections, though, stems from the broad range of lies that could be covered and the potential controversies about what is the truth, what is a factual claim, and what is opinion. Narrower restrictions might pose fewer problems.

This is particularly so with regard to lies about the when, where, how, and who of elections: For instance, lies about when polls close, where one can vote, whether one can vote online, by mail, and the like, and who is eligible to vote. These lies can generally be narrowly defined and tend to be easily verifiable; and many such lies are likely to happen shortly before the election, when established alternative institutions—election officials, candidates, the media, and others—might not have the time to undo the effects of the lie.

To be sure, the distinction between such statements and political lies more broadly is not completely sharp. In New York Times v. Sullivan, the allegation about King having been arrested seven times by the police was also a narrow and easily verifiable claim, and yet that too was seen as categorically not punishable. Moreover, there will sometimes be controversy about whether a particular statement is an obvious joke, as with this meme, which is the foundation of a criminal prosecution:

(This is similar to the disputes that often arise about whether an allegedly libelous statement is satire.) And there may also be controversies about what the true meaning of an ambiguous statement might be.

C. Lies about highest-level government officials/large corporations

This analysis suggests that perhaps, on balance, statements about government officials who are at the highest levels of government should also be categorically immune from liability. There is a comparable risk that prosecutors, judges, and jurors will be too partisan and emotionally invested to adjudicate these claims correctly. The media may well have enough incentive to figure out the truth about this; and such journalistic fact-finding, however imperfect, may be better than (and certainly much quicker than) jury fact-finding. The same may be true about statements about large corporations, especially given that those corporations have ample resources that they can quite legitimately use to rebut such allegations.

Perhaps on this point, even if not as to lower-level officials, the New York Times v. Sullivan concurrences were correct. At the same time, it may be good to have crisper lines here, such as between libels of the government and libels of individuals or nongovernmental entities, as opposed to lines between super-high-level officials and merely ordinary officials.

D. Statements about science, government, and the like in specific libel or fraud lawsuits

We see, then, that courts are generally allowed to determine whether a statement about an individual is true, or whether a statement said to government officials is true, or whether a statement aimed at getting money is true. They are generally not allowed to determine whether a statement about the government, or about history or science, is true (at least in the context of punishing speech, as opposed to litigating nonspeech claims). What happens, though, at the intersection of those two categories?

Say, for instance, that someone testifies before a grand jury—or even just tells government investigators—“I saw a police officer beat John Smith,” even without any reference to a specific police officer. If a prosecutor believes that the witness knowingly lied, could such a statement lead to a perjury or false statement prosecution? Or say that someone makes the same claim in a fundraising letter (“Give money to the Anti-Police-Brutality Foundation, so we can deal with abuse such as the police beating of John Smith”). If a prosecutor or a consumer protection agency claims that the author knowingly lied, could such a statement lead to a fraud prosecution?

Or let’s take an example of a speech that is not within an existing First Amendment exception, but that would likely be restrictable under intermediate scrutiny, the test adopted by the controlling Alvarez concurrence: Someone files an unsworn complaint with the police department, claiming the police beat him. If the prosecutor believes this is a knowing lie, could this lead to a prosecution for filing a false police report? (Three dissenting justices of the Minnesota Supreme Court recently suggested the answer is “no.”)

The same questions arise not just with statements alleging government misconduct, but also with statements about “history, the social sciences, the arts, and other matters of public concern.” The government may not ban newspaper articles about whether or not human activity is warming the Earth, or about whether there are biological differences between the sexes’ or races’ cognitive faculties. But say that a scientist testifies in court that his study showed that there are or are not such biological differences (perhaps in supporting or opposing an argument that a disparity between an employer’s workforce and its applicant pool stems from real differences in applicant quality and not from discrimination). Could he be prosecuted for perjury if there is evidence that he had deliberately falsified his research results?

Or say the scientist applies for a grant, or seeks contributions from the public, based on his claimed medical discoveries—could he be prosecuted for fraud if it is shown that he has admitted to friends that the discoveries are fake, or at least that the effects he claims are overstated? The Third U.S. Circuit Court of Appeals, for instance, has held that a writer of a book about the pope could be prosecuted for defrauding his publisher by falsely claiming to have conducted interviews with his subject. But several judges dissented, arguing that courts ought not pass judgment on the accuracy of claims in books, even when the falsehoods were seen as defrauding not just consumers but also the publishers.

In all of these situations, one can still say that “it is perilous to permit the state to be the arbiter of truth”—just as it would be perilous if the government tried to generally ban lies about itself, about science, about history, and about other “matters of public concern.” Prosecutors, judges, and juries may often be unreliable evaluators of what is a knowing lie and what is an unorthodox truth in such matters. And that’s true in a perjury, fraud, or false statement prosecution as much as in a hypothetical prosecution for seditious libel, racial libel, or Holocaust denial.

Thus, for instance, say someone wants to raise money for a foundation that would reject claims that the Ottoman government deliberately engaged in genocide of Armenian civilians during World War I. Doubtless in his fundraising letters, the speaker would assert that the Armenian genocide did not in fact happen. Yet if the government then prosecutes the speaker for fundraising fraud, this would pose much the same kind of threat to free debate about history that the Court was concerned about in Alvarez.

At the same time, disabling the government from prosecuting such lies (or, in some instances, allowing civil liability based on such lies) when those lies lead to tangible financial injury, or interfere with court proceedings or police investigations, can leave a good deal of harm unpunished and undeterred. This is why when it comes to individual libel, the harms of which are intensely focused on particular people, the Court allowed liability for knowing lies and rejected an argument for categorical protection. Again, the concurring justices in New York Times v. Sullivan made such an argument, precisely on the grounds that the legal system can’t be trusted in determining the truth or falsehood of allegations on matters of public concern (especially allegations about public figures). But the majority refused to go along.

Likewise, when it comes to religious claims, the Court has long recognized that the government may not be an arbiter of truth. But the Court has nonetheless allowed fraud prosecutions based on a showing that someone seeking religious donations is knowingly lying about his supposed religious experiences and miraculous healing powers. That was the holding of United States v. Ballard in rejecting a Free Exercise Clause defense to a fraud prosecution, but the logic of the Court’s analysis would apply to a Free Speech Clause defense as well.

This question has so far been little explored, and my views on it are tentative. But it seems to me that such lies should be punishable at least when they (1) constitute fraud, perjury, false statement to government officials, or some other generally punishable sort of lie, and (2) are claimed to be based on personal knowledge.

Lies that are supposedly based on personal knowledge—“I saw the police beat John Smith,” “I conducted my experiments this way,” “I have healed the sick by laying on my hands,” “I interviewed the pope for my book”—tend to be particularly dangerous. Speakers who claim firsthand knowledge of some incident are essentially claiming to have superior expertise about the incident: They were present, and (usually) very few other people were. As a result, people considering the matter tend to feel inclined to take such eyewitness claims seriously.

In such situations, it’s especially important that lies be deterred and punished by the legal system. The lies are potentially more harmful, or at least more directly harmful, than mere allegations in newspapers because they involve attempts to get money, attempts to deceive the judicial process, or attempts to distract government officials in their investigations. And the lies are potentially more harmful than broad assertions not based on personal knowledge (e.g., claims that the Armenian genocide did not happen), precisely because the speaker’s claim of firsthand knowledge is especially likely to be believed and is especially hard to rebut.

On the other hand, someone who is yet another voice on the subject of the Armenians in World War I, the JFK assassination, or racial differences is only one source among many. Most listeners, especially ones whose job it is to evaluate his statement (such as government officials) or whose money is on the line (such as book publishers), will give the person’s claims comparatively little credit—why credit them more than all the other experts’ claims?—and will find it relatively easy to find other sources on the subject, precisely because the speaker is just one claimed expert among many.


The Court has never precisely explained when lies are constitutionally protected and when they are punishable. But the particular lines that it has drawn seem generally consistent with a comparative institutional approach to responding to lies. Government determination of which assertions are false and should therefore be punished is always perilous. When institutions—scholars, the government as speaker, the media, perhaps opposing election campaigns—there is a way to avoid the peril while still rebutting the lies. It’s imperfect, but it’s better than the alternative of government coercion; in such a situation, “the fitting remedy for” lies, as well as for “evil counsels,” is rebuttal.

But in other situations, when the harm from lies is serious and alternative institutions for rebutting the lies aren’t likely to exist, the government can indeed try to deter the lies by the threat of criminal prosecution or civil liability. That explains the constitutionality of properly limited libel law, and of the laws punishing fraud, perjury, and the like. And that can help decide where the lines can be drawn in the areas that remain unsettled.


Many thanks to Katy Glenn Bass, Genevieve Lakier, and Catherine J. Ross for their helpful comments on the paper.


Printable PDF 


© 2022, Eugene Volokh.


Cite as: Eugene Volokh, When Are Lies Constitutionally Protected?, 22-10 Knight First Amend. Inst. (Oct. 19, 2022), [].