Misinformation Is Not a Legal Category, and Government Should Stop Pretending

There is a word that appears in op-eds, task forces, grant proposals and congressional letters with the confidence of a statute, and it appears nowhere in the United States Code as a category of forbidden speech. The word is "misinformation." Everyone is against it. Nobody can tell you what it is. It is the rare enemy that every committee has sworn to defeat and none has managed to describe, like a dragon with a very thorough press office.

That is not a rhetorical flourish. It is the whole problem, and this post is a rebuttal. First I will state the case for government action against misinformation as fairly as I can. Then I will take it apart in order.

Define the terms first

Terms matter, because the slipperiness of this debate lives in its vocabulary.

Falsehood is a statement that does not match the facts. Misinformation, in the way the word is commonly used, is a falsehood spread without intent to deceive. Disinformation is a falsehood spread on purpose. Fraud is a falsehood told to obtain money or property that the victim relies on to his harm. Defamation is a false statement of fact that damages a particular person's reputation. Perjury is a lie under oath.

The first two words are descriptions of speech in the broad. The last three are legal categories. They have elements, defences, burdens of proof, and a long line of cases. The difference is not a technicality. A legal category tells a citizen, ahead of time, what he may not do, and lets a court decide whether he did it. A description tells a citizen only that a powerful person disapproves.

Hold that distinction in your head. Everything below depends on it.

The case for official action, stated at its strongest

Here is the best version of the other side, and I want its defenders to recognise themselves in it.

Democratic self-government depends on citizens who can distinguish true claims from false ones. Falsehoods can now travel to millions of people in hours, and a claim that is believed can change a vote, empty a bank, or end a life. A government that does nothing while a lie spreads is abandoning its citizens to those who profit from deception. The First Amendment protects the marketplace of ideas, but a marketplace needs rules against counterfeits, and misinformation is a counterfeit. Other free democracies regulate it. Surely the United States, which regulates false advertising and securities fraud, can regulate the dangerous falsehoods that threaten the democracy itself. If the choice is between a small amount of official discretion and a large amount of public confusion, the argument goes, the discretion is the lesser evil.

That is not a foolish argument. Its concerns are real, and it deserves an answer. It gets five.

The answer, in five numbered points

One: the Supreme Court has already said false speech is generally protected

In United States v. Alvarez (2012), the Court struck down part of the Stolen Valor Act of 2005, which made it a crime to falsely claim to have received military decorations, including the Medal of Honor. Xavier Alvarez, an elected member of a California water district board, had falsely said he was a recipient of the Medal of Honor. He lied, repeatedly and brazenly. As résumé padding goes, it was the equivalent of claiming you invented the wheel and also the road. His lie was about the most honoured thing in American military life. And six Justices still voted that the statute violated the First Amendment.

Justice Kennedy's plurality opinion rejected the government's claim that false statements of fact are a category outside First Amendment protection, and it declined to read earlier cases as saying so. The plurality said that the Court had never endorsed the notion that false speech, as such, lacks protection. Justice Breyer, concurring in the judgment with Justice Kagan, reached the same result by a different route: he would have applied intermediate scrutiny and found the statute's reach too broad, in part because a narrower statute, one limited to harms like fraud, would do the job. Justice Alito, joined by Justices Scalia and Thomas, dissented, arguing that the Act reached only a narrow class of false statements of fact.

Notice what that means. If a man may lie about the Medal of Honor without the criminal law reaching him on that ground alone, then a statute against "misinformation" about a policy dispute has a very steep hill to climb. The Stolen Valor Act targeted a discrete, verifiable lie about a specific, checkable fact, and it lost. A campaign against "misinformation" targets a fuzzy class of claims, many of them contested, about matters on which the facts are in dispute. If the first fails, the second does not stand a chance.

The plurality also made a point that deserves more attention. The Court observed that the government had not shown that counterspeech would be inadequate. The remedy for Alvarez was exposure, and exposure worked: he was exposed by the press and public ridicule before he was ever prosecuted. The truth was available. It had only to be spoken.

Two: the exceptions are narrow, defined and old

The Court in Alvarez did not say that all lies are free. It said that the unprotected categories involving falsity are those where the falsehood is tied to a legally cognizable harm and the category has been recognised for a long time.

Defamation requires a false statement of fact about an identifiable person, and in cases involving public officials and public figures, New York Times v. Sullivan (1964) added that the speaker must have acted with "actual malice," meaning knowledge of falsity or reckless disregard for the truth. Fraud requires a misrepresentation made to obtain something of value that induces reliance. Perjury requires an oath. False statements to federal investigators and false claims to obtain government benefits are punishable because they interfere with the government's own proceedings. Impersonating an officer is punishable because it is a fraud on the person who obeys. Incitement, under Brandenburg v. Ohio (1969), requires advocacy directed to producing imminent lawless action and likely to produce it. True threats are punishable because they are acts of intimidation, not arguments.

Every item on that list has elements. Every one requires a showing that the falsehood did something, to somebody, that the law recognises as a harm: robbed him, defamed him, endangered him, deceived a tribunal. None says "wrong, and widely shared." The mere wrongness of a claim is not the offence. This is the point the misinformation campaign cannot get around. It does not want to punish fraud, because fraud is already illegal. It wants to reach the claims that fall outside those categories, the ones that are false but harm no particular victim in a way the law recognises. That is precisely the territory the law has deliberately left to argument.

Three: "misinformation" has no definition a court could apply

Ask the following questions of any proposal.

Who decides? A federal agency? A panel of experts? A private company acting on a government tip?

By what standard? A claim is false only if it is contrary to the facts. But many of the claims at issue in the great political controversies are not facts. They are predictions, forecasts, interpretations, and value judgments. "This policy will raise prices" is a prediction. "This law is unconstitutional" is a legal opinion. "This programme works" is a judgment about evidence on which experts divide. A statute that said "no false claims about the economy" would require a judge to rule on which economist was lying.

When does falsehood become "dangerous"? The word that the argument leans on is harm, but harm in the sense of a persuaded citizen voting for the wrong candidate is the whole point of an election.

A statute that tried to settle which economist was lying would need a judge with a crystal ball and a very patient clerk. The vagueness is not a drafting flaw. It is intrinsic to the project. A law that is clear enough to be applied by a court is a law against fraud or defamation, and we have those. A law that reaches beyond them is a law that must be applied by a person with discretion, and discretion over speech is a power the Constitution takes away.

Four: the Sedition Act of 1798 is what this looks like when it is tried

We have run this experiment. In 1798, with the country anxious about a possible war with France and about foreign agents, the Federalist Congress passed the Sedition Act. It made it a crime to write, print or publish "any false, scandalous and malicious writing" against the government, Congress or the President with intent to defame them or bring them into contempt. It had a truth defence, which was more than English law gave, and its defenders said, accurately, that it only punished falsehood.

That was the argument. Only falsehood. Only false and malicious. And the people who administered it decided what was false. Prosecutions fell on Republican newspaper editors who criticised the Adams administration, and on Matthew Lyon, a Vermont congressman who was convicted and jailed and then re-elected while in jail. The statute expired in 1801. Jefferson pardoned those convicted. In New York Times v. Sullivan, the Court observed that although the Sedition Act was never tested in the Supreme Court, the attack on its validity "has carried the day in the court of history."

The lesson of 1798 is not that its sponsors were villains. They believed they were fighting falsehood in a dangerous time. The lesson is that "only falsehood" is not a limit. It is a license, because the official who applies it is always sure that the claim he dislikes is the false one. For the whole story see The Sedition Act of 1798 Jailed a Congressman Who Won Reelection From His Cell.

Five: official truth changes, and the people who doubted it were sometimes right

This is the point I would put to every advocate of official misinformation policy, and I would put it gently: are you sure that what officials believe today will be what officials believe in ten years?

Consider two cases from the public record, which I mention because they involve government accounts of events that were later contradicted, not because anyone acted in bad faith. In August 1964, the administration reported attacks on American destroyers in the Gulf of Tonkin, and Congress passed the Tonkin Gulf Resolution on the strength of the account. Material declassified in later decades, including a National Security Agency historical review, indicated that the second reported attack, on August 4, did not occur as described. And in 1971 the Pentagon Papers, a classified study of the Vietnam war, were published by newspapers over the objections of the government, which sought to stop them. The Supreme Court refused to enjoin publication in New York Times Co. v. United States. Those who spoke against the official account in each case were, at the time, speaking against what government said was true.

Now imagine a standing apparatus of the kind some propose, with authority to label, to downrank, or to "correct" claims contradicting the official account. In the 1960s it would have been aimed at the doubters. The better lesson is that you cannot design a referee that is only sometimes wrong in a direction you approve. Official error is a permanent feature of government, and the free exchange of dissent is the means by which it is caught. A society that licenses the correction of dissent has disabled its error-detection system. For a related argument about how government pressure on platforms works, see Murthy v. Missouri Lost on Standing, Yet the Record It Built Still Stands.

The Founding view: why this is not a close call

The First Amendment says: "Congress shall make no law ... abridging the freedom of speech, or of the press." It does not say "except false speech." The Founders knew that falsehood existed. They had read their own newspapers, which were savage, partisan, and often wrong. They did not respond by creating an authority to sort the true papers from the false ones. They responded by trusting the public.

Madison, in his Report on the Virginia Resolutions of 1800, argued against the Sedition Act on the ground that the power of the press to examine public men and measures is essential, even though it comes with abuses. That is the Founders' wager: the abuses of free debate are less dangerous than the abuses of a government that decides which debates are allowed. It is a wager, and wagers can lose. But the other side of the bet is a government with a monopoly on the definition of truth, and history has not been kind to that side.

I would add a point from the other direction, one that conservatives who like the idea of official action on a topic they dislike should keep in mind. Power to police "misinformation" does not stay with the officials you approve. It passes with each election to people whose view of which claims are dangerous may be very different from yours. A rule you would not give to your opponents is a rule you should not give to anyone. See also Milton's Areopagitica of 1644 Is Still the Best Case Against Licensing Speech for the oldest version of the argument, written in 1644.

What government may do instead

Opposing a misinformation bureaucracy is not the same as opposing government honesty. Here is what is left to government, and it is not nothing.

• It may speak: publish accurate data, correct its own errors and answer claims about its programmes, as a participant and not a judge.
• It may enforce existing law: prosecute fraud, perjury, defamation claims brought by injured persons, impersonation, and true threats.
• It may be transparent: publish what it asks of platforms and why, so that citizens can see any request that touches speech, a principle the Court reinforced in NRA v. Vullo. See NRA v. Vullo and the Unanimous Rule Against Government Jawboning.
• It may teach: support the civic education that lets citizens evaluate claims themselves.

And citizens may do what they have always done: argue back, check sources, and decline to share what they cannot verify. The right to be wrong is the same right as the right to be right. They arrive together.

The verdict

"Misinformation" is a useful word in conversation and a useless one in law. It collapses three different things: a statement that is false, a statement that is harmful, and a statement an official dislikes. The law has real tools for the second class where the harm is concrete, and it has none for the third, and it is right to have none. A government that treats the first as if it were the second is a government that has mistaken its own opinions for the public interest.

The honest answer to a falsehood is a truer statement, said louder, and the honest answer to a contested claim is an argument. Neither requires a licence from an agency.

The case in brief

1. "Misinformation" is a description, not a legal category, and it has no elements, defences or burden of proof that a court could apply.
2. United States v. Alvarez (2012) held that even a flagrant lie about the Medal of Honor could not be criminalised merely for being false, and counterspeech was the preferred remedy.
3. The truly unprotected categories, fraud, defamation, perjury, incitement and true threats, each require a concrete harm and are already enforceable.
4. The Sedition Act of 1798 shows that "only falsehood" is a licence for the official who decides what is false, and history has judged it harshly.
5. Official accounts have been revised later, as the Gulf of Tonkin and the Pentagon Papers show, so a society that suppresses dissent from the official account has disabled its own error-correction.

What you can do: Read the Constitution First Amendment text, and when a bill or agency proposal promises to fight "misinformation," ask your representatives, through my representatives, what the definition is and who applies it.