Anti-Libel Injunctions and the Criminal Libel Connection

My new article, forthcoming in the University of Pennsylvania Law Review late this year -- I'd love to get feedback, while there's still plenty of time to edit it.

Here's the Introduction (minus the footnotes)—you can find the whole text here:

[* * *]

An injunction against libel, which carries the threat of prosecution for criminal contempt, is like a miniature criminal libel law—just for a particular defendant, and just for statements about a particular plaintiff. That is its virtue. That is its danger. And that is the key to identifying how the First Amendment and equitable principles should constrain such injunctions.

From the 1960s to the 1990s, libel was conventionally understood to be controlled (to the extent that it can be controlled) by the threat of civil damages. Criminal libel was seen as an anachronism. Injunctions were seen as unavailable in libel cases, whether because "equity will not enjoin a libel," because anti-libel injunctions were unconstitutional prior restraints, or for some related reason. My sense is that many people still assume this is so.

When one considers the libel cases from the mainstream First Amendment casebooks—New York Times Co. v. Sullivan, Gertz v. Robert Welch, Inc., and Dun & Bradstreet v. Greenmoss Builders—focusing on damages makes sense. For libels by a newspaper, magazine, or credit rating agency, damages are likely both a fair remedy and a reasonable deterrent. Criminal liability seems like overkill, and an injunction is usually pointless: The defendants there aren't likely to keep saying false things about the plaintiff in any event, especially after a libel judgment, so nothing will need enjoining. Print defamation is generally a short sharp shock, which causes damage in a way that an injunction can't stop, but that damages might deter. Even defamation in a credit report will usually stop when the credit agency is shown its error (and especially when it is ordered to pay damages).

But the judgment-proof libeler, always a hazard, has become still more common—and more dangerous—in the Internet age. The Internet lets speakers publish libels at little cost to a potentially broad audience, and these libels can cause enduring damage. Every time someone types a plaintiff's name into Google, the libels pop up again. Moreover, 47 U.S.C. § 230(c)(1) generally immunizes intermediaries, such as search engines or online service providers, that do have money. In any practical sense, damages awards do not leave plaintiffs in such cases with "an adequate remedy at law"; damages cannot be collected from the judgment-proof, and cannot effectively deter them. If libelers who lack money are to be deterred, criminal punishment is the one tool that can do the job.

Consider, then, several different ways that such criminal punishment can be threatened. Assume that judgment-proof Don says Paula cheated him in business, and Paula thinks he's lying. We can imagine several possible responses:

Criminal libel: Paula goes to the prosecutor, who tells Don, "We still have a criminal libel law in this state; I think your statements about Paula are lies, and if you keep libeling her, I'll prosecute you for criminal libel." That doesn't violate the First Amendment, as I'll discuss in Part I, though it may be condemned as too likely to chill speech (especially since prosecutors can prosecute under such statutes even without giving warning).

The catchall injunction: Paula goes to court, and gets an injunction against Don saying, "You may not libel Paula, or you will be prosecuted for criminal contempt." That, I'll argue in Part II, also doesn't violate the First Amendment, though it may also chill speech too much, and appellate courts generally frown on such injunctions (though some trial courts have imposed them).

The specific preliminary injunction: Paula goes to court, and gets a preliminary injunction against Don saying, "You may not say that Paula has cheated you in business, or you will be prosecuted for criminal contempt." Appellate courts generally condemn such injunctions (though again some trial courts have imposed them). Though the injunction is less chilling than criminal libel law, it fails to offer some of the important procedural protections than criminal libel law does (as Part III discusses). In particular, such a specific preliminary injunction lets speech be suppressed based on just a likelihood-of-success-on-the-merits preliminary finding, rather than a full decision on the merits, following a trial.

The specific permanent injunction: Paula goes to court, and gets a permanent injunction against Don saying, "You may not say that Paula has cheated you in business, or you will be prosecuted for criminal contempt." More than thirty states have allowed such injunctions, at least in some situations, and only six have generally rejected them. If "equity will not enjoin a libel" was ever a firm rule, it isn't so now. But, I'll argue in Part IV, that these injunctions also fail to provide certain important procedural protections.

The hybrid permanent injunction: Paula goes to court, and gets a permanent injunction against Don saying, "You may not libelously say that Paula has cheated you in business, or you will be prosecuted for criminal contempt." This sort of injunction, I'll argue in Part V, can provide the procedural protections that criminal libel law and catch-all injunctions offer—chiefly because the injunction by its terms only punishes speech if it's found libelous both at the injunction hearing and at the ultimate criminal contempt trial—but at the same time has the narrower chilling effect that characterizes the specific permanent injunction.

The hybrid preliminary injunction: Paula goes to court, and gets a preliminary injunction against Don saying, "You may not libelously say that Paula has cheated you in business, or you will be prosecuted for criminal contempt." I'll argue in Part VI that this also provides the constitutionally required procedural protections (unlike the widely condemned specific preliminary injunctions), but at the same time protects Paula against libel more quickly.

One way of understanding this is by focusing on exactly what kind of speech each remedy actually criminalizes:

Criminal libel law	All statements found by jury to be libelous beyond a reasonable doubt
Catchall injunction	All statements by Don about Paula found by jury at contempt trial to be libelous beyond a reasonable doubt
Specific preliminary injunction	Specific statements by Don about Paula found by judge, based on abbreviated hearing, to probably be libelous
Specific permanent injunction	Specific statements by Don about Paula found by judge at trial to be libelous by a preponderance of the evidence
Hybrid permanent injunction	Specific statements by Don about Paula found by judge at trial to be libelous by a preponderance of the evidence and then found by jury at contempt trial to be libelous beyond a reasonable doubt
Hybrid preliminary injunction	Specific statements by Don about Paula found by judge, based on abbreviated hearing, to probably be libelous and then found by jury at contempt trial to be libelous beyond a reasonable doubt

I will argue that properly crafted criminal libel laws and catchall injunctions are constitutional, though probably too broad as a policy matter; that specific injunctions, permanent or preliminary, are unconstitutional; and that hybrid injunctions, permanent or preliminary, are constitutional and may indeed be well-advised.

Properly crafted anti-libel injunctions are thus permissible under the First Amendment, if a state chooses to implement them—as some state courts and state legislatures have done. (I set aside here injunctions that forbid more than just the libelous statements; those are generally unconstitutionally overbroad, and I discuss them in a separate article.) But a court considering whether to allow such injunctions requires a difficult judgment about state law, again precisely because the injunctions effectively create a mini-criminal-libel law.

For instance, about a dozen states have criminal libel laws, and most of those states at least occasionally use them. A properly crafted anti-libel injunction would thus cut out the prosecutor's role, and the opportunity for prosecutors to use their discretion to decline to launch a criminal libel prosecution: A contempt-of-court prosecution for violating an injunction can be started by the court itself—or, in some states, even by the plaintiff—with no need for prosecutorial approval. As I'll discuss in Part VII, courts need to decide whether this is a feature or a bug.

In Part VIII, I'll turn to the many other states have repealed their criminal libel laws. Should courts view the legislative judgment behind repealing criminal libel laws as condemning all criminal punishment for libel, in which case even the narrow injunctions should be unavailable? Or should they view the legislative judgment as condemning only the broad chilling effect of normal criminal libel laws, in which case the narrow injunctions would be permissible? These are hard questions to answer, but state courts need to ask them when deciding whether to recognize a novel remedy that seems to recriminalize what the legislature decriminalized.

As I'll discuss in Part IX, many libel cases end up in federal court because of diversity of citizenship between the parties. Even if a federal court concludes that an injunction in such a case would be consistent with the First Amendment, it should also (following Erie) consider whether such an injunction is consistent with state law, or with how it expects state courts will develop that law.

Finally, in Part X, I'll talk about how this analysis can apply to injunctions in other tort cases: false light, slander, trade libel, slander of title, interference with business relations, and the disclosure of private facts.