Proposed "information privacy" rules that give us the power to "control . . . information about ourselves" sound undeniably appealing. The First Amendment, however, generally bars the government from "control[ling the communication] of information," either by direct regulation or through the authorization of private lawsuits. This article argues that: (1) While privacy protection secured by contract turns out to be constitutionally sound, broader information privacy rules are not easily defensible under existing free speech law. (2) Creating new free speech exceptions to accommodate information privacy speech restrictions could have many unfortunate and unforeseen consequences. Most of the justifications given for information privacy speech restraints are directly applicable to other speech control proposals that have already been suggested, and accepting these justifications in the attractive case of information privacy speech restrictions would create a powerful precedent for those other restraints.

INTRODUCTION

I. INFORMATION PRIVACY SPEECH RESTRICTIONS

II. CONTRACT

A. Permissible Scope

B. Limitations

C. Government Contracts

D. Contracts with Children

A. Intellectual Property Rules as Speech Restrictions

B. Existing Restrictions as Supposed Precedents

1. Copyright law

2. Trademark law

3. Right of publicity law

4. Misappropriation and trade secret law

5. Summary

C. Functional Arguments for Upholding Information Privacy Speech Restrictions Under a Property Theory

1. Avoiding "free-riding" and unjust enrichment

2. Internalizing costs and maximizing aggregate utility

D. The Potential Consequences

A. What "Commercial Speech" Means

B. The Risks to Other Speech

A. The Argument

B. Theoretical Objections

C. Doctrine

D. The Experience Under the Two "Public Concern" Doctrines

E. Potential Consequences

1. Direct analogies

2. Indirect influence

A. Countervailing Constitutional Rights

B. Dignity, Emotional Distress, and Civil Rights

C. Keeping the Internet Attractive to Consumers

D. Preventing Misconduct and Crime

1. Discrimination

2. Fraud and violent crime

          Privacy is a popular word, and government attempts to "protect our privacy" are easy to endorse. Government attempts to let us "control . . . information about ourselves"⁽¹⁾ sound equally good: Who wouldn't want extra control? And what fair-minded person could oppose requirements of "fair information practices"?⁽²⁾

          The difficulty is that the right to information privacy--my right to control your communication of personally identifiable information about me--is a right to have the government stop you from speaking about me. We already have a code of "fair information practices," and it is the First Amendment, which generally bars the government from controlling the communication of information (either by direct regulation or through the authorization of private lawsuits⁽³⁾), whether the communication is "fair" or not.⁽⁴⁾ While privacy protection secured by contract is constitutionally sound, broader information privacy rules are not easily defensible under existing free speech law.

          Of course, the Supreme Court and even lower courts can create new First Amendment exceptions or broaden existing ones; and if the courts did this for information privacy speech restrictions, I can't say that I'd be terribly upset about the new exception for its own sake. Speech restrictions aimed at protecting individual privacy just don't get my blood boiling. Maybe they should, but they don't. Perhaps this is because, from a selfish perspective, I'd like the ability to stop others from talking about me, and while I wouldn't like their stopping me from talking about them, the trade-off might be worth it.

          Nonetheless, I'm deeply worried about the possible downstream effects of any such new exception. Most of the justifications given for information privacy speech restraints are directly applicable to other speech controls that have already been proposed. If these justifications are accepted in the attractive case of information privacy speech restrictions, such a decision will be a powerful precedent for those other restraints and for still more that might be proposed in the future.

          Thus, for instance, some argue that information privacy laws are defensible because they protect an intellectual property right in one's personal information.⁽⁵⁾ Such arguments don't fit well into the intellectual property exceptions to the First Amendment, which generally don't entitle anyone to restrict the communication of facts. And if we are to consider extending the existing exceptions, we should also consider that an intellectual property rights rationale is already being used as an argument for other speech restrictions: the proposed database protection law, the attempts to expand the right of publicity, and more. Before wholeheartedly endorsing the principle that calling certain information "intellectual property" lets the government restrict speech communicating that information, we should think about the consequences of such an endorsement.

          Similar problems confront the arguments that information privacy speech restrictions are constitutional because they restrain only commercial speech,⁽⁶⁾ restrain only speech that is not on matters of public concern,⁽⁷⁾ are narrowly tailored to a compelling government interest in protecting people's dignity, emotional tranquility, or safety,⁽⁸⁾ are needed to protect a countervailing civil right,⁽⁹⁾ or pass muster under a "context-sensitive balancing."⁽¹⁰⁾ First, for these arguments to succeed, existing First Amendment precedents would have to be substantially stretched. Second, the stretching may make the doctrine loose enough to give new support to many other restrictions. Bans on sexually themed speech might become justified under a "no public concern" rationale. Campus speech codes might be justified under a "countervailing civil right" rationale or a "narrowly tailored to a compelling government interest" rationale. Restrictions on online discussion about economic matters or on consumer complaints might be justified under a broadened commercial speech rationale. Restrictions on online distribution of information about encryption or drugs might be justified under a crime prevention rationale. And who knows what might be allowed under "context-sensitive balancing," which has in practice long been a tool for judges to justify a wide range of speech restrictions?

          In making these arguments, I will try to identify concrete, specific ways--doctrinal, political, and psychological--in which upholding certain kinds of information privacy speech restrictions could affect the protection of other speech. I will try to avoid making general slippery slope arguments of the "today this speech restriction, tomorrow the Inquisition" variety; the recognition of one free speech exception certainly does not mean the end of free speech generally, or else all would have been lost long ago. But slippery slope concerns are still quite sensible, especially when accepting a proposed speech restriction entails accepting a principle that is broader than the particular proposal and that can logically cover many other kinds of restraints.⁽¹¹⁾ Our legal system is based on precedent. Our political life is in large measure influenced by arguments by analogy. And many people's normative views of free speech are affected by what courts say: If the legal system accepts the propriety of laws mandating "fair information practices," people may becomes more sympathetic to legal mandates of, for instance, fair news reporting practices or fair political debate practices.⁽¹²⁾

          This article is an attempt to consider, as concretely as possible, the possible unintended consequences of various justifications for information privacy speech restrictions. I ultimately conclude that these consequences are sufficiently troubling that I must reluctantly oppose such information privacy rules. But I hope the article will also be useful to those who are committed to supporting information privacy speech restrictions, but would like to design their arguments in a way that will minimize the risks that I identify; and even to those who welcome the possibility that information privacy speech restrictions may become a precedent for other restrictions, because they believe the Court has generally gone too far in protecting, say, nonpolitical speech or speech that injures the dignity of others. Thinking ahead about the possible unintended implications of a proposal--even, and perhaps especially, if it seems viscerally appealing--is always worthwhile.

          My analysis throughout this article will focus on the government acting as sovereign, restricting what information nongovernmental speakers may communicate about people. I thus exclude restrictions that the government imposes on its own agencies, such as Freedom of Information Act provisions that prevent government revelation of certain data,⁽¹³⁾ or IRS or census rules that prohibit the communication of some tax or census data to other government agencies or to the public.⁽¹⁴⁾ Government agencies do not have free speech rights against their own governments; for instance, federal agencies must comply with congressional mandates, and creatures of the state such as city or county governments cannot claim rights against the state legislature.⁽¹⁵⁾ Whether speech by state agencies may be restrained by the federal government is a tougher question, but one that's beyond the scope of this article.⁽¹⁶⁾ By focusing on communication by nongovernmental speakers--reporters, businesspeople, private detectives, neighbors--I limit the inquiry to people and organizations that indubitably have free speech rights.

          I also exclude restrictions that the government imposes as an employer (e.g., telling its employees that they may not reveal confidential information learned in the course of employment), or as a contractor putting conditions on the communication of information that it has no constitutional duty to reveal (e.g., telling people who want certain lists from the Federal Election Commission that they may only get them if they promise not to use those lists for certain purposes,⁽¹⁷⁾ or telling litigants that they will get discovery materials only if they promise not to reveal them⁽¹⁸⁾). The government has long been held to have much broader powers when it's acting as employer or contractor, imposing constraints on those who assume them in exchange for government benefits or for access to government records, than when it's acting as sovereign, controlling the speech of private citizens.⁽¹⁹⁾ The unconstitutional conditions doctrine may impose some limits even on the government acting as employer or as contractor, but I will set these matters aside for purposes of this article.

          I also focus only on restrictions on communication. Other things that are often called privacy rules--the right to be free from unreasonable governmental searches and seizures, the right to make certain decisions about one's life without government interference, the right not to have people listen to you or watch you by going onto your property, the right not to have people electronically eavesdrop on your conversations, the requirement that credit bureaus notify consumers when credit reports about them are prepared, and the like--are outside the scope of my discussion.⁽²⁰⁾ Some of these laws, for instance restraints on government snooping or control, pose no First Amendment problems. For other laws, such as restrictions on nongovernmental gathering of information through nonspeech means, the First Amendment rules are unclear; but it is clear that the analysis of restrictions on information gathering is different from the analysis of restrictions on speech.⁽²¹⁾ It is the latter doctrine that is most fully developed, and that provides the most protection against government restrictions.

          These three exclusions merely reflect the fact that the strongest protection of free speech has long been seen as arising when the government is acting as sovereign, restricting the speech of private parties. And within this zone lie a variety of current and proposed speech restrictions:

          1. The "disclosure" tort, which bars the public dissemination of "nonnewsworthy" personal information that most people would find highly private,⁽²²⁾ and more specific state laws that forbid some such communications, for instance criminal laws forbidding the publication of the names of rape victims.⁽²³⁾ The uniting principle here is that it is particularly embarrassing to reveal a certain narrow range of information about people, for instance their medical histories, their criminal histories, their sexual practices, the images of their naked bodies, the contents of their conversations with their lawyers or psychiatrists, or possibly some of their reading or viewing habits.⁽²⁴⁾ These laws generally bar the communication of such information to the public, precisely because it's the publicizing of such potentially embarrassing information--either to large groups of people or possibly to smaller groups (friends, neighbors, and business associates) whose opinion the subject especially values--that is usually seen as especially offensive.

          2. Proposed restrictions on communication of all sorts of information about people, including matters that are not generally seen as especially private, for instance the food or clothes they buy, the stores (online or offline) they've shopped at, and so on.⁽²⁵⁾ Some such information may be embarrassing, but these laws do not focus on that; rather, they cover all information about a person, or at least all information that was gathered in a particular way (for instance, through online business transactions with that person).⁽²⁶⁾ And because embarrassment isn't the major concern, these laws also apply to communications aimed at fairly narrow groups of recipients about whose opinion most people care little--for instance, communications to another business that wants to sell things to you. The felt injury here is the perceived indignity or intrusion flowing from the very fact that people are talking about you or learning about you, and not the embarrassment flowing from the fact that people are learning things that reflect badly on you.

          3. Finally, a narrow range of restrictions aimed at preventing people from communicating information that might put others in danger of crime, for instance (in some contexts) the names of witnesses or jurors,⁽²⁷⁾ or databases of people's social security numbers that some can use to engage in fraud.⁽²⁸⁾

          Each of these categories covers some restrictions that are imposed only on the subject's business partners (for instance, bans on lawyers revealing information about their clients, or bans on businesses revealing information about their customers) and other restrictions that are imposed on everyone (for instance, bans on the media publishing embarrassing information that they learned from third parties, or property rights in information that bind everyone without regard to whether they've entered into any contracts). And of course these categories may overlap: Some restrictions aim at preventing embarrassment, preventing crime, and preventing communications about people more broadly.

A. Permissible Scope

          To begin with, one sort of limited information privacy law--contract law applied to promises not to reveal information⁽²⁹⁾--is eminently defensible under existing free speech doctrine. The Supreme Court explicitly held in Cohen v. Cowles Media that contracts not to speak are enforceable with no First Amendment problems.⁽³⁰⁾ Enforcing people's own bargains, the Court concluded (I think correctly), doesn't violate those people's rights, even if they change their minds after the bargain is struck. Some have criticized this conclusion on the grounds that it slights the interests of the prospective listeners, and this criticism has some force. Still, I think that ultimately the free speech right must turn on the rights of the speakers, and that it's proper to let speakers contract away their rights--and certainly this is the view that the Cohen v. Cowles Media Court took. Insisting that people honor their bargains is a constitutionally permissible "code of fair practices," whether information practices or otherwise.

          And such protection ought not be limited to express contracts, but should also cover implied contracts (though, as will be discussed below, there are limits to this theory). In many contexts, people reasonably expect--because of custom, course of dealing with the other party, or all the other factors that are relevant to finding an implied contract⁽³¹⁾--that part of what their contracting partner is promising is confidentiality. This explains much of why it's proper for the government to impose confidentiality requirements on lawyers, doctors, psychotherapists, and others: When these professionals say "I'll be your advisor," they are implicitly promising that they'll be confidential advisors, at least so long as they do not explicitly disclaim any such implicit promise.⁽³²⁾

          Laws that explicitly infer such contracts from transactions in which there's no social convention of confidentiality are somewhat more troublesome, especially if they require relatively formal disclaimers. Imagine, for instance, a law providing that all questions by reporters will be interpreted as implicitly promising not to quote the source by name in a published article, unless the source consents in writing after being given full disclosure of the true purpose for which the quote is to be used. Or consider a law providing that people who buy a product implicitly promise to give the seller equal space to respond to any negative article they publish about the product, unless the seller consents in writing after being given full disclosure of the true purpose for which the product is being bought.⁽³³⁾ Though journalists could avoid the restriction by getting the requisite explicit consent, the request for the consent may deter many of the sources and especially many of the sellers; and this in turn may deter journalists from publishing hostile reviews or stories that include quotes which show the sources in a bad light.

          These concerns may justify treating the Cohen v. Cowles Media principle as applicable only to those implied contracts where confidentiality really is part of most people's everyday expectations. This would mean the implicit contract theory could uphold laws that by default prevent lawyers, doctors, psychiatrists, sellers of medical supplies, and possibly sellers of videos and books from communicating information about their customers; but it wouldn't uphold laws that by default prevent reporters (who are notorious for communicating embarrassing things, not keeping them confidential) from revealing what was said to them, prevent consumers from reviewing products, or prevent sellers of groceries or shoes from communicating who bought what from them. I doubt that most of us expect that someone selling us our food is implicitly promising to keep quiet about what they sold us.⁽³⁴⁾

          On the other hand, I'm not sure that such a narrow application of Cohen v. Cowles Media is proper or ultimately workable. It's often hard to determine exactly what most people expect. When someone buys a video, especially a video whose title he wouldn't want associated with his name, he probably assumes that the video store won't publicize the purchase, at least in part because a video store that does publicize such purchases would lose a lot of business.⁽³⁵⁾ But is he assuming that the video store is promising not to publicize such a purchase? He probably isn't even thinking about this.⁽³⁶⁾

          If he is assuming such a promise, is he assuming that the video store is promising not to communicate information about such a purchase at all, or only promising not to pass it along to the public or his neighbors, while reserving the right to communicate it to others in the same business? Again, most buyers probably have not even thought about the matter. One advantage of statutory default rules is precisely that they clarify people's obligations instead of leaving courts to guess what people likely assumed.

          So I tentatively think that a legislature may indeed enact a law stating that certain legislatively identified transactions should be interpreted as implicitly containing a promise of confidentiality, unless such a promise is explicitly and prominently disclaimed by the offeror, and the contract together with the disclaimer is accepted by the offeree.⁽³⁷⁾ True, this might justify laws that treat reporters as implicitly promising that they won't reveal or even quote their sources, which troubles me. But so long as the implicit promise is genuinely disclaimable, I'm not too troubled. Even if this might eventually lead to the reporter hypothetical, I don't think too much would be lost; and what is gained from allowing statutorily defined default nondisclosure rules is the clear enforceability of promises that often are reasonably inferred by one of the contracting parties, and that can be important parts of the bargain.

          Furthermore, though Cohen v. Cowles Media involved traditional enforcement of a promise through a civil suit, there should be no constitutional problem with the government enforcing such promises through administrative actions, or using special laws imposing presumed or even punitive damages for breaches of such promises. I suspect that even with purely contractual remedies, the threat of class action suits could be a powerful deterrent to breaches of information privacy contracts by e-commerce sites, especially since the suits would create a scandal: In the highly competitive Internet world, a company could lose millions in business if people hear that it's breaking its confidentiality promises. But I think it would be constitutional for the government to try to increase contractual compliance either by providing an extra incentive for aggrieved parties to sue or by bringing a complaint itself. Though breach of contract has traditionally been seen as a purely private wrong, to be remedied through a private lawsuit, it's similar enough--especially when it's willful--to fraud or false advertising that there's nothing startling about a government agency such as the Federal Trade Commission prosecuting some such breaches itself.⁽³⁸⁾

          The great free speech advantage of the contract model is that it does not endorse any right to "stop people from speaking about me." Rather, it endorses a right to "stop people from violating their promises to me." One such promise may be a promise not to say things, and perhaps there may even be special defaults related to such promises or special remedies for breaches of such promises. But in any event, the government is simply enforcing obligations that the would-be speaker has himself assumed.⁽³⁹⁾ And such enforcement, in my view, poses little risk of setting a broad precedent for many further restrictions, precisely because it is founded only on the consent of the would-be speaker, and thus cannot justify the many other restraints--such as the Communications Decency Act, database protection legislation, and so on--to which the speaker has not consented.

B. Limitations

          Contract law protection, though, is distinctly limited, in two ways.

          First, it only lets people restrict speech by parties with whom they have a speech-restricting contract, express or implied.⁽⁴⁰⁾ If I make a deal with a newspaper reporter under which he promises not to identify me as a source, I can enforce the deal against the reporter and the reporter's employer, whom the reporter can bind as an agent. But if a reporter at another news outlet learns this information, then that outlet can publish it without fear of a breach of contract lawsuit. Likewise, there are no First Amendment problems with an employer suing an employee for breach of an express or implied nondisclosure agreement, but if the employee leaks the information to a newspaper, the employer can't sue that newspaper, at least under the Cohen v. Cowles Media theory.⁽⁴¹⁾ The newspaper simply hasn't agreed to anything that would waive its First Amendment rights, which is the premise on which Cohen v. Cowles Media rests. The disclosure tort would similarly not be justifiable under a contract theory.

          Second, Cohen v. Cowles Media cannot validate speech-restrictive terms that the government compels a party to include in a contract; the case at most validates government-specified defaults that apply unless the offeror makes clear that these terms aren't part of the offered deal. Thus, while the government may say "Cyberspace sales contracts shall carry an implied warranty that the seller promises not to reveal the buyer's personal information," it may not add "and this implicit warranty may not be waived, even by a prominent statement that is explicitly agreed to by a customer clicking on an 'I understand, and agree to the contract in spite of this' button."

          This flows directly from the rationale on which Cohen v. Cowles Media rests: "The parties themselves . . . determine the scope of their legal obligations, and any restrictions which may be placed on the publication of truthful information are self-imposed."⁽⁴²⁾ A merchant's express promise of confidentiality is "self-imposed"; so, one can say, is an implicit promise, when the merchant had the opportunity to say "by the way, I am not waiving my rights to speak about this transaction and am thus not promising confidentiality" but didn't do so. But when someone is legally barred from communicating, even if he explicitly told his contracting partner that he was making no such promise, then such an obligation is hardly "self-imposed" or determined by mutual agreement.

          Thus, I certainly do not claim that a contractual approach to information privacy, even with a large dollop of implied contract, is a panacea for information privacy advocates. As Paul Schwartz and others have pointed out, there is much that information privacy advocates may want but that contract will not provide.⁽⁴³⁾ I claim only that contractual solutions are a constitutional alternative and may be the only constitutional alternative, not that they are always a particularly satisfactory alternative.

C. Government Contracts

          Cohen v. Cowles Media does not decide to what extent the government, acting as contractor, may require people to sign speech-restrictive contracts as a condition of getting data from the government itself. This question raises thorny issues of unconstitutional conditions and often of the government's right to restrict access to government records that have historically been in the public domain (such as court records). Unfortunately, the Supreme Court case that some thought would help resolve this matter was decided on procedural grounds,⁽⁴⁴⁾ and the dicta in the many opinions in that case shed little light on exactly where the Court would have come down had it confronted the question on the merits.⁽⁴⁵⁾ I deal with this issue by setting it aside.

D. Contracts with Children

          Finally, this discussion of contracts presupposes that both parties are legally capable of entering into the contract and of accepting a disclaimer of any implied warranty of confidentiality. If a cyber-consumer is a child, then such an acceptance might not be valid. This is also a difficult issue, but one that is outside the scope of this Article.⁽⁴⁶⁾

A. Intellectual Property Rules as Speech Restrictions

          Partly because of the limitations of the contract theory, many information privacy advocates argue that people should be assigned a property right in personal information about themselves.⁽⁴⁷⁾ Such a property approach would bind everyone, and not just those who are in contractual privity with the person being talked about. Database operators would have to stop communicating information about people unless people give permission, even though the database operators have never promised, expressly or implicitly, to keep silent. Likewise, people could stop newspapers from publishing stories about them, even if the information was gleaned through interviews with third parties or was taken (with no contractual constraints) from public records.⁽⁴⁸⁾

          Calling a speech restriction a "property right," though, doesn't make it any less a speech restriction, and it doesn't make it constitutionally permissible. Broad, pre-New York Times v. Sullivan libel laws can be characterized as protecting a property right in reputation; in fact, some states consider reputation a property interest.⁽⁴⁹⁾ The right to be free from interference with business relations, including interference by speech urging a boycott as in NAACP v. Claiborne Hardware,⁽⁵⁰⁾ is often seen as a property right.⁽⁵¹⁾ A recent attempt at banning flag burning rested on the argument that the flag is the intellectual property of the United States, and that flag desecration thus violated property rights.⁽⁵²⁾ Restrictions on speech that uses cultural symbols in ways that the cultures find offensive might likewise be reframed as property rights in those symbols.⁽⁵³⁾ A ban on all unauthorized biographies, whether of former child prodigies,⁽⁵⁴⁾ movie stars, or politicians, can be seen as securing a property interest in the details of those people's lives. Similarly, an early right of publicity case took the view that people who aren't public figures have the exclusive right to block all photos and portraits of themselves, with no exceptions for news stories.⁽⁵⁵⁾

          Each of these "property rights," though, would remain a speech restriction.⁽⁵⁶⁾ A property right is, among other things, the right to exclude others;⁽⁵⁷⁾ an intellectual property right in information is the right to exclude others from communicating the information--a right to stop others from speaking. Like libel law, intellectual property law is enforced almost entirely through private litigation, but like libel law, it's still a government-imposed restriction on speech.⁽⁵⁸⁾ Some such restrictions may be permissible because there's some substantive reason why it's proper for the government to restrict such speech, but not because they are intellectual property rights.

          The question isn't (as some suggest) "who should own the property right to personal information?"⁽⁵⁹⁾ Rather, it's whether personal information should be treated as property at all--whether some "owner" should be able to block others from communicating this information, or whether everyone should be free to speak about it.

B. Existing Restrictions as Supposed Precedents

          The Court has, of course, upheld some intellectual property rights against First Amendment challenge, acknowledging that they are speech restrictions but holding that those restrictions were constitutional. In all these precedents, though, the Court has stressed a key point: The restrictions did not give the intellectual property owners the power to suppress facts. And this power to suppress facts is exactly the power that information privacy speech restrictions would grant.⁽⁶⁰⁾

1. Copyright law.

          But the main reason Harper & Row gave for this conclusion is that copyright law does not give anyone a right to restrict others from communicating facts or ideas. "[C]opyright's idea/expression dichotomy strike[s] a definitional balance between the First Amendment and the Copyright Act by permitting free communication of facts while still protecting an author's expression."⁽⁶²⁾ "No author may copyright his ideas or the facts he narrates."⁽⁶³⁾ Copiers "possess[] an unfettered right to use any factual information revealed in [the original]," though they may not copy creative expression.⁽⁶⁴⁾ There ought not be "abuse of the copyright owner's monopoly as an instrument to suppress facts."⁽⁶⁵⁾ "In view of the First Amendment protections already embodied in the Copyright Act's distinction between copyrightable expression and uncopyrightable facts and ideas," copyright law is constitutional.⁽⁶⁶⁾ Under the copyright exception to free speech protection, then, speech that borrows creative expression is restrictable, but speech that borrows only facts remains free.

          This limitation on the copyright exception is both theoretically and practically significant. Theoretically, this limitation is what leaves speakers ample alternative channels for communicating their message--speakers still possess "an unfettered right to use any factual information" that they please. Practically, people do indeed take advantage of this limitation. If a historian spent years of effort uncovering some remarkable, hitherto unknown facts, you may freely use those facts, as historians indeed do (though ethical rather than legal concerns may dictate that the users give credit to the original discoverer). Exactly where to draw the line between idea and expression is sometimes uncertain, but there are fewer uncertainties about the line between fact and expression; people who don't care about using the original author's rhetorical flourishes can definitely communicate facts that they've learned from others' work.

2. Trademark law

          Even the Gay Olympics case,⁽⁷⁰⁾ which involved an unusually broad quasi-trademark law that gave the U.S. Olympic Committee the exclusive right to use the word "Olympic" for advertising and promotional purposes, stressed this point: "By prohibiting the use of one word for particular purposes, neither Congress nor the USOC has prohibited the [plaintiff] from conveying its message."⁽⁷¹⁾ The case did not involve any congressional attempt to let the USOC stop people from discussing the Olympics, conveying facts about the Olympics, writing fiction about the Olympics, and so on.⁽⁷²⁾ Even given this limitation, the law considered in the Gay Olympics case has been criticized as going too far,⁽⁷³⁾ and I generally agree with these criticisms. But even if the law improperly gave the USOC too much power, it didn't give it the power to stop the communication of facts.

3. Right of publicity law.

          The Supreme Court has never confronted the broader right to restrict speech that uses one's name or likeness; Zacchini explicitly stressed that it wasn't deciding anything about this right,⁽⁷⁸⁾ and though some courts and commentators have omitted this critical limitation and have cited Zacchini as generally "hold[ing] that the right of publicity is constitutional,"⁽⁷⁹⁾ such a characterization is mistaken. But even to the extent that lower courts have recognized such a right, they too have adopted limiting principles that keep the right from restraining the communication of facts.

          To begin with, though the right of publicity is sometimes described as a right to stop others from using one's name, likeness, and other attributes of identity "in commerce" or "for trade purposes,"⁽⁸⁰⁾ courts and legislatures have long recognized that use of name or likeness "in news reporting, commentary, entertainment, or in works of fiction or nonfiction"⁽⁸¹⁾ must be excluded. These uses are sold in commerce and in trade, but they are nonetheless protected from right of publicity claims, in large part because of free speech concerns.⁽⁸²⁾ The right is not allowed to stop the communication of facts about a celebrity, even if it is allowed to block advertising or merchandising that merely tries to associate the advertiser or the consumer with a celebrity.

          Moreover, even the use of name or likeness in an advertisement that is incidental to the permitted uses--for instance, a billboard advertising an unauthorized biography, which will necessarily use the subject's name and probably likeness--is likewise excluded from the right of publicity, though it's clearly "in commerce" and "for trade purposes."⁽⁸³⁾ This again relates directly to the need to prevent the suppression of facts. Letting Elizabeth Taylor block the unauthorized use of her name in ads for clothing would rarely substantially interfere with the manufacturer's ability to convey the facts about the clothing. Letting her block the use of her name in ads for an unauthorized biography, however, would mean that the biographer couldn't communicate to potential buyers the critical fact that the book is about Taylor.

          The right of publicity may have gotten too big,⁽⁸⁴⁾ but even it basically respects the principle that there ought to be no "abuse of the [intellectual property] owner's monopoly as an instrument to suppress facts";⁽⁸⁵⁾ supporters of property rights in facts thus can't get much analogical support out of it. For whatever it's worth, the few cases that have considered right of publicity claims based on the sale of databases containing personal information have rejected such claims.⁽⁸⁶⁾

4. Misappropriation and trade secret law.

          The first such rule is the right to be free from "unfair" misappropriation of hot news (and possibly of other information). This right was recognized by the Court in 1918 in International News Service v. Associated Press as a matter of pre-Erie federal common law,⁽⁸⁷⁾ but has been mostly rejected since then, most prominently by the Restatement (Third) of Unfair Competition.⁽⁸⁸⁾

          Perhaps because this tort has largely (though not entirely⁽⁸⁹⁾) withered, the Court has never decided whether it passes muster under the First Amendment. Certainly the 1918 decision recognizing the tort didn't confront a First Amendment defense, and in any event First Amendment protections have been dramatically strengthened since then. I believe that if the Court does confront this question, it should conclude that such a right to stop others from communicating hot news is indeed an unconstitutional content-based restriction on fully protected speech.⁽⁹⁰⁾

          The second such quasi-property right is secured by trade secret law. Trade secret protection generally flows from a contract, express or implied, between the trade secret owner and the defendant who is threatening to use or expose the secret;⁽⁹¹⁾ in such a case, Cohen v. Cowles Media strongly suggests that the defendant can be held to the bargain.⁽⁹²⁾ Occasionally, trade secret claims may be based on illegal acquisition (for instance, through a trespass) by the defendant; certainly such acquisition can be punished without First Amendment difficulties.⁽⁹³⁾

          The serious First Amendment problems arise when a trade secret owner seeks to restrict the speech of those who are not in contractual privity with it,⁽⁹⁴⁾ for instance when a company whose employees leaked secret information to a newspaper wants to enjoin the newspaper from publishing the information. The newspaper has never promised anyone not to speak about this, so Cohen v. Cowles Media doesn't apply; the speech restriction can be justified only on the theory that the leaker's initial violation of his confidentiality promise bars otherwise innocent third parties from reporting on the leaked information. The same issue arises in other confidential information contexts, for instance when a newspaper publishes information illegally leaked by a government employee, or illegally taped by someone who then passed along the tape recording.

          The Supreme Court has never decided whether such speech restrictions are constitutional, and lower courts are divided on the subject. I think those courts that come out against such speech restrictions have it right: Speech by people who have never promised to remain quiet about something may not be suppressed simply because someone else wrongfully revealed the information to them. Newspapers must be able to publish leaked information (at least absent some overwhelming national security concerns), even if the leaker breached a contract or even broke the law; a contrary rule would dramatically undermine newspapers' ability to report.⁽⁹⁵⁾ Intercepting confidential communications is properly outlawed, but a newspaper need not stay silent about such communications if they come into the newspaper's hands.⁽⁹⁶⁾ People shouldn't be legally forbidden from telling their friends the truth about someone's medical condition (for instance, that the friend's prospective lover is suffering from a contagious disease) even if the information originally came from a source who had no right to reveal it (such as the prospective lover's doctor).

          Thus, there are no existing First Amendment exceptions that justify restrictions on communication of hot news and restrictions on the publication of illegally leaked facts. One could, of course, argue that the Court should create such new exceptions, but one can't argue that these exceptions already provide support for information privacy speech restrictions. Rather, as I argue below in Part III.D, it is Supreme Court recognition of a property-rights-based First Amendment exception for information privacy speech restrictions that would substantially strengthen the calls for a hot news exception and an illegally leaked facts exception.

5. Summary.

C. Functional Arguments for Upholding Information Privacy Speech
Restrictions Under a Property Theory

1. Avoiding "free-riding" and unjust enrichment.

          This, though, cannot be the justification for restricting speech, unless we are willing to dramatically redefine free speech law. Newspapers and radio and TV news programs, after all, make billions from stories that are made possible only by the existence of their subjects. The essence of news is precisely the reporting of things done or discovered by others; the essence of the news business is profiting from reporting on things done or discovered by others. But news organizations generally don't pay a penny to the subjects of their stories--in fact, it is seen as unethical for news organs, though not entertainment organs, to pay subjects.⁽⁹⁹⁾ Likewise, unauthorized biographers and historians make money from publishing information about others, information that only exists because those people exist. Comedians who tell jokes about people make a living from those they mock.⁽¹⁰⁰⁾

          In a sense, all these speakers are free-riding: They are taking advantage of something that relates to someone else and that exists only because of that other person's existence, and they aren't paying that person for it (though they are usually investing a good deal of time, money, and effort in the project--this free-riding is certainly not mere literal copying). But our legal system correctly allows a great deal of free-riding. It has never been a principle of tort law that all free-riding is illegal, or that all such enrichment is unjust. In the words of the Restatement (Third) of Unfair Competition,

          [T]he principle of unjust enrichment does not demand restitution of every gain derived from the efforts of others. A small shop, for example, may freely benefit from the customers attracted by a nearby department store, a local manufacturer may benefit from increased demand attributable to the promotional efforts of a national manufacturer of similar goods, and a newspaper may benefit from reporting on local athletic teams. Similarly, the law has long recognized the right of a competitor to copy the successful products and methods of others absent an infringement of patent, copyright, or trademark rights.⁽¹⁰¹⁾

          And it has certainly not been a principle of free speech law that speech may be restricted simply to assure the subject of the speech a piece of the profits.

          What intellectual property law has generally tried to prevent is not free-riding as such, but free-riding of a particular kind: the use not just of something that relates to another, but the use of the product of another's substantial labor, and even that only in limited cases.⁽¹⁰²⁾ Such a use runs the risk of dramatically diminishing the incentive to engage in such labor, which is what makes the defendant's enrichment socially harmful rather than merely unjust in some abstract moral sense. This concern is at the heart of copyright law,⁽¹⁰³⁾ of the right to prevent the unauthorized transmission of an entire act,⁽¹⁰⁴⁾ and to a large extent of trade secret law. But this concern does not apply to personal information about people, where the incentive arguments don't really apply.

          Again, I stress that my critique here only relates to the intellectual property justification for information privacy speech restrictions; perhaps there are some other justifications that can support such speech restraints. But the fact that information distributors are profiting while the subjects of the information does not itself provide such support.

2. Internalizing costs and maximizing aggregate utility.

          The principle of free speech law, though, is that speakers do not have to internalize all the felt costs that flow from the communicative impact of their speech. The NAACP didn't have to internalize the tangible economic (not just emotional) cost that its boycott imposed on the Claiborne County merchants.⁽¹⁰⁷⁾ Movie producers don't have to internalize the tangible cost that their movies impose on victims of viewers who commit copycat crimes.⁽¹⁰⁸⁾ Cohen, Johnson, and Hustler didn't have to internalize the emotional distress cost that their speech inflicted on passersby or on its subject.⁽¹⁰⁹⁾

          Again, if there's an independent reason why this speech should be treated differently from other speech, for instance because it falls within some new free speech exception, then the law may require that its costs be internalized. But the desire to maximize aggregate social utility doesn't itself justify a new exception; on the contrary, it's only the new exception that would legitimize speech restraints aimed at maximizing aggregate social utility.

D. The Potential Consequences

          I have explained why I think that merely calling information privacy speech restrictions "property rights" doesn't advance the First Amendment inquiry, why such speech restrictions aren't justifiable under any existing intellectual property exceptions, and why such monopolies in facts, not just expression, are theoretically troubling.⁽¹¹⁰⁾ Of course, despite all this, the Court is always free to carve out a new First Amendment exception or broaden an existing one; my goal now is to explain why I think this would be a bad idea.

          Speech that reveals private information is not the only speech that some want to restrict under the property rights model. As many leading commentators have recently argued, we are now in the midst of a broad movement that uses intellectual property rhetoric to broaden people's rights to restrict others' speech.⁽¹¹¹⁾ The proposed database protection legislation would give database owners a form of property right in collections of information.⁽¹¹²⁾ Some recent cases have revived the misappropriation tort, recognizing a property right in news.⁽¹¹³⁾ Many recent cases have broadened trademark owners' rights to restrict parodies and other transformative uses (though fortunately some courts seem to be resisting this trend).⁽¹¹⁴⁾ Copyright terms are being lengthened and some argue that fair use is being unduly contracted.⁽¹¹⁵⁾ The right of publicity is growing to include any advertising, merchandising, and even interior decor that reminds people of a celebrity, even if it doesn't use the celebrity's name or likeness.⁽¹¹⁶⁾

          Many have criticized this creeping propertization of speech, often on First Amendment grounds.⁽¹¹⁷⁾ They have decried the tendency of many courts to merely label speech restrictions "property rules" as if such a relabeling could eliminate the First Amendment objections.⁽¹¹⁸⁾ They have pointed out that cases upholding the propriety of some speech restrictions--such as the core of copyright law, traditional trademark law aimed at preventing consumer confusion, or the right to control the rebroadcast of one's entire act--don't necessarily validate all new restrictions that one might call "copyright," "trademark," or "right of publicity" (much less "intellectual property" generally).⁽¹¹⁹⁾

          But if the arguments that "it's not a speech restriction, it's an intellectual property rule" or "the Supreme Court has upheld property rights in information, so property rights in information are constitutional" are accepted for information privacy speech restrictions, they will be considerably strengthened as to the other restrictions, too. If, for instance, courts hold that information privacy speech restrictions are proper because they merely "internalize th[e] cost" of their speech "by paying those whose data is used,"⁽¹²⁰⁾ it will be easy to argue the same as to other "data" that someone may say is his. Likewise, if courts hold that such speech restrictions are permissible because the restrictions encourage "a process of voluntary
exchange, [that,] like the free market generally, would assure that 'human satisfaction as measured by aggregate consumer willingness to pay . . . is
maximized,'" the same argument could apply to broad new rights in all sorts of information.⁽¹²¹⁾

          Of course, courts already can, if they really want to, uphold new intellectual property rules by analogy to the existing old ones; but the creation of yet another kind of intellectual property speech restriction--and one that promises to be quite popular--will strengthen the argument. Ask yourself: Would the courts be less likely to accept the notion of property in personal information if trademark and right of publicity had never existed, and the only intellectual property speech restriction were copyright? Probably yes; there are too many distinctions between personal information and copyrightable expression for this one analogy to be that helpful. But as other potential analogies are added, the argument becomes easier--one can say "this proposal is sound because it's like precedent A in one respect, like precedent B in another respect, and like precedent C in a third respect," so even if the proposal is unlike any particular precedent, it can be seen by observers as similar to their aggregate. If this is so, then the case for new intellectual property speech restraints would be further strengthened by the recognition of yet one more kind of such speech restriction to which people can analogize.⁽¹²²⁾

          Moreover, as I've argued, a new exception for a property right in personal information would be the first (but I fear not the last) First Amendment authorization for a property right in pure facts. Right now the database protection proposals are being confronted with the objection that the law does not generally recognize intellectual property rights that restrict communication of facts.⁽¹²³⁾ The analogy to copyright law actually works against those proposals, because they seek to protect exactly what the Court in Feist Publications, Inc. v. Rural Telephone Service Co.⁽¹²⁴⁾ said copyright doesn't protect, and they seek to do exactly what the Court in Harper & Row said would violate the First Amendment--use an "[intellectual property] monopoly as an instrument to suppress facts."⁽¹²⁵⁾ But if information privacy speech restrictions are upheld, they would provide an excellent new analogy for the database protection bill supporters. The same is true for the asserted right to property in hot news, which is today subject to powerful free speech attack,⁽¹²⁶⁾ but which would be strengthened if the courts accept another property right in facts.

          Now perhaps my parade of horribles isn't so horrible; maybe we should have more property rights in facts, which is to say restrictions or speech that communicates facts. Or if I am right to be skeptical of such new property rights, perhaps supporters of property rights in personal information can come up with a narrow justification for those particular rights that will provide little precedential support for the other proposals. Nonetheless, people who are worried about the general trend towards propertization of information should look very carefully at even those proposals that might at first seem benign and even just; such proposals could have effects far beyond the context in which they are first suggested.

A. What "Commercial Speech" Means

          Some argue that sale of information about customers is restrictable because it fits within the "commercial speech" doctrine.⁽¹²⁷⁾ The Court's definition of "commercial speech," though, isn't (and can't be) simply speech that is sold as an article of commerce: Most newspapers, movies, and books are articles of commerce, too, but they remain fully protected.⁽¹²⁸⁾ Likewise, speech can't be commercial just because it relates to commerce, or else the Wall Street Journal, union leaflets and newsletters,⁽¹²⁹⁾ newspaper reviews of commercial products,⁽¹³⁰⁾ and speech by disgruntled consumers criticizing what they consider poor service by producers⁽¹³¹⁾ would be deprived of full constitutional protection.

          Rather, the Court's most common definition of commercial speech is speech that explicitly or implicitly "propose[s] a commercial transaction."⁽¹³²⁾ Commercial advertisements for products or services are classic examples. So are stock prospectuses, which propose the purchase of stock; this is why fairly heavy SEC regulation of speech in such prospectuses is largely permissible, while similar SEC regulation of newsletters or newspapers that discuss stocks is not.⁽¹³³⁾ At the outer boundary, a company's publications that generally discuss a kind of product without mentioning the company by name--for instance, a contraceptive producer's pamphlets discussing contraception generally, rather than just the producer's own devices--also qualify as commercial speech.⁽¹³⁴⁾ Query, though, how far this goes: It's not clear, for instance, that a book touting the health benefits of wine should be treated differently depending on whether its author owns a leading winery.

          The Court has at times suggested that the commercial speech category may also generally cover speech that is "related solely to the economic interests of the speaker and its audience,"⁽¹³⁵⁾ and some lower courts have accepted this definition.⁽¹³⁶⁾ But this can't be right. Consider again the newspaper that discusses business affairs, almost entirely in order to make money by helping its readers do well in business. Consider a product review written by its author because he wants to be paid, published by the newspaper because it wants to keep its paying subscribers, and read by readers because they want to know how to best spend their money. Consider a union buying TV ads urging people to "Buy American" because that's the best way of maintaining the viewers' (and the union members') standard of living.

          Such economic commentary, it seems to me, is as protected as political, religious, social, or artistic commentary. That it has to do with the listeners' economic interests merely highlights its importance--for most people, economic well-being is more important than politics, art, social concerns, or often even religion, and speech on economic matters often has more effect on the nation than does most art or theology, or even much political debate. The speech may not be "political" in the narrow sense of the word, but (as I discuss further in Part IV), the Court has long recognized that strong First Amendment protection extends far beyond politics. Nor does the speech implicate the concerns about fraud in a particular commercial transaction that have been seen as justifying the regulation of commercial advertising. In fact, every one of the Court's dozens of commercial speech cases has involved speech that advertises a product or service;⁽¹³⁷⁾ and the last decade's precedents, which have generally been shifting in the direction of more protection even for speech that is classified as "commercial speech," have stressed the "proposes a commercial transaction" formulation and largely ignored the "solely economic interests" test.⁽¹³⁸⁾

          Under the "speech that proposes a commercial transaction" analysis, communication of information about customers by one business to another is not commercial speech. It doesn't advertise anything, or ask the receiving business to buy anything from the communicating business.⁽¹³⁹⁾ It poses no special risk of the speaker misleading or defrauding the listener, beyond those risks present with fully protected speech generally. The recipient business does intend to use the information to more intelligently engage in commercial transactions, but that's equally true of businesspeople reading Forbes.⁽¹⁴⁰⁾

          Some might argue that there's something inherently un-speech-like in corporations communicating to other corporations, but there's no reason why this would be so. To begin with, the corporate status of the speaker or the listener can't be relevant; surely it can't matter for privacy purposes whether customer information is communicated by and to corporations, partnerships, or sole proprietorships. And the Court has specifically held that speech doesn't lose its constitutional protection because the speaker is a corporation,⁽¹⁴¹⁾ which makes sense for various reasons, among them that almost all media organizations and many nonprofit political advocacy groups are corporations.

          Even if we recast the claim as focusing on businesses communicating to other businesses, the fact is that businesses don't communicate--people communicate. When the managers of Acme Software, at their CEO's urging, read the Wall Street Journal so they can apply what they learn to their business decisions, this isn't "the Wall Street Journal communicating to Acme." It's people at the Journal--the editors, who direct the creation of a joint product by many writers--communicating to people who run Acme. When a scientist working in industry sends the results of his experiments to another scientist also working in industry, the communication may be said to be between their employers (since for both scientists it's part of their jobs), but it's also between people. Likewise, it is no less speech when a credit bureau sends credit information to a business. The owners or managers of a credit bureau are communicating information to decisionmakers, such as loan officers, at the recipient business.⁽¹⁴²⁾

          It's true that in such cases, neither the speaker nor the listener intend to communicate an ideological message through the information, but that's just because the information is fact, not idea. Likewise, in many such cases, neither the speaker nor the listener sees this factual communication as implementing or furthering some ideology, in part because it's just their job. In some cases, though, the people will see the communication as a means of implementing some ideology--"we report the news because the truth is sacred," "we make the wheels of business run more smoothly," "we want to advance the progress of science," "we help protect you from deadbeats because failure to repay a loan is a form of fraud that we want to stop." Many businesspeople genuinely believe that their work is not just a job but part of a broader mission to improve society; it's a peculiar conceit of some professional would-be opinion molders to think that they alone really believe in what they're doing, and that everyone else is only in it for the money. I suspect that the ideological commitment of a typical newspaper reporter who's writing, say, product reviews or local crime stories is not much different from the ideological commitment of a typical businessperson. And this fact helps explain why speech is protected without regard to the speaker's or the listener's ideological motivations.

          Of course, even if speech that communicates personal information is seen as "commercial speech," restrictions on such speech will still have to face considerable scrutiny. Whether they will pass such scrutiny is hard to tell, since commercial speech scrutiny is so notoriously vague.⁽¹⁴³⁾ But this question is actually somewhat tangential to my main point. To me, the main problem with treating speech that communicates personal information as "commercial speech" is not that this will put such speech at more risk of restriction. Rather, it is that stretching the definition of "commercial speech" will put a wide range of other speech at risk, too.

B. The Risks to Other Speech

          Consider a recent example of the government trying to regulate cyberspace speech about economic matters on the grounds that it's "commercial speech." In Taucher v. Born, several operators of commodities-themed Web sites successfully sued to set aside a prior restraint system which bars people from distributing for profit any unlicensed speech that relates "to the value of or the advisability of commodity trading" or that contains "analyses or reports" about commodities.⁽¹⁴⁴⁾ And the license that speakers must get to be allowed to speak isn't just a modest tax; the Commodities Futures Trading Commission can refuse a license if it finds "good cause" to do so, and speaking without a license is illegal. Nor is this speech restriction limited to individualized, person-to-person professional advice: The regulation is broad enough to cover people who "never engage in individual consultations with their customers" and who "under no circumstances make trades for their customers."⁽¹⁴⁵⁾

          The law essentially restricts the Web equivalent of books and newspapers about commodity training--it's as if the government claimed the right to refuse the Wall Street Journal a license to publish articles about the market. As it happens, the law specifically excludes publishers who publish such data "incidental[ly]" as part of a broader news enterprise of "general and regular dissemination,"⁽¹⁴⁶⁾ so the Journal can sleep easy--and the CFTC can sleep easy without the risk of incurring the ire of established, powerful news organs. But under the logic of the law, newspapers and book publishers could also be subject to a prior restraint system, just as the small commodities-focused electronic publishers were subject to it until the court's ruling.

          The CFTC argued that speech about commodities is mere "commercial speech," but the court correctly rejected this:⁽¹⁴⁷⁾ "The plaintiffs' publications in this case do not propose any commercial transaction between the plaintiffs and their customers."⁽¹⁴⁸⁾ If, however, the commercial speech doctrine had been extended to cover the sale of speech about a business's clients, the court's decision might well have been different. After all, the Web business journalist who writes about commodities is likewise selling information that's primarily of economic concern, and that has little to do with broad political debates. If that's enough to deny free speech protection to communications about customers, it may be enough to deny such protection to communications about commodities.

          Consider another example: disgruntled homebuyers putting up signs criticizing the developer that sold them their homes, or consumers leafleting outside a business that they claim sold them defective goods, often hoping that the business will give them a refund or at least will do a better job in the future. In cyberspace, the analogy would be consumers putting up a http://www.[businessname]sucks.com site or circulating messages to a long list of acquaintances or to a Usenet newsgroup.

          In my view, the First Amendment fully protects such speech that is aimed at creating public pressure on someone to do what you think is right, even in economic contexts--that, after all, is what much advocacy is about.⁽¹⁴⁹⁾ The fact that the speech exposes alleged problems with a product and aims at redressing an economic harm should not strip it of protection. Again, for many people problems with their homes and redress for shoddy wares are more important than problems with politicians and redress for shoddy policies, and far more important than art, entertainment, or many other kinds of fully protected speech.

          If the consumer's speech is an intentional lie (or perhaps in some circumstances if it's merely negligently false), the business can sue for libel; false statements of fact, whether on economic matters or not, lack constitutional protection.⁽¹⁵⁰⁾ But the law shouldn't impose extra restrictions on the speech just because the speech deals with economic issues. It shouldn't, for instance, punish true speech on the grounds that it interferes with a business's prospective economic advantage.⁽¹⁵¹⁾ It shouldn't impose prior restraints such as preliminary injunctions on the speech, even if the court tentatively concludes that the speech is probably false.⁽¹⁵²⁾ And even if the speech is found to be in error, the law shouldn't impose liability unless some fault on the speaker's part is shown. Though some such speech restrictions may be permissible as to commercial speech,⁽¹⁵³⁾ they're not permissible as to noncommercial speech; and under current doctrine, consumer criticisms aren't commercial speech because they don't propose a commercial transaction between the speaker and the listener.⁽¹⁵⁴⁾

          Again, though, a broadening of the commercial speech doctrine would jeopardize speech of this sort. If communicating information about a person's bad credit record is mere "commercial speech," then communicating information about a business's bad service record should be, too. Both, after all, involve speech on economic matters. Both involve speech that's primarily of economic interest to listeners. Both are motivated by the speaker's economic interest--either a desire to get money from the buyer of the information, or a desire to get redress from the business. Either both are commercial speech or neither is.

          In a free and competitive economy, people naturally want to talk about economic matters. Often their motives for such speech are largely economic: They want to learn how to make more money. They want to persuade people that some course of action is economically better. They want to alert people to what they think are others' dishonest business practices. Giving the government an ill-defined but potentially very broad power to restrict such speech--not just speech that proposes a commercial transaction between speaker and listener and thus directly implicates the risk of fraud--risks exposing a great deal of speech to government policing.⁽¹⁵⁵⁾

A. The Argument

          One feature of virtually all information privacy proposals (except those built on a contract model) is their distinction between speech on matters of public concern and speech on matters of private concern.⁽¹⁵⁶⁾ Even people who argue that newspapers should be forbidden from publishing a private person's long-ago criminal history or a politician's sexual orientation would probably agree that they have a right to publish the politician's criminal history, no matter how old. Warren and Brandeis would have called this a "matter which is of public or general interest";⁽¹⁵⁷⁾ others call it "political speech" or "speech on matters of public concern" or "newsworthy" material.

          Speech that fits within these labels, they would argue, is constitutionally protected, while speech that is merely of private concern is not protected, at least against information privacy speech restrictions. But this approach, I will argue, is theoretically unsound; it is precedentially largely unsupported; in the few circumstances in which it has been endorsed, it has proven unworkable; and, if adopted, it would strengthen the arguments for many other (in my view improper) speech restrictions.

B. Theoretical Objections

          Under the First Amendment, it's generally not the government's job to decide what subjects speakers and listeners should concern themselves with.⁽¹⁵⁸⁾ A private concern exception essentially says "you have no right to speak about topics that courts think are not of legitimate concern to you and your listeners," a view that's inconsistent with this understanding.⁽¹⁵⁹⁾

          A clear example of the danger of such government power comes in a disclosure tort case, Diaz v. Oakland Tribune.⁽¹⁶⁰⁾ Diaz, the first woman student body president at a community college, was a transsexual, and the Oakland Tribune published this fact. Diaz sued, and the court of appeals held that her lawsuit could go forward; if a jury found that Diaz's transsexuality wasn't newsworthy, she could prevail.⁽¹⁶¹⁾ As usually happens in these cases, the court didn't define newsworthiness but left it to the jury, subject only to the instruction that "[i]n determining whether the subject article is newsworthy you may consider [the] social value of the fact published, the depth of the article, [its] intrusion into ostensibly private affairs, and the extent to which the plaintiff voluntarily acceded to a position of public notoriety."⁽¹⁶²⁾ But the court did stress that a jury could find that the speech wasn't newsworthy: "[W]e find little if any connection between the information disclosed and Diaz's fitness for office. The fact that she is a transsexual does not adversely reflect on her honesty or judgment."⁽¹⁶³⁾

          Now I agree with the court's factual conclusion; people's gender identity strikes me as irrelevant to their fitness for office. But other voters take a different view. Transsexuality, in their opinion, may say various things about politicians (even student body politicians): It may say that they lack attachment to traditional values, that they are morally corrupt, or even just that they have undergone an unnatural procedure and therefore are somehow tainted by it. These views may be wrong and even immoral, but surely it is not for government agents--whether judges or jurors--to dictate the relevant criteria for people's political choices, and to use the coercive force of law to keep others from informing them of things that they may consider relevant to those choices.⁽¹⁶⁴⁾ I may disagree with what you base your vote on, but I must defend your right to base your vote on it, and the right of others to tell you about it.

          This is the clearest example of a court using the public concern test to usurp what should be a listener's and speaker's choice, but other public disclosure cases raise similar problems. Consider, for instance, the criminal history cases, in which some courts held that it was illegal for newspapers to print information about "long past" criminal activity by people who are now supposedly rehabilitated and are leading allegedly blameless lives. The leading such case is Briscoe v. Reader's Digest Association, in which Reader's Digest was held liable for revealing that Briscoe had eleven years earlier been convicted of armed robbery (a robbery that involved his fighting "a gun battle with the local police").⁽¹⁶⁵⁾

          The court acknowledged that the speech, while not related to any particular political controversy, was newsworthy; the public is properly concerned with crime, how it happens, how it's fought, and how it can be avoided.⁽¹⁶⁶⁾ Moreover, revealing the identity of someone "currently charged with the commission of a crime" is itself newsworthy, because "it may legitimately put others on notice that the named individual is suspected of having committed a crime,"⁽¹⁶⁷⁾ thus presumably warning them that they may want to be cautious in their dealings with him.

          But revealing Briscoe's identity eleven years after his crime, the court said, served no "public purpose" and was not "of legitimate public interest"; there was no "reason whatsoever" for it.⁽¹⁶⁸⁾ The plaintiff was "rehabilitated" and had "paid his debt to society."⁽¹⁶⁹⁾ "[W]e, as right-thinking members of society, should permit him to continue in the path of rectitude rather than throw him back into a life of shame or crime" by revealing his past.⁽¹⁷⁰⁾ "Ideally, [Briscoe's] neighbors should recognize his present worth and forget his past life of shame. But men are not so divine as to forgive the past trespasses of others, and plaintiff therefore endeavored to reveal as little as possible of his past life."⁽¹⁷¹⁾ And to assist Briscoe in what the court apparently thought was a worthy effort at concealment, the law may bar people from saying things that would interfere with Briscoe's plans.

          Judges are of course entitled to have their own views about which things "right-thinking members of society" should "recognize" and which they should forget; but it seems to me that under the First Amendment members of society have a constitutional right to think things through in their own ways. And some people do take a view that differs from that of the Briscoe judges: While criminals can change their character, this view asserts, they often don't. Someone who was willing to fight a gun battle with the police eleven years ago may be more willing than the average person to do something bad today, even if he has led a blameless life since then (something that no court can assure us of, since it may be that he has continued acting violently on occasion, but just hasn't yet been caught).

          Under this ideology, it's perfectly proper to keep this possibility in mind in one's dealings with the supposedly "reformed" felon. While the government may want to give him a second chance by releasing him from prison, restoring his right to vote and possess firearms, and even erasing its publicly accessible records related to the conviction, his friends, acquaintances, and business associates are entitled to adopt a different attitude. Most presumably wouldn't treat him as a total pariah, but they might use extra caution in dealing with him, especially when it comes to trusting their business welfare or even their physical safety (or that of their children) to his care.⁽¹⁷²⁾ And, as Richard Epstein has pointed out, they might use extra caution in dealing with him precisely because he has for the last eleven years hidden this history and denied them the chance to judge him for themselves based on the whole truth about his past.⁽¹⁷³⁾ Those who think such concealment is wrong will see it as direct evidence of present bad character (since the concealment was continuing) and not just of past bad character.

          Revealing Briscoe's name, under this view, may have little to do with broad political debates, but it is still of intense and eminently legitimate public concern to one piece of the public: people who know Briscoe, the very same group whose ignorance Briscoe seemed most concerned about preserving.⁽¹⁷⁴⁾ These members of the public would use this information to make the decision, which is probably more important to them than whom they would vote for next November, about whether they could trust Briscoe in their daily dealings.

          This isn't speech on political matters, but rather on what I might call "daily life matters." Under the First Amendment, which protects movies, art, jokes, and reviews of stereo systems,⁽¹⁷⁵⁾ such speech on daily life matters is at least equally worthy. At least as much as those kinds of protected speech, daily life matter speech--communication related to "the real, everyday experience of ordinary people"⁽¹⁷⁶⁾--indirectly but deeply affects the way we view the world, deal with others, evaluate their moral claims on us, and even vote; and its effect is probably greater than that of most of the paintings we see or the editorials we read. Consider how much our view of crime and punishment, secrecy and publicity, and many other topics would be indirectly influenced--towards greater liberalism, conservatism, or something else--by the knowledge that some of our seemingly law-abiding neighbors have been concealing a criminal past.⁽¹⁷⁷⁾

          In any event, which viewpoint about our neighbors' past crimes is "right-thinking" and which is "wrong-thinking" is the subject of a longstanding moral debate. Surely it is not up to the government to conclude that the latter view is so wrong, that Briscoe's conviction was so "[il]legitimate" a subject for consideration, that the government can suppress speech that undermines its highly controversial policy of forgive-and-forget. I can certainly see why all of us might want to suppress "information about [our] remote and forgotten past[s]" in order "to change . . . others' definitions of [ourselves]."⁽¹⁷⁸⁾ But in a free speech regime, others' definitions of me should primarily be molded by their own judgments, rather than by my using legal coercion to keep them in the dark.⁽¹⁷⁹⁾

          The same goes for databases of personal information as much as for news stories about such information. Many such databases--for instance, credit history databases or criminal record databases--are used by people to help them decide whom it is safe to deal with and who is likely to cheat them. Other databases, which contain less incriminating information, such as a person's shopping patterns, may be less necessary for self-protection; but of course for the same reason the data stored in them will also generally be much less embarrassing to their subjects, which makes the supposed harm to the subjects of the communication of such data much smaller. And in any event, even this data is of direct daily life interest to its recipients, since it helps them find out with whom they should do business.

          In some instances, it may be quite unlikely that certain speech would be useful to the listeners either for political purposes or for daily life purposes; this largely has to do with information that shows people in ridiculous, embarrassing, or demeaning contexts without revealing any useful new information about them. Everybody knows that I go to the bathroom; printing a picture of me on the toilet would embarrass me not because it reveals something new about me, but because it shows me in a pose that by cultural convention is seen as ridiculous or undignified.

          This may explain cases such as Daily Times Democrat v. Graham,⁽¹⁸⁰⁾ where a newspaper was held liable for printing a picture of a woman whose dress was accidentally blown up over her waist, and it may partly explain why most people would gladly restrict the nonconsensual publication of photographs of people naked or having sex with their spouses.⁽¹⁸¹⁾ These pictures aren't embarrassing because of the facts they reveal (except in rare cases where they show embarrassing deformities); everyone knows that we're all naked underneath our clothes, and that spouses generally have sex. Rather, they are embarrassing because these poses are conventionally seen as lacking in dignity. Whatever else sex may be, it isn't dignified, and while we may have little concern about our dignity while engaging in the act privately, this lack of concern may stem precisely from the fact that we know other people aren't watching.

          But while there may be a narrow zone of fairly uncontroversially non-public-concern topics, the danger is that the vague, subjective "public concern," "newsworthiness," or "legitimate public interest" test will flow far beyond this zone; and as Briscoe and Diaz, among others, show, this danger has materialized. This risk may be enough to abandon the test altogether, and it is certainly enough to demand that the test be rephrased as something much clearer and narrower before it is accepted.

          We can all think of examples of entertainment that has no connection to public issues, but Winters v. New York was right to conclude that entertainment should be protected despite this, because "[t]he line between the informing and the entertaining is too elusive for the protection of [the] basic right [of free speech]."⁽¹⁸²⁾ If the word "fuck" were forcibly expurgated from public debate, discussion would likely not be substantially impoverished, but Cohen v. California was right to conclude that the word should be protected despite this, because otherwise "no readily ascertainable general principle [would] exist[] for stopping short of" far broader restrictions.⁽¹⁸³⁾ If vitriolic, relatively nonsubstantive parodies such as the one in Hustler v. Falwell were banned, "public discourse would probably suffer little or no harm," but the Court correctly refused to uphold such a ban, since it could find no "principled standard to separate" them from speech that had to be protected.⁽¹⁸⁴⁾ Likewise, the notion that otherwise protected speech should be restrictable when it doesn't relate to matters of public concern strikes me as so potentially broad and so vague that it deserves to be abandoned, even if it would yield the right results in a narrow subset of the cases in which it would be applied.⁽¹⁸⁵⁾

C. Doctrine

          That, then, is why I think the public concern test is theoretically unsound. The doctrinal discussion is easier: Though the Court has often said in dictum that political speech or public-issue speech is on the "highest rung" of constitutional protection,⁽¹⁸⁶⁾ it has never held that there's any general exception for speech on matters of "private concern." Political speech, scientific speech, art, entertainment, consumer product reviews, and speech on matters of private concern are thus all doctrinally entitled to the same level of high constitutional protection, restrictable only through laws that pass strict scrutiny.

          The two situations where the Court has adopted a public concern / private concern distinction are narrow exceptions to this general principle. The first such exception, established in Connick v. Myers, is that the government acting as employer may freely restrict speech on matters of private concern by its employees.⁽¹⁸⁷⁾ The government's power as employer to fire its employees for what they say has always been far greater than its power to fine or imprison private citizens for what they say, and the Connick Court explicitly stressed that private-concern speech remains protected against the government acting as sovereign.⁽¹⁸⁸⁾ The restriction on such speech by government employees was justified only by the special role of the government acting as employer, in which the government's interest in efficient day-to-day operation would make it infeasible to let people sue the government over every discharge that was based on any sort of speech.

          The second exception, established in Dun & Bradstreet v. Greenmoss Builders, is that plaintiffs in libel cases involving false statements on matters of purely private concern may be awarded punitive and presumed damages without a showing of actual malice.⁽¹⁸⁹⁾ This, though, also came in a context where the government has special power to restrain speech: restrictions on false statements of fact.⁽¹⁹⁰⁾ Such statements, the Court has held, have "no constitutional value"⁽¹⁹¹⁾; any protection they get stems from the need to prevent the undue chilling of true statements, which are indeed constitutionally protected.⁽¹⁹²⁾ Dun & Bradstreet thus says little about the propriety of applying the "private concern" test to speech that, unlike false statements of fact, is presumptively constitutionally valuable.⁽¹⁹³⁾

          And Dun & Bradstreet's reasoning confirms that the lower protection given to private-concern speech flows precisely from the speech being false and thus presumptively unprotected. The economic interests of the speaker and its audience, the Court argued, warrant no special protection when "the speech is wholly false."⁽¹⁹⁴⁾ Likewise, the "chilling" effect on constitutionally protected true statements would be minimal because accurate credit reports are "hardy and unlikely to be deterred," are "more objectively verifiable," and are in any case likely to have been heavily verified by successful credit agencies.⁽¹⁹⁵⁾ Neither verifiability nor the market pressure for accuracy is relevant outside the context of restrictions on false statements of fact.

D. The Experience Under the Two "Public Concern" Doctrines

          In practice, neither of these doctrines has been a success story for the public concern test. As many critics have pointed out, the government employee private concern doctrine has proven both vague to the point of indeterminacy and extremely broad.⁽¹⁹⁶⁾ Much speech that would clearly fit within a normal reading of the words "public concern" has been found to be of purely private concern and therefore unprotected, with seemingly little justification other than the desire to make life easier for government employers confronted with troublemaking employees.

          Connick itself found that speech among District Attorney's office employees about "the confidence and trust that [employees] possess in various supervisors, the level of office morale, and the need for a grievance committee" was "not of public concern," hardly a commonsense reading of the term "public concern." And in trying to flesh the test out further, the Court could only say that it was supposed to turn on the "content, form, and context" of the speech, an approach that virtually guarantees that the inquiry will be both unpredictable and little related to the phrase "public concern."⁽¹⁹⁷⁾

          Later cases have likewise found, for instance, that speech criticizing the way a dean runs a public university department,⁽¹⁹⁸⁾ alleging race discrimination by a public employer,⁽¹⁹⁹⁾ and criticizing the way the FBI decides whom to lay off⁽²⁰⁰⁾ was not "of public concern," though other cases reached opposite results on seemingly similar facts.⁽²⁰¹⁾ Whether or not the government should have the power to dismiss employees for such speech, surely the government ought not have the power to censor such speech by citizens at large on the grounds that it's supposedly of insufficient "public concern."

          Under Dun & Bradstreet, the concept of "speech of purely private concern" has ended up similarly vague, and has sometimes covered speech that clearly seems to be of public concern under any normal definition of the term:⁽²⁰²⁾ for instance, speech discussing the competence of psychologists to whom children are sent by government-run schools,⁽²⁰³⁾ the business practices of car dealers,⁽²⁰⁴⁾ and alleged misconduct by the owner of a gymnastics school.⁽²⁰⁵⁾ Again, perhaps it's permissible to allow presumed and punitive damages for false statements on such topics, but surely it would be unconstitutional to restrict true statements on these matters on the grounds that they aren't of "public concern."

          The experience of the public concern test in these two areas thus suggests that the theoretical criticisms of the public concern / private concern distinction are sound: There's a substantial practical risk of the courts finding too much speech to be of "private concern," and while some facially vague and broad tests have the merit of being tied to an existing body of clarifying and narrowing caselaw, that's hardly the case here. Maybe for want of anything better, the public / private concern distinction may remain sensible as to the genuinely hard and necessarily vague government employee speech cases, but its track record hardly seems to encourage expanding it elsewhere.

E. Potential Consequences

1. Direct analogies.

          The classic example is sexually themed speech. A recurring argument in favor of restrictions on such speech, from pornography to art to sexual humor, is that such speech has little to do with self-government, politics, or any of the important, legitimate topics of public debate.⁽²⁰⁶⁾ What, the argument goes, is lost if such speech is restrained, especially if the restraint serves noble goals such as preserving morality, preventing antisocial attitudes, and shielding children against improper influences? Not political debate, not scientific discourse, just people saying and listening to things that they have no really good reason to say and listen to.

          The more courts endorse some speech restrictions on the grounds that the First Amendment doesn't protect speech that's "not of legitimate public interest," the stronger this pro-restriction argument will be in other cases. Right now, the two areas where the courts have accepted a "public concern" test are at least cabinable as involving areas outside the core of First Amendment protection: restrictions imposed by the government acting as employer, where the government has always had a relatively free hand, and restrictions on false statements of fact, which already constitute a First Amendment exception. Analogies between, say, the Communications Decency Act and those restrictions can be rebutted by pointing out that the CDA involves the government acting as sovereign, restricting otherwise constitutionally protected speech.

          Say, though, that courts accept a private concern justification for restrictions on speech that reveals personal information, which are restrictions on otherwise constitutionally protected speech imposed by the government acting as sovereign. Supporters of restrictions on sexually themed speech would then acquire several useful related arguments.

          First, they would be able to argue that there is already a general "no public concern" exception to free speech protection.⁽²⁰⁷⁾ Second, they could point to the information privacy speech restrictions as a specific precedent in favor of similar restrictions on sexually themed speech: Both, after all, involve restrictions on otherwise valuable speech imposed by the government acting as sovereign, and sexually themed speech, they'd argue, is no more important than are politicians' sexual identities or neighbors' criminal pasts. If courts accept the argument that personally identified data is unprotected because (1) it is not communicated "for its expressive content at all," (2) it is only "a tool for processing people, not a vehicle for injecting communication into the 'marketplace of ideas,'"⁽²⁰⁸⁾ and (3) "in the ways that matter, [it isn't] really 'speech' at all,"⁽²⁰⁹⁾ some will quickly argue that sexually themed speech (1) is not communicated for its expressive content at all, (2) is only a tool for sexually arousing people, not a vehicle for injecting communication into the marketplace of ideas, and (3) in the ways that matter, isn't really speech at all.⁽²¹⁰⁾ What's more, information privacy speech restrictions are likely to prove quite popular; what better way to support your argument for restrictions on other "no public concern" speech than by analogizing not just to technical, little-known restrictions but to a widely liked and viscerally appealing one? Third, the precedential value of the government employee speech cases and libel cases would itself be strengthened. Right now these cases can be limited on the grounds that they don't involve the government as sovereign restricting otherwise valuable speech, but once those cases are accepted as an analogy for information privacy speech restrictions, such a limitation will be lost.

          Those who want to protect sexually themed speech will try to distinguish it from speech that reveals private information. The definition of sexually themed speech, they'll argue, is either so vague or so broad that it includes matters that are of clearly legitimate public interest--discussions of sexually transmitted diseases, political statements about sexual matters that rely on graphic sexual imagery for their force, or moral or scientific statements about certain sexual subjects that are best made frankly and not through sanitized euphemism.⁽²¹¹⁾ But the same, of course, is true of speech that communicates others' personal information, which often can be either of public interest or of daily life interest. If this argument is rejected for private information speech, it will also be easier to reject for sexually themed speech.

          Likewise, opponents of restrictions on sexually themed speech will argue that the government has no business deciding which topics are "legitimate" and which aren't--that the First Amendment leaves this decision to speakers and listeners, not government officials. But again, if this argument is rejected for speech that reveals private information, and the government does get to decide that people really have no business talking about certain topics, the argument will also be much easier to reject for sexually themed speech.

          Any new "no public concern" exception will help support other restrictions, too. Restrictions on profanity and on flag burning have been urged on the grounds that the speech is not really necessary for the communication of important ideas;⁽²¹²⁾ campus speech codes have often been defended on the same grounds.⁽²¹³⁾ Though people have the right to express offensive or bigoted ideas, the argument goes, profanity, flag burning, and slurs don't really add anything much to such expression; the idea can still be expressed just as well without this valueless component. Bans on such speech, the argument might go, "would not damage the communication of a message," just as some argue that information privacy speech restrictions are constitutional because "[r]estraints on the circulation of personal information would not damage the communication of a message."⁽²¹⁴⁾ If courts accept the notion that publishing people's names in news stories can be restricted because the "need of the people to be informed of matters of general or public interest" could be "served as well without identifying" "the people concerned"⁽²¹⁵⁾ they will also be likely to uphold other government attempts to excise offensive and supposedly valueless components of other speech.⁽²¹⁶⁾

          Similarly, businesses criticized by disgruntled consumers have already argued that such consumer criticism doesn't relate to speech on matters of genuinely "public concern," and should therefore be restrictable even if it's true or if it's mere opinion.⁽²¹⁷⁾ Allowing tort liability under the disclosure tort for speech on supposedly "private matters" (such as a person's criminal history or failure to pay his debts⁽²¹⁸⁾) would provide strong support for allowing tort liability under the intentional interference tort for speech on "private matters" (such as a business's unfair practices or breaches of warranty).

2. Indirect influence.

          First, "privacy" is a word with many meanings, and with such words both judges and laypeople often shift from one meaning to the other even in cases where the meanings have little in common. Consider how often privacy arguments commingle the Griswold/Roe constitutional right of decisional privacy, the Fourth Amendment right to privacy from physical government intrusion, and the four distinct privacy torts, even though these doctrines are at best distant cousins.⁽²¹⁹⁾ Or consider how often Zacchini v. Scripps-Howard Broadcasting Co. is cited for the proposition that a broad right of publicity is constitutional,⁽²²⁰⁾ even though the case itself upheld only a narrow and unusual subset of the right of publicity--the right to block the rebroadcast of an entire act--on grounds that are specific to this narrow right and with the specific statement that it wasn't deciding the constitutionality of the broader right of publicity.⁽²²¹⁾ Our legal system (and perhaps human nature) operates by analogy, and analogies that rely on multiple meanings of the same word are unusually powerful.

          Because of this, once restrictions on people's speech are accepted in the name of "privacy," people will likely use them to argue for other restrictions on "privacy" grounds, even when the matter involves a very different sort of "privacy." For instance, many people have already urged restrictions on sexually themed speech on the grounds that it invades people's "privacy" by being accessible in their homes (and thus in a way intruding on their seclusion), by being accessible to their children (and thus interfering with their "privacy" right to familial autonomy), or by lowering the moral tone of society in a way that affects people's most private relationships.⁽²²²⁾

          Second, a strong free speech principle necessarily requires the protection of speech that many sincerely believe is evil and dangerous. One way of mustering support for this principle, both among courts and among the public, is to stress that all sorts of groups are in this boat: If people are upset that the speech they hate is protected, they should take comfort in the fact that speech that they may like and that other people hate is protected, too.⁽²²³⁾

          The converse of this, though, is that people's willingness to accept protection of the speech they hate decreases as they see new exceptions carved out for restrictions on other speech which they may see as much less harmful. We see this reaction already: Why should the harm that racist advocacy imposes on its victims remain unremedied, some supporters of campus speech codes ask, when harms to copyright owners, to libel victims, and the like have been found to justify punishment?⁽²²⁴⁾ One article even makes the same argument in favor of information privacy speech restrictions themselves: "If the powerful may exert property rights or invoke contractual obligations to prevent or limit speech" (alluding to the existing free speech exceptions for contract law, trademark law, and contract law), "so too may others" asserting informational privacy rights.⁽²²⁵⁾

          But the longer the list of permissible restrictions, the stronger these arguments for further restrictions will be. Imagine that the Court upholds information privacy speech restraints. Why should the harm to my child and my family stemming from the child's exposure to online indecency remain unprevented, some may then argue, when the indignity that someone feels from having his shopping habits communicated by one business to another justifies a speech restriction? Both, after all, involve nonpolitical speech. Neither involves t