No CrossRef data available.
AUTONOMY AND INFORMATIONAL PRIVACY, OR GOSSIP: THE CENTRAL MEANING OF THE FIRST AMENDMENT
Published online by Cambridge University Press: 04 June 2004
Extract
My thesis is simple. The right of informational privacy, the great modern achievement often attributed to the classic Samuel Warren and Louis Brandeis article, “The Right to Privacy” (1890), asserts an individual's right not to have private personal information circulated. Warren and Brandeis claimed that individual dignity in a modern society requires that people be able to keep their private lives to themselves and proposed that the common law should be understood to protect this dignity by making dissemination of private information a tort. As broadly stated, this right not to have private information distributed directly conflicts with a broadly conceived freedom of speech and of the press. My claim is that, in cases of conflict, the law should reject the Warren and Brandeis innovation. Speech and press freedom should prevail; the privacy tort should be ignored. This conclusion requires a normative argument concerning the appropriate basis and status of speech freedom that this essay will not really provide but for which I have argued elsewhere. Here, instead, I will describe that theory of speech freedom, explore its implications for informational privacy, and finally suggest some reasons to think that rejection of the privacy tort should not be so troubling and is, in fact, pragmatically desirable.
Information
- Type
- Research Article
- Information
- Copyright
- © 2004 Social Philosophy and Policy Foundation
“Gossip is no longer the resource of the idle and of the vicious, but has become a trade… .”1
Samuel Warren and Louis Brandeis, “The Right to Privacy,” Harvard Law Review 4 (1890): 193–220.
Book dedication: “To Vito Russo, in gratitude for much good conversation and great gossip… .”2
Larry Gross, Contested Closets (Minneapolis: University of Minnesota Press, 1993), dedication page.
“I find that when I am gossiping about my friends as well as my enemies I am deeply conscious of performing a social duty; but that when I hear they gossip viciously about me, I am rightfully filled with righteous indignation.”3
Max Gluckman, “Gossip and Scandal,” Current Anthropology 4, no. 3 (1963): 307–16, at 315.
I. Introduction
My thesis is simple. The right of informational privacy, the great modern achievement often attributed to the classic Samuel Warren and Louis Brandeis article, “The Right to Privacy” (1890),4
Warren and Brandeis, supra note 1. This article has been described as “perhaps the most famous and certainly the most influential law review article ever written.” Melville B. Nimmer, “The Right of Publicity,” Law and Contemporary Problems 19 (1954): 202, 203.
C. Edwin Baker, Human Liberty and Freedom of Speech (Oxford: Oxford University Press, 1989); C. Edwin Baker, “Harm, Liberty, and Free Speech,” Southern California Law Review 70 (1997): 979.
The essay proceeds in the following way. Section II describes different possible informational privacy rights, identifying the one at stake in this essay. Section III describes two conceptions of autonomy, showing that both speech freedom and informational privacy serve the ultimately most important substantive conception, which is characterized as “meaningful” autonomy. Section III argues, however, that the right of free speech is better seen as based on the other, “formal” conception of autonomy, and that this grounding implies overriding the legal protection of informational privacy recommended by Warren and Brandeis. Thus, Section IV describes various ways that the law could treat informational privacy, but recommends treatment consistent with the formal right of free speech. Section V argues that protection of speech freedom leaves many ways of serving informational privacy fully available. Speech freedom turns out to be fully compatible with prohibiting possibly the most common and important ways in which informational privacy is invaded. This discussion should relieve some of the anticipated resistance to my thesis. Finally, Sections VI and VII offer two pragmatic reasons to find the thesis acceptable. Section VI argues that the popular appeal of privacy may be to some significant degree misguided. Section VII argues that speech that invades privacy—speech that I illustrate with gossip but that also includes other forms of individual expression as well as media invasions of privacy—serves valuable functions that help make the constitutional status of free speech appealing and explicable. Thus, Sections V, VI, and VII together should make more plausible both the thesis of this essay and the formal conception of autonomy on which I claim the right of free speech is based.
II. Informational Privacy
Informational privacy involves (some) limitation on inspection, observation, and knowledge by others. The appropriate legal response to claims for informational privacy depends, of course, on the more specific content of the claims. Among other possibilities, informational privacy could refer to one, or some combination, of the following: (1) inalienable private information, that is, categories of information that are not permissibly exposed to or held by anyone other than the original holder or, in the case of joint holders, possibly resulting from joint participation in a private activity, by any nonintimate; (2) control over initial disclosure of information; as an almost necessary corollary, this right implies the availability of a range of meaningful contexts in which private information can be created, discovered, or used in ways that do not necessarily result in any disclosure; or (3) control over further uses and dissemination of (private) information after an initial disclosure, that is, a virtual property right in private information quite analogous to various intellectual property rights.6
Julie Cohen, “Examined Lives: Informational Privacy and the Subject As Object,” Stanford Law Review 52 (2000): 1373–1438
Cohen v. Cowles Media Co., 501 U.S. 663 (1991) (promissory estoppel can apply to newspaper that promised confidentiality).
Seattle Times Co. v. Rhinehart, 467 U.S. 20 (1984) (newspaper can be ordered not to publish information obtained through discovery).
Bartnicki v. Vopper, 532 U.S. 514 (2001).
The first of the three conceptions of informational privacy that I enumerated above—information inalienability—seems overtly contrary to the individual agency of the person required to keep information private. For example, the U.S. military's “Don't ask, don't tell” policy prevents a gay male or lesbian from disclosing his or her homosexuality, thereby coercively creating a degree of information inalienability. Still, information inalienability might be thought in some contexts to serve (a certain conception of) personhood, a decent society, or other values. If so, this service could provide a rationale for this interpretation of informational privacy. Privacy would be, to some limited extent, inalienable. This approach to privacy comes close to the premise behind child pornography laws, although that premise may be crucially affected by the added paternalistic assumption that children are not capable of giving appropriate consent to disclosure and that any parent or guardian who gives consent is not acting properly in the child's interest.10
The Court relied upon concerns about a child's participation in the making of child porn and with the availability of the permanent record of the child's participation to justify the law in New York v. Ferber, 458 U.S. 747 (1982), harms that distinguish this case from Ashcroft v. Free Speech Coalition, 535 U.S. 234 (2002) (striking down prohibition of computer-created child porn in which no actual child was used).
Barnes v. Glenn Theatre, Inc., 501 U.S. 560 (1991); City of Erie v. Pap's A.M, 529 U.S. 277 (2000) (upholding bans on public nudity as applied to dancers in dance halls or adult establishments). Both cases generated strong dissents. The dissenters distinguished public nudity before unconsenting adults, who might be considered to be viscerally assaulted by the nudity, and before exclusively consenting parties such as in a theater, where the state interest seems directed specifically at stopping expressive communication. (Elsewhere, I have suggested that in an advocacy as opposed to an entertainment context, nudity should sometimes be protected even in relation to those who are offended: Baker, Human Liberty, supra note 5, at 135, 173–78, 306 n.27, 318 n.29; and C. Edwin Baker, “The Evening Hours During Pacifica Standard Time,” Villanova Sports and Entertainment Law Journal 3 [1996]: 45.) The plurality in Pap's A.M. attempted to meet the objection that the law aimed at suppressing communication by arguing that, instead, the state interest was related to preventing “secondary effects,” that is, effects not dependent on whether anyone received the message and not involving any condemnation of the communicative exposure. The purported secondary effect would occur if people who may not even have received the communication come to the area and engage in activities that the state properly restricts, such as prostitution. Previously, purported secondary effects have only justified “zoning” the expression in a manner hoped to reduce these bad effects, which makes sense of the fact that secondary effects cases are analyzed much like “time, place, or manner” cases. No case prior to Pap's A.M. used secondary effects analysis to entirely bar the expression, a point emphasized by the dissent. Arguably, the plurality in Pap's A.M. only makes doctrinal sense if the complete ban on intentionally appearing in public in a “state of nudity” was not a complete ban on the particular expression. The plurality argues this is so. The dancer could make the same communication, the plurality implied, because she could be almost nude, a view ridiculed by the dissent, which claimed audiences distinguish dancing nude from dancing with “pasties and G-strings.” The plurality seems overtly inconsistent with Cohen v. California, 403 U.S. 15 (1971), which emphasized not only that “the Constitution leaves matters of taste and style … largely to the individual” but also that the First Amendment protects the important emotive function of using a particular word, “fuck.” Cohen also noted that any bar on the word's use runs the danger of suppressing ideas.
If, as claimed here, a person has a right to disseminate private information about another person, surely she should have the same right to disseminate information about herself. An obviously essential individual power is the capacity to reveal or expose oneself, be it to one's lover or to the world, at least to the extent that the person can find effective means to disseminate her verbal or pictorial self-portrait. If taken to its extreme, a regime of inalienable privacy would mandate invisibility, a social erasure practiced in some totalitarian countries and possibly the experience of many subordinated peoples.12
Cf. Ralph Ellison, The Invisible Man (New York: Random House, 1952).
But cf. Anita L. Allen, “Coercing Privacy,” William and Mary Law Review 40 (1999): 723–57, 740, 752–53.
Largely because of its obvious potential to conflict directly with speech freedom, but also because it is often seen as the most important innovation of the Warren and Brandeis article and possibly the major development of the common law in the twentieth century, the third sense of informational privacy will be the focus of this essay; that is, this essay will focus on control over further dissemination once someone else holds one's personal information. More specifically, my question is whether an adequate and appropriate legal response to informational privacy claims should be limited mostly to maintaining or strengthening privacy in the second sense—control over initial disclosure and the maintenance of private spaces—and, in any event, should not extend to a more general, property-like claim to control others' dissemination of private information about oneself. My thesis is that a person's control over other people's dissemination of private information about the person, purportedly recognized under the rubric of protecting privacy, should be rejected in any case where it would restrict freedom of speech or of the press as properly understood. But the complete story will be more complex, including caveats on when limitations on further dissemination are not contrary to First Amendment requirements.
To jump ahead, the direction of the argument for these caveats will be twofold. First, as opposed to the strong sense of information inalienability rejected above—for example, a person cannot appear nude, even in an enclosed public space—there is also a more limited notion of market inalienability.14
Margaret Jane Radin, Contested Commodities (Cambridge, MA: Harvard University Press, 1996), 18–20.
III. Two Conceptions of Autonomy
A. Meaningful autonomy
‘Effective’ or ‘meaningful’ individual agency or autonomy might be described, loosely, as a capacity (including the necessary opportunities) to lead a meaningfully self-authored life without unnecessary or inappropriate frustration by others. The precise formulation or definition of the term ‘?meaningful autonomy’ is not a major concern here. Rather, my point is that any such notion is not an on/off variable but a matter of degree. A key function of social policy and of the legal order ought to be to create the conditions that enhance meaningful autonomy. Meaningful autonomy is also concrete; the concern is not with having maximum abstract freedom of choice but with having opportunities to make those choices one actually wants to make. Enhancing effective autonomy for one person will often impair it for another. Conflicts are inevitable. A legal order that provides one person with great wealth, for example, will normally advance her effective autonomy, though this often means someone else will have less wealth, with the consequence that the latter's effective autonomy will be restricted. Recognizing that one of two claimants has a right to a plot of land (or a bank account) will typically advance the winner's capacity to lead a self-authored life while having the opposite effect on the losing claimant.
Many, many things—education, John Rawls's primary goods, sensible environmental policies—including informational privacy, can serve effective individual agency. Elimination of “informational” privacy was a major evil imagined in George Orwell's 1984, a fictional horror not unlike what technology is rapidly making possible today. However, informational privacy is only one of many resources that can serve meaningful autonomy. To pick the key category of this essay, consider, for example, speech and information. Some have invoked as a central rationale for freedom of speech the premise that (effective) individual autonomy requires information and access to varying viewpoints.15
Thomas Scanlon, “A Theory of Freedom of Expression,” Philosophy & Public Affairs 1, no. 2 (1972): 204–26; David A. Strauss, “Persuasion, Autonomy, and Freedom of Expression,” Columbia Law Review 91 (1991): 334–71; and Richard H. Fallon, Jr., “Two Senses of Autonomy,” Stanford Law Review 46 (1994): 875–905.
Anita L. Allen, supra note 13; Julie Cohen, supra note 6.
The claim that meaningful autonomy requires privacy often involves assertions that for development, experimentation, and repose, individuals need the capacity to shield themselves, at various times and places and to varying degrees, from exposure to the critical eyes of the world. A common assertion is that the public sphere depends on and is, in this sense, parasitic on the private sphere (and, many commentators go on to add, vice versa).17
This is similar to Habermas's claim that private autonomy and public autonomy are co-original. Jürgen Habermas, Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy, trans. William Regh (Cambridge, MA: MIT Press, 1996), 104, 121–22, 263, 314, and 454. Individual rights are created by the exercise of public autonomy (the liberty of the ancients), but the exercise of public autonomy requires autonomous individual rights holders. See infra note 145.
Long a subject of science fiction, this need for privacy in order to develop as a full person was the theme of two recent popular Hollywood movies, The Truman Show (1998) and Ed TV (1999).
First Amendment doctrine recognizes this need for privacy, emphasizing the role that anonymity played in revolutionary pamphleteering and, more generally, its role in creating a willingness to engage in political activities. See NAACP v. Alabama ex rel Patterson, 357 U.S. 449 (1958) (First Amendment protects against state demand for disclosure of membership lists); Talley v. California, 362 U.S. 60 (1960) (ban on anonymous leafletting is unconstitutionally overbroad); Brown v. Socialist Workers '74 Campaign Committee, 459 U.S. 87 (1982) (Socialist Workers Party has right not to report the identity of campaign contributors); McIntyre v. Ohio Elections Commission, 514 U.S. 334 (1995) (person has right to distribute anonymous campaign literature). See generally Seth F. Kreimer, “Sunlight, Secrets, and Scarlet Letters: The Tension between Privacy and Disclosure in Constitutional Law,” University of Pennsylvania Law Review 140 (1991): 1–147. Fifth Amendment limits on compelled disclosure might also be seen, in part, as protecting these interests.
See cases cited in note 19.
Even this second sense of privacy, disclosure control, would be potentially problematic if there is a constitutional, presumably a First Amendment, right of access to information—a right said to serve individuals' autonomy interests.21
See Scanlon, Strauss, and Fallon articles, supra note 15.
Justice Potter Stewart, “Or of the Press,” Hastings Law Journal 26 (1975): 631–37; Baker, Human Liberty, supra note 5, chap. 10.
See, e.g., Buckley v. Valeo, 424 U.S. 1 (1976) (upholding disclosure requirements).
Even if meaningful autonomy requires both access to some sorts of information about others and some capacity to limit disclosure, surely autonomy or agency is not robbed of value simply because a person does not possess all possible information or complete control over access to information about herself. Maximizing both access to information and the capacity to limit disclosure would be contradictory. Adding to one person's privacy decreases another person's information. The same informational privacy that contributes to one person's meaningful autonomy can undermine that of another. It is equally undesirable to try to maximize either access to information or privacy. Not only does a person not need either in maximal amounts, but meaningful autonomy probably does not require that a person always have any specifiable type of information about others available or about herself hidden.
Still, both cloaking and making available particular categories of information may be valuable. Whether meaningful autonomy will be increased more by one or the other will characteristically vary not only among cultures and between various subgroups of a society, but also among different people within any subgroup or for the same person over time. These variations reflect the fact that both information availability and informational privacy relate to autonomy more as instrumental supports than as defining elements. Neither information nor privacy is, itself, a self-authored activity (though the activity of disclosure or hiding can be). Rather, both information and privacy are resources that make people more capable of various meaningful or valuable self-authored choices. Typically a person has more power of self-authorship when she has more of both the privacy and the information that she values. In contrast, for this person, gaining more of the wrong sort of privacy or information is at best a diversion, and sometimes can be a burden that reduces her capacity for meaningful self-authorship. Given that obtaining more of privacy or information imposes costs of various sorts, both on the person gaining the resource and on others, and given that obtaining either resource will vary in significance, the socially ideal amount and type of informational privacy or information availability will be controversial and contested.
B. Formal autonomy
The claim in the previous section is that both information (including privacy-invading information) and informational privacy instrumentally support meaningful autonomy. However, a potentially controversial distinction can be made between ‘abstract autonomy’, or what I will often call ‘formal autonomy’, and ‘meaningful autonomy’, a distinction roughly equivalent to what Rawls called “liberty” and the value or “worth of liberty.”24
John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), 204.
Although I will not defend them here,25
See Baker, Human Liberty, supra note 5; Baker, “Harm,” supra note 5.
In free speech literature, my distinction between meaningful and formal autonomy parallels Richard Fallon's distinction between “descriptive” and “ascriptive” autonomy. See Fallon, supra note 15. According to Fallon, in the past, First Amendment autonomy theorists have mostly ignored ascriptive autonomy in favor of descriptive autonomy, but that free speech doctrine ought to respond to both. Fallon observes that the two conceptions place sometimes conflicting demands on the legal order—and descriptive autonomy can lie on both sides of an issue. He then concludes that the proper approach balances all the autonomy claims (although when he does the balancing, it seems he usually favors the strongest claims made on behalf of descriptive autonomy). In my view, many scholars (including myself) whom Fallon characterizes as advancing negative liberty, descriptive autonomy claims were in fact offering ascriptive autonomy theories. See C. Edwin Baker, “Realizing Self-Realization: Corporate Political Expenditures and Redish's ‘The Value of Free Speech’”, University of Pennsylvania Law Review 130 (1982): 646–77 (criticizing Martin H. Redish for adopting a descriptive conception of autonomy). Although Fallon is right that both types of autonomy are important for law and policy, I provide a different account of their proper relationship. Namely, I propose that, rather than balance what are in effect apples and oranges, the law ought consistently to respect ascriptive (or formal) autonomy as a trump or side constraint in developing ways to promote descriptive (or meaningful) autonomy. I suggest that this approach is both more consistent with First Amendment law and more normatively defensible, although both points are disputed.
Different theoretical accounts could be given for the status of formal autonomy. The account that I have found persuasive sees the legitimacy of the legal order as dependent on respecting people as equal and autonomous agents.27
C. Edwin Baker, “Foundations of the Possibility of Legitimate Law” (unpublished essay, 2003).
In many respects, this view of two conceptions of autonomy repeats approaches that are relatively common in legal and political theory. The theory of criminal punishment that views legitimate punishment as limited by a principle of proportionality, but as extending no further than is beneficial for the collective welfare, embodies a formal conception of autonomy as a restraint and a more utilitarian element that controls within the bounds allowed by the constraint.28
See H. L. A. Hart, Punishment and Responsibility: Essays in the Philosophy of Law (New York: Oxford University Press, 1968), 1–27; and Kent Greenawalt, “Punishment,” in Joshua Dressler, Encyclopedia of Crime and Justice, 2d ed. (New York: Macmillan, 2002), 1282–94, at 1289.
C. Edwin Baker, “Sandel on Rawls,” University of Pennsylvania Law Review 133 (1985): 895–928.
See, e.g., Rainer Forst, Contexts of Justice: Political Philosophy beyond Liberalism and Communitarianism, trans. John M. M. Farrell (Berkeley: University of California Press, 2002).
The law respects formal autonomy to the extent that it meets two conditions. First, it must allocate ultimate control over a person's mind, body, and property to that person, except when that person would use her body or property to interfere with another's legitimate realm of decision-making control. Second, the law must not aim at eliminating or suppressing people's freedom to make decisions about behavior or values. These requirements have clear implications for speech, namely, that a person should be able to decide for herself what to say. These requirements, however, imply nothing about whether she will have the capability to say something, which would require, for example, knowledge on her part and which is more a matter of meaningful autonomy.
Obviously, legal recognition of formal autonomy (or liberty) is not necessarily absolute. A person could have autonomy in respect to some choices, could be free to say some things, and not others. Thus, as with meaningful autonomy, a person could have more or less formal autonomy. Nevertheless, formal autonomy is unlike meaningful autonomy, which conceptually could not be provided in any complete sense, and it is unlike either information availability or privacy, which may justifiably not be maximized (not only for cost reasons but also because of their diverse relation to meaningful autonomy). Decisions to impose direct legal limits on formal autonomy are not a necessary part of a legal order and should be especially troublesome. Such restrictions appear aimed at serving collective purposes by means of unnecessarily disrespecting or forbidding self-authorship rather than, for example, by necessarily distributing inherently limited resources. Thus, laws aimed at restricting choice are unlike laws selecting among different distributions, which inevitably affect different people's meaningful autonomy differently but do not themselves restrict formal autonomy, that is, individuals' choices about themselves and their property.
Legal rules that limit liberty (or formal autonomy, as described here) are not a logically necessary part of a legal order, while legal rules that limit the value or “meaningfulness” of a person's liberty—and that limit it differently for different people—are inherent to any legal order.31
The points merely asserted here and in the next two paragraphs are developed in Baker, “Harm,” supra note 5.
In contrast to these allocative or distributive laws are a second sort, which I label ‘general prohibitions’. These laws prohibit certain choices on the part of everyone: for example, no one can engage in sodomy, say bad things about the president or the country, eat pork, drink bourbon, or charge a higher price. These general prohibitions are directed at making some choices unavailable, which usually means they are directed at limiting formal autonomy. Moreover, unlike allocative rules, general prohibitions are not an inherent or logically required aspect of a legal order. Thus, a commitment to respect formal autonomy suggests the following possible general principle: do not allow general prohibitions even when a general prohibition predictably increases the worth of liberty to many people more than it decreases the worth of liberty to others. I have defended such a principle in those contexts where formal liberty seemed to be actually at stake, that is, where the general prohibition was properly seen as blocking valued exercises of autonomy as opposed to merely making certain choices instrumentally more difficult, which distributional rules regularly do. Thus, my reformulated general principle is: a state acts improperly when its aim is to suppress individual choice as a means of carrying out even the state's good aims. This principle of liberty must be contrasted with the propriety of (sometimes) protecting a person's chosen actions from interference by another, which recalls the distributive or allocative issue.
One additional characteristic of this description of respect for formal autonomy merits attention. This view centrally identifies the person with agency, with action, and with the possibility of choice. In a sense, this is an activity view of personhood: it favors a person's activity of speech over the status of being unknown. This perspective accords, I think, with the view of Justice William J. Brennan, who, after asserting that “freedom of speech is itself an end,” went on to say that “freedom of speech is … intrinsic to individual dignity,” and who characterized “a democracy like [the United States]” as one “in which the autonomy of each individual is accorded equal and incommensurate respect.”32
Herbert v. Lando, 441 U.S. 153, 184 n.1 (1979) (Brennan, J., dissenting in part).
Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555, 585 (1980) (Brennan, J., concurring in the judgment).
Warren and Brandeis, supra note 1, at 205. Lillian BeVier suggests that privacy is the equivalent of speech as a constitutive element of liberty. I find this to be a strange conception of liberty. It implies that liberty is not merely a power of choice about one's own action. Rather, it identifies liberty either with a state of affairs (i.e., the information is not known) or with power over other people's acts (i.e., they cannot speak what they know). This notion of liberty, however, may explain why, in her contribution to this volume, BeVier does not find copyright's restriction on people's speech choices in conflict with the First Amendment, a view that I find equally strange.
Though in this essay my concern is primarily to explore the significance for informational privacy of a commitment to speech freedom, some comment on these (at least partially) conflicting views about what should be legally “inviolate” is warranted. Essentially, favoring “choice” over “personality” privileges a view of the fundamental aspect of personhood as activity rather than something static. To assert as basic a person's right to have a characterization of her personality unchallenged by others' expression is an assertion of power over others—in practice over their speech choices but in ambition over even their mental views. Though recognizing a person's legitimate interest in others' choices, the claim that a person should have this type of power over others seems very problematic. It is difficult to see why the legitimacy of a legal order would depend on its recognition of power over others. Certainly, it is not a power that a person would have in the absence of a legal grant. Thus, my premise will remain that formal autonomy of the sort described, which includes a person's choices about her own speech, is plausibly something that the legal order should treat as inviolate. In contrast, any notion of personality that includes control over others' speech is not a plausible candidate for treatment as inviolate. However, the latter view of personality still implicates often supremely important interests that merit various forms of incomplete legal as well as informal, customary support.
Whether or not formal autonomy as a conception of autonomy is appealing in general, if we narrow our gaze just to speech freedom it is clear that this kind of autonomy has the virtue that it could be, whether or not it should be, fully guaranteed. A person, whenever she is in a place where she has an unrestricted right to be, could be free to say whatever she wants. Of course, allocation rules will affect where (or when) a person can say particular things. Still, there is no category of “content” that a person necessarily must be prohibited from saying in order to authorize another person to make choices for herself, for example, whether or not to say the same thing, to say something else, or to say nothing. Censorship—i.e., prohibitions on people's choices to communicate particular content—amounts to general prohibitions. (I put aside a preliminary theoretical inquiry, presumably informed by why speech should be protected, of identifying particular behavior as “speech” in a normatively or constitutionally relevant sense. Critics of free speech absolutism almost always mischaracterize the absolutists' position by assuming that absolutists believe that the First Amendment protects every verbal or vocal act, a view clearly repudiated by every prominent First Amendment absolutist. For example, no prominent First Amendment absolutist of whom I am aware ever defended First Amendment protection for perjury or commercial speech. Rather, absolutists claim that properly protected speech choices should not be suppressed on the basis of an instrumental balancing analysis of the sort that enamors the current Supreme Court majority.)
Of course, a formal autonomy speech right may not lead to much meaningful autonomy if a speaker never has access to the resources (e.g., educational, informational, conceptual, experiential) that are needed to have anything meaningful to say or the material resources that are necessary to communicate with her intended audience. Still, many civil libertarians (including me) incline toward absolutely protecting this type of formal autonomy against abridgement. Opponents of this position typically argue that such a principle should at best be a rule of thumb or a rebuttable presumption. That is, they argue that the claim of formal autonomy (or liberty) should be rejected when its recognition detracts seriously enough, as it sometimes does, from other people's meaningful autonomy, or when it seriously interferes with a more egalitarian provision for meaningful autonomy. In the context of privacy concerns, the opponents of absolutism might claim that speech freedom can detract from meaningful autonomy by allowing one person to expose personal information about another, and that this intrusion may justify limiting speech freedom.
In contrast to formal autonomy-based speech rights, meaningful autonomy-based, speech-related “interests” cannot be fully or absolutely protected or served. This conclusion should be obvious from our discussion, since provision of the same information that invades one person's privacy (and undermines her meaningful autonomy) can affirmatively serve someone else's meaningful autonomy. Both having information and having privacy are autonomy-based interests, but full provision of one necessarily limits the other: privacy limits information, sometimes access to information limits privacy. The conflict can be resolved only by a decision, presumably a policy decision, that tries best to balance or accommodate the competing interests.
As noted, my premise here is that the central justification for the constitutional status of freedom of speech relates to a need for the law to respect individual autonomy in the formal sense of protecting a person's choice of what to say (or her choice to listen to a willing speaker).35
See Baker, Human Liberty, supra note 5.
For a differing view, see articles listed in note 15. Also Martin H. Redish, “The Value of Free Speech,” University of Pennsylvania Law Review 130 (1982): 591–645.
Cf. Carlin Meyer, “Sex, Sin, and Women's Liberation: Against Porn-Suppression,” Texas Law Review 72 (1994): 1097–1201.
IV. Privacy and Autonomy in Law
I have argued that informational privacy can contribute to meaningful autonomy, and so can information that exposes private matters. Some degree of privacy may be an essential aspect of human dignity, although there is no reason to believe that privacy about any particular fact is essential. Cultures and individuals will vary in respect to the information that they find most important to keep private. Policy analysis should evaluate the significance of both informational privacy and information exposure in particular contexts. Policymakers can often choose which to favor, as illustrated by the trade-offs that generated the Privacy Act of 197438
5 U.S.C. 552(a).
5 U.S.C. 552. The Freedom of Information Act provides a right of access to much of the information maintained in federal agency records. The Privacy Act exempts from disclosure much information that would intrusively and arguably unnecessarily expose private information about individuals.
C. Edwin Baker, “Posner's Privacy Mystery and the Failure of Economic Analysis of Law,” Georgia Law Review 12 (1978): 475–95.
Nevertheless, sometimes favoring or disfavoring a particular version of informational privacy can be a constitutional matter. Consider, first, two provisions of the U.S. Constitution that provide important protections of informational privacy. Protecting informational privacy is probably part of the explanation for, and certainly one consequence of, the Fourth and Fifth Amendments.41
But compare Lessig's observation concerning the need to determine whether the point of these amendments is to prevent intrusions, or to prevent insults to a dignity-based informational privacy, or substantively to limit government power, or, presumably, some combination of these goals. Lawrence Lessig, Code and Other Laws of Cyberspace (New York: Basic Books, 1999), 146–50.
Second, control over disclosure may be a significant aspect of First Amendment holdings protecting a person's right of anonymity, although whether the First Amendment justification for protecting anonymity really involves a general concern with a person's autonomous control over exposure of private information is doubtful. Often the justification involves anonymity's instrumental effect of making speech acts less costly to the speaker and, hence, of preventing the loss of publicly available speech due to the “chilling effect” of exposure. The goal may not be informational privacy itself but, rather, having more speech and, as a result, more information within a marketplace of ideas.42
NAACP v. Alabama ex rel Patterson, 357 U.S. 449 (1958); Brown v. Socialist Party, 459 U.S. 87 (1982); McIntyre v. Ohio Elections Commission, 514 U.S. 334 (1995).
Cf. Bowen v. Roy, 476 U.S. 693 (1986) (rejecting claim that a child receiving AFDC benefits had a religious free exercise claim to stop the federal government from requiring an identification number for receipt of benefits, but suggesting that the child might have a right not to use the number herself). In a demonstration at Stanford University in 1969, participants decided to give the movement's name in response to requests by the authorities that the demonstrators identify themselves. This tactic was neither an attempt to deceive nor a maneuver to hide (taking pictures was not discouraged) but a political statement of solidarity.
Sometimes a right not to disclose one's identity or other personal information obviously serves a person's instrumental need for a secluded expressive space in which to develop and to define herself. Consider recognition of an especially strong copyright in unpublished letters.44
Cf. Salinger v. Random House, Inc., 811 F.2d 90 (2d Cir. 1987) (recognizing strong rights in respect to unpublished letters, relying on Harper & Row, Publishers, Inc. v. Nation Enterprises, 471 U.S. 539 [1985]), with 17 U.S.C. § 107 (as amended in 1992 to make fair use of unpublished materials more acceptable).
U.S. Const., art. I, § 8 (Congress has power “to promote the progress of science and useful arts by securing for limited times to authors … the exclusive right to their respective writings …”). This “limited grant … is intended to motivate the creative activity of authors…. The monopoly … thus rewards the individual author in order to benefit the public.” Harper & Row Publishers, Inc. v Nation Enterprises, 471 U.S. at 546 (1985) (quoting Sony Corporation of America v. Universal City Studios, Inc., 464 U.S. 417, 429, 477 [1984] [dissenting opinion]).
Alternatively, First Amendment rights might disfavor informational privacy by trumping some privacy-affirming policies. Consider two possible types of First Amendment claims. First are claims made in behalf of a right to know or to have access to information, even when this information might be considered private. Imagine, for example, a claim that the public and/or the media have a right to enter a prison in a manner that a prisoner views as infringing upon her already-infringed-upon privacy. Although in closely divided decisions the Supreme Court basically rejected this claim, certainly there is plausibility to the idea that there is a constitutional right to access some information that someone else considers private.46
Pell v. Procunier, 417 U.S. 817 (1974); Saxbe v. Washington Post Co., 417 U.S. 843 (1974); Houchins v. KQED, Inc., 438 U.S. 1 (1978).
Globe Newspaper Co. v. Superior Court, 457 U.S. 596 (1982).
Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555 (1980).
The second type of First Amendment claim is the limit on privacy policies that is central to this essay. Formal autonomy—or, more specifically, speech freedom—allows a person to say what she wants. This right, I claim, exists even when her speech causes harm, for example, even if the speech exposes information about another that the other wants to keep private.
Constitutional rights are often usefully viewed as trumps or side constraints. A privacy trump, such as a Fourth or Fifth Amendment right, requires the legal order to protect or favor privacy over certain forced information exposures. The claim here is the opposite: that a First Amendment trump limits the ways that the law can protect privacy. The formal conception of autonomy centers on the agent being the final authority over decisions about herself or, in the case of speech, about her speech. This corresponds to the common conception that people should be free to say what they want and to listen to what someone else with a right to speak wants to disclose. The right exists even if, as is often the case (sometimes intentionally), her speech is instrumentally harmful to another. Moreover, respect for a person's expressive autonomy should mean that the person is free to listen and observe in places where she has a right to be and among people with whom she has a right to interact in order to learn more and then speak about it. If recognized as an aspect of freedom of speech, this autonomy-based speech right would mean that the law cannot protect privacy by limiting people's speech.
The argument for speech rights of media entities differs somewhat from the argument for individual speech rights. The arguments differ because institutional entities like press enterprises have no intrinsic autonomy claims.49
See Frank I. Michelman, Brennan and Democracy (Princeton, NJ: Princeton University Press, 1999), 12–14. There is generally no reason to think that the typical market-oriented, corporate media entity composed of numerous persons expresses or represents the unified, autonomous views of individuals involved in the enterprise, individuals who lose no individual expressive rights merely because of some regulation of the collective entity. Still, this claim may be less true in the case of voluntary expressive associations organized around the participants' solidaristic aims, which is why the Supreme Court has treated their First Amendment claims more generously than those of commercial or market-oriented corporations. Cf. Austin v. Michigan State Chamber of Commerce, 494 U.S. 652 (1990) with Federal Election Commission v. Massachusetts Citizens for Life, Inc., 479 U.S. 238 (1986).
See Stewart, supra note 22; Vince Blasi, “The Checking Value in First Amendment Theory,” American Bar Foundation Research Journal (1977): 521; and Baker, Human Liberty, supra note 5, chap. 10.
But see note 52. Justice Stewart's view that the press has greater speech rights (Stewart, supra note 22), a view never accepted by the Court, creates the absurd image of a person being liable for reading aloud to her breakfast companion a newspaper story that the newspaper had a right to print.
I can think of one exception. Although copyright restricts the freedom of both individuals and the press to say or print what they want, the press's constitutional role as a provider of information and vision may be adequately protected by a combination of an expansive fair use privilege and the idea/expression distinction. In combination, these guarantee a right to copy the idea or facts but not the actual words of a copyrighted item. In contrast, copyright should be unconstitutional under the First Amendment to the extent that it prevents an individual from expressing herself by repeating or distributing specific copyrighted expression. C. Edwin Baker, “First Amendment Limits on Copyright,” Vanderbilt Law Review 55 (2002): 891–951.
Rights other than speech rights may differ. The government may have power to legislate regarding media structure or ownership in order to make the media better serve its public functions without having an analogous power over individuals. C. Edwin Baker, “Turner Broadcasting: Content-Based Regulation of Persons and Presses,” Supreme Court Review (1994): 57–128. And the First Amendment may be a source of defensive rights that protect the institutional integrity of press entities, such as a “reporter's privilege” not to disclose a secret source, again without analogy to any individual rights. Baker, Human Liberty, supra note 5, chap. 10.
I put aside whether the First Amendment restricts the state's power to limit information-gathering activities. For example, does the First Amendment prevent the state from forbidding all information-gathering trespasses or, more interestingly, forbidding an inquirer from asking specific people, maybe jurors, for information about their deliberations? Likewise, I put aside whether the First Amendment limits the government's ability to impose damages or punishment on people for the dissemination of information that was initially acquired illegally, either by the speaker or by others.54
For skeptical conclusions, see Bartnicki v. Vopper, 532 U.S. 514 (2001) (invalidating restriction on publication of information of public importance that was obtained through illegal wiretap); Nebraska Press Ass'n v. Stuart, 427 U.S. 539 (1976) (Brennan, J., et al, concurring) (indicating injunction against publication of information about the accused was improper no matter how shabbily the information was obtained); Food Lion v. Capital Cities/ABC, Inc., 194 F.3d 505 (4th Cir. 1999) (rejecting publication damages for communicative content obtained through illegal means).
Cohen v. Cowles Media Co., 501 U.S. 663 (1991). Generally, confidentiality—a restriction on speech—can be required of an employee or a person with an agency relationship to another.
But cf. Jerry Kang, “Information Privacy in Cyberspace Transactions,” Stanford Law Review 50 (1998): 1193–1294.
Cf. Smith v. Daily Mail Publishing Co., 443 U.S. 97 (1979) (juvenile); Florida Star v. B. J. F., 491 U.S. 524 (1989) (rape victim).
V. Protection of Privacy
Unrestricted speech freedom does not leave informational privacy without protection. It only requires that the law not use a particular means—that is, abridging speech freedom—to protect privacy. The structure of this requirement is quite conventional in First Amendment contexts. The Court routinely holds that the government can pursue various goals, even the goal of effectively restricting what people or the press are able to communicate in order to preserve secrecy, confidentiality, or privacy, as long as the means do not involve restricting speech. Most obviously, the law may leave a person free to communicate whatever she knows but leave her unable to communicate certain information by denying her access to it. One permissible reason to deny her access is precisely to prevent her from being able to communicate certain information.
Thus, the Court would undoubtedly and unanimously hold that the state's interest in the confidentiality of preliminary investigations of judicial fitness justifies the state's nondisclosure (i.e., its preservation of the secrecy or privacy) of complaints about fitness, of the existence of an investigation, or of the information gathered therein. The state could reject requests by the public or the press for this information. The purpose of nondisclosure is to achieve state aims of informational privacy by disabling people from engaging in speech with particular content: what the press does not know, it cannot report. Nevertheless, in Landmark Communications v. Virginia (1978),58
435 U.S. 829 (1978).
443 U.S. 97 (1979).
491 U.S. 524 (1989).
Both the rules that the Court would presumably uphold (government nondisclosure) and the rules that it struck down in Landmark, Florida Star, and Daily Mail (prohibitions on communications) aim at restricting dissemination of the same information. Both sets of rules, if allowed and effective, would accomplish the same end: preventing communication. The most salient difference is that the permissible restrictions directly limit only access to a resource, namely information, that is instrumentally useful to speakers and arguably to listeners (i.e., to their effective autonomy), while the impermissible restrictions explicitly limit a speaker's choice to speak or publish (i.e., limit her formal autonomy). The state acts properly in basing its information policies on instrumentalist, policy judgments about the value of privacy as compared to the value of information accessibility. The Court, however, blocks the execution of the state policy if the policy is carried out by means that violate formal autonomy. (Note that because both sets of rules serve the same state interest and thwart the exposure of the same information, the distinction between the rules is difficult—although maybe on rule-utilitarian grounds, possible—to explain on a marketplace-of-ideas theory of the First Amendment, which values speech instrumentally in terms of its contribution to the information available to people. In contrast, the difference is easily explained on grounds of respecting speakers' autonomy.)
This example makes clear that privacy is not without statutory protection. Privacy may also gain a degree of constitutional status due to people's episodic opportunities to be in nonpublic spaces. Consider, as I previously pointed out, the Fourth Amendment's protection of the home and of private papers, and the Fifth Amendment's protection in one special context of the right not to self-disclose. Other means exist to promote a world in which privacy is treated with greater regard. I share to some degree Warren's and Brandeis's tastes for more serious content in newspapers, although at dinner among colleagues, it is easy to see that even educated people (whether or not they are as aristocratic as Warren and Brandeis were) gossip, especially about academics who are not present. Possibly the government should try to structure media industries so that they would favor more “serious” content.61
Although the Court in Turner Broadcasting System v. FCC, 512 U.S. 622 (1964) appeared to rule this purpose constitutionally impermissible, all the Supreme Court's prior cases involving structural regulation of the media suggest the opposite. Baker, “Turner Broadcasting,” supra note 53.
For the best development of this claim, see Diane L. Zimmerman, “Requiem for a Heavyweight: A Farewell to Warren and Brandeis's Privacy Tort,” Cornell Law Review 68 (1983): 291–367.
The law can intervene to protect informational privacy at distinct junctures and in different ways. Most generally, the legal order can (1) restrict or regulate the initial alienation of private information; (2) restrict or regulate the gathering of information, most obviously by aiding a person in keeping information from originally being nonconsensually exposed; (3) prohibit further dissemination of information to which the subject objects; and (4) restrict or regulate particular uses (other than dissemination) of information that has been gathered. That is, the law can protect informational privacy by restricting the alienation, gathering, dissemination, and use of information. The first of these means, at least if applied to an individual's speech about herself, violates her speech rights, although I will suggest a different conclusion in the context of market alienations. The third of these means, at least as applied to individual speakers or to the media, violates the free speech principles assumed here. Subsections A and B below explore the second and fourth means of protecting informational privacy.
A. Restrictions on gathering private information
Many restrictions on gathering information are uncontroversial. As noted previously, the Fourth and Fifth Amendments impose gathering restrictions at least on government actors. (Conversely, the Sixth Amendment grants the accused a right to compulsory process for obtaining witnesses in his favor; this information-gathering right potentially intrudes into informational privacy and has been extended by other discovery rules and provisions for compulsory process.) One function of private property may be to protect a physical zone of privacy.63
C. Edwin Baker, “Disaggregating the Concept of Property in Constitutional Law,” in Nicholas Mercuro and Warren J. Samuels, The Economics of Legal Relationships, vol. 4, The Fundamental Interrelationships between Government and Property (Stamford, CT: JAI Press, 1999).
Protection of “private zones” does not mean that the government should have unrestricted authority to limit gathering information. Prohibitions on journalists or presumably on anyone else observing, “monitoring,” or even photographing a person when she appears in public realms should be (and probably are) impermissible, even though the observations necessarily expose information.64
Cf. Galella v. Onassis, 487 F.2d 986 (2d Cir. 1973).
But cf. Hill v. Colorado, 530 U.S. 703 (2000).
As noted, the Court has struck down applications of laws against publishing the name of juvenile defendants or rape victims, at least it has when the name has been “lawfully obtained.”66
Smith v. Daily Mail Publishing Co., 443 U.S. 97 (1979); Florida Star v. B. J. F., 491 U.S. 524 (1989).
Nevertheless, the Court has written its opinions in this area very narrowly. In troubling dicta, the Court in Florida Star, while protecting publication of the rape victim's name, left open the question of whether “the Constitution permitted a State to proscribe receipt of [this] information,” and, if it did, whether it could then prohibit publication.67
Florida Star, 491 U.S. at 536.
Two additional problematic questions concern illegally obtained information. First, no one doubts that a person is liable for any tort or crime committed while gathering information, but can she also be punished or have damages increased for communicating the information that she obtained illegally?68
Cf. Food Lion v. Capital Cities/ABC Inc., 194 F.3d 505 (4th Cir. 1999) (rejecting, on First Amendment grounds, publication damages resulting from dissemination of illegally obtained information), with Dietemann v. Time, Inc., 449 F.2d 245 (9th Cir. 1971) (allowing damages for tort to be enhanced due to publication).
Pearson v. Dodd, 410 F.2d 701 (D.C. Cir. 1969) (rejecting damages for publication where publisher did not commit the illegal intrusion).
New York Times v. United States, 403 U.S. 713 (1971).
Food Lion, 194 F.3d 505.
Lower courts have struggled without arriving at a uniform answer to either question raised in the previous paragraph. Recently, in Bartnicki v. Vopper (2001),72
532 U.S. 514 (2001).
Long the doctrinal norm, this principle has been rejected in the arguably unique circumstances of child porn. New York v. Ferber, 458 U.S. 747 (1982).
Drawing on distinctions possibly important for copyright, a different analysis is possible. Arguably, copyright should be understood to violate the First Amendment if it prevents a person in a noncommercial context from saying whatever she likes, including the entire content of a copyrighted item. Such a restriction directly infringes the speaker's liberty, in particular, her freedom of speech, on the basis of content. (Note, although closer to existing law than might be at first imagined, the analysis here does not purport to track existing case law but represents my attempt to show the constitutionally legitimate scope of copyright law.)74
Baker, supra note 52.
In contrast, copyright's goal of providing an incentive for valuable creative behavior properly protects people's expressive works from commercial appropriation without consent, while appropriately broad interpretations of two doctrines—'fair use’ and the noncopyrightability of ‘facts and ideas', only of expression—prevents any real limitation on the instrumentally justified constitutional role of the press. Since the media has no autonomy interests and often no constitutional interest in the precise expressive formulation of its product, but only a constitutional interest in being able to provide the public with information or vision, copyright as limited by a broad interpretation of these two doctrines does not abridge the press's constitutionally protected role. The Court's distinction in Bartnicki relating to material of public concern is roughly analogous. Explicit photographic images and private diary language can provide the public with valuable perspectives on human alternatives, but it is unclear that for this the media needs illegally obtained content. Either fictionalized or consensually obtained material would arguably suffice. Thus, prohibiting media reproduction of illegally obtained content that lacks “public concern” may be appropriate.75
The argument in the text would not seem confined to illegally obtained information. For instance, the state could argue that the name of a rape victim or the name of person who in the distant past engaged in some disreputable behavior is or is no longer a matter of public interest. So far, courts have, in my view properly, mostly rejected these arguments. Cox Broadcasting Corp. v. Cohn, 420 U.S. 469 (1975); Hayes v. Alfred A. Knopf, Inc., 8 F.3d 1222 (7th Cir. 1993). The information may add to the journalistic or scholarly integrity of the reports as well as provide truthful information about individuals that other persons may want.
See Eastwood v. Superior Court, 149 Cal. App. 3d 409 (1983); Hoffman v. Capital Cities/ABC, Inc., 33 F. Supp. 2d 867 (C.D. Cal. 1999). This point does not suggest that a right to protect one's image from unwanted use is wise policy even in these contexts. See Michael Madow, “Private Ownership of Public Image: Popular Culture and Publicity Rights,” California Law Review 81 (1993): 125–240.
Additional questions arise about which both the law and normative theory seem unclear. Even if it is granted that people have an autonomy-based First Amendment right to gather information by either observing or listening to another person in public, does this right restrict the extent to which the state (1) can limit when space is considered “public” or (2) restrict use of technologically enhanced means of gathering information? My tentative answer to both questions is that existing law rightfully assumes that the state has such policymaking authority and that there is no persuasive First Amendment basis for objection.
Note that the two questions are parallel. The normative premise behind the idea of a public space is not that all information should necessarily be public but that, when in a public space, people should have an anarchic right to provide, gather, and receive information. There does not seem to be, however, any logical or natural way to determine in particular instances the borders of the public sphere in which this right of a flesh and blood person obtains. The informational-privacy-related reason to protect private property is to provide places beyond the peering eyes and listening ears to which a person is exposed in public. The law, though, could reduce protection of privacy by eliminating or limiting trespass laws to certain types of private property.77
Cf. Florida Publishing v. Fletcher, 340 So. 2d 914 (1976); Prahl v. Brosamle, Case No. 152-062, Circuit Court, Dane County, August 28, 1982, cited in Marc A. Franklin, David A. Anderson, and Fred H. Cate, Mass Media Law, 6th ed. (New York, Foundation Press: 2000), 570. Neither the notion of private property nor the Constitution requires a trespass doctrine like that common in most American jurisdictions. A court developing an implied consent for reporters' presence on private property, such as developed in Fletcher or Prahl, obviously does not assume that the owner would have actually consented in these circumstances, but, rather, assumes a public policy justification for treating reporters' presence as “custom.”
Information that is only available to listeners or observers through technologically enhanced means should not be understood as inherently “in” a public space. Rather, allowing nonconsensual use of technology amounts to saying that the information it generates is effectively in a public space, while limiting the technology's use amounts to the opposite, treating the information as in a private sphere and the technology's use as trespassing. Both property rules and rules about technology use construct realms where people either are or are not protected from nonconsensual exposure. Both property and technologically enhanced observation could be regulated to protect privacy so that access only comes with permission, or could be left unrestricted so people are exposed. Just as the legal creation of both public space and private property is sociologically important, it seems likely that the law ought not automatically treat all the different technologies or different contexts of their use in the same way. Certainly, the value of privacy justifies treating some uses of technology as the equivalent of an invasion of a legally protected private sphere.
Telephone conversations are typically carried on cable or fiber optic lines that cross public space or on electromagnetic radio waves. Technology may allow a nonparticipant to hear (and record) conversations, although these eavesdroppers are as uninvited as if the conversation occurred in one's private home. Existing wiretap laws sensibly treat conversations floating across public spaces as private, even if an intruding listener is traveling on a public road and the participants in the conversation are in a public park using cell phones.
Available technology allows a person receiving a call to identify the caller or, under readily available current technology, at least to identify the listed owner of the caller's phone. To protect the caller's privacy, could the government restrict the use of this “caller ID” technology, leaving it to the initiator of the call to voluntarily disclose her identity? The general assumption emphasized above is that people cannot be restricted from using their eyes and ears to identify a person. Putting aside the policy merits of this assumption, the question now is, does this freedom apply when the observation only occurs through use of technology? My claim has been that neither the interest in privacy nor the right to gather information automatically determines the best answer. A person should not automatically be assumed to have appeared in public for observation because she uses a telephone. Rather, whether she has so appeared should depend on society's policy choices about borders.
Even before the recent commercialization of “caller ID” technology, the architecture of the existing phone system allowed telephone companies to identify the phone initiating a call in order to route the recipient's reply to the right phone and to direct the phone bill to the responsible party. That is, the technology required a caller to identify herself (or at least her phone) to the phone company and its record keepers. Any law that limits a phone company's freedom to further utilize or communicate these records amounts to declaring the phone call private, which would be like treating it as having occurred on visually or aurally secluded private property controlled by the speakers, despite the caller having in a sense “voluntarily” exposed herself to the phone company. The law protects privacy in the call by restricting the phone company's speech. At first glance, the restraint might seem analogous to, but even less permissible than, restraints on further communications of illegally obtained information. The cases are distinguishable for two reasons, however. First, the telephone company should not be treated as having First Amendment speech rights. Second, the law can be treated as, in effect, creating a mandatory confidentiality term to the contract between the phone user and the company.
Likewise, the government should be able to protect privacy by declaring that disclosures functionally necessary for computer interactions are not to be treated as in a public space or as having been, in a constitutionally relevant sense, voluntarily exposed. Law professor Jerry Kang persuasively argues that the government should make the default rule be that the observer (the other computer or its owner) or the owner of the transmission system cannot permissibly “know” (make use of) the information except as needed to engage in the particular communicative interaction then occurring.78
Kang, supra note 56.
See Ann Bartow, “Our Data, Ourselves: Privacy, Propertization, and Gender,” University of San Francisco Law Review 34 (2000): 633–704.
Technology, of course, changes many things. Not only does technology enhance the power to hear or observe, it also allows making a permanent record of what is currently available to a person to see, hear, and repeat or publish verbatim. Some states outlaw taping without the consent of all speakers. Since my only point here is that protection of a broad speech freedom, which includes the right to disseminate “private information” about another, does not prevent appropriate legal regulation of the information-gathering process, I leave aside the issue of whether this regulation of technology should be permissible. Journalists have argued to no avail that these laws make their reporting less accurate because they must rely on memory and notes, and, even when their stories are accurate, the laws leave the media much more vulnerable to defamation suits because it is more difficult to prove the accuracy of their reports.80
Dietemann v. Time, Inc. 449 F.2d 245 (9th Cir. 1971); Shevin v Sunbeam Television Corp., 351 So. 2d 723 (Fla. 1977).
Diane Leenheer Zimmerman, “I Spy: The Newsgatherer under Cover,” University of Richmond Law Review 33 (2000): 1185–1231 (review and tentative First Amendment critique of these restraints).
B. Uses and users of information
Even though a party properly gathers or observes personal information about another, the use of such information presents an additional issue. An employer might freely observe another's race or sex but be forbidden to use that information in her hiring decisions. Free speech issues arise when the government regulates further communication of the information. Still, although the contexts are variable and sometimes raise controversial constitutional issues, restrictions on other uses usually are and should be permitted.
Consider cases where information initially was legally obtained but by means that were not themselves constitutionally protected, for example, it was not gathered within constitutionally protected conversation or through constitutionally protected observation in a public sphere. Under such circumstances, the source—the government or private party—may have imposed as a condition of access that the information only be used for particular purposes or that it not be used in particular ways. Often the government restricts the use of personnel files, sometimes just to the uses that justified gathering the information in the first place. It regularly provides certain government employees with personal or private information about other employees or about members of the public, but limits any further disclosure of the records apart from the purpose for which they were provided.82
Cf. Snepp v. United States, 444 U.S. 507 (1980) (federal government can require former CIA employee not to disclose information obtained during employment without approval of agency).
467 U.S. 20 (1984).
See Seattle Times, 467 U.S. at 37 (Brennan, J., and Marshall, J., concurring). Analogous to the unconstitutional condition doctrine, restrictions imposed by employers on further use by employees or restrictions imposed by contract could be unenforceable because contrary to public policy. A court could have easily decided to protect publication of the source's name in Cohen v. Cowles Media Co., 501 U.S. 663 (1991), or information provided by whistle-blowers on this nonconstitutional ground.
In the late 1990s, lower courts invalidated a law that, though it made available names and addresses of arrestees to private parties for many purposes, including media exposure, denied the information to recipients who intended to use the information directly or indirectly to sell a product or service. In Los Angeles Police Department v. United Reporting Publishing Corp. (1999), the Supreme Court reversed.85
528 U.S. 32 (1999).
See supra note 84.
Baker, Human Liberty, supra note 5, chap. 9.
Central Hudson Gas & Electric Corp. v. Public Service Commission of New York, 447 U.S. 557 (1980).
In addition to regulating certain uses of “private” information (for example, for hiring decisions) and imposing appropriate conditions on governmentally supplied information, another basis for restricting use may be especially important: the government's general authority to regulate businesses and commercial practices, including those of professionals. Securities laws illustrate that this power extends to regulating speech involved in this business. Disclosures are required—and sometimes prohibited. Insider use of information is forbidden. In the legal world, also, government regularly restricts speech and the use of information. Although sometimes cast as a disagreement over the propriety of regulating speech that is inconsistent with a defense attorney's role, arguably the disagreement that ensued among the justices over the standard for regulating a defense attorney's out-of-court speech was primarily over their differing visions of a defense attorney's role.89
Gentile v. State Bar of Nevada, 501 U.S. 1030 (1991) (5–4 decision upholding particular restriction on defense attorney's out-of-court speech). The case involved a lawyer's public statements at a news conference that he called to defend his client and to attack the veracity of the police. The Nevada Supreme Court had issued rules preventing defense attorneys from making public statements that might taint the jury pool. The attorney's case against the state disciplinary board's recommendation that he be reprimanded eventually reached the U.S. Supreme Court.
See, e.g., Baker, Human Liberty, supra note 5, chap. 9.
Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748 (1976); First National Bank of Boston v. Bellotti, 435 U.S. 765 (1978); Burt Neuborne, “The First Amendment and Government Regulation of Capital Markets,” Brooklyn Law Review 55 (1989): 5–63.
If commercial entities can be limited in ways that would be impermissible to restrict noncommercial or media entities, determining what constitutes a media entity will sometimes be crucial in determining the permissibility of privacy regulations. One possibility is that First Amendment protection extends to any information provider, any entity that sells or has information (or opinion or art or music) as its product. This, however, seems too broad. Lawyers and doctors and accountants are largely information providers but are not likely candidates to be characterized as “the press.” Alternatively, the protected entity could be one that offers nonindividualized communications as its product and makes this expression widely available to an interested public. The Investment Act of 1940 restricts who can engage in the investment-advice business, but exempts those who give advice by means of a bona fide publication of general and regular circulation. The Court majority in Lowe v. Securities and Exchange Commission (1985),92
472 U.S. 181 (1985).
Lewis Publishing Co. v. Morgan, 229 U.S. 288 (1913). See generally, C. Edwin Baker, Advertising and a Democratic Press (Princeton, NJ: Princeton University Press, 1994), 118–37.
851 F.2d 365 (D.C. Cir. 1988).
Despite this “commercial speech” explanation for the court's conclusion, the court's actual analysis was based on “government's broad power to regulate the securities industry.” Clearly, however, the court was misguided to imply any broad, general governmental authority to regulate individuals' noncommercial speech or media speech about the securities industry.
The difficulty of deciding what counts as media may apply to (commercially accessible) databases. They are part of an information-supplying or communications industry. Should they either always or sometimes be treated as constitutionally protected media? Maybe not. In Dun & Bradstreet v. Greenmoss Builders, Inc. (1985),96
472 U.S. 749 (1985).
Gertz v. Robert Welch, Inc., 418 U.S. 323 (1974).
Hadden View Investment Co. v. Coopers & Lybrand, 70 Ohio St. 2d 154, 436 N.E.2d 212 (1982) (accountant malpractice).
Gutter v. Dow Jones Inc., 22 Ohio St. 3d 286, 490 N.E.2d 898 (1989) (liability for error in stock table denied in contrast to liability for accountant error).
Cf. Fair Credit Reporting Act (1970) and Consumer Credit Reporting Act of 1996, 15 U.S.C. § 1681–1681t.
Though Greenmoss could be understood to be about regulating a nonmedia business, a general power to regulate databases raises a somewhat different issue. Modern computer-based technology makes databases increasingly cheap to construct and easy to use. In finding a “personal privacy” interest in information that had previously been made public and was now stored on an FBI database containing rap sheets maintained on 24 million persons, the Court properly observed that “there is a vast difference between the public records that might be found after a diligent search of courthouse files, county archives, and local police stations throughout the country and a computerized summary located in a single clearinghouse of information.”101
United States Dep't of Justice v. Reporters Committee for Freedom of the Press, 489 U.S. 749 (1989).
My thinking here was prompted by Cohen, supra note 6.
As to the first context, using computer technology in processing and organizing information about a person is in many respects parallel to using technology to gather information on her that would not otherwise be available to our eyes and ears. Arguably, both uses of technology should be equally subject to regulation. Or maybe an even closer analogy is tape recording. Even when a person has a constitutional right to hear and publish verbatim the contents of a speech, current doctrine permits a legal prohibition on recording it without the consent of the speaker, thereby preventing the listener from making the most accurate use of information that she rightfully possesses. In each of these cases, information about a person has been exposed, but the law would restrict technology that makes the information more usable. Nevertheless, it might be wise to resist these pro-regulatory conclusions. Not having a general right to process, comprehend, or access information that one rightfully has (and then to communicate it) seems to contradict the freedom of the intellect that surely is part of the meaning of free speech. Could a newspaper be barred from keeping and using files of its past stories? If not, could it be prohibited from using computers to do so at a lower cost than before? Can individuals, for example, sociologists or historians, be prohibited from engaging in such activities? If the answer to these questions is “no,” although the issue is not beyond doubt, then general restrictions—for example, restrictions that apply not just to certain businesses or to certain uses but to all individuals and the press—on entering information into databases seem very questionable. Moreover, as to the second context, the same arguments apply to regulation of the maintenance of databases.
Regarding the third and fourth contexts—laws restricting creation or uses of databases for certain commercial purposes—the permissibility of such laws may well be the same as the permissibility of regulating other commercial practices. All commercial practices, including the creation and use of databases, are generally subject to regulation while noncommercial and media uses of information are properly protected unless a particular use (e.g., for criminal purposes) is independently outlawed.103
But cf. the government's questionable concession in United Reporting, supra note 85, that if the commercial user “independently acquires the data, the First Amendment protects its right to communicate it to others.” 528 U.S. at 45 (Stevens, J., dissenting).
All scope for a right of publicity is critiqued in Michael Madow, supra note 76. My claim here is that the constitutional critique should only apply to nonadvertising appropriations of another person's image.
If the restrictions described in the above paragraph are acceptable, then the implications are huge for greater protection of privacy. Those espousing privacy values often state them in sweeping, general terms as a purported right of a person to control (usually personal or private) information about herself. However, the issue is mostly inflamed by people's sense of being ever more subject to manipulation, harassment, or targeting (e.g., unwanted phone solicitations) by market entities (or by government). If gathering, assembling, and using private information for commercial purposes were limited, the major concerns of privacy advocates would be met. These rules could also dramatically reduce the incentives for construction of offending databases, thereby further reducing the threat to privacy.
Regulation of market activities might even go a step further. General rules forbidding alienation of private information impermissibly restrict a person's speech about herself. I suggested, however, that the law could establish a default position that information exposed by engaging in digital communications could only be used for functionally necessary aspects of the particular interaction involved. Still, commercial entities may be in a powerful strategic position to obtain consent to unrestricted use of the information in exchange for something they offer, such as access to their Web site, individualized service, a lower price or free use, or whatever. Standard reasons to regulate market transactions—for example, unequal information, structures that create inappropriate amounts of power of one party over the other, negative impact on third parties—could possibly justify prohibitions on market alienation of private information. A person would be left free, though, to publish personal information or even to give it to a particular commercial entity, except not in exchange for some benefit.
Many privacy advocates will not be satisfied with the protections defended here. Nevertheless, much of the rise in popular concern about disappearing privacy apparently, and quite reasonably, relates to fears about the collection and manipulative use of information by corporate and governmental bodies. Basically, my claim is that most privacy-protection policies that relate to these threats, although possibly difficult to secure politically, do not conflict with First Amendment principles. The area of real conflict between free speech and privacy is much more limited. The abstract principle that requires respect for people as autonomous agents in control of their own speech choices dictates that, in this discrete context, privacy claims should lose. Sections VI and VII will argue that this result is pragmatically justified on the basis of plausible, though inevitably inconclusive, consequentialist considerations. However, before taking up that issue, a final means of protecting privacy, which turns out to be fully consistent with the First Amendment, merits comment.
C. Strongest protector: Nonlegal norms
Probably the greatest protection of informational privacy comes through voluntary decisions not to disclose “private” information about another person or, possibly more often, to monitor closely when and to whom to make disclosures. These decisions are often quite rigidly determined by nonlegal social norms, such as the injunction not to “snitch.” All communities value privacy, although to varying degrees and in relation to different information. Communities predictably develop different privacy-protecting social norms or practices that embody their judgments about privacy. Robert Post is surely correct about the necessity of “civility rules” for the maintenance of communities, communities that are themselves necessary for autonomous individuals to develop.105
Robert C. Post, “The Social Foundations of Defamation Law: Reputation and the Constitution,” California Law Review 74 (1986): 691; Robert Post, Constitutional Domains: Democracy, Community, Management (Cambridge, MA: Harvard University Press, 1995). For a liberal account of the necessity of community for development of individual autonomy and its implications for political theory, see also Will Kymlicka, Liberalism, Community, and Culture (Oxford: Oxford University Press, 1989).
The mistake would be to assume that these civility rules require legal or other formal enforcement mechanisms. As an empirical matter, I expect that reliance on social practice will maintain sufficient civility rules to fully meet Post's concern with having an environment in which people can develop as (autonomous) individuals. In contrast, legal enforcement of civility rules that protect privacy against offending communications, as well as enforcement of most other civility rules relating to expression, may make communities more rigid, oppressive, and slow to adopt useful change.
Contrary to frequent communitarian characterizations of liberal theory, my claims on behalf of the First Amendment never assume that individuals are factually “unencumbered” or that effective or meaningful autonomy can be a context-free conception. Both legal and societal respect for formal autonomy and practical realization of meaningful autonomy are historical achievements, not abstract realities that provide an ahistorical basis for some mythical social contract.106
See Baker, “Sandel on Rawls,” supra note 29.
The voluntariness of language rules does not mean that grammar and shared word meanings are not of major significance or a proper subject of government policy. Public education and, in a curiously different way, public support for the arts both, at times, exemplify such policies. (This fact is roughly analogous to the point of the two preceding subsections.) Information privacy can be a matter of policy embodied in legal rules even if not protected directly by law that prohibits offending speech choices. Civility rules—group choices—about what information should be private and private in what ways or to what extent, operate within a legal structure that protects against certain information-collection devices (e.g., secretly taping what happens in the bedroom) and that regulates various commercial uses of information.
These nonlegal norms protecting privacy can be both extraordinarily important and quite effective. They often protect secretive individuals, adding to their freedom. They both maintain and are maintained by group solidarity. Legal scholar Randall Kennedy describes how most African Americans know of other blacks who “pass.” Despite African Americans' general disapproval of passing and despite their capacity to prevent it by exposure, social norms effectively prohibit exposure, especially exposure to whites.107
Randall Kennedy, “Racial Passing,” Ohio State Law Journal 62 (2001): 1145–93, 1171–73. Even if African Americans mostly condemn passing, many also view it as a method to flout and subvert silly but oppressive racist laws and norms. Id. at 1169–70. Kennedy observes that “Langston Hughes repeatedly defended passing as a joke on racism.” Id.
Id. at 1171 (citing Shirlee Taylor Haizlip's account). Cf. NAACP v. Claiborne Hardware Co., 458 U.S. 886 (1982) (among constitutionally protected methods of exercising power, boycotters of white businesses applied pressure on black “violators” by reading aloud their names in church and by publishing their names in a local black newspaper).
The account offered by communications professor Larry Gross of the history and debates over “outing” of gays makes the point even more powerfully.109
This paragraph is based entirely on Gross, supra note 2.
Id. at 9. Corpses may lie with both alternatives. Gross notes the suggestion that lack of positive role models contributes to the severely disproportionate number of suicides among gay youth. Id. at 126.
The lesson of this history of outing, as well as the history of passing, might be threefold. First, informational privacy can be extraordinarily important and valuable even when, as the example of passing illustrates, the legal order would happily condone disclosure. Second, especially among those who have the information and, thus, often the power111
The capacity of someone with information to “out” another should not be assumed too quickly. A common mainstream view has been that homosexuality is not so bad as long as kept secret. The “Don't ask, don't tell” policy had not yet been invented by the U.S. military when Gross wrote his book, but he fully explained the appeal of such a policy. Gross, supra note 2, at 144–52. The mainstream press reflected this view by engaging in “inning,” that is, by refusing to indicate a person's homosexuality in contexts where it would indicate a person's heterosexuality. Similarly, the press often refused to report as “news” that a major figure who assumed a public heterosexual image was actually gay, while routinely reporting other personal “news” about the figure.
This point should not obscure the fact that “outing” in neighborhoods, employment, and social contexts of gays, usually by nongays, also occurs and often has tremendously harmful consequences. At the end of my discussion of gossip below in the text, I note that gossip has overtly negative as well as useful consequences.
VI. Overvaluation of Privacy
My abstract thesis is that formal autonomy, or free speech rights, should operate as a side constraint on policy formulation, that is, it should act as a “trump.” Speech rights should prevail even when they run roughshod over people's desire to keep information private. This thesis rejects giving people any property right in information about themselves that can be used to control the communicative choices of other individuals or the media.
Nevertheless, informational privacy is extraordinarily important to individuals and to communities. The information that should be kept private and the appropriate methods of guarding this privacy, however, are matters on which people and communities will differ. Societal self-definition involves creating particular types of communities, encouraging certain values and practices, and even supporting certain images of persons. In formulating privacy policies, all of these considerations, as well as concerns about efficiency and distribution, should have a role. Of course, the proper content and weight of these efficiency, justice, and self-definitional concerns, even as goals, are contestable. Nevertheless, elsewhere I have argued that their appropriate elaboration would lead to considerable legal protection for the privacy of persons and much less protection for the privacy of instrumentally valued, institutional creations such as corporations (or governments).113
Baker, “Posner's Privacy Mystery,” supra note 40. This argument responds to the opposite conclusion offered in Richard Posner, “The Right of Privacy,” Georgia Law Review 12 (1978): 393–422.
Now, I want to take a different tack. This section suggests that privacy, no matter how essential for people's flourishing, currently is in danger of being seriously overrated, at least in many contexts. Thus, my claim here begins an inevitably incomplete pragmatic defense of the thesis that, when at stake, formal autonomy should “trump” privacy or, if the trumping conception is resisted, that formal autonomy justifies great caution before being compromised by pro-privacy policies.
Warren and Brandeis's encomiums for a right of informational privacy have resounded more in the court of public opinion than in the courts of law. Currently, great popular support exists for a right to privacy from the snooping eyes of government, of the media, and of corporate marketers. Also, people commonly disapprove of, for somewhat different reasons, neighborhood or office “gossip.” Although I certainly cannot prove it, I am inclined to believe that there has been a recent surge in the popular appeal of informational privacy.114
Zimmerman, “I Spy,” supra note 81.
Zimmerman, “Requiem for a Heavyweight,” supra note 62.
Greater popularity may reflect the plausible view that the increase in commercial use of newly cheapened, individualized marketing strategies and mounting encroachments by government have put privacy in greater danger today than ever before. Resistance to the “disciplinary agencies” of government and the market is surely progressive and justified. Possibly the central factor in their greater power, as well as a prime generator of increased fears, is the incredible advances in computer and related digital technologies that dramatically reduce the cost of collecting information and, possibly more ominously, lower the cost of indexing, storing, and retrieving it.116
Oscar H. Gandy, Jr., The Panoptic Sort: A Political Economy of Personal Information (Boulder, CO: Westview Press, 1993).
Nevertheless, another possible basis for the growing popularity of privacy rights may be a sign of cultural sickness. Troublingly, the high valuation of privacy may reflect increasing desires to withdraw from civil society and especially from the public sphere. This increasingly common preference for virtually complete withdrawal stands in dramatic contrast to the classic vision of a private realm as a necessary base to which a person periodically repairs, but always with the hope and expectation of returning to the public world. My fear is that the current positive valuation of privacy reflects a society in which all value is increasingly seen as located in private life. To an increasing extent, people seem to find all meaning in private interactions of family and other personal associations or, even more disconcertingly, in a more purely commodity-oriented world of private consumption with value largely based on wealth and material goods.117
I am not in a position to prove this characterization to those who do not see it. Still, declining levels of voting, the media's reduced emphasis on policy-oriented or political news, the view that society's problems are more a matter of charity or private action than public policy, a cynical view of politicians and public servants, the so-called “bowling alone” phenomenon, as well as interpretations of popular culture, are among the features of social life that I would examine for evidence. Battles over taxes (or the size of government) can be seen as disagreements about whether marginal value lies more in public or private expenditures of resources and energies. A retreat to private life, if it exists, may be a uniformly global phenomenon or it may be that the United States is alone or at least in the vanguard, in which case comparative evidence would be informative. For example, the only reports that appeared in the mainstream American media on the 2003 Finnish parliamentary elections emphasized the lack of policy differences or issues, other than how to respond to increased unemployment, between the two dominant parties, which produced a voter turnout of only 70 percent. Lizette Alvarez, “Finnish Center Party Edges Past Social Democrats in Election,” New York Times, March 17, 2003, late edition–final, sec. A, p. 2, col. 3. Only by going beyond U.S. media, say, by viewing BBC online, could one discover that another prominent issue was the losing Social Democrats' inadequate opposition to President Bush's military policy toward Iraq. A Finnish academic told me, however, that the real story was the dramatic loss of seats by the conservative party, which apparently fell from favor because of the unpopularity of its promise of a tax cut, while both leading leftist parties picked up seats.
Withdrawal represents a direct threat to the First Amendment values of dissent and challenge to the status quo.118
Steven H. Shiffrin, Dissent, Injustice, and the Meanings of America (Princeton, NJ: Princeton University Press, 1999).
Report of The Committee on Homosexual Offenses and Prostitution, The Wolfenden Report, authorized American ed. (New York: Stein and Day, 1963). The report stimulated a famous debate between Patrick Devlin and H. L. A. Hart. See Patrick Devlin, The Enforcement of Morals (Oxford: Oxford University Press, 1965); and H. L. A. Hart, Law, Liberty, and Morality (Stanford, CA: Stanford University Press, 1963).
Gross, supra note 2, at 144–50. See also Larry Cata Becker, “Exposing the Perversions of Toleration: The Decriminalization of Private Sexual Conduct, the Model Penal Code, and the Oxymoron of Liberal Toleration,” Florida Law Review 45 (1993): 755–802.
Gross, supra note 2, at 149–50.
As I previously noted, during the late 1980s and 1990s when “outing” was a central topic within the gay community, the matter of informational privacy almost inevitably came to the fore. Describing it as a right/left split, Gross noted two views of outing. First, many gay advocates argued that “the primary issue for the gay movement (perhaps the only issue) was the protection of the right of privacy.”122
Id. at 146.
Id. at 152.
Id. at 127.
Id. at 146.
Gross observed that publicity played a key role in the process of societal change in attitudes toward heterosexual cohabitation. He concluded that there is “no way to move beyond [antigay] attitudes without accepting and acting on the presumption that homosexuality must be seen and treated the same as heterosexuality…. [I]t behooves … us … to act accordingly, and outing—or [as journalist and activist Gabriel Aotello calls it], equalizing—is a means to this end.”126
Id. at 169–70.
Id. at 172 (quoting Benjamin Schatz, “Should We Rethink the Right to Privacy?” Advocate, February 1991).
Another illustration of the devaluing of public life can be found in a rarely commented upon aspect of libel law. A society that highly valued public life and civic participation would presumably reward and encourage, not penalize and burden, participation in public life. Although such a society would maintain and respect a private sphere to which all people could repair for relief, provisions, contemplation, and amusement, it would hardly encourage a total retreat to a “purely” private life. People in public life share with people who shun public roles a legitimate interest in maintaining their reputations and in conducting their private lives free from prying eyes. Actually, the parallel is not exact. For a person in public life, the consequences of injury to her reputation among people whom she does not know, with whom she does not work, and whom she will probably never meet, are normally much more severe than is a similar injury for those who do not enter public life. Reputation among strangers is often a major asset, a virtual currency, for people in public life, but it is of little instrumental significance for a purely private person. To oversimplify, reputation is truly valuable for a private person only among her close compatriots, while for the public person it is also extraordinarily valuable among strangers. Media libels carry a greater risk of influencing the perceptions of a public figure in the minds of strangers than of friends. Strangers will have fewer or no firsthand alternative bases on which to evaluate the libel, fewer opportunities to hear the libeled person's defense, and often less incentive to inquire further. Thus, as compared to libels communicated to friends or even acquaintances, libels communicated to strangers are more likely to be determinative of their views. In other words, the legitimate interest in protection from media libels seems to be much greater in the case of people in public life.
Yet, existing defamation law reverses both my normative premises and my descriptive hypotheses. Rather than reward a person for becoming a public official or public figure by specially honoring her legitimate interests (e.g., in being spoken about truthfully), defamation law burdens such a person with greater legal vulnerability to libelous falsehoods.128
Gertz v. Robert Welch, Inc., 418 U.S. 323 (1974).
The Court has not ruled on the application of this tort to private persons. However, in Hustler Magazine v. Falwell, 485 U.S. 46 (1988), while applying the Sullivan standard of “actual malice” to inflictions of emotional distress, the Court emphasized that the case involved public figures.
This result does not reflect merely the judgment that (hopefully truthful) information about public issues is especially important, and, thus, the First Amendment should be especially protective in this context. Private figures receive more legal protection than public figures from injurious speech even if the speech involves reporting on vitally important public issues. In Gertz, the Court reversed the normative principles that I suggested above and said “private individuals are … more deserving of recovery,” (emphasis added) apparently because private persons have made no attempt to assume “an ‘influential role in ordering society.’ ” To this lack of involvement, the Court contrasted “[a]n individual who decides to seek governmental office,” who, the Court says, “must accept certain necessary consequences.”130
Gertz, 418 U.S. at 344–45.
In discussions of informational privacy, ‘private’ refers to a characteristic of the information. In contrast, defamation law introduces a somewhat different, dual conception of ‘private’. First, it raises a question about whether the information at issue involved “matters of public importance,” a characteristic similar to that which fuels informational privacy analyses. However, courts usually wish to avoid the appearance of ideological regulation of the marketplace of ideas. Courts should be, and often have been, loath to say that any content that a newspaper decided to publish was not about a matter of public importance.131
Dun & Bradstreet v. Greenmoss Builders, Inc., 472 US 749 (1985), is a rare exception where the Court, despite the dissent's devastating critique, characterized the content as private.
These two usages, privacy of the information and of the person, are connected in at least two ways. First, a person might want not to be a public figure precisely because she values control over her personal or private information, and she reasonably expects that it will be more difficult (even putting the law aside) to have this control if she becomes a public figure. A public figure loses her privacy both because the law protects it less and because people have a greater interest in knowing personal facts about her. Second, treatment of informational privacy as especially valuable may make for a culture that is more inclined to praise and reward private life, while the society that values and rewards active participation in the public sphere, although not ignoring the value of informational privacy, is likely to rank fame higher and informational privacy lower in its list of values. In any event, a positive valuation of civic participation and public excellence provides a reason for caution about maintaining too great an emphasis on informational privacy, except, as noted in Section V, when that emphasis relates specifically to preventing manipulation and control by governmental or corporate entities.
VII. The Value of Gossip
The common practice most at war with informational privacy is gossip, though the notion of gossip may understate the area of my concern. Warren and Brandeis may be right about gossip being the activity of the idle,132
Warren and Brandeis, supra note 1, at 196.
John Sabini and Maury Silver, Moralities of Everyday Life (Oxford: Oxford University Press, 1982), 92–94.
Nothing here turns on a more precise definition of gossip; Sabini and Silver devote most of a chapter to describing the concept. Id. at 89–106. I should emphasize that nothing about gossip implies anything about whether it is or is not accurate, but the gossip that I defend here is only gossip that is not subject to the critique of being knowingly or recklessly false.
This section draws heavily on Sabini and Silver, supra note 133; Gluckman, supra note 3.
Even if gossip occurs most often when people are idle, so does the preponderance of most people's other conversation, including political discussions. Though Warren and Brandeis did have some legitimate complaints, their association of gossip with idleness probably represented an attempt to belittle the activity. Nevertheless, the productive work done by gossip is multifarious, important, and often political in significant ways. Gossip provides a major mechanism for teaching social norms, often helping to show the norms' real, as opposed to perceived, weight. Psychologists John Sabini and Maury Silver observe that gossip is almost surely the way by which most people would unpack the ambiguities implicit in an abstract ethical injunction such as “Sex must be part of a meaningful relationship.”136
Sabini and Silver, supra note 133, at 100–101.
Gluckman, supra note 3.
Patricia Meyer Spacks, Gossip (New York: Alfred A. Knopf, 1985), 5. See also James C. Scott, Domination and the Arts of Resistance (New Haven, CT: Yale University Press, 1990), 142–43.
See, e.g., Spacks, supra note 138, at 46. She claims that “gossip gives voices to the dominated as well as the dominant.” Id. at 263.
Id. at 256.
Gross, supra note 2, at 125.
Id. at 124 (quoting Andrea Weiss, “'A Queer Feeling When I Look at You': Hollywood Stars and Lesbian Spectatorship in the 1930s,” in Christine Gledhill, ed., Stardom: Industry of Desire [London: Routledge, 1991], 283–99).
Formulating, debating, teaching, and changing the norms of social life may be the most important social function of gossip. Certainly, there is more to be said for gossip, even more than noting its apparently universal appeal, an appeal almost as universal as that of sex, which is often its subject. Gossip is an essential means of communication. “[S]ex and gossip alike comprise modes of intimate communication,” both of which are widely available to the dispossessed and marginalized as well as the powerful, are self-expressive, and are thereby “unpredictable and uncontrollable.”143
Spacks, supra note 138, at 40. The role of sex as a form of communication is the reason that I have given for providing First Amendment protection to consensual sex. See, e.g., C. Edwin Baker, “Op-Ed: First Amendment Protection for Gays,” New York Times, July 27, 1991, late–edition final, p. 23.
This essay places on center stage a conception of formal autonomy or liberty of choice that is often criticized as too individualistic.144
I have replied to this criticism when leveled against John Rawls in Baker, “Sandel on Rawls,” supra note 29.
Benjamin Constant, “The Liberty of the Ancients Compared with That of the Moderns,” (1819), in Biancamaria Fontana, ed. and trans., Political Writings (Cambridge: Cambridge University Press, 1988), 309–28. The “liberty of the ancients” consisted in “active and constant participation in collective power.” Id. at 316. My claim is that gossip has important similarities.
C. Edwin Baker, “Campaign Expenditures and Free Speech,” Harvard Civil Rights–Civil Liberties Law Review 33 (1998): 1–55.
First Amendment scholar Thomas Emerson listed four functions or values of speech that justified constitutional protection. His third function emphasized political participation in decision-making by all members of society. But this significant aspect of speech freedom, Emerson argued, “carries beyond the political realm. It embraces the right to participate in the building of the whole culture….”147
Thomas I. Emerson, The System of Freedom of Expression (New York: Random House, 1970), 7.
See Kenneth Karst, “Boundaries and Reasons: Freedom of Expression and the Subordination of Groups,” Illinois Law Review 1990 (1990): 95–149.
Baker, Human Liberty, supra note 5, chap. 4.
Beyond the observation that individual behavioral choices, including speech choices, create the social realm that in turn influences further individual choices, is the more “dialogic” question: How do people discuss and evaluate these choices? Possibly the most common and central means is gossip: “Sally did x; that's pretty bad! Or is it? She was faced with y. What do you think?” A negative verdict on the behavior may well sway (or punish) Sally and lead others to avoid (or hide) doing x. Alternatively, frequent gossip about y may result in the rejection or relaxation of the established social prohibition on x.
Thus, any proponent of protecting only or primarily political speech should have a hard time ruling out protection of gossip or other presentations of private information. In New York Times v. Sullivan (1964),150
376 U.S. 254 (1964).
Harry Kalven, Jr., A Worthy Tradition: Freedom of Speech in America (New York: Harper & Row, 1988), 63.
See Harry Kalven, Jr., “The New York Times Case: A Note on ‘The Central Meaning of the First Amendment,’ ” Supreme Court Review (1964): 191–221, at 221.
This may be a slight overstatement. Vince Blasi suggests that the checking function of the First Amendment argues for the propriety of giving challengers in electoral contests a right-to-reply such as the one struck down in Miami Herald Publishing Co. v. Tornillo, 418 U.S. 241 (1974), although he thought the checking function was not served by giving such a right to incumbents. Blasi, supra note 50. However, Kalven, who might have implied this view by suggesting that rejection of seditious libel is definitive of democracy, in fact saw the key to New York Times v. Sullivan in its adoption of Alexander Meiklejohn's political speech theory of the First Amendment. See Kalven, supra note 152.
Rosenbloom v. Metromedia, 403 U.S. 29 (1971); Gertz v. Robert Welch, Inc., 418 U.S. 323 (1974) (Brennan, J., dissenting).
A possible explanation is that absolutists do not credit “lies” as speech under the First Amendment. Absolutists never suggest, for example, that perjury or fraud constitute protected speech. If so, then they still might protect lies about political matters for prophylactic reasons. For example, they might distrust either state officials' or jurors' evaluation of the intentionality of falsehoods made during partisan debate. Falsehoods usually are knowing lies when made by the opposition but, when “we” make them, the falsity is always accidental and made in good faith! For this reason, absolutists might protect these, but only these, purportedly “knowing” falsehoods.
Another way to get to protection of gossip is to ask: Who is to decide what matters are of public importance? For the government, of which the courts are one branch, to define public importance detracts from the potential public sphere—a point Robert Post fondly makes.156
Post, Constitutional Domains, supra note 105, 119–78.
Anderson v. Fisher Broadcasting Companies, 300 Ore. 452 (1986).
Justice Linde's point illustrates not only how judgments about newsworthiness are controversial and ideological, but also how the capacity to contradict those judgments is central to the capacity to use speech to challenge the status quo, a point made more concretely in my earlier discussions of “outing.” In Linde's opinion, editors make the decision about “public importance” when they decide whether to include a story. Of course, their occasional dissent from established conventions158
Gross noted that in the debates on outing at the first convention of the National Lesbian and Gay Journalists Association in 1992, “it was easy to tell who was on which team by the uniforms: those in favor are most often clad in multiple earrings and sassy T-shirts. Those opposed wear suits and ties.” Gross, supra note 2, at 151. Still, one wonders which side was most committed to “official” journalistic norms that emphasize truth-telling and the public's right to know. In 1990, Michelangelo Signorile observed that to print a story “about a closeted gay man's woman friend as his lover … is applauded,” but “if you print the truth you are deemed ‘frightening and offensive.’ ” Id. at 60 (quoting Signorile).
Karst, supra note 148; Shiffrin, supra note 118.
Both informational privacy and the right to gossip can support meaningful autonomy. Both privacy and the ability to expose are resources or forms of power. Unlike material wealth, lawmaking power, or instruments of violence, the direct or “natural” connection to the person of both privacy and the ability to expose it necessarily results in their comparatively equal distribution. Thus, those who are committed to a democratic or egalitarian distribution of power and capabilities would have reason to value legal recognition of both. This point, however, does not explain why, in cases of conflict, the speech claim—the right to gossip—should trump the interest in privacy. Note, however, that absent (illegal) coercion, people usually can choose, initially, not to disclose information about themselves. The priority of speech freedom does not deprive people of this right. They and others can continue to avoid (or limit) disclosure or, alternatively, can choose to speak. Respect for autonomy should be seen as requiring that the choice between these alternatives be left to the individual, which is a voluntarist method of determining what information will be made available to a public. Similarly, the political as well as the autonomy-based claim on behalf of speech is that speech is a power that people should have and be able to use to try to change the world. The debates and practices about “outing” illustrate not that people will necessarily make the right choices—for what is “right” is inherently contestable—but that they will take their responsibilities in the exercise of this speech power seriously.
I would be remiss if I did not note the underside of gossip, although neither my comments here nor my earlier defense of gossip does either side justice. Gossip is often unfair in two ways and undesirable in a third. First, gossip is frequently inaccurate. Moreover, in contrast to falsehoods published in the media, these inaccuracies can be particularly hard for the unfairly treated person to discover and, thus, to refute. Nevertheless, the argument here only defends “true” gossip. Inaccurate gossip does not so much violate informational privacy as create a defamation type of injury. Application of Sullivan's actual malice standard to the inaccuracies in reports about a family's response to being taken hostage (a so-called “false light” privacy case) follows easily from either an autonomy notion of individual speech freedom or a standard conception of the media's speech rights.160
Time, Inc. v. Hill, 385 U.S. 374 (1967).
Second, even when true, gossip can be unfair. It can treat as important something about a person that should not, at least not now, be relevant for the person's public persona or for most other evaluative purposes. Gossip can be a true report, but out of context. Most unfairly, even true gossip can provide a hard-to-challenge opportunity for prejudice to operate, or can stimulate or reinforce prejudices, often to the distinct disadvantage of members of vulnerable groups.
Third, gossip can unfairly divert attention from what should be important about a person or divert attention from society's real problems. Is gossip the opiate of the masses? Personally, I find sensationalism, a part of the news at least since Roman times,161
John D. Stevens, Sensationalism and the New York Press (New York: Columbia University Press, 1991), 6–7.
Although much can be said—and too much was said—about the Bill Clinton/Monica Lewinsky affair, a common observation was that the drawn-out exposé fascinated and engaged even many people who reported that the affair and the cover-up did not determine, possibly was not even relevant to, their view of whether President Clinton should have remained in office. I am hesitant about the observation that I shall next make, partly because I did not follow either scandal too closely. Still, I wonder about differences between the impact of revelations about Clinton's sexual activities and the earlier sexual harassment allegations leveled against Clarence Thomas at the time of his nomination to the Supreme Court. Arguably, discussion of allegations against Thomas helped put sexual harassment on the cultural map, a desirable result. Arguably, too, discussion of Clinton's escapades constituted an implicit debate about how society ought to react when a public official's personal life intersects with his public role, possibly a debate moving us closer to toleration. If there were these positive effects, then privacy-invading gossip and media sensationalism may have led to a useful discursive evaluation of norms. Admittedly, I continue to believe that both educational practice and social norms should encourage more interest in substantive public policy issues, which elites like me consider central. And I think governmental media policy should encourage the development of media institutions with less sensationalist, more serious (as well as more interesting and culturally playful) foci. Still, the affirmative reasons favoring gossip suggest that these choices are appropriately both contextual and contestable. This contextual inconclusiveness provides further pragmatic grounds, despite gossip's underside, to reject legal restriction of so-called non-newsworthy, utterly offensive disclosures of private information.
VIII. Conclusion
Keeping all or some of the power to expose or to use private information out of the hands of the “disciplinary agencies” of government and profit-oriented enterprises is often desirable. The press, however, must be exempted from this conclusion since it constitutes a crucial part of the public sphere. The propriety of legislative limitations on the power of government and market enterprises is similar to keeping from these institutions the power to be secretive. Limits on market entities' use of private information and Privacy Acts limiting government have the same legitimacy as the Freedom of Information Act, open meeting laws, and modern demands for transparency. Government's capacity to have easy, cheap, and not really consensual access to database information containing detailed personal profiles is leading us down a slippery slope at the end of which lies government decrees ordering people to wear an identifying star. Before reaching that dire end, government intrusiveness contributes to the creation of a timorous, docile population that tries to avoid any behavior that might, say, fit the profile of a terrorist. Surely individual autonomy must receive some protection from both governmental and corporate infringements of informational privacy.
In contrast, sometimes privacy may need to be breached if social progress is to be achieved. As Randall Kennedy's stories of “passing” and Larry Gross's account of “outing” illustrate, both privacy and exposure are forms of power that can be used by members of marginalized groups to pursue their cultural, political, and personal goals. (Of course, both privacy and exposure are also used by dominant groups to help maintain their favored norms.) Gossip teaches and maintains, but also helps to change norms of social relations, as it serves to reinforce the identity and cohesion of the oppressed.
My normative claim has been that speech freedom, including freedom to expose any private information that a person knows, is an aspect of formal autonomy that government must respect if it is to remain legitimate in its pursuit of conditions that make meaningful autonomy possible. My pragmatic claim has been that society and, especially, its oppressed segments benefit by leaving speech choices—whether to choose privacy or exposure—largely free of legal limitation. Speech freedom is a relatively egalitarian power that people can and will claim and use.
Footnotes
An earlier version of this essay was presented to Martha Nussbaum and David Strauss's Law and Philosophy Workshop at the University of Chicago Law School in 2000. I thank the participants in that workshop, fellow contributors to this volume, Michael Madow, Diane Zimmerman, and the editors of Social Philosophy & Policy for helpful comments, questions, and encouragement.
You have
Access
- 11
- Cited by