Introduction
Fear can be a powerful silencer. Speakers may be thwarted not only by direct force but also when they check themselves because they anticipate adverse consequences. Some assessment of costs and benefits is involved whenever anyone decides to communicate, of course. That is normal and actually valuable. Yet acute anxiety, caused by the realistic prospect of violence or other grave harm, differs from ordinary consequential reasoning, even if both result in silence.
Today, speakers seem to be hesitating with concerning frequency. Their reticence is understandable, because disagreement and its consequences have become severe in certain settings. Acute fear of speaking has affected those on the right and on the left, though not perhaps in the same way or to the same degree. Consider an example at Columbia University. Reportedly, a truck with a billboard bearing the words “Columbia’s Leading Antisemites,” alongside the names and faces of students and faculty, appeared in Morningside Heights and drove slowly around campus. The truck targeted Columbia affiliates who allegedly had signed a statement of solidarity with Palestinians in Gaza. The stunt was reported to be a project of a conservative media group, which also published a website that listed students and faculty members and was updated regularly. People named on the website were said to be members of various campus groups that were engaged in protests against Israel’s military actions. The website asked readers to send messages to Columbia’s board of trustees urging them to “take a stand” against “these hateful individuals.” The group also purchased domain names that corresponded to the actual names of several students and faculty on the list, and it planned to send the truck to the targets’ homes. Two Columbia law students who were named had job offers withdrawn by law firms, according to news outlets covering the story.
Although debates over Israel’s military actions in Gaza are particularly fierce, they are not unique. Fear of speaking seems to have intensified as political conflict has escalated in the United States and elsewhere. As differences have deepened and political identities have tribalized, speakers noticeably have withdrawn, nervous about the possible results. Some antagonism is inherent in healthy democratic discourse, and it is not cause for regret. Criticism is a feature of deliberation, and it is valued by First Amendment traditions. But anticipation of systematic violence is something of a different order, at least arguably. Intimidation like that can degrade democratic discourse and political cooperation—or that at least is the worry with respect to freedom of expression.
Technological changes have contributed to the climate of anxiety, of course. Although some of the activity at Columbia was analog—it took the form of a truck circling campus—other aspects have leveraged the efficiency of digital media. Today, any utterance can be preserved and disseminated, instantly and cheaply. Anonymity reduces accountability for the intimidation; though anonymity also can serve freedom of expression, its possible piercing can disincentivize debate. A feeling of surveillance can result—the sense that something you say can provoke reprisal that is utterly devastating.
Although this is hard to document, the university classroom itself shows signs of being impoverished by the effects of systemic fear. On questions of political controversy, students appear reluctant to volunteer views that even conceivably could expose them to retribution or stigmatization. Faced with a choice between the exploration of ideas that entails the risk of retribution and the safety of silence, many students opt for the latter, it seems. And that is true of at least some students on the right and left.
Professor Helen Norton’s insightful essay for this symposium explores a related dynamic surrounding the law of stalking. For purposes of this Essay, what her sophisticated analysis illustrates is a more general phenomenon, namely that speech generating fear can itself have a silencing effect. Although the constitutional debate surrounding stalking manifestly concerns the expressive rights of the stalker, its latent lesson is that there are speech interests on both sides—that the victims of harassment and intimidation themselves can become muzzled. Understanding that dynamic complicates any consideration of the First Amendment of fear.
This short Essay seeks to make modest progress on understanding and approaching the relationship between extreme fear and freedom of speech. Part I draws inspiration from Judith Shklar, who famously built a liberal political theory designed to shield citizens from fear. Though her theory was concerned solely with government oppression, and though she built up only a minimal kind of political morality, it can be complicated and complemented to include private sources of intimidation and, relatedly, to comprehend a positive government obligation to ensure the basic social and material conditions for a healthy speech environment. Part II then applies this political conception to the problem of doxing. State statutes regulating doxing already exist, though they mostly have not yet been tested for adherence to the First Amendment. Whether and how those laws comport with the right to freedom of expression is a complicated issue, one that must include consideration of the expressive interests of the targets or victims of doxing, as well as of the perpetrators, and it must involve the social and economic power relationships that constitute and distort the expressive environment.
I. The Politics of Fear
A place to start is with perhaps the most prominent political theorist of fear, Judith Shklar. Reading her later work today is bracing—it elicits a jolt of recognition. At the most basic level, Shklar seeks to organize a conception of liberalism around a summum malum, namely “cruelty and the fear it inspires, and the very fear of fear itself.” Cruelty of the sort that concerns her is systematic; it is not haphazard but instead it entails the “deliberate infliction of physical, and secondarily emotional, pain” as an exercise of power by those in positions of strength against those in positions of weakness.
Terror and acute anxiety are bad, supremely bad, partly because they interfere with freedom, understood as the ability to direct one’s own life. Here is a key passage: “Every adult should be able to make as many effective decisions without fear or favor about as many aspects of her or his life as is compatible with the like freedom of every other adult. That belief is the original and only defensible meaning of liberalism.” She explains that liberalism aims to lift the burden of political anxiety from the shoulders of adults, who then can order their lives according to their own beliefs, wants, and needs.
Shklar’s chain of reasoning—from basic security, to the fear of its violation, to the fear of fear itself, and then to the connection of security to the exercise of basic freedom—resonates. She connects power differentials and their abuse to human emotion, and she connects psychological security to the ability of individuals to function as citizens in a political community. A commitment to personal security is political insofar as it concerns a necessary condition of the community—of the project of democratic cooperation.
For theorists of free speech, it is a short step from Shklar’s liberalism to the realization that speakers cannot be free if they are fearful of physical violence or of power exercised against them in ways that threaten their safety. Is this silencing systematic? Shklar is not simply concerned with insecurity, again, but with insecurity that issues from power differentials and is patterned. Silencing that results from fear today could possibly be considered systematic, in a sense. Or you could say that some silencing is systematic in this way. In a radically polarized political climate, the content and viewpoint that risks retribution is foreseeable—and it is precisely this predictability that creates the conditions for censorship. If speakers did not know what positions or politics would endanger themselves or their families, they would not be able to avoid them. But because such viewpoints are foreseeable, and to the extent they are, they can be silenced.
Shklar is focused on public power, which she identifies with government action. And her sharp distinction between public and private power is recognizably liberal. To some degree, Shklar’s work nevertheless remains relevant here, even strictly construed, because it applies to certain violations of freedom of expression that are instigated by government actors. In this pattern, a public official identifies a political enemy with the knowledge and expectation that followers will harass and intimidate the targeted person, terrifying them into submission and silence. So, although the proximate harm is caused by private persons, the coordination and instigation come from politicians.
Yet taking a broader view, Shklar’s focus on state action neglects instances in which private actors threaten potential speakers without any apparent or actual coordination by government figures. Because the effect is often the same, and because the topics involved may well be matters of public concern, her neglect of nonpublic exercises of power to systematically silence people limits the usefulness of her insights in today’s speech environment.
It is true that Shklar acknowledges the relevance of some social and institutional conditions for the exercise of individual freedom. For example, she emphasizes some differences between her theory and Isaiah Berlin’s negative liberty, which is otherwise quite similar. She sees the importance of protecting not just negative liberty as such, but also the conditions that make its exercise possible, and she realizes that in this way negative liberty is a necessary but not sufficient condition for personal freedom. “No door is open in a political order in which public and private intimidation prevail,” she says, and therefore it is important to identify the “institutional characteristics of a relatively free regime,” including mechanisms for the dispersal of power—social as well as strictly political power.
In a prominent critique, Sam Moyn portrays the liberalism of fear as an instance of what he calls Cold War liberalism. Chastened and traumatized by the wars and totalitarianisms of the twentieth century, Shklar pessimistically “dropped any radical expectations of improvement” and retreated to defending minimal pluralism against the persistent threat of violence. Abandoning the aspirations of her earlier work, Shklar adopted a “ ‘survivalist’ approach to political theory,” one resigned to hope only for “damage control.” This form of liberalism offered few resources to resist the rise of libertarianism and neoliberalism, though it was distinct from both.
We need not accept the limits of the liberalism of fear. To the degree that Shklar herself is focused solely on precarity caused by government policing, we can expand her insights and apply them to nongovernmental sources of insecurity. Political and constitutional theory can assimilate the insight that freedom of speech, like other basic liberties, cannot be merely formal but must be real for the actual human beings living in historically specific social situations. It is essential for the meaningful exercise of freedom of speech for certain essential social and economic conditions to obtain. Some of these conditions are egalitarian, and some are sufficientarian, as argued in other work. Here, the specific point is that speakers do not have a meaningful ability to express themselves freely if they are subject to fundamental physical and psychological insecurity. This is not just the healthy fear of avoidable pain, as Shklar emphasizes, but the systematic circumstance of political polarization and power exertion that predictably suppresses particular viewpoints.
The First Amendment of fear qualifies as a political conception because it attends to social and economic power and locates solutions, ultimately if not exclusively, in the state. Although the closest causes of anxiety today often are other citizens, leveraging digital media and other technologies of terror, they are unlikely to be stopped solely by private means. Government has an obligation to ensure people’s security, and their sense of security, as a condition of meaningful political participation as cogovernors in a democracy. Whether that obligation itself has constitutional force is an interesting but different question, and regardless government ought to be constitutionally permitted to pursue the structural conditions for real
people’s meaningful exercise of the fundamental right to freedom of expression.
Is the political commitment to freedom from fear possible to implement in law, given existing First Amendment doctrine? There is some doubt. Part II explores that question in the context of state statutes criminalizing or otherwise regulating doxing.
II. Doxing and Silencing
In the little space that is left in this Essay, let’s consider attempts by law to guard against a particular cause of fear, namely doxing. These attempts face serious challenges under the First Amendment, since doxing typically constitutes speech that does not necessarily or obviously fall within an existing category of unprotected expression, and since it is regulated based on its content. From the perspective of the First Amendment of fear, this legal circumstance could be seen as a matter of regret, insofar as doxing itself can have a powerful silencing effect on those it targets. There may be no way to assimilate that insight into the existing structure of free speech doctrine. Yet this Part cautiously explores one possible pathway.
Doxing can be understood in several ways. The Oxford English Dictionary defines it as “[t]he action or process of searching for and publishing private or identifying information about a particular individual on the internet, typically with malicious intent.” At root, and colloquially, the practice involves publishing identifying information about someone in order to facilitate harm of that person by third parties. Yet not every element of this understanding is essential; arguably, the trucks that circled campuses displaying the names and images of students constituted doxing even though they did not involve the internet. And additional elements may be necessary, such as a particular level of mens rea, or a requirement that the information not already be publicly available, or that the target is not a public official.
California has an influential criminal statute that includes a specification of what constitutes doxing. It prohibits “electronically distribut[ing]” through various means “personal identifying information” of another person “with intent to place another person in reasonable fear for his or her safety, or the safety of the other person’s immediate family” and “for the purpose of imminently causing that other person unwanted physical contact, injury, or harassment . . . by a third party.” Among the exceptions is the distribution of information with the target’s consent.
Though the California criminal statute does not use the term doxing, it has been understood to regulate that activity. A newer state law provides a civil cause of action for “doxing,” which it defines using much the same language as the criminal provision. Legislative history shows that state lawmakers intended to provide a civil cause of action for doxing, referencing the criminal statute. Also, Stanford University’s “Anti-Doxxing Policy” appears to be modeled on the California statutes and uses substantially the same definition of the prohibited activity.
Notable here is California’s use of the term “fear” to indicate the harm that it seeks to protect against. Apparently, the state believes that disabling anxiety on the part of victims is serious enough to warrant a criminal prohibition. And the statute recognizes that the electronic distribution of personal information has the power to generate a specific kind of harm, and to an extraordinary degree. Yet the statute also limits itself to fear of “physical contact, injury, or harassment,” not just anticipation of political criticism or even social ostracism, without more. Though there is considerable variation among state doxing statutes on this and other questions, the California approach is among the most straightforward. Also notable is that the California statute does not exempt news reporting, perhaps because lawmakers reasoned that the intent requirement would not be satisfied and therefore an explicit exemption was unnecessary.
Could the California statute withstand a First Amendment challenge? The difficulty, of course, is that the distribution of personal identifying information could itself be seen to be expression, or expressive conduct. And if the Speech Clause is implicated, then plausibly it requires strict scrutiny of the California statute, which regulates on the basis of content. After all, the statute only prohibits a specific kind of speech, namely the distribution of certain identifying information using a particular medium. And whether the statute is narrowly tailored to a compelling government interest is unclear.
Under one theory, the California statute might be constitutional because it regulates a type of threat. On this approach, the regulated content would fall within a traditional category of unprotected speech. Compare California’s threat statute. It criminalizes threatening another person with “death or great bodily injury” with the specific intent that the statement be taken as a threat and under circumstances that convey a specific and immediate danger so that the target “reasonably . . . [is] in sustained fear for his or her own safety or for his or her immediate family’s safety.” Threat statutes do typically guard against a particular kind of “fear.” Conceivably, California’s doxing statute regulates a subset of threat—a specific kind of statement, made through a particular medium, that causes the target to feel insecure, particularly with regard to safety. Both statutes have an intent requirement and they both apply only to reasonable fears and imminent dangers. Neither requires the violence to eventuate because both recognize that the fear itself is harmful.
To be sure, there are differences that may be important. Crucially, the doxing statute shields against statements that risk harm not by the speaker, but by a third party. For this reason, it could be reasonably argued that incitement is the category of unprotected speech that is more closely analogous to doxing. But incitement does not centrally involve fear, which seems essential to the harm of doxing.
If it is correct that doxing can count as a type of threat, then its prohibition could be seen as having a kind of derivative constitutionality. Threats constitute a category of unprotected speech, under established Supreme Court doctrine. If California has criminalized a species of threat, then its doxing statute could survive under that First Amendment doctrine. Importantly, the state would have to require at least a recklessness level of mens rea to avoid chilling protected activity. But because the doxing statute requires an “intent” and a “purpose,” it could be construed to clear that bar.
A possible doctrinal objection is that the Supreme Court protected an early form of doxing in NAACP v. Claiborne Hardware. That decision is mainly known for its holding that a civil rights boycott of white-owned businesses was constitutionally protected. Less well known is the Court’s holding that recording the names of customers who violated the boycott, reading those names at meetings, and publishing the names in a newspaper could not be punished. If boycott enforcers from the NAACP intended to threaten these customers by causing them to fear that they would be identified to third parties who would harm them, then their activity of recording names looks like doxing. And the Court held that the organizers “admittedly sought to persuade others to join the boycott through social pressure and the ‘threat’ of social ostracism” but that “[s]peech does not lose its protected character . . . simply because it may embarrass others or coerce them into action.” Violence did occur in the course of the boycott, but still, the Court held that publishing names of boycott violators could not be punished. This could be understood as precedent for protecting doxing.
Yet the Court also reaffirmed that both violence and the threat of violence are unprotected and could be punished. What the First Amendment requires is precision, not protection of threatening speech.
When [violence or a threat] occurs in the context of constitutionally protected activity . . . ‘precision of regulation’ is demanded. Specifically, the presence of activity protected by the First Amendment imposes restraints on the grounds that may give rise to damages liability and on the persons who may be held accountable for those damages.
In the context of doxing, this may provide good reason to require an elevated level of mens rea. The lesson may be that while speech that identifies perceived wrongdoers cannot be punished, particularly when it addresses matters of public concern, speech that identifies individuals for the specific purpose of eliciting violence can be prohibited because criminalization with that elevated level of intent constitutes “precision of regulation.” Whether the “doxing truck” at Columbia could be criminalized under this approach would depend, in part, on whether it was operated with the requisite level of intent.
Another possible objection is that punishing only doxing that is specifically intended to elicit fear of violence, and that qualifies as a type of threat, does not match people’s common understanding of doxing. On this objection, it is undesirable for there to be a mismatch between the social meaning of a term like doxing and the legal prohibition that seeks to address the harm. That is reasonable. If the truck at Columbia were not motivated by the requisite intent, such that it did not constitute doxing as a legal matter, some would view that as a fault of the statute. One response is that not protecting against doxing at all, or doing so only through existing statutes designed for other purposes, also fails to match people’s reasonable expectation that the law should address serious harms. Another response is that many people seemingly do think that doxing involves “malicious intent” in the words of the Oxford definition, if not necessarily intent relative to violence. So maybe the mismatch is not so egregious.
Another mismatch is that the proposal here only prohibits doxing that creates a fear of violence, not also harassment. That is because it is unclear whether a statute that guarded against fear of harassment, without more, would fall within the category of unprotected speech for true threats. So here too, there is a potential mismatch between the proposal and colloquial understandings of doxing—an awkwardness that may simply be a cost of fitting this particular protection against disabling fear into the existing constitutional doctrine.
This proposal would bring the regulation of doxing within the unprotected category of threatening speech. Still, it is unfortunate that First Amendment law is being understood to require strict scrutiny of all speech regulations that fall outside a recognized category of unprotected speech, such as threats. As Genevieve Lakier and Evelyne Douek have argued, the Supreme Court has moved away from its traditional practice of evaluating speech regulations with respect to First Amendment values and competing considerations. In this context, as noted, doxing regulation could serve important free speech values, especially by protecting victims against the sort of disabling hesitation that effectively silences them. So it could be said that in at least some cases there are expressive interests on both sides of a statute like California’s. A full consideration of values would take that symmetry into account.
Yet, for now, unless and until there is a meaningful change in the ideology or composition of the Roberts Court, the binary approach to speech doctrine must be taken as a fixed feature of constitutional law. And under that approach, a doxing statute like this one can best survive if it is understood to regulate a subset of threatening speech.
Conclusion
From the perspective of the First Amendment of fear, government ought to be at least permitted, if not required, to safeguard the conditions under which people have a meaningful and not just formal ability to participate in democratic discourse and otherwise express themselves freely. One obstacle to that freedom is the systematic apprehension of speaking on certain topics. Unfreedom of this kind is worth protecting against. Government has the ability and the responsibility to ensure expressive security—not freedom from fear of criticism or rebuke, but freedom from systematic fear of violence, at the very least. Arguably, this kind of safety is essential to expression. Doxing legislation may be one example of government protection that, despite facing constitutional hurdles, can be crafted so that it does more to promote than to frustrate First Amendment imperatives.
The First Amendment of Fear
Introduction
Fear can be a powerful silencer. Speakers may be thwarted not only by direct force but also when they check themselves because they anticipate adverse consequences. Some assessment of costs and benefits is involved whenever anyone decides to communicate, of course. That is normal and actually valuable. Yet acute anxiety, caused by the realistic prospect of violence or other grave harm, differs from ordinary consequential reasoning, even if both result in silence.
Today, speakers seem to be hesitating with concerning frequency. Their reticence is understandable, because disagreement and its consequences have become severe in certain settings. Acute fear of speaking has affected those on the right and on the left, though not perhaps in the same way or to the same degree. Consider an example at Columbia University. Reportedly, a truck with a billboard bearing the words “Columbia’s Leading Antisemites,” alongside the names and faces of students and faculty, appeared in Morningside Heights and drove slowly around campus.1Esha Karam, ‘Doxxing Truck’ Displaying Names and Faces of Affiliates It Calls ‘Antisemites’ Comes to Columbia, Colum. Spectator (Oct. 25, 2023, 11:45 AM), https://www.columbiaspectator.com/news/2023/10/25/doxxing-truck-displaying-names-and-faces-of-affiliates-it-calls-antisemites-comes-to-columbia [https://perma.cc/37K2-QSXM]. For reporting on a similar incident, see Anemona Hartocollis, After Writing an Anti-Israel Letter, Harvard Students Are Doxxed, N.Y. Times (Oct. 18, 2023, 5:03 AM), https://www.nytimes.com/2023/10/18/us/harvard-students-israel-hamas-doxxing.html [https://web.archive.org/web/20231018090959/https://www.nytimes.com/2023/10/18/us/harvard-students-israel-hamas-doxxing.html]. The truck targeted Columbia affiliates who allegedly had signed a statement of solidarity with Palestinians in Gaza. The stunt was reported to be a project of a conservative media group, which also published a website that listed students and faculty members and was updated regularly.2Karam, supra note 1. People named on the website were said to be members of various campus groups that were engaged in protests against Israel’s military actions. The website asked readers to send messages to Columbia’s board of trustees urging them to “take a stand” against “these hateful individuals.”3Id. The group also purchased domain names that corresponded to the actual names of several students and faculty on the list, and it planned to send the truck to the targets’ homes. Two Columbia law students who were named had job offers withdrawn by law firms, according to news outlets covering the story.4Id. For reporting on self-censorship surrounding the military actions in Gaza, see Emily Nayyer, Surveys Reveal Rising Student and Faculty Concern About Censorship, Self-Censorship Post-October 7, FIRE (July 12, 2024), https://www.thefire.org/news/surveys-reveal-rising-student-and-faculty-concern-about-censorship-self-censorship-post [https://perma.cc/5RJW-WXFC] (reporting the results of a survey that found increased censorship and self-censorship among students concerning the war in Gaza); Shibley Telhami & Marc Lynch, Middle East Scholar Barometer #7 (May 23-June 6, 2024), https://criticalissues.umd.edu/sites/criticalissues.umd.edu/files/November%202023%20MESB%20Results.pdf [https://perma.cc/8YKA-2GSS] (reporting the results of a poll conducted by the University of Maryland and George Washington University, finding that seventy-five percent of scholars of the Middle East responded “Yes” when asked, “Do you feel the need to self-censor when speaking about the Palestinian-Israeli issue in an academic or professional capacity?”).
Although debates over Israel’s military actions in Gaza are particularly fierce, they are not unique. Fear of speaking seems to have intensified as political conflict has escalated in the United States and elsewhere.5Again, the phenomenon probably is not limited to one political camp, however asymmetric it may be. It is possible to imagine a situation where a public identification, accompanied by a charge of racism or bigotry, could be intended to elicit violence by third parties. The 2020 Central Park incident was meaningfully different, both because the intent of the person who posted the video did not seem to be malicious, and because the speech the video depicted was not on a matter of public concern. But a variation on that incident could be invented that would constitute doxing. Olivia Land, NYC’s ‘Central Park Karen’: I still live in hiding three years after viral video, N.Y. Post (Nov. 7, 2023), https://nypost.com/2023/11/07/metro/central-park-karen-still-hiding-3-years-after-viral-video [https://perma.cc/5LUY-VDNN]. As differences have deepened and political identities have tribalized, speakers noticeably have withdrawn, nervous about the possible results. Some antagonism is inherent in healthy democratic discourse, and it is not cause for regret. Criticism is a feature of deliberation, and it is valued by First Amendment traditions. But anticipation of systematic violence is something of a different order, at least arguably. Intimidation like that can degrade democratic discourse and political cooperation—or that at least is the worry with respect to freedom of expression.
Technological changes have contributed to the climate of anxiety, of course. Although some of the activity at Columbia was analog—it took the form of a truck circling campus—other aspects have leveraged the efficiency of digital media. Today, any utterance can be preserved and disseminated, instantly and cheaply. Anonymity reduces accountability for the intimidation; though anonymity also can serve freedom of expression,6See McIntyre v. Ohio Elections Comm’n, 514 U.S. 334, 342 (1995) (describing a First Amendment tradition of protecting anonymous speech in the United States). its possible piercing can disincentivize debate. A feeling of surveillance can result—the sense that something you say can provoke reprisal that is utterly devastating.
Although this is hard to document, the university classroom itself shows signs of being impoverished by the effects of systemic fear. On questions of political controversy, students appear reluctant to volunteer views that even conceivably could expose them to retribution or stigmatization. Faced with a choice between the exploration of ideas that entails the risk of retribution and the safety of silence, many students opt for the latter, it seems. And that is true of at least some students on the right and left.
Professor Helen Norton’s insightful essay for this symposium explores a related dynamic surrounding the law of stalking.7Helen Norton, Fear and Free Speech, 98 S. Cal. L. Rev. 1351 (2025). For purposes of this Essay, what her sophisticated analysis illustrates is a more general phenomenon, namely that speech generating fear can itself have a silencing effect. Although the constitutional debate surrounding stalking manifestly concerns the expressive rights of the stalker, its latent lesson is that there are speech interests on both sides—that the victims of harassment and intimidation themselves can become muzzled.8To get a sense of that debate, see the majority and dissenting opinions in Counterman v. Colorado, 143 S. Ct. 2106 (2023), and Genevieve Lakier & Evelyn Douek, The First Amendment Problem of Stalking: Counterman, Stevens, and the Limits of History and Tradition, 113 Calif. L. Rev. 149, 195–203 (2025) (endorsing Justice Sotomayor’s concurring opinion in Counterman). Understanding that dynamic complicates any consideration of the First Amendment of fear.
This short Essay seeks to make modest progress on understanding and approaching the relationship between extreme fear and freedom of speech. Part I draws inspiration from Judith Shklar, who famously built a liberal political theory designed to shield citizens from fear.9A representative work is Judith N. Shklar, The Liberalism of Fear, in Liberalism and the Moral Life 21, 29 (Nancy L. Rosenblum ed., 1989). Though her theory was concerned solely with government oppression, and though she built up only a minimal kind of political morality, it can be complicated and complemented to include private sources of intimidation and, relatedly, to comprehend a positive government obligation to ensure the basic social and material conditions for a healthy speech environment. Part II then applies this political conception to the problem of doxing. State statutes regulating doxing already exist, though they mostly have not yet been tested for adherence to the First Amendment.10For examples of state doxing statutes, see infra notes 34, 37. For cases testing doxing statutes, see Kratovil v. City of New Brunswick, 261 N.J. 1 (2025) (holding that a New Jersey law that shielded an official from publication of their exact home address was narrowly tailored to a compelling state interest); Atlas Data Privacy Corporation v. We Inform, LLC, 758 F. Supp. 3d 322 (D.N.J. 2024) (also upholding the New Jersey statute); DeHart v. Tofte, 326 Or. App. 720 (Ct. App. Ore. 2023) (declining to apply a doxing law on expressive grounds); Publius v. Boyer–Vine, 237 F. Supp. 3d 997 (E.D. Cal. 2017) (ruling unconstitutional a statute that protected against the publication of certain identifying information about certain officials, upon request); Brayshaw v. City of Tallahassee, 709 F. Supp. 2d 1244 (N.D. Fla. 2010) (overturning on constitutional grounds a conviction for publishing identifying information about a police officer); Ostergren v. Cuccinelli, 615 F.3d 263 (4th Cir. 2010) (holding that a free-speech advocate’s publication of social security numbers was protected); Sheehan v. Gregoire, 272 F. Supp. 2d 1135 (W.D. Wash. 2003) (invalidating a state statute prohibiting publication of identifying information about certain public officials). Whether and how those laws comport with the right to freedom of expression is a complicated issue, one that must include consideration of the expressive interests of the targets or victims of doxing, as well as of the perpetrators, and it must involve the social and economic power relationships that constitute and distort the expressive environment.
I. The Politics of Fear
A place to start is with perhaps the most prominent political theorist of fear, Judith Shklar. Reading her later work today is bracing—it elicits a jolt of recognition. At the most basic level, Shklar seeks to organize a conception of liberalism around a summum malum, namely “cruelty and the fear it inspires, and the very fear of fear itself.”11Shklar, supra note 9, at 29. Cruelty of the sort that concerns her is systematic; it is not haphazard but instead it entails the “deliberate infliction of physical, and secondarily emotional, pain” as an exercise of power by those in positions of strength against those in positions of weakness.12Id.
Terror and acute anxiety are bad, supremely bad, partly because they interfere with freedom, understood as the ability to direct one’s own life. Here is a key passage: “Every adult should be able to make as many effective decisions without fear or favor about as many aspects of her or his life as is compatible with the like freedom of every other adult. That belief is the original and only defensible meaning of liberalism.”13Id. at 21. She explains that liberalism aims to lift the burden of political anxiety from the shoulders of adults, who then can order their lives according to their own beliefs, wants, and needs.14Id. at 31 (explaining that liberalism restricts itself to politics and seeks “to lift the burden of fear and favor from the shoulders of adult women and men, who can then conduct their lives in accordance with their own beliefs and preferences”).
Shklar’s chain of reasoning—from basic security, to the fear of its violation, to the fear of fear itself, and then to the connection of security to the exercise of basic freedom—resonates. She connects power differentials and their abuse to human emotion, and she connects psychological security to the ability of individuals to function as citizens in a political community.15Id. at 29 (“Systematic fear is the condition that makes freedom impossible . . . .”). A commitment to personal security is political insofar as it concerns a necessary condition of the community—of the project of democratic cooperation.16Id. (“[W]hen we think politically, we are afraid not only for ourselves but for our fellow citizens as well. We fear a society of fearful people.”).
For theorists of free speech, it is a short step from Shklar’s liberalism to the realization that speakers cannot be free if they are fearful of physical violence or of power exercised against them in ways that threaten their safety. Is this silencing systematic? Shklar is not simply concerned with insecurity, again, but with insecurity that issues from power differentials and is patterned. Silencing that results from fear today could possibly be considered systematic, in a sense. Or you could say that some silencing is systematic in this way. In a radically polarized political climate, the content and viewpoint that risks retribution is foreseeable—and it is precisely this predictability that creates the conditions for censorship. If speakers did not know what positions or politics would endanger themselves or their families, they would not be able to avoid them. But because such viewpoints are foreseeable, and to the extent they are, they can be silenced.
Shklar is focused on public power, which she identifies with government action.17Id. at 21 (“[W]hile the sources of social oppression are indeed numerous, none has the deadly effect of those who, as the agents of the modern state, have unique resources of physical might and persuasion at their disposal.”). And her sharp distinction between public and private power is recognizably liberal.18Cf. Corey Robin, Fear: The History of a Political Idea 14–15 (2004) (describing and complicating Shklar’s exclusive focus on government, and showing how that focus relates to her liberalism). To some degree, Shklar’s work nevertheless remains relevant here, even strictly construed, because it applies to certain violations of freedom of expression that are instigated by government actors. In this pattern, a public official identifies a political enemy with the knowledge and expectation that followers will harass and intimidate the targeted person, terrifying them into submission and silence. So, although the proximate harm is caused by private persons, the coordination and instigation come from politicians.
Yet taking a broader view, Shklar’s focus on state action neglects instances in which private actors threaten potential speakers without any apparent or actual coordination by government figures. Because the effect is often the same, and because the topics involved may well be matters of public concern, her neglect of nonpublic exercises of power to systematically silence people limits the usefulness of her insights in today’s speech environment.
It is true that Shklar acknowledges the relevance of some social and institutional conditions for the exercise of individual freedom. For example, she emphasizes some differences between her theory and Isaiah Berlin’s negative liberty, which is otherwise quite similar. She sees the importance of protecting not just negative liberty as such, but also the conditions that make its exercise possible, and she realizes that in this way negative liberty is a necessary but not sufficient condition for personal freedom.19Shklar, supra note 9, at 28. “No door is open in a political order in which public and private intimidation prevail,” she says, and therefore it is important to identify the “institutional characteristics of a relatively free regime,” including mechanisms for the dispersal of power—social as well as strictly political power.20Id. (emphasis added); see id. at 30–31 (embracing property rights as a mechanism for the dispersal of power).
In a prominent critique, Sam Moyn portrays the liberalism of fear as an instance of what he calls Cold War liberalism. Chastened and traumatized by the wars and totalitarianisms of the twentieth century, Shklar pessimistically “dropped any radical expectations of improvement” and retreated to defending minimal pluralism against the persistent threat of violence.21Samuel Moyn, Before—and Beyond—the Liberalism of Fear, in Between Utopia and Realism: The Political Thought of Judith N. Shklar 24, 24 (Samantha Ashenden & Andreas Hess eds., 2019). Abandoning the aspirations of her earlier work, Shklar adopted a “ ‘survivalist’ approach to political theory,” one resigned to hope only for “damage control.”22Id. at 25; Shklar, supra note 9, at 27 (“We say ‘never again,’ but somewhere someone is being tortured right now, and acute fear has again become the most common form of social control. To this the horror of modern warfare must be added as a reminder. The liberalism of fear is a response to these undeniable actualities, and it therefore concentrates on damage control.”). This form of liberalism offered few resources to resist the rise of libertarianism and neoliberalism, though it was distinct from both.23Cf. Daniel McAteer, A Conversation with Samuel Moyn: The Cold War and the Canon of Liberalism, Univ. of Oxford: Ctr. for Intell. Hist. (Apr. 1, 2022), https://intellectualhistory.web.ox.ac.uk/article/a-conversation-with-samuel-moyn-the-cold-war-and-the-canon-of-liberalism [https://perma.cc/597Y-CEJ9] (arguing that in “the Cold War” period, “you get a much more libertarian framing of liberalism”).
We need not accept the limits of the liberalism of fear. To the degree that Shklar herself is focused solely on precarity caused by government policing, we can expand her insights and apply them to nongovernmental sources of insecurity. Political and constitutional theory can assimilate the insight that freedom of speech, like other basic liberties, cannot be merely formal but must be real for the actual human beings living in historically specific social situations.24Nelson Tebbe, A Democratic Political Economy for the First Amendment, 105 Cornell L. Rev. 959, 974–80 (2020). For a new, important account of why legal rights often are formal, not real, see Jeremy Kessler, Law and Historical Materialism, 74 Duke L.J. 1523, 1527–1538 (2025). It is essential for the meaningful exercise of freedom of speech for certain essential social and economic conditions to obtain. Some of these conditions are egalitarian, and some are sufficientarian, as argued in other work.25Tebbe, supra note 24, at 967. Here, the specific point is that speakers do not have a meaningful ability to express themselves freely if they are subject to fundamental physical and psychological insecurity. This is not just the healthy fear of avoidable pain, as Shklar emphasizes, but the systematic circumstance of political polarization and power exertion that predictably suppresses particular viewpoints.
The First Amendment of fear qualifies as a political conception because it attends to social and economic power and locates solutions, ultimately if not exclusively, in the state. Although the closest causes of anxiety today often are other citizens, leveraging digital media and other technologies of terror, they are unlikely to be stopped solely by private means. Government has an obligation to ensure people’s security, and their sense of security, as a condition of meaningful political participation as cogovernors in a democracy. Whether that obligation itself has constitutional force is an interesting but different question, and regardless government ought to be constitutionally permitted to pursue the structural conditions for real
people’s meaningful exercise of the fundamental right to freedom of expression.
Is the political commitment to freedom from fear possible to implement in law, given existing First Amendment doctrine? There is some doubt. Part II explores that question in the context of state statutes criminalizing or otherwise regulating doxing.
II. Doxing and Silencing
In the little space that is left in this Essay, let’s consider attempts by law to guard against a particular cause of fear, namely doxing. These attempts face serious challenges under the First Amendment, since doxing typically constitutes speech that does not necessarily or obviously fall within an existing category of unprotected expression, and since it is regulated based on its content.26For decisions considering the constitutionality of doxing laws, see supra note 10. From the perspective of the First Amendment of fear, this legal circumstance could be seen as a matter of regret, insofar as doxing itself can have a powerful silencing effect on those it targets.27Although doxing nearly always constitutes speech, it does not always single out its targets because of their expression. Even when it does not, however, it can exert a silencing effect. There may be no way to assimilate that insight into the existing structure of free speech doctrine. Yet this Part cautiously explores one possible pathway.
Doxing can be understood in several ways. The Oxford English Dictionary defines it as “[t]he action or process of searching for and publishing private or identifying information about a particular individual on the internet, typically with malicious intent.”28Doxing, Oxford Eng. Dictionary, https://doi.org/10.1093/OED/6624632723 [https://perma.cc/V3TS-EJ3G]. At root, and colloquially, the practice involves publishing identifying information about someone in order to facilitate harm of that person by third parties. Yet not every element of this understanding is essential; arguably, the trucks that circled campuses displaying the names and images of students constituted doxing even though they did not involve the internet.29See, e.g., Karam, supra note 1 (using the phrase “doxxing truck”). And additional elements may be necessary, such as a particular level of mens rea, or a requirement that the information not already be publicly available, or that the target is not a public official.
California has an influential criminal statute that includes a specification of what constitutes doxing. It prohibits “electronically distribut[ing]” through various means “personal identifying information” of another person “with intent to place another person in reasonable fear for his or her safety, or the safety of the other person’s immediate family” and “for the purpose of imminently causing that other person unwanted physical contact, injury, or harassment . . . by a third party.”30Cal. Penal Code § 653.2(a) (West 2024). The statute also, but separately, prohibits distributing “an electronic message of a harassing nature about another person, which would be likely to incite or produce that unlawful action.” Id. Among the exceptions is the distribution of information with the target’s consent.31See id.
Though the California criminal statute does not use the term doxing, it has been understood to regulate that activity. A newer state law provides a civil cause of action for “doxing,” which it defines using much the same language as the criminal provision.32Cal. Civ. Code § 1708.89(a)(1) (West 2024). Here is the language:
“Doxes” means an act when a person, with intent to place another person in reasonable fear for their safety, or the safety of the other person’s immediate family, by means of an electronic communication device, and without consent of the other person, and for the purpose of imminently causing that other person unwanted physical contact, injury, or harassment, by a third party, electronically distributes, publishes, emails, hyperlinks, or makes available for downloading, personal identifying information, including, but not limited to, a digital image of another person, or an electronic message of a harassing nature about another person, which would be likely to incite or produce that unlawful action. Legislative history shows that state lawmakers intended to provide a civil cause of action for doxing, referencing the criminal statute.33A.B. 1979, 2023–2024, Reg. Sess. (Cal. 2024). Also, Stanford University’s “Anti-Doxxing Policy” appears to be modeled on the California statutes and uses substantially the same definition of the prohibited activity.34Anti-Doxxing Policy, Stanford Univ. Bull., https://bulletin.stanford.edu/academic-polices/student-conduct-rights/anti-doxxing [https://perma.cc/L5D5-4UE6]; see David Cremins, Defending the Public Quad: Doxxing, Campus Speech Policies, and the First Amendment, 76 Stan. L. Rev. 1813, 1821 (2024) (noting that Stanford’s anti-doxing provision was modeled on California’s law, and that it passed the Faculty Senate with “near-unanimous support”).
Notable here is California’s use of the term “fear” to indicate the harm that it seeks to protect against. Apparently, the state believes that disabling anxiety on the part of victims is serious enough to warrant a criminal prohibition. And the statute recognizes that the electronic distribution of personal information has the power to generate a specific kind of harm, and to an extraordinary degree. Yet the statute also limits itself to fear of “physical contact, injury, or harassment,” not just anticipation of political criticism or even social ostracism, without more. Though there is considerable variation among state doxing statutes on this and other questions, the California approach is among the most straightforward.35For an example of a state statute that regulates doxing by name, see Wash. Rev. Code § 4.24.792 (2024). For an example of a law that does not use the term and is narrower in that it only applies to the disclosure of telephone numbers and home addresses, see Tex. Penal Code Ann. § 42.074 (West 2023). Also notable is that the California statute does not exempt news reporting, perhaps because lawmakers reasoned that the intent requirement would not be satisfied and therefore an explicit exemption was unnecessary.
Could the California statute withstand a First Amendment challenge?36The statute has been applied by courts, none of which have reached the constitutional question. Dziubla v. Piazza, 273 Cal. Rptr. 3d 297, 306–07 (2020); People v. Shivers, 186 Cal. Rptr. 3d 352, 356–358 (2015); see Cremins, supra note 34, at 1819 (“Since its passage in 2008, Section 653.2 has apparently never been challenged on First Amendment grounds . . . .” (footnote omitted)). The difficulty, of course, is that the distribution of personal identifying information could itself be seen to be expression, or expressive conduct. And if the Speech Clause is implicated, then plausibly it requires strict scrutiny of the California statute, which regulates on the basis of content.37See Cremins, supra note 34, at 1823, 1824 n.51 (noting that doxing rules single out speech on the basis of content). After all, the statute only prohibits a specific kind of speech, namely the distribution of certain identifying information using a particular medium. And whether the statute is narrowly tailored to a compelling government interest is unclear.38For examples of decisions finding that strict scrutiny was satisfied, see Kratovil, 261 N.J. at 26 (“we hold that when it enacted Daniel’s Law, the Legislature carefully calibrated the statute to serve a state interest of the highest order by the least restrictive means”), and Atlas Data Privacy Corp., 758 F. Supp. 3d at 337.
Under one theory, the California statute might be constitutional because it regulates a type of threat. On this approach, the regulated content would fall within a traditional category of unprotected speech. Compare California’s threat statute. It criminalizes threatening another person with “death or great bodily injury” with the specific intent that the statement be taken as a threat and under circumstances that convey a specific and immediate danger so that the target “reasonably . . . [is] in sustained fear for his or her own safety or for his or her immediate family’s safety.”39The relevant section of the statute reads, in full:
Any person who willfully threatens to commit a crime which will result in death or great bodily injury to another person, with the specific intent that the statement, made verbally, in writing, or by means of an electronic communication device, is to be taken as a threat, even if there is no intent of actually carrying it out, which, on its face and under the circumstances in which it is made, is so unequivocal, unconditional, immediate, and specific as to convey to the person threatened, a gravity of purpose and an immediate prospect of execution of the threat, and thereby causes that person reasonably to be in sustained fear for his or her own safety or for his or her immediate family’s safety, shall be punished by imprisonment in the county jail not to exceed one year, or by imprisonment in the state prison.
Cal. Penal Code § 422(a) (West 2024). Threat statutes do typically guard against a particular kind of “fear.”40Counterman v. Colorado, 143 S. Ct. 2106, 2114 (2023) (“True threats subject individuals to ‘fear of violence’ and to the many kinds of ‘disruption that fear engenders.’ ” (quoting Virginia v. Black, 538 U.S. 343, 360 (2003))). Conceivably, California’s doxing statute regulates a subset of threat—a specific kind of statement, made through a particular medium, that causes the target to feel insecure, particularly with regard to safety. Both statutes have an intent requirement and they both apply only to reasonable fears and imminent dangers. Neither requires the violence to eventuate because both recognize that the fear itself is harmful.
To be sure, there are differences that may be important. Crucially, the doxing statute shields against statements that risk harm not by the speaker, but by a third party. For this reason, it could be reasonably argued that incitement is the category of unprotected speech that is more closely analogous to doxing.41Under this alternative, a doxing conviction would have to meet the Brandenburg test, according to which “the constitutional guarantees of free speech and free press do not permit a State to forbid or proscribe advocacy of the use of force or of law violation except where such advocacy is directed to inciting or producing imminent lawless action and is likely to incite or produce such action.” Brandenburg v. Ohio, 395 U.S. 444, 447 (1969). But incitement does not centrally involve fear, which seems essential to the harm of doxing.
If it is correct that doxing can count as a type of threat, then its prohibition could be seen as having a kind of derivative constitutionality. Threats constitute a category of unprotected speech, under established Supreme Court doctrine.42Counterman, 143 S. Ct. at 2113 (“True threats of violence, everyone agrees, lie outside the bounds of the First Amendment’s protection.”). If California has criminalized a species of threat, then its doxing statute could survive under that First Amendment doctrine. Importantly, the state would have to require at least a recklessness level of mens rea to avoid chilling protected activity.43Id. But because the doxing statute requires an “intent” and a “purpose,” it could be construed to clear that bar.44Cal. Penal Code § 653.2(a) (West 2024).
A possible doctrinal objection is that the Supreme Court protected an early form of doxing in NAACP v. Claiborne Hardware.45NAACP v. Claiborne Hardware Co., 458 U.S. 886 (1982). That decision is mainly known for its holding that a civil rights boycott of white-owned businesses was constitutionally protected.46Id. at 914. Less well known is the Court’s holding that recording the names of customers who violated the boycott, reading those names at meetings, and publishing the names in a newspaper could not be punished.47Id. at 925–26. If boycott enforcers from the NAACP intended to threaten these customers by causing them to fear that they would be identified to third parties who would harm them, then their activity of recording names looks like doxing. And the Court held that the organizers “admittedly sought to persuade others to join the boycott through social pressure and the ‘threat’ of social ostracism” but that “[s]peech does not lose its protected character . . . simply because it may embarrass others or coerce them into action.”48Id. at 910–11. Violence did occur in the course of the boycott, but still, the Court held that publishing names of boycott violators could not be punished. This could be understood as precedent for protecting doxing.
Yet the Court also reaffirmed that both violence and the threat of violence are unprotected and could be punished.49Id. at 916 (“[T]here is no question that acts of violence occurred. No federal rule of law restricts a State from imposing tort liability for business losses that are caused by violence and by threats of violence.”). What the First Amendment requires is precision, not protection of threatening speech.
When [violence or a threat] occurs in the context of constitutionally protected activity . . . ‘precision of regulation’ is demanded. Specifically, the presence of activity protected by the First Amendment imposes restraints on the grounds that may give rise to damages liability and on the persons who may be held accountable for those damages.50Id. at 916–17 (quoting NAACP v. Button, 371 U.S. 415, 438 (1963)).
In the context of doxing, this may provide good reason to require an elevated level of mens rea.51Cf. id. at 919 (noting that, in the context of those who associate with others who commit violence, “to punish association with such a group, there must be clear proof that a defendant specifically intends to accomplish the aims of the organization by resort to violence”) (alterations and internal quotation marks omitted). The lesson may be that while speech that identifies perceived wrongdoers cannot be punished, particularly when it addresses matters of public concern, speech that identifies individuals for the specific purpose of eliciting violence can be prohibited because criminalization with that elevated level of intent constitutes “precision of regulation.” Whether the “doxing truck” at Columbia could be criminalized under this approach would depend, in part, on whether it was operated with the requisite level of intent.
Another possible objection is that punishing only doxing that is specifically intended to elicit fear of violence, and that qualifies as a type of threat, does not match people’s common understanding of doxing. On this objection, it is undesirable for there to be a mismatch between the social meaning of a term like doxing and the legal prohibition that seeks to address the harm. That is reasonable. If the truck at Columbia were not motivated by the requisite intent, such that it did not constitute doxing as a legal matter, some would view that as a fault of the statute. One response is that not protecting against doxing at all, or doing so only through existing statutes designed for other purposes, also fails to match people’s reasonable expectation that the law should address serious harms. Another response is that many people seemingly do think that doxing involves “malicious intent” in the words of the Oxford definition, if not necessarily intent relative to violence.52See supra text accompanying note 28. So maybe the mismatch is not so egregious.
Another mismatch is that the proposal here only prohibits doxing that creates a fear of violence, not also harassment.53Cremins argues that doxing statutes should only punish in which there is a threat of physical contact or injury. Cremins, supra note 34, at 1827–29, 1832. That is because it is unclear whether a statute that guarded against fear of harassment, without more, would fall within the category of unprotected speech for true threats.54Note that the California threat statute only protects against fear of “death or great bodily injury” or lack of safety of self or family. See supra text accompanying note 39. So here too, there is a potential mismatch between the proposal and colloquial understandings of doxing—an awkwardness that may simply be a cost of fitting this particular protection against disabling fear into the existing constitutional doctrine.
This proposal would bring the regulation of doxing within the unprotected category of threatening speech. Still, it is unfortunate that First Amendment law is being understood to require strict scrutiny of all speech regulations that fall outside a recognized category of unprotected speech, such as threats. As Genevieve Lakier and Evelyne Douek have argued, the Supreme Court has moved away from its traditional practice of evaluating speech regulations with respect to First Amendment values and competing considerations.55Lakier & Douek, supra note 8, at 216 (taking no position on the constitutionality of laws against doxing, inter alia, but arguing that the issue should be confronted “head-on,” and not through the “distorted kaleidoscope” of current doctrine, with its categories of unprotected speech and its assumption that all speech regulation outside them will draw strict scrutiny); see id. at 217 (arguing that “the First Amendment provides more latitude to legislatures to protect individuals from this kind of fear-inducing speech than a superficial reading of the Court’s recent precedents implies”). In this context, as noted, doxing regulation could serve important free speech values, especially by protecting victims against the sort of disabling hesitation that effectively silences them. So it could be said that in at least some cases there are expressive interests on both sides of a statute like California’s. A full consideration of values would take that symmetry into account.
Yet, for now, unless and until there is a meaningful change in the ideology or composition of the Roberts Court, the binary approach to speech doctrine must be taken as a fixed feature of constitutional law. And under that approach, a doxing statute like this one can best survive if it is understood to regulate a subset of threatening speech.
Conclusion
From the perspective of the First Amendment of fear, government ought to be at least permitted, if not required, to safeguard the conditions under which people have a meaningful and not just formal ability to participate in democratic discourse and otherwise express themselves freely. One obstacle to that freedom is the systematic apprehension of speaking on certain topics. Unfreedom of this kind is worth protecting against. Government has the ability and the responsibility to ensure expressive security—not freedom from fear of criticism or rebuke, but freedom from systematic fear of violence, at the very least. Arguably, this kind of safety is essential to expression. Doxing legislation may be one example of government protection that, despite facing constitutional hurdles, can be crafted so that it does more to promote than to frustrate First Amendment imperatives.
98 S. Cal. L. Rev. 1413
Download
* Jane M.G. Foster Professor of Law, Cornell Law School. Thanks to Evelyn Douek, Eugene Volokh, participants in the Southern California Law Review symposium, and the staff of the law review for comments and suggestions. The author served as an observer for a committee of the Uniform Law Commission on doxing, but the arguments here do not reflect the views of the committee.
Like this article?