Must free speech be harmful? That is, must the freedom of speech protect harmful speech? Popular discourse in the United States often assumes that it must.1 Discussions about hate speech or false speech frame harm as the price we pay for freedom. Meanwhile, several distinguished scholars have also asserted that the right of freedom of speech must include protection for harmful speech.2 They claim that, in order to be either conceptually or normatively significant, any plausible speech right must protect harmful speech.3 In other words, freedom of speech must include harm. If it does not, it is not doing its job.

This assertion has the distinction of being at once an old saw, a sophisticated philosophical argument, and a fairly stunning claim. A right that must encompass harmful conduct? Why would people say that harm protection is a necessary feature of a right? And are they correct?

Protection of harmful conduct is not a necessary feature of any right, including a free speech right. Considering the relationship between free speech rights and harm clarifies the structure of rights and shows that, conceptually speaking, there is no reason to conclude that free speech must protect harmful conduct in order to be meaningful. That is our choice, and one that few other cultures make, at least in such strong terms.4 This is not to say that protection for harmful conduct is never justified. It is just not inevitable in the ways commonly asserted by judges, scholars, and the citizenry. Here, I will identify the arguments made in support of the view that free speech must include harm, argue that protection for harmful conduct is not inherent in the structure of rights generally, and argue finally that special rights such as freedom of speech need not include protection for harmful conduct.

I.     Perspectives on Rights and Harm

Various important thinkers have asserted that freedom of speech must protect activity that risks causing harm to other people. In various ways, they have asserted such protection as a necessary criterion for any plausible free speech right. Some identify particular activities and conclude that any plausible free speech principle must protect them, whether because of their relationship to democratic self-governance, or to autonomy, or to some other value. Some of the identified activities involve risks of harm. For example, lawyers and scholars commonly conclude that freedom of speech must protect dangerous or incendiary political speech.5 Such speech carries obvious risks of physical and other types of harm, risks of the kind that make other conduct regulable. Despite these risks, many stipulate that freedom of speech must protect incendiary political speech—not necessarily to the level of immunity, but to a higher degree than similarly risky non-speech activity. These types of arguments have a long doctrinal pedigree, back to Justice Holmes’s shift from treating subversive speech like any other subversive conduct to reformulating the “clear and present” danger test to protect speech to a higher degree.6 The result of such reasoning is that some speech that carries a risk of harm—a high enough risk to justify regulation of non-speech conduct—will be protected. When that risk manifests in harm, the speaker will be insulated from liability. This is one way in which the freedom of speech can require protection of harmful conduct.

Others begin by developing conceptual criteria for a free speech right and conclude that freedom of speech must include protection for harmful conduct.7 One version of this argument contends that free speech that does not protect harm is not very meaningful. For example, Professor Thomas Scanlon, in discussing immunity for harm-causing speech, says that it is “the existence of such cases which makes freedom of expression a significant doctrine.”8 Similarly, in considering whether to “accept the principle that speech may be restricted when it causes harm to others,” Frederick Schauer concludes: “[y]et then what [would be] the point of a principle of free speech?”9 On this view, if freedom of speech did not protect harmful conduct, it would not be a significant right.

Kent Greenawalt makes a similar argument and places it in the context of a larger theory of rights. Greenawalt begins with a minimal principle of liberty, which protects all harmless conduct. Within that context, he develops criteria for a free speech right:

As far as speech is concerned, the minimal principle of liberty establishes that the government should not interfere with communication that has no potential for harm. To be significant, a principle of freedom of speech must go beyond this, positing constraints on the regulation of speech which are more robust.10

On this view, a free speech right that did not protect harmful conduct would not just be insignificant; it would be superfluous.

Like Greenawalt, Ronald Dworkin reasons from a general theory of rights to the conclusion that free speech must protect harmful speech.11 Dworkin defines free speech as a “special right,” one that pertains to a limited set of activities, in contrast to a “general right” that covers all activities.12 Mill’s liberty principle, which protects all activity to the extent that it does not cause harm, is a prime example of a general liberty right.13 Dworkin explains:

Freedom of speech is a special right: government may not infringe that special freedom unless it has what American lawyers have come to call a “compelling” justification. Speakers may not be censored even when what they say may well have bad consequences for other people . . . The right to free speech can be abridged only in emergencies: only to prevent, again in a phrase beloved of American lawyers, a clear and present—and, we might add, grave—danger.14

Dworkin’s framework thus contemplates that free speech will protect harmful conduct, because only a compelling justification can support a restriction of speech, and a compelling justification generally requires something like a clear, present, and grave danger. Risks that do not meet this bar go unregulated, and protected speech therefore may cause harm to others.15

These scholars range in their approaches. Dworkin appears to make protection of harmful conduct an inevitable feature of a special right such as freedom of speech. The others do not suggest that a free speech right covering only harmless conduct is conceptually impossible, just that such a right would be essentially meaningless.16 What all these claims have in common, however, is the idea that freedom of speech should protect harmful conduct. If a free speech right does not, it is either a conceptual failure or insignificant.

II.   On Rights and Harm

Must rights protect harmful conduct? If so, then freedom of speech must include protection of harmful speech. The short answer is, no, rights need not protect harmful conduct. A right may exist for reasons having nothing to do with protection for harmful conduct, and limits on governmental action can constrain the government without affording direct protection for harmful conduct. A Millian general liberty right to engage in harmless conduct is an example: it expressly does not extend to harmful conduct but is a right nonetheless.17

Yet perhaps it is not so simple. Much depends on what “harm” means. Every right has a correlative of some kind that places limitations on other actors: a Hohfeldian claim-right correlates with a duty toward the rightsholder; an immunity correlates with a disability.18 When the right in question is a right held against the government, the existence of the right inevitably limits the government’s scope of action in some way. Perhaps it prevents the government from regulating certain conduct at all; perhaps, more modestly, it requires the government to give certain types of justifications or to withstand heightened scrutiny for its actions.19 One could argue that any limitation on governmental action involves harm because it reduces the government’s ability to achieve an interest. If this is the case, then not only do free speech rights protect harmful conduct: all rights protect harmful conduct.

I raise this possibility to set it aside. Limitations on the government’s means of pursuing an interest may reduce its ability to achieve that interest. Such limitations on government, however, do not inherently count as “harms” in the sense of setbacks to interests. Thus, for example, a free speech right that protects people’s ability to post political signs on their property will frustrate the government’s pursuit of its interest in reducing visual clutter.20 In this case, the freedom of speech reduces the ability of government to pursue a legitimate objective, but it does not involve harm in the sense of setbacks to interests. It seems implausible, then, to say that all rights cause harm because they limit the government’s ability to pursue its objectives.

But what if any limitation on the government’s ability to pursue its interests necessarily posed harm to third parties, even if indirectly? If this were so, then rights generally would cause harm, because rights would place limitations on the government, which would inevitably, though indirectly, pose harm to others. For example, Fourth Amendment rights against unreasonable search and seizure and Fifth Amendment rights against self-incrimination both limit how the government may pursue its interest in law enforcement. This could well mean that some private individuals are harmed, for example by becoming victims of someone who was not convicted. If every governmental interest ultimately worked to protect third parties from harm, then all rights would indirectly risk harm to others.

The problem is that the premise is not true: not all government interests ultimately protect third parties from harm. The government’s asserted interest in reducing visual clutter is one example.21 So are many other interests that the government has advanced in the First Amendment context: administrative interest in accurate recordkeeping,22 interest in protecting the flag as a symbol of national unity,23 and interest in “maintaining the parks in the heart of our Capital in an attractive and intact condition.”24 Protecting such interests bears no necessary relation to preventing harm in the sense of setbacks to individual interests, and constraining the government’s pursuit of these interests bears no necessary relation to harm creation.

Thus, harmful conduct is not a necessary feature of all rights. Even if one were to conclude, contrary to my argument above, that the limitation on pursuit of the government’s interests constitutes “harm,” or that all rights do pose indirect risks of harm to third parties, commentators’ contentions about free speech are still different. Those who claim free speech must protect harmful conduct have in mind conduct that poses obvious, direct risks of harm, mostly physical harm, to individuals. The typical examples—incendiary speech,25 false speech,26 advocacy of law violation,27 speech that poses risks short of a clear and present danger28 —all contemplate risk of physical harm to other people. As Dworkin puts it, what speakers say “may well have bad consequences for other people” and yet still is protected by freedom of speech.29 The basic claim is that the freedom of speech protects conduct that poses direct harm to third parties and that this feature is crucial to the significance of the right.

III.  On Free Speech and Harm

When articulated in this way—that free speech must affirmatively protect conduct that poses direct and obvious risks of harm to other people—the contention starts to sound unusual. Protection of harmful conduct is not a predicate for other rights, such as voting rights, rights of sexual intimacy, or rights of religious exercise. Some of these rights, as implemented, may lead to harm, and some conceptions of them may have harm as an unavoidable incidental effect. No one, however, starts from the premise that these rights must protect harmful conduct or else they have no meaning. It is a curious feature of free speech rights that people regularly describe them this way.30

Is there something special about a free speech right that requires protection of harmful conduct in order for it to be significant? It seems not. It is possible to conceive of something properly deemed a speech right that does not protect harmful conduct and that would still be meaningful. For example, one could frame speech as a positive right, which would require the government to provide speech opportunities. A government could be obligated to provide speech opportunities for citizens without protecting their speech when it causes harm. Or imagine formulating a free speech right in a primary or secondary school setting. One valuable part of an education is for students to learn how to formulate ideas and arguments and to learn how to listen to others. A classroom might be obligated to provide these opportunities to students. At the same time, however, students might have no right to be disruptive of the educational endeavor or derogatory or cruel toward each other. Both the civic formulation and the classroom formulation involve rights properly designated as free speech rights, and those rights seem significant enough to identify and discuss. Perhaps we might think that a free speech right in civil society should go farther than this, and certainly American free speech jurisprudence does.31 But it need not in order to be worth talking about. Indeed, many nations around the world subject free speech rights to proportionality and balancing tests that come much closer to limiting free speech with some kind of Millian harm principle, and yet those nations still find free speech worth singling out.32

One argument for the harm-protecting view is that protecting harmful conduct is necessary to differentiate free speech from a general liberty right. Greenawalt, for example, contends that, in order to be significant, a free speech right must protect conduct that would otherwise not be protected by a general liberty principle.33 For Greenawalt, this translates into protection for harmful speech. Dworkin also seems to take for granted that speech rights must afford more protection than general rights, and for him, like Greenawalt, this seems to translate into protection for harmful conduct.34 Note that, if this is true, it is true of all special rights, not just speech rights: all of them would have to be differentiated from general liberty through protection of harmful conduct.

This is not true, however, for two reasons. First, even if we accept Dworkin and Greenawalt’s premise that special rights have to serve some function above and beyond general liberty, that could take forms other than protection for harmful conduct. Again, a positive right is an example: an affirmative obligation to provide speech opportunities would do something beyond what a general liberty right does, without protecting harmful conduct. If it is important to Greenawalt and Dworkin that a special right do something more than a general liberty right, there are other things for it to do besides protect harmful conduct.

Second, and more importantly, special rights need not do more than general rights. If a right has a special relationship to a particular value, it may be important to single it out in order to identify that special relationship, regardless of whether the right ultimately affords additional protection. For example, one might think that certain sexual activities are not harmful—say, the decision to use contraception or to engage in fully consensual sexual activity between adults. If such activity is not harmful, then it will be protected by a general liberty right. But we might still want to recognize a special right of sexual autonomy, rather than lumping this conduct in with all the other harmless activity covered by general liberty. We might think a right of sexual autonomy should be recognized for historical or pragmatic reasons—that is, because the state has a distinctive history of attempting to regulate this particular harmless conduct.35 Alternatively, we might think sexual autonomy bears a special relationship to a larger underlying value—personal autonomy, self-development, or the like. For either reason, we could rightfully conclude that our taxonomy of rights should identify a special right of sexual autonomy. Recognition of such a right is appropriate, regardless of whether it protects any more conduct than would otherwise be protected. Recognition of the special right makes clear that violating that right infringes freedom in more than one way: it implicates the values underlying the general right of liberty and those underlying a special right of, say, sexual autonomy. Thus, regulating consensual sexual activity is wrongful in a different way from regulating, say, harmless activities such as hopscotch or handball.

We might also think about rights of religious free exercise. Imagine a view that religious observance is distinctive from other forms of activity (a commonly held view that I am simply stipulating here). Imagine a right of religious free exercise that protects religious observance, but not when a particular practice poses harm to other people. Suppose further that the state will not make positive provisions for free exercise opportunities because of establishment concerns. Incidentally, this is a fairly accurate description of American jurisprudence. Regulation that incidentally burdens religion is permitted when justified by legitimate government interests; preventing harm to third parties is an obviously legitimate governmental interest.36 In addition, the Establishment Clause has placed limits on statutory religious accommodations that would burden third parties.37

Against the backdrop of a right of general liberty, this free exercise right accomplishes exactly nothing: it protects no more activity than general liberty already protects, and it imposes no additional positive obligations on the state. If, however, we think that free exercise of religion is importantly distinctive, then it deserves to be singled out for special recognition, even if that recognition does not result in additional protection. If we think that prohibiting harmless religious activity is more wrongful than, or wrongful in a different way from, prohibiting harmless activity generally—for example, that banning the wearing of a yarmulke while playing handball is wrongful in a different way from banning the playing of handball—then we have reason to recognize a right of religious free exercise. Doing so articulates that targeting harmless religious free exercise implicates another value besides whatever value supports the right of general liberty. Regulation that targets harmless religious practice is impermissible not only because it fails the general liberty principle but also because—given the distinctiveness of religion—targeting harmless religious practice is a particularly wrongful thing for the government to do.

The fact that two rights protect the same activity to the same degree does not make one of the rights superfluous. The fact that harmless speech—or harmless religious exercise or sexual activity—would be protected by a general liberty right does not make the special right meaningless and unnecessary. The question is whether we have a reason to identify the activity as distinctive—to distinguish it from the other activities covered by the general right. Whether a special right is appropriate will depend upon the distinctiveness of the activity, not upon whether the special right would ultimately afford more robust protection. This approach makes clear that, even if there were no general liberty right (or other broad principle encompassing the conduct), the narrower right would still exist. This approach also pushes us to identify all the reasons we have for recognizing rights. This seems like a salutary feature for a conception of rights.38

IV.  Conclusion

Must free speech protect harmful conduct? No—not in order for the right to qualify as a special right, and not for it to be significant enough to discuss. Yes, the more robust a free speech right is, the more attention it is likely to demand, but that is not the same as being significant. Free speech can be a sufficiently significant doctrine without requiring protection for harmful conduct. Deductions to the contrary are not correct.

That leaves us with the inductive approach—the conclusion of some jurists and scholars that certain forms of harmful conduct, such as incendiary speech, must be protected by any plausible speech right.39 This approach does not begin with the premise that free speech must protect harmful conduct in order to be meaningful. It looks for speech activities that require protection and finds that some include harmful conduct. This is reasoning not from the structure of rights but from the particular types of speech that seem normatively worthy of protection. Some of these intuitions might bear out: the proposition that the government should not be able to deny people benefits based on their political affiliation—finally embraced by the Supreme Court after many years of struggle—may seem an important commitment, even if it does have the tendency to risk certain types of harm.40 Other harm-causing propositions may be less defensible. The point is that these are discrete normative judgments, which should be continually questioned and defended, and which should not find refuge in general statements that free speech inevitably involves harm.

Our free speech tradition involves strong protections, including protections for various forms of harmful conduct. My point here is not to say that all of those protections are unjustified; it is simply to say that they are not inevitable. Despite our societal insistence to the contrary, a right of free speech need not include protection for harmful conduct. Nothing about the structure of rights generally compels that, and nothing about the structure of free speech rights does either. Our insistence that free speech must protect harmful conduct is a product of our weighing of speech and harm, weighing that deserves the same level of “uninhibited, robust, and wide-open” inquiry that the First Amendment prescribes for other subjects.41

  • 1See, e.g., Garrett Epps, Free Speech Isn’t Free, The Atlantic (Feb. 7, 2014), ht‌‌t‌p‌s‌:‌/‌/‌w‌w‌w‌.‌t‌h‌‌‌e‌‌‌a‌‌tl‌a‌ [‌3‌J‌J‌]‌.‌
  • 2[1]See infra Part I.
  • 3Id.
  • 4Compare Féret v. Belgium, App. No. 15615/07, Eur. Ct. H.R. (July 16, 2009), h‌t‌t‌p‌:‌/‌/‌h‌u‌do‌c‌.‌e‌c‌h‌r‌ [] (holding that there was no violation of Article 10 (freedom of expression) of the European Convention on Human Rights for convicting appellant based on distributed leaflets publicly inciting discrimination and hatred) with Brandenburg v. Ohio, 395 U.S. 444 (1969) (prohibiting laws restricting incitement to hatred or violence unless the words or expression at issue create a serious and imminent risk of harmful conduct). See also Frederick Schauer, The Exceptional First Amendment, in American Exceptionalism and Human Rights 29 (Michael Ignatieff ed., 2005).
  • 5See, e.g., Vincent Blasi, The Pathological Perspective and the First Amendment, 85 Colum. L. Rev. 449, 449–50 (1985) (“[T]he overriding objective at all times should be to equip the first amendment to do maximum service in those historical periods when intolerance of unorthodox ideas is most prevalent and when governments are most able and most likely to stifle dissent systematically.”); see also Mary Ellen Gale and Nadine Strossen, “The Real ACLU,” 1 Yale J. L. & Feminism, 161, 173 (1989) (“If free speech is to have meaning, it must encompass ‘freedom for the thought that we hate,’ freedom for the idea, opinion or expression that is unpopular, divergent, degraded, derided, dangerous, or even pornographic or obscene.”); John Rawls, Political Liberalism 348–56 (1993) (arguing that political speech merits protection unless it poses a clear and present danger to the democratic order); Seana Valentine Shiffrin, A Thinker-Based Approach to Freedom of Speech, 27 Const. Comment. 283, 285 (2011) (“[A] decent regime of freedom of speech must provide a principled and strong form of protection for political speech and, in particular, for incendiary speech.”).
  • 6Compare Frohwerk v. United States, 249 U.S. 204 (1919), Debs v. United States, 249 U.S. 211 (1919), and Schenck v. United States, 249 U.S. 47 (1919), with Abrams v. United States, 250 U.S. 616 (1919) (Holmes, J., dissenting).
  • 7See, e.g., Thomas Scanlon, A Theory of Freedom of Expression, 1 Phil. & Pub. Aff. 204, 208 (1972); Ronald Dworkin, Religion Without God 131 (2013).
  • 8Scanlon, supra note  NOTEREF _Ref32948353 \h  \* MERGEFORMAT 7 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300350033000000 , at 204.
  • 9See Frederick Schauer, Must Speech Be Special?, 78 Nw. U. L. Rev. 1284, 1294 (1983); see also Frederick Schauer, Free Speech on Tuesdays, 34 L. & Phil. 119, 135 (2015) (stating that a plausible free speech right must protect harmful conduct, while reserving judgment on whether such a right can ultimately be successfully delineated); Frederick Schauer, The Phenomenology of Speech and Harm, 103 Ethics 635, 652 (concluding that, if a free speech right principle exists, it necessarily entails protection for activity that causes harm).
  • 10Kent Greenawalt, Speech, Crime, and the Uses of Language 10 (1989).
  • 11See Dworkin, supra note  NOTEREF _Ref32948353 \h  \* MERGEFORMAT 7 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300350033000000 , at 129­­–31.
  • 12Id. at 129–30.
  • 13John Stuart Mill, On Liberty, in On Liberty, Utilitarianism, and Other Essays 15 (Mark Philp & Frederick Rosen, ed. 2015) (“[T]he principle [of liberty] requires liberty of tastes and pursuits; of framing the plan of our life to suit our own character; of doing as we like, subject to such consequences as may follow: without impediment from our fellow-creatures, so long as what we do does not harm them, even though they should think our conduct foolish, perverse, or wrong.”).
  • 14Dworkin, supra note  NOTEREF _Ref32948353 \h  \* MERGEFORMAT 7 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300350033000000 , at 131–32.
  • 15Id.
  • 16See, e.g., Scanlon, supra note  NOTEREF _Ref32948353 \h  \* MERGEFORMAT 7 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300350033000000 , at 204; Schauer, supra note  NOTEREF _Ref48974515 \h  \* MERGEFORMAT 9 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600340038003900370034003500310035000000 , at 1294.
  • 17Mill, supra note  NOTEREF _Ref44587197 \h  \* MERGEFORMAT 13 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600340034003500380037003100390037000000 , at 165.
  • 18Wesley Newcomb Hohfeld, Some Fundamental Legal Conceptions as Applied in Judicial Reasoning, 23 Yale L.J. 16 (1913) (describing different types of rights and their correlatives).
  • 19See, e.g., Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n, 447 U.S. 557, 573 (1980) (requiring government regulations of commercial speech to pass intermediate scrutiny).
  • 20See, e.g., City of Ladue v. Gilleo, 512 U.S. 43 (1994) (invalidating residential signage ban).
  • 21See id.; see also Members of the City Council v. Taxpayers for Vincent, 466 U.S. 789, 795 (1984); Reed v. Town of Gilbert, 576 U.S. 155, 171–72 (2015).
  • 22See United States v. O’Brien, 391 U.S. 367, 379–80 (1968).
  • 23See Texas v. Johnson, 491 U.S. 397, 410 (1989); United States v. Eichman, 496 U.S. 310, 314 (1990).
  • 24Clark v. Cmty. for Creative Non-Violence, 468 U.S. 288, 288 (1984).
  • 25See NAACP v. Claiborne Hardware Co., 458 U.S. 886, 888–89 (1982).
  • 26See, e.g., United States v. Alvarez, 567 U.S. 709 (2012).
  • 27See Brandenburg v. Ohio, 395 U.S. 444, 477 (1969).
  • 28See Schenck v. United States, 249 U.S. 47 (1919).
  • 29Dworkin, supra note  NOTEREF _Ref32948353 \h  \* MERGEFORMAT 7 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300350033000000 , at 131.
  • 30See Scanlon, supra note  NOTEREF _Ref32948353 \h  \* MERGEFORMAT 7 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300350033000000 , at 208; Dworkin, supra note  NOTEREF _Ref32948353 \h  \* MERGEFORMAT 7 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300350033000000 , at 131.
  • 31See supra notes  NOTEREF _Ref44590489 \h  \* MERGEFORMAT 22 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600340034003500390030003400380039000000 – NOTEREF _Ref44590496 \h  \* MERGEFORMAT 24 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600340034003500390030003400390036000000 and accompanying text.
  • 32See, e.g., Bernhard Schlink, Proportionality in Constitutional Law: Why Everywhere but Here?, 22 Duke J. Comp. & Int’l L. 291, 295–96 (2012) (noting that the German constitutional model, which protects fundamental freedoms but simultaneously empowers the legislature to limit it, has been exported to most European countries, as well as Israel, Canada, and South Africa).
  • 33See Greenawalt, supra note  NOTEREF _Ref32948397 \h  \* MERGEFORMAT 10 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300390037000000 , at 10.
  • 34See Dworkin, supra note  NOTEREF _Ref32948353 \h  \* MERGEFORMAT 7 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300350033000000 , at 130–31.
  • 35See Schauer, Free Speech on Tuesdays, supra note  NOTEREF _Ref48974515 \h 9 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600340038003900370034003500310035000000 (considering historical and pragmatic reasons for singling out special rights).
  • 36Emp’t Div. v. Smith, 494 U.S. 872, 883 (1990).
  • 37See, e.g., Estate of Thornton v. Caldor, 472 U.S. 703, 710–11 (1985); see also Micah Schwartzman, Nelson Tebbe, and Richard Schragger, The Costs of Conscience, 106 Ky. L. J. 781, 788 (2017).
  • 38A further argument might be that a special right necessarily offers further protection because it requires a higher level of scrutiny than a general right. This is not necessarily true, for the reasons I outline in Kendrick, supra note  NOTEREF _Ref55039005 \h † 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600350035003000330039003000300035000000 , 108–09. It is possible for the requirements of both a general and a special right to be satisfied by a single justification, and in a world of perfect information, no additional scrutiny would be required to ensure that the justification was as sincere and satisfactory for the special right as for the general right. To the extent that courts impose higher scrutiny on special rights (which they do not always do), this could be a matter of institutional design in response to imperfect information. Thus, in practice, special rights might receive more scrutiny, but it is not at all clear that they conceptually must in the way that, say, Dworkin seems to contemplate. See Dworkin, supra note  NOTEREF _Ref32948353 \h  \* MERGEFORMAT 7 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600330032003900340038003300350033000000 , at 131.
  • 39Shiffrin, supra note  NOTEREF _Ref44591875 \h  \* MERGEFORMAT 5 08D0C9EA79F9BACE118C8200AA004BA90B02000000080000000D0000005F00520065006600340034003500390031003800370035000000 , at 285.
  • 40Compare United States v. Robel, 389 U.S. 258 (1967) (rejecting the contention that defense facility employment could be predicated on mere political affiliation) with, e.g., Speiser v. Randall, 357 U.S. 513 (1958) (upholding denial of state tax exemption for veterans to veteran who refused to attest to political affiliation).
  • 41New York Times Co. v. Sullivan, 376 U.S. 254, 270 (1964).