True Threats: Voter Intimidation and the Constitution


Daniel P. Tokaji


I. Introduction

II. Congressional Power

III. The First Amendment

IV. Conclusion

I. Introduction

Voter intimidation is a recurrent problem that, like so many other aspects of our election system, provokes sharply polarized reactions along party lines. Of particular current interest are the activities of Tea Party-affiliated groups like True the Vote that ostensibly seek to promote electoral integrity but, in the eyes of critics, threaten to intimidate racial minorities, students, and other Democratic-leaning voters.[1] There is a long history of voter intimidation in the United States, as well as federal efforts to stop it.[2] Congress enacted the Enforcement Acts of 1870 and 1871 in response to the Ku Klux Klan’s often violent intimidation of African American voters.[3] Despite these laws, voter intimidation played an important role in the mass disenfranchisement of racial minorities through much of the United States, starting in the nineteenth century and continuing through most of the twentieth century.

Contemporary incidents of voter intimidation are nowhere near the scope and severity of those which emerged after the Civil War and persisted until the Voting Rights Act of 1965 (“VRA”). Still, there is reason to be concerned, as Ben Cady and Tom Glazer argue in Voter Strike Back.[4] Their article exposes underutilized tools that might be brought to bear against contemporary voter intimidation. Of particular interest is section 11(b) of the VRA, which prohibits intimidating, threatening, or coercing voters.[5] Cady and Glazer persuasively argue that this statute was designed to dispense with any requirement of intent, including both racially discriminatory intent and an intent to intimidate. Private groups like True the Vote may therefore violate section 11(b) even if they do not intend to discriminate or intimidate. It is sufficient, according to Cady and Glazer, that they engage in tactics that are “objectively intimidating.”[6] The authors go on to explain how section 11(b) might apply in various contexts, including aggressive poll-watching, challenges to voter eligibility, threats away from the polling place, and employer coercion.[7]

Although Cady and Glazer’s interpretation of voter intimidation statutes is persuasive, it raises thornier constitutional problems than they acknowledge. There are two distinct constitutional difficulties, both of which would have to be overcome by plaintiffs seeking to challenge alleged voter intimidation. The first is the scope of Congress’s power, particularly over purely private actors in purely state and local elections. There is no problem applying these statutes to state actors like election officials, police officers, and poll workers, but voter intimidation by private individuals and groups is a different matter. The second difficulty is that some applications of anti-intimidation statutes might violate First Amendment rights. What seems like intimidation to a would-be voter may well be free speech in the mind of a True the Vote volunteer. The pivotal question under the First Amendment is the scope of the “true threats” exception, an issue the Court considered but ultimately avoided last term in Elonis v. United States.[8] There is no doubt of section 11(b)’s consistency with the First Amendment where there is an intent to intimidate voters through a threat of physical violence. It is less clear whether section 11(b) may constitutionally be applied where non-violent harms are threatened or the intent to intimidate is lacking. However, because it provides only civil remedies, the statute is probably consistent with the First Amendment.

II. Congressional Power

The problem of congressional power arises from the Civil Rights Cases, which famously—some would say infamously—held that Congress’s Fourteenth Amendment enforcement power extends only to state action and not to private action.[9] While the Civil Rights Cases did not involve race discrimination in voting, the Fifteenth Amendment (like the Fourteenth) is by it terms limited to government action.[10] As I have previously explained, federal voting rights statutes enforce two distinct constitutional rights.[11] One is the right to be free from intentional race discrimination in voting. The other is “the right to vote as such,” as Richard Pildes has termed it[12]—that is, the right to be free from laws or practices that impede participation. These constitutional rights can generally be violated only by government action and not by purely private action.

There are exceptions to the state action requirement in which people or entities that are not traditional government actors have been held to violate Fourteenth or Fifteenth Amendment rights. One exception applies when private actors perform a traditional public function, such as the Texas Democratic Party whose exclusion of blacks from the party’s primary election was successfully challenged in the White Primary Cases.[13] The other exception applies where there is entanglement between the government and private actors, such as when a state law encourages or facilitates private discrimination.[14] These exceptions might apply in some instances of voter intimidation—for example, when one of the major parties is involved in voter caging,[15] or where poll watchers are working in concert with poll workers to frighten racial minorities away from the polls. They would not, however, extend to the actions of purely private persons who are not working in concert with state or local officials.

Suppose, for example, that a private citizen pays for billboards in Latino neighborhoods, proclaiming that voting by those with an outstanding child support order is illegal and punishable by twenty years imprisonment.[16] Further suppose that the statements are untrue and intended to discourage eligible citizens from voting. So long as there is no support or encouragement from the government, there is no state action and the citizen’s actions would not violate the Fourteenth or Fifteenth Amendment. It is doubtful that the Supreme Court would find that the enforcement clauses of either amendment furnish Congress with authority to prohibit voter intimidation by such private actors.

Presumably for this reason, Cady and Glazer rely on the Elections Clause in Article I, section 4 of the Constitution as the source of congressional authority for section 11(b) and other anti-intimidation laws. The Elections Clause gives Congress the power to “make or alter” the rules governing congressional elections. In Ex Parte Yarbrough, the Supreme Court upheld the voter intimidation provision of the Enforcement Act of 1871, commonly known as the KKK Act, under the Elections Clause, as applied to an election in which both congressional and state candidates were on the ballot.[17] The Court recently reaffirmed the broad scope of Congress’ Elections Clause authority in Arizona v. Inter-Tribal Council of Arizona, holding that states must comply with the National Voter Registration Act’s requirement that they “accept and use” the federal registration form for federal elections.[18]

Cady and Glazer are not the first to rely on the Elections Clause as an alternative to the Fourteenth and Fifteenth Amendments. Several scholars have previously argued that the Elections Clause provides a constitutional source of authority for some applications of the Voting Rights Act.[19] The problem, however, is that Congress’s Elections Clause power extends only to congressional elections; it cannot justify the application of federal anti-intimidation laws to the many elections in which only state and local candidates are on the ballot. Although Cady and Glazer suggest that the Necessary and Proper Clause would likely provide authority for the application of section 11(b) to local questions, they do not explain—and it is difficult to imagine—how this would be so, given that the Elections Clause, by its unambiguous terms, is limited to congressional elections.

It follows that most but not all conceivable applications of federal voter intimidation statutes would fall within Congress’ congressional authority. The Fourteenth and Fifteenth Amendment give Congress the power to provide remedies for voter intimidation by state and local officials. They may also justify the statute’s application to major parties and other people or groups who are either performing traditional public functions or entangled with the state. The Elections Clause would authorize Congress to provide remedies against purely private actors, but only for congressional elections. Voter intimidation statutes are therefore constitutional as applied to government action or a federal election, but may well exceed Congress’ authority as applied to purely private actions in purely state and local elections.

III. The First Amendment

Even if most applications of federal voter intimidation statutes fall within the scope of congressional power under the Elections Clause, they must still comport with the First Amendment. Given Cady and Glazer’s persuasive argument that section 11(b) requires neither proof of the intent to discriminate nor proof of the intent to intimidate, this interpretation raises the question whether some applications of section 11(b) would run afoul of free speech rights. A prohibition on “intimidation” would almost surely be deemed a content-based regulation of speech, hence subject to strict scrutiny unless it falls within some categorical exception.[20]

As with the question of congressional power, the difficulty of the First Amendment issue varies depending on the context. There is no serious constitutional issue with applying section 11(b) to state actors, who would be hard-pressed to make a persuasive free speech argument when it comes to the performance of their official duties.[21] A poll worker, for example, could not plausibly claim that she has a constitutional right to tell all white citizens to go away when they come out to vote. Designated challengers and poll watchers may also be deemed state actors rather than private actors endowed with speech rights when they intimidate voters at the polls.

The more difficult First Amendment questions involve purely private actors. Suppose for example that True the Vote members pass out flyers at predominantly black polling places stating that anyone with an outstanding warrant or criminal conviction is ineligible to vote and subject to prosecution for voter fraud. Suppose further that the statements are untrue, but that there is no evidence that they were intended to intimidate voters. The mere fact that the statements are false does not deprive them of constitutional protection, as the Supreme Court recently clarified in United States v. Alvarez.[22] The Court struck down the Stolen Valor Act’s criminal prohibition on false statements regarding military honors. Although there was no majority opinion, a majority of justices expressed the view that falsity alone is insufficient to put speech outside the protection of the First Amendment.[23] It is possible to imagine an exception for false statements that discourage people from voting, but the current Court has demonstrated little inclination to carve out new exceptions to the First Amendment.[24]

In Burson v. Freeman, the Court upheld a ban on electioneering within 100 feet of a polling place as narrowly tailored to protect the right to vote and electoral integrity.[25] It is less certain whether broader restrictions on speech surrounding elections are constitutional. The pivotal question is whether section 11(b)’s prohibition on voter intimidation falls within the “true threats” exception, under which genuine threats of violence are unprotected by the First Amendment. The seminal case is Watts v. United States,[26] involving a statute criminalizing knowing and willful threats to injure or kill the President. The Court said that the statute “is constitutional on its face,” but could not be applied to defendant’s statement at an anti-war rally, “If they ever make me carry a rifle the first man I want to get in my sights is L.B.J.”[27] Such crude hyperbole, the Court held, was not truly threatening and therefore not proscribable.

The Court refined the true threats doctrine in Virginia v. Black,[28] involving a state law that criminalized cross burning with the intent to intimidate. It defined true threats as “statements where the speaker means to communicate a serious expression of an intent to commit an act of unlawful violence to a particular individual or group of individuals.”[29] The Court offered three reasons why true threats fall outside the protection of the First Amendment: (1) to protect individuals from the fear of violence, (2) to protect them from the disruption that this violence engenders, and (3) to protect against the possibility that the threatened violence will actually occur.[30] It went on to define intimidation “in the constitutionally proscribable sense” as a kind of true threat “where the speaker directs a threat to a person or group of persons with the intent of placing the victim in fear of bodily harm or death.”[31] The Court held cross burning with the intent to intimidate falls within this definition and therefore may be proscribed by a properly drawn statute. However, a plurality found the jury instruction given under Virginia’s statute constitutionally defective because it made cross burning prima facie evidence of the intent to intimidate.[32]

There are two significant difficulties in determining whether the true threats doctrine would apply to alleged acts of intimidation that might be prosecuted under section 11(b). The first concerns the nature of the threat. In Black, the Court expressly stated that true threats of violence are unprotected by the First Amendment. But what if something other than violence is threatened? Voter intimidation may sometimes involve violence, but not always—as in the example of threats of criminal prosecution for voter fraud. It is not at all clear whether threats of something other than violence fall within the true threats doctrine. There is a strong argument that they should, given that non-violent threats may discourage eligible citizens from voting as much as threats of violence. However, the Court’s articulated definition of true threats in Black refers exclusively to violence, seeming to exclude other threatened harms. This narrow definition may be attributable to the facts of Black, which involved cross-burning and the implied threats of physical violence—often actualized—that historically tend to accompany it. For this reason, the Court’s characterization in Black of the true threats doctrine should not necessarily be understood to exclude harms other than physical violence.

In the end, the question whether non-violent harms fall within the true threats exception will probably turn on which of the above three reasons for the doctrine the Court deems most salient. If the doctrine is primarily used to prevent fear and attendant disruption (the first two rationales), threats of nonviolent harms may be just as bad as threats of violence. In the context of voter intimidation, the disruption is particularly noxious, given the fundamental character of the right to vote. But if the doctrine is primarily aimed at preventing physical violence, then other kinds of threats will probably be deemed insufficient.

The second difficulty in applying the true threats doctrine to voter intimidation is whether it covers speech that is not intended to intimidate. Some perceived forms of voter intimidation are not actually intended to discourage eligible citizens from voting. Those who post billboards warning of voter fraud prosecutions for illegal voting, pass out fliers threatening deportation of noncitizens who vote, or aggressively monitor polling places in certain neighborhoods might well intend to promote electoral integrity rather than to intimidate eligible citizens from voting.

The Supreme Court has not definitively resolved the question whether the true threats exception requires intent, and there is a longstanding split in the circuits as Cady and Glazer note.[33] The language in Black tends to support a narrower view of the doctrine, stating that the speaker must have the “intent of placing the victim in fear of bodily harm of death” for his expression to be deemed a true threat.[34] Here again, however, it is important to remember the context of Black. That case involved a criminal prosecution for burning a cross with the intent to intimidate. By contrast, only civil remedies are available under section 11(b).[35] That is clear from section 12(b) of the VRA, which allows criminal penalties for violation of section 11(a) (the prohibition on state officials refusing to allow an eligible person to vote) but does not mention section 11(b).[36]

An intent to intimidate might well be constitutionally required for criminal prosecutions, but not for civil actions. The most recent true threats case decided by the Court, Elonis v. United States,[37] suggests a distinction between criminal and civil actions. Elonis was a criminal case arising from crude social media postings that were perceived as threatening his estranged wife and others.[38] The defendant was convicted of violating a federal statute criminalizing “any threat to injure the person of another” transmitted through interstate commerce.[39] Although the defendant argued that the First Amendment required the intent to threaten, the Court did not reach that question. Instead, it concluded that the criminal statute should be understood to impose a scienter requirement—specifically, that the communication was “for the purpose of issuing a threat, or with knowledge that the communication will be viewed as a threat.”[40] The Court relied on the longstanding principle that a “guilty mind” is required for criminal convictions, even when a scienter requirement is not explicitly included in the statute, for its interpretation of the statute in Elonis.[41] Construing the criminal statute to require either the purpose to threaten of knowledge that it would be viewed a threat, the Court reversed the conviction without reaching the First Amendment issue.

The Court’s decision in Elonis provides some reason to believe that, where threats are concerned, a higher standard might be required for criminal liability than civil liability. On this view, section 11(b) would be constitutional, even in the absence of an intent to intimidate voters, because it is enforceable only through civil actions and remedies. While the precise scope of the true threats doctrine remains uncertain, there is good reason to believe that many if not all applications of the statute would be consistent with the First Amendment.

IV. Conclusion

Cady and Glazer have done an admirable job of explaining how existing federal statutes might be deployed to stop contemporary voter intimidation. They persuasively argue that section 11(b) of the VRA does not require intent, either to intimidate or to discriminate. Their interpretation of the statute, however, raises serious constitutional questions regarding both Congress’ enforcement power and the First Amendment. Most applications of existing anti-intimidation laws probably fall within the scope of congressional power under either the Fourteenth Amendment, the Fifteenth Amendment, or the Elections Clause, but their constitutionality is doubtful with respect to purely private actors in purely state and local elections. Section 11(b) also raises serious First Amendment issues in some contexts. There would be no free speech violation if the statute were applied to election officials or to private parties who intentionally intimidate voters with threats of violence. But the statute’s constitutionality is less certain in cases involving statements by private persons that are not intended to threaten violence. The best view is probably that the statute is constitutional as applied to such cases, insofar as only civil remedies are available. While this point is contestable, there can be no doubt that the issue of voter intimidation highlights the need for the Supreme Court to clarify the scope of the true threats doctrine.

Charles W. Ebersold and Florence Whitcomb Ebersold Professor of Constitutional Law, The Ohio State University, Moritz College of Law.

1. Ben Cady & Tom Glazer, Voters Strike Back: Litigation Against Modern Voter Intimidation, 39 N.Y.U. Rev. L. & Soc. Change 173, 177 (2015).

2. Id. at 181–91.

3. Id. at 183–87.

4. Id.

5. 52 U.S.C. § 10307(b).

6. Cady & Glazer, supra note 1, at 205.

7. Id. at 216–22.

8. 135 S. Ct. 2001 (2015).

9. 109 U.S. 3 (1883) (striking down the Civil Rights Act of 1875’s prohibition on private discrimination in public accommodations); see also United States v. Harris, 106 U.S. 629 (1883) (striking down criminal statute prohibiting private parties from conspiring to deny equal protection).

10. U.S. Const., amend. XIV (“The right of citizens of the United States to vote shall not be denied or abridged by the United States or by any state on account of race, color, or previous condition of servitude.”) (emphasis added).

11. Daniel P. Tokaji, Applying Section 2 to the New Vote Denial, 50 Harv. C.R.-C.L. L. Rev. 439, 468–73 (2015); Daniel P. Tokaji, Intent and Its Alternatives: Defending the New Voting Rights Act, 58 Ala. L. Rev. 350, 368–74 (2006) [hereinafter Intent].

12. Richard H. Pildes, The Future of Voting Rights: From Anti-Discrimination to the Right to Vote, 49 How. L.J. 741, 760 (2006).

13. Smith v. Allwright, 321 U.S. 649 (1944).

14. See, e.g., Reitman v. Mulkey, 387 U.S. 369 (1967) (finding state action where a ballot initiative “would encourage and significantly involve the State in private discrimination”); Burton v. Wilmington Parking Auth., 365 U.S. 715 (1961) (finding state action where municipal parking authority leased space to private restaurant that discriminated on the basis of race); Shelley v. Kramer, 334 U.S. 1 (1948) (finding state action where a state court enforces racially restrictive covenants among private parties).

15. Voter caging is the tactic of trying to remove from the rolls voters who are presumed to be unfriendly, typically by sending mail to voters likely to be aligned with the other party and then challenging those for whom the mailing was returned as undeliverable.

16. For a real-life example of billboards that seem intended to discourage people from voting, see James O’Toole, Clear Channel Takes Down Voter Fraud Billboards, CNNMoney (Oct. 22, 2014),

17. 110 U.S. 651, 660–62 (1884).

18. 133 S.Ct. 2247 (2013).

19. I am one of them. See Intent, supra note 11, at 365–68; see also Gabriel J. Chin, Section 5 of the Voting Rights Act and the ‘Aggregate Powers’ of Congress over Elections (U.C. Davis Legal Studies Research Paper No. 313, 2012),; Samuel Issacharoff, Beyond the Discrimination Model in Voting, 127 Harv. L. Rev. 95, 107­13 (2012); Franita Tolson, Reinventing Sovereignty?: Federalism as a Constraint on the Voting Rights Act, 65 Vand. L. Rev. 1195, 1207­32 (2012).

20. See Reed v. Town of Gilbert, 135 S. Ct. 2218, 2226 (2015) (content-based speech restrictions must be narrowly tailored to a compelling interest); see also id. at 2227 (“Government regulation of speech is content based if a law applies to particular speech because of the topic discussed or the idea or message expressed.”).

21. Public employees do enjoy some protection for speech on “matters of public concern.” Pickering v. Bd. of Educ., 391 U.S. 563 (1968) (holding that the First Amendment prohibits dismissal of public school teacher for a letter to the editor criticizing school board and superintendent); see also Connick v. Myers, 461 U.S. 138 (1983) (holding that a public employee’s speech was not protected because it did not involve a matter of public concern). There may be some cases in which an election official or other public employee can claim that alleged intimidation of voters was speech on a matter of public concern, but such cases are likely to be quite rare.

22. 132 S. Ct. 2537 (2012).

23. Id. at 2545; id. at 2553–55 (Breyer, J., concurring).

24. In addition to Alvarez, see Brown v. Entm’t Merch.s Ass’n, 131 S. Ct. 2729 (2011) (striking down state law prohibiting violent video games), and United States v. Stevens, 559 U.S. 460 (2010) (striking down statute prohibiting depictions of animal cruelty).

25. The plurality applied strict scrutiny. 504 U.S. 191, 206, 211 (1992) (Blackmun, J.). Justice Scalia concurred in the judgment, believing that the regulation need only be reasonable and viewpoint-neutral to survive constitutional scrutiny. Id. at 214 (Scalia, J., concurring).

26. See Watts v. United States, 394 U.S. 705 (1969).

27. Id. at 706.

28. 538 U.S. 343 (2003).

29. Id. at 359.

30. Id. at 360.

31. Id.

32. 538 U.S. at 363–67 (O’Connor, J., joined by Chief Justice Rehnquist, Justice Stevens, and Justice Breyer). Justices Scalia and Thomas believed that the cross-burning statute should not be struck down on its face but that the case should be sent back to the Virginia Supreme Court for it to construe the statute. Id. at 368 (Scalia, J., concurring). Justices Souter, Kennedy, and Ginsburg believed the statute unconstitutional for another reason, that it impermissibly discriminated based on content within a category of proscribable speech. Id. at 380 (Souter, J., concurring).

33. Cady & Glazer, supra note 1, at 209 & 209 nn.234 & 236. Although they say that the split emerged “[i]n the wake of” Black, the circuits were actually divided even before that case. See, e.g, United States v. Orozco-Santillian, 903 F.2d 1262, 1265 (9th Cir. 1990) (defining question as “whether a reasonable person would foresee that the statement would be interpreted by those to whom the maker communicates the statement as a serious expression of intent to harm or assault”) (emphasis added); United States v. Kelner, 534 F.2d 1020, 1027 (2d Cir. 1976) (“So long as the threat on its face and in the circumstances in which it is made is so unequivocal, unconditional, immediate and specific as to the person threatened, as to convey a gravity of purpose and imminent prospect of execution, the statute may properly be applied.”) (emphasis added).

34. 538 U.S. at 360.

35. The VRA expressly authorizes civil actions by the Attorney General for violations of section 11, including 11(b)’s prohibition on voter intimidation. 52 U.S.C. § 10308(d). Cady and Glazer take the position that there is a private right of action for injunctive relief but not damages under section 11(b). See Cady & Glazer, supra note 1, at 207. That is less certain. They cite one district court case that allowed a private claim for injunctive relief, James v. Humpheries Cnty. Bd. of Election Comm’rs, 384 F. Supp. 114 (N.D. Miss. 1974), and one circuit court case that disallowed a private claim for damages, Olagues v. Rossoneillo, 770 F.2d 792, 804–05 (9th Cir. 1985). There is no express private right of action, so the availability of both injunctive relief and damages would hinge on whether an implied right of action lies. In addition, plaintiffs might make claims for both damages and injunctive relief under section 1983, where state or local officials are alleged to engage in voter intimidation. For a more detailed discussion of the availability of private rights of action in voting rights cases, both implied and under section 1983, see Daniel P. Tokaji, Public Rights and Private Rights of Action: The Enforcement of Federal Election Laws, 44 Ind. L. Rev. 113 (2010).

36. 52 U.S.C. § 10308(a) (“Whoever shall deprive any person of any right secured by … section 10307(a) of this title [section 11(a) of the VRA], shall be fined not more than $5,000, or imprisoned not more than five years, or both.”).

37. 135 S. Ct. 2001 (2015).

38. Id.

39. Id. at 2008 (quoting 18 U.S.C. § 875(c)).

40. Id. at 2012.

41. Id. at 2009.

Suggested Reading