A version of this essay was presented at Columbia Law Review’s 2018 symposium, “A First Amendment for All? Free Expression in an Age of Inequality,” co-sponsored by the Knight First Amendment Institute and the Center for Constitutional Governance.

Over the past four decades, the political economy of the First Amendment has undergone a significant shift. If in the early twentieth century winners in First Amendment cases tended to be representatives of the marginalized and the disenfranchised, these days, they are much more likely to be corporations and other powerful actors. This Essay excavates the causes of that change and suggests how it might be reme­died. It argues that the shift in First Amendment political economy is not primarily a consequence of the overly expansive scope of current free speech law—as some have argued. Nor is it a product of the Court’s free speech libertarianism. What it reflects instead is the Court’s embrace over the past several decades of a highly formal conception of the First Amendment equality guarantee. If the Court once interpreted the First Amendment to require, or at least permit, substantive equality of expres­sive opportunity, today the Court insists that the First Amendment guarantees—and guarantees only—formally equal treatment at the gov­ernment’s hands. It is this shift, this Essay argues, that has pro­duced a free speech jurisprudence that tends to favor the powerful and the propertied. By examining its causes and excavating areas of free speech law in which the Court has attempted to vindicate a more substantive conception of expressive equality, this Essay begins the work of charting out an alternative, more antisubordinating First Amendment.


    1. Campaign Finance Regulation
    2. Time, Place, and Manner Cases
    3. Public Employee Speech
    1. Speech on Private Property
    2. Media-Access Cases
    3. Employer Speech



Over the past forty years, the political economy of the First Amendment has undergone a significant shift. In the early and mid-twentieth century, litigants that won First Amendment cases tended to be civil rights groups like the NAACP, 1 See, e.g., NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 460, 466 (1958) (holding that disclosure of the names of NAACP members to the state of Alabama would violate the organization’s “freedom to engage in association for the advancement of beliefs and ideas”). proponents of minority religions, 2 See, e.g., W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 642 (1943) (holding that a policy compelling Jehovah’s Witnesses to recite the pledge of allegiance violated their First Amendment rights). and other representa­tives of the marginalized and the disenfranchised. 3 See, e.g., De Jonge v. Oregon, 299 U.S. 353, 365–66 (1937) (holding that Oregon’s Criminal Syndicalism Act was unconstitutional as applied to a member of the Communist Party who was convicted for conducting a peaceful meeting under the auspices of the party). These days, the winners in First Amendment cases are much more likely to be corporations and other economically and politically powerful actors. 4 See John C. Coates IV, Corporate Speech & the First Amendment: History, Data, and Implications, 30 Const. Comment. 223, 248 (2015) (analyzing “data from Supreme Court and Circuit Court decisions to illustrate how recently the corporate takeover of the First Amendment has occurred, and how pervasively and systematically corporations have been using the First Amendment to achieve de- or re-regulatory goals”). The result is that today the First Amendment often serves as the “primary guarantor of the privileged” rather than the champion of the powerless it used to be. 5 Mark Tushnet, An Essay on Rights, 62 Tex. L. Rev. 1363, 1386 –92 (1984).

Scholars have provided two explanations for the change. Some have argued that it is a consequence of the decision to extend constitutional protection to commercial speech and corporate speakers. 6 See, e.g., Daniel J.H. Greenwood, First Amendment Imperialism, 1999 Utah L. Rev. 659, 659 (“The First Amendment threatens to swallow up all politics. . . . Increasingly, it acts as a bar to governmental action not just with regard to the issues of conscience and religious practice with which it began, but far into the realm of economic regula­tion . . . .”); Jedediah Purdy, Neoliberal Constitutionalism: Lochnerism for a New Economy, 77 Law & Contemp. Probs., no. 4, 2014, at 195, 203 (arguing that decisions like Citizens United v. FEC, 558 U.S. 310 (2010), demonstrate “the Court’s march away from a principle that it accepted with the New Deal: Buying and selling enjoy no special constitu­tional status”). By inter­preting the guarantee of freedom of speech too expansively, they argue, the Court has allowed the First Amendment to be “hijacked” by corpora­tions and other business groups and to be turned into a tool of economic deregula­tion and corporate power. 7 Tim Wu, The Right to Evade Regulation: How Corporations Hijacked the First Amendment, New Republic (June 3, 2013), []. Others attribute the shift in the demo­graphics of the First Amendment’s beneficiaries to the excessive libertarianism of the con­temporary Court. They argue that the Court’s tendency to treat free speech interests as more important than almost any other competing interest has produced a First Amendment jurispru­dence that is favorable for corporations, relative to everyone else. 8 See, e.g., Tabatha Abu El-Haj, “Live Free or Die”—Liberty and the First Amendment, 78 Ohio St. L.J. 917, 922–23 (2017) (arguing that the “libertarian First Amendment” that has emerged in recent years poses a real threat to the ability of the regu­latory state to per­form its core functions); Kathleen M. Sullivan, Two Concepts of Freedom of Speech, 124 Harv. L. Rev 143, 145 (2010) (arguing that decisions such as Citizens United represent the “triumph of the libertarian over the egalitarian vision of free speech”); Morgan N. Weiland, Expanding the Periphery and Threatening the Core: The Ascendant Libertarian Speech Tradition, 69 Stan. L. Rev. 1389, 1397 (2017) (arguing that contemporary First Amendment law relies on a libertarian vision of the First Amendment that “represents a radical break from the republican and liberal traditions on which it draws” by “subordinat­ing [listener rights] to corporate speech rights and eventu­ally nullifying them altogether”).

This Essay suggests a third explanation. It argues that the shift in the First Amendment’s political economy is not entirely—or even primar­ily—a consequence of the process Professor Daniel Greenwood has called “First Amendment imperialism,” which he describes as the “rapid expansion [of the First Amendment] into areas long thought impervious to consti­tutional law,” particularly areas of “economic regulation . . . the courts had abandoned to the legislatures after the Lochner disaster.” 9 Greenwood, supra note 6, at 659–60 (footnote omitted). Nor is it a product of the Court’s excessive libertarianism. Indeed, this Essay challenges the idea that the contemporary Court is particularly libertar­ian when it comes to freedom of speech.

What this shift reflects instead is the Court’s embrace over the past sev­eral decades of a highly formal conception of the First Amendment equality guarantee. Since the New Deal period, the Court has recognized that implicit in the First Amendment guarantee of expressive liberty is a guarantee of expressive equality—that freedom of speech means not only the right to speak but the right to speak on equal terms as other speakers. 10 See Kenneth L. Karst, Equality as a Central Principle in the First Amendment, 43 U. Chi. L. Rev. 20, 21 (1975). Over time, however, the Court has significantly changed its understand­ing of what this means.

For much of the twentieth century, the Court interpreted the guar­antee of expressive equality in a manner that was sensitive to the eco­nomic, political, and social inequalities that inhibited or enhanced expression. 11 Id. It interpreted the First Amendment, for example, to require that those who lacked other means of expressing themselves be granted access to publicly important spaces (including privately owned public spaces) to do so. 12 See infra notes 121–134 and accompanying text. It also struck down laws that, although in principle appli­cable to all, had a disparate impact on the ability of the poor and the powerless to communicate. 13 See infra notes 39–42 and accompanying text. And it refused to invalidate on First Amendment grounds laws that restricted the speech of the powerful in an effort to enhance the speech of the powerless. 14 See infra notes 43–44 and accompanying text. It interpreted the First Amendment, in other words, to guarantee—or at least permit—a rough kind of substantive equality in expressive opportunity.

Since the 1970s, however, the Court has moved increasingly far away from this context-sensitive, substantive-equality-promoting view of the First Amendment. It has rejected the idea that courts should take into account inequalities in economic and political power when interpreting the First Amendment command. 15 See, e.g., Buckley v. Valeo, 424 U.S. 1, 48–49 (1976) (per curiam). It has also, for the most part, rejected the idea that the First Amendment permits the government to limit the speech of wealthy or powerful speakers in order to enhance the speech of others. 16 Id.; see also Ariz. Free Enter. Club’s Freedom Club PAC v. Bennett, 564 U.S. 721, 736–40 (2011); Davis v. FEC, 554 U.S. 724, 742 (2008). Instead, it has interpreted the guarantee of expressive equal­ity to require—and to require only—formally equal treatment at the gov­ernment’s hands.

It is this change in the Court’s conception of what it means to guarantee expressive equality that is largely responsible, this Essay argues, for the “corporate takeover” of the First Amendment. 17 Coates, supra note 4, at 265. And it is a prob­lem, not only because it means in practice that the First Amendment fre­quently fails to protect the expressive freedom of those who lack the eco­nomic resources to communicate effectively in our highly commodi­fied public sphere. Indeed, this shift is so troubling because it under­mines as a result the robust and inclusive public debate that the First Amendment is supposed to make possible.

Taking stock of the present state of free speech jurisprudence thus requires taking stock of this change in the Court’s understanding of expressive equality. Doing so also obviously has normative implications. If the problem posed by the contemporary free speech doctrine is simply that it renders too much ordinary economic regulation subject to judicial scrutiny and that it makes that judicial scrutiny too demanding when it applies, then the obvious response is to narrow the scope of the First Amendment (to decolonize the empire, in other words) and to weaken the intensity of its protections. But if the problem with contemporary free speech doctrine is an egalitarianism that tends to favor both govern­ment and private power, what is needed is not a weaker and a narrower First Amendment but a different First Amendment—one that functions better to protect the expressive freedom of the powerless. What needs to change, in other words, is not the strength of the speech right but its meaning.

This Essay begins the work of charting out that alternative—an “anti­subordinating” First Amendment—by pointing to the areas of case law in which the Court has attempted to vindicate a more substantive conception of expressive equality. It proceeds in three parts. Part I describes the recent shift in free speech jurisprudence toward a more formal conception of expres­sive equality. Part II canvasses some of the doctrinal areas in which the Court once sought to vindicate substantive equality. Part III addresses some of the arguments that might be made against an antisubordinating approach.

I. Two Visions of Equality

It is by now well known that, when it comes to the Equal Protection Clause, there has been significant and enduring disagreement about what it means to guarantee equal protection of the law. 18 See Jack M. Balkin & Reva B. Siegel, The American Civil Rights Tradition: Anticlassification or Antisubordination?, 58 U. Miami L. Rev. 9, 10 (2003) (describing the “antisubordination” and “anticlassification” approaches, as well as the Supreme Court’s adoption of the latter). On one view—a view that is, at present, embraced by a majority of the members of the Supreme Court—what the Equal Protection Clause guarantees is formal equality. 19 See Kimberly Jenkins Robinson, The Constitutional Future of Race-Neutral Efforts to Achieve Diversity and Avoid Racial Isolation in Elementary and Secondary Schools, 50 B.C. L. Rev. 277, 315 (2009) (“The Court’s current approach to equal protec­tion, which has been labeled an antidiscrimination, anticlassification, or color-blind approach, emphasizes the impropriety of government use of racial classifications.”). On this view, the government violates its obligation to provide equal protection when it treats people differently because of their race, gender, or a limited number of other morally irrelevant and historically freighted “suspect” categories—at least absent an extremely good reason for doing so. 20 See Balkin & Siegel, supra note 18, at 10 (“Roughly speaking, [the anticlassifica­tion] principle holds that the government may not classify people either overtly or surreptitiously on the basis of a forbidden category: for example, their race.”). The government lives up to its equality obligations, in con­trast, when it treats individuals of different races or genders the same; that is, when it adopts, in relation to its citizens, “color-blind” 21 See, e.g., Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 748 (2007) (plurality opinion) (Roberts, C.J.) (asserting that “[t]he way to stop discrim­ination on the basis of race is to stop discriminating on the basis of race”); Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 291 (1978) (plurality opinion) (Powell, J.) (“Racial and ethnic distinctions of any sort are inherently suspect and thus call for the most exact­ing judicial examination.”). or “gender-blind” 22 See, e.g., United States v. Virginia, 518 U.S. 515, 531 (1996) (“Parties who seek to defend gender-based government action must demonstrate an ‘exceedingly persuasive justification’ for that action.” (quoting Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 273 (1979))). goggles.

On another view—one articulated in earlier cases such as Strauder v. West Virginia 23 100 U.S. 303 (1879). and Shelley v. Kraemer, 24 334 U.S. 1 (1948). and embraced today by a minority of Justices on the Court 25 See Reva B. Siegel, From Colorblindness to Antibalkanization: An Emerging Ground of Decision in Race Equality Cases, 120 Yale L.J. 1278, 1305 (2011) (“The four dissenting Justices in Parents Involved express key tenets of the antisubordination under­standing of Brown.”). —what equal protection requires is something more substantive. On this view, the government fails to provide equal protection of the law when it allows its institutions to be used to perpetu­ate the de facto, even if not de jure, second-class status of some members of society. 26 Kraemer, 334 U.S. at 19–21 (holding that judicial enforcement of racially restric­tive covenants violates the Equal Protection Clause by using “the full coercive power of government to deny [to some] . . . rights of ownership or occupancy enjoyed as a matter of course by other citizens of different race or color”); Strauder, 100 U.S. at 308 (asserting that the purpose of the Equal Protection Clause is to protect ex-slaves and their descend­ants from “legal discriminations [that] . . . lessen[ ] the security of their enjoyment of the rights which others enjoy” and from “discriminations which are steps towards reducing them to the condition of a subject race”). What determines, on this view, whether a law equally protects is not the form it takes or the purposes that motivate it but its effects on a complex social environment. As a result, formally identical treatment does not necessarily satisfy equal protection, nor does formally dissimilar treatment necessarily violate it. In fact, some proponents of this more context-sensitive conception of equal protection have argued that not only does the Fourteenth Amendment not prohibit the government from making suspect distinctions; in some cases, it may mandate them—when, for example, there is no other way the government can avoid perpetu­ating the second-class status of members of historically disenfranchised groups. 27 David A. Strauss, The Myth of Colorblindness, 1986 Sup. Ct. Rev. 99, 128 (“It [is] plausible [to conclude] . . . that affirmative action is constitutionally required [because] it is divisive and harmful to society to exclude nearly all blacks from important social institu­tions and benefits, even when that exclusion is the product of race-neutral criteria.”).

As should be obvious from these thumbnail descriptions, whether a court adopts one or the other view of the Equal Protection Clause will affect the outcome it reaches in a wide variety of cases. In cases involving challenges to race-based affirmative action laws and policies, as well as in cases involving constitutional challenges to formally neutral laws that have a significantly disparate impact on one group or another, the differ­ence between the formal and substantive—or what sometimes get called the “anticlassificatory” 28 Genevieve Lakier, Reed v. Town of Gilbert, Arizona, and the Rise of the Anticlassificatory First Amendment, 2016 Sup. Ct. Rev. 233, 236. and “antisubordinating” 29 Ruth Colker, Anti-Subordination Above All: Sex, Race, and Equal Protection, 61 N.Y.U. L. Rev. 1003, 1010 n.18 (1986). —conceptions of equal protection really matters. 30 See Balkin & Siegel, supra note 18, at 11 (explaining that “depending on how the Court dealt with . . . facially neutral practices with a disparate impact on racial minorities, the Constitution would either rationalize or destabilize the practices that sustained the racial stratification of American society”). Indeed, scholars have convincingly argued that the Court’s embrace of a highly formal interpretation of the Equal Protection Clause has significantly undermined its effectiveness as a safe­guard of substantive equality. 31 For a strong articulation of this point, see Reva Siegel, Why Equal Protection No Longer Protects: The Evolving Forms of Status-Enforcing State Action, 49 Stan. L. Rev. 1111, 1129 (1997); see also Ian Haney-López, Intentional Blindness, 87 N.Y.U. L. Rev. 1779, 1877 (2012) (“Intentional blindness, as both doctrine and justificatory rhetoric, stands today as a prelude to even more unjust racial politics.”). One could put the point more strongly: that the Court’s embrace of a formal rather than a substantive concep­tion of equality has turned the Equal Protection Clause into a powerful mechanism of disequalization: a tool that can be, and frequently is, used to defend entrenched social hierarchies rather than to challenge them. 32 This is obviously true of the affirmative action cases, but the same can also be said of cases such as McCleskey v. Kemp, 481 U.S. 279 (1987), which validate the “equality” provided by facially neutral laws. For a full account of this argument, see Randall L. Kennedy, McCleskey v. Kemp: Race, Capital Punishment, and the Supreme Court, 101 Harv. L. Rev. 1388, 1389 (1988).

A similar story can be told about the First Amendment. Although the text of the First Amendment speaks only of liberty, not equality, since the early twentieth century, the Court has recognized that implicit in the guarantee of freedom of speech is a guarantee of expressive equality. It has insisted, in other words, that freedom of speech means not only that one possesses some quantum of liberty to speak but that one has the same liberty to speak as do others. It has concluded, as a result, that the government may not restrict speech unequally even when it might be able to do so equally. 33 This is the lesson from decisions such as Niemotko v. Maryland, 340 U.S. 268, 272 (1951), and Cantwell v. Connecticut, 310 U.S. 296, 310 (1940), which struck down overly discretionary licensing laws not because they restricted too much speech but because they could be too easily used to discriminate.

The Court has interpreted the First Amendment to guarantee expres­sive equality because it has recognized that laws that treat speakers une­qually threaten one of the First Amendment’s central purposes: namely, ensuring that public debate on public matters is “uninhibited, robust, and wide-open.” 34 N.Y. Times Co. v. Sullivan, 376 U.S. 254, 270 (1964). This is because when the government uses its coercive power to advantage or disadvantage certain ideas or certain speakers, the effect may be to drive some ideas out of the public realm altogether, thereby diminishing the “uninhibited” and “wide-open” nature of the public debate. 35 Simon & Schuster, Inc. v. Members of N.Y. State Crime Victims Bd., 502 U.S. 105, 116 (1991) (noting that discriminatory laws “raise[ ] the specter that the government may effectively drive certain ideas or viewpoints from the marketplace”). Laws that deprive some speakers of expressive opportunities that others possess, the Court has suggested, also insult the dignity of those they target by implying—and sometimes outright declaring—that some ideas and opinions are more valuable than others. 36 Id. (asserting that “[t]he constitutional right of free expression is . . . intended to remove governmental restraints from the arena of public discussion . . . in the belief that no other approach would comport with the premise of individual dignity and choice upon which our political system rests” (second alteration in original) (internal quotation marks omit­ted) (quoting Leathers v. Medlock, 499 U.S. 439, 448–49 (1991))); see also Elena Kagan, Private Speech, Public Purpose: The Role of Governmental Motive in First Amendment Doctrine, 63 U. Chi. L. Rev. 413, 511 (1996) (arguing by analogy to equal protection juris­prudence that discriminatory regulations of speech are impermissible because the govern­ment’s treatment of some speakers as “less intrinsically worthy . . . register[s] a kind of disrespect that . . . renders [the regulations] improper”). Consequently, the Court has insisted for decades now that the First Amendment guarantees not only freedom of speech but also what Justice Marshall described in Police Department of Chicago v. Mosley as “equality of status in the field of ideas.” 37 408 U.S. 92, 96 (1972) (internal quotation marks omitted) (quoting Alexander Meiklejohn, Political Freedom: The Constitutional Powers of the People 27 (1948)).

Like the guarantee of equal protection, the guarantee of “equality of status in the field of ideas” could obviously be interpreted in multiple ways. It could be interpreted as a guarantee of substantive equality. That is to say, it could be interpreted to mean that differently positioned mem­bers of the public enjoy roughly equal opportunity to express themselves publicly. Alternatively, it could be interpreted more formally, to require—but to require only—that the government not treat speakers differently because of who they are or what they have to say.

For much of the twentieth century, the Court tended to employ the first, more substantive, conception of expressive equality. This is not to say that the early- and mid-twentieth-century Court had no objection to laws that subjected different speakers to different rules based on the con­tent of their speech. To the contrary: In case after case, it made clear that the government could not treat speakers unequally because it disliked their message or their viewpoint. In most of these cases, however, the groups targeted by the discriminatory speech laws were those at the bottom of the political and social hierarchies: Jehovah’s Witnesses, for example, or Communists. 38 See, e.g., Saia v. New York, 334 U.S. 558, 559 (1948); Cantwell v. Connecticut, 310 U.S. 296, 300 (1940); Herndon v. Lowry, 301 U.S. 242, 245 (1937); Stromberg v. California, 283 U.S. 359, 362 (1931). By prohibiting formally unequal treatment, the Court also promoted substantive equality.

In other cases, the Court struck down laws that satisfied formal equality (that is, they applied equally to all and were not motivated by discrimina­tory animus) but nevertheless imposed a much greater burden on the expressive freedom of the poor and powerless than they imposed on oth­ers. In Martin v. City of Struthers, for example, the Court struck down a local ordinance that prohibited distributors of handbills and circulars from ringing the doorbell or knocking on the door of any house in town with the purpose of distributing their materials. 39 319 U.S. 141, 146 (1943). It did so because it found that the ordinance prohibited a mode of communica­tion that played a vital role in democratic politics—that was, in particular, “essen­tial to the poorly financed causes of the little people”—without adequate justifi­cation. 40 Id. In its heckler veto cases, meanwhile, the Court held that other­wise-constitutional breach of peace laws could not be used to penal­ize those whose speech incited a hostile audience reaction, except when doing so was absolutely necessary to preserve the peace. 41 See, e.g., Cox v. Louisiana, 379 U.S. 536, 551 (1965); Edwards v. South Carolina, 372 U.S. 229, 238 (1963); Terminiello v. Chicago, 337 U.S. 1, 37 (1949). It did so because it believed that any other rule would allow “standardiza­tion of ideas either by legislatures, courts, or dominant political or commu­nity groups”—it would allow those who possessed the power of the crowd (those who belonged to locally “dominant . . . groups”) to use the instru­mentali­ties of the govern­ment to silence those who lacked this power. 42 Terminiello, 337 U.S. at 4–5 (emphasis added). The Court deviated from this rule once, in Feiner v. New York, 340 U.S. 315 (1951), in an opinion that relied, ultimately, on a formal equality argument. Indeed, the primary justification the majority provided for why the arrest and conviction of the unpopular speaker could be sustained in that case was that “there was no evidence which could lend color to a claim that the acts of the police were a cover for suppression of petitioner’s views and opinions.” Id. at 319–20. Because there was no evi­dence that the government intended to discriminate against the speaker because of his views, the Court concluded that the First Amendment was not offended. Justice Black wrote a blistering dissent, in which he warned of the danger that the majority’s analysis posed to the freedom of speech of unpopular speakers. See id. at 328–29 (Black, J., dissenting) (“[T]oday’s holding means . . . minority speakers can be silenced . . . . Criti­cism of public officials will be too dangerous for all but the most courageous. . . . [W]hile previous re­straints probably cannot be imposed on an unpopular speaker, the police have discretion to silence him as . . . the customary hostility to his views develops.”). In later years, the Court largely heeded Black’s warnings and returned to the approach employed several years before Feiner in Terminiello. See, e.g., Cox, 379 U.S. at 552; Edwards, 372 U.S. at 238.

In these and other cases, the Court recognized that even when the government did not treat speakers differently because it disliked their viewpoint or their identity, its actions could have a disparate impact on the ability of some to communicate, given underlying inequalities in access to expressive resources or social capital. It recognized, further­more, that this disparate impact, if sufficiently substantial, could violate the consti­tu­tional guarantee of freedom of speech. This is a vision of the First Amendment as the safeguard of more—and in some cases less—than formal equality.

Indeed, not only did the Court strike down formally neutral laws that had a disparate effect on the ability of the “little people” to express themselves; it also upheld formally unequal laws that, by treating differ­ent speakers differently, attempted to compensate for differences in eco­nomic and political power. Perhaps the most famous instance of this is the deci­sion in Red Lion Broadcasting Co. v. FCC in 1967, about which I will say more later. 43 395 U.S. 367 (1969); see also infra notes 150–159 and accompanying text. But also notable in this regard is the 1957 decision in United States v. UAW-CIO, in which the Court refused to invalidate on First Amendment grounds a section of the Taft–Hartley Act that prohib­ited unions—but only unions—from spending money on election-related speech in order to “protect the political process from . . . the corroding effect of money employed in elections by aggregated power.” 44 352 U.S. 567, 582 (1957).

Here and elsewhere, the Court adopted what we might call, borrow­ing from the equal protection scholarship, an “antisubordinating” view of the First Amendment. It did not always do so self-confidently or with great clarity. In fact, at times, the Court appeared entirely unable to explain why it was interpreting the First Amendment as it did. In UAW-CIO, for example, Justice Frankfurter’s majority opinion elaborated at great length on the history of campaign finance regulation in the United States but was entirely unable to explain why the ban on union spending did not violate the First Amendment. 45 Instead, after a lengthy discussion of the pressing concerns that motivated Congress to regulate campaign spending, Frankfurter argued that the Court neither needed to nor should analyze the constitutional issues raised by the case and noted some of the questions that it might address on future occasions. Id. at 591–92. The questions Frankfurter listed strongly suggested, however, that he believed that the ban could be profitably challenged only as applied, rather than facially, though he did not explain why this might be. See id. at 592 (noting as questions that might profitably be explored on future occasions: “Did it constitute active electioneering or simply state the record of particular candidates on economic issues? Did the union sponsor the broadcast with the intent to affect the results of the election?”). It is not hard to understand, how­ever. If the purpose of the First Amendment is to foster “the widest possi­ble dissemination of information from diverse and antagonistic sources”—as the Court declared it to be in Associated Press v. United States in 1945 46 326 U.S. 1, 20 (1945). and would reiterate on many occasions subsequently—then laws that restrict the expressive activity of powerful actors in order to foster participation from less powerful, but potentially more numerous, more diverse, and more antagonistic others do not necessarily threaten First Amendment values; instead, they may help promote them. 47 For later iterations of the claim made in Associated Press, see CBS. v. Democratic Nat’l Comm., 412 U.S. 94, 183 (1973); Citizen Publ’g Co. v. United States, 394 U.S. 131, 139 (1969); N.Y. Times Co. v. Sullivan, 376 U.S. 254, 266 (1964). This is certainly what the opinion in Associated Press suggested. 48 Associated Press, 326 U.S. at 20 (“Freedom of the press from governmental inter­ference under the First Amendment does not sanction repression of that freedom by pri­vate interests.”).

That Frankfurter was unwilling to endorse this view of the First Amendment, explicitly at least, suggests how difficult at least some members of the Court found the antisubordinating approach when what it entailed were constraints on private speech. It was well and good to say that the government could not regulate speech when doing so had a dis­parate impact on the expressive freedom of the “little people.” It was quite another thing, however, to say that the expressive freedom of powerful private actors could be curtailed when doing so made it easier for other, less powerful speakers to express themselves. Taken to its logi­cal extreme, this principle called into question the distinction between public and private power that is a constitutive feature of the modern First Amendment tradition. 49 See infra notes 138–140 and accompanying text.

Yet, notwithstanding the evident discomfort that some members of the Court demonstrated toward the broader implications of the antisub­ordinating view, in the decades after UAW-CIO was handed down, the Court repeatedly upheld laws that compensated for inequalities in social, political, and economic power by limiting the expressive freedom of powerful private speakers. 50 One can include on this list, besides Red Lion and UAW-CIO, NRLB v. Gissel Packing Co., 395 U.S. 575, 616 (1969), which I discuss below. See infra notes 183–186 and accompanying text. More generally, it crafted a free speech juris­prudence that was sensitive to economic and political inequality—that assessed the constitutionality of state action by examining the context in which it operated, and its effects, as well as its motivations and its form.

Beginning in the 1970s, however, the Court turned away from this power- and context-sensitive approach to the interpretation of freedom of speech. In many different areas of law, it began to insist—in contrast to the earlier cases—that the First Amendment poses little bar to well-intentioned government actions that have a disparate impact on the ability of some to communicate. 51 See infra notes 57–60 and accompanying text. It has instead interpreted the First Amendment to prohibit—and, for the most part, to prohibit only—gov­ernment actions that treat some speakers differently because of the viewpoint or subject matter of their speech, or because of other “suspect” characteris­tics, such as their institutional identity or their wealth. 52 See infra notes 68–74 and accompanying text.

The result has been to limit the effectiveness of the First Amendment as a tool for protecting the expressive freedom of those at the bottom of the economic and social hierarchies—those whose speech is most likely to be constrained by forces other than the discriminatory animus of government actors. It has also turned the First Amendment into a barrier to legislative efforts to protect the expressive freedom of the (relatively) poor and (relatively) powerless by limiting the expressive freedom of the richer and more powerful.

A. Campaign Finance Regulation

This is most obviously true when it comes to the Court’s campaign finance jurisprudence. Campaign finance regulation in the United States has always been justified, in whole or in part, by a desire to promote a rough kind of substantive equality in the market for political influence. 53 As early as the late nineteenth century, reformers argued that legislation was necessary to prevent wealth—particularly corporate wealth—from undermining the ability of the nonwealthy to influence politics. This sentiment was vigorously expressed by statesman Elihu Root in a 1894 speech that proved influential to the Court’s mid-twentieth-century cam­paign finance jurisprudence. See United States v. UAW-CIO, 352 U.S. 567, 571 (1957) (“The idea [behind campaign finance restrictions] is to prevent . . . the great railroad companies, the great insurance companies, the great telephone companies, the great aggregations of wealth from using their corporate funds . . . to send members of the legislature . . . to vote for their protection and the advancement of their interests . . . .” (second alteration in original) (internal quotation marks omitted) (quoting Elihu Root, Addresses on Government and Citizenship 143 (Robert Bacon & James Brown Scott eds., 1916))). By limiting the amount of money that can be spent on election-related speech, campaign finance laws attempt to ensure that the wealthy and the not so wealthy have at least somewhat equal opportunity to use money to influence both who is elected and how they govern once elected. Cam­paign finance has tended to promote this goal, however, by restricting the spending of only certain kinds of political actors—cor­porations, for exam­ple, and unions. Even when they apply a formally neutral rule, cam­paign finance laws uniquely burden the wealthy and are intended to do so.

For decades now, critics have argued that the fact that they inten­tionally burden only some speakers and not others means that campaign finance laws violate the First Amendment. Justice Douglas wrote a vigor­ous dissent in UAW-CIO in which he argued that the majority should have struck down the ban on union campaign spending on First Amendment grounds. 54 See UAW-CIO, 352 U.S. at 597 (Douglas, J., dissenting). Douglas adamantly rejected the argument that the ban could be justified by the need to prevent unions from using their considerable economic power to exert an “undue and disproportionate influence upon federal elections.” 55 Brief for the United States at 51, 52–57, UAW-CIO, 352 U.S. 567 (No. 44), 1956 WL 89052. Douglas argued that the fact that “one group or another . . . is too powerful,” like the fact that one group or another “advo­cates unpopular ideas,” cannot constitutionally justify preventing it from speaking or expending money on speech. 56 See UAW-CIO, 352 U.S. at 597 (Douglas, J., dissenting).

Until the early 2000s, this argument utterly failed to convince the Court that campaign finance laws necessarily violated the First Amendment—or even that that they did so in most cases. Although the Court acknowl­edged that constitutional interests were implicated when the government limited how much corporations or unions could spend on election-related speech, it insisted that “considerable deference” be given to the legislative judgment that “the particular legal and economic attributes of corpora­tions and labor organizations” justify their “particu­larly careful regulation.” 57 FEC v. Nat’l Right to Work Comm., 459 U.S. 197, 209–10 (1982).

This deference was not unlimited. In Buckley v. Valeo, the Court held that the government could not enact campaign finance laws in order to “equal­iz[e] the relative ability of individuals and groups to influence the outcome of elections.” 58 424 U.S. 1, 48 (1976) (per curiam). “[T]he concept that government may restrict the speech of some elements of our society in order to enhance the rela­tive voice of others,” the Court asserted, “is wholly foreign to the First Amendment.” 59 Id. at 48–49. Instead, the government had to demonstrate that cam­paign finance laws furthered some other, more compelling interest, such as the interest in pre­venting corruption or the appearance of corruption. 60 See id. at 25–27. It also had to show that the laws were “closely drawn” to advance those ends. 61 Id. at 25.

In later cases the Court interpreted this standard relatively loosely. 62 See Richard L. Hasen, Rethinking the Unconstitutionality of Contribution and Expenditure Limits in Ballot Measure Campaigns, 78 S. Cal. L. Rev. 885, 891 (2005) (dis­cussing the four post-Buckley cases—which he called the “New Deference Quartet”—in which the Court “markedly lowered the bar for upholding the constitutionality of cam­paign finance regulations in candidate campaigns”). And notwithstanding Buckley’s evident discomfort with the substantive-equality-promoting aspects of campaign finance law, the Court subse­quently interpreted the government’s interest in preventing political cor­ruption broadly, to include efforts to prevent the “corrosive and dis­torting effects of immense aggregations of wealth . . . accumulated with the help of the corporate form” from undermining the integrity of the demo­cratic political process. 63 Austin v. Mich. Chamber of Commerce, 494 U.S. 652, 660 (1990), overruled by Citizens United v. FEC, 558 U.S. 310 (2010); see also FEC v. Mass. Citizens for Life, Inc., 479 U.S. 238, 257 (1986) (describing the desire to “regulate the ‘substantial aggregations of wealth amassed by the special advantages which go with the corporate form of organization’” as an anticorruption interest (quoting FEC v. Nat’l Right to Work Comm., 459 U.S. 197, 207 (1982))). In effect, the Court defined the anti­corruption interest to include an interest in promoting equality. 64 David A. Strauss, Corruption, Equality, and Campaign Finance Reform, 94 Colum. L. Rev. 1369, 1369 n.1 (1994). One can well under­stand why. As Professor David Strauss has argued, corrup­tion is a serious concern in the campaign finance context largely because of the significant economic inequality that characterizes American society. 65 Id. at 1370 (arguing that corruption in the campaign finance context “is a deriva­tive problem” and that “[i]f somehow an appropriate level of equality were achieved, much of the reason to be concerned about corruption would no longer exist”). It is only in a context in which economic resources are unevenly distributed that the ability of some voters to use those resources to buy political influ­ence poses a serious threat to the representativeness of the system writ large. This means that a constitutional rule that recognizes, as Buckley did, that the government has a compelling interest in combating corruption because of the threat that corruption poses to “the integrity of our system of representative democ­racy” can easily be interpreted to mean that the government also has a compelling interest in combatting the influence of money on politics. 66 Buckley v. Valeo, 424 U.S. 1, 26–27 (1976) (per curiam). Certainly this is how the Court in later cases inter­preted the Buckley rule. The result was that, notwithstand­ing its strict lan­guage, in practice Buckley was not inter­preted as an insur­mountable con­straint on the govern­ment’s ability to limit the campaign spending of the very wealthy. 67 See, e.g., McConnell v. FEC, 540 U.S. 93, 203 (2003) (“Since our decision in Buckley, Congress’ power to prohibit corporations and unions from using funds in their treasuries to finance advertisements expressly advocating the election or defeat of candi­dates in federal elections has been firmly embedded in our law.”), overruled by Citizens United, 558 U.S. 310; Nixon v. Shrink Mo. Gov’t PAC, 528 U.S. 377, 391–92 (2000) (“While Buckley’s evidentiary showing exemplifies a sufficient justification for contribution limits, it does not speak to what may be necessary as a minimum.”).

And then Citizens United v. FEC came along. 68 558 U.S. 310. In that decision, the Court famously—perhaps infamously—held that the government’s inter­est in safeguarding elections from the corrupting and “distorting” effects of corporate wealth not only wasn’t compelling; it wasn’t even legiti­mate. 69 Id. at 348–50. Instead, the Court held that the government had a compelling interest in preventing only quid pro quo corruption—the exchange of money for specific favors—and its appearance. 70 Id. at 359 (“The hallmark of corruption is the financial quid pro quo: dollars for political favors[.]” (internal quotation marks omitted) (quoting FEC v. Nat’l Conservative Political Action Comm., 470 U.S. 480, 497 (1985))). “The fact that speakers may have influence over or access to elected officials does not mean that th[o]se officials are corrupt,” Justice Kennedy asserted for the Court. 71 Id. The Court also held that the government could not impose greater re­strictions on corporate speech than on the speech of natural individu­als. 72 Id. at 356 (overruling Austin v. Mich. Chamber of Commerce, 494 U.S. 652 (1990), on the grounds that the First Amendment prevents “the Government [from] sup­press[ing] political speech on the basis of the speaker’s corporate identity”). This is because, the Court explained, when the government “tak[es] the right to speak from some and giv[es] it to others, [it] deprives the dis­advantaged person or class of the right to use speech to strive to establish worth, stand­ing, and respect for the speaker’s voice.” 73 Id. at 340–41. It deprives the disadvantaged speakers, in other words, of “equality of status in the field of ideas.” 74 Police Dep’t of Chi. v. Mosley, 408 U.S. 92, 96 (1972) (internal quotation marks omitted) (quoting Alexander Meiklejohn, Political Freedom: The Constitutional Powers of the People 27 (1948)).

In an article published shortly after the decision was handed down, Professor Kathleen Sullivan argued that the majority opinion in Citizens United reflected the triumph of a libertarian, as opposed to an egalitarian, conception of the First Amendment guarantee of freedom of speech. 75 Sullivan, supra note 8, at 145. But in fact the central value the opinion vindicated was equality, not liberty. As Justice Stevens pointed out in his dissent, the “basic premise under­lying the Court’s ruling [was] its iteration, and constant reiteration, of the proposition that the First Amendment bars regulatory distinctions based on a speaker’s identity.” 76 Citizens United, 558 U.S. at 394 (Stevens, J., dissenting). Indeed, the Court did not hold that the govern­ment could never restrict election-related speech. It merely held that the government could not restrict election-related speech in order to pro­mote certain voices over others. It could not enact campaign-finance laws, in other words, to advance what the Court clearly considered to be a dis­criminatory aim (that is, reducing the political influence of the wealthy). Nor could it use discriminatory—that is, speaker-based—means of advanc­ing its legitimate aims. 77 Id. at 340 (“Premised on mistrust of governmental power, the First Amendment . . . [p]rohibit[s] . . . restrictions distinguishing among different speakers, allowing speech by some but not others.”). Indeed, the Court insisted that the gov­ernment could employ speaker distinctions only when doing so was necessary to the functioning of government institutions. 78 Id. at 341 (“The Court has upheld a narrow class of speech restrictions that oper­ate to the disadvantage of certain persons, but these rulings were based on an interest in allowing governmental entities to perform . . . functions . . . [that] cannot operate with­out some restrictions on particular kinds of speech.”).

What Citizens United demonstrates, therefore, is less a shift toward a libertarian conception of freedom of speech than a shift in the Court’s understanding of what an egalitarian First Amendment requires. In lieu of the more contextual approach it had previously taken, the Court now insisted on a much more formal equality rule: one that not only pre­vented the government from employing speaker-based distinctions when it regulated political spending but also defined the government’s anticorruption interest so narrowly as to prevent the enforcement of even many non-speaker-based laws that were intended to limit the influ­ence of wealth on politics. 79 See Michael S. Kang, The End of Campaign Finance Law, 98 Va. L. Rev. 1, 5 (2012). The result was a significant increase in spending on election campaigns—much of it coming from wealthy donors. 80 See, e.g., Bob Biersack, 8 Years Later: How Citizens United Changed Campaign Finance, Open Secrets (Feb. 7, 2018), [] (document­ing the growing influence of outside spending relative to total federal campaign spending).

The Court’s recent campaign finance jurisprudence thus provides a stark illustration of how the insistence on formal equality can actively interfere with the government’s ability to promote substantive equality—in this case, substantive equality of opportunity in the market for political influence. The law of campaign finance is not, however, the only area of First Amendment law in which the Court has embraced a much more formal conception of expressive equality than was previously the case, with similarly problematic consequences. Instead, the shift evident in Citizens United can be espied in many different areas of free speech doctrine.

B. Time, Place, and Manner Cases

Consider, for example, the body of First Amendment cases dealing with laws that regulate the time, place, or manner of speech. For over seventy years, the Court has recognized that the government may restrict when, where, and how individuals express themselves publicly, so long as the restrictions it imposes serve a legitimate—that is, noncensorial—pur­pose, are reasonably limited, and do not “unfair[ly] discriminat[e]” against any particular group or individual. 81 Cox v. New Hampshire, 312 U.S. 569, 576 (1941) (“If a municipality has authority to control the use of its public streets for parades or processions, as it undoubtedly has, it cannot be denied authority to give consideration, without unfair discrimination, to time, place and manner in relation to the other proper uses of the streets.”). Over time, however, the Court has interpreted the prohibition against discriminatory time, place, and manner laws in markedly different ways.

Originally, the Court interpreted the antidiscrimination principle to mean that the government could not enact a time, place, or manner law if doing so made it significantly harder for a particular subset of speakers to express themselves publicly. In Martin v. Struthers, the Court struck down a ban on the door-to-door distribution of pamphlets because it found that being able to distribute pamphlets door-to-door was crucial to the “poorly financed causes of little people.” 82 319 U.S. 141, 146 (1943). In the hotly contested license-tax cases it decided around the same time, members of the Court bitterly divided over whether the government could ever impose license taxes on expressive activity; however, all nine Justices agreed that the Court could not impose license fees that were too steep for some to pay. 83 See, e.g., Murdock v. Pennsylvania, 319 U.S. 105, 111 (1943) (“Freedom of speech, freedom of the press, freedom of religion are available to all, not merely to those who can pay their own way.”); Jones v. City of Opelika, 319 U.S. 103, 134–35 (1943) (Frankfurter, J., dissenting) (arguing that license taxes should be considered constitu­tional unless so large as to be “oppressive in their effect upon [expressive] activities”).

In these and other cases, the Court made clear that laws that regu­lated the time, place, or manner of speech were constitutional only if they left open what it later described as “ample alternative channels” for affected parties to communicate. 84 See, e.g., Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, Inc., 425 U.S. 748, 771 (1976) (“Whatever may be the proper bounds of time, place, and manner restrictions on commercial speech, they are plainly exceeded by this Virginia statute, which singles out speech of a particular content and seeks to prevent its dissemination completely.”). The Court also insisted that the gov­ernment’s justifications for regulating the time, place, and manner of speech had to be closely scrutinized to ensure that the real motivations for the law were not, in fact, a desire to suppress disfavored messages or—equally prob­lematic—unsubstantiated stereotypes about those they regulated. 85 Police Dep’t of Chi. v. Mosley, 408 U.S. 92, 100–01 (1972) (striking down a picket­ing law that prohibited nonlabor pickets after finding no evidence that nonlabor speech was, as the government claimed, categorically more likely to result in disruption than labor speech). For more discussion of this point, see Lakier, supra note 28, at 287–89.

The Court did not always enforce these requirements very rigor­ously. 86 See Geoffrey R. Stone, Content-Neutral Restrictions, 54 U. Chi. L. Rev. 46, 50 (1987) (“The Court does not seriously inquire into the substantiality of the governmental interest, and it does not seriously examine the alternative means by which the government could achieve its objectives. As a result, when the Court applies this standard, it invariably upholds the challenged restriction.”). And by the 1980s, it had almost entirely stopped enforcing them. 87 The one, notable exception to this general rule is the 1994 decision in Ladue v. Gilleo, which struck down a local ordinance that banned lawn signs because it provided insufficient alternatives for impacted citizens to communicate, particularly when they were “persons of modest means or limited mobility.” 512 U.S. 43, 57 (1994). The fact that the ordinance threatened the rights of homeowners—that is, property owners—may help explain the unusual rigor of the Court’s analysis. Although in theory, the government continued to have to demonstrate that its regulations left open ample alternative channels of communica­tion and represented a narrowly tailored response to an empirically verifiable problem, in practice, the Court tended to be “extraordinarily lenient” when determining whether these requirements were satisfied. 88 Robert Post, Recuperating First Amendment Doctrine, 47 Stan. L. Rev. 1249, 1263 (1995). It focused its analysis of the constitutionality of time, place, and manner laws instead on an orthogonal question: Namely, was the law in question “content-based” or “content-neutral”? In other words, did the law restrict speech because of its content or because of some other characteristic (for example, its location or its volume)? If the former, the Court struck the law down unless the government was able to show that it served a compelling interest by the least restrictive means possible. 89 United States v. Playboy Entm’t Grp., Inc., 529 U.S. 803, 813 (2000). If the latter, the Court often applied very deferential scrutiny. 90 Leslie Kendrick, Content Discrimination Revisited, 98 Va. L. Rev. 231, 237 (2012) (“[C]ontent-neutral laws receive what the Court calls ‘intermediate scrutiny,’ in practice a highly deferential form of review which virtually all laws pass.” (quoting Turner Broad. Sys., Inc. (Turner II) v. FCC, 520 U.S. 180, 189 (1997))).

In effect, the Court redefined what it means to say that a time, place, or manner law discriminates. No longer did the discrimination inquiry turn on an analysis of the law’s effects or the substantiality of its justifica­tions; what mattered instead were the formal distinctions the government employed and the legislative purposes that motivated the law. The conse­quence was, as in the campaign finance cases, to transform the guarantee of expressive equality into a guarantee of formal equality rather than something more substantive. So long as the government did not intend to suppress speech because of its content or employ an explicitly content-based distinction, the Court found that it could constitutionally deprive those it regulated of any realistic opportunity to express themselves. 91 See, e.g., Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989) (“A regulation that serves purposes unrelated to the content of expression is deemed neutral, even if it has an incidental effect on some speakers or messages but not others.”); City of Renton v. Playtime Theatres, Inc., 475 U.S. 41, 54 (1986) (concluding that a zoning law that applied to adult movie theatres left open ample channels of communication even though it left only five percent of city territory zoned for the theatres, none of which included—accord­ing to the Court of Appeals—any “commercially viable” land). It also could regulate on the basis of largely unsubstantiated apprehensions of harm. 92 For example, in Clark v. Community for Creative Non-Violence, the Court upheld a Park Service regulation that prevented protestors from sleeping in two national parks because the regulation reasonably advanced the government’s interest in administrative efficiency and would “limit the wear and tear on park properties.” 468 U.S. 288, 297–99 (1984). The majority reached this conclusion notwithstanding the fact that the govern­ment provided no evidence that allowing protestors to sleep in the park would present administrative problems or meaningfully increase wear and tear. Id. at 310–11 (Marshall, J., dissenting).

The result was in some respects the opposite of the result in the cam­paign finance cases: It made it easier for the government to justify regula­tions that had the effect of limiting certain speakers’ speech but not others’. But the mechanism was the same. Just as was true of its campaign finance cases, in its time, place, and manner cases, the Court now focused its analy­sis of the constitutionality of a given law almost entirely on the nature of the relationship it established between the government and the speaker rather than the context in which it operated or its effects.

The consequence of this transformation in the time, place, and manner cases was to make it significantly more difficult for litigants to use the First Amendment to challenge regulations that imposed a dispar­ate burden on their ability to communicate publicly—at least when those regulations did not happen to take a content-based form. Consider in this respect the 1984 decision in Members of the City Council v. Taxpayers for Vincent. 93 466 U.S. 789 (1984). The case involved a Los Angeles ordinance that absolutely prohibited the posting of signs on public property anywhere in the city. 94 Id. at 791–92. The plaintiffs—a group who wanted to put up signs to support an (ulti­mately unsuccessful) candidate for City Council—argued that the ordi­nance was unconstitu­tional because it restricted too much speech and did so to further an in­sufficiently weighty purpose: namely, the elimina­tion of “visual clutter” in the city. 95 Id. at 795–96. The Ninth Circuit agreed and struck the ordinance down, but the Supreme Court reversed. The Court acknowl­edged that in the past it “ha[d] shown special solicitude for forms of expression [such as signs] that are much less expensive than feasible alternatives and hence may be important to a large segment of the citi­zenry.” 96 Id. at 812 n.30. The Court noted, however, that “this solicitude has practical bound­aries.” Id. It nevertheless upheld the ordinance because it found there to be “not even a hint of bias or censorship in [its] enactment or enforce­ment” and that those affected by the ordinance could still communicate their messages by other means—for example, by posting signs on private property or by using handbills. 97 Id. at 804, 812. In fact, though, as Justice Brennan pointed out in dissent, the alternative channels of communication the Court pointed to were far from “ample” and far from adequate as alter­natives. 98 Id. at 819–20 (Brennan, J., dissenting). After all, not everyone has access to pri­vate property on which to post. 99 Id. at 820. Meanwhile, handbilling is a far more labor-intensive and expen­sive form of communication than the post­ing of signs. 100 Id. The result of the Court’s holding was therefore to make it much more difficult for poorly funded political groups like the plaintiffs to make use of a “critical [and inexpensive] . . . means of communication.” 101 Id. at 819.

Vincent demonstrates the danger to expressive interests posed by the formalism of the Court’s approach to the discrimination inquiry in its time, place, and manner cases. It makes clear how a First Amendment analysis that focuses—to the exclusion of almost all else—on the question of whether state action is motivated by discriminatory animus, or takes a content-based form, can end up rubberstamping what are in fact highly repressive speech policies. In fact, this type of analysis prevents courts from taking seriously the possibility that government officials might have “strong incentives to overregulate even in the absence of an intent to censor particular views.” 102 This is the case, Justice Marshall argued, because of the political realities of gov­ernment. Clark v. Cmty. for Creative Non-Violence, 468 U.S. 288, 315 (1984) (Marshall, J., dissenting) (arguing that the “incentive [to overregulate] stems from the fact that . . . the political power of [the general public] is likely to be far greater than that [of those who seek to use a particular forum for First Amendment activity]”).

C. Public Employee Speech

Consider one final example of the Court’s formalist shift: the public-employee speech cases. Until 2006, courts faced with First Amendment challenges involving the speech of government employees resolved these challenges by applying the Pickering balancing test—so named for the Warren Court decision in which the test was first applied. 103 Pickering v. Bd. of Educ., 391 U.S. 563 (1968). The test allowed the government to discipline employee speech when its interest “in promoting the efficiency of . . . public services” outweighed the em­ployee’s constitutionally protected interest “in commenting upon matters of public concern.” 104 Id. at 568. Although the Pickering test allowed the gov­ernment, when acting as employer, to restrict significantly more speech than it could when acting as sovereign, it prohibited the government from sanc­tioning employee speech merely because it disliked it or believed that it cast the government into disrepute. Private employers, of course, were not so limited. Nevertheless, because the Court recognized that a major purpose of the First Amendment is to “protect the free discussion of governmental affairs”—including the free discussion of “the manner in which government is operated or should be operated”—the Pickering test did not grant government employers the same free­dom. 105 Mills v. Alabama, 384 U.S. 214, 218 (1966). The doctrine recognized, in other words, the important institu­tional differences that distinguished government employers from private employers.

It did so, that is, until 2006, when, in Garcetti v. Ceballos, the Court significantly limited the reach of the Pickering test. 106 547 U.S. 410 (2006). In that case, a major­ity of Justices on the Court—in fact, the same Justices who signed on to Justice Kennedy’s opinion in Citizens United—held that Pickering does not apply when employees speak pursuant to their official duties because such speech is categorically beyond the scope of First Amendment concern. 107 Id. at 421.

The arguments the Court provided to explain why such speech is not entitled to constitutional protection are not very persuasive. Justice Kennedy, who wrote the opinion for the Court, just as he did in Citizens United, argued that the exemption of such speech from constitutional pro­tection was necessary to promote the efficiency of government ser­vices. 108 Id. at 418 (“Government employers, like private employers, need a significant degree of control over their employees’ words and actions [because] without it, there would be little chance for the efficient provision of public services.”). But of course, Pickering already required courts to consider the govern­ment’s interest in the efficient provision of services and to honor it when that interest outweighed the free speech interests at stake. 109 Pickering, 391 U.S. at 568.

Justice Kennedy also argued that limiting constitutional protection for employee speech made pursuant to official duties was necessary to prevent excessive judicial intervention into the inner workings of the gov­ernment. 110 Garcetti, 547 U.S. at 423. In theory, perhaps. And yet judges had had nearly forty years between the articulation of the Pickering balancing test and its constriction in Garcetti to intrusively intervene into the government work­place if they wanted to. There was no evidence that they had taken the opportunity to do so. Instead, by all accounts, judicial review of government-employee cases, including on-the-job speech cases, tended to be quite deferential. 111 See, e.g., Eugene Volokh, Freedom of Speech and Workplace Harassment, 39 UCLA L. Rev. 1791, 1820 n.130 (1992) (noting that “[t]he government as employer can ban even political discussions by its employees if they ‘disrupt[ ] the work of the office,’ or ‘discredit[ ] the office’ before the public, or ‘demonstrate[ ] a character trait that [makes the speakers] unfit to perform [their] work’” (quoting Rankin v. McPherson, 483 U.S. 378, 389 (1987))).

Finally, Justice Kennedy argued that Pickering need not apply to speech made pursuant to an employee’s official duties because when the employer disciplines the employee for that speech, the employer “does not infringe any liberties the employee might have enjoyed as a private citizen” but simply “exercise[s] . . . control over what the employer itself has commissioned or created.” 112 Garcetti, 547 U.S. at 421–22. This, too, is unpersuasive as an explana­tion for the change. After all, as the Court acknowledged, the Pickering test was not intended to protect only the individual employee’s rights. It was also intended to protect “the public’s interest in receiving the well-informed views of government employees engaging in civic discussion.” 113 Id. at 419. For the public, it should make no difference if the employee’s speech occurs pursuant to a job-related duty; the only thing that should matter is whether it touches on a genuine matter of public concern.

The only persuasive explanation for the change Garcetti wrought to public-employee speech doctrine is one that is not explicit in the text but suggested throughout it: namely, that in limiting the scope of the Pickering test, the Court sought to place government employers on a roughly equal footing with private employers vis-à-vis their employees. As Professor Kermit Roosevelt notes, public-employee speech doctrine has always “attempt[ed] to promote two different kinds of equality simultaneously”: 114 Kermit Roosevelt III, Not as Bad as You Think: Why Garcetti v. Ceballos Makes Sense, 14 U. Pa. J. Const. L. 631, 633 (2012).

First, the Court wants to promote equality between government and private employers with respect to control over the workplace and employee performance: The government employer should have manage­rial author­ity that at least resembles that of the private employer. Second, it wants to maintain equality between government employees and other citizens: Government employees should not be worse off in constitutional terms, namely, they should not be required to surrender their First Amendment rights as a condition of public employment. 115 Id.

Pickering reconciled these competing interests by means of a balanc­ing test. Garcetti does so by establishing a bright-line rule that grants gov­ernment employers the same total authority to discipline employees for their job-related speech that private employers have traditionally possessed.

The opinion thus articulates what we might think of as an equality principle—albeit one that is in many respects the inverse of the equality principle articulated in Citizens United. If Citizens United stands for the proposition that the government has no right to distinguish between pri­vate actors when it regulates the public sphere, 116 Citizens United v. FEC, 558 U.S. 310, 394 (2010) (Stevens, J., concurring in part and dissenting in part) (“The basic premise underlying the Court’s ruling is its iteration, and constant reiteration, of the proposition that the First Amendment bars regulatory distinctions based on a speaker’s identity, including its ‘identity’ as a corporation.”). Garcetti stands for the proposition that the government enjoys the same right as private employ­ers to control what occurs in the (now quasi-private) institutions it oper­ates. These propositions are complementary. Indeed, the logic of Garcetti explains the Court’s conclusion in Citizens United that, although speaker-based distinctions are not permitted when the government regulates the public forum, they are permitted when it regulates its own institutions. But if the one rule empowers private actors, the other does the reverse.

In both cases, however, the principle operates by ignoring important economic, political, or—in this case—institutional differences between the groups it equates for purposes of First Amendment analysis. Here, as elsewhere, the consequence has been to limit litigants’ ability to use the First Amendment as a tool for challenging entrenched power. Indeed, in the ten or so years since Garcetti was handed down, courts have routinely denied protection to whistleblowing employees who bring governmental misconduct to light. 117 Helen Norton has documented the effects of Garcetti on whistleblowers in a num­ber of recent articles. See, e.g., Helen Norton, Constraining Public Employee Speech: Government’s Control of Its Workers’ Speech to Protect Its Own Expression, 59 Duke L.J. 1, 4–5 (2009) (“Courts’ unblinking deference . . . allows government officials to punish, and thus deter, whistleblowing.”); Helen Norton, Shining a Light on Democracy’s Dark Lagoon, 61 S.C. L. Rev. 535, 546 (2010).

Garcetti and Citizens United demonstrate the problem with conceiving of the Court’s recent free speech jurisprudence as libertarian. In both opinions, the interest that informs the analysis is not liberty but equality. Further, the Court conceptualizes that interest formally rather than sub­stantively—that is to say, by focusing almost exclusively on the nature of the relationship between the government and the speaker rather than the context in which the speech occurs or the effects of the government’s actions on the broader speech environment. What both opinions make visible, in other words, is a commitment to formal equality that in prac­tice frequently redounds to the benefit of those with power. This is not sur­prising. After all, a conception of equality that requires the govern­ment to treat the powerful and the powerless alike provides powerful speakers (including powerful government speakers) tremendous oppor­tunity to use their economic and political resources not only to promote their own views but also to limit the ability of others to express theirs. For those committed to a view of the First Amendment as a check on—rather than a handmaiden to—power, this state of affairs is deeply troubling.

II. The Antisubordinating First Amendment that Was

The good news is that things have been, and therefore could again be, different. As the previous Part suggests, the tendency of contempo­rary free speech law to reinforce rather than combat existing inequalities in wealth and power is the product of a relatively recent shift in the Court’s conception of what it means to guarantee “equality of status in the field of ideas,” 118 Police Dep’t of Chi. v. Mosley, 408 U.S. 92, 96 (1972) (internal quotation marks omitted) (quoting Alexander Meiklejohn, Political Freedom: The Constitutional Powers of the People 27 (1948)). not an inevitable feature of the modern First Amendment tradition. Earlier cases—even the relatively restricted sample of cases resolved by the Supreme Court—sometimes applied a substantially different approach to the First Amendment analysis than the approach courts take today. This is true not only of the three areas of free speech law canvassed in the previous Part, although it is certainly true of those areas. But in other areas of First Amendment jurisprudence, as well, one can discern in the older cases an often significantly different understanding of what it means to guarantee freedom of speech: one much more willing to take into account inequalities in economic and political power and much more sensitive to the disparate effects that formally neutral and well-intentioned laws can have on the ability of the “little people” to communicate.

These earlier cases suggest what an antisubordinating First Amendment might look like. They also make clear what the previous Part already hinted at: namely, that the regressive tendencies of contemporary First Amendment law cannot solely, or even primarily, be blamed on the “im­perial” or libertarian tendencies of the Roberts Court—that is to say, on its tendency to interpret the guarantee of freedom of speech too expan­sively and rigorously. This is because, as they show, our contempo­rary First Amendment is in many respects not very expansive and not always very rigorous. It is in many respects much narrower and much weaker than the First Amendment that existed fifty years ago—at least with respect to certain kinds of free speech claims (those that push against, rather than reinforce, property rights). What this suggests is that the answer to the ills that beset contemporary free speech law is not less constitu­tional protection for speech but a different kind of constitutional protec­tion: one that reduces, rather than reinforces, the inequalities in expressive opportunity that are a consequence of the highly, and increas­ingly, unequal distribution of economic and political power in the United States.

A. Speech on Private Property

To see the far-from-imperial nature of contemporary free speech jurisprudence, one need only look at the cases involving First Amendment rights of access to private property. Today, speakers have virtually no fed­eral constitutional right of access to privately owned spaces, like shopping malls. 119 See Hudgens v. NLRB, 424 U.S. 507, 521 (1976) (holding that the “constitutional guarantee of free expression has no part to play” in cases involving rights of access to pri­vate property). The only exception to this general rule is when the private prop­erty owner exercises “powers traditionally exclu­sively reserved to the State.” 120 Jackson v. Metro. Edison Co., 419 U.S. 345, 352 (1974). This was not always the rule, however. In Marsh v. Alabama, the Court held that a state trespass law could not con­stitutionally be used to exclude a Jehovah’s Witness who wished to distrib­ute religious literature on the sidewalks of a company-owned town. 121 326 U.S. 501, 509–10 (1946). Two decades later, in Amalgamated Food Employees Union Local 590 v. Logan Valley Plaza, Inc., the Court similarly concluded that state courts could not constitutionally enjoin members of a union from picketing a store located in a privately owned shopping mall. 122 391 U.S. 308, 319–20 (1968), overruled by Hudgens v. NLRB, 424 U.S. 507 (1976). During this period, lower courts also found that speakers possessed a First Amendment right of access to migrant labor camps, as well as to certain portions of their employer’s property. See, e.g., NLRB v. Lake Superior Lumber Corp., 167 F.2d 147, 150–51 (6th Cir. 1948); Folgueras v. Hassle, 331 F. Supp. 615, 623 (W.D. Mich. 1971); People v. Rewald, 318 N.Y.S.2d 40, 45–46 (Cty. Ct. 1971).

Today, these cases are usually celebrated—or derided—for establish­ing (in the case of Marsh) or extending (in the case of Logan Valley) the principle that when a private actor performs a sufficiently important gov­ernmental function, it occupies the role of a state actor for constitutional purposes. 123 See, e.g., Jody Freeman, The Private Role in Public Governance, 75 N.Y.U. L. Rev. 543, 578 n.131 (2000) (noting Marsh as the first application of the public-function test for state action); Frederick F. Schauer, Hudgens v. NLRB and the Problem of State Action in First Amendment Adjudication, 61 Minn. L. Rev. 433, 453–55 (1977) (criticizing Logan Valley for extending the governmental-function test too far). The Court has repeatedly encouraged this framing. 124 When citing Marsh, this is invariably how the Court describes its holding. See, e.g., Edmonson v. Leesville Concrete Co., 500 U.S. 614, 621–22 (1991) (describing Marsh as standing for the proposition that state action applies when “the actor is performing a traditional governmental function”); Lloyd Corp. v. Tanner, 407 U.S. 551, 562 (1972) (finding that Marsh “simply held that where private interests were substituting for and performing the customary functions of government, First Amendment freedoms could not be denied where exercised in the customary manner on the town’s sidewalks and streets”). However, it represents a significant distortion of what the opinions actually say.

In fact, neither decision depended on the conclusion that the pri­vate property owner occupied the role of state actor. To the contrary, the Court assumed in both cases that it was the judicial enforcement of a state property law that satisfied the First Amendment’s state action require­ment, not the actions of the private property owner. 125 In his majority opinion in Marsh, Justice Black framed the issue raised by the case as “whether a State, consistently with the First and Fourteenth Amendments, can impose criminal punishment on a person who undertakes to distribute religious literature on the premises of a company-owned town contrary to the wishes of the town’s management.” Marsh, 326 U.S. at 502. In his dissenting opinion, Justice Reed framed the issue as whether the First Amendment required a state to “commandeer, without compensation, the private property of other citizens” to “furnish[ ] the opportunity for information, education and religious enlightenment.” Id. at 515 (Reed, J., dissenting). Despite important differences between the two framings, both assumed that the state court, not the private company, constituted the state actor to whom the First Amendment applied. A similar assumption informed the majority and Justice Black’s dissent in Logan Valley. Indeed, neither opinion even raised the possibility that the state court’s decision to enjoin picketing in the mall failed to satisfy the First Amendment’s state action requirement. See Logan Valley, 391 U.S. at 309 (framing the issue as whether “the decisions of the state courts [to enjoin the peti­tioners’] picketing as a trespass [were] violative of their rights under the First and Fourteenth Amendments”); id. at 329 (Black, J., dissenting) (arguing that “the lower court’s injunction [was] valid under the First Amendment”). What both deci­sions depended on instead was finding that the private property at issue served an important “public function”—and did so for the eco­nomic benefit of its owner. 126 See Logan Valley, 391 U.S. at 318 (“The general public has unrestricted access to the mall property. The shopping center here is clearly the functional equivalent of the business district of Chickasaw involved in Marsh.”); Marsh, 326 U.S. at 507–08 (“Whether a corporation or a municipality owns or possesses the town the public . . . has an identical interest in the functioning of the community [such] that the channels of communication remain free. . . . The ‘business block’ serves as the community shopping center and is freely accessible and open . . . .”). What this meant, the Court concluded in both cases, was that the owner’s property rights had “become circum­scribed by the statutory and constitutional rights of those who use[d its property].” 127 Logan Valley, 391 U.S. at 325 (citing Marsh, 326 U.S. at 506).

In holding as much, the Court in some senses merely extended a principle it developed in a series of early-twentieth-century cases dealing with speech on publicly owned streets, parks, and sidewalks: namely, that when it comes to spaces that serve as important sites of public expression (what would come to be known as “public forums”), the public’s right of access outweighs the right of the property owner to exclude. 128 See, e.g., Jamison v. Texas, 318 U.S. 413, 413–14 (1943); Hague v. Comm. for Indus. Org., 307 U.S. 496, 515–16 (1939); Schneider v. State, 308 U.S. 147, 164–65 (1939). Professor Harry Kalven first coined the term “public forum.” See Harry Kalven, Jr., The Concept of the Public Forum: Cox v. Louisiana, 1965 Sup. Ct. Rev. 1, 11–12 (“[I]n an open democratic society the streets, the parks, and other public places are an important facility for public discussion and politi­cal process. They are in brief a public forum that the citizen can commandeer . . . .”). But it also pushed strongly against a formalist conception of the First Amendment equality guarantee. After all, in neither case was there any suggestion that the state courts charged with enforcing the private owner’s property rights did so in a discriminatory manner or that the laws themselves were content based. Nevertheless, in both cases, the Court held that the courts’ actions violated the First Amendment because its effect was to make some members of the public—those who because of geography or economic circumstances patronized privately owned, as opposed to gov­ernment-owned, public spaces—second-class citizens when it came to their First Amendment rights.

Indeed, both opinions were very explicit about the fact that the constitutional problem with the government’s actions in both cases was its impact on the substantive equality of those it affected. “Many people,” Justice Black noted in his opinion in Marsh, “live in company towns.” 129 Marsh, 326 U.S. at 508. Were it the case that the principles that governed the public forum cases did not protect them merely because they spoke on privately owned streets, he argued, it would be much more difficult for these “free citizens” to access the uncensored information necessary to “make decisions [about] the welfare of [their] community and nation.” 130 Id. It would block up the “channels of communication” in those spaces in which residents of company towns happened to live and work, thereby depriving them of access to the same information that those who hap­pened to patronize publicly owned spaces enjoyed—and this the Court refused to allow. 131 Id. at 507. “There is no more reason for depriving these people of the liberties guaranteed by the First and Fourteenth Amendments,” Black insisted, “than there is for curtailing these freedoms with respect to any other citizen.” 132 Id. at 508–09.

Justice Marshall, who wrote the opinion for the Court in Logan Valley, similarly noted that denying speakers a constitutional right of access to privately owned shopping malls would make it much more diffi­cult “for workers seeking to challenge substandard working conditions, consumers protesting shoddy or overpriced merchandise, and minority groups seeking nondiscriminatory hiring policies” at a mall shop to do so than it was for those who wished to communicate the exact same message about a shop located on a public street. 133 Amalgamated Food Emps. Union Local 590 v. Logan Valley Plaza, Inc., 391 U.S. 308, 324–25 (1968), overruled by Hudgens v. NLRB, 424 U.S. 507 (1976). “Neither precedent nor pol­icy,” Marshall concluded, “compels a result so at variance with the goal of free expression and communication that is the heart of the First Amendment.” 134 Id. at 325.

By reconceiving Marsh as a state action case, the Court significantly limited its reach and transformed its meaning. It rendered it simply another illustration of the well-established formal-equality rule that content-based discrimination at the hands of a state actor violates the First Amendment guarantee, except in the most exceptional circumstances. Transforming Marsh into a state action case also made it easy to explain why Logan Valley was incorrect and had to be overruled—which it was, just eight years after it was handed down. 135 See Hudgens, 424 U.S. at 520–21. After all, running a shopping mall is not an important state function, let alone something that has tended to be exclusively (if ever) performed by the government. Read on their own terms, however, Marsh and Logan Valley point to a much more expansive conception of the guarantee of expressive equality: one that prohibits the enforcement of even well-intentioned laws that have a disparate effect on the ability of some members of the public to access the “channels of communication” or otherwise participate in public debate.

Scholars have criticized both decisions, but particularly Logan Valley, for defining the scope of the First Amendment’s application too elas­tically. Professor Frederick Schauer, for example, has argued that the Logan Valley Court erred by extending First Amendment access rights to speakers on private property even when the owner of that property does not possess “powers equivalent to those of the state.” 136 Schauer, supra note 123, at 454. This is a mistake, Schauer argues, because it threatens the vitality of the marketplace of ideas by depriving private property owners of a right the First Amendment permits them: namely, the right to discriminate against speech because they dislike it, fear it, or for any other reason. 137 Id. at 450.

It is certainly true that a central assumption of modern free speech law is that the First Amendment is not intended to prohibit private dis­crimination. Quite the contrary: By prohibiting the government from dis­criminating when it comes to speech, the First Amendment is sup­posed to encourage private discrimination by protecting speakers against retalia­tion from the government for the expressive choices they make and the stances they choose to embrace or reject. As Schauer notes: “[T]he ideals of free public debate and a marketplace of ideas presume that there will be partisanship and preference for some ideas over others”—that private persons can and will discriminate, even though the government won’t and shouldn’t. 138 Id.

But nothing in either Marsh or Logan Valley contradicts this general principle. Because neither decision presumed that the private property owner was a state actor, neither decision requires that the same constitu­tional constraints that apply to the government as owner of public prop­erty apply to those who own company towns or shopping malls. In fact, in Logan Valley, the Court explicitly acknowledged as much. 139 Logan Valley, 391 U.S. at 319 (“[I]t may well be that respondents’ ownership of the property here in question gives them various rights, under the laws of Pennsylvania, to limit the use of that property by members of the public in a manner that would not be permissible were the property owned by a municipality.”). The Court reaffirmed this point several years later when it held that, even if mall owners could not be compelled by the First Amendment to grant access to their property, they might be compelled to do so by their state constitutions. 140 PruneYard Shopping Ctr. v. Robins, 447 U.S. 74, 81 (1980) (“Our reasoning in Lloyd . . . does not ex proprio vigore limit the authority of [a] State . . . to adopt in its own Constitution individual liberties more expansive than those conferred by the Federal Constitution.”). The Court made clear, however, that this could be true only in cases in which the grant of access did not undermine property owners’ expressive freedom by forcing the owners to associate with views they reject. 141 Id. at 87. It recognized that private property owners enjoy a degree of constitutional protection that the government, when it regulates the public forum, absolutely does not possess. 142 Indeed, if anything can be said to be a core principle of modern free speech law, it is that the government may not exclude speech from the public forum to promote its own ideas. See, e.g., W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 642 (1943) (“If there is any fixed star in our constitutional constellation, it is that no official, high or petty, can prescribe what shall be orthodox in politics, nationalism, religion, or other matters of opinion or force citizens to confess by word or act their faith therein.”).

Marsh and Logan Valley do not, in other words, obliterate the distinc­tion between state and private action upon which the First Amendment depends. What they suggest instead is that the scope of First Amendment application need not be coterminous with state property rules and that facially neutral and well-intentioned laws might violate the First Amendment because of their disparate effects.

Properly understood, both cases represent in this respect what we might describe with not too much overstatement as a radical—and much more speech-protective—alternative to the current approach to constitu­tional claims of access to private property. It is an approach, nonetheless, that emerges out of, and is entirely compatible with, the foundational assumptions of modern free speech law—chief among these being that the purpose of the First Amendment is to create an “uninhibited, robust, and wide-open” 143 N.Y. Times Co. v. Sullivan, 376 U.S. 254, 270 (1964). public sphere in which all kinds of viewpoints, and classes of people, can participate.

B. Media-Access Cases

A similarly substantive vision of the First Amendment’s equality guar­antee—and a similar willingness to prioritize the expressive rights of the public over private property rights—characterizes the sequence of cases, beginning with Red Lion Broadcasting Co. v. FCC in 1969 144 395 U.S. 367 (1969). and ending with the decision in Turner Broadcasting System, Inc. v. FCC in 1997, 145 520 U.S. 180 (1997). in which the Court held that Congress could require media companies to transmit speech not of their choosing when doing so promoted an “uninhibited,” as opposed to a “monopoliz[ed],” marketplace of ideas. 146 Red Lion, 395 U.S. at 388–90; see also Denver Area Educ. Telecomms. Consortium, Inc. v. FCC, 518 U.S. 727 (1996); Turner Broad. Sys., Inc. v. FCC (Turner I), 512 U.S. 622 (1994); Columbia Broad. Sys., Inc. v. Democratic Nat’l Comm., 412 U.S. 94 (1973). These are amazing cases to read today, given the Court’s renewed insistence that “[t]he concept that government may restrict the speech of some . . . in order to enhance the relative voice of others” is one that “is wholly foreign to the First Amendment.” 147 Citizens United v. FEC, 558 U.S. 310, 349–50 (2010) (internal quotation marks omitted) (quoting Buckley v. Valeo, 424 U.S. 1, 48–49 (1976) (per curiam)); see also Ariz. Free Enter. Club’s Freedom Club PAC v. Bennett, 564 U.S. 721, 741 (2011). In fact, all of these cases endorse, to one degree or another, this supposedly foreign concept.

This is obviously true of Red Lion. In that case, the Court upheld against constitutional challenge a series of FCC regulations that required broad­casters to give those they criticized, or the opponents of political candidates they endorsed, a right of reply. 148 Red Lion, 395 U.S. at 373–74. A Pennsylvania radio station and its president argued that the rules violated the First Amendment by denying them rights that other speakers enjoyed: namely, the right to “refus[e] in [its] speech or other utterances to give equal weight to the views of [its] opponents.” 149 Id. at 386 (“No man may be prevented from saying or publishing what he thinks, or from refusing in his speech or other utterances to give equal weight to the views of his opponents. This right, they say, applies equally to broadcasters.”). Their claim, in other words, was that the rules under­mined the constitutionally protected right of broadcasters to privately discriminate, thus violating their formal equality. The Court rejected this argument. It concluded that the rules “enhance[d] rather than abridge[d] the freedoms of speech and press protected by the First Amendment.” 150 Id. at 375. It justified this conclusion on a number of grounds.

First, it argued that radio licenses were scarce and the privilege of owning one was not constitutionally guaranteed. The fact that only rela­tively few people could exercise the right to speak in this manner, the Court argued, meant that those who did could be required to “share [their] frequency with others and to conduct [themselves] as a proxy or fiduciary with obligations to present those views and voices which are rep­resentative of [the] community and which would otherwise, by necessity, be barred from the airwaves.” 151 Id. at 388–89. The Court concluded, in other words, that under conditions of scarcity, the voices of some could be restricted in order to enhance the voices of others. It reached this conclusion because it believed—not implausibly—that any other conclusion would threaten the vitality of the marketplace of ideas by allowing it to be dominated by a few powerful voices.

This is a tenable argument—and it was certainly one that, at the time Red Lion was handed down, had been made before. Twenty-five years ear­lier, Justice Frankfurter had relied on a similar scarcity rationale to sus­tain another set of FCC rules against First Amendment challenge. 152 See Nat’l Broad. Co. v. United States, 319 U.S. 190, 226–27 (1943) (“Unlike other modes of expression, radio inherently is not available to all. That is its unique characteris­tic, and that is why . . . it is subject to governmental regulation. . . . The right of free speech does not include, however, the right to use the facilities of radio without a license.”).

The problem with the argument was that, at the time Red Lion was decided, it was no longer true that radio licenses were scarce—or, at least, that the scarcity of radio licenses made radio any different than other media of communication. As the radio station noted in its brief to the Court, by 1967, there were over three times as many commercial radio and television stations in the United States as there were daily news­papers. 153 Brief for Petitioners at 35, Red Lion, 395 U.S. 367 (No. 2), 1968 WL 129369 (reporting census findings that, as of 1967, there were “6253 commercial radio and televi­sion stations in the United States, compared to only 1754 daily newspapers, a ratio of bet­ter than 3½ to 1”). The Court’s insistence that, notwithstanding these facts, radio remained a scarce re­source produced significant criticism, both at the time and in the years to follow. 154 See, e.g., Lee C. Bollinger, Jr., Freedom of the Press and Public Access: Toward a Theory of Partial Regulation of the Mass Media, 75 Mich. L. Rev. 1, 10 (1976); Thomas W. Hazlett, Physical Scarcity, Rent Seeking, and the First Amendment, 97 Colum. L. Rev. 905, 913 (1997); L.A. Powe, Jr., “Or of the [Broadcast] Press,” 55 Tex. L. Rev. 39, 55–57 (1976).

In fact, the scarcity of radio licenses was not the only reason the Court provided to justify its conclusion that the FCC rules enhanced, rather than abridged, freedom of speech, although it was certainly the one the Court emphasized the most. The opinion also noted that “[e]ven where there are gaps in spectrum utilization”—and therefore, licenses that were not scarce—“the fact remains that existing broadcasters have often attained their present position because of their initial govern­ment selection in competition with others before new technological advances opened new opportunities for further uses.” 155 Red Lion, 395 U.S. at 400. It continued:

Long experience in broadcasting, confirmed habits of listeners and viewers, network affiliation, and other advantages in pro­gram procurement give existing broadcasters a substantial advantage over new entrants, even where new entry is techno­logically possible. These advantages are the fruit of a preferred position conferred by the Government. Some present possibility for new entry by competing stations is not enough, in itself, to render unconstitutional the Government’s effort to assure that a broadcaster’s programming ranges widely enough to serve the public interest. 156 Id.

As this passage makes clear, the purported scarcity of radio licenses was not crucial to the Court’s holding in the case. What was crucial was the Court’s belief that inequalities produced by both technological and his­torical factors had given the established networks significant power to decide what the public heard when it turned on the radio. The Court recognized that, given barriers of entry, this disparity in power was hard to redress by means of market competition. And the Court acknowledged that the government itself had played an important role in fostering ine­qualities in the industry through its initial conferral of licenses. It was in this context that the Court concluded that the First Amendment not only allowed the FCC to mandate a right of reply but might even require it. As Justice White wrote in his majority opinion:

It is the purpose of the First Amendment to preserve an unin­hib­ited marketplace of ideas in which truth will ultimately pre­vail, rather than to countenance monopolization of that market, whether it be by the Government itself or a private licensee. . . . It is the right of the public to receive suitable access to social, political, esthetic, moral, and other ideas and experiences which is crucial here. That right may not constitutionally be abridged either by Congress or by the FCC. 157 Id. at 390 (citations omitted).

This is an unabashedly antisubordinating conception of the First Amendment—one that adamantly rejects the idea that what freedom of speech requires is formally equal treatment of all speakers. To the con­trary: Like the most radical voices in the affirmative action debate, this conception suggests that vindicating the constitutional guarantee of equality might require policies designed to ameliorate significant inequali­ties in access to valuable resources—at least in cases where the govern­ment, in its initial conferral of property rights, helped create those inequalities. It certainly makes clear that, in cases where inequalities in power and resources undermine the vibrancy and diversity of public debate, the govern­ment can restrict the speech of some in order to enhance the speech of others.

It is also, of course, a vision of the First Amendment that the Court soon renounced—not only in Buckley but also two years prior to that, in Miami Herald Publishing Co. v. Tornillo, when it struck down a Florida law that imposed a right-of-reply requirement on newspapers. 158 418 U.S. 241, 258 (1974). As Justice Burger made clear in his opinion in that case, the dysfunctions in the newspaper industry that the Florida right-of-reply law was designed to ameliorate were remarkably similar to the dysfunctions in the radio business that led the FCC to enact the rules challenged in Red Lion. Just like the radio industry, the newspaper industry was also highly concen­trated. 159 Id. at 249–50 (“The elimination of competing newspapers in most of our large cities . . . [is an] important component[ ] of this trend toward concentration of control of outlets to inform the public.”). Moreover, Burger noted that the same interests that owned radio and television stations also frequently owned newspapers too—the industries were not merely parallel, they were tightly interlinked. Id. As a result, here too “the power to inform the American people and shape public opinion” had been “place[d] in a few hands.” 160 Id. And just as was true of radio, the problem of concentration in the newspaper industry was not easily solved by the ordinary practices of market compe­tition, given the steep barriers to entry into the industry. 161 See id. at 251. Nevertheless, in Tornillo, the Court concluded that the First Amendment prevented Florida from doing virtually anything to correct those problems and insisted that in this context, it was the right of the speaker, not the lis­tener, that was paramount. “A newspaper,” Burger wrote, “is more than a passive receptacle or conduit for news, comment, and advertising.” 162 Id. at 258. It instead is the product of “editorial control and judg­ment.” 163 Id. What this meant was that, just as the First Amendment forbade the government from telling individual persons what they could say, so too it prevented the government from limiting the ability of newspaper editors (companies, really) to determine what the newspaper would say.

It is hard to imagine a more dramatic rejection of the logic of Red Lion. In place of that opinion’s concern with inequalities in access and power, Burger’s opinion in Tornillo articulated a strict formal-equality rule. Indeed, Burger made clear that even if Florida’s right-of-reply law imposed no costs on newspaper editors—and therefore did not chill expres­sion by forcing editors to choose between publishing critical opin­ions and having to give up valuable space to replies (as the Florida law was said to do)—a right-of-reply requirement would offend the First Amendment by infringing on the editorial freedom of newspapers. 164 Id.

As such, Tornillo appeared to provide newspapers a powerful shield against not only right-of-reply laws but all kinds of access requirements. 165 See, e.g., Passaic Daily News v. NLRB, 736 F.2d 1543, 1557 (D.C. Cir. 1984) (relying on Tornillo to deny the NLRB the power to order a company to reinstate a news­paper columnist whose column was taken away in retaliation against his union activities). And not only newspapers. Nothing in its logic depended on the fact that the media corporation involved in the case was a newspaper. What mat­tered instead was that the corporation performed an editorial func­tion. 166 See Tornillo, 418 U.S. at 256–58. And indeed, in subsequent years, courts relied on the decision to invali­date access laws that applied to other kinds of media, as well as non­media corporations. 167 See, e.g., Pac. Gas & Elec. Co. v. Pub. Utils. Comm’n of Cal., 475 U.S. 1, 11–12 (1986) (electric utility); Zhang v. Inc., 10 F. Supp. 3d 433, 436–37 (S.D.N.Y. 2014) (search engine).

Tornillo’s reach remained limited in two important respects, however. First, the opinion did not overrule (or in fact mention) Red Lion. In sub­sequent cases, the Court affirmed that Red Lion remained good law. 168 See FCC v. Nat’l Citizens Comm. for Broad., 436 U.S. 775, 794–95 (1978). This meant that the FCC retained authority to regulate broadcast media in order to vindicate the right of the public to “an uninhibited market­place of ideas.” 169 Red Lion Broad. Co. v. FCC, 395 U.S. 367, 390 (1969). Second, in subsequent years, the Court proved unwill­ing to extend Tornillo as far as its logic suggested.

Consider in this respect the Court’s 1994 and 1997 decisions in Turner Broadcasting System, Inc. v. FCC. 170 Turner II, 520 U.S. 180 (1997); Turner I, 512 U.S. 622 (1994). In that case, the Court was con­fronted with a First Amendment challenge to a federal law that required cable companies to devote a portion of their channels to local broadcast networks. 171 See Turner I, 512 U.S. at 633–34. The government argued that Red Lion should control the analysis and that the law was therefore constitutionally unproblematic because, although it constricted the freedom of cable providers, it pro­moted a vibrant marketplace of ideas by preventing cable companies from using their market power to drive broadcast competitors out of business. 172 See id. at 637. The cable companies argued instead that Tornillo should con­trol and that the law was therefore unconstitutional. 173 Id. at 653.

The Court rejected both arguments. It noted that technological advances meant that soon there “may be no practical limitation on the number of speakers who may use the cable medium,” and concluded that Red Lion was therefore “inapt.” 174 Id. at 639. It argued, meanwhile, that Tornillo did not control, due to the “important technological difference between news­papers and cable television,” specifically noting that “[a]lthough a daily newspaper and a cable operator both may enjoy monopoly status in a given locale, the cable operator exercises far greater control over access to the relevant medium.” 175 Id. at 656. This was the case, the Court explained, be­cause “[w]hen an individual subscribes to cable, the physical connection between the television set and the cable network gives the cable operator bottleneck, or gatekeeper, control over most (if not all) of the television programming that is channeled into the subscriber’s home.” 176 Id. Instead, the Court upheld the law as a reasonable means by which Congress attempted to further what it described as “a governmen­tal purpose of the highest order”: namely, ensuring “the widest possible dissemination of information from diverse and antagonistic sources.” 177 Turner II, 520 U.S. 180, 190–92 (1997).

The arguments the Turner Court made to distinguish the case from Tornillo are, to put it mildly, unconvincing. As Professor Yochai Benkler has pointed out, it is not the fact that they use physical cables to communicate to their subscribers that vests cable companies with what the Court called “gatekeeper . . . control” of a “critical pathway of communication.” 178 See Yochai Benkler, Free as the Air to Common Use: First Amendment Constraints on Enclosure of the Public Domain, 74 N.Y.U. L. Rev. 354, 374 (1999). After all, competitors could simply install their own cables in potential subscribers’ houses, thereby making it much easier for customers to switch providers. What gives cable companies such power are instead the economic conditions of the industry—specifically, the “large fixed costs of wiring a city, and the relatively low incremental costs of distributing information once a city is wired.” 179 Id. And certainly Justice Burger would have been surprised to learn that the very serious problems of concentration in the newspaper industry he documented in his opin­ion in Tornillo were not so serious because they did not have a physical cause.

In fact, the opinion represents an approach to First Amendment media-access questions that is strikingly similar to the reasoning of Red Lion—and one that is quite difficult to reconcile with the analysis in Tornillo. Even as it denied that the deferential approach the Court took in Red Lion to regulations aimed at fostering diversity in the broadcast media could be applied to regulations of cable, the Court employed essentially the same argument as that made in Red Lion to uphold the cable law against constitutional challenge. “The potential for abuse of [the cable companies’] private power over a central avenue of communi­cation cannot be overlooked,” Justice Kennedy asserted in his majority opinion in Turner. 180 Turner I, 512 U.S. at 657. “The First Amendment’s command that govern­ment not impede the freedom of speech does not disable the gov­ern­ment from taking steps to ensure that private interests not restrict . . . the free flow of information and ideas.” 181 Id.

As in Red Lion, the Court recognized that the government could limit the expressive freedom of some, when they possessed dis­proportion­ate gatekeeping power, in order to promote the interests of the public writ large. 182 See id. at 656 (“[S]imply by virtue of its ownership of the essential pathway for cable speech, a cable operator can prevent its subscribers from obtaining access to pro­gramming it chooses to exclude.”). As in Red Lion, it attempted to cabin the reach of its holding by linking it to the unique physical characteristics of the medium in ques­tion. And, as in Red Lion, its attempt to do so proved far from persuasive. Just as was true of Red Lion, however, the fact that the Turner Court misidentified the cause of the inequality it recognized takes away nothing from its legal argument. It merely suggests that the principle on which it relied should apply more broadly than the Turner Court was willing to acknowledge.

Like Red Lion, the decision in Turner is best understood as express­ing a principle that is not easily cabined: namely, that constraints on the expressive freedom of individual speakers may be permissible when those speakers possess significant power to dominate the channels of communi­cation and the regulations reasonably can be expected to pro­mote (rather than hinder) a diverse marketplace of ideas. Construed as such, the deci­sion—like the decision in Red Lion—echoes the approach to the First Amendment analysis applied by the Court in Marsh and Logan Valley. Like those decisions, it forgoes formal equality to promote a more egal­itarian and therefore more vibrant public sphere.

C. Employer Speech

Consider one last example of the antisubordinating tradition in First Amendment law—this time, a case that is far less beloved by progressives than Marsh and Red Lion. This is the decision in NLRB v. Gissel Packing Co., in which the Court upheld NLRB regulations that prohibited employers, during the course of union elections, from communicating to employees either a “threat of reprisal or force or promise of benefit.” 183 395 U.S. 575, 617 (1969) (internal quotation marks omitted) (quoting National Labor Relations Act, Pub. L. No. 74-189, 49 Stat. 449 (codified as amended at 29 U.S.C. § 158(c) (2012))).

In upholding the regulation, the Court departed quite significantly from the principles that ordinarily govern the regulation of political speech. Although First Amendment doctrine has recognized since the early twentieth century that “true threats” may be punished without vio­lating the freedom of speech, by the time Gissel Packing was handed down, it was clear that the category of “true threats” was much narrower than the relatively broad category of threatening communications the NLRB regulations prohibited. 184 By 1969, the Court had made clear that the low-value category of “true threats” did not extend to even “vituperative [and] abusive” speech that was “expressly condi­tional.” Watts v. United States, 394 U.S. 705, 708 (1969) (per curiam). However, the employer’s speech in Gissel Packing was expressly conditional. As Professor Julius Getman notes, “[A]t no point did the employer directly threaten to close the plant or take eco­nomic reprisals in retaliation for the employees’ voting for representation. Indeed, his comments were all premised upon the likelihood of a union-called strike.” Julius Getman, Labor Law and Free Speech: The Curious Policy of Limited Expression, 43 Md. L. Rev. 4, 6 (1984). Meanwhile, nothing in the cases held that the government could prohibit litigants involved in electoral speech from seeking to persuade voters to choose one option over another. Thus, if the regulation were to be applied beyond the context of a union election, it would have undoubtedly been deemed unconstitutional.

The Court nevertheless concluded that the regulation did not uncon­stitutionally abridge the employer’s expressive freedom, given the context in which it applied. Chief Justice Warren explained:

Any assessment of the precise scope of employer expres­sion . . . must be made in the context of its labor relations setting. Thus, an employer’s rights cannot outweigh the equal rights of the employees to associate freely, . . . [a]nd any balanc­ing of those rights must take into account the economic dependence of the employees on their employers, and the nec­essary ten­dency of the former, because of that relationship, to pick up in­tended impli­cations of the latter that might be more readily dismissed by a more disinterested ear. 185 Gissel Packing, 395 U.S. at 617.

What is so interesting about this passage is its recognition of the possibility that, given the hierarchical and economically dependent nature of the relationship between the employer and the employee, the same words might mean something different (and be intended to mean some­thing different) than they would mean in another context. This fact led the Court to conclude that rules developed to govern political speech out­side the workplace—that is to say, in a context in which “the inde­pend­ent voter may be freer to listen more objectively and employers as a class freer to talk”—were not appropriate to apply in the workplace. 186 Id. at 617–18.

Gissel Packing represents in this respect something like the anti-Garcetti: Like Garcetti, it acknowledges that the hierarchical nature of the workplace requires the application of different constitutional principles than apply in the public sphere, where individuals confront one another as citizens and equals. But rather than interpreting this to mean that em­ployers should enjoy almost unbounded freedom to dictate the terms and conditions of employment, it reaches the opposite conclusion. It construes the First Amendment to allow a greater imposition on the em­ployer’s freedom than would otherwise be the case. It recognizes, in other words, the serious problem to expressive freedom that economic inequality may represent and, like Marsh and Logan Valley and Red Lion, shapes the First Amendment rules that apply to respond to this problem. The result is, once again, a free speech jurisprudence that does not treat all speakers alike but rather allows legislators and courts to take into account the economic and political attributes that constrain, or empower, their expression.

III. The Antisubordinating First Amendment that Could Be

The discussion in the preceding Part is not meant to be taken as a comprehensive cataloging of the cases in which the Court has inter­preted the First Amendment to promote substantive, as opposed to for­mal, equality. To the contrary: There likely are other areas of free speech doctrine in which the Court has at times departed from a strict rule of formal equality.

What the discussion in the preceding Part is meant to do is to make clear that the anticlassificatory approach that currently dominates has not always been so hegemonic. Even today, the formal-equality principles that the Court claims have always guided the First Amendment—for exam­ple, the principle forbidding the government from “restrict[ing] the speech of some . . . in order to enhance the . . . voice of others” 187 Buckley v. Valeo, 424 U.S. 1, 48–49 (1976) (per curiam). —can­not fully explain the unruly and capacious body of law that gives effect to the constitutional guarantee of expressive freedom. After all, Marsh, Red Lion, Turner, and Gissel Packing remain good law, even if the reach of some of these cases has been blunted, their meaning reshaped by their subsequent interpretation.

This suggests that critics of the Court may have to complicate their analysis of its failings. Professor Owen Fiss argued in his 1986 essay Free Speech and Social Structure that modern free speech doctrine (what he called the “Free Speech Tradition”) is fundamentally flawed due to its tendency to make two largely incorrect assumptions. 188 Owen M. Fiss, Free Speech and Social Structure, 71 Iowa L. Rev. 1405, 1414 (1986). The first is that protecting individual autonomy inevitably ensures the “uninhibited, robust, and wide-open” public debate the First Amendment is supposed to guarantee. 189 Id. at 1410 (internal quotation marks omitted) (quoting N.Y. Times Co. v. Sullivan, 376 U.S. 254, 270 (1964)). Indeed, Professor Fiss argued that “[t]he Tradition assumes that by leaving individuals alone, free from the menacing arm of the policeman, a full and fair consideration of all the issues will emerge. The premise is that autonomy will lead to rich public debate.” Id. When power is distributed unequally, Fiss argued, this is unlikely to be the case. Those with power will use their autonomy to monopo­lize expressive resources just like they monopolize other resources, thus denuding the quality of public debate. 190 Id. (arguing that protecting speaker autonomy may not only “be insufficient to insure a rich public debate” but “might even become destructive of that goal”). The second incorrect assump­tion that underpins the modern Free Speech Tradition, Fiss argued, is that the primary danger to free expression comes from the gov­ernment. 191 Id. at 1414 (“Under the received Tradition, free speech becomes one strand . . . of a more general plea for limited government.”). But in fact, Fiss pointed out, in our highly commodified public sphere, private parties have the ability to shape in very significant ways the overall speech environment. This means that private parties pos­sess similar, in some contexts greater, power to stifle the voices of others than the govern­ment. 192 Id. (arguing that “[t]he state of affairs protected by the first amendment can just as easily be threatened by a private citizen as by an agency of the state” because the state’s “peculiar kind of power is not needed to curb and restrict public debate”). Given its failure to grasp these two essential facts about expres­sive freedom, Fiss argued that a “radical break” with the Free Speech Tradition was necessary. 193 Id. at 1417.

Fiss is certainly correct that both of these assumptions play an important role in the modern tradition. They help justify the different rules that the Court tends to apply to public and private speakers, the strin­gent restrictions placed on content-based regulation of the public forum, and the long-standing discomfort the Court has evinced toward redis­tributive speech policies.

The assumptions are nevertheless far from uncontested, as the media-access cases demonstrate vividly. Indeed, in these and the other cases dis­cussed in Part II, the Court recognized the possibility that the autonomy of private actors could be constrained, consistent with the Free Speech Tradition, when that autonomy poses a real threat to the robust­ness and inclusivity of public debate. It also assumed that courts would take into account the overtly private rules that govern property and con­tractual relationships when interpreting freedom of speech. In other words, it assumed that the First Amendment protected speakers against more than intentional discrimination at the hands of the state.

That the Free Speech Tradition is somewhat more complex than Fiss and other progressive critics have made it out to be should not be terribly surprising. After all, the primary justification that the New Deal Court provided for applying heightened judicial scrutiny in First Amendment cases was that doing so was necessary to protect the democratic system of government from capture by those with economic, political, or social power. 194 This idea was articulated most famously by Justice Brandeis’s assertion in his famous concurrence in Whitney v. California that freedom of speech was intended to guard against “the occasional tyrannies of governing majorities.” 274 U.S. 357, 375–76 (1927) (Brandeis, J., concurring). But it infused the New Deal cases as well. It helps explain, for example, why the Court primarily relied upon First Amendment cases as support for its assertion in the second paragraph of the famous Footnote Four of United States v. Carolene Products Co. that heightened scrutiny is appropriate when laws “restrict[ ] those political processes which can ordinarily be expected to bring about repeal of undesirable legisla­tion.” 304 U.S. 144, 152 n.4 (1938) (citing, among other cases, De Jonge v. Oregon, 299 U.S. 353 (1937), Near v. Minnesota ex rel. Olson, 283 U.S. 697 (1931), Whitney v. California, 274 U.S. 357 (1927), and Gitlow v. New York, 268 U.S. 652 (1925)). The idea that judicial scrutiny is appropriate when necessary to reinforce and defend the practices of democratic representation against capture by dominant groups is most often invoked to justify heightened scrutiny in equal protection cases; however, as Professor G. Edward White has shown, the Court initially made this argument in its First Amendment cases, particularly those involving freedom of speech. See G. Edward White, The First Amendment Comes of Age: The Emergence of Free Speech in Twentieth-Century America, 95 Mich. L. Rev. 299, 327–28 & n.83 (1996). Consequently, a central principle of the New Deal Court’s First Amendment cases was the idea that “[f]reedom of speech . . . [must be] available to all, not merely to those who can pay their own way.” 195 Murdock v. Pennsylvania, 319 U.S. 105, 111 (1943). It would be strange, given all this, had the Court not paid attention to social context or material inequality when interpreting the First Amendment’s command. In this respect, it is the current, much more formalist approach taken by the Court to free speech questions that is much more difficult to recon­cile with the New Deal Court’s insistence that the First Amendment was meant not only to guarantee an individual right to auto­nomy in thought and expression but also to facilitate and safeguard a par­ticular kind of social institution: namely, the democratic public sphere.

This is not to say that the earlier age was a golden era of progressive judges, vindicating the public good in the name of the First Amendment. But it does mean that there are resources within the modern tradition that those dissatisfied with the current doctrinal arrangement can use to make free speech jurisprudence more protective of the expressive free­dom of the “little people” and less protective of corporate power.

Think, for example, of what it could mean for the regulation of pro­test if courts were to treat the “ample alternative channels of commun­ication” requirement that is recited in every time, place, or man­ner case as a stringent constraint on the government’s powers rather than a parch­ment barrier. Surely one consequence would be to make it considerably harder to justify the “free speech cages” and other content-neutral devices that modern municipalities commonly employ to corral dissent at major political events. 196 See, e.g., ACLU of Colo. v. City & Cty. of Denver, 569 F. Supp. 2d 1142, 1175 (D. Colo. 2008) (upholding a content-neutral law that prevented protestors from being within sight or earshot of the Democratic National Convention and discussing other cases in which similarly restrictive content-neutral protest laws were upheld). See generally Timothy Zick, Speech and Spatial Tactics, 84 Tex. L. Rev. 581, 581–82 (2006) (criticizing the use of facially neutral time, place, and manner laws as “powerful weapon[s] of social and political control”).

Or consider what it might mean to the regulation of the digital eco­nomy if the First Amendment were recognized to grant a right of access to privately owned but socially important sites of public expression, such as Facebook, similar to that which Logan Valley extended to those who wished to speak in privately owned shopping malls. Scholars routinely dismiss the possibility that any First Amendment constraints apply, or should apply, to internet companies such as Google and Facebook. They argue that it would be inappropriate, and a very bad idea, to impose on these companies the same duties of nondiscrimination that apply to the government when it regulates speech. 197 That there is discrimination in how private media companies like Facebook mod­erate internet speech is evident. See Kate Klonick, The New Governors: The People, Rules, and Processes Governing Online Speech, 131 Harv. L. Rev. 1598, 1653–55 (2018). For a recent argument against extending First Amendment scrutiny to the actions of Facebook and Google that assumes that doing so would require designating these companies as state actors, see Tim Wu, Knight First Amendment Inst., Is the First Amendment Obsolete? 22–23 (2017),
%20Wu%20Is%20the%20First%20Amendment%20Obsolete.pdf []. Professor Tim Wu’s answer to the question he poses is yes—but it need not be.
This is probably true—although one should not overstate how broadly that prohibition applies. 198 As Part I makes clear, the government enjoys considerable power to discriminate, both in institutions that are considered nonpublic forums and in public forums—when it can do so via content-neutral rules. But Logan Valley suggests that the First Amendment limits the actions of pri­vate actors only when they stand in the government’s shoes. It, and Marsh before it, suggest the possibility of a more nuanced constitutional rule.

Of course, one might worry that interpreting the First Amendment in this more genuinely imperial manner would unduly limit the freedom of powerful parties—be they the government or private companies—and thereby undermine their ability to maintain the “good order upon which [civil liberties] ultimately depend.” 199 Cox v. New Hampshire, 312 U.S. 569, 574 (1941). This concern is understandable. Freedom of expression is not, obviously, the only interest at stake in First Amendment cases. Speech can threaten, defame, clutter, and terrorize. But to say that courts could, and should, interpret the First Amendment’s command in a manner that is less constrained by the requirement of for­mal equality than contemporary free speech law is not to say that the right to speak should always trump all other rights or regulatory interests. Precisely the reverse: It is to say that courts could, and should, engage in a far more realistic analysis than they currently do of the political, eco­nomic, and social realities that impede, or enable, the “uninhibited, robust, and wide-open” public debate that the First Amendment is sup­posed to make possible 200 N.Y. Times Co. v. Sullivan, 376 U.S. 254, 270 (1964). —and develop rules in response.

To this, one might object that requiring judges to take serious account of economic, political, and social context would give them too much discretionary power or require too much of them. One might query whether judges have the capacity to engage in the contextual, ulti­mately sociological analysis required to effectuate an antisubordinating vision of the First Amendment. This concern also has merit. Certainly, cases such as Red Lion should give one pause. After all, even those who generally sup­ported the approach that the Court took in that case acknowledge that the set of policies it upheld may have been normatively undesirable; by imposing costs on broadcasters who criticized political candidates, they may have denuded the quality of public debate. 201 See, e.g., Fiss, supra note 188, at 1419–20 (noting that policies like the Fairness Doctrine “seek[ ] to enhance public debate by forcing . . . broadcasters . . . to present opposing sides of an issue . . . , but the fear is that [they] might work in the opposite direction, . . . by discouraging [broadcasters] from taking chances, and by undermining norms of professional independence”); Karst, supra note 10, at 49 (cautioning that a “right of reply” requirement for broadcasters “will give added encouragement to an editorial blandness already promoted by the broadcasters’ commercial advertisers”). This suggests that a doctrine that allows more active government intervention in the media environment might produce, in some cases, normatively undesirable results.

The problem is that there likely is no better alternative—at least if one takes off the table dramatic changes in economic policy that would reduce economic inequality and ensure that underlying differences in the distribution of property rights do not seriously undermine the robust­ness and inclusiveness of public debate. 202 Even in a context of significantly reduced economic inequality, one might worry about social and political cleavages that make it difficult for some speakers to participate in public debate. See Fiss, supra note 188, at 1412. I leave that to one side, however, given that the kinds of redistributive economic policies that would make it necessary to think seriously about noneconomic sources of inequality are unlikely to be enacted any time soon. Simply relying on the market­place of ideas to solve the problem is no solution because, as Fiss noted in his critique of the Free Speech Tradition, in many cases, the market­place is the problem. 203 See id. at 1413. And although some participants in this Symposium have suggested self-regulation as an alternative solution to the obvious failure of contemporary free speech law, it is hard to see why the corporate interests of powerful private actors would sufficiently align with the public’s expressive interests to make self-regulation an adequate sub­stitute for the carrot and stick of judicial intervention. 204 See Jack M. Balkin, Free Speech Is a Triangle, 118 Colum. L. Rev. 2011, 2025–28 (2018). In arguing for self-governance, Balkin assumes that the rules that govern classic public forums would necessarily govern internet domains. However, this need not be the case, as this Essay makes clear.

In the end, vindicating the values that the Court has repeatedly said it believes the First Amendment is intended to safeguard—chief among these, the cultivation of a robust and inclusive public sphere that is neces­sary for the maintenance of a healthy democratic system of govern­ment—may just require courts to engage in the difficult task of balancing the often-competing constitutional interests at stake when the govern­ment regulates speech either directly or indirectly.

The anticlassificatory approach that the Court employs today in both its free speech and equal protection cases is very appealing because it produces what appear to be crystal-clear rules. Absolutes are frequently very comforting. In fact, one might question how clear the anticlassifica­tory approach is—particularly when it requires courts to ignore features of the economic and political landscape that they believe (correctly) should matter to the analysis. Turner provides a good example of how, even within a nominally anticlassificatory framework, courts can reach the results they want.

Even if we assume, however, that the anticlassificatory framework provides rules that do cabin judicial discretion more than the alternative approach, it does so only by outsourcing all the difficult questions—who shall speak? how shall they speak? what speech shall be permitted?—to property and contract law, or (in the case of government institutions) to internal regulations. Yet there is no reason to believe that the rules that gov­ern those areas of the law were designed with any concern for free speech values.

It is certainly true, as the Court reminded us in Citizens United, that the First Amendment is predicated on a mistrust of government power. 205 Citizens United v. FEC, 558 U.S. 310, 340 (2010). But it is far from obvious that a formal rule that cabins judicial discretion provides any greater protection from the misuse of government power than the alternative. In fact, the history recounted in this Essay suggests that the opposite is true.

This is not to say that it will be easy to persuade the Court that the First Amendment should be interpreted in light of a substantive concep­tion of equality. The transformation of the First Amendment that this Essay calls for may simply not be politically feasible right now. And yet there is value in remembering both what the First Amendment has been and what it may be again. Doing so reminds us that the free speech guarantee is susceptible to multiple interpretations and that the disequal­izing tendencies of contemporary free speech law are neither necessary nor inevitable.


Supporters of the Roberts Court are fond of saying that it is the most speech-protective Court in history. 206 See, e.g., Joel M. Gora, Free Speech Matters: The Roberts Court and the First Amendment, 25 J.L. & Pol’y 63, 75 (2016). As this Essay has suggested, this is quite simply not so. In many respects, the Warren Court was much more solicitous of the rights of speakers than the Court is today. And we have come far from the days when the Court routinely claimed that, when compared to property rights, First Amendment rights occupied a “pre­ferred position” in our constitutional scheme due to their importance to the democratic system of government. 207 See, e.g., Saia v. New York, 334 U.S. 558, 562 (1948) (“Courts must balance the various community interests in passing on the constitutionality of local regulations . . . . But in that process they should be mindful to keep the freedoms of the First Amendment in a preferred position.”); Marsh v. Alabama, 326 U.S. 501, 509 (1946) (“When we balance the Constitutional rights of owners of property against those of the people to enjoy freedom of press and religion, as we must here, we remain mindful of the fact that the latter occupy a preferred position.”); Murdock v. Pennsylvania, 319 U.S. 105, 115 (1943) (holding that the fact that a license tax was “nondiscriminatory is imma­terial” and that “[s]uch equality in treatment does not save the ordinance” because “[f]reedom of press, freedom of speech, freedom of religion are in a preferred position”). It is true that today the First Amendment provides a powerful protection to those who wish to use their property for expressive purposes—and to guard against govern­ment efforts to regu­late or restrict those uses—but it provides shoddy protection to everybody else. This is a problem, not only because it means that constitutional pro­tection is afforded to those who least need it. It is a problem because it also means that the First Amendment is unable to effectively achieve its core purposes.

This Essay argues that scholars need not reinvent the wheel to con­struct a First Amendment doctrine that does a better job of ensuring that free speech rights are—in practice and not just in theory—“available to all, not merely to those who can pay their own way.” 208 Murdock, 319 U.S. at 111. Instead, they can—and perhaps should—look to the First Amendment’s past as a guidepost for its future. Even if the Roberts Court is unlikely to counte­nance the reinvigoration of half-buried or firmly cabined precedents, the complex history of the Court’s engagement with substantive equality pro­vides scholars and advocates with a vision of a First Amendment that is more sensitive to private power and more attentive to the effects—and not just the purpose—of government action. The antisubordinating vision of the First Amendment sketched out in this Essay may be in exile now, but it need not be forever.