Jump to ContentJump to Main Navigation
Extending Rights' ReachConstitutions, Private Law, and Judicial Power$

Jud Mathews

Print publication date: 2018

Print ISBN-13: 9780190682910

Published to Oxford Scholarship Online: April 2018

DOI: 10.1093/oso/9780190682910.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 18 August 2019

State Action and Constitutional Containment

State Action and Constitutional Containment

(p.117) 5 State Action and Constitutional Containment
Extending Rights' Reach

Jud Mathews

Oxford University Press

Abstract and Keywords

This chapter explores the U.S. Supreme Court’s use of, and departures from, the state action rule. It begins by reconstructing the state action rule’s origins in the Civil Rights Cases. From the late nineteenth century onward, the state action rule served as a constitutional containment device, bolstering the Court’s monopoly over constitutional interpretation and eliminating uncomfortable questions about what rights meant for the ordering of American society. A changing political context and the emergence of new normative demands in the twentieth century put this regime under pressure, which the Court managed through a series of strategic evasions of the state action rule, even while pledging fealty to it.

Keywords:   United States, U.S. Supreme Court, state action, twentieth century, constitutional rights, constitutional containment

Birth of the State Action Rule

Coming not long after a stepped-up enforcement bill failed in Congress and President Ulysses S. Grant declined to send federal troops to Mississippi to stop violence against freedmen, the Cruikshank decision completed a troika of events that dramatically illustrated the abandonment of Reconstruction by all three branches of the national government (Foner 2002).1 But Cruikshank is important also because it contained the seeds of the state action doctrine that the Court more fully developed in the Civil Rights Cases.2 The Civil Rights Cases mark the most important moment in the Court’s postwar constitutional settlement, and they are worth examining in some detail.

In March 1883, the Court heard five consolidated cases that involved the exclusion of African Americans from inns, theaters, and the railroad, in violation of the Civil Rights Act of 1875. Only one of the cases came from the Deep South, where the Act was going effectively unenforced (Goldstone 2012, 104, 118).

The central question for the Court, wrote Justice Bradley, was whether Congress had the power to pass the law. The statute’s defenders relied principally on the Fourteenth Amendment. While acknowledging that Section 1’s litany of things that “no State shall” do is “prohibitory in its character, and prohibitory upon the States,” Justice Bradley noted that “[i]t is State action of a particular character that is prohibited. Individual invasion of individual rights (p.118) is not the subject-matter of the amendment.”3 The contours of the right guaranteed by Section 1 of the Fourteenth Amendment also circumscribe the scope of Congress’s power to enforce the amendment under Section 5:

[T]he last section of the amendment invests Congress with power to enforce it by appropriate legislation. To enforce what? To enforce the prohibition. To adopt appropriate legislation for correcting the effects of such prohibited State law and State acts, and thus to render them effectually null, void, and innocuous. This is the legislative power conferred upon Congress, and this is the whole of it.4

Congress cannot act here, precisely because the requisite state action is lacking:

[U]ntil some State law has been passed, or some State action through its officers or agents has been taken, adverse to the rights of citizens sought to be protected by the Fourteenth Amendment, no legislation of the United States under said amendment, nor any proceeding under such legislation, can be called into activity.5

Congress may not use the Fourteenth Amendment to pass legislation that “cover[s] the whole domain of rights appertaining to life, liberty, and property, defining them and providing for their vindication. That would be to establish a code of municipal law regulative of all private rights between man and man in society.”6 The Civil Rights Act “does not profess to be corrective of any constitutional wrong committed by the States,” but “proceeds ex directo to declare that certain acts committed by individuals shall be deemed offenses, and shall be prosecuted and punished by proceedings in the courts of the United States.”7 Hence, it exceeds the scope of Congress’s power under the amendment.8

The Civil Rights Cases were recognized at the time, and have been recognized since, as a watershed in American rights jurisprudence.9 As a matter of (p.119) black letter constitutional law, the Civil Rights Cases are recognized as the origin of the state action rule, even if the decision did not explicitly reject the state neglect concept. A standard contemporary hornbook for law students offers the following lapidary formulation of this rule: “Nearly all of the rights guaranteed by the Constitution to individuals are protected only against interference by government” (Emanuel 2012, 435). But the Civil Rights Cases decision is not as straightforward as it might appear at first glance.

Start with the term “state action” itself. The natural question to ask is, “state action” as opposed to what: what is the alternative? Logically, there are two: there is nonstate action, and there is state inaction (West 2003). Statements in the Civil Rights Cases such as “[i]ndividual invasion of individual rights is not the subject-matter of the amendment” draw the reader’s focus to the first of these oppositions, contrasting action of the state to action of nonstate actors. In other words, they suggest that the Court must choose between holding that the U.S. Constitution may be violated by actions of the state only and holding that the Constitution may be violated by both state action and the acts of nonstate actors. Framed in this way, the U.S. Supreme Court’s judgment amounts only to a rejection of direct horizontal effect and seems uncontroversial.

But this framing obscures the second opposition, between state action and state inaction. In this framing, the question becomes: may the Constitution be violated by state action only, or also by state inaction? Or, to particularize it to the issue in the Civil Rights Cases, can the right to equal protection of the laws impose on the state affirmative obligations: for instance, to proscribe or punish certain forms of discriminatory treatment? This question is trickier. As described above, the contemporaneous conventional wisdom seemed to be that the Fourteenth Amendment encompassed state neglect: that state inaction could be, in a word, actionable under the Fourteenth Amendment.

The question of just how far the state’s obligations ran, however, was more controverted, as discussed above. But Justice Bradley framed matters in the Civil Rights Cases in a way that permitted the Court to sidestep these difficult issues. In his opinion, Justice Bradley nowhere explicitly repudiated the statement he had made to Judge Woods a dozen years earlier: that “denying the equal protection of the laws includes the omission to protect as well as the omission to pass laws for protection.” Rather, he changed the conversation, shifting the focus from the role of the state to the proximate source of the injuries for which redress was sought (here, the owners of private establishments who denied entry to African Americans). In this way, the Civil Rights Cases prefigure the peculiarities of Flagg Brothers a century later. The Court in Flagg Brothers focused on whether the actions of the warehousemen could be attributed to the state, rather than asking whether the challenged statute, which was undeniably an action of the state, was consistent with the Constitution.

(p.120) The state action rule, as introduced in the Civil Rights Cases, reduced the potency of rights as a tool for challenging unequal institutions and practices. It marks an important moment in more than one narrative of institutional and moral failure in American history. It is the culmination of the judicial abandonment of the Reconstruction project, and the recently freed slaves it had sought to help, which was part of the abandonment by each of the branches of the national government.10 It also anticipated the Supreme Court’s wholesale capitulation to the rise of Jim Crow in the South more than a decade later, in Plessy v. Ferguson.11

But even if the Court had wanted to pursue an equality agenda, it is by no means clear how successful it would have been in the face of widespread hostility by the governments of the South and without support from the political branches of the federal government.12 Political support for enforcing Reconstruction rights was spent well before the Civil Rights Cases were argued to the Court. As Pamela Brandwein has argued,

It seems clear that the Court, in this highly unstable political environment, lacked the clout to lead with an assertive racial egalitarianism. The Court faced institutional concerns regarding the preservation of its legitimacy. Recounting Colfax or repeating Justice Woods would have highlighted Southern maladministration of law as a rights denial. And as events made plain, the Court could not count on the executive branch to undertake the broad prosecutorial efforts necessary to remedy this rights denial, i.e., to enforce the Court’s rules. By recounting Colfax or repeating Justice Woods, the Court thus risked appearing ineffectual. The Court’s modulated expression of state neglect, then, can be seen as an “available device for not doing”; that is, for not putting itself out in front of the political branches during a period of instability when four Republican presidents (Grant, Hayes, Arthur, Harrison) tacked away and then back toward rights enforcement. This secured key legal principles without risking the influence of the Court (Brandwein 2007, 372) (internal citation omitted).

(p.121) The Court “risked appearing ineffectual” if it declared that citizens were entitled to rights that the state governments violated at every turn with the tacit acquiescence of federal law enforcement. Instead, the Court declared “mission accomplished” and left the field, asserting that the abuses in the South did not implicate constitutional rights.13

Taking a more expansive approach to constitutional rights would also have reduced the Supreme Court’s control over the interpretation and application of those rights. Suppose the Supreme Court had signaled in the Civil Rights Cases that the Fourteenth Amendment imposed some affirmative duties on the states to protect citizens against discrimination. Questions about what the Constitution concretely required of the states would be litigated not only in federal but also state courts, with limited opportunities for federal review.14 The Supreme Court could have substantial difficulty corralling disparate views on the reach of rights into coherent doctrine, as it might also had it afforded Congress broad latitude to use its Fourteenth Amendment enforcement power to define the meaning of the equality right in its own terms. Competing constructions of rights could both challenge the Court’s cassational capacity and drag it into contentious conflicts it would prefer to avoid in an era of highly charged politics.

The state action doctrine helped the Supreme Court to consolidate its interpretative authority over the Constitution. Other moves the Court made during this same period also strengthened its position. In particular, the Supreme Court had normalized judicial review of federal statutes in the years immediately after the Civil War. Even as the Supreme Court bobbed and weaved on the most high-profile issues concerning the war and Reconstruction, it struck down ten federal statutes in whole or part between 1865 and 1873, in contrast to only two previously (in Marbury and Dred Scott)15 (Kutler 1968, 114; Horwitz (p.122) 1992, 19). The cases included United States v. Dewitt,16 the first case in which the Court struck down a federal law on Commerce Clause grounds, and Hepburn v. Griswold,17 which invalidated the issuance of paper money as inconsistent, not with any specific provision, but with the “spirit” of the Constitution.

In any other period of U.S. history, these cases probably would have received more attention because they raised important issues of federal-state relations and the separation of powers. But with war and Reconstruction cases drawing disproportionate scrutiny, the Court quietly laid the groundwork for judicial supremacy in these and other rulings.18 The Court was laying to rest the departmentalist views of constitutional interpretation that had had such currency in the antebellum years.

The Court also marked out the borders of its bailiwick by sharpening the hitherto hazy distinction between the law of the U.S. Constitution and what it called “general constitutional law.” In Loan Association v. Topeka, from 1874, the Supreme Court was still willing to invalidate a state tax imposed for the benefit of a private business, not because it violated any textual provision of the Constitution but because “there are such rights in every free government beyond the control of the State. . . . The theory of our governments, State and National, is opposed to the deposit of unlimited powers anywhere.”19 But just four years later, the Court was far more punctilious about the nature of federal constitutional law. In Davidson v. Louisiana, the Court declined to strike down a state tax as a violation of due process.20 The Court distinguished its resort to the general common law in Loan Association on the ground that the Court had been reviewing a lower federal court, not a state court, in the earlier case. The (p.123) Court’s ostensible restraint in Davidson at the same time underlines the idea, commonplace now but not at the time, that the law of the U.S. Constitution is a body of positive law, which is distinct and separate from general principles and inherited legal traditions and over which the Supreme Court has final interpretive authority. When first confronted with the opportunity to give horizontal effect to constitutional rights, the U.S. Supreme Court demurred. The Court framed the issue in terms of a dichotomy between state action and nonstate action and held categorically that constitutional rights did not apply to the latter. The discriminatory treatment African Americans endured, especially in the South, fell far short of the constitutional values embodied in the Reconstruction Amendments, and using rights to bridge the gap would have been consistent with the widespread contemporary understanding that rights could require positive action by the state. But in the context of post-Reconstruction politics, doing so was not in the Court’s interest. While federal courts had new formal powers to control how federal law was applied in the states, owing to the postwar expansions of federal jurisdiction, there was no political support for advancing the protection of civil rights. A broad declaration of the scope of rights would be undercut by the intransigence of Southern states and the indifference of the Congress and the President.

By contrast, the state action rule yielded functional advantages for the Court. By framing the question of rights’ horizontal effect in terms of an opposition between state action and nonstate action, the Court altogether avoided entanglement in difficult, contentious questions about when state inaction might be constitutionally actionable. And confining the effect of rights to the vertical citizen-state relationship aided the Court in its efforts to differentiate constitutional law from common law and establish itself as the ultimate authority on the former.

State Action Under Pressure

In the decades after the Civil Rights Cases, the state action rule effectively shut down any and all claims that the states bore obligations, derived from constitutional rights, to protect the interests of individuals against the acts of other nonstate actors. Under the banner of the state action rule, the Supreme Court held not only that states were not required to take action to protect individuals against unequal treatment but also that the federal government was not permitted to protect individuals against such treatment.21 The doctrine, only (p.124) rarely subjected to critical examination, served for many years as a powerful constitutional containment device.22 The point is not that the Supreme Court was categorically hostile to new rights claims; after all, the “Lochnerism” of the early twentieth century depended on finding new rights to economic liberty in the Fourteenth Amendment. Rather, the state action requirement blocked rights claims to—as opposed to rights claims against—state interventions in private orderings. And by preemptively ruling out of bounds many potential rights claims as well as many potential uses of the congressional enforcement powers, it helped ensure that the Supreme Court would dictate the course of constitutional change.

In the years following World War II, however, the Court’s state action framework came under increasing pressure. In the wake of a war against Nazi and fascist enemies, and in the context of a Cold War against a repressive Soviet state, a civil liberties and equal rights agenda gained political ground in the United States. Popular opposition mounted to racial discrimination in particular, championed by an active civil rights movement. The Supreme Court played an important role in pushing the national debate on civil rights forward, most explosively with its 1954 decision in Brown v. Board of Education. Since the Supreme Court gave up its role as legislative censor in the social and economic field in the late 1930s, it had begun defining its identity, with increasing clarity, as a rights-protecting court.

The Court was in the vanguard of the broader attitudinal and political changes, in particular with respect to racial equality. Mid-century opinion polls show a steady increase among Americans sympathetic to civil rights from the 1940s until the 1960s (Best and Radcliff 2005, 84–85). The political branches also took action. Soon after the war’s end, President Harry S. Truman advanced the cause of civil rights with an executive order ending racial discrimination in the military, which was largely executed by President Dwight D. Eisenhower (Nichols 2007).23 Congress passed its first civil rights legislation in over eighty years in 1957, after years of hard-fought effort by liberal Democrats that finally overcame the opposition of Southern Democrats.

Indeed, political divides over civil rights and civil liberties increasingly (though not exclusively) broke along sectional lines. More and more, the South played the role of a regional holdout from an emerging national consensus in favor of greater rights protections. And the structure of American federalism (p.125) constrained national institutions in their ability to bring recalcitrant states and localities in line with national norms. Congress’s reach was restricted by the limitation of its legislative powers to those subjects enumerated in Article I of the Constitution plus the Reconstruction Amendments. Congress would famously overcome these limitations to some extent through an expansive construction of its Commerce Clause power, aided and abetted by the Supreme Court. But even the power to regulate interstate commerce would not permit Congress to penetrate the innermost citadels of the private law with legislation that gave concrete expression to the values of freedom and equality.

Federal courts likewise had a limited capacity to countermand state practices. State courts were the final authorities of matters on state law, and in the absence of federal legislation, only the Constitution trumped state law and the decisions of state courts.24 And so, the Supreme Court increasingly turned to the Constitution. Starting after the end of the Second World War, the Supreme Court began selectively deploying constitutional rights horizontally to fill the normative gap between national standards and local (principally Southern) outliers.

The Supreme Court’s softening of the strong state action norm is an important part of what I call the second postwar constitutional settlement.25 In judicial decisions and political discourse alike, American constitutionalism came to be increasingly identified with the protection of individual rights. Richard Primus emphasizes how the postwar valorization of human rights, and their identification with constitutional rights, served to distinguish the United States from its international foes, the Nazis and Soviets (Primus 1999, 177–203). Interestingly, in the United States, the threats to rights came primarily from subnational political units: the states and, in particular, Southern states. Accordingly, promoting rights in the United States meant nationalizing the protection of rights. The Supreme Court did this, not least, by formally extending the reach of rights to the states. Between 1948 and 1969, the Supreme Court handed down eleven decisions “incorporating” rights in the first eight amendments to the Constitution (p.126) against the states through the operation of the Fourteenth Amendment Due Process Clause (Primus 1999, 69). Granting Congress more leeway to legislate under the banner of rights and relaxing the state action requirement reinforced the conception of the Constitution as a rights-protecting pact and the conception of the Supreme Court as a rights-protecting court.

The postwar emphasis on rights, and the Supreme Court’s close identification with rights protection, also provided a new rationale for judicial supremacy. The Supreme Court must have the last word on the meaning of rights to prohibit other actors, particularly at the state level, from defining rights down. The twinning of judicial supremacy and rights protection in the Court’s normative narrative came through vividly in the iconic 1958 decision Cooper v. Aaron.26 When the governor and legislature of Arkansas denied that they were bound by the Supreme Court’s decision in Brown v. Board of Education that school segregation was unconstitutional, the Court emphatically rejected the claim. In a decision signed by each Justice individually, the Court declared not only that the Constitution was the supreme law by virtue of Article VI, but that “the federal judiciary is supreme in the exposition of the law of the Constitution.”27 And in the conclusion of the opinion, the Court asserted that extending the mantle of supremacy from the Constitution to the Court was essential to the realization of rights:

The principles announced in that decision and the obedience of the States to them, according to the command of the Constitution, are indispensable for the protection of the freedoms guaranteed by our fundamental charter for all of us. Our constitutional ideal of equal justice under law is thus made a living truth.28

A tide of judicial decisions and legislative actions advanced antidiscrimination and civil liberties norms for a generation after the end of World War II. And—most relevant for present purposes—the Supreme Court made a series of ad hoc alterations to the state action rule during that same period, to reach and remedy the increasingly anomalous sectional divergences that otherwise lay beyond the reach of federal powers. Because these alterations were ad hoc, the Supreme Court never articulated a theory of what, exactly, rights entitled citizens to under the law. It is the ad hoc quality of the Court’s doctrinal excrescences that earned state action doctrine the sobriquet of a “conceptual disaster area.”

(p.127) But the lack of a general theory that would explain when the effects of rights would reach into the private law also permitted the state action norm to continue to function as a constitutional containment device, which reinforced the Court’s supremacy. The Court did not hem itself in with a theory: instead, rights would reach into the private law when the Court said they would.

The variation in how the Supreme Court approached state action over time furnishes material for a comparison. Why was it that the Supreme Court used constitutional rights to intercede in matters governed by private law following the Second World War but not the Civil War? Of course, much had changed between the early 1880s (when the Supreme Court first rejected the horizontal effect of rights in the Civil Rights Cases) and the late 1940s (when the Court first blocked judicial enforcement of private discrimination). But the key, I argue, is that the strategic logic to the Court’s use of constitutional rights changed dramatically in that period.

After Reconstruction, granting horizontal effect to rights meant taking on a mission the weakened Court could not accomplish: to root out discrimination in the South, without the backing of the other branches of government or a powerful public constituency. After World War II, and especially as the Civil Rights movement gained traction, the Supreme Court was in the vanguard of a broader political and social change in the politics of civil rights and civil liberties. The cause of robust rights protection attracted significant constituencies, inside and out of government. As the Supreme Court got out of the business of policing economic regulation, it made rights central to its agenda and identity. Among national institutions, it was only the Supreme Court that could reach and rectify local discrimination under the auspices of the private law, and only through resort to constitutional rights. When the Court did so, it demonstrated its unique value-added to American government while advancing its agenda.

The next sections illustrate developments in state action doctrine and in processes of lawmaking, in three areas. The first section, devoted to the horizontal application of equality rights, is the longest, because this is the area in which the Supreme Court has deviated most from the state action requirement. I argue that this is where we should expect to see rights given horizontal effect, since states were blocking the advance of a civil rights agenda that had the Court’s support, and only resort to constitutional rights could breach the fortress of federalism. Still, the Court did not formally jettison the state action rule, but worked around it, avoiding a still broader transformation in how constitutional law is made. I discuss how the Court did this with a focus on voting and housing discrimination. The Court also licensed a degree of legislative constitutionalism in this area, allowing Congress some latitude to legislate in the service of equality values, and I discuss the key decisions on Congress’s enforcement powers, which transformed the restrictive approach the Court had taken in (p.128) the Civil Rights Cases. But here as well, the Court’s tolerance for constitutional pluralism was limited, and I argue that the Court justified antidiscrimination legislation with reference to the rights enforcement powers of Congress only when no other avenue was open.

There are a number of areas where private legal arrangements can implicate the freedom of expression, but the Supreme Court has most decisively brought constitutional rights to play in the context of libel law. This also makes sense, I argue, since libel law more directly implicates the Court’s civil libertarian concerns than the others. The discussion here focuses on the Supreme Court’s landmark New York Times v. Sullivan case, and its transformation of libel law. By contrast, I argue, the Court has no incentive to constitutionalize internal family legal relations, and the Court has not done so. The DeShaney case illustrates how the Court has left the regulation of this area to state lawmakers. The Court will bring the Constitution to bear on family law matters only when an alleged rights violation can be traced directly back to the act of a state—for instance, state law’s exclusion of same-sex couples from marriage rights.


It is not surprising that cases involving racial discrimination play an outsized role in American state action doctrine. Here is an area where the normative gap between localized practices (especially in the Jim Crow South) and emerging national equality norms yawned especially wide. Moreover, the constitutional guarantee of equal protection could be plausibly interposed to fill this gap. And crucially, the Supreme Court could draw on institutional and popular support for expanding the rights agenda to combat discrimination.

The Supreme Court began moving into the civil rights space, slowly at first, in the 1940s. In the 1960s the Court gave its blessing to a far-reaching legislative antidiscrimination agenda that transformed state and local law. But the legislation left some private-party relationships under the law untouched. And in these areas, the Supreme Court repeatedly used constitutional rights to remedy these wrongs, adding twists and turns to state action doctrine in the process.

The discussion here of the Court’s state action jurisprudence in the antidiscrimination context is necessarily selective, given how much of the development of state action doctrine was driven by racial discrimination cases. I focus here on landmark cases and the relationship between the Court’s state action doctrine and the legislative antidiscrimination agenda. I argue that the Court promoted substantive equality values, and at the same time maintained its own privileged position with respect to rights, by using rights selectively to ground its decisions when no other source of law would, and seeking nonrights sources for congressional power to promote equality when possible.

(p.129) White Primaries

The first major crack in the state action rule appeared in the context of a long-running series of conflicts over the “white primaries” in Texas. Because the state Republican parties in the South had collapsed in the wake of Reconstruction, the Democratic primaries were the only real competitive elections in many Southern states. A 1923 statute in Texas expressly excluded blacks from participating in the primaries. In the unanimous 1927 decision Nixon v. Herndon, the Supreme Court struck down the statute, ruling that it violated the Fourteenth Amendment.29

Nixon v. Herndon did not pose a difficult state action problem: the target of the challenge was a statute, and it was unambiguously discriminatory. But the state’s efforts to circumvent Herndon required the Court to confront successively more difficult state action issues. In the wake of Herndon, the Texas legislature rewrote the statute to provide that the executive committee of each party could set eligibility requirements for primary voting. In 1932, the Supreme Court struck down this arrangement as well. Justice Cardozo wrote for the unanimous Court that state parties, in prescribing eligibility requirements for voting, were exercising power delegated by the state, and therefore subject to the Fourteenth Amendment.30 Within a month of the Supreme Court’s decision, the state Democratic convention of Texas had adopted on its own volition a resolution limiting membership in the party, and participation in the primary, to whites.

The state Democratic Party had taken up where the legislature had left off. The effect was exactly the same—only whites could vote in the primary—but since the discrimination was now effected by a voluntary political association, the Court could reach it only by rejecting, ignoring, or finessing the state action doctrine. At first, the Court blinked. In the 1935 case Grovey v. Townsend, a unanimous Court held that the Fourteenth Amendment did not reach the party’s regulation of its own membership.31

But significantly, nine years later, the Court did an about-face. In Smith v. Allwright, the Court overruled Grovey and invalidated the Texas Democratic Party’s exclusion of nonwhites from primary participation.32 The Court now found that rules restricting participation in primaries fell within the ambit (p.130) of state action, even if imposed by parties, owing to the primaries’ tight integration into the machinery of public elections. In Justice Reed’s words for the Court, “this statutory system for the selection of party nominees for inclusion on the general election ballot makes the party which is required to follow these legislative directions an agency of the state in so far as it determines the participants in a primary election.”33

Why the switch between Grovey and Allwright? The nine years separating the cases nicely illustrate some of the political and institutional currents that were beginning to change the logic of requiring state action in the context of racial discrimination. For starters, institutional support for dismantling race restrictions on voting was growing within the executive branch. Frank Murphy, who became Attorney General in 1939, established a Civil Liberties Unit within the Department of Justice (DOJ). The Civil Liberties Unit undertook a long-term, incremental litigation strategy to rooting out voting discrimination (Valelly 2004, 160). Among other things, the DOJ persuaded the Supreme Court that the Fifteenth Amendment did apply to primaries in the 1941 case United States v. Classic,34 a decision that provided important support for the Court’s holding in Allwright. At the same time, interest group and social movement pressure was beginning to grow, resulting in more litigation and refinements to the legal arguments. Thurgood Marshall, who led the NAACP’s litigation strategy against white primaries and argued Smith v. Allwright, hammered home the argument that the white primary was an “integral part of the election machinery of Texas” (Valelly 2004, 161). And the Court itself was changing. Remarkably, between 1935 and 1944, Franklin Roosevelt made eight appointments to the Supreme Court. Roosevelt’s appointments included Justices “determined to deepen the Court’s civil liberties and civil rights orientation” (Valelly 2004, 158).

The nascent civil rights agenda became even more mainstream, and more politically institutionalized, in the second half of the 1940s. Encouraged by his advisers to reach out to black voters or else lose the 1948 election, President Truman began pushing a civil rights agenda near the midpoint of his first term (Valelly 2004, 165). The landmark report of Truman’s Committee on Civil Rights, released in 1947, identified elements of racial hierarchy throughout American society and the economy, and urged their abolition. Truman announced the policy to end segregation within the armed forces and civil service in 1948. At the 1948 convention, liberal Democrats seized the momentum and won acceptance for an aggressive civil rights plank in the party’s platform, precipitating a walkout by the Southern “Dixiecrats.” These moves catalyzed (p.131) a broader realignment that made African Americans a crucial constituency of the national Democratic Party. The politics of civil rights were thus changing fast. As a result, and unlike after Reconstruction, efforts by the Supreme Court to make civil rights—and especially voting rights—effective for African Americans would henceforth find substantial constituencies within the political branches.

Against this backdrop, the Supreme Court decided its final “white primary” case in 1953. In Terry v. Adams, stopping race-based exclusion from voting required the Court to deviate still further from strict adherence to the state action rule. In Fort Bend County, Texas, the Jaybird Democratic Association, “a self-governing voluntary club,” held pre-primary elections prior to the official Democratic primaries.35 Since 1889, the winner of the Jaybird primary almost invariably went on to enter the Democratic primary, to run without opposition, and to win the general election. Black voters in Ford Bend County challenged their exclusion from the Jaybird primary under the Fifteenth Amendment.

The Court sustained the challenge, but divided on the reasoning. Writing for himself and two others, Justice Black seemed to toggle between two rationales, neither stated with precision. The first was that the Jaybird primary itself constituted state action. By virtue of its decisive role in selecting public officials, the Jaybird primary was an “election” within the meaning of Fifteenth Amendment.36 The second was that the Fifteenth Amendment imposed an affirmative duty on the state to bar the discrimination practiced by the Jaybird club:

For a state to permit such a duplication of its election processes is to permit a flagrant abuse of those processes to defeat the purposes of the Fifteenth Amendment. . . . It violates the Fifteenth Amendment for a state, by such circumvention, to permit within its borders the use of any device that produces an equivalent of the prohibited election.37

Writing for himself and three others, Justice Clark concluded that the Jaybird club constituted a political party, so that the case was covered by Allwright. Justice Frankfurter found that the state action requirement was met because some county officials participated in the Jaybird primary, and Justice Minton dissented, finding no state action present.

The white primary cases together illustrate how the growing political and legal potency of civil rights claims put pressure on the state action requirement. (p.132) The successive rounds of litigation amounted to a dialectical process, wherein Texas distanced its official state institutions further and further from voting exclusion, drawing the Court deeper and deeper into private associations in an effort to root out discrimination. And importantly, the Court was not alone in pushing civil rights into new terrain. Private and public streams of litigation, ably tended by the NAACP and the Justice Department, fed the Court a steady diet of arguments and incrementally built an infrastructure of useful precedents. More broadly, by the early 1950s, the cause of civil rights was beginning to find both partisan and popular constituencies.

In going after discrimination outsourced to private associations, the Supreme Court did not abandon the venerable state action framework, unleashing consequences that would be hard to predict and control. Rather, the Court produced ad hoc rationalizations of its results, generating confusion along the way. In the Terry opinions alone, we can discern four distinct arguments why the state action requirement is met—and one why it is not.

In some respects, Terry showcased rights’ profound disruptive potential. The Court permitted the Fifteenth Amendment voting right to be used not just as a shield but as a sword and, moreover, a sword wielded against a private party. The defendant Adams was president of the Jaybirds. The suit against him sought damages as well as injunctive relief.38 Moreover, the Court treated the right against race-based voting discrimination as an entitlement, not a value. Once the right is implicated, it’s not subject to balancing against other interests: it dictates the outcome of the case.

But at the same time, the broader impact of the decision was blunted by the Court’s failure to embed it in a satisfactory theory to explain why, when, and how constitutional rights could have horizontal effect. The Court did not provide much of a guide for lower courts, legislatures, or litigants to use in further elaborating what rights meant for the interactions of private individuals. Ultimately, the impact of Terry did not extend far beyond the peculiar phenomenon of the Jaybird club and similar organizations.39

Racially Restrictive Covenants

Terry shows the Supreme Court at mid-century reaching out to strike down consequential acts of discrimination initiated by private actors without reconceiving the state action framework inherited from the Civil Rights Cases, but by (p.133) making ad hoc exceptions to it. The Court’s most famous state action decision, Shelley v. Kraemer, offers an even better illustration of the phenomenon.

The case played out against the background of racial residential segregation in the United States, a phenomenon the Supreme Court itself had played some part in shaping. In the 1917 case Buchanan v. Worley, the Supreme Court struck down, as violating the Fourteenth Amendment, a Louisville, Kentucky, ordinance that made it unlawful for “any colored person” to move into a block with a majority of white residents.40 Racially restrictive covenants—agreements inserted into deeds that barred the sale of property to African Americans or other minority groups—were a private law alternative to municipal ordinances. Restrictive covenants existed before Buchanan was decided, but their adoption spread in the 1920s, not just through the South but through the North and Midwest, especially in cities that received large numbers of African Americans who left the South as part of the Great Migration (Jones‐Correa 2000, 551–59). The Supreme Court declined to hear a challenge to the constitutionality of restrictive covenants in 1926, on the grounds that the record did not present a substantial question of constitutional law, in Corrigan v. Buckley.41 Residential segregation was bolstered by the federal government dating back to the New Deal, when the Home Owners Loan Corporation, a forerunner to the Federal Housing Administration, made neighborhood “stability” (including racial homogeneity) a key desideratum for federal mortgage underwriting standards (Jones‐Correa 2000, 565–66).

Two decades after Corrigan, the Court granted review of two state supreme court decisions, one from Missouri and the other from Michigan, that upheld restrictive covenants against constitutional challenge. Both of the state supreme courts had sustained the validity of the covenants on the grounds that state action was lacking: the covenants represented merely private agreements.42 Both state supreme courts had also rejected arguments that the covenants were invalid as contrary to public policy.

Overturning the rulings on the basis that this latter conclusion was wrong—in other words, holding racially restrictive covenants to be incompatible with (p.134) public policy—was not an option for the U.S. Supreme Court. The supreme courts of Missouri and Michigan are the final authorities on the law of Missouri and Michigan, respectively, and the content of public policy is a matter of state law. The U.S. Supreme Court could trump their rulings only by bringing the Constitution into play. Accordingly, the briefs of the petitioners and a number of amici curiae, including the United States, furnished the Court with arguments why the enforcement of such covenants ran afoul of the Constitution.

In his opinion for a unanimous (if short-staffed) Court,43 Chief Justice Vinson ruled that the Constitution forbade the enforcement of the racially restrictive covenants. Vinson’s opinion addressed the state action issue head on. Citing to Buchanan, Vinson observed that racial restrictions on the sale of property would be clearly unconstitutional if adopted by state statute or local ordinance. By contrast, these privately adopted covenants by themselves would not violate the Constitution, if “effectuated by voluntary adherence to their terms.”44 Here, however, “there was more”: “These are cases in which the purposes of the agreements were secured only by judicial enforcement by state courts of the restrictive terms of the agreements.”45 As the Court framed it, the central question in the case became whether judicial enforcement of the covenants counted as “state action.”

Framed this way, the Court’s question was easy to answer in the affirmative. The Court unleashed a cavalcade of quotations and citations to the effect that judicial action is state action.

But this conclusion is equal parts uncontroversial and irrelevant. The discriminatory choice at the root of the constitutional violation here is the action of private parties: it is the inclusion of a racially restrictive covenant in the deed. It can’t be that judicial involvement introduces the element of state action into every legal relationship of private parties, such that the norms governing the parties are attributed to the state, or else the state action requirement becomes a nullity.

So why, then, is judicial enforcement of a racially restrictive covenant unconstitutional? This is the real question at the heart of the case, and the Court dodges it, using the state action doctrine in a kind of sleight of hand. In this respect, Shelley resembles Flagg Brothers, though the misdirection works to opposite ends in the two cases. In Flagg Brothers, the Court rested its ruling on the common-sensical proposition that the warehouseman’s sale of property (p.135) is not an action of the state, drawing attention away from what the state did do: authorizing the sale under statute. In Shelley, the Court points to judicial enforcement of the covenant to unleash the horizontal effect of the Fourteenth Amendment. But what is lacking is a substantive theory of the equality right to explain why it bars enforcement of discriminatory agreements drafted by private parties and to indicate prospectively when the logic of Shelley does and does not apply. The Court’s ruling here is strong medicine—it overwrites the content of agreements whose validity is unimpeached within the terms of private law—but it is administered without prescription.

Legislative Constitutionalism and Its Limits

The cause of civil rights gained significant political traction in the 1960s, finding champions in the White House and winning the support of majorities in Congress. In the same years, the Supreme Court advanced civil rights judicially in ways that complemented and, to some extent, went beyond the efforts of the political branches. The Court’s role involved, first of all, getting out of Congress’s way. To an unprecedented degree, the Court permitted Congress to use its enforcement power under the Reconstruction Amendments to define what constitutional equality rights would mean on the ground. But even so, the Court was only willing to let Congress share in the task of constitutional construction within limits. Also in this period, the Court invoked more exceptions to the state action norm, without ever abandoning the state action framing, to reach pockets of racial discrimination beyond the scope of federal legislation.

As I argued above, introducing the state action doctrine had originally allowed the Supreme Court to contain the disruptive potential of constitutional rights and to consolidate its own authority in a period of institutional vulnerability. Rights were easier to manage when they operated in one dimension only: the vertical dimension, containing the relationship between citizen and state. Bringing rights to heel ex ante, by narrowing their scope of application, kept the Court from being dragged, through litigation, into unwelcome disputes over discriminatory social institutions. It also helped the Court establish a strong brand of judicial supremacy, in which rights functioned as entitlements with determinant content and the Court was the authoritative arbiter of that content.

In its 1960s decisions on Congress’s power to enforce the Reconstruction Amendments, the Court relaxed, to some degree, its monopoly over authoritative constitutional interpretation. The Court did so most consequentially in the context of the Voting Rights Act of 1965 (VRA). The disenfranchisement of African Americans in the former states of the Confederacy was extensive and enduring: in many places, voting rolls in the early 1960s were nearly as lily-white (p.136) as they had been in the 1890s. Southern states used numerous means to block access to the ballot, including literacy tests and other voting requirements that were facially neutral but applied in a discriminatory fashion. An earlier legislative effort, the Civil Rights Act of 1957, had limited success in rooting out these practices. The 1965 Act went further, effectively putting elections in the South into federal receivership. In addition to banning the use of any voting qualification, standard, practice, or procedure resulting in the denial or dilution of the right to vote on account of race, including a number of specified procedures, the VRA required that jurisdictions meeting certain criteria have any changes to their voting rules precleared by the Justice Department to ensure against discriminatory effects. The states originally covered by the preclearance formula, which hinged on voting practices and participation rates, included most of the states of the former Confederacy, and none outside of it.

In two important cases from 1966, the Supreme Court upheld the VRA against claims that it exceeded Congress’s authority. In South Carolina v. Katzenbach, the Court held that Congress was authorized under the enforcement clause of the Fifteenth Amendment even to ban practices that the Court itself had held did not violate the right to vote.46 South Carolina invoked the Supreme Court’s original jurisdiction to challenge the preclearance provisions, arguing that the VRA was not “appropriate” legislation to enforce voting rights within the meaning of Section 2 of the Fifteenth Amendment, because it proscribed more than denials of the vote based on race. Chief Justice Warren’s opinion for the Court rejected that argument. After recounting Congress’s difficulties in securing effective voting rights in the face of Southern efforts to undermine them, the Court concluded that Congress must have broad discretion in choosing the means to secure the right. Chief Justice Warren invoked the Supreme Court’s most famous case about the powers of Congress:

The basic test to be applied in a case involving § 2 of the Fifteenth Amendment is the same as in all cases concerning the express powers of Congress with relation to the reserved powers of the States. Chief Justice Marshall laid down the classic formulation, 50 years before the Fifteenth Amendment was ratified:

“Let the end be legitimate, let it be within the scope of the constitution, and all means which are appropriate which are plainly adapted to that end, which are not prohibited, but consist with the letter and spirit of the constitution, are constitutional.”47

(p.137) The message was clear: the Court would judge Congress’s choice of means in enforcing rights just as with respect to Congress’s other enumerated powers: with ample deference.

Katzenbach v. Morgan raised a challenge to VRA Section 4(e), which barred the use of literacy tests as voting qualifications against persons educated in Puerto Rico.48 The complication here was that the Supreme Court had recently ruled that literacy tests do not violate the Fourteenth Amendment.49 The question was whether Congress was nonetheless authorized to ban the use of literacy tests under its power to enforce the Fourteenth Amendment.50 In order to answer yes, the Court could not rely on the logic of South Carolina: that Congress could create a protective buffer, in order to protect the right at its core. The Fifteenth Amendment guarantees a right to vote without regard to race, and to safeguard that right, Congress could proscribe facially neutral practices that had been used to undercut that right. But by contrast, the Supreme Court had ruled that the Constitution permitted literacy tests as voting qualifications. Banning the use of literacy tests against those schooled in Puerto Rico did not obviously serve to vindicate any constitutional right as identified by the Supreme Court. Sustaining Section 4(e) would seem to require granting Congress some latitude to define the Equal Protection right for itself, in opposition to the Supreme Court’s law on point.

The Supreme Court turned back the challenge to the law. In places, Justice Brennan’s opinion for the Court suggests that Congress might get to decide for itself what the right to equal protection required. Congress might have good reasons for restricting the use of literacy tests, according to Justice Brennan. Congress could have questioned whether New York’s test was in fact motivated by prejudice, or “whether denial of a right deemed so precious and fundamental in our society was a necessary or appropriate means of encouraging persons to learn English, or of furthering the goal of an intelligent exercise of the franchise.”51 Elsewhere in the opinion, Justice Brennan links the ban on literacy tests, somewhat tenuously, back to other Fourteenth Amendment concerns. Broader ballot access for Puerto Ricans could translate into “enhanced political power,” which would in turn “be helpful in gaining nondiscriminatory treatment in public services for the entire Puerto Rican community.”52

(p.138) Justice Harlan, joined by Justice Stewart, dissented. In proscribing a practice the Court had found consistent with the Fourteenth Amendment, Congress was not enforcing the amendment but “defin[ing] the substantive scope of the Amendment.”53 And this Congress could not do: “That question [whether a practice violates equal protection] is one for the judicial branch ultimately to determine.”54

Arguably Jones v. Alfred H. Meyer Co., decided in 1968, represented the high-water mark for the Court’s tolerance of legislative constitutionalism.55 The case concerned a challenge to Section 1 of the Civil Rights Act of 1866, since codified as 42 U.S.C. Section 1982. Section 1982 guaranteed “all citizens . . . the same right, in every State and Territory, as is enjoyed by white citizens thereof to inherit, purchase, lease, sell, hold, and convey real and personal property.” When Joseph Jones and his wife were refused the chance to purchase a home in a St. Louis subdivision because they were black, they sued for damages and injunctive relief under Section 1982. In Hurd v. Hodge,56 a companion case to Shelley v. Kraemer, the Supreme Court had ruled that Section 1982 could be used as a shield, to block the rescission of a home sale to African American buyers.57 Hurd relied on the Shelley logic, that the judicial enforcement of a restrictive covenant added the element of state action. But that rationale was not available here, since the plaintiffs were attempting to use Section 1982 as a sword not a shield. Section 1982 was the basis for a lawsuit brought by one private party against another, for damages and injunctive relief. And the claimed source for Congress’s authority to pass this provision was the Thirteenth Amendment enforcement power. The Thirteenth Amendment abolished slavery and involuntary servitude. Obviously, housing discrimination was not slavery. The Supreme Court could sustain the law only by giving Congress a fairly free hand in legislating under the banner of rights.

And in an opinion by Justice Stewart, the Court did. First, the Court addressed the horizontal effect issue, concluding that “the fact that § 1982 operates upon the unofficial acts of private individuals, whether or not sanctioned by state law, presents no constitutional problem.”58 The right against slavery, (p.139) unlike other rights, had been understood from the first to be a right erga omnes, directly binding private and public actors alike. But how did Congress’s power to enforce that right reach private real estate conveyances? The Court cited to the Civil Rights Cases for the proposition that the Thirteenth Amendment clothed Congress “with power to pass all laws necessary and proper for abolishing all badges and incidents of slavery in the United States.”59 Congress has the power “rationally to determine what are the badges and the incidents of slavery, and the authority to translate that determination into effective legislation.”60 And Congress’s choice here was rational, because “whatever else they may have encompassed, the badges and incidents of slavery—its ‘burdens and disabilities’— included restraints upon ‘those fundamental rights which are the essence of civil freedom, namely, the same right . . . to inherit, purchase, lease, sell and convey property, as is enjoyed by white citizens.’ ”61 “[W]hen racial discrimination herds men into ghettos and makes their ability to buy property turn on the color of their skin,” Justice Stewart concluded, “then it too is a relic of slavery.”62

These cases all show the Court permitting Congress to share in the task of giving content to equality rights. The Court relaxed the strict hierarchy of constitutional interpretation, just as it relaxed its adherence to the state action norm, in ways that pushed principles of racial antidiscrimination deeper into the fabric of American society. But in other cases from the same period, the Court resisted an expansive approach to Congress’s enforcement powers. Notably, the Court sustained the landmark Civil Rights Act (CRA) of 1964, which barred discrimination in public accommodations, as an exercise of Congress’s Commerce Clause power rather than its enforcement power under the Fourteenth Amendment.63

The biggest obstacle to rooting the decision in the enforcement power was the state action doctrine: the CRA banned discrimination by private parties. In a concurrence, Justice Douglas argued that the Court should have relied at least in part on Section 5 of the Fourteenth Amendment in sustaining the statute. The CRA was appropriate legislation to enforce the equal protection right because the requisite state action was present, in the availability of judicial enforcement of private discrimination, à la Shelley v. Kraemer. But the (p.140) majority opted to reach its result via an expansive approach to the Commerce Clause, rather than either granting Congress more leeway to define rights for itself (which would trench on the Court’s domain) or generalizing the Shelley approach to state action (which would theoretically make all private discrimination constitutionally actionable). Taken together, these cases suggest that the Supreme Court permitted legislative constitutionalism only as a last resort: when no other plausible argument would sustain important antidiscrimination legislation.

The State Action Doctrine and the Civil Rights Movement

Notwithstanding Congress’s legislative promotion of civil rights, there was still significant societal discrimination that was not reached by the federal statutes, especially but not exclusively in the South. And without a federal hook, these would remain matters of state law, presided over by state legislators and judges who, especially in the South, were ferociously hostile to civil rights. But the Warren Court repeatedly stepped in to measure practices in the states against federal equality norms, using the tool that unlocked federal jurisdiction: the Constitution.

Constitutionalizing issues beyond the reach of federal civil rights legislation meant finding exceptions to the strict state action requirement. And a great deal of the Court’s state action doctrine—which amounts to a series of exceptions to and qualifications of the strict state action rule—was forged in the 1960s, in cases involving racial discrimination (Fee 2004, 578; Eskridge 2002, 2331–32). A sample of prominent cases gives a sense for the various workarounds the Court devised to the strict state action rule. In Burton v. Wilmington Parking Authority64 (which preceded the Civil Rights Act), the Court ruled that a private restaurant could not deny service to customers on the basis of race, because it rented space in a public parking facility. In Evans v. Newton,65 the Court held that a park, originally devised to a city with the proviso that it be used by white people only, retained its public character and could not discriminate on the basis of race, even after the city resigned trusteeship and the park passed into private hands. And in Reitman v. Mulkey,66 the Court invalidated a California referendum that barred state interference with property owners’ freedom to sell, rent, or lease their property to whomever they choose: the referendum, (p.141) which voided existing antidiscrimination statutes, would “involve the State in private racial discriminations to an unconstitutional degree.”67

The 1964 case Bell v. Maryland offers a revealing glimpse into the Supreme Court’s strategic approach to state action issues.68 The Court was asked to pass on the constitutionality of the criminal trespass convictions of twelve African American students who had staged a sit-in at a Baltimore restaurant in 1960. From the perspective of state action doctrine, this case posed a hard problem. The law of trespass does not itself embody a discriminatory norm: the discrimination comes entirely from the restaurant owner. This situation is not so very far afield from the classic example of private discrimination beyond the reach of constitutional scrutiny: a homeowner’s use of trespass laws to permit whites but not African Americans onto his property (Tushnet 2003, 81). It would be difficult for the Court to vacate the convictions without repudiating the state action rule.

In fact, the papers of the Justices indicate that some were willing to do that. Justice Clark circulated a draft opinion that would extend constitutional scrutiny to “usages . . . which have become a kind of common law.”69 Others on the Court voted to retain the state action rule and uphold the convictions. The issue closely divided the Court, and changes of heart gave first one position and then the other a tentative 5–4 majority. Ultimately, Justice Brennan balked at rejecting the state action rule out of concern over the effect such a ruling would have on the ongoing effort in Congress to craft civil rights legislation. A sure-to-be-controversial show of judicial power at this moment might shake loose Southern senators from the coalition moving the Civil Rights Act toward passage (Ackerman 2014, 146). Instead, the Court opted to stay its hand. The Court remanded the case to Maryland with instructions to consider the convictions in light of a recently passed state antidiscrimination statute. Weeks later, Congress did pass the Civil Rights Act, which proscribed the sort of discrimination at issue in Bell legislatively, on a national basis. Bell and the cases discussed above together illustrate the pattern: the Warren Court generally turned to the Constitution to fill the normative gaps only when private discrimination tolerated by state law was beyond the reach of federal legislation.

The Court’s horizontal deployment of rights, together with Congress’s antidiscrimination legislation, contributed to a far-reaching transformation of American society, in which open and notorious racial discrimination was dramatically reduced. And then, the moment passed. The retirement of Earl (p.142) Warren and his replacement with Warren Burger as Chief Justice in 1969 marked the end of the era in many respects. The Burger Court would prove itself far less interested than its predecessor in advancing the reach of rights. Changes on the Court mirrored, and were partially the result of, broader political changes. The last major piece of civil rights legislation was the Civil Rights Act of 1968, which banned discrimination in the housing market and was less popular among whites nationally than the 1964 Act, which had mostly affected the South. Campaigning for the presidency that year, Richard Nixon was critical of the Warren Court and promised to appoint “strict constructionists” to the Court (Engel 2011, 302–05). And once elected, Nixon appointed Justices who helped to wind down the state action innovations of previous years, including when racial discrimination was involved.70 Rather than repudiating the new branches of state action doctrine outright, the Court more often added exceptions to the exceptions, further contributing to the subject’s reputation as an unruly mess.

Freedom of Expression

There are a number of areas in American law where private law rules and arrangements can implicate the freedom of expression. In contract law, individuals may be subject to legal liability for speech in violation of confidentiality agreements (Garfield 1998). The tort of invasion of privacy also can make it actionable for one private party to communicate information about another private party (Emerson 1979). The tort of libel can do the same, as can the tort of intentional infliction of emotional distress.

Ordinarily, these are all areas in which the legal relations of nonstate parties play out under the auspices of the common law, as curated by state judges. And the Supreme Court has largely permitted common law regulation to run its course, except insofar as the common law threatens to block the flow of information on matters of public concern.71 The Court has confronted this (p.143) problem in the context of libel and (to a lesser degree) the intentional infliction of emotional distress. The Warren Court significantly constitutionalized libel law in particular. It makes sense that the Court would relax the state action norm in the context of these forms of private speech regulation more so than others. Libel cases, especially involving issues of public concern, implicated the Warren Court’s push to expand civil rights and civil liberties more directly than the other classes of cases.

The Supreme Court had been alternating between periods of beefing up its protection for speech rights and backpedaling since the late 1930s (Shapiro 1966). By 1962, the high-protectionist position secured a majority on the Court (Siegel 2006, 362). From that point forward, the Court’s jurisprudence consistently reflected the classic civil libertarian principle that the law could not be used to exclude views, even unpopular ones, from the public sphere (Van Alstyne 2006).72 In the Court’s typical speech cases, the threat to freedom of expression came from government actors: school boards,73 Congress,74 state lawmakers,75 and so forth. Neither the enforcement of confidentiality agreements nor actions for invasion of privacy were likely to pose much of a threat to core First Amendment values, as seen from the civil libertarian perspective: the first mostly concerned matters of commerce, and the second involved disclosures with little relevance to public affairs. But libel actions, if brought by public figures, had the potential to chill debate on matters of public concern.

When this possibility presented itself, the Court did not hesitate to expand the scope of First Amendment protections to bring libel within its purview. The Supreme Court made this move in the landmark 1964 case of New York Times Co. v. Sullivan.76 The Supreme Court’s turn to the Constitution in New York Times was almost overdetermined. The case not only involved political speech but race and civil rights, issues central to the Court’s rights-protecting role (Feeley 2006; Kalven 1965). The case concerned an advertisement that ran in the New York Times on March 29, 1960. The ad, which was raising funds (p.144) for Martin Luther King, Jr.’s legal defense,77 described civil rights abuses in the South, including a few specific incidents in Montgomery, Alabama. L.B. Sullivan, the police commissioner in Montgomery, brought a libel suit against the newspaper and four African American ministers, alleging that the ad contained falsehoods that damaged his reputation.78

Sullivan was not mentioned by name in the advertisement—indeed, no officials from Alabama were. Nonetheless, the case was tried (in a segregated courtroom), and an Alabama jury returned a $500,000 verdict against the defendants, which the Alabama Supreme Court upheld. The size and circumstances of the verdict were unprecedented. The common law defined defamatory statements as those that tend to subject a person to hatred, contempt, or ridicule (Forer 1987, 52). While this cast a very wide net in theory, in practice, libel suits generally involved expense, delay, and small recoveries, leading one authority to opine that “[n]o other formula of the law promises so much and delivers so little” (Green 1960, 907).79

But the Sullivan verdict was enormous, and it appeared to be the harbinger of more to come. After Sullivan filed suit against the Times and the ministers, so did four other Alabama officials, including the governor, for damages totaling $3 million (Lewis 1992, 13).80 A second jury returned a $500,000 verdict for the mayor of Montgomery a few months after the Sullivan trial. These damages were so substantial as to pose an existential threat to the New York Times. And Sullivan’s success at trial inspired a frenzy of imitators seeking to shut down news coverage of civil rights topics: by the summer of 1964, Southern officials had filed libel actions seeking more than $300 million in damages from the press (Lewis 1992, 36). Southern officials had retooled libel litigation into a very potent threat to the civil rights movement and the freedom of the press. And because libel law was state law, there was no scope for federal court oversight of these lawsuits—unless they raised a constitutional dimension.

(p.145) The New York Times sought a writ of certiorari from the Supreme Court, raising the argument that the verdict violated the paper’s First Amendment right to freedom of the press.81 The Court granted review and reversed in a unanimous opinion authored by Justice Brennan. The Court acknowledged that it had never before found libel actions to raise constitutional concerns, but also noted that this case was different. Never before had the Court “sustained the use of libel laws to impose sanctions upon expression critical of the official conduct of public officials.”82 The difference matters, because “freedom of expression upon public questions is secured by the First Amendment.”83 The law of libel as applied in Alabama courts would undermine First Amendment freedoms by imposing “the pall of fear and timidity . . . upon those who would give voice to public criticisms.”84 The Constitution, therefore, bars recovery for libel by a public official unless he can prove the offending statement was made with “actual malice,” which is to say, either knowledge that it was false or reckless disregard for its possible falsity.85

But before coming to this conclusion, the Court first had to confront the state action issue, which was one ground on which the Alabama Supreme Court rejected the defendants’ constitutional claims. The state action rule was inapposite here, according to the Supreme Court:

Although this is a civil lawsuit between private parties, the Alabama courts have applied a state rule of law which petitioners claim to impose invalid restrictions on their constitutional freedoms of speech and press. It matters not that the law has been applied in a civil action and that it is common law only, though supplemented by statute. The test is not the form in which state power has been applied but, whatever the form, whether such power has in fact been exercised.86

And so the Supreme Court brought libel actions within the scope of constitutional scrutiny. In reality, the case did not pose a very difficult state action problem. The Court was not facing, as in Shelley, a court merely standing ready to (p.146) enforce agreements whose contents were dictated by private parties. The contents of Alabama libel law were the product of Alabama courts. The courts, in effect, furnished to private parties a muzzle they could use to shut up other private parties. The state’s hand would be somewhat easier to see had it expressed its policy legislatively rather than in judge-made common law. But the Court noted that a rule of law is no less an instrument of the state for being embodied in common law, instead of a statute.87 A comparison with Flagg Brothers is instructive. In that case, the state furnished a legal device that affected private-party legal relations—by way of a statute, no less—and yet there, the Court could see no state action. The Court is simply more likely to see state action when bringing the Constitution to bear suits its purposes.

And the Sullivan decision closed a potential private law loophole to its First Amendment jurisprudence that threatened to undermine core constitutional values. Sullivan was a landmark—“an occasion for dancing in the streets,” in the words of First Amendment éminence grise Alexander Meikeljohn—and it effectively neutralized the libel suit as a weapon against the civil rights movement (Forer 1987, 17). Sullivan broke the state action barrier for libel actions, so that all constitutional objections to libel suits henceforth would have to be entertained on the merits. And while the Supreme Court’s First Amendment libel jurisprudence grew more fragmented over time (Emerson 1970, 525), the Court by and large continued to follow the civil libertarian lodestar, protecting speech more aggressively when it implicated political debate and matters of public interest. In the absence of actual malice, speech was protected against libel claims by public officials (and soon, public figures)88 when it addressed questions of public concern. The Court extended the mantle of First Amendment protection against claims of intentional infliction of emotional distress under (p.147) parallel circumstances.89 The First Amendment posed less of a bar to recovery when plaintiffs were not public figures.90

The Sullivan decision was powerful, but also surgical: it stopped libel actions from impeding the civil rights movement and from putting the brakes on other political debates. One reason that the case’s impact was so tightly targeted is that the Supreme Court defined the horizontal effect of the First Amendment right in terms of an entitlement rather than a value. The Supreme Court did discuss the values that underlie the needs for First Amendment protection. But rather than leaving it there, and letting other courts balance those values against competing considerations in future cases, the Supreme Court articulated a bright-line rule. Specifically, the First Amendment grants a privilege to publish good faith criticisms of government officials (Tribe 1988, 864–65). True to form, even as it expanded the scope of constitutional protection, the Supreme Court did so by way of a tailored doctrinal formula that kept the Court in the driver’s seat of constitutional development.

Family Law

A discussion of the Supreme Court’s use of constitutional rights to adjust legal relations within the family will necessarily be short, because the Supreme Court has almost never used constitutional rights to adjust legal relations within the family. The explanation for the Court’s inactivity on this front is not that relations within the family do not potentially implicate constitutional values; they do. Consider parental restrictions of children’s speech, discriminatory terms in family trusts, and testamentary restraints on remarriage. To take one concrete example, the will of Frank Mandelbaum, which was contested in 2012, used the threat of disinheritance to attempt to coerce his gay son to marry a woman (Popoff 2014).91 Constitutional values—of free expression, equality, freedom of choice—are at stake within the family.92

(p.148) The Supreme Court’s choice not to intervene in this domain, like the silence of the dog that did not bark, signifies something.93 Constitutionalizing the legal rules that govern family member interactions would not meaningfully advance a normative vision or strategic agenda that the Court has promoted. But doing so would bring the Court into conflict with state courts and legislatures, and would multiply the number of sites where constitutional law is made in ways that the Court might find difficult to control.

Moreover, rather than breaking the family open to public law values, the clear trend in the Court’s jurisprudence in the past half-century has been to emphasize the private character of the family sphere and its consequent immunity from some forms of state intervention. This was the message of Griswold v. Connecticut,94 Roe v. Wade,95 and Lawrence v. Texas,96 among other cases.97 In these and similar cases, the Court had chosen to focus on threats to liberty emanating from outside the family, rather than those coming from within the family. Threats from outside—government efforts to dictate sexual practices, reproductive choices, or family living arrangements—are more readily identifiable as problems within the civil libertarian template and are also easier for courts to manage.

The net result of the Court’s forbearance was to leave to state legislatures the project of updating family law to better comport with changing norms. This is a role that states have long occupied. State legislatures in the nineteenth century had eliminated the institution of coverture, in which women’s legal personality was effectively dissolved upon marriage, and twentieth-century legislatures relaxed the legal regulation of sexual practices and liberalized divorce. William Eskridge and John Ferejohn argue that these developments amount to “small ‘c’ constitutional change” (2010, 238). Demographic and technological changes, social movements, administrative and legislative action, even Alfred (p.149) Kinsey all contributed more to the “constitutional shift” they describe than the Supreme Court.

The clearest point of entry for the Court to constitutionalize family law, even to a limited extent, would be to recognize that individuals’ rights can oblige the government to provide protection that involves interventions in familial relations. The Court roundly rejected this possibility in DeShaney v. Winnebago County Department of Social Services.98 The facts of DeShaney are brutally sad. A divorce court had awarded custody of Joshua DeShaney to his father. State social services repeatedly received credible evidence that Joshua was being beaten by his father. Social services made visits to Joshua’s home and recommendations to improve the home environment, but did not remove Joshua from his father’s custody, even as the evidence suggested abuse was continuing. At age four, Joshua was beaten so severely by his father that he fell into a coma. The cumulative effects of Joshua’s abuse resulted in profound permanent brain damage. Joshua’s mother brought suit on her own and his behalf against the Winnebago County Department of Social Services and several employees, alleging that their failure to intervene amounted to a deprivation of substantive due process, in violation of the Fourteenth Amendment. The district court granted summary judgment for the defendants, and the court of appeals affirmed.

In their brief to the Supreme Court, the DeShaneys argued that the Fourteenth Amendment creates an affirmative duty to protect children. This would amount to indirect horizontal effect: the Constitution obligates the state to regulate the family—for instance, by terminating an abusive parent’s custodial or parental rights—in ways that impact the legal relations of its members. The Supreme Court rejected the argument and affirmed the lower court. Writing for a six-Justice majority, Chief Justice Rehnquist leaned hard on the action part of the state action requirement:

[N]othing in the language of the Due Process Clause itself requires the State to protect the life, liberty, and property of its citizens against invasion by private actors. The Clause is phrased as a limitation on the State’s power to act, not as a guarantee of certain minimal levels of safety and security. It forbids the State itself to deprive individuals of life, liberty, or property without “due process of law,” but its language cannot fairly be extended to impose an affirmative obligation on the State to ensure that those interests do not come to harm through other means. . . . Its purpose was to protect the people from the State, not to ensure that the State protected them from (p.150) each other. The Framers were content to leave the extent of governmental obligation in the latter area to the democratic political processes.

Consistent with these principles, our cases have recognized that the Due Process Clauses generally confer no affirmative right to governmental aid, even where such aid may be necessary to secure life, liberty, or property interests of which the government itself may not deprive the individual.99

The Court also rejected the idea that the state had incurred affirmative obligations vis-à-vis Joshua DeShaney by undertaking to protect him. The majority also recognized the possibility of a normative gap between the accountability provided by statutory law and the justice that “natural sympathy” demands, but denied that it should be filled by the Constitution. Instead, it should be filled by lawmaking at the state level:

The people of Wisconsin may well prefer a system of liability which would place upon the State and its officials the responsibility for failure to act in situations such as the present one. They may create such a system, if they do not have it already, by changing the tort law of the State in accordance with the regular lawmaking process. But they should not have it thrust upon them by this Court’s expansion of the Due Process Clause of the Fourteenth Amendment.100

DeShaney illustrates the Court’s antipathy to any conception of rights that might obligate the state to intervene within families. The Supreme Court has shown itself to be allergic to positive conceptions of constitutional rights in general. Even the dissenters rejected the idea that the Constitution imposes an affirmative obligation on the state to intervene within families. The Court argues that positive conceptions of rights are not part of our constitutional tradition, but as the previous chapter suggests, the reality is actually more complicated. It is true, on the other hand, that positive rights pose a significant management challenge for the Court. To the extent that constitutional rights impose affirmative duties on the state, countless thousands of legislative and executive actions and omissions take on a new constitutional salience. It is hard for a tribunal to exercise any control over such an unruly mass of actions, and since the first postwar constitutional settlement, the Supreme Court has made controlling constitutional interpretation central to its role in the American state. To the (p.151) extent that extending the Constitution’s reach is advantageous, it is far easier for the Court to do so through ad hoc adjustments to the state action doctrine. And in the context of family law, there are simply few normative or strategic imperatives even to do this.101


This chapter has argued that state action doctrine has proved an effective tool for the Supreme Court’s management of (potentially unruly) constitutional rights. The stark formulation of the general state action rule allowed the Court to keep a lid on constitutional pluralism, while the Court has licensed deviations from the rule to avoid unpalatable results and keep states in line. The Court has never articulated a theory of when, exactly, constitutional rights require states to curb private behavior, but this vagueness has its use as well, in securing to the Supreme Court the role of sole authoritative state action gatekeeper.

The state action rule was born out of the Supreme Court’s denial of horizontal effect in the years after Reconstruction. Then for a generation after the end of World War II, the Court applied rights horizontally in a number of high-profile contexts, all while maintaining the rhetorical insistence on a state action requirement. Extending the reach of rights into private orderings served the Court’s interests in the second period in a way that it had not in the first. Crucially, there was much broader political and institutional support for the nationalization of rights protection in the years after the Second World War than in the years after construction. After retiring from its (self-appointed) pre-New Deal role as council of revision for economic legislation, the Court found a new role for itself in the American political economy, one in which the promotion of civil rights and civil liberties was central to its raison d’être. As the preceding sections have argued, the Court relaxed the state action rule where doing so furthered this agenda.

At times, the Supreme Court permitted rights to disrupt private-party relationships forcefully—for instance, by generating causes of action against private parties for legal and injunctive relief. Most of the time, though, constitutional rights had no effect on private parties. State action doctrine itself did not offer an adequate explanation for when rights would and would not be permitted to matter and, as a result, has justly received low marks for coherence. Rights (p.152) apply to private-party relations when the Supreme Court says they do. In a sense, the decisionism of the state action jurisprudence is the obverse face of the judicial supremacy narrative that the Court continues to cultivate. If the Court is the oracle of the Constitution, whose rulings are supreme just because they are the Court’s, it has no need to hem itself in with an intelligible theory of horizontal effect. State action doctrine continues to function as a constitutional containment device even while the Supreme Court licenses departures from it.102 Although the Court allowed Congress some room to share in the work of defining rights to a limited extent, this did not unleash a more general constitutional cascade that fundamentally blurred the categories of ordinary and constitutional law, or the roles of court and legislature.

In recent years, prominent scholars have argued that the lines between ordinary and constitutional law are blurred in the United States as a practical matter, and that statutes have a constitutional character. There are different aspects to, and variants of, the argument.103 William Eskridge and John Ferejohn coined the term “super-statutes” to describe a class of legislative enactments that enjoy a kind of quasi-constitutional stature (Eskridge and Ferejohn 2001).There are two chief respects in which these statutes (which by Eskridge and Ferejohn’s lights include the Sherman Antitrust Act, the National Labor Relations Act, the Civil Rights Act of 1964, and many more) resemble constitutional commitments. First, like the Constitution, they are deeply entrenched in our law, and therefore hard to change.104 Second, they are distinguished from ordinary statutes by their substantive importance. In Eskridge and Ferejohn’s words, “[i]n America’s republic of statutes, republican deliberation over fundamental national commitments has migrated, relatively speaking, away from Constitutionalism and towards legislative and administrative constitutionalism” (Eskridge and Ferejohn 2010, 16). Eskridge and Ferejohn claim that these statutes exercise a (p.153) gravitational force on other law and argue that their special stature justifies a special solicitude from courts, which includes liberal, purposive construction and ample deference to administrative interpretations.105

In a similar vein, Bruce Ackerman argues that “Our Living Constitution” lives on through statutory law, which changes over time to meet the fundamental challenges the polity faces even as the formal Constitution remains frozen (2007). Ackerman, like Eskridge and Ferejohn, believes that “[t]he Supreme Court has an institutional obligation to recognize that superprecedents crystallize fixed points in our constitutional tradition, and should not be overruled or ignored in the course of doctrinal development” (2007, 1752). But while Eskridge and Ferejohn treat legislative constitutionalism as a fait accompli, Ackerman is less sanguine. Notwithstanding their institutional entrenchment, super-statutes in Ackerman’s treatment come off as more vulnerable to judicial undermining, and Ackerman is more critical of the Court’s handling of them, especially in recent years (2014, 328–37).

It is useful to consider these claims in a comparative context, and this project helps us to do that. Eskridge and Ferejohn are certainly correct that many of the most fundamental collective choices of the American people are expressed in statutes. In that sense, and because their institutionalization makes them “sticky,” it is reasonable to consider landmark statutes to have a small-c constitutional character. But what comparison with Germany points up is how these policy processes are cut off from the conceptual and institutional resources of constitutionalism proper. In Germany, the high courts’ constitutional jurisprudence has treated rights as the embodiments of values that pervade the whole legal system. This turns important legislative choices into the site for multi-branch “constitutional dialogues.” This means, on the one hand, that legislatures self-consciously engage in constitutional deliberation, and on the other, that when courts are called to intervene, they do so with some deference to the choices legislatures have made. Consider, for instance, how civil courts, the national legislature, and the Federal Constitutional Court collectively dragged German family law into compliance with the Basic Law’s gender equality mandate over a period of years. By comparison, the United States experiences a much more limited form of legislative constitutionalism.

Also, super-statutes enjoy their superpowers at the sufferance of courts. As a descriptive matter, Eskridge and Ferejohn have argued that courts—including the Supreme Court—by and large do treat super-statutes with extra deference, owing to their “quasi-constitutional” stature (Eskridge and Ferejohn 2001, 1217, (p.154) 1246–63). Ackerman is more skeptical, and his skepticism seems increasingly warranted. While the Supreme Court may in the past have shown compunction about sweeping aside the “quasi-constitutional” conclusions of the legislative branch, it shows little now.

The Court’s 2013 decision regarding Congress’s enforcement powers, Shelby County v. Holder, illustrates the point and offers a bookend to our discussion of earlier Voting Rights Act cases.106 When Congress reauthorized the Voting Rights Act in 2006, leaving in place the existing preclearance formula, Shelby County, a covered jurisdiction in Alabama, cried foul, arguing that Justice Department review of its voting practices could no longer be justified as a measure to enforce the right to be free from race-based voting discrimination.

The Supreme Court agreed. While the VRA’s coverage formula may have originally tracked voting discrimination, the country had changed dramatically since 1965, but the formula had not changed at all. The Warren Court’s permissive approval of legislation with a rational relation to the rights in question was replaced with a newfound parsimony. If Congress was to subject the voting practices of states to extra oversight, Chief Justice Roberts wrote for the 5–4 majority, it must do so on the basis of “current data reflecting current needs.” In other words, legislation that was at one time a valid exercise of Congress’s enforcement power will cease to be if Congress fails to update its statute to mirror the changing shape of the problem.

In the United States, the formal hierarchies of constitutional law over ordinary law, and court over legislature as steward of constitutional meaning, functionally dictate how policy-relevant decisions are made to a greater degree than in Germany or Canada. In the domain of ordinary law, the courts are assigned the role of legislative amanuensis, discerning and applying congressional intent. But when the Constitution is “switched on,” judicial construction displaces congressional deliberation entirely. The ordinary law discourse and the constitutional law discourse are largely separate conversations.

And yet the relationship between constitutional and ordinary law can be dynamic and interactive, notwithstanding the fact that they have a strictly hierarchical relationship as a formal matter. Indeed, the common law tail can even sometimes wag the constitutional law dog. When ordinary law fails to meet potent demands for the protection of individuals’ interests, pressure can build (p.155) on courts to turn to the Constitution, even when the threats to those interests come from other individuals. In this way, and in an inversion of the official hierarchy in which constitutional law reigns supreme, ordinary law can shape what constitutional law becomes. So, for instance, in New York Times v. Sullivan, the deficiencies of the law of defamation draw the First Amendment into tort law. So functionally, constitutional law can have a supplementary relationship to ordinary law, even if constitutional doctrine has no way to account for this. The Constitution is pressed into this role of private law gap-filler only on occasion, notably to bring regional outliers back into line with national norms.

The gap between doctrine and reality is the cost of preserving the Court’s two postwar constitutional settlements. Since the late nineteenth century, the Supreme Court has articulated a narrative for itself that fused constitutional supremacy and judicial supremacy. In the mid-twentieth century, it added to this narrative a role for itself as defender of civil liberties. Maintaining its role in the driver’s seat required that the Court exercise tight control over how constitutional rights could be deployed, even as the Court itself sometimes applied rights horizontally in furtherance of its agenda. Whatever else can be said about the state action doctrine, over the long haul, it has performed as an effective constitutional containment device. (p.156)


(1.) For details on the so-called Force Bill, see Gillette (1980, 280–99).

(2.) See, for instance, 92 U.S. at 554 (“The fourteenth amendment prohibits a State from depriving any person of life, liberty, or property, without due process of law; but this adds nothing to the rights of one citizen as against another. It simply furnishes an additional guaranty against any encroachment by the States upon the fundamental rights which belong to every citizen as a member of society.”).

(3.) Civil Rights Cases, 109 U.S. 3, 11 (1883).

(4.) Id.

(5.) Id. at 13.

(6.) Id.

(7.) Id. at 14.

(8.) The Court went on to hold that the Thirteenth Amendment did not authorize the Civil Rights Act either.

(9.) See Fairman (2009, 568–85) (describing contemporaneous media responses to the decision).

(10.) Justice Bradley’s eagerness to turn the page on Reconstruction is almost palpable in the Civil Rights Cases. He writes: “When a man has emerged from slavery, and by the aid of beneficent legislation has shaken off the inseparable concomitants of that state, there must be some stage in the progress of his elevation when he takes the rank of a mere citizen, and ceases to be the special favorite of the laws, and when his rights as a citizen, or a man, are to be protected in the ordinary modes by which other men’s rights are protected.” 109 U.S. at 25.

(11.) 163 U.S. 537 (1896).

(12.) See generally Rosenberg (2008).

(13.) Intimidation already reduced the number of claims brought under the Civil Rights Act to a handful, rendering it a “dead letter” even before the Supreme Court issued its opinion (Foner 2002, 552). The Supreme Court’s decision spared the national government the embarrassment of having the law on the books any more.

(14.) These questions could end up decided by a state court depending on the manner in which they were raised. As the Supreme Court first held in Metcalf v. City of Watertown, 128 U.S. 586 (1888), a case only “arises under” federal law for the purposes of the federal subject matter jurisdiction statute if federal law is a necessary element of the plaintiff’s case. So if, for instance, a defendant in a trespass action asserted that his ejection from an inn on account of his race was unconstitutional, he would have to litigate the issue in state court, since the federal (constitutional) issue only arises as part of the defense. The Supreme Court would be the only federal court that could exercise jurisdiction, and it could only do so after the state appeals process had run its course.

(15.) Dred Scott v. Sandford, 60 U.S. 393 (1857).

(16.) 76 U.S. 41 (1870).

(17.) 75 U.S. 603 (1870), overruled by the Legal Tender Cases, 79 U.S. 457 (1871).

(18.) Stanley Kutler’s summation of these developments is worth quoting at some length:

Perhaps the greatest relevance of the judicial review cases is as an adumbration of future developments. The role and record of the Supreme Court as a legislative censor in the late nineteenth century are clearly established. The activities of the post–Civil War decade, however, might be seen as a foundation for the Court’s later behavior and character. The regular pattern of asserting the propriety of judicial review, the willingness to examine legislative purpose, and the perfection of vested rights characterized the Chase Court’s decisions almost as much as those of later years. While the history of judicial review properly goes back to Marbury v. Madison, and beyond, the Court’s performance from 1865 to 1873 marked a significant deviation. The sheer volume alone is different. But more important was the political and public acceptance of this judicial function as a standard for measuring the legitimacy of power (Kutler 1968, 125).

(19.) 87 U.S. 655, 662, 663 (1874).

(20.) 96 U.S. 97, 105 (1878).

(21.) See, e.g., Hodges v. United States, 203 U.S. 1 (1906).

(22.) In the 1920s, Robert Hale voiced criticisms of the state action doctrine that would find a wider hearing only decades later (1923).

(23.) As Nichols notes, Eisenhower may have made his most lasting contribution to civil rights through his appointments to the federal bench, including the Supreme Court.

(24.) The Supreme Court’s epic 1938 decision Erie Railroad Co. v. Tompkins, 304 U.S. 64, had a role to play in this as well. As discussed above, under Swift v. Tyson, federal courts sitting in diversity jurisdiction (i.e., hearing nonfederal claims between citizens of different states) would resolve those disputes with reference to federal common law: that is, federal courts would decide for themselves the appropriate common law rule of decision in the case. In Erie, the Supreme Court overruled Swift, rejecting the idea that there is any such thing as general federal common law, with the result that diversity cases would have to be decided with reference to the state law rules.

(25.) Richard Primus refers to it as the “third transformation” in the history of American rights, which played out between 1948 and 1969. This third transformation was in part a “struggle over the meaning of the first two”—namely, the Bill of Rights and Reconstruction (Primus 1999, 70).

(26.) 358 U.S. 1.

(27.) Id. at 18.

(28.) Id. at 19–20.

(29.) 273 U.S. 536. Oliver Wendell Holmes wrote for the Court. The Court did not reach the question whether the Fifteenth Amendment applied to primaries.

(30.) 286 U.S. 73 (1932).

(31.) 295 U.S. 45 (1935).

(32.) 321 U.S. 649 (1944).

(33.) Id. at 663 (emphasis added).

(34.) 313 U.S. 299 (1941).

(35.) 345 U.S. 461, 463 (1953).

(36.) Id. at 468.

(37.) Id. at 469.

(38.) The trial court concluded that the plaintiffs had waived their claim for damages, and the plaintiffs did not appeal on this point. Terry v. Adams, 90 F. Supp. 595 (S.D. Tex. 1950).

(39.) The Democratic Clubs of South Carolina, for one, had functioned much like the Jaybird club. See Baskin v. Brown, 174 F.2d 391 (4th Cir. 1949).

(40.) 245 U.S. 60 (1917). Incidentally, Buchanan was the first major victory of the NAACP.

(41.) 271 U.S. 323 (1926). Corrigan concerned the enforcement of a restrictive covenant in Washington, D.C., which was considered exempt from the Fourteenth Amendment’s commands to the states.

(42.) Kraemer v. Shelley, 355 Mo. 814 (1947); Sipes v. McGhee, 316 Mich. 614 (1947). In Kraemer, the plaintiffs had sought an injunction to block a pending sale of a property to an African American family through a white intermediary. In Sipes, the plaintiffs sought to have the covenant enforced against an African American family that had already purchased and begun occupying the property in question.

(43.) Justices Reed, Jackson, and Rutledge recused themselves, presumably because they owned property subject to racially restrictive covenants.

(44.) Shelley v. Kraemer, 334 U.S. 1, 13 (1948).

(45.) Id. at 14.

(46.) 383 U.S. 301 (1966).

(47.) Id. at 326 (citing McCulloch v. Maryland, 17 U.S. 316, 421 (1819)).

(48.) 384 U.S. 641 (1966).

(49.) Lassiter v. Northampton County Board of Elections, 360 U.S. 45 (1959).

(50.) The Fourteenth Amendment was implicated and not the Fifteenth because literacy tests did not implicate a denial of the right to vote on account of race.

(51.) 384 U.S. at 654.

(52.) Id. at 653.

(53.) Id. at 668 (Harlan, J., dissenting).

(54.) Id. at 667 (Harlan, J., dissenting).

(55.) 392 U.S. 409 (1968).

(56.) 334 U.S. 24 (1948).

(57.) Hurd relied on § 1982 and not the Equal Protection Clause because the case arose in Washington, D.C., and the Fourteenth Amendment did not apply to Washington. The Court would later devise a workaround to this anomaly in Bolling v. Sharpe, 347 U.S. 497 (1954).

(58.) 392 U.S. at 438.

(59.) Id. at 439 (quoting Civil Rights Cases, 109 U.S. 3, 20 (1883)) (emphasis added by Jones).

(60.) Id. at 440.

(61.) Id. at 441 (quoting Civil Rights Cases, 109 U.S. at 22).

(62.) Id. at 442–43.

(63.) Heart of Atlanta Motel Inc. v. United States, 379 U.S. 241 (1964).

(64.) 365 U.S. 715 (1961).

(65.) 382 U.S. 296 (1966).

(66.) 387 U.S. 369 (1967).

(67.) Id. at 378–79.

(68.) 378 U.S. 226 (1964).

(70.) See Moose Lodge No. 107 v. Irvis, 407 U.S. 163 (1972) (holding that state’s issuance of a liquor license did not render racial discrimination by private fraternal organization subject to constitutional scrutiny).

(71.) Largely, but not completely. For instance, in 1967 a closely divided Supreme Court overturned a damages award against Life magazine, in an invasion of privacy case brought by members of a family whose experience as crime victims was presented, in exaggerated form, in the pages of the magazine. Time Inc. v. Hill, 385 U.S. 374 (1967). See also Cox Broad. Corp. v. Cohn, 420 U.S. 469 (1975) (barring recovery for invasion of privacy when broadcaster obtained name of deceased rape victim from public court records).

There are other contexts in which the Supreme Court confronted state action issues and speech claims. Notably, in Marsh v. Alabama, 326 U.S. 501 (1946), the Supreme Court held that the restriction of religious speech in a company-owned town violated the constitutional right to free speech. Marsh was significant in establishing a “public function” exception to the state action rule. But the case was somewhat idiosyncratic, since the facts the Court confronted were rather extreme: a company that restricted door-to-door solicitations owned an entire town.

(72.) For the classic articulation of the civil libertarian free speech position, see Chafee (1920).

(73.) W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

(74.) Yates v. United States, 354 U.S. 298 (1957).

(75.) Speiser v. Randall, 357 U.S. 513 (1958).

(76.) 376 U.S. 254.

(77.) King was prosecuted by the State of Alabama for felony tax evasion; this was the first felony tax evasion prosecution in the state’s history. King was acquitted.

(78.) Sullivan claimed that the ministers, whose names appeared in the advertisement, had endorsed its message, although they offered uncontroverted evidence that they had not known about or approved of the ad. The inclusion of the ministers did serve a more practical end: their presence in the suit defeated removal of the matter from state court to federal court. In the absence of a federal issue, cases can be removed when the plaintiffs and defendants are citizens of different states. But a case can be removed only when all of the plaintiffs are from different states than all of the defendants, and the ministers were citizens of Alabama.

(79.) Quoted in Emerson (1970, 519).

(80.) The Alabama governor, John Patterson, added Martin Luther King, Jr. as a defendant as well.

(81.) The paper, represented by eminent law professor Herbert Wechsler, also argued that the Alabama court’s assertion of personal jurisdiction over the paper was unconstitutional.

(82.) 376 U.S. at 268.

(83.) Id. at 269.

(84.) Id. at 278.

(85.) Id. at 229.

(86.) Id. at 265.

(87.) The Court did, perhaps, overstate the extent to which “the state rule of law” it identified really was the established law of Alabama. The Supreme Court identified several features of Alabama’s rule of liability. To wit: publications are libelous per se if they impute misconduct to officials in their office; when the plaintiff is a public official, criticism of his agency is enough to find he suffered reputational harm; once libel per se is established, a defendant’s only defense is that the statements were true in every respect; general damages may be presumed. In fact, pre-Sullivan libel precedents in Alabama were not so draconian as the Supreme Court’s précis suggests. But in justifying its ruling, the Alabama Supreme Court had itself insisted that the judgment against the Times was amply supported by precedents. And if the Alabama Supreme Court was going to oversell its own precedents, the U.S. Supreme Court was only too happy to go along, since doing so bolstered its claim that the constitutional problem was firmly fixed in “the law” of Alabama.

(88.) Curtis Pub. Co. v. Butts, 388 U.S. 130 (1967).

(89.) Hustler Magazine, Inc. v. Falwell, 485 U.S. 46 (1988).

(90.) Gertz v. Robert Welch, Inc., 418 U.S. 323 (1974).

(91.) The dispute was resolved by settlement.

(92.) A rare exception to the Court’s forbearance is Troxel v. Granville, 530 U.S. 57 (2000), in which the Court invalidated, as applied, a state statute permitting courts to grant visitation rights to any person at any time when visitation serves the child’s best interests. A court order under the statute had granted visitation rights to grandparents against the mother’s wishes, and the Court held that this violated parental rights, protected by the Fourteenth Amendment, to make certain child-rearing choices for their children. The Court’s decision is perhaps best understood as a one-off effort to correct what the court perceived as a serious abuse of discretion on the part of the trial judge, when the Court lacked jurisdiction to review the exercise of discretion as such (Ross et al. 2014, 614).

(93.) “ ‘Is there any point to which you would wish to draw my attention?’

‘To the curious incident of the dog in the night-time.’

‘The dog did nothing in the night-time.’

‘That was the curious incident,’ remarked Sherlock Holmes.” Doyle (1892, 196–97), cited in Chisom v. Roemer, 501 U.S. 380, 396 n.23.

(94.) 381 U.S. 479 (1965).

(95.) 410 U.S. 113 (1973).

(96.) 539 U.S. 558 (2003).

(97.) See, e.g., Moore v. City of East Cleveland, 431 U.S. 494 (1977) (invalidating city ordinance limiting occupancy of any dwelling to members of a family, narrowly defined).

(98.) 489 U.S. 189 (1989).

(99.) Id. at 195–96.

(100.) Id. at 203.

(101.) There is, of course, a hugely important constitutional family law issue that has engaged the Court: the constitutionality of same-sex marriage restrictions. But since the question concerns governmental definitions of marriage, the same-sex marriage cases do not implicate the state action rule. Matters are different in Canada, as Chapter Seven explains.

(102.) One context in which the Supreme Court has applied rights horizontally in recent years is in reviewing punitive damages awards. The Court has held that excessive punitive damage awards amount to a violation of due process and struck down the awards recovered by private parties. See State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408 (2003); BMW of N. Am., Inc. v. Gore, 517 U.S. 559 (1996). This line of cases has been viewed by some popular and academic observers as part of the broader efforts by the Rehnquist and Roberts Courts to create a more favorable legal climate for business (Rosen 2008).

(103.) These claims are also related to a broader body of work on popular constitutionalism, which denies the starring role in American constitutional development to courts, in favor of democratic institutions. Leading voices in this direction include Balkin (2011), Kramer (2004), Tushnet (2000), and West (2003).

(104.) They are entrenched not least because they have been institutionalized in agencies dedicated to maintaining the regimes that they establish.

(105.) See also Lupu (1993) (arguing that certain statutes “revolve in constitutional law orbits” and should be interpreted accordingly).

(106.) Shelby County was a Fifteenth Amendment case; the Court had already revisited its Fourteenth Amendment enforcement jurisprudence in the 1997 case of City of Boerne v. Flores. In City of Boerne, the Supreme Court declared that there must be “a congruence and proportionality” between measures adopted by Congress pursuant to its Fourteenth Amendment rights enforcement power and the rights themselves—as defined by the Court. 521 U.S. 507, 520 (1997).