December 01, 2015 Urban Lawyer

United States Supreme Court Justices' Voting in Systemic Racial Discrimination Cases in Education

by Lewis M. Wasserman & John P. Connolly
In the Classroom

In the Classroom

Lewis M. Wasserman is an Associate Professor of Educational Leadership and Policy Studies at the University of Texas at Arlington. He holds a Ph.D. in psychology from Hofstra University, Hempstead, New York, and a J.D. from St. John’s University School of Law, Jamaica, New York. He engaged in the private practice of law for 23 years in New York, specializing in public education and civil rights litigation, before entering the academy. He also teaches Education Law at SMU Dedman College of Law in Dallas. His most recent article, Overcoming Obstacles to Religious Exercise in K-12 Education, was published in Notre Dame Law School’s Journal of Legislation
in 2014.

John P. Connolly is a political scientist and statistician with the Office of Information Technology at the University of Texas at Arlington. He earned his Ph.D. in political science from the University of Texas at Dallas in 2013. He also holds a Master of Liberal Studies from Southern Methodist University, a Master of Business Administration from University of Dallas, and a Bachelors of Engineering from the University of Strathclyde in Scotland. His essay on political ideology and issue constraint is scheduled for publication in the International Encyclopedia of the Social and Behavioral Sciences, in March 2015.

The authors gratefully acknowledge the special contribution of Pamela L. Steen, Esq., Member of the New York Bar, for her helpful commentary and criticism during the production of the manuscript.

On May 17, 1954, United States Supreme Court Chief Justice Earl Warren delivered a unanimous ruling in the landmark civil liberty case Brown v. Board Of Education.1 The Court held that de jure segregation of public schools was a violation of the Equal Protection Clause of the Fourteenth Amendment and was therefore unconstitutional.2 In Brown, the Chief Justice worked hard to obtain unanimity,3 fearing that a divided decision would produce backlash.4 Despite his success in bringing about a unified Court, his expectations about minimizing resistance were misplaced.5

A substantial number of scholars assert that in the post-Brown period the Supreme Court has largely gutted the Equal Protection Clause, leaving Brown’s promise unfulfilled.6 This asserted trend corresponds more or less in time to the diminution of liberal perfectionist influences on the Court7 — Justices who believed that progressive ideals, such as racial integration, should guide constitutional interpretation — represented by the departures of Chief Justice Warren and Justices William Brennan, Thurgood Marshall, and William O. Douglas,8 and the elevation of “minimalist”9 and “fundamentalist”10 Justices to the Court, commencing during President Nixon’s administration.11

In this article, we trace the course of liberal perfectionist jurisprudence at the Supreme Court in the post-Brown period, explaining the ground for its decline, and account for this trend using empirical methodology derived largely from political science. We argue that those who assert Brown’s promise has been unfulfilled have relied on a perfectionist interpretation of the Constitution that gives more weight to Brown’s aspirational statements and moral conclusions than the decision’s specific holding. This specific interpretation resulted in utopian expectations about what Brown promised, leading ultimately to disappointment for those adhering to this view.

In Part I, after examining Brown, we follow its arc to its peak enforcement period, in or about 1973. In Part II, titled Conservative Retrenchment, we show how liberal perfectionist aspirations in racial matters were thwarted from 1974 to present. During this period, we see at the Supreme Court (1) an acceptance of systemic unequal funding of educational opportunity, (2) establishment of stringent criteria to prove unlawful race discrimination;12 and (3) a limitation on the permissible scope of remedial orders for purposeful-systemic racial segregation.

In Parts III and IV, we analyze case developments at the Court regarding benign race-based discrimination as a tool for voluntary K-12 desegregation and college and university admissions. In each area we see reluctance at the Court to support such measures.13 Moreover, we observe in the states a counter-movement to defeat benign discrimination efforts, mostly through the process of amending state constitutions.14

In Part V, we briefly examine the nomination and confirmation process during the Nixon and Reagan administrations, paying particular attention to the goals each president had respecting his choices and the factors that facilitated or impeded those selections.

In Part VI, we examine empirically the Justices’ voting in systemic racial discrimination cases in the post-Brown period using the attitudinal model of judicial voting developed in political science literature. In furtherance of this objective we studied the 552 votes rendered by the Supreme Court Justices for the period under review. Using statistical techniques that account for correlated voting patterns arising from repeated observations of the same Justice, we examine the effect of Justices’ ideologies, their appointment dates, and legal precedent, on their liberal or conservative voting.15

Among our findings are large effects associated with Justices appointed during the Nixon and Reagan appointment eras.16 We also find a precedent effect associated with the seminal 1974 case of Milliken v. Bradley,17 which narrowed the permissible scope of remedial measures for purposeful segregation in public schools.18 In Part VII, we synthesize the foregoing sections and comment on the prospects for changes in the Court following the 2014 midterm elections and the 2016 race for the White House.

Part I.  Brown and the Era of Stringent Enforcement

A.  Brown v. Board of Education

During the 1952 Term the Supreme Court granted review in five consolidated cases that challenged the “separate but equal” doctrine in the context of elementary and secondary education.19 At that time, seventeen states had mandatory public school segregation,20 four had permissive statutes,21 and South Carolina and Georgia had announced they would abolish public schools if segregation were banned.22 Moreover, the District of Columbia was practicing segregation in its public schools.23

The school-system defendants in Brown furnished inferior public education to black citizens in virtually every relevant respect. For example, in the challenge to the South Carolina system,24 it was alleged that white schools had one teacher for every twenty-eight pupils, while black schools had one teacher for every forty-seven students; white schools were brick and stucco, while black schools were made of wood; white schools had indoor plumbing, while black schools had outhouses.25

The Justices then sitting could not agree on a decision and set the cases for reargument during the following term.26 The Justices asked the parties to brief several questions for that argument, focusing on the intent of the framers of the Fourteenth Amendment during the 1868 ratification period.27 The cases were argued on October 13, 1953, and, through the concerted efforts of Chief Justice Warren (who replaced Chief Justice Vinson following Vinson’s heart attack during the summer preceding the 1953 term), all of the Justices were persuaded to join in a unanimous opinion,28 holding that “separate but equal” was impermissible in the context of public education.29

On May 17, 1954, the Supreme Court issued its decision in Brown.30 The opinion authored by Chief Justice Warren notably began by explaining that the constitutionality of segregation could not be resolved on the ground of original intent.31 Warren explained that historical sources were inconclusive and that the enormous changes in the nature of education in the years between 1868 and 1954 made history of little use in resolving the question.32 In reaching its conclusion, the Court did not focus on the inequalities between the black and white schools in the cases before it; instead, it assumed that the disparities between

the Negro and white schools involved have been equalized, or are being equalized, with respect to buildings, curricula, qualifications and salaries of teachers, and other ‘tangible’ factors. Our decision therefore, cannot turn on merely a comparison of those tangible factors in the Negro and white schools involved in each of the cases. We must look instead to the effect of segregation itself on public education.33

Based on an assumption of factual equality,34 the Court stated the issue as: “Does segregation of children in public schools solely on the basis of race, even though the physical facilities and other ‘tangible’ factors may be equal, deprive children of the minority group of equal educational opportunities?”35

In answering this question the Court stated unambiguously that state mandated segregation inherently stamps black children as inferior and impairs their educational opportunities.36 The Court concluded: “To separate them from others of similar age and qualifications solely because of their race generates a feeling of inferiority as to their status in the community that may affect their hearts and minds in a way unlikely ever to be undone.”37 The Court supported this conclusion by relying on social science literature, which purported to show that segregation causes black children to feel inferior and interferes with their learning.38 Finally, the Court asserted “that in the field of public education the doctrine of ‘separate but equal’ has no place. Separate educational facilities are inherently unequal.”39 Despite its strong words the Court did not prescribe a remedy, but asked for the argument on that issue later that term.40 The following term, in Brown II,41 the Court remanded the cases to the lower courts to use traditional equity principles to fashion remedies “to admit to public schools on a racially nondiscriminatory basis with all deliberate speed the parties to these cases.”42

Even among its supporters, Brown I drew criticism, some scholars arguing that the use of social science studies to support its conclusion — especially the research it used — weakened the force of the decision and opened the Court to broad-based criticism as to its methodology. Professor Mark G. Yudof showed, for example, that “[v]irtually everyone who has examined the question now agrees that the Court erred” in relying upon the social science data it employed,43 and that the Court created confusion between what the law is and what the law should be.44 Yudof sees a justifiable skepticism about social sciences in the law in the post-Brown era, especially as to the use of social science as a tool in reaching factual conclusions in complex policy areas.45 Nevertheless, he supports the decision. Yudof agrees with others that Brown’s “antidiscrimination principle, embod[ies] the ethical assertion that race-dependent government decisions should be disfavored, [and] is a principle of wide moral appeal that properly informs the interpretation and application of the Equal Protection Clause of [the Fourteenth Amendment].”46

In the five cases the Brown Court reviewed, each case involved purposeful government sanctioned separation of the races.47 Ordinary methods of judicial construction should help us determine what Brown meant and whether ensuing decisions at the Court are consistent with Brown’s holding. In performing this function we are required to determine (1) the facts of the case, and (2) the Court’s pronouncements of its decision, along with any justifications for the decision, where such justifications relate to those facts.48

Accordingly, the following principles may be derived from the Court’s opinion: (1) The meaning of the Equal Protection Clause cannot be resolved on the ground of original intent;49 (2) equality of tangible services does not satisfy equal protection requirements in an environment infected by de jure segregation; (3) separating school children from others of similar age and qualifications solely because of their race is unconstitutional; (4) the “separate but equal” doctrine has no place in our constitutional order; and (5) intentional segregation is harmful to the hearts and minds of its victims.50 In the cases that follow in Parts II-IV, we do not find any inconsistencies with this strict reading of the Brown ruling.

Notably, Brown did not address how factors such as voluntary or exclusionary residential housing patterns and lower school funding in districts containing substantial minority populations implicated Equal Protection concerns. Moreover, it did not answer the question of how pre-existing school boundaries and the State’s role in creating them fit into Brown’s exposition of the Equal Protection Clause, at least when they were not designed to separate the races.

Although Brown addressed the unlawfulness of de jure segregation, it did not address whether states had a constitutional obligation to insure blending of the races in K-12 settings, when segregation came about for reasons unrelated to state action. These might include, for example, choices animated by differences in wealth, social class, or even private racial animosity. Those who read Brown’s holding to include these issues go beyond its reach.

Finally, Brown did not address the question of whether the Equal Protection Clause had anything to say about assignment of students to schools, or educational programs based on qualifications not intentionally designed to eliminate blacks from participation.

B.  Stringent Enforcement: 1954 - 1973

The Court’s selection of the ambiguous phrase “all deliberate speed” in Brown II 51 must have been influenced by the strong reaction in the states over the dismantling of purposeful segregation.52 Indeed, southern states openly and aggressively resisted compliance with Brown. This defiance manifested in state legislatures adopting resolutions of court “nullification” and “interposition,” which declared that the Supreme Court’s decisions were without effect.53

Neither Brown I’s inspiring language nor Brown II’s “deliberate speed” language were helpful in providing guidance to lower courts in how to evaluate Equal Protection violations on the one hand, or how to fashion remedies once a violation was established, on the other hand. Thus, it was inevitable that these problems would be visited upon the Court for resolution. Indeed, the Court faced the daunting task of establishing standards for proving discrimination in the school context,54 determining what courts can and cannot do and under what circumstances,55 and deciding when federal desegregation remedies should be ended,56 among other things.

The Court first confronted the wide-scale resistance to the Brown decisions in Cooper v. Aaron, which arrived at the Supreme Court in 1958.57 There, the Little Rock, Arkansas school system had been ordered to desegregate during the 1957-58 school year, but the Governor had instructed the Arkansas National Guard to keep blacks out.58 The Little Rock school system then asked for a stay of the integration plan. In its opinion the Court began by declaring:

As this case reaches us it raises questions of the highest importance to the maintenance of our federal system of government. It necessarily involves a claim by the Governor and Legislature of a state that there is no duty on state officials to obey federal court orders resting on this Court’s considered interpretation of the United States Constitution.59

In rejecting the State’s claim that it did not have to comply with the Supreme Court’s decision, the Court said:

[Marbury] declared the basic principle that the federal judiciary is supreme in the exposition of the law of the Constitution, and that principle has ever since been respected by this Court and the Country as a permanent and indispensable feature of our constitutional system. It follows that the interpretation of the Fourteenth Amendment enunciated by this court in the Brown case is supreme law of the land No state legislature or executive or judicial officer can war against the Constitution without violating his undertaking to support it.60

Thus the Court strongly reaffirmed Brown and said that it could not be nullified either “openly and directly by state legislators or state executive or judicial officers”61 or “indirectly by them through evasive schemes for segregation whether attempted ‘ingeniously or ingenuously.’ ”62

Despite Cooper’s statement of the law, the efforts to avoid desegregation in the South were unrelenting. In 1963, for example, enforcing its statement of “what the law is,” in Goss v. Board of Education,63 the Supreme Court declared unconstitutional Knoxville, Tennessee’s law which allowed students who were assigned to new schools as part of desegregation to transfer from schools where they were a racial minority to ones where they would be in the racial majority.64 This rule would have allowed white students placed in a predominantly black school to transfer back to a white school and black students who were placed in a predominantly white school to do the same.65 The Court found this unconstitutional because “[i]t is readily apparent that the transfer system . . . lends itself to perpetuation of segregation.”66

Griffin v. County School Board67 illustrates both the resistance to Brown and the Court’s commitment to strong enforcement of its decrees. There, faced with an order to desegregate, a county school board refused to appropriate funds for the operation of its public schools, effectively shutting them down.68 However, it gave tax credits for contributions to private white schools.69 These private school students became eligible for county and state tuition grants one year later.70 Public schools outside the county continued to operate as they had previously.71 The Court found that it was unconstitutional for school systems to close rather than desegregate and meanwhile contribute to the support of private white schools.72 The Court stated that “[t]here has been entirely too much deliberation and not enough speed in enforcing [Brown].”73 Concluding the program violated Equal Protection strictures, the Court explained: “Whatever nonracial grounds might support a State’s allowing a county to abandon public schools, the object must be a constitutional one, and grounds of race and opposition to desegregation do not qualify as constitutional.”74

In 1968, in Green v. County School Board,75 the Court declared unconstitutional a “freedom-of-choice" plan in a district that had previously been declared unlawfully segregated.76 Such plans were a common approach used to frustrate desegregation.77 Green involved a school system of about 1300 students in rural Virginia, about half of whom were black.78 Notably, “[t]here [was] no residential segregation in the county; persons of both races reside[d] throughout.”79 The county had only two schools, each one serving the whole county.80 Under the freedom-of-choice plan, first and eighth graders had an affirmative duty to select which school they would attend.81 Students in the other grades were allowed, but not required, to select a school.82 Students who made no selection were assigned by the state board of education to the school they previously attended.83

Three years after the plan was enacted, no white student was attending the black school and only 15% of black students were attending the white school.84 The purpose of this plan was apparent, even to the most na¨ıve observer. The Court said:

[Ten] years after Brown II directed the making of a “prompt and reasonable start[,]” [adopting this freedom-of-choice plan was a] deliberate perpetuation of the unconstitutional dual system [which] can only have compounded the harm of such a system. . . . The burden on a school board today is to come forward with a plan that promises realistically to work, and promises realistically to work now.85

The Court asserted emphatically that school boards have “the affirmative duty to take whatever steps might be necessary to convert to a unitary system in which racial discrimination would be eliminated root and branch,” and held that this “‘freedom-of-choice’ plan” violated Equal Protection.86

The Green Court described its authority and duty in broad terms such that they included rendering a decree that insofar as possible eliminated the discriminatory effects of the past, and barred similar discrimination in the future.87 Green said that dual systems are ones that are “racially identifiable.”88 The goal was, in Green’s terms, “a system without a ‘white’ school and a ‘Negro’ school, but just schools.”89 The Court set out six factors that made a school racially “identifiable.” These are (1) composition of the student bodies, (2) faculty, (3) staff, (4) transportation programs, (5) extracurricular activities, and (6) physical facilities.90

The unconstitutional de jure segregation in Green was relatively easy to resolve. After all, there was no residential segregation and only two schools to contend with. Therefore, student assignment by neighborhood would give a rather straightforward remedy,91 arguably leading to a fix for the other five Green factors.92

Justice Brennan, author of this unanimous opinion, said at the outset that the issue for decision was whether the school board had complied with Brown by “determining admission to the public schools on a non- racial basis.”93 If so, it is hard to see how, on its face, the plan failed the test, since the parents, who were private actors, were given a choice of which school to attend.94 This was seemingly a racially neutral choice, which all parents, black and white, were entitled to make.95 Green’s remedial order must then mean something different.

A more likely explanation of the integration remedy in Green was that it was necessary to eliminate the effects of past practices already proven. Indeed, since Green held that the school system remained a dual system, the decree was necessary to obtain compliance with the Court’s previous order.96 In this sense, then, integration is a remedy tethered to the original constitutional violation, not an independent constitutional right.97

Three years later, in Swann v. Charlotte-Mecklenburg Board of Education,98 the Court reiterated its goal of “eliminat[ing] from the public schools all vestiges of [S]tate-imposed segregation,”99 when it unanimously upheld a lower court’s broad remedial order after school authorities defaulted in their obligation to give acceptable remedies for prior violations.100

In Swann the students remained largely segregated in 1969, despite a 1965 desegregation plan based on geographic zoning with a free-transfer provision.101 After the school board failed to produce an adequate plan, the district court imposed one, which grouped several outlying elementary schools with each black inner-city school and required extensive busing.102 The plan also required that as many schools as practicable reflect the white to black ratio that existed in the district as a whole.103 The new plan was challenged as too burdensome.104

In the process of imposing its broad remedial order the Court found it necessary to “defin[e] in more precise terms [than it had before] the scope of the duty of school authorities and district courts in implementing” the decision in Brown.105

Swann, building on Green, clarified that when school authorities fail to devise effective remedies for State-imposed segregation, the district courts may constitutionally (1) Order the school districts to assign teachers so as to achieve faculty desegregation; (2) forbid patterns of school construction and abandonment that serve to perpetuate or reestablish a dual system; (3) impose racial quotas, when used not as inflexible requirements, but as a starting point for the shaping of a desegregation plan;106 (5) alter school attendance zones, including the grouping and pairing of noncontiguous zones; and (6) require busing to a school not closest to the students’ homes, in order to achieve desegregation, except when the travel time is excessive.107

Notably, the Swann Court explained a limit on its remedial powers. It said:

Absent a constitutional violation there would be no basis for judicially ordering assignment of students on a racial basis. All things being equal, with no history of discrimination, it might well be desirable to assign pupils to schools nearest their homes. But all things are not equal in a system that has been deliberately constructed and maintained to enforce racial segregation. The remedy for such segregation may be administratively awkward, inconvenient, and even bizarre in some situations and may impose burdens on some; but all awkwardness and inconvenience cannot be avoided in the interim period when remedial adjustments are being made to eliminate the dual school systems.108

Finally, the Swann Court imposed further constraints on its remedial authority. It said:

It does not follow that the communities served by such systems will remain demographically stable, for in a growing, mobile society, few will do so. Neither school authorities nor district courts are constitutionally required to make year-by-year adjustments of the racial composition of student bodies once the affirmative duty to desegregate has been accomplished and racial discrimination through official action is eliminated from the system. This does not mean that federal courts are without power to deal with future problems; but in the absence of a showing that either the school authorities or some other agency of the State has deliberately attempted to fix or alter demographic patterns to affect the racial composition of the schools, further intervention by a district court should not be necessary.109

While the North Carolina State Board of Education in Swann was submitting racial unity plans to the district court on the remand from the Supreme Court, the North Carolina legislature was enacting the Anti-Busing Law, which forbade assignment of any student on account of race or for the purpose of creating a racial balance in the schools.110 When Swann returned to the Supreme Court, that Court held that the State’s absolute prohibition against transportation of students assigned on the basis of race or for the purpose of creating a racial balance or ratio was an unconstitutional impairment of desegregation remedies.111 The Court said that state policy must give way when it hinders or obstructs the operation of a unitary school.112 Indeed, the Court concluded that the policy adopted by the state, when considered “against the background of segregation,” would have rendered illusory the intent of Brown.113

The Swann Court asserted that in cases where purposeful discrimination was established, its first duty was to eliminate “invidious racial distinctions”114 and eliminate the vestiges of the former de jure system. And it would look to the Green factors in determining how to go about it.115 Moreover, Swann states that where de jure segregation was previously established, a presumption operates that the segregated pat- terns are attributable to those unlawful acts.116 The Swann Court said:

No per se rule can adequately embrace all the difficulties of reconciling the competing interests involved; but in a system with a history of segregation the need for remedial criteria of sufficient specificity to assure a school authority’s compliance with its constitutional duty warrants a presumption against schools that are substantially disproportionate in their racial composition. Where the school authority’s proposed plan for conversion from a dual to a unitary system contemplates the continued existence of some schools that are all or predominately of one race, they have the burden of showing that such school assignments are genuinely nondiscriminatory. The court should scrutinize such schools, and the burden upon the school authorities will be to satisfy the court that their racial composition is not the result of present or past discriminatory action on their part.117

Unfortunately, the Swann Court did not elaborate much on how a school district can overcome the presumption against it, or what criteria might apply in linking past discriminatory effects to present conditions. Its reference in the quoted sections above to highly mobile societies118 and its eschewing year-to-year adjustments strongly suggests that the Court wanted to avoid perpetual oversight of school operations, even where past purposeful discrimination was proven.119

In Keyes v. School District No. 1,120 the parents of Denver school children sought desegregation of the Park Hill area schools, which comprised only part of the Denver city school system.121 The parents argued that the school board, using various techniques — including gerrymandering attendance zones, manipulating school site locations, and neighborhood school assignment policy — created or maintained racially and ethnically segregated schools.122 Upon securing an order of the district court directing relief, the parents sought to expand their suit to secure desegregation of the remaining schools of the Denver school district, particularly those in the core city area.

The district court denied the further relief, holding that the deliberate racial segregation of the Park Hill schools did not prove that a similar violation was occurring in connection with the core city schools.123 The district court nevertheless found that the segregated core city schools were educationally inferior to “white” schools elsewhere in the district and, relying on Plessy v. Ferguson,124 ordered the respondents to give substantially equal facilities to those schools.125 This latter relief was reversed by the Tenth Circuit Court of Appeals, which affirmed the Park Hill ruling and agreed that Park Hill segregation, even though deliberate, proved nothing regarding an overall policy of segregation.126

In a seven-to-one ruling written by Justice Brennan, the Supreme Court held that proof that school authorities have pursued an intentional segregative policy in a “substantial portion” of the school district will support a finding by the trial court of the existence of a dual system, absent a showing that the district is divided into clearly unrelated units.127 Pursuant to this determination, the Court directed that on remand the district court should decide initially whether the School Board’s deliberately segregative policy respecting the Park Hills schools caused the whole Denver school district to be a dual one.128 Moreover, the Court held that where a policy of intentional segregation has been proved with respect to a significant portion of the school system, the burden is on the school authorities (regardless of claims that their “neighborhood school policy” was racially neutral) to prove that their actions related to other segregated schools in the system were not likewise motivated by a segregative intent.129 In explaining the ruling Justice Brennan stated:

This is not a case, however, where a statutory dual system has ever existed. Nevertheless, where plaintiffs prove that the school authorities have carried out a systematic program of segregation affecting a substantial portion of the students, schools, teachers, and facilities within the school system, it is only common sense to conclude that there exists a predicate for a finding of the existence of a dual school system. Several considerations support this conclusion. First, it is obvious that a practice of concentrating Negroes in certain schools by structuring attendance zones or designating ‘feeder’ schools on the basis of race has the reciprocal effect of keeping other nearby schools predominantly white. Similarly, the practice of building a school — such as the Barrett Elementary School in this case — to a certain size and in a certain location, ‘with conscious knowledge that it would be a segregated school,’ has a substantial reciprocal effect on the racial composition of other nearby schools. So also, the use of mobile classrooms, the drafting of student transfer policies, the transportation of students, and the assignment of faculty and staff, on racially identifiable bases, have the clear effect of earmarking schools according to their racial composition, and this, in turn, together with the elements of student assignment and school construction, may have a profound reciprocal effect on the racial composition of residential neighborhoods within a metropolitan area, thereby causing further racial concentration within the schools.130

Thus, Keyes held that “a finding of intentionally segregative school board actions in a meaningful portion of a school system . . . creates a presumption that other segregated schooling within the system is not adventitious” and creates “a prima facie case of unlawful segregative design.” This shifts the burden to the school system to prove “that other segregated schools within the system are not also the result of intentionally segregative actions.”131

Justice Brennan further explained that in discharging the burden of showing that segregated schooling (at the other locations) “is not the result of intentionally segregative acts,” school authorities may not rely on some “allegedly logical, racially neutral explanation for their actions[,]” but must adduce enough proof “to support a finding that segregative intent was not among the factors that motivated their actions.”132 Where the school district fails to meet its burden, school authorities have the affirmative duty to effectuate transition of the entire school district to a racially nondiscriminatory system.133

Requiring the Denver school system as a whole to overcome a presumption of wrongdoing when it behaved unlawfully in one of part its system, might be considered a drastic remedy to some people residing in other parts of the system who did not create the injury inflicted. Notable in this respect was Justice Rehnquist’s sole dissent in Keyes. He stated:

Underlying the Court’s entire opinion is its apparent thesis that a district judge is at least permitted to find that if a single attendance zone between two individual schools in the large metropolitan district is found by him to have been ‘gerrymandered,’ the school district is guilty of operating a ‘dual’ school system, and is apparently a candidate for what is in practice, a federal receivership. Not only the language of the Court in the opinion, but its reliance on the case of Green v. County School Board indicates that such would be the case. It would therefore presumably be open to the District Court to require, inter alia, that pupils be transported great distances throughout the district to and from schools whose attendance zones have not been gerrymandered. Yet, unless the Equal Protection Clause of the Fourteenth Amendment now be held to embody a principle of ‘taint,’ found in some primitive legal systems but discarded centuries ago in ours, such a result can only be described as the product of judicial fiat.134

Finally, in 1973, in Norwood v. Harrison,135 the Court held the Equal Protection Clause prohibited a state from subsidizing private schools that engaged in race discrimination, in order to circumvent Brown.136 It held that the State may not, at least in the absence of a unitary system, grant tangible, specific financial aid — for example, free books or tuition grants — to private segregated schools.137 The Court stated, “[t]he existence of a permissible purpose cannot sustain an action that has an impermissible effect.”138

Norwood is notable because it (1) eschewed, on Equal Protection grounds, the use of statutes for which no discriminatory purpose could be established, and (2) encouraged lower courts to look askance at any activity which appeared to manifest a discriminatory purpose in school systems where courts had determined previously that invidious discrimination was practiced.139

Norwood thus created an interventionist role for the courts wholly different than one based on purposefully discriminatory state action. Decisions such as Norwood, and Keyes before it, for example, suggest that segregation is inherently harmful and therefore unconstitutional. If this is what these cases imply, they vary from the authors’ understanding of what Brown held.140 To the best of our knowledge, no decision has gone so far as to directly equate de facto segregation with an Equal Protection violation, even in the rights-driven enforcement period between 1954 and 1973.

The era of perfectionism was not destined to last long, however, in light of the significant personnel changes that were on the Court’s horizon. These changes began to occur following President Nixon’s election in 1968 and his nominations for the Court. In 1969, Chief Justice Warren Burger replaced Earl Warren; in 1970, Justice Harry Blackmun replaced Justice Abe Fortas; and in 1972, Justice Lewis Powell replaced Justice Hugo Black, and Justice William Rehnquist replaced Justice John Marshall Harlan II.141 These changes augured a conservative turn in race jurisprudence that is still unfolding. Emblematic of these changes are San Antonio Independent School District. v. Rodriguez,142 Milliken v. Bradley,143 Board of Education v. Dowell,144 Freeman v. Pitts,145 and Missouri v. Jenkins.146 Collectively these decisions sent a clear message of restraint to the lower courts when it came to desegregation orders. These and other cases are examined in the next Part.

Part II. Conservative Retrenchment: 1974-2014

A.  Systemic Unequal Funding of Educational Opportunity

The Court in Rodriguez147 placed at least one nail in perfectionism’s coffin.148 There, the Court rejected the claim that Texas’ public education finance system violated the Fourteenth Amendment’s Equal Protection Clause by failing to distribute funding equally among its school districts.149 In Rodriguez, the school districts received their funding from both a State program designed to establish a minimally adequate education in every school and supplemental revenue based on local property taxes levied by school districts.150 The San Antonio Independent School District, acting on behalf of students whose families resided in the district, challenged this funding scheme by arguing that it disadvantaged such students educationally because their schools lacked the vast property tax base that other Texas districts had available to them.151 The plaintiffs asserted that reliance on assessable property caused severe interdistrict disparities in per-pupil expenditure.152

In its five-to-four decision, the Court refused to examine the Texas funding scheme with strict scrutiny since there is no fundamental right to education in the Constitution and the funding system did not systematically discriminate against all poor people in Texas.153 The Court said that given the similarities between the Texas system and those in other states, the funding scheme was not so irrational as to be invidiously discriminatory within the meaning of the Equal Protection Clause.154 Justice Powell pointedly argued that on the question of wealth and education, “the Equal Protection Clause does not require absolute equality or precisely equal advantages.”155

Rodriguez effectively eliminated any cause of action for Equal Protection violations based on unequal funding.156 Giving a distinctly federalist slant to its decision,157 the Court referred educational policy decisions back to the states.158 But Rodriguez left the door slightly ajar by pointing out that the plaintiffs did not claim that Texas school children received an inadequate education relative to State requirements, only that it was unequal. The Supreme Court has yet to rule on a case where an assertion of inadequate education is made.159

Rodriguez has spawned seemingly endless claims in state courts under state constitutions, which continue to this day.160 Some scholars have linked racial composition of the district, especially in urban settings, with lack of success in achieving education finance reform in state supreme courts, contrasting this with better results for white non-urban reformers.161 Recently, education finance scholar R. Craig Wood has recommended analytic approaches linking race and educational adequacy, which he argues may lead to better outcomes in the courts for poor urban blacks making Equal Protection Clause claims.162 It is too early to know whether these recommendations will bear fruit.

B.  The Doctrine of Discriminatory Purpose and Effect

Any discussion of the Court’s treatment of systemic racial discrimination claims must include Washington v. Davis163 and Arlington Heights v. Metropolitan Housing Corp.,164 two non-education cases, both of which had a profound impact on the course of systemic racially based Equal Protection claims in public education.

Washington v. Davis, a seven-to-two decision,165 held that the procedures and written personnel test employed by the District of Columbia Police Department did not automatically constitute racial discrimination under the Equal Protection Clause merely because they produced racially disproportionate outcomes.166 The Court stated: “The central purpose of the Equal Protection Clause of the Fourteenth Amendment is the prevention of official conduct discriminating on the basis of race” and “disproportionate impact . . . alone . . . does not trig- ger the rule that racial classifications are to be subjected to the strictest scrutiny and are justifiable only by the weightiest of considerations.”167 The Court reasoned that the D.C. police department’s procedures did not have discriminatory intent and were facially neutral measures of employment qualification.168

In explaining the meaning of discriminatory intent for Equal Protection purposes, the Court stated four years after Davis that plaintiffs must prove the challenged action was taken “‘because of,’ not merely ‘in spite of,’ its adverse effects.”169 In other words,“‘[d]iscriminatory purpose’ . . . implies more than intent as volition or intent as awareness of consequences.”170 Since it was extremely unlikely that state and local legislators would include language that manifested a discriminatory purpose, this meant claims alleging discriminatory purposes would be extremely hard to establish.

Arlington Heights v. Metropolitan Housing Corp.,171 decided by the Court in 1977, illustrates the difficulty in establishing intent or purpose, relative to an Equal Protection violation. There, the Metropolitan Housing Development Corporation (MHDC) contracted with a religious order to build racially integrated low and moderate income housing in Arlington Heights.172 When MHDC applied for the necessary zoning permits, authorizing a switch from single to multiple family classification, Arlington Heights’ planning commission denied the request.173 Acting on behalf of itself and several minority members, MHDC challenged Arlington Heights’ denial as racially discriminatory.174 On appeal from an adverse district court decision, the court of appeals reversed and the Supreme Court granted Arlington Heights certiorari.175

The Supreme Court, reversing and remanding to the court of appeals for further consideration, held that MHDC failed to establish Arlington Heights’ racially discriminatory intent or purpose.176 While acknowledging that Arlington Heights’ zoning denial may result in racially disproportionate impact, the evidence in the record did not show that this was Arlington Heights’ deliberate intention.177

The Court explained that although disproportionate impact is not irrelevant to an Equal Protection Clause claim, it is not determinative in establishing invidious racial discrimination.178 Because many factors influence decision making by legislators and administrators, often involving competitive policy choices, courts generally “refrain from reviewing the merits of [such] decisions, absent a showing of arbitrariness or irrationality.”179 That said, the Court explained that racial discrimination should not be treated as just another competing factor.180 “When there is [evidence] that discriminatory purpose has been a motivating factor in the decision, judicial deference is no longer justified.”181

The Arlington Heights Court suggested how a challenger may prove purposeful discrimination. This showing may include (1) an impact which is so clearly discriminatory as to allow no other explanation than that it was adopted for impermissible purposes;182 (2) the historical events leading to the decision;183 (3) departures from the normal procedural sequence as well as substantive departures from normative practices; and (4) the legislative or administrative record, particularly when they are based on contemporary statements made by members of the body during the challenged decision making process.184 Unquestionably, this meant that, except in the most egregious cases, race-based Equal Protection pleading was effectively a dead letter.185

Finally, Arlington Heights said that where the proof indicates that a decision is motivated in part by a racially discriminatory purpose, the burden would shift to the government to prove that “the same decision would have resulted, even had the impermissible purpose not been considered.”186 The Arlington Heights Court, however, did not describe how the government might prove that it would have taken the same action anyway. This question was answered in part in Hunter v. Underwood,187 an eight-to-zero decision.

In Hunter, the Supreme Court considered an Alabama law that permanently denied the right to vote to anyone convicted of a crime involving “moral turpitude,” including misdemeanors.188 The Court held the law unconstitutional.189 Relying on the burden-shifting approach described in Arlington Heights, the Hunter Court stated: “Once racial discrimination is shown to have been a ‘substantial’ or ‘motivating’ factor behind enactment of the law, the burden shifts to the law’s defenders to demonstrate that the law would have been enacted without this factor.”190 The evidence in Hunter indicated that excluding a misdemeanant from voting had a substantial discriminatory impact against blacks and that racial discrimination was a key purpose of the legislature when the law was adopted in 1901.191 The Court found no persuasive evidence that the law would have been adopted without this motivation, and thus concluded that it was unconstitutional because “its original enactment was motivated by a desire to discriminate against blacks on account of race and the section continues to this day to have that effect.”192

The following principles emerge from this judicial thicket. First, if a law is racially neutral, a challenger must show a discriminatory purpose and a discriminatory effect to state an Equal Protection claim.193 Second, if plaintiffs provide proof that a discriminatory purpose animated the government’s action, then the government has the responsibility to show that it would have taken the same action, regardless of race or national origin.194 Third, if the court accepts the government’s justification, only rational basis review is used195 and in virtually all cases the government will prevail. Fourth, if the court is convinced that there is a discriminatory purpose motivating the enactment, but the government fails to convince the court it would have taken the action anyway, the law is treated as an unlawful race or national origin classification and will be invalidated.196 Finally, in such a case the formal application of strict scrutiny is unnecessary because persuading the court that the only purpose behind the law is discriminatory forecloses the government’s ability to show a compelling purpose for it.197

C.  Remedial Orders for Purposeful Invidious Discrimination

An important case, albeit less heralded than the seminal Milliken decision,198 is Gilmore v. City of Montgomery, Alabama.199 Gilmore was decided about thirty days before Milliken, at a time when the city was under an order to desegregate parks and its school system. Gilmore held that the city’s allocation of exclusive possession and control of its recreational facilities to white-only private schools and groups affiliated with such schools, for official athletic contests and similar functions, had the effect of creating enclaves of segregation, deprived blacks of equal access to parks and recreational facilities, and was properly enjoined by the lower courts.200

The Court reasoned that “the exclusive use and control of city recreational facilities, however temporary, by private segregated schools were little different from the city’s agreement with the YMCA to run a ‘coordinated’ but, in effect, segregated recreational program.”201 “This use,” said the Court, “carried the brand of ‘separate but equal’ and, in the circumstances of this case, was properly terminated by the [d]istrict [c]ourt.”202 Most importantly, in the eyes of the Court, allowing such exclusive use tends to perpetuate a dual school system which was constitutionally impermissible.203 Nevertheless, in a less emphasized part of the decision, Gilmore illustrates the Court’s tightening the reins on lower courts’ remedial authority in cases implicating systemic K-12 racial discrimination.

The Gilmore Court refused to enjoin the City from granting permits to segregated private school groups when that use was coupled with use by other, presumably nonsegregated organizations.204 The Court asserted that the record contained insufficient facts from which to predicate a legal judgment as to whether uses of city facilities, when made in common with other groups, in some way constituted state action which would ascribe to the city the discriminatory actions of the private groups.205 In reaching its conclusion, the Court balanced the rights of private persons to freedom of association, including racially exclusionary practices (which are not subject to Fourteenth Amendment Equal Protection constraints), and state action, which is subject to Equal Protection proscriptions.206

The city’s willingness to allow segregated use of its parks, as long as the use was not exclusive, comes dangerously close to the state action the Court condemned in its pre-Brown ruling in McLaurin v. Oklahoma State Regents, where the University of Oklahoma Law School separated a black student from whites in the classroom and other activities.207 Such rules were predicated on “separate but equal” grounds.

The racial separation in McLaurin resulted only from state action and in this way, the case dovetails well with the part of Gilmore’s ruling forbidding the grant of exclusive access to city facilities to racially segregated groups. However, so much of the order which seemed to allow granting non-exclusive access to segregated groups must have caused deep consternation to liberal perfectionist interest groups, especially since Montgomery was still under a desegregation order.208 In any event, this line drawing signaled a limitation on the types of remedies the Court would apply in jurisdictions where de jure racial segregation was previously established, foreshadowing later decisions by the Court.209

In Milliken v. Bradley,210 the record revealed that the Detroit, Michigan public schools had a long history of public and private discrimination that helped to produce residential segregation, although the city did not have a history of formal segregation ordered by it, or al- lowed by law.211 The plaintiffs and their parents claimed that the school board’s imposition of school attendance zones over the existing segregated residential pattern had produced an unconstitutional school system in the city.212 Moreover, they alleged the school board’s policy in school construction and its approval of optional attendance zones in fringe areas, unconstitutionally perpetuated segregated schools.213 The court of appeals, affirming the district court’s order for a busing plan which included fifty-three suburban districts surrounding Detroit,214 held that a constitutionally adequate system of public schools could not be established within the Detroit school district’s geographic limits and that a multidistrict plan was necessary to remediate the injuries inflicted within Detroit.215 When the case reached the Supreme Court, the issue was whether the remedy for de jure segregation practiced in a large urban school district could include inter-district busing where there was no proof that the surrounding districts participated in the inner city’s unlawful discrimination.216

Writing for the Court in a five-to-four decision, Chief Justice Burger stated that the multidistrict remedy was “wholly impermissible” and held that “[b]efore the boundaries of separate and autonomous school districts may be set aside by consolidating the separate units for remedial purposes or by imposing a cross-district remedy, it must first be shown that there has been a constitutional violation . . . that produces a significant segregative effect in another district.”217 Thus, the Court concluded that “without an interdistrict violation and interdistrict effect there is no constitutional wrong calling for an interdistrict remedy.”218 Moreover, the Court, admonishing the lower tribunals, said that the neighboring districts were denied a meaningful opportunity to present evidence or be heard on the propriety of a multidistrict remedy, or on the question of constitutional violations by those districts.219

The Milliken Court asserted that the constitutional right of black children residing in the Detroit public school district was only to attend unitary school systems in that district, and unless officials drew lines in a discriminatory fashion, or arranged for white students residing in the district to attend schools in neighboring districts, the outlying districts were under no constitutional duty to make provisions for the students to attend such schools.220

In Milliken, the five majority votes came from new Nixon appointees, Justices Warren E. Burger, Harry Blackmun, Lewis F. Powell, Jr. and William F. Rehnquist and Eisenhower appointee Justice Potter Stewart (concurring and filing an opinion).221 The dissenters were: Justice William O. Douglas, a Roosevelt appointee; Justice Byron White, a Kennedy appointee; Justice William Brennan, an Eisenhower appointee; and Justice Thurgood Marshall, a Johnson appointee.222

Several aspects of Milliken warrant emphasis. First, the Court explained that its decision would be guided by the principle that the scope of the remedy could not exceed the scope of the violation.223 This rule had played little role in cases like Swann224 and Keyes,225 where the remedy clearly exceeded the scope of the initial constitutional violation. In Milliken, however, the Court placed greater emphasis on the importance of local control, enabling it to tightly link the violation and its cure.226 Thus, unless the evidence showed, for example, that school district boundaries were purposefully gerrymandered to promote segregation, cross-district busing could not be used to fix the violations which occurred within the city only.227

A second aspect of Milliken warranting emphasis is that the de facto-de jure distinction undergirds the Court’s holding.228 This meant that those seeking to impose broad remedies would have to overcome the hard fact that much segregation was attributable to housing discrimination, not policies set up by suburban school districts to segregate their schools.229

In other words, to the extent that suburban school districts might otherwise be motivated to segregate their schools, discriminatory housing practices did that work for them.230 This meant racial gerrymandering became unnecessary and the case for plaintiffs much more difficult to prove.231 Moreover, even if some derivative wrongdoing could be attributed to the state through local entities, it is undeniable that socioeconomic factors, including wealth and private discrimination, contributed to the inability of minority families to live in certain locations.232 Thus, by maintaining the de facto-de jure distinction, the Milliken Court effectively sealed off the suburbs from racial integration.233

Swann and Milliken thus gave a powerful one-two punch to defeat desegregation in urban school districts. Swann compelled all-out busing in large urban districts, giving urban middle class whites a reason to leave the city, while Milliken became a guarantor that those who made it to the surrounding suburbs would be spared from busing.234

Thus, within a year after its decision in Norwood,235 the Court in Milliken placed substantial limits on the power of courts to “impose a multidistrict, area-wide remedy to a single-district de jure segregation problem.”236 The Gilmore237 Court then signaled its reluctance to attribute Equal Protection violations, when private and governmental entities acted jointly in taking actions which seemed to perpetuate pre-existing segregation, at least when competing First Amendment associational values were implicated.238

The range of permissible remedial orders was further narrowed in Pasadena City Board of Education v. Spangler.239 In Spangler, the school district was under a court order that required, “beginning with the 1970-1971 school year, [that] there would be no school ‘with a majority of any minority students.’ ”240 The defendant Board of Education complied with the order and assigned students in a racially neutral manner, thereby meeting the “no majority” requirement.241 During the following school year, however, the school system had an increasing number of schools that were not in compliance with this requirement.242 This failure was not the result of the school board’s purposeful discrimination, but rather from random changes in residential housing patterns in the district.243 In this light, the school board asked the court to relieve it from the standing order for continuous judicial oversight and racial balancing.244

Justice Rehnquist, writing for the majority, held that once schools achieve court-ordered desegregation of student populations, the court may not require school officials to make yearly alterations to student assignment plans in order to maintain a strict numerical ratio of majority-minority students.245 Such ratios may be used by the courts only as guidelines for the initial transition from unlawfully segregated to unitary schools.246 The Court reasoned that the Fourteenth Amendment focuses on official governmental action, not conduct engaged in by private actors.247 Indeed, “having once implemented a racially neutral attendance pattern in order to remedy the perceived constitutional violations on the part of the defendants, the [d]istrict [c]ourt had fully performed its function of providing the appropriate remedy for previous racially discriminatory attendance patterns.”248

Spangler’s clear message was that the federal judicial role was a limited one: the federal courts’ job does not include perpetual policing of school districts to guard against resegregation.249 Moreover, the Court made it clear that the Equal Protection Clause neither bans segregation,250 nor assures actual equality in public education.251

Board of Education of Oklahoma City Public Schools v. Dowell,252 decided by a five-to-three vote, involved the question of when a school district operating under a desegregation injunction may be relieved from that order.253 In 1972, a federal district court had issued an injunction requiring the Board of Education of Oklahoma City Public Schools to implement a so-called “Finger Plan” under which the district bussed black students to white schools.254 In 1977, the district court withdrew its enforcement of the plan, declaring that the Board had complied with the plan and reached “unitary” racial composition.255 In 1984, the Board passed the Student Reassignment Plan (“SRP”), which reduced busing in an effort to lessen travel time for black students.256 The original group protesting segregation sought to restore the court-ordered desegregation, claiming that the school system again became segregated.257 The district court denied the application, but on appeal the United States Court of Appeals for the Tenth Circuit ruled that the injunction was never formally removed.258 On reconsideration, the district court again declined to restore the injunction because it found the original “Finger Plan” unworkable.259 Yet the court of appeals reinstated the injunction stating that the school district had an affirmative duty to disestablish the dual system.260

Chief Justice Rehnquist, writing for the majority, held that “federal supervision of local school systems [has always] been intended as a temporary measure to remedy past discrimination.”261 Once this school system shows earnest compliance with an injunction, the federal court can consider removing it.262

That said, Rehnquist wrote that the district court’s finding in 1977 that the plan had achieved its goal of racially unitary schools was not clear enough to dissolve the 1992 injunction.263 Instead, the proper analysis for the district court to apply would be to remove the injunction upon a factual finding “that the school system was [presently] being operated in compliance with the Equal Protection Clause, and that it was unlikely that the Board would return to its former ways.”264 If the Board satisfied these criteria, the district court could remove the injunction and consider the SRP separately on its own merits.265

In other words, the problem of resegregation, to the extent that it existed, would involve pleading and proving a separate claim independent of the initial one dating back to the early 1970s.266 Justice Rehnquist indicated that local districts were entitled to be relieved of burdensome desegregation orders displacing local authority at the point they can show that the “vestiges of past discrimination had been eliminated to the extent practicable.”267 Thus, dissolution of a desegregation decree was allowed after local parties had operated in compliance with it for a reasonable period.268

Neither a school board’s, nor a district court’s, duty ends with the initial desegregation order.269 Rather, there is a “continuing duty [for school officials] to eliminate the system–wide effects of earlier discrimination and to create a unitary school system untainted by the past.”270 Likewise, the district court “retain[s] jurisdiction until it is clear that [S]tate-imposed segregation has been completely removed.”271 The goal of the district court, nevertheless, is to return “schools to the control of local authorities at the earliest practicable date.”272 In discharging its duty, the district court must apply to the offending district the so-called “Green factors.”273

In Freeman v. Pitts,274 the Supreme Court faced the question of whether a district court is allowed to withdraw, in part, its supervision of a school system that is under a desegregation decree in the areas in which the school system complies with its decree, even if it does not comply in every area.275 There, the district court in 1969 ordered the DeKalb County School System (DCSS) to eliminate its previous de jure racial segregation and its inequitable byproducts, and maintained judicial control over DCSS’s compliance.276 However, in 1986, DCSS officials filed a motion for an order declaring that the DCSS officially desegregated and withdrawing court supervision.277 The district court found that DCSS was successful in four of the six Green categories; consequently, the district court decided to relinquish control over the DCSS in the four areas, while it maintained control and mandated further improvements in the areas in which segregation still existed.278

Both the respondents, the black school children and their parents, and the petitioners, the DCSS officials, appealed the district court’s decision, and the United States Court of Appeals for the Eleventh Circuit reversed the district court, holding that the district court should have authority over the DCSS until it is fully desegregated in all six categories for several years.279 In a unanimous decision reversing the court of appeals, authored by Justice Anthony M. Kennedy, the Court held that a district court has discretion to order an incremental or partial withdrawal of its control over a system in the categories in which the system has complied with the court-awarded desegregation plan.280 The Court contended that this incremental approach gives district courts a systematic and orderly means to their ultimate goal: to fully withdraw supervision.281 In addition, the Court reasoned that this procedure allows the district court to focus more attention and resources where they are needed most.282

Pitts was an extension of Board of Education v. Dowell283 since it eased the ability of school districts to be relieved of prior desegregation decrees. Perhaps more importantly, Pitts’ unanimity seems to represent a growing realization by the entire Court that, by continuing its long-term receivership of public education systems, it was contributing to a failed social experiment in a society with a highly mobile population and too many moving parts to control.284 Moreover, Pitts seems indicative of the Court’s heightened deference to local decision making in an emerging federalist era.

Missouri v. Jenkins examined the scope of district courts’ remedial powers in addressing the effects of previously established de jure segregation and the circumstances under which school districts could be released from federal court supervision.285 In an opinion delivered by Chief Justice Rehnquist on a five-to-four vote, Jenkins, like Pitts, further circumscribed the permissible scope of remedial orders, signaling the Court’s further disengagement from school district oversight.286

Jenkins’ history was a long one, but suffice it to say that it was not until 1977 that a federal district court ordered the desegregation of Kansas City, Missouri Public Schools (“KCMPS”).287 It was apparent that the federal court’s desegregation order made a difference. In 1983, twenty-four schools in the district had a black student enrollment of 90% or more.288 By 1993, no elementary level students attended a school with an enrollment that was 90% or more black.289 At the middle school and high school levels, the percentage of students attending schools with black student enrollment of 90% or more declined from about 45% to 22%.290

In Jenkins, Missouri challenged the district court’s orders requiring the state to (1) fund salary increases for virtually all instructional and non-instructional staff within the KCMSD and (2) continue to fund remedial “quality education” programs because student achievement levels were still “at or below national norms at many grade levels.”291

In affirming the orders, the court of appeals rejected the State’s argument that the salary increases exceeded the district court’s remedial authority because they did not directly address and relate to the State’s constitutional violation: that is, its operation, before 1954, of a segregated school system within the KCMSD.292 The court of appeals said, among other things, that the increases were designed to eliminate the vestiges of State-imposed segregation by improving the “desegregative attractiveness” of the district and by reversing a “white flight” to the suburbs.293 The court of appeals also approved the district court’s implicit rejection of the State’s request for a determination of partial unitary status, in reliance on Freeman v. Pitts,294 with respect to quality education programs.295

The Supreme Court, in an opinion written by Chief Justice Rehnquist, ruled in favor of the State on every issue.296 There were three principal parts of the Court’s holding. First, the Court ruled that the district court’s order, that tried to attract nonminority students from outside the district, was impermissible in scope because there was no proof of an interdistrict violation.297 Second, the Court ruled that the district court lacked authority to order an increase to teacher salaries, although the district court believed that an across-the-board salary increase would attract teachers, a factor it considered essential for desegregation.298 The Supreme Court concluded it was not necessary or permissible as a remedy.299 Third, the Court ruled that the continued disparity in test scores between black and white students did not justify continuance of the federal court’s desegregation order.300 The Court wrote that the Constitution requires equal opportunity, not equality in results.301 Therefore, disparities between black and white students on standardized tests are not a sufficient basis for concluding that desegregation had not been achieved.302 The Court held that once a desegregation order is complied with, the federal court effort should be ended.303

Jenkins commands that release from federal court supervision is informed, in part, by (1) whether there has been compliance with the decree in those aspects of the school system where federal supervision is to be withdrawn; (2) whether retention of judicial control is necessary or practicable to achieve compliance in other facets of the school system; and (3) whether the district has shown to the public and to the parents and students of the once disfavored race its good faith commitment to the whole decree and to those statutes and constitutional provisions that were the predicate for judicial intervention in the first place.304

Ultimately, the Court stated, the question is whether the constitutional violator has complied in good faith with the decree since it was entered, and whether the vestiges of discrimination have been eliminated to the extent practicable.305

Thus, Milliken v. Bradley, Board of Education of Oklahoma City Public Schools v. Dowell, Freeman v. Pitts, and Missouri v. Jenkins, taken together, gave powerful tools to school districts previously segregated by law or practice, to show that unlawful school segregation no longer existed. Necessarily, this discouraged lower courts from undertaking remedial work for past purposeful segregation.306 Moreover, these cases encouraged district courts to decline to impose remedial measures, even where unlawful discrimination is presently found. The lower courts have accepted this invitation.307 These factors, when combined with seminal decisions such as Arlington Heights308 and Washington v. Davis,309 making systemic race-based Equal Protection claims in public education extraordinarily difficult to prove, effectively closed the door on efforts to desegregate America’s schools, leaving only the most determined liberal perfectionists to pursue this elusive goal.310

Part III.  Benign Race-Based Classifications and K-12 Desegregation

The developments described so far address race-based Equal Protection claims of invidious discrimination, and limits on courts’ remedial authority for purposeful K-12 segregation. These cases do not encompass the capacity of school boards to remedy racial segregation in the absence of identified purposeful governmental action. As the preceding cases presaged, such voluntary efforts were not likely to receive a positive reception at the Supreme Court and, indeed, they have not.

The Court’s most recent decision involving voluntary integration efforts is Parents Involved in Community Schools v. Seattle School District No. 1.311 The Parents Involved cases emanated from Louisville, Kentucky, and Seattle, Washington, where each school district adopted plans which used race as one factor in assigning students to schools to achieve greater racial diversity.312

Louisville’s program encompassed grades kindergarten through twelve.313 That district had been previously segregated by law and subject to a judicial desegregation order.314 However, the order had been lifted not long before Louisville adopted its own desegregation plan.315 In other words, Louisville had attained unity when it adopted its plan.316 Although Seattle had never been segregated by law, it nevertheless adopted a plan that used race as a factor in assigning students to high schools to achieve greater racial diversity.317

In a five-to-four decision, the Parents Involved Court found both plans to be unconstitutional.318 The five Justices in the majority agreed that since racial criteria were part of the plans they must be subjected to strict scrutiny, even if using race to achieve desegregation.319 However, those five Justices divided on the reasons for entry of the judgment. The Roberts four-Justice plurality found that Seattle and Louisville lacked a compelling interest for their desegregation programs since racial criteria could only be used to remedy constitutional violations.320 Moreover, these Justices argued that diversity in classrooms was not a sufficiently compelling interest to meet strict scrutiny standards. Chief Justice Roberts wrote:

However closely related race-based assignments may be to achieving racial balance, that in itself cannot be the goal, whether labeled “racial diversity” or anything else. To the extent the objective is sufficient diversity so that students see fellow students as individuals rather than solely as members of a racial group, using means that treat students solely as members of a racial group is fundamentally at cross-purposes with that end.321

Thus, the plurality view would completely preclude the use of race in voluntary student assignments.322 Justice Thomas made a similar point in a separate concurrence.323

Justice Kennedy joined the majority’s judgment, agreeing that strict scrutiny applied.324 However, Kennedy disagreed with the Roberts plurality that achieving racially diverse schools was not a compelling governmental interest.325 In this regard Kennedy wrote in his concurrence: “In the administration of public schools by the state and local authorities it is permissible to consider the racial makeup of schools and to adopt general policies to encourage a diverse student body, one aspect of which is its racial composition.”326

Despite his disagreement with the plurality on this point, Kennedy concurred in the judgment, since he thought Seattle and Louisville failed to implement their programs in a way that was “narrowly tailored.”327 He argued that in assigning individual students, race “may be considered legitimate only if [it is] a last resort to achieve a compelling interest,” and that the school districts failed to carry their burden in this regard.328 Thus, an understanding of what narrow tailoring means in the context of voluntary integration programs is the most important lesson to be derived from this decision.

Kennedy identified several alternatives that school systems can use to achieve greater racial diversity in their schools:

School boards may pursue the goal of bringing together students of diverse backgrounds and races through other means, including strategic site selection of new schools; drawing attendance zones with the general recognition of the demographics of neighborhoods; allocating resources for special programs; recruiting students and faculty in a targeted fashion; and tracking enrollments, performance, and other statistics by race. These mechanisms are race conscious but do not lead to different treatment based on a classification that tells each student he or she is to be defined by race, so it is unlikely any of them would demand strict scrutiny to be found permissible.329

Parents Involved is a paradigmatic illustration of how the Court’s members view K-12 segregation in both its constitutional and policy dimensions. The views of Chief Justice Roberts and Justices Scalia, Thomas, and Alito regarding the constitution depart markedly from the dissenters, as well as from Kennedy’s. These four assert that the Constitution demands complete colorblindness and minimization of the importance of diversity in public schools as a constitutional value absent proof of intentional racial segregation.330 Moreover, they assign little relevance to Brown in voluntary desegregation efforts.331 Justice Roberts said:

Before Brown, schoolchildren were told where they could and could not go to school based on the color of their skin. The school districts in these cases have not carried the heavy burden of demonstrating that we should allow this once again — even for very different reasons. For schools that never segregated on the basis of race, such as Seattle, or that have removed the vestiges of past segregation, such as Jefferson County, the way “to achieve a system of determining admission to the public schools on a nonracial basis,” is to stop assigning students on a racial basis. The way to stop discrimination on the basis of race is to stop discriminating on the basis of race.332

Justice Breyer wrote a lengthy dissent to the judgment, in which Justices Stevens, Souter, and Ginsburg joined.333 Among the concerns raised by the dissenters were how American public schools had become increasingly racially segregated and the adverse effects of the judgment on existing desegregation plans.334

Justice Breyer argued that the majority was required to show more deference to legislative decisions locally enacted, and in what may be fairly regarded as liberal perfectionist rhetoric said: “To invalidate the plans under review is to threaten the promise of Brown. The plurality’s position, I fear, would break that promise. This is a decision that the Court and the Nation will come to regret.”335

The importance of Justice Kennedy’s concurring opinion cannot be overstated. His brand of narrow tailoring is one that allows direct use of race, but only as a last resort.336 In essence, he asserts that school districts may use not only race-neutral, but race-conscious measures to integrate their schools.337 This would include, for example, creating attendance zones and building site locations with an eye toward integration.338 Justice Kennedy also encourages using proxies for race such as place of residence and socioeconomic factors to facilitate racial integration.339 Thus, Justice Kennedy effectively imposes an exhaustion of remedies requirement that must be satisfied before race qua race may be employed. Even then, however, race should not be dispositive but one of several factors guiding student assignment and admission.340

Although Justice Kennedy’s opinion stands alone, it is controlling for the Court on the question of when and how school districts may take race into account in assigning students. Because of his vote, Justice Kennedy saved voluntary integration plans from extinction, albeit subjecting them to his own brand of strict scrutiny. Yet to date, Justice Kennedy has never approved a benign discrimination program adopted by states or their subdivisions.341 The dissenters would have given far more latitude to school districts to consider race than would Justice Kennedy, but were cabined by Justice Kennedy’s approach and had to take what they could get.342

It is perhaps worthwhile to look back and consider that during the Court’s peak enforcement period343 the Court required that race be taken into account in efforts to attain unitary status.344 Of course, this meant classifying persons by race and assigning them accordingly, and indeed racial classifications were used in countless desegregation cases.345 Now, under Parents Involved, absent exhaustion of race “neutral” proxies, school districts may not take race into account in student diversity plans.346 Of course, there is a difference between the earlier de jure invidious discrimination cases and the later de facto benign discrimination ones, which accounts for these differences, but there is more at work here than the distinction between governmental and privately created segregation.

During the peak enforcement period the Court explicitly endorsed voluntary integration efforts. In Swann, for example, the Court said school districts could, of their own volition, assign students so that each school featured “a prescribed ratio of Negro to white students reflecting the proportion for the district as whole.”347 In Parents Involved, Chief Justice Roberts dismissed such language as nonbinding dicta.348 That said, there is a strong case to be made that the Swann Court would have upheld plans like those devised in Parents Involved.349 The likelihood of such plans receiving the Court’s imprimatur today has wholly vanished. This difference is probably not best explained by the logic of jurisprudential niceties, but the Supreme Court appointment process used by presidents during various intervals in the post-Brown period.350 We examine this issue in our statistical analysis in Part VI, infra.

Part IV.  Benign Race-Based Discrimination in College and University Admissions351

It was not until 1989 that the Supreme Court, in Richmond v. J.A. Croson Co.,352 expressly held that strict scrutiny should be used in evaluating State and local benign race discrimination programs.353 There, the Court invalidated a Richmond, Virginia plan to set aside 30% of public works monies for minority-owned businesses.354 About one year later the Court, in a five-to-four decision in Metro Broadcasting, Inc. v. Federal Communications Commission, held that congressionally approved “affirmative action” programs need to meet only intermediate scrutiny.355 As described below, this shift would be short lived.

In 1995, in Adarand Constructors, Inc. v. Pena, the Supreme Court said: “[A]ll racial classifications, imposed by whatever federal, state, or local governmental actor, must be analyzed by a reviewing court under strict scrutiny.”356 Notably, between the Court’s decision in Metro Broadcasting in 1990, and Adarand in 1995, four of the Justices in the Metro Broadcasting majority resigned, but none of those in the dissent resigned. In Adarand, the four dissenters from Metro Broadcasting joined with Justice Thomas to form a majority to overrule Metro Broadcasting,357 concluding that “[f]ederal racial classifications, like those of a [s]tate, must serve a compelling governmental interest, and must be narrowly tailored to further that interest.”358

Between 2003 and 2013, the United States Supreme Court was given the opportunity in three cases to apply the principles announced in J.A. Croson and Adarand to race-conscious college and university admission decisions. These were Grutter v. Bollinger,359 Gratz v. Bollinger,360 and Fisher v. University of Texas.361 The Court’s opinion in each of these cases is analyzed below.362

Grutter, which gave the green light to benign race discrimination in college and university admissions,363 produced strong backlash in Michigan, resulting in a state constitutional amendment known as Proposal 2 which forbade racial preferences in college and university admissions.364 This in turn produced litigation challenging Proposal 2 on Equal Protection grounds in a case titled Schuette v. Coalition to Defend Affirmative Action,365 which reached the Supreme Court in 2014.366 We begin this Part by addressing decisions that concern race-conscious admissions practices and then turn to state countermeasures to those practices using Schuette as our template.367 In furtherance of this goal we look back on the 1978 Regents of the University of California v. Bakke368 decision, which substantially influenced the outcome of these later cases.

A.  Race-Conscious College and University Admissions

In Bakke, the Court examined an Equal Protection challenge to a benign medical school admissions program that allocated sixteen slots in the entering class of 100 for minority students.369

There was no majority opinion. Four Justices — Brennan, White, Marshall, and Blackmun — asserted that intermediate scrutiny was the appropriate standard to apply to racial classification benefitting minority students.370 They argued that “a number of considerations . . . lead us to conclude that racial classifications designed to further remedial purposes ‘must serve important governmental objectives and must be substantially related to achievement of those objectives.’ ”371 These four Justices voted to uphold the medical school’s affirmative action program.

Justices Stevens, Burger, Stewart and Rehnquist concluded that the affirmative action program violated Title VI of the 1964 Civil Rights Act, which prohibited discrimination by institutions receiving federal funds.372 These Justices did not reach the constitutional issue or dis- cuss the level of scrutiny. However, five Justices — Brennan, White, Marshall, Blackmun, and Powell — rejected this view arguing that the analysis under Title VI and the Constitution is identical.373

Most importantly, however, Justice Powell, writing only for himself in Bakke, said that strict scrutiny should be used for affirmative action and the medical school’s admissions program failed this test. He said that “[r]acial and ethnic distinctions of any sort are inherently suspect and thus call for the most exacting judicial examination.”374 Powell held that although the quota employed in the set-aside was unconstitutional, it was permissible for race to be used as one factor in admissions decisions to enhance diversity.375

Thus, the vote in Bakke was five to four invalidating the racial set aside with Justices Powell, Stevens, Burger, Rehnquist, and Stewart subscribing to this view, but also five to four holding that it is permissible for universities to use race as a factor in admissions to increase diversity. The latter view was supported by Justices Powell, Brennan, Marshall, White, and Blackmun. In the end, Justice Powell’s standalone view would emerge as the most important influence in the Court’s later college and university decisions about race-conscious decision making. It is to those decisions that we now turn.

In Grutter, a white Michigan resident applied for admission to the University of Michigan’s law school.376 Grutter had a 3.8 undergraduate GPA and an LSAT score of 161.377 The law school acknowledged that it used race as a factor in making admissions decisions, asserting its practice served an educationally compelling interest in achieving diversity among its student body.378 The court of appeals had held that Justice Powell’s opinion in Bakke379 constituted a binding precedent establishing diversity as a compelling governmental interest, sufficient under strict scrutiny review, to justify the use of racial preferences in admissions.380 The court of appeals also held that the law school’s use of “critical mass” criteria in determining whether enough diversity was achieved was not the functional equivalent of a quota.381

In a five-to-four decision, delivered by Justice O’Connor, the Court held that the Equal Protection Clause does not prohibit the law school’s narrowly tailored use of race in admissions decisions to further a compelling interest in “obtaining the educational benefits that flow[ed] from a diverse student body.”382 The Court reasoned that, because the law school conducts highly individualized review of each applicant, no acceptance or rejection is based automatically on a variable such as race, and that this process ensures that all factors that may contribute to diversity are meaningfully considered alongside race.383 Justice O’Connor wrote, “[I]n the context of its individualized inquiry into the possible diversity contributions of all applicants, the [l]aw [s]chool’s race-conscious admissions program does not unduly harm nonminority applicants.”384

In Gratz v. Bollinger, the Court confronted a complaint by two applicants who were rejected for admission to the University of Michigan’s undergraduate College of Literature, Science and the Arts.385 They asserted the admission process to which they were subjected ran afoul of Equal Protection constraints.386 The University admitted that it used race as a factor in making admissions decisions, arguing that it served a compelling interest in providing educational benefits through achieving diversity among its student body.387 In addition, the University had a policy to admit virtually all qualified applicants who were members of one of three select racial groups: African Americans, Hispanics, and Native Americans.388 They were considered to be underrepresented on the campus.389

In a six-to-three decision, authored by Chief Justice William Rehnquist, the Court held that the University of Michigan’s use of racial preferences in undergraduate admissions violated both the Equal Protection Clause and Title VI.390 While rejecting the argument that diversity cannot constitute a compelling governmental interest,391 the Court reasoned that the automatic distribution of twenty points, or one-fifth of the points needed to guarantee admission, to every single “underrepresented minority” applicant, solely because of race, was not narrowly tailored and did not give the individualized consideration Justice Powell contemplated in Bakke.392 Chief Justice Rehnquist wrote: “[B]ecause the University’s use of race in its current freshman admissions policy is not narrowly tailored to achieve [the University’s] asserted compelling interest in diversity, the admissions policy violates the Equal Protection Clause.”393

Yet little solace was afforded to liberal perfectionists in Chief Justice Rehnquist’s acceptance in Gratz that diversity can be a compelling interest. He would seemingly allow diversity criteria to be used only as a remedy for past proven discrimination by the agency using race-conscious criteria, not for general societal discrimination.394

Grutter-Gratz thus followed the path laid down by Justice Powell in Bakke about twenty-five years earlier: diversity is a compelling interest in college and university admission because it promotes educational benefits, and these institutions may make race conscious decisions to ensure such diversity.395 However, where quota or numerical goals are established, they run afoul of Equal Protection requirements.396

Distinguishing between “critical mass” on the one hand, and numerical set-asides on the other hand, however, is no easy task, especially where an admissions committee uses race as a factor in selecting students while keeping track of the number of minority students in order to know if it has reached this critical mass.397 Unless judges simply take colleges and universities at their word and become obsequious in their review of the conduct of admission officials, it seems to us that critical mass and quotas are distinctions without much difference.

When Grutter and Gratz were decided in 2003, “seven of the nine [J]ustices saw no difference” between critical mass and a quota.398 “Four Justices — Rehnquist, Scalia, Kennedy, and Thomas — would have struck down both programs; three Justices — Stevens, Souter and Ginsburg — would have upheld both. Only Justices O’Connor and Breyer saw a meaningful difference.”399

Since 2003, significant personnel changes have occurred on the Court: Chief Justice Roberts replaced Chief Justice Rehnquist, Justice Alito replaced Justice O’Connor, Justice Sonia Sotomayor replaced Justice Souter, and Justice Elena Kagan replaced Justice Stevens.400 It will be recalled that in Parents Involved,401 Justices Roberts, Alito, Scalia, and Thomas rejected racial diversity as a compelling state interest.402 Since Justice Kennedy concluded in both Grutter and Gratz the admission programs reviewed in each case were unconstitutional, it was uncertain how he would vote when Parents Involved reached the Court.

In Parents Involved, Justice Kennedy and the four dissenters accepted the notion that diversity in elementary and secondary schools was a compelling interest.403 Moreover, Justice Kennedy not only placed greater emphasis on narrow tailoring, but also defined this prong as including a requirement for exhaustion of facially neutral criteria before race-conscious decision making could be used.404 In his calculus, Justice Kennedy counted as “neutral” factors that were related to race, but not explicitly so.405 Therefore, the government would have to show that it exhausted these “neutral” factors as part of its proof when race-conscious programing was challenged.

Justice Kennedy rationalized this approach by purporting to follow precedent.406 By giving school districts extraordinary deference on the compelling interest prong of strict scrutiny, Justice Kennedy moved the playing field to “narrow tailoring.”407 Justice Kennedy’s approach set the stage for how the Court would handle Fisher v. University of Texas,408 the next affirmative action dispute it confronted involving college and university admissions.

By way of background on the Fisher case, in 1997, in an effort to improve the proportion of racial minorities in the university undergraduate population at the University of Texas, the Texas Legislature enacted a law requiring the University of Texas to admit all high school seniors ranked in the top 10% of their high school class.409 After finding that differences still remained between the proportion of racial minorities in the applicant pool and those in the university undergraduate population, the University decided to modify its “race-neutral” admissions policy.410 The policy continued to admit all in-state students who graduated in the top 10% of their high school classes.411 For the rest of the in-state freshman class, the University would consider race as a factor in admission.412

Abigail Fisher, a white female, applied to the undergraduate program in 2008.413 Fisher was not in the top 10% of her class, so she competed for admission with other non-top 10% in-state applicants.414 The University of Texas denied Fisher’s application.415 Fisher filed suit against the University and related defendants, asserting that the use of race as a consideration in admission decisions violated the Equal Protection Clause of the Fourteenth Amendment.416 Both the federal district court and the United States Court of Appeals for the Fifth Circuit upheld the program.417

At the Supreme Court the Justices reaffirmed that strict scrutiny was the guiding principle when race was used as a factor in university admissions.418 The Court upheld Fisher’s claim insofar as she argued that the lower courts erred in their application of strict scrutiny’s narrow tailoring requirement.419 Notably, Fisher did not challenge the University’s contention that the educational benefits of diversity were compelling and the Court did not revisit that important issue.420

Delivering the opinion for a seven-to-one Court, Justice Kennedy wrote that the court of appeals did not hold the University of Texas admissions policies to a standard of strict scrutiny, so the judgment was incorrect.421 Justice Kennedy said that courts must determine whether race conscious policies are “precisely tailored to serve a compelling governmental interest.”422 If the policy is not “precisely tailored to serve a compelling governmental interest,” it may not be considered in the admissions process.423

Justice Kennedy explained that it was the duty of the reviewing court to verify that the University policy in question was necessary to achieve the benefits of diversity and that no race-neutral alternative would give the same benefits.424 This was a reformulation of the exhaustion criteria he applied in Parents Involved. Because the lower courts in Fisher did not properly apply the narrow tailoring inquiry, the Court remanded to the Fifth Circuit for it to answer this question under the proper standard.425

Justice Scalia wrote a concurring opinion in which he argued that “[t]he Constitution prohibits governmental discrimination on the basis of race.”426 However, because this case did not ask the Court to overrule precedent that allowed universities to consider diversity a compelling interest that justified race-based admissions policies, he joined the majority’s opinion in full.427

In his separate concurrence, Justice Thomas wrote that he joined the majority’s opinion that the lower courts did not sufficiently apply strict scrutiny, but he also argued that “the Equal Protection Clause of the Fourteenth Amendment prohibits a state’s use of race as a factor in higher education admissions.”428 He further argued that there is nothing “necessary” about the benefits that flow from racial diversity in higher education, so there is no compelling state interest to promote it.429

Justice Ginsburg, the sole dissenter in Fisher, argued that the University’s admission policy treats race as merely one factor in the overall decision to admit a student, which is permissible under previous judicial precedent.430 She also argued that the Equal Protection Clause does not require that state universities be blind to the history of overt discrimination, and that it is preferable that they explicitly include race as a factor in admission decisions, rather than try to obfuscate its role.431 Indeed, Justice Ginsburg asserted that “only an ostrich could regard the supposedly neutral alternatives as race unconsciousness.”432

Ginsburg explained that “[i]t is race consciousness, not blindness to race, that drives such plans.”433 However, the other Justices in Fisher ignored Justice Ginsburg’s remarks. More than twenty years earlier Justice Sandra Day O’Connor, writing for the Court in Croson, first suggested that a “race-neutral alternative” could be a policy that is facially neutral, but designed to achieve a racial goal.434 Justice Ginsburg’s rejection of this kind of thinking is undeniably sound. As a matter of logic, how can an approach whose purpose is to achieve an increase in racial enrollment be considered “race-neutral,” when the reason it is selected is because of its correlation with race?435

Moreover, since the Texas Top Ten Percent Plan’s purpose was adopted to increase minority enrollment, the motives for the legislation were necessarily infused with race-consciousness, although the policy was written in facially neutral terms.436 Of course, the Fisher Court did not confront this issue, since Fisher did not challenge the University of Texas’ Top Ten Percent Plan.437

There is a credible case to be made that Justice Kennedy’s shifting the Equal Protection focus to “narrow tailoring” begs a more important question: does the government’s interest in educational benefits include racial diversity, assuming that the educational benefits of diversity are compelling? If racial diversity is a compelling interest, then it is curious that one would choose racial proxies to achieve this goal rather than a more direct method.438 Michael Rosman has argued the use of “racially motivated manipulation of facially neutral criteria to achieve a racial goal” is akin to “requiring the serious consideration of a slow-moving alternative to achieving a speed goal.”439

Ultimately, proponents of the prevailing affirmative action process, as manifested in cases like Fisher, must confront the questions of (1)  how to operationally define benefits which result from diversity; (2)  whether these benefits may be derived from sources such a class and other nonrace-conscious criteria; and (3) whether any measureable educational benefits derive from the use of racial criteria. By eschewing the importance of these questions and instead deferring to colleges and universities on the compelling governmental interest prong of Equal Protection law, the Supreme Court may have abdicated its responsibility.

B.  Counter-Measures to Benign Racial Discrimination

In Schuette v. Coalition to Defend Affirmative Action, the Supreme Court, in a highly fragmented six-to-two decision, rejected the claim that an amendment to Michigan’s state constitution, known as Proposal 2, which prohibited race and sex based discrimination and preferential treatment in public university admissions decisions, violated the Equal Protection Clause of the Fourteenth Amendment.440

Proposal 2 was a direct response to Grutter, discussed in the previous section, which upheld the use of race as one factor among many in law school admissions to ensure educational diversity.441 Proposal 2 was approved by 58% of Michigan voters.442 The United States Court of Appeals for the Sixth Circuit ruled, en banc, in an eight-to-seven vote, that Proposal 2 violated the Equal Protection Clause by running afoul of the so-called “political restructuring doctrine.”443 The court concluded that Proposal 2 made it more difficult for disfavored minorities to participate in the give and take of politics, insofar as it affected college and university admissions.444 This was because they were relegated to the cumbersome process of constitutional amendment to affect changes in university admissions policies, while other groups such as alumni and students with special talents, for example, would not have to go through that process to affect the admission process.445

The issue, as framed by the Supreme Court, was “not the permissibility of race-conscious admissions policies under the Constitution but whether, and in what manner, voters in the [s]tate may choose to prohibit the consideration of racial preferences in governmental decisions, in particular with respect to school admissions.”446

Writing for a three-Justice plurality,447 Justice Kennedy asserted the decision was a modest one, explaining that “this case was not about the constitutionality of race-conscious admissions,” but rather about whether the “voters in the States may choose to prohibit the consideration of such racial preferences” in the decisions of governmental bodies, specifically with respect to school admissions.448 The plurality held that the attempt to define and protect interests based on race ran the risk of allowing the government to classify people based on race and therefore perpetuate the same racism such policies were meant to alleviate.449 Justice Kennedy said that while voters certainly may determine that some race-based preferences should be adopted, it is not the role of the courts to disempower voters from making such a choice.450 If certain issues were decided to be too sensitive to be addressed by voters, it would be denying the voters their right to debate and act through the lawful democratic process.451

Chief Justice Roberts concurred, arguing that the use of racial preferences might reinforce racial awareness and therefore “do more harm than good.”452 Justice Scalia, concurring in the judgment, argued that (1)   a state law that gave equal protection by not allowing the use of racial preferences at least facially did not violate the Constitution; (2)   judges should not be in the position of dividing the country into “racial blocs” in determining what policies are in each one’s interests; and (3) there was no reason to allow local subordinate authorities like universities to have more power over the use of race-based preferences than the voters of the state, and since the amendment in question prohibited the use of racial preferences, it gave equal protection under the law rather than denying it.453

Justice Breyer wrote a separate opinion concurring in the judgment in which he argued that, while the Constitution “allows local, state, and national communities” to implement “narrowly tailored, race-conscious” policies, it is the voters and not the court who should determine the merits of such strategies.454 Justice Breyer contended that the amendment better allowed for this process to take place because it took the power to decide whether to implement race conscious policies away from unelected officials and placed it firmly “in the hands of the voters.”455 Justice Scalia, joined by Justice Thomas, separately concurred in the result. They asserted that although the constitutional amendment would most likely not survive under the political-process doctrine, that doctrine should be overruled since it is inconsistent with established Equal Protection law. Equal Protection law requires that for plaintiff to prevail, the challenged state action must reflect a discriminatory purpose, a condition plainly absent in Schutte.456 Justice Sotomayor wrote a vigorous dissent in which Justice Ginsburg joined.457

Schuette’s outcome was predictable. After all, between Brown’s 1954 issuance and 2014, the Supreme Court had never held that benign discrimination benefitting racial minorities was constitutionally required under the Equal Protection Clause. It would seem to follow that when Michigan voters amended their constitution to prohibit the use of racial preferences by the State and its creatures, there was little a federal court could or should do to stop it.

Schuette’s message was not lost on the Court’s recently seated liberals who still envisioned a realization of what they considered Brown’s ideals.458 Schuette has a decidedly federalist slant to it, deferring to states’ rights emanating from the Tenth Amendment, including the will of the voters. Accordingly, Schuette made clear it was unwilling to decide for independent state entities, at what level in their decision-making hierarchies, policy about diversity initiatives should be made. This is especially so when, as there, the state constitutional amendment procedure was followed. Moreover, Schuette is a decidedly minimalist decision, cautious not to enlarge the civil liberties-based edifice established during the stewardship of Chief Justice Warren and continued to a lesser extent by Chief Justice Burger and the Warren Court holdovers.

Part V.  The Supreme Court Selection Process in the Nixon and Reagan Administrations

In this Part we examine events that occurred in the Nixon and Reagan years. This examination was triggered by our sense that the four Nixon nominees (Justices Burger, Blackmun, Powell, and Rehnquist) and three Reagan nominees (Justices O’Connor, Scalia, and Kennedy) seated on the Court had a significant impact on the way systemic racial discrimination claims were adjudicated in the post-Brown period. The results of this inquiry will help in explaining our theoretical approach and give a basis for the empirical analyses we conduct in the next Part.

A.  The Nixon Appointments

During his 1968 campaign for the presidency, Richard Nixon promised to appoint conservative Southerners and “strict constructionists.”459 In the wake of the resignations of Chief Justice Warren and Associate Justices Abe Fortas, Hugo Black and John Marshall Harlan II, during Nixon’s first term, Nixon had opportunities to fulfill this promise. When Warren resigned in 1969, Nixon soon settled on Warren Burger, a conservative federal judge from the Eighth Circuit, whose negative views on the Warren Court had drawn Nixon’s attention.460 Burger was confirmed in the Senate by a 74-3 vote.461

Although his nomination of Burger was successful, Nixon’s next two appointees to fill Abe Fortas’ seat were blocked. Nixon first chose Clement Haynsworth in 1969, and then G. Harold Carswell in 1970, for Fortas’ seat.462 These nominees were rejected by the Democratic controlled Senate by 45-55 and 45-51 votes, respectively.463

Following these defeats Nixon nominated Burger’s colleague from Minnesota, Eighth Circuit Judge Harry Blackmun.464 Nixon believed at the time that Blackmun was at bottom a “strict constructionist” and indeed Blackmun’s judicial record was conservative, though less so than Haynsworth’s or Carswell’s.465 Blackmun was confirmed by the Senate with a 94-0 vote.466 As we will see, Nixon’s assumptions about Blackmun were not entirely justified and would become even less so.

The following year Justices Black and Harlan announced their retirements within one week of each other, forcing the Court to start the 1971 term with only seven Justices.467 In October 1971, Nixon announced the nominations of Lewis Powell and William Rehnquist to replace Justices Black and Harlan.468

Powell was confirmed in 1971 with an 89-1 vote and Rehnquist by a 68-26 vote.469 Although Powell was a relative moderate compared to Carswell and Haynsworth, Powell was nevertheless a prominent conservative lawyer from the South, albeit moderate on civil rights.470 Nixon’s “real coup” however, was in enabling Rehnquist’s appointment, especially since he was a young and committed conservative activist from within the Nixon administration itself.471

This brief recounting reveals that Nixon faced considerable constraints in getting his “strict constructionists” on the Court. These obstacles were mainly created by the Democrats who controlled the Senate during these years.472 Nevertheless, relative to the Warren Court’s liberal perfectionists and holdovers from the Roosevelt era, Nixon had to consider these appointments better than what preceded them, albeit not as conservative as the Justices he would have preferred. Although he did not achieve all of his goals, Nixon’s efforts would soon be rewarded.

B.  The Reagan Appointments

Reagan’s first Supreme Court selection was conservative-leaning jurist Sandra Day O’Connor, chosen because he made a campaign pledge to appoint the first woman to the Court.473 Reagan might well have obtained a more conservative jurist but for this obligation.474 Moreover, Reagan stumbled in 1987 by nominating Robert Bork for the Court over more moderate — or at least less controversial — conservative jurists.475 Bork’s defeat by a 42-58 vote was the largest one in history.476

Bork had been “one of the most outspoken critics of the Warren and Burger Courts.”477 The Reagan Administration gave insufficient attention to the obvious fact that the Democrats had regained control of the Senate in 1986 and the signs that an intense confirmation battle would ensue over any nominee associated with new right Republicans.478 Moreover, Bork lacked the smoothness of, say, a John Roberts, had an extensive paper trail that was used to his disadvantage by his opponents, and was perhaps tone deaf during the confirmation process.479 Moreover, unlike the O’Connor nomination in 1981, the Rehnquist nomination in 1986, and the Scalia nomination in 1986, the vacancy created by Justice Powell’s resignation was considered a pivotal one since Justice Powell had supplied the crucial fifth vote in cases upholding abortion, affirmative action, and other controversial social issues.480

Reagan again faltered with the nomination of Douglas Ginsburg to the Court, not because of his lack of conservative bona fides, but on the political level. Ginsburg was insufficiently vetted, requiring him to withdraw from consideration because of personal indiscretions.481 Following these defeats the fallback choice was Anthony Kennedy who, while a conservative, is generally considered to be more centrist than conservative contemporaries, such as Robert Bork, Douglas Ginsburg, John Roberts, Samuel Alito, or Clarence Thomas.482

Despite these mistakes, the insight and substantial success that Reagan’s White House achieved in the confirmation of Justices consistent with its ideological goals was perhaps unparalleled in history.483

Part VI.  Empirical Analysis of Justices’ Voting: 1954 - 2014

A.  Overview

The discussion thus far has focused on the legal evolution of United States Supreme Court decisions pertaining to systemic racial discrimination in public education as well as related decisions from the Court, which we deemed to be important in terms of the overall trajectory of the Court following Brown. We now turn to an empirical analysis of the relevant data to explain the main factors at work in the decline of liberal perfectionist jurisprudence during this period.


The dataset is derived from the individual votes rendered between 1954 and 2014 across the sixty-five cases listed in Appendix A. These cases pertain to racial distinctions and the Equal Protection Clause in public education settings decided by the Court after argument. We included in our data base six decisions that we considered to have significant impact on the education decisions we reviewed, although they did not arise in educational settings.484 We adopt Brown I in 1954 as our starting point since it marks, among other things, the dawn of what we have termed a perfectionist vision of the United States Constitution on matters concerning education and race;485 that is, a conviction that the goal of a racially integrated society should guide constitutional interpretation. The evolution of the cases largely reflects, first, questions concerning the enforcement of Brown; second, the scope of remedial orders to achieve racial integration; third, the criteria for lifting remedial orders imposed on school districts; and finally, questions concerning affirmative action programs, culminating with Fisher v. University of Texas in 2013, and the backlash to Grutter, in Schuette v. Coalition to Defend Affirmative Action in 2014. The latter includes issues of State control over such policies.


The dependent variable is a dichotomous indicator of each individual voting decision across all the cases, coded as being either in a “liberal” or a “conservative” direction. Of a potential 585 votes, 33 are missing at random, leaving a total of 552 actual voting decisions to analyze. Of these, 351 were cast in a liberal direction while 201 were in a conservative direction. These represent 64% and 36% of the votes cast, respectively. Only 33 Justices were involved in casting the full 552 votes. As such, our statistical analysis, which follows accounts for the possibility that voting decisions, in part, reflect idiosyncratic attributes of Justices voting on a repeated basis.

While the coding of our dependent variable requires a degree of familiarity with each individual case, we do not believe that our allocation of any of the votes to one of the two categories raises any serious issues concerning subjective judgment. At first blush this may be surprising, given the somewhat vague character of the terms “liberal” and “conservative.”

Conceptually, what we mean by liberal voting is something more specific; namely, voting in a “liberal perfectionist” direction. Ideally it would be possible to translate this concept into an operational rule. For example, we considered coding “liberal” as a pro-plaintiff decision and “conservative” as pro-defendant. This rule appears to work well for cases concerning issues of segregation but not, unfortunately, for cases concerning benign racial discrimination/ affirmative action. In short, a liberal perfectionist believes in working towards a society free of racial discrimination and is willing to tolerate certain discriminatory effects in pursuit of that end. This view contrasts with Chief Justice Roberts’ opinion that “[t]he way to stop discrimination on the basis of race is to stop discriminating on the basis of race.”486 The liberal perfectionist, as we see it, views this tautological reasoning as arising out of a failure to distinguish between discrimination as an end itself versus discrimination as a byproduct of attaining a greater good.

In the following analysis, we code a liberal vote as zero and a conservative vote as one. While this choice is strictly arbitrary from a technical perspective, it means that positive effects of our explanatory variables are to be interpreted in terms of a departure from the liberal perfectionist norm that ensued for close to twenty years following Brown.


Our explanatory model of judicial decision making is partly influenced by the attitudinal theory of judicial behavior popularized in political science literature.487 According to Maltzman, the attitudinal model has always been viewed as the “culmination of the behavioral revolution as applied to the study of judicial politics.”488 Like Maltzman, we treat this framework with some caution. While the concept of ideology is central to our analysis, we seek to probe beyond truisms concerning the respective dispositions of Democrats and Republicans to contextualize any such ideological effects. As such, we first establish a baseline acontextual attitudinal model with a single predictor — the judge’s political ideology — operationalized in terms of an ex ante measure;489 specifically, the party of the appointing president. Consistent with the attitudinal model, we expect Justices appointed by Republican Party presidents to vote in a more conservative direction than Justices appointed by Democratic Party presidents in terms of statistical averages. In turn, we adopt this as our “baseline” model of decision making.

The foregoing analysis of the cases suggested a turning point in the progress of liberal perfectionism beginning in the early 1970s. In 1968, Richard Nixon ascended to the Presidency and thereafter had the fortuitous opportunity to appoint four new Justices to the Supreme Court: first, a new Chief Justice in Warren E. Burger (replacing the outgoing liberal perfectionist, Earl Warren), followed by Justices Harry Blackmun, Lewis F. Powell, Jr., and, finally, William Rehnquist.490 As we have illustrated, the progress of liberal perfectionism then ran into various obstacles pertaining to the nature of remedial orders relating back to Brown.491 Only one Justice, John Paul Stevens, was appointed in the period between Nixon’s resignation and Ronald Reagan becoming President in 1981.492 In turn, President Reagan had the opportunity to appoint three new Justices, and there are strong grounds to believe that these appointments also reflected a desire to increase conservative voting tendencies.493

Figure 1 illustrates how the role of ideology appears to have changed over time with respect to our cases under consideration. The chart shows the difference between the number of liberal and conservative votes cast each year by party of the appointing president, so that here, a positive difference indicates a liberal preponderance of votes and a negative difference indicates a conservative preponderance. Before 1972, there was not any clear difference between the voting directions of Democratic and Republican Party appointees. In fact, both types of appointees voted in a liberal direction on average, up until 1971. Thereafter, one begins to see the two voting in opposite directions.

We contend that the forces which operate at points in time create “perfect storms” that offer extraordinary opportunities for presidents to impact the direction of the Court. As such we contextualize our attitudinal model by including interaction effects between ideology and appointment eras, beginning in 1968 when Nixon took office and consider these effects for the Reagan era as well. We consider the possibility that an emphasis on judicial ideology considered in isolation may fail to account for episodic openings for political opportunism. We refer to the latter as “appointment era effects.” That said, we recognize that a presidential headwind can only do so much, and that its power depends on the President’s particular interests, how power is exercised, and the composition of the Senate.

The attitudinal approach is qualified here in another respect. We do not argue that legal precedent has no significance. Indeed, we have emphasized what we consider to be the potentially enduring impact of key cases following Brown. Milliken, in 1974, is one whose legal impact we argued was of special significance: we expect that this case had an impact that persisted over and above the Justices’ individual ideological dispositions, signaling a break from the perfectionist disposition of the Court that prevailed during the Warren era. In explaining the individual voting decisions, we thus include an indicator of whether the case was decided before or after the date of Milliken, expecting that decisions made after this date tended to go in a more conservative direction reflecting a new legal precedent.

Figure 1: Difference in the Number of Liberal and
Conservative Votes Cast Each Year by Ideology:  

Our perspective, then, is that the “attitudinal” versus “legal” approaches represent something of a false dichotomy.494 We assert that it is the interplay between the two that potentially offers a richer interpretation than that of a crude acontextual attitudinal model.

Before proceeding to the question of statistical inference, however, we first examine what the descriptive statistics suggest about the relationship between our key predictors of interest and judicial voting decisions.

B.  Nixon, Reagan, and the Role of Ideology


Our first variable of interest is the ideology of the Justices, operationalized in terms of the party affiliation of the appointing President. In line with the political science perspective that attaches great weight to the attitudinal theory of judicial behavior, ideology ought to have a strong impact on Justices’ individual voting decisions.495 Figure 2 breaks down voting direction by ideology.

Of the votes cast since Brown in 1954, 89% percent of votes cast by Justices appointed by Democratic presidents were in a liberal direction compared with 50% of votes cast by Justices appointed by Republican presidents. What remains to be investigated, however, is whether these differences in proportions reflect a phenomenon that has been relatively constant over time, or something that is better explained in terms of episodic success in tying voting outcomes to ideological or political party aspirations.


The first period of interest is the Nixon era, where two factors are noted. The first concerns Nixon’s own selection of Justices that potentially changed the political complexion of the previous liberal perfectionist Warren Court, what we call an “appointment era” effect. The second concerns new legal precedents established in this era that had potentially enduring impacts over and above that of changes in the ideological complexion of the Court, what we call a “legal precedent” effect. Each is considered in turn.

Figure 2: Individual Voting Direction Count Based on the
Party Affilation of the Appointing President

a.  The Nixon Appointees

We first consider the possibility of a Nixon “appointment era” effect pertaining to those Justices appointed by President Nixon. Figure 3 breaks down voting direction based on whether the Justice was appointed by President Nixon or by another President.

Consistent with the hypothesized appointment-era effect, a higher proportion of Nixon-era appointees voted in a conservative direction than those Justices appointed by Presidents other than Nixon. Whereas Justices appointed by Presidents other than Nixon cast 72% of their votes in a liberal direction, the pattern is reversed in the case of Nixon appointees, who cast only 39% of their votes in a liberal direction. Note here that the comparison is with non-Nixon appointees, which also includes other Republican Party appointees, including those appointed by President Reagan. As such, one would expect Nixon appointees to be even more strongly differentiated if we also controlled for Reagan appointees, an issue we explore in the inferential analysis to follow.

Figure 3: Individual Voting Direction Counts Based on
Whether or Not the Justice was Appointed by Nixon

b.  Milliken v. Bradley

We also consider a legal precedent effect associated with the Nixon era following the influential Milliken v. Bradley case, decided twenty years after Brown, and which we have argued erected obstacles in the path of the Court’s earlier perfectionist trajectory. Figure 4 breaks down voting direction based on whether a vote was cast before or after Milliken.496

Consistent with the hypothesized legal precedent effect, one sees a decisive shift in voting patterns in the direction of more conservative voting following Milliken relative to the prior period. Whereas 88% of the votes cast in the period from Brown to Milliken were in a liberal direction, only 44% of later votes were cast in a liberal direction.

Figure 4: Individual Voting Direction Count by Decision
Date Before and After Milliken


We next consider our other seminal event of interest: President Reagan’s inauguration and any effects on voting direction that may have ensued. Figure 5 compares the votes of those appointed by President Reagan with those appointed by other presidents.

Again consistent with a hypothesized appointment era effect, one observes a striking difference in the proportionate voting habits based on the appointing president. Whereas 70% of the votes cast by Justices appointed by Presidents other than Reagan were in a liberal direction, a mere 6% of the votes cast by Reagan appointees were in a liberal direction. As in the previous breakdown of Nixon era appointees, we are here comparing Reagan appointees against all other appointees, including other Republican Party appointees. As such, we would expect these differences to be accentuated further when controlling for other conservative appointees.

Figure 5: Individual Voting Direction Count Based on
Whether or Not the Justice was Appointed by Reagan


Here we pause to consider a possible objection to our operationalization of ideology. As discussed, we have coded ideology in terms of the party affiliation of the appointing President. One possible alternative operationalization would be in terms of the party affiliation of the Justice him or herself, in the cases where it is known. Of particular concern might be the case of Justice Brennan who, while personally a Democrat, was appointed by President Eisenhower, and is thus coded as a Republican according to our operational rule.497 This is even more problematic given that Justice Brennan is considered to be one of the last liberal perfectionists on the Court and, further, participated in more of our cases under consideration than any other Justice.498 However, in the empirical analyses to follow, it was found that even when recoding the ideology of Brennan from a Republican to a Democrat, we find no substantive difference in our results.

The distinction between our operationalization of ideology in terms of the party affiliation of the appointing President, and the ideology of the Justice, is nonetheless conceptually interesting. If one is concerned with the ability of political parties to influence future voting decisions, then it is the former effect that is of interest, while it is the latter that is arguably more relevant if one is primarily concerned with the role of ideology qua ideology.

However, we caution that recoding Justice Brennan would introduce a new ambiguity of its own. To be consistent, it might be insisted that we recode all Justices whose personal party affiliation differed from that of the appointing President. This would require, for example, that the Nixon appointee Justice Lewis F. Powell — a Southern Democrat — be recoded as well. Yet it could be credibly argued that the Justices Brennan and Powell cases are substantially different and require different treatment. Justice Powell was a Southern Democrat, which was not at that time incompatible with a conservative ideology, and thus he need not necessarily be recoded as a Democrat, if the interest is indeed in the effect of ideology qua ideology.499 Given this ambiguity, we return to our original operationalization of ideology in terms of the party affiliation of the appointing President.

C.  Estimating the Effect Sizes

We now proceed to the use of inferential statistical techniques to evaluate the effect sizes of our predictors. As the outcome is dichotomous (either conservative or liberal), one possible approach would have been to employ logistic regression.500 However, the dataset here also includes repeated observations of the same Justice across different decisions. This is illustrated in Figure 6, which shows the number of votes cast by the most frequently reoccurring Justices in the dataset.

In these circumstances, one would expect the observations pertaining to the same Justice to be more alike than observations pertaining to other Justices. However, a technique such as logistic regression would not take account of this correlated aspect of the data and, as such, is likely to yield invalid standard errors.501 An appropriate technique to address this issue involves the estimation of solutions to generalized estimating equations (GEEs).502 In the following analysis, we apply the GEE approach to three statistical models, starting first with an acontextual “baseline” model, followed by a series of models that integrate predictors pertaining to our specific arguments concerning the evolution of liberal perfectionism, Brown to the present.503 The GEE estimates for the three models are summarized in Table 1. These estimates are akin to those that would be obtained from a logistic regression. As such, they do not lend themselves to intuitive interpretation in comparison with, for example, estimates from ordinary least squares regression. They can, however, be used to compute estimated average probabilities (along with associated 95% confidence intervals). These are depicted graphically below in the discussion of each specific model.

Figure 6: A Count of the Number of Votes Cast by the Most Frequent Voters   

Table 1: Estimated GEE Coefficients and Associated Standard Errors for the Three Models


The first model could be regarded as a “crude” attitudinal model that pays attention to neither legal factors nor the kinds of episodic opportunities to reconfigure the political complexion of the Court which we earlier referred to as “appointment era” effects. As such, our baseline model has a single predictor: the ideology of the Justice. According to this baseline model, and consistent with our review of the descriptive statistics, ideology has a strong effect on voting decisions. As shown in Figure 7, the estimated probability of a Democratic appointee voting in a conservative direction is 0.11, while the estimated probability of a Republican appointee voting as such is 0.5.

Figure 7: Probability of a Conservative Vote by Ideology with 95% Confidence Intervals


We now augment the baseline model by adding one more predictor, which is an indicator of whether the case was decided before or after Milliken, which we have argued set an important legal precedent concerning remedial orders relating back to Brown. Figure 8 shows the estimated probability of voting in a conservative direction before and after Milliken while controlling for the effect of ideology.

Here the estimated probability of voting in a conservative direction is 0.14 prior to Milliken, and 0.51 thereafter. Additionally, while the probability difference in voting direction by ideology decreases from 0.39 to 0.31, the overall explanatory power of the model effectively doubles, based on commonly used criteria.504

Figure 8: Probability of a Conservative Vote pre-Milliken versus post-Milliken with 95% Confidence Intervals


Our final model examines whether the role of ideology can be attributed specifically to Nixon and Reagan era appointees. As such, we replace the original ideology predictor with two new predictors indicating whether the Justice in a particular voting outcome was appointed by President Nixon or President Reagan, respectively.

This final model offers a substantial improvement in explanatory power compared with the previous models.505 The three predictors have large independent effects (Figure 9). The largest such effect pertains to Reagan era appointees. The estimated probability of non-Reagan appointees voting in a conservative direction was 0.30 while the estimated probability of a Reagan appointee voting in a conservative direction was 0.88. Similarly, the estimated probability of a non-Nixon appointee voting in a conservative direction was 0.26 while that of a Nixon appointee was 0.60. We infer from these results that the effect of ideology, operationalized in terms of the party affiliation of the appointing President, is better explained in terms of our Nixon and Reagan era appointment effects.

Figure 9: Probability of a Conservative Vote by Key Predictors with 95% Confidence Intervals

The legal precedent set by Milliken continues to have its own independent effect over and above that of ideology. Whereas the estimated probability of a conservative voting decision was 0.20 pre-Milliken, the estimated probability of a conservative voting decision post- Milliken was 0.49.

In sum, we find substantial “appointment era” effects associated with the Nixon and Reagan eras, and a legal precedent effect associated with Milliken. Before the early 1970s, political ideology had no effect in the cases under consideration (Figure 1), but took on a more prominent role thereafter, as the legal cases delved into issues that appeared to be wider in scope than that articulated in Brown.

However, we find that the role of ideology is even better explained in terms of our “appointment era” effects. The largest such effect is that of the Reagan appointment era. This is somewhat surprising since two of his appointees, Justices O’Connor and Kennedy, are often depicted as moderates or “minimalists.”506 Reagan’s appointment of Justice O’Connor was in part an obligation to fulfill a campaign pledge to nominate a woman to the Supreme Court, while Justice Kennedy was only a backup choice following the unsuccessful nominations of Robert Bork and the apparently more conservative nominee, Douglas Ginsburg.507 Yet in terms of the voting decisions in the cases under our consideration, Justices O’Connor and Kennedy rank as among the most conservative of our Justices, voting in a conservative direction 90% and 94% of the time, respectively, albeit less conservatively than Reagan’s other appointee Justice Scalia, who voted in a conservative direction 100% of the time.

Similarly, while we found a Nixon era appointment effect, back- ground considerations suggest that this might have been even more substantial but for Nixon’s failure to anticipate the evolution of Justice Harry Blackmun,508 who voted in a conservative direction only 33% of the time, and his poor choices in the Haynsworth and Carswell nominations.509 In contrast, Nixon’s other appointees, Justices Burger, Powell, and Rehnquist voted in a conservative direction 55%, 77%, and 85% of the time, respectively.

We also note that Democratic Party appointees, not just Republican Party appointees, have exhibited more conservative voting habits since the early 1970s (Figure 1). Perhaps what we have been referring to throughout as “liberal perfectionist” was simply old school liberalism of the Warren Court, while what we now call liberal blurs a distinction between old school (liberal) “perfectionists” and contemporary (liberal) “minimalists.” Certainly the classic perfectionist vision of using the United States Constitution to redress past racial injustices has been markedly diminished, even if one still hears echoes of the perfectionist vision in the voices of Justices Breyer, Ginsburg,510 and Sotomayor.511

However, Justice Ginsburg is now eighty-one years old while Justice Breyer is seventy-seven. For this reason, their respective retirement plans have taken on some urgency in liberal legal circles.512 Had the Democrats retained control of the Senate in the recent congressional elections, President Obama would have been freer to replace either with like-minded Justices. As it is, he will now be constrained by a Republican-controlled Senate so that, in the event of either of the two retiring, any replacement would most likely be of the minimalist-centrist variety, which would shift the Court’s ideological center further to the right.513 Yet if Justices Ginsberg and Breyer choose to stay on, and the Republicans capture The White House in 2016, it is quite plausible that their subsequent retirements could lead to a Court with seven conservatives and only two liberals, Justices Sotomayor and Kagan. In this event, the much-heralded pivotal influence of Justice Kennedy to prevent the demise of affirmative action in cases like Fisher could become irrelevant, since only one more vote is currently needed for the Court to hold that educational diversity is not a compelling governmental interest. This would mean that affirmative action in colleges and universities — and other settings — would become unconstitutional as violating the Equal Protection Clause, a ruling that would effectively kill off the liberal perfectionist vision, once and for all.

The timing of Justice Kennedy’s retirement — he will be eighty years old in 2016 — could still be relevant, however. Although frequently depicted as a “minimalist” and “centrist,” our own analysis confirms that he has been a reliable conservative voter in the cases under consideration. As such, his own retirement would seem to be the Democrats’ only near-term opportunity to nudge the Court leftwards.514

How all this will unfold is impossible to predict. However, it may well turn out that future analysts will look back and observe brand new appointment era effects arising of the kind we demonstrated in our analyses of the Nixon and Reagan eras.

VII.  Conclusion

The following principles may be derived from the United States Supreme Court’s unanimous 1954 ruling in the landmark civil rights case Brown v. Board of Education: (1) The meaning of the Equal Protection Clause cannot be resolved on the ground of original intent; (2) equality of tangible services does not satisfy Equal Protection requirements in an environment infected by de jure segregation; (3) separating school children from others of similar age and qualifications solely because of their race is unconstitutional; (4) the “separate but equal” doctrine has no place in our constitutional order; and (5) intentional segregation is harmful to the hearts and minds of its victims.

In light of commentary lamenting the failure of the Supreme Court to fulfill Brown v. Board of Education’s promise, we traced legal developments at the Court in cases involving systemic racial discrimination claims in the post-Brown period, and investigated how these decisions came about. Among other things, we argued that the promise read into Brown by many has proven to be misplaced and relies on a “liberal perfectionist” vision of the United States Constitution, which goes beyond the strictly legal aspects of the Brown ruling.

We showed that after a period of fairly stringent enforcement of its desegregation decrees between 1954 and 1973, the Court, in its public education decisions rendered between 1974 and the present, (1) clarified standards for bringing systemic Equal Protection claims; (2) limited the scope of permissible remedial measures for proven systemic race discrimination; and (3) imposed stringent standards for benign K-12 racial balancing regimes and affirmative action admission efforts in colleges and universities. We argued that these results are consistent with the meaning of Brown.

After explaining these developments in jurisprudential terms, we examined them empirically, based on the attitudinal model of judicial voting developed in political science literature. In furtherance of this objective we studied the 552 votes rendered by the Supreme Court Justices for the period under review. Using statistical techniques that account for correlated voting patterns arising from repeated observations of the same Justice, we examined the effect of Justices’ ideologies, their appointment dates, and legal precedent on their liberal or conservative voting.

Arguing that the dichotomy between legal-attitudinal models of decision making at the Court is false and unsupportable, we showed (1) large appointment era effects associated with the Justices appointed by Richard Nixon and Ronald Reagan; (2) significant precedential effects attributable to the 1974 case of Milliken v. Bradley, which are independent of ideology as measured by party of the appointing President; and that (3) Democratic Party, as well as Republican Party, appointees have voted more conservatively since the early 1970s, but that before that time political ideology played no role in the voting in the cases under consideration.

Finally, we examined the current composition of the Court and con- sider possible effects of the mid-term 2014 Senate races and the 2016 presidential election on systemic racial discrimination cases based on which justices retire.

Appendix A: United States Supreme Court Decisions in Data Base

  1. Brown v. Bd. of Educ. (Brown I), 347 U.S. 483 (1954).
  2. Id. at 493, 495.
  4. See id. at 686 (“He wanted a single, unequivocating opinion that [left] no doubt that the Court had put Jim Crow to the sword.”); id. at 690 (discussing the Justices’ realization that they had to balance the firmness of their decision with its implementation so as to minimize arousing passions about such a controversial social policy as school desegregation); DAVID M. O’BRIEN, STORM CENTER 323 (Aaron Javsicas, ed., 9th ed. 2011) (“[T]he Warren Court waited a year before issuing, in Brown II [349 U.S. 294 (1955)], its mandate for ‘all deliberate speed’ in ending racial segregation in public education. Resistance to the social policy announced in Brown I was expected. A rigid timetable for desegregation would only intensify opposition.”).
  5. See generally KLUGER at 751-89 (presenting an overview of the fifty-year post-Brown era); Adam Liptak, On Supreme Court, Does 9-0 Add Up to More than 5-4?, N.Y. TIMES (August 11, 2014),
  7. We borrow the phrase “liberal perfectionists” from CASS R. SUNSTEIN, RADICALS IN ROBES: WHY EXTREME RIGHT-WING COURTS ARE WRONG FOR AMERICA (2005). Sunstein states that “[m]any American liberals are willing to . . . recognize or create new rights of many different kinds. When liberal perfectionists are committed, in principle, to a right, they often want the Supreme Court to say that that right is part of the Constitution.” Id. at xii. Sunstein is highly critical of perfectionism, whatever its nature. In the context of this article we define “liberal perfectionists” as those persons who believe that achieving racial integration is an overriding constitutional concern following the demise of the “separate but equal” doctrine in Brown. See, e.g., Ivan E. Bodensteiner, The Supreme Court as a Major Barrier to Racial Equality, 61 RUTGERS L. REV. 199, 203 (2009) (“[T]here is nothing in either the history of the Equal Protection Clause or the generally unfavorable pre-Brown decisions that precludes an interpretation of the Fourteenth Amendment that would promote actual equality, instead of only formal equality.”); see also Eric Schnapper, Affirmative Action and the Legislative History of the Fourteenth Amendment, 71 VA. L. REV. 753 (1985).
  8. See SUNSTEIN, supra note 7, at 32. See generally THOMAS M. KECK, THE MOST ACTIVIST SUPREME COURT IN HISTORY 38-103 (2004) (describing the expansion of rights-based jurisprudence during these periods).
  9. See SUNSTEIN, supra note 7, at 27-30. “Minimalism” suggests a “one step at a time” approach to decision making. Id. at 29. Minimalist judges seek to “nudge” the law along, rather than make sweeping “earthquake-like” changes. Id. at 30. They tend to write opinions narrowly to fashion them to the specific facts before the court, relying more upon precedent. Id. at 29. Sunstein places Justices Sandra Day O’Connor and liberal Ruth Bader Ginsburg in the minimalist category. Id. at 29-30. He places Justice Anthony Kennedy in this category as well. Id. at 31. Sunstein considers the minimalist approach the most sensible in that “it makes it possible for people to agree when agreement is necessary, and unnecessary for people to agree when agreement is impossible.” Id. at 28.
  10. See SUNSTEIN, supra note 7, at 53-77. Sunstein is sharply critical of members whom he refers to as “fundamentalist.” Id. at 25-27. These thinkers, according to Sunstein, believe that judges should not depart from the original understanding of those who ratified the Constitution in 1789, the Bill of Rights in 1791, or the Fourteenth Amendment in 1868. Id. at 26. He assigns Justices Scalia and Thomas to this category. Id. Although we do not necessarily subscribe to Sunstein’s policy views or legal conclusions, we find the categories he created to serve as useful metaphors in describing the Justices’ approaches to race infused issues and the Equal Protection Clause.
  11. See SUNSTEIN, supra note 7, at xii. Sunstein also uses a “majoritarian” label to describe Justices who defer to democratic outcomes, unless they are plainly unconstitutional. Id. Justices characterized in this way exercise much restraint in exercising the power of judicial review. Id. In respect of the focus of this article, majoritarians would permit affirmative action programs and allow states to forbid same-sex marriage. Sunstein characterizes Oliver Wendell Holmes as a majoritarian, asserting that he was perhaps the greatest figure in American legal history. Id. Sunstein does not assign this label to any Justice now on the Court. Throughout the article we use Sunstein’s categories to help illustrate approaches taken by the Court in its decision making, recognizing that no Justice will fit into one category all of the time; to a degree, they are a shorthand for a Justice’s general approach to solving complex constitutional questions.
  12. The Court has emphasized, for example, the distinction between de facto and de jure segregation, spurring much debate about the proper role of the Equal Protection Clause in assuring equality. For different perspectives on these constitutional values compare Frank Goodman, De Facto School Segregation: A Constitutional and Empirical Analysis, 60 CALIF. L. REV. 275, 319 (1972), with David Strauss, Discriminatory Intent and the Taming of Brown, 56 U. CHI. L. REV. 935, 962 (1989).
  13. See infra notes 311-50 and accompanying text (race-based voluntary desegregation policies); infra notes 351-458 and accompanying text (affirmative action college and university admission programs).
  14. See infra notes 440-58 and accompanying text.
  15. See infra notes 484-515 and accompanying text.
  16. This corresponds to appointments made between 1968 and 1971, and 1981 and 1988, respectively.
  17. 418 U.S. 717 (1974). See infra notes 210-38 and accompanying text (examining Milliken’s facts and the legal principles for which it stands).
  18. See infra notes 210-38 and accompanying text.
  19. Brown I, 347 U.S. 483 (1954).
  20. These were Alabama, Arkansas, Delaware, Florida, Mississippi, Missouri, North Carolina, Oklahoma, Georgia, Kentucky, Louisiana, Maryland, South Carolina, Tennessee, Texas, Virginia, and West Virginia. Luther A. Huston, High Court Bans School Segregation; 9-to-0 Decision Grants Time to Comply, N.Y. TIMES, May 17, 1954, at A1, available at
  21. These were Kansas, New Mexico, Arizona, and Wyoming. Wyoming, however, had never exercised this authority. Id.
  22. Id.
  23. The District of Columbia’s schools had been “segregated since the Civil War days under laws passed by Congress.” Id.
  24. Briggs v. Elliot, 342 U.S. 350, decided with Brown v. Bd. of Educ. (Brown II), 349 U.S. 294 (1955).
  25. KLUGER, supra note 3, at 331-32.
  26. See KECK, supra note 8, at 47-48 (describing Justice Felix Frankfurter’s successful and persistent effort in urging his fellow Justices to delay action in Brown I after argument during the 1952 Term, in the hopes of keeping the decision narrow and unanimous).
  27. See KLUGER, supra note 3, at 617-19 (describing the questions to be answered the following term); see also KECK supra note 8, at 48 (noting it was Justice Frankfurter and his law clerk Alexander Bickel who drafted the five questions for reargument the following term).
  28. KLUGER, supra note 3, at 696-99.
  29. See Brown I, 347 U.S. at 493-95.
  30. Id. at 486.
  31. Id. at 492-93.
  32. Id. at 490-92. But see Michael McConnell, Originalism and the Desegregation Decisions, 81 VA. L. REV. 947, 1140 (1995) (concluding that school segregation was understood during Reconstruction to violate principles of equality of the Fourteenth Amendment).
  33. Brown I, 347 U.S. at 492.
  34. Id. at 492 n.9.
  35. Id. at 493.
  36. Id. at 494.
  37. Id.
  38. See id. at 494 n.11.
  39. Brown I, 347 U.S. at 495.
  40. Id. at 495-96.
  41. Brown v. Bd. of Educ. (Brown II), 349 U.S. 294 (1955).
  42. Id. at 301.
  43. Mark G. Yudof, School Desegregation: Legal Realism, Reasoned Elaration, and Social Science Research in the Supreme Court, 42 LAW & CONTEMP. PROBS. 57, 70 (1978) (finding consensus among experts that the proffered evidence was methodologically unsound and concluding the damage of the dual school system, and the systematic treatment of blacks as inferior beings, is a historical not an empirical truth).
  44. Id. at 69.
  45. See id. at 69-73.
  46. Id. at 57-58. The connection between Brown’s moral aspects and Equal Protection is widely recognized. See, e.g., J. HARVIE WILKINSON, III, SERVING JUSTICE: A SUPREME COURT CLERK’S VIEW 130-33 (1974); Charles Black, The Lawfulness of the Segregation Decisions, 69 YALE L.J. 421, 427 (1960).
  47. Brown I, 347 U.S. at 487-88.
  48. See generally Michael Abramowicz & Maxwell Stearns, Defining Dicta, 57 STAN. L. REV. 953 (2005) (defining the difference between holding and dicta); Shawn J. Bayern, Case Interpretation, 36 FLA. ST. U. L. REV. 125 (2009) (determining case meaning outside of the dichotomy of holding vs. dicta); Michael C. Dorf, Dicta and Article III, 142 U. PA. L. REV. 1997 (1994) (focusing on Article III as a method of distinguishing holding from dicta); Pierre N. Leval, Judging under the Constitution: Dicta about Dicta, 81 N.Y.U. L. REV. 1249 (2006) (arguing that treating dicta as binding law is nonsensical); Judith M. Stinson, Why Dicta Becomes Holding and Why it Matters, 76 BROOK. L. REV. 219 (2010) (identifying the causes of the conflation of dicta and holding).
  49. But see McConnell, supra note 32, at 1140 (reaching the opposite conclusion).
  50. Brown I, 347 U.S. at 494-95.
  51. 349 U.S. at 301.
  52. See id. at 300-01 (recognizing difficulties in implementing desegregation and directing lower courts to keep jurisdiction to ensure compliance with Brown I’s commands).
  53. Perhaps the strongest formal statement made in support of segregation was a “Declaration of Constitutional Principles,” sometimes referred to as the Southern Manifesto, endorsed by ninety-six congressmen. See 102 CONG. REC. 4461-4464 (daily ed. Mar. 12, 1956) (statement of Sen. Strom Thurmond during 84th Cong) (vigorously attacking Brown). See also Alvin Schuster, 96 in Congress Open Drive to Upset Integration Ruling, N.Y. TIMES, March 12, 1956, at A1.
  54. For example, was only purposeful discrimination covered? See, e.g., Washington v. Davis, 426 U.S. 229, 239-242 (1976). Or could mere discriminatory effects be sufficient to establish an Equal Protection violation? See, e.g., Yick Wo v. Hopkins, 118 U.S. 356 (1886).
  55. It was obvious that simply enjoining de jure segregation would not address most of the problems. The range of issues was enormous. For example, how would the courts address faculty and staff assignments, transportation, extra-curricular activities, facilities and student assignments? See generally, Green v. Cnty. Sch. Bd., 391 U.S. 430, 435 (1968); Bd. of Educ. v. Dowell, 498 U.S. 237 (1991).
  56. For instance, once a school district came under federal court supervision for its Equal Protection violations, what rules would apply in determining whether it later achieved unity, that is, when it did all it could do to fix the prior segregation caused by official action? See, e.g., Freeman v. Pitts, 503 U.S. 467, 490-91 (1992).
  57. 358 U.S. 1 (1958).
  58. Ultimately, this resulted in President Eisenhower federalizing the Arkansas National Guard and using regular federal forces to remove obstructions to school integration in Little Rock. See Anthony Lewis, Eisenhower on Air, N.Y. TIMES, September 24, 1957, at A1.
  59. Cooper, 358 U.S. at 4.
  60. Id. at 18.
  61. Id. at 17.
  62. Id. at 17 (quoting Smith v. Texas, 311 U.S. 128, 132 (1940)).
  63. 373 U.S. 683 (1963).
  64. See id. at 684-85, 688.
  65. See id. at 686.
  66. Id.
  67. 377 U.S. 218 (1964).
  68. See id. at 220-23, 230-31.
  69. See id. at 223-25.
  70. See id.
  71. See id. at 221-25.
  72. Id. at 225, 231-32.
  73. Id. at 229.
  74. Id. at 231.
  75. 391 U.S. 430 (1968). The Supreme Court had previously held a state-wide system initially established and maintained under Virginia constitutional and statutory provisions was unconstitutional. See Davis v. Cnty. Sch. Bd., 103 F. Supp. 337 (1952), decided with Brown I.
  76. See Green, 391 U.S. at 431-33.
  77. Id. at 430. In Green, the county school board did not try to desegregate its schools until 1965, when, after the law suit was filed, it adopted the freedom-of-choice plan. Id. at 433. The plan was adopted in order to remain eligible for federal financial aid. Id.
  78. Id. at 432 (approximately 740 students were black and 550 were white).
  79. Id. (emphasis added).
  80. Id.
  81. Id. at 433-34.
  82. Id.
  83. Id.
  84. See id. at 441.
  85. Id. at 438-39. The immediacy of the requirement of schools to integrate took on even greater urgency as time progressed. In Alexander v. Holmes Cnty. Bd. of Educ., 396 U.S. 19 (1969) and Carter v. W. Feliciana Parish Sch. Bd., 396 U.S. 290 (1970), the Court held that even a few months delay in desegregation to avoid disruption during the school year was impermissible and it peremptorily ordered immediate desegregation.
  86. Green, 391 U.S at 437-38. In Raney v. Board of Education, springing from a setting similar to Green, the Court invalidated another freedom-of-choice plan. See Raney v. Bd. of Educ., 391 U.S. 443 (1968). There, during the plan’s first year of operation, “[t]he number of [students] applying for enrollment in the fifth, tenth, and eleventh grades at [the white school] exceeded the number of places available;” the applications of twenty-eight black students were denied. Id. at 446. The excluded students sought injunctive relief against their required attendance at the all-black school; the provision of inferior school facilities in the all-black school; and the district’s continued operation of a racially segregated school system. Id. at 446. The Court held that in light of the district’s long history of racially segregated schools, the freedom-of-choice plan was insufficient to convert it into a unitary nonracial school system. Id. at 447.
  87. Green, 391 U.S. at 437 n.4.
  88. Id. at 435, 440-41.
  89. Id. at 442.
  90. Id. at 435. Although Green might be read to suggest these six factors apply in determining systemic race-based Equal Protection violations, Green’s procedural posture as an enforcement proceeding suggests otherwise. Taking Justice Brennan at his word that the case was about compliance with Brown, id. at 431-32, the Court’s holding is about how the Green factors apply to remedial orders, see id. at 441-42. Although these factors are obviously relevant to establishing a purposeful violation with discriminatory effects, that is not what Green is about.
  91. See id. at 441-42.
  92. See id.
  93. Id. at 431-32 (quoting Brown II, 349 U.S. at 300-01).
  94. Id. at 433-34. The Green Court emphasized its goal was to eliminate racially identifiable schools. Yet, the remedy it imposed necessarily involved racial identity. Since this was an enforcement action implementing a prior desegregation order, the onus would be on the school district to comply with the order. This might explain the Court’s comment that the district unlawfully relied on the parents to dismantle the dual system when it was the district’s responsibility to fulfill. Id. at 441-42.
  95. See id. The Court does see that there might be other reasons that make school choice plans illegitimate. These could include, relative to their attendance at formerly white schools: fear of retaliation and hostility from white parents; known incidents of violence visited upon black children in other places; improper influence of white officials discouraging black children from making a transfer; and incidents of poverty, such as inadequate attire leading to embarrassment, and perhaps fear of academic competition, among other things. See id. at 441 n.5.
  96. See id. at 441-42.
  97. See id. at 437 (distinguishing between the requirement for universal and compulsory integration and a racially nondiscriminatory system based on “established unconstitutional deficiencies”).
  98. 402 U.S. 1 (1971).
  99. Id. at 15.
  100. See id. at 20-22, 32.
  101. See id. at 6-8.
  102. See id. at 6-11.
  103. See id. at 21-26 (describing the use of racial ratios as a permissible initial guide to desegregation, but indicating they need not be rigidly applied).
  104. Id. at 22-31.
  105. Id. at 6. Charlotte-Mecklenburg encompassed the city of Charlotte and surrounding Mecklenburg County. Together it included an area of 550 square miles spanning about 22 miles east-west and 36 miles north-south. Id. During 1968-69 it served more than 84,000 pupils in 107 schools. About 71% of the students were white and about 29% black. Id. at 6-7.
  106. The Court stated this “does not mean that every school in every community must always reflect the racial composition of the school system as a whole” and, in fact, “some small number of one-race, or virtually one-race, schools within a district” may be unavoidable. Id. at 24, 26. However, the Court cautioned that such a result always should receive close judicial review in a school system that was once segregated by law. Id.
  107. Swann, 402 U.S. at 22-31. The Court said that bus transportation is an important “tool of school desegregation” and is a constitutionally acceptable remedy unless “the time or distance of travel is so great as to either risk the health of the children or significantly impinge on the educational process.” Id. at 30-31.
  108. Id. at 28.
  109. Id. at 31-32. This language narrowing the scope of the remedies against systems where de jure segregation was practiced was lost neither on school districts subject to desegregation orders nor on Justices required to vote on systemic race discrimination claims aimed at dismantling segregated schools. See infra notes 459-83 and accompanying text.
  110. See N.C. State Bd. of Educ. v. Swann, 402 U.S. 43, 44-45 (1971).
  111. See id. at 45-46.
  112. Id. at 46.
  113. Id. at 45-46.
  114. Swann v. Charlotte-Mecklenburg, 402 U.S. at 18.
  115. See id. at 15-31.
  116. Id. at 26.
  117. Id.
  118. See supra notes 86-94 and accompanying text.
  119. Swann is distinguishable from most of the cases we review since it encompassed urban and suburban communities thereby giving many opportunities to impose desegregation remedies which could be effective See James E. Ryan, The Real Lessons of School Desegregation, in FROM SCHOOLHOUSE TO COURTHOUSE: THE JUDICIARY’S ROLE IN AMERICAN EDUCATION 73, 83-84 ( Joshua M. Dunn & Martin R. West, eds., 2009) [hereinafter The Real Lessons] (arguing that Swann belies the claims that busing may never be an effective remedy for intentional segregation).
  120. 413 U.S. 189 (1973).
  121. Id. at 191-92.
  122. Id. at 192.
  123. See id. at 192-93.
  124. 163 U.S. 537 (1896).
  125. Keyes, 413 U.S. at 193-94.
  126. Id. at 194-95.
  127. See id. at 201-03. Justice Byron White took no part in the decision.
  128. Id. at 204.
  129. See id. at 207-15.
  130. Id. at 201-02 (footnote and citation omitted).
  131. Id. at 208 (emphasis added).
  132. See id. at 208-09.
  133. Id. at 213.
  134. Id. at 257 (Rehnquist, J., dissenting) (internal citation omitted).
  135. 413 U.S. 455 (1973).
  136. See id. at 468-69.
  137. See id. at 464-66.
  138. Id. at 466.
  139. Since 1940 the State had been buying textbooks and lending them free to students in both public and private schools, without reference to whether any participatingprivate school had racially discriminatory policies. Once the integration era began, the number of private, nonsectarian schools had increased from 17 in the 1963-64 school year (enrolling 1446 white students), to 155 in the 1970-71 school year (enrolling about 42,000 white students). See id. at 457. The creation and enlargement of these private schools was in direct response to the desegregation of public schools. Thousands of students who were attending private, all-white schools received free textbooks. See id. at 457-58, 459 n. 4. Although 90% of the state’s school children still attended public schools, some school districts had lost all of their white students to private, segregated schools. Notably, there was no proof that absent the free textbooks, any children would withdraw from segregated private schools and enroll in unitary, public schools. Despite this lack of evidence, the Court concluded that the State may not induce, encourage, or promote private persons to accomplish what it may not constitutionally accomplish itself. The Court said that the state provision of free texts may not be essential to the continued operation of private segregated schools, but it does constitute substantial state support of discrimination; therefore, the Fourteenth Amendment prohibits it. Id. at 463-70 (emphasis added). The Court remanded the case for a determination as to which private schools might be entitled to receive state support under the statute. Id. at 471.
  140. See supra notes 19-42 and accompanying text. See also ALEXANDER BICKEL, THE SUPREME COURT AND THE IDEA OF PROGRESS 119-120 (1st ed. 1970).
  141. See generally LEE EPSTEIN & THOMAS G. WALKER, CONSTITUTIONAL LAW FOR A CHANGING AMERICA 785-91 (8th ed. 2013) (listing membership in each Natural Court since the Roosevelt appointments). In 1975, President Ford nominated Justice John Paul Stevens to replace Justice William O. Douglas, id., one of the Justices we earlier characterized as liberal perfectionist. There were probably few commentators at the time of Stevens’ appointment who would have predicted how liberal a jurist he would turn out to be.
  142. 411 U.S. 1 (1973).
  143. 418 U.S. 717 (1974).
  144. 498 U.S. 237 (1991).
  145. 503 U.S. 467 (1992).
  146. 515 U.S. 70 (1995).
  147. 411 U.S. 1 (1973).
  148. See generally Susan H. Bitensky, We “Had a Dream” in Brown v. Board of Education. . ., 196 DET. C.L. MICH. ST. U. L. REV. 1, 15 (1996) (“[I]t is a composite of factors that have turned Brown’s promise more into words of aspiration than commitment. But Rodriguez, without fanfare, and under cover of Milliken . . . has figured into the process.”).
  149. See Rodriguez, 411 U.S. 1, 18, 28-29, 33-35.
  150. See id. at 7-17 (describing the history and then current sources of revenue for Texas school districts).
  151. Id. at 4-5, 16-17.
  152. Id. at 16.
  153. See id. at 36-40. See generally id. at 33-40 (explaining the reasoning for the decision). The Rodriquez opinion was written by Nixon appointee Justice Lewis F. Powell. Justices Burger, Blackmun, and Rehnquist joined in the opinion. Justice Stewart, an Eisenhower appointee, concurred and filed a separate opinion. Justices Brennan, White, Douglas, and Marshall dissented.
  154. Id. at 44.
  155. Id. at 24.
  156. See John Dinan, School Finance Litigation: The Third Wave Recede, in FROM SCHOOLHOUSE TO COURTHOUSE, supra note 119, at 97 (asserting that Rodriquez brought a halt to federal constitutional challenges to state education finance systems).
  157. Id.
  158. Id.
  159. For a decision holding out the possibility of a federal Due Process or Equal Protection claim based on education falling below state adequacy standards, see Papasan v. Allain, 478 U.S. 265 (1986) (examining a claim that petitioners had been denied their right to a minimally adequate education, but also that such a right was fundamental and that because the right had been infringed, the State’s action should be reviewed under strict scrutiny).
  160. Recent articles examine such claims. See, e.g., Carlee Poston Escue et al., Some Perspectives on Recent School Finance Litigation, 268 EDUC. L. REP. 601 (2011); Patrick A. O’Reilly, State Constitutional Law-Education Funding-Evidence of a Correlation between Poor School Funding and Poor Student Performance Must Be Presented to Render a School Funding System Unconstitutional: Davis v. State, 804 N.W.2d 618 (S.D. 2011), 43 RUTGERS L.J. 835 (2013); R. Craig Wood & William E. Thro, Originalism and the State Education Clauses: The Louisiana Voucher Case as an Illustration, 302 EDUC. L. REP. 875 (2014).
  161. See, e.g., Yohance C. Edwards & Jennifer Ahern, Unequal Treatment in State Supreme Courts: Minority and City Schools in Education Finance Litigation, 79 N.Y.U. L. REV. 326, 346-53 (2004). There, the authors assert that the predominant race of the plaintiff school districts, school district setting (urban-other), and the number of districts involved in the litigation, rather than traditional “legal” factors—constitutional language, wave of litigation (equity or adequacy era), wealth gap between schools in the state, percentage of local revenue from local sources, per-pupil spending, average teacher salaries, judicial selection method, liberalism in the state, percentage of the state population that is urban, percentage of the state population that is minority, median household income in the state, and political culture in the state—are better predictors of the outcome of education finance litigation. Moreover, the authors see that urban minority districts fare especially poorly in education finance litigation. Id.
  162. See R. Craig Wood, A Critique of the Federal Challenge to Financing Public Education along Racial Lines in Lynch v. Alabama: How the Plaintiffs, the Defendants, and the Federal District Court Erred in Examining the Funding of Public Education in Alabama, 1 EDUC. LAW AND POL. REV. 123, 123-71 (2014).
  163. 426 U.S. 229 (1976).
  164. 429 U.S. 252 (1977).
  165. Davis, 426 U.S. 229 (1976).
  166. Id. at 239-42. Chief Justice Burger and Justices Stewart, White, Blackmun, Powell, Rehnquist and Stevens formed the majority. Only Brennan and Marshall dissented.
  167. Id. at 239, 242 (internal citation omitted).
  168. Id. at 248-50. Although in Davis Justice Stevens concurred in the result, he saw “the line between discriminatory purpose and discriminatory impact is not nearly as bright, and perhaps not quite as critical, as the reader of the Court’s opinion might assume.” Id. at 254. He asserted that not every disproportionate impact gives rise to a constitutional claim, but that it may do so in situations where that impact is dramatic. In Davis, he considered the challenge insufficient because the test in issue merely required the applicants to meet a minimum standard of literacy and the test was used throughout the federal service. Id. at 254-55.
  169. Pers. Adm’r. of Mass. v. Feeney, 442 U.S. 256, 258 (1979). In Feeney, the plaintiffs alleged that the credit given to honorably discharged military veterans disadvantaged females in the application process. Justice Stevens voted with the majority because the number of males disadvantaged by the law was sufficiently large and close to the number of disadvantaged females to refute the claim that the move was intended to benefit males as a class or harm females as a class. Id. at 281. In a similar vein, Justice Stewart said that the law placed many men who were not veterans at a disadvantage as well, and that the distinction made in the law was between veterans and nonveterans, not between men and women. Id. at 257.
  170. Id. at 279.
  171. 429 U.S. 252 (1977).
  172. See id. at 255-57.
  173. See id. at 257-58.
  174. Id. at 254.
  175. Id. at 254-55.
  176. See id. at 268-71. Justice Powell’s five-to-three majority opinion was joined by Chief Justice Burger and Justices Stewart, Blackmun, and Rehnquist. Justices Brennan, White and Marshall dissented.
  177. See Arlington Heights, 429 U.S. at 269-70.
  178. See id. at 265-68.
  179. Id. at 265.
  180. Id. at 265.
  181. Id. at 265-66.
  182. Id. at 264-65. The Arlington Heights Court said: “The impact of the official action—whether it ‘bears more heavily on one race than another’—may provide an important starting point. Sometimes a clear pattern, unexplainable on grounds other than race, emerges from the effects of the state action even when the governing legislation appears neutral on its face." Id. at 266 (citations omitted). One way of proving a discriminatory purpose is to show a statistical pattern that can be explained only by a discriminatory purpose. See, for example, Gomillion v. Lightfoot, 364 U.S. 339 (1960), where black voters challenged the government’s redrawing of boundaries to exclude them from participating in the election. There, Tuskegee, Alabama, was transformed from a square shape into a 28-sided figure. Id. at 339. All but four or five of 400 blacks in the City were placed outside its boundaries, but no whites were excluded. Id. The court said that the “conclusion would be irresistible, tantamount for all practical purposes to a mathematical demonstration, that the legislature is solely concerned with segregating white and colored voters by fencing Negro citizens out of town so as to deprive them of their pre-existing municipal vote.” Id. at 341. See also, Castenada v. Partida, 430 U.S. 482 (1977). There, 79% of the county’s population was Spanish-surname, but only 39% of grand jurors were Spanish-surname from 1960 to 1972. This was sufficient to establish a prima facie case of discrimination and shift the burden to the government to offer a race-neutral explanation for these statistics. See id. at 492-501 (explaining the use of statistical evidence in support of an actionable Equal Protection claim).
  183. Arlington Heights, 429 U.S. at 267. “The Court said the historical background of the decision is one evidentiary source, particularly if it reveals a series of official actions taken for invidious purposes . . . . The specific sequence of events leading up to the challenged decision also may shed some light on the decision-maker’s purposes.” Id. The Court illustrated this point by using Guinn v. United States as an exemplar. Id. at 266 (citing Guinn v. U.S. 238 U.S. 347 (1915)). In Guinn, the Court declared unconstitutional an Oklahoma law that required a literacy test as a condition of voting, but in effect excluded white citizens from this requirement through a “grandfather” clause. See Guinn 238 U.S. at 367-68. Although the law expressly prohibited discrimination based on race and voting, it also created a literacy test that had an exemption for those eligible to vote in 1866 and their descendants. Id. at 358. The obvious purpose and clear effect was to disenfranchise blacks.
  184. Arlington Heights, 429 U.S. at 268.
  185. For later Supreme Court decisions vacated in light of Arlington Heights, see Metro. Sch. Dist. v. Buckley, 429 U.S. 1068 (1977) and Sch. Dist.v. U.S., 433 U.S. 667 (1977).
  186. Arlington Heights, 429 U.S. at 270 n. 21.
  187. 471 U.S. 222 (1985). Associate Justice Lewis Powell did not participate in the decision.
  188. Id. at 223.
  189. Id. at 225.
  190. Id. at 228.
  191. See id. at 227, 232-33.
  192. Id. at 233.
  193. See supra notes 165-70 and accompanying text (discussing Washington v. Davis); supra notes 171-86 and accompanying text (discussing Arlington Heights v. Metro. Hous. Corp.); supra notes 187-92 and accompanying text (discussing Hunter v. Underwood).
  194. Hunter, 471 U.S. at 228; see also Mt. Healthy City Sch. Dist. Bd. of Educ. v. Doyle, 429 U.S. 274, 281-87 (1977) (applying burden shifting in context of First Amendment Free Speech dispute).
  195. See ERWIN CHEMERINSKY, CONSTITUTIONAL LAW: PRINCIPLES AND POLICIES, ASPEN STUDENT TREATISE SERIES 717-18 (4th ed. 2013) (explaining effects of government satisfying its burden).
  196. See id. at 746-47.
  197. Id. This in essence represents the “but-for” aspect of Equal Protection as enunciated by the Court.
  198. Milliken v. Bradley, 418 U.S. 717 (1974).
  199. 417 U.S. 556 (1974).
  200. See id. at 566-67.
  201. Id. at 557 (emphasis added).
  202. Id.
  203. See id. at 568-69.
  204. See id. at 569-74.
  205. Id. at 570.
  206. See id. at 556-58.
  207. McLaurin v. Okla State Regents, 339 U.S. 637 (1950).
  208. See Gilmore, 417 U.S. at 559-61.
  209. Since the First Amendment protects the association rights of private groups to discriminate, where the exclusionary practices are integral to expressive activity, the core factual question in Gilmore was whether the city as a matter of practice granted exclusive access to the parks for expressive activity. If so, the remand to the lower courts might have been unnecessary. See, e.g., Hurley v. Irish-American Gay, Lesbian and Bisexual Groups of Boston, 515 U.S. 557 (1995) (holding that a private group that organized a St. Patrick’s Day parade could exclude a gay, lesbian, and bisexual groups from participation; the unanimous decision said that organizing a parade is an inherently expressive activity and those doing so have a right to exclude messages inimical to their own); see also Kenneth Karst, The Freedom of Intimate Association, 89 YALE L. J. 624 (1980) (extensively exploring the constitutional values of freedom of association).
  210. 418 U.S. 717 (1974).
  211. See id. at 723-26.
  212. See id. at 724-27.
  213. See Bradley v. Milliken, 338 F. Supp. 582, 585-87 (E.D. Mich. 1971).
  214. See Bradley v. Milliken, 484 F.2d 215, 221-38, 239-41 (6th Cir. 1973) (en banc), aff ’g. 338 F. Supp. 582, 585-87 (E.D. Mich. 1971).
  215. See id. at 241-49.
  216. See Milliken, 418 U.S. at 741-45.
  217. Id. at 744-45.
  218. Id. at 745.
  219. Id.
  220. See id. at 745-47.
  221. See id.
  222. See Milliken, 418 U.S. at 781.
  223. Id. at 750.
  224. See supra notes 98-119 and accompanying text.
  225. See supra notes 120-39 and accompanying text.
  226. See Milliken, 418 U.S. at 742-43, 750-52; see also supra notes 105-19 and accompanying text.
  227. See Milliken, 418 U.S. at 745.
  228. See id. at 745-47.
  229. See, e.g., James E. Ryan, The Supreme Court and Voluntary Integration, 121 HARV. L. REV. 131, 140 (2007) (observing that with the de facto-de jure distinction in tact plaintiffs had to show that local or state officials took actions or devised policies with the intent of segregating, not just schools, but districts; “such proof was hard to come by, in part because housing discrimination kept most African-Americans out of the suburbs.”).
  230. See id.
  231. See id.
  232. See The Real Lessons, supra note 119, at 83 (“The causes of residential segregation are numerous and intertwined and include such factors as economics, preferences, and private discrimination.”).
  233. Justice Stewart recognized this issue. In Milliken he wrote cogently on this point. He said that if state officials conspired in purposeful segregation by enacting racially discriminatory housing or zoning laws, then enforcement through interdistrict busing or redrawing district boundaries might be appropriate. Milliken, 418 U.S. at 755 (Stewart, J., concurring). Although in Milliken he obviously considered this argument, he dismissed it based on his reading of the evidence, and thus became the fifth vote in the majority.
  234. See James E. Ryan, Schools, Race, and Money, 109 YALE L. J. 249, 275-84 (1999).
  235. See supra notes 135-40 and accompanying discussion.
  236. See, Ivan E. Bodensteiner, The Supreme Court as the Major Barrier to Racial Equality, 61 RUTGERS L. REV. 199, 209 (2009) (arguing that Milliken undermined Brown’s promise that “the doctrine of ‘separate but equal’ has no place” in the “field of public education” and “[s]eparate educational facilities are inherently unequal.” Brown I, 347 U.S. at 495). The same argument would apply to the Court’s Gilmore analysis, at least in terms of segregative effects.
  237. See supra notes 199-206 and accompanying text.
  238. See supra notes 199-206 and accompanying text.
  239. 427 U.S. 424 (1976).
  240. Id. at 424.
  241. Id.
  242. Id.
  243. Id. at 436.
  244. Id. at 431.
  245. Spangler, 427 U.S. at 436. Justices Burger, Stewart, White, Blackmun, and Powell formed the Spangler majority. Justices Brennan, Marshall and Stevens dissented.
  246. Id. at 434.
  247. See id. at 435-36.
  248. Id. at 436-37.
  249. Id.
  250. Id. at 434. See also supra notes 140-51 and accompanying text.
  251. See supra notes 130-39 and accompanying text.
  252. 498 U.S. 237 (1990).
  253. Id. at 240.
  254. Id. at 241.
  255. See id. at 241-42.
  256. See id. at 242-43.
  257. See id. at 242-43.
  258. Id. at 243-44.
  259. Id. at 243.
  260. Id. at 244.
  261. Id. at 247.
  262. See id. at 248-49.
  263. Id. at 246.
  264. Id. at 248.
  265. See id. at 250.
  266. See id. at 249-50.
  267. Id. at 245, 250.
  268. Id. at 249-50. Plainly, Dowell places substantial procedural barriers on putative plaintiffs, since they would be required to establish a new violation independent of the old one, despite a history of intentional segregation in the jurisdiction.
  269. 498 U.S. at 244 (holding that where a district court order did not formally dissolve a desegregation decree it remains in effect, even if the district court finds that the school system achieved unitary status).
  270. Ross v. Houston Indep. Sch. Dist., 699 F.2d 218, 225 (5th Cir. 1983) (citing Swann, 402 U.S. at 15).
  271. Green, 391 U.S. at 439 (citing Raney v. Bd. of Educ., 391 U.S. 443, 449 (1968)).
  272. Freeman v. Pitts, 503 U.S. 467, 490 (1992).
  273. See Green, 391 U.S. at 435; supra notes 86-90 and accompanying text.
  274. 503 U.S. 467 (1992).
  275. Id. at 471, 485.
  276. The 1969 plan abolished a prior “freedom of choice” scheme and adopted a neighborhood school attendance plan that had been proposed by DCSS and accepted by the Department of Health, Education and Welfare subject to minor modifications and incorporated into a consent decree. Under the plan all of the former black schools were closed and their students were reassigned among the remaining neighborhood schools and the district court retained jurisdiction. Id. at 473.
  277. Id. at 471.
  278. See id. at 481-485.
  279. See id.
  280. Id. at 489-90.
  281. Id.
  282. Id. at 493. Despite the unanimity in the decision, there is a counter-argument to be made. Under the Pitts doctrine DCSS was entitled to relief from the pupil assignment and facilities provisions of the prior order, but not the teacher assignment mandate. Since DCSS had complied with the facilities part of the Green criteria in the prior decree, it was entitled to be relieved of that portion of the order as well. One difficulty with this piecemeal approach was a then current controversy over a new facilities plan, which the plaintiffs contended ran afoul of Equal Protection strictures. See id. at 492-99. For the plaintiffs this meant commencement of more proceedings, which in all likelihood would take years to complete. Moreover, piecemeal adjudication may not have as many virtues as the Court contended. In a very real way school districts operate as hydraulic systems with one part necessarily affecting the other. By not insisting on satisfaction of all the Green factors at once time, the Pitts approach could arguably result in exacerbation of racial segregation in the schools.
  283. See supra notes 252-69 and accompanying text.
  284. Pitts, 503 U.S. at 495 (“Where resegregation is a product not of state action but of private choices, it does not have constitutional implications. It is beyond the authority and beyond the practical ability of the federal courts to try to counteract these kinds of continuous and massive demographic shifts. To attempt such results would require ongoing and never-ending supervision by the courts of school districts simply because they were once de jure segregated. Residential housing choices, and their attendant effects on the racial composition of schools, present an ever-changing pattern, one difficult to address through judicial remedies.”).
  285. 515 U.S. 70 (1995).
  286. Id. Voting with the Chief Justice were Justices O’Connor, Scalia, Kennedy and Thomas. Dissenting were Justices Stevens, Souter, Ginsburg, and Breyer.
  287. See id. at 74.
  288. Id. at 115 (Thomas, J., concurring).
  289. CHEMERINSKY, supra note 195, at 748.
  290. Id. (describing effects of prior decisions on integration in Kansas City).
  291. Jenkins, 515 U.S. at 73, 80.
  292. Id. at 81.
  293. Id. at 80-81.
  294. 503 U.S. at 491.
  295. See Jenkins, 515 U.S. at 81-82.
  296. See CHEMERINSKY, supra note 195, at 808 (analyzing the decision).
  297. See Jenkins, 515 U.S. at 92-99 (eschewing the concept of desegregative attractiveness as a guiding principle in this case, characterizing it as an unjustified interdistrict remedy). In any case, Chief Justice Rehnquist relied on Milliken v. Bradley in Jenkins to overturn the lower court’s use of this tool, since there was only proof of an intradistrict violation.
  298. Id. at 100.
  299. Id. (stating that this was “too far removed from an acceptable implementation of a permissible means to remedy previous legally mandated segregation.”).
  300. See id. at 100-03.
  301. Id. at 100-01.
  302. Id. at 101 (stating the required question which the lower courts failed to answer was “whether the reduction in achievement by minority students attributable to prior de jure segregation has been remedied to the extent practicable.”).
  303. Id. at 102.
  304. Id. at 89 (citing Pitts, 503 U.S. at 491).
  305. Id. at 89.
  306. KLUGER, supra note 3, at 765-774 (describing the judicial retreat from desegregation efforts and the re-segregation of America’s schools).
  307. See, e.g., Belk v. Charlotte-Mecklenburg Bd. of Educ., 269 F.3d 305 (4th Cir. 2001); Taylor v. Ouachita Parish Sch. Bd., 965 F. Supp.2d 758 (W.D. La. 2013).
  308. See supra notes 171-86 and accompanying text.
  309. See supra notes 165-70 and accompanying text.
  310. See, e.g., KECK, supra note 8, at 176 (“The Court’s increasingly narrow vision of constitutional equality was reinforced . . . by its conclusion . . . that the effort to integrate the public schools had proved beyond the capacity of the federal courts. In this view, once school authorities had stopped formally segregating students on the basis of race, the constitutional violation had ceased, and federal court supervision should be ended.”).
  311. 551 U.S. 701 (2007).
  312. Id. at 709. For an interesting discussion of voluntary integration plans, see Neil S. Siegel, Race-Conscious Student Assignment Plans: Balkanization, Integration, and Individualized Consideration, 56 DUKE L.J. 781, 835-38 (2006).
  313. See Parents Involved, 551 U.S. at 718.
  314. Id. at 715.
  315. See id. at 716.
  316. Id.
  317. See id. at 711-14.
  318. See id. at 747-48.
  319. Chief Justice Roberts said: “It is well established that when the government distributes burdens or benefits on the basis of individual racial classifications, that action is reviewed under strict scrutiny.” Id. at 720.
  320. See id. at 723-32. Chief Justice Roberts was joined by Justices Scalia, Thomas and Alito. Justice Kennedy concurred in part, but also concurred only in the judgment in part. Therefore, Kennedy’s separate opinion is essential in determining the scope and impact of the decision.
  321. Id. at 733.
  322. See id.
  323. See id. at 752.
  324. See id. at 783 (Kennedy, J., concurring in part and concurring in the judgment).
  325. Id. at 783 (Kennedy, J., concurring in part and concurring in the judgment).
  326. Id. at 788 (Kennedy, J., concurring in part and concurring in the judgment).
  327. See id. at 783, 789.
  328. Id. at 790. All five Justices in the majority agreed that the school districts failed to show that race-neutral means cannot achieve desegregation.
  329. Id. at 789 (emphasis added).
  330. See id. at 729-33 (Roberts, C.J.), 748 (Thomas, J., concurring).
  331. Id. at 747-48.
  332. Id. (citation omitted).
  333. Id. at 803-68.
  334. See id. at 804-23 (describing the problem).
  335. Id. at 868 (Breyer, J., dissenting). Although Breyer’s language sounds in perfectionist rhetoric, we find it difficult to classify him in Sunstein’s nosology.
  336. See id. at 787-90.
  337. See id. at 788-90.
  338. Id. at 789.
  339. See id. at 789, 798.
  340. See id. at 786-90.
  341. See City of Richmond v. J.A. Croson Co., 488 U.S. 469 (1989); Adarand Constructors Inc. v. Pena, 515 U.S. 200 (1995); Grutter v. Bollinger, 539 U.S. 306 (2003); Gratz v. Bollinger, 539 U.S. 244 (2003); and Fisher v. University of Texas, 133 S. Ct. 2411 (2013).
  342. See Parents Involved, 551 U.S. at 735-71 (Breyer, J., dissenting).
  343. See supra notes 19-146 and accompanying text.
  344. Id.; See also The Real Lessons, supra note 119, at 88-90.
  345. See The Real Lessons, supra note 119, at 83-84.
  346. See supra notes 311-42 and accompanying text.
  347. Swann, 402 U.S. at 16.
  348. See Parents Involved, 551 U.S. at 737.
  349. See id. at 798-804 (Stevens, J., dissenting).
  350. The Parents Involved majority was formed by Justices appointed by Presidents Ronald Reagan, George H.W. Bush and George W. Bush.
  351. We use the phrase “benign discrimination” to refer to legislative and executive actions undertaken to help historically disadvantaged groups, especially, but not only blacks, in gaining increased access to governmental benefits, such as contracts, employment and particularly, education. We use this phrase in contradistinction to “invidious discrimination,” which refers to actions intended to exclude minorities from access to governmental benefits. Use of these terms is solely descriptive. We recognize that many Americans consider “benign” discrimination a contradiction in terms, for the very reasons that it disadvantages majority applicants. Usually we use “benign discrimination” rather than “affirmative action” because we think the phrase more accurately describes the process that occurs both as to its beneficiaries and losers in a zero-sum game. The term “affirmative action” was first employed by President Kennedy in an executive order that directed government contractors to not discriminate against applicants or employees about their race. Since then, the term’s meaning has morphed into what we call “benign discrimination.” See Affirmative Action|Overview, NATIONAL CONFERENCE OF STATE LEGISLATURES (Feb. 7, 2014),
  352. 488 U.S. 469 (1989).
  353. In Croson, Chief Justice Rehnquist and Justices O’Connor, White, Kennedy, and Scalia wrote or joined in opinions declaring that strict scrutiny was the appropriate test in evaluating so-called affirmative action plans. Id. at 472-74. Justice Marshall, perhaps resigned to the inevitable dissented saying: “Today, for the first time, a majority of this Court has adopted strict scrutiny as its standard of Equal Protection Clause review of race-conscious remedial measures.” Id. at 551 (Marshall, J. dissenting).
  354. Id. at 477-78.
  355. 497 U.S. 547, 596-97 (1990). In Metro Broadcasting, the decision was written by Justice Brennan and joined by Justices White, Marshall, Blackmun, and Stevens. The dissenters were Justices O’Connor, Kennedy, Scalia, and Rehnquist.
  356. 515 U.S. 200, 227 (1995).
  357. CHEMERINSKY, supra note 195, at 754 (commenting on personnel changes on the Court during this period).
  358. Adarand, 515 U.S. at 235.
  359. 539 U.S. 306 (2003).
  360. 539 U.S. 244 (2003).
  361. 133 S. Ct. 2411 (2013), vacating and remanding 644 F.3d 301 (5th Cir. 2011), and on remand, 758 F.3d 633 (2014) (holding that UT Austin’s undergraduate admission plan satisfies Equal Protection requirements).
  362. The effectiveness of affirmative action as a policy question is outside the scope of this article. For interesting coverage of some of the issues related to college and university admission, see Dan Slater, Does Affirmative Action Do What It Should?, N.Y. TIMES SUNDAY REVIEW, March 17, 2013, at SR 1. For interesting scholarly perspectives on this subject, see THOMAS J. EPENSHADE & ALEXANDRA WALTON RADFORD, NO LONGER SEPARATE, NOT YET EQUAL: RACE AND CLASS IN ELITE COLLEGE ADMISSION AND CAMPUS LIFE (2009); Daniel E. Ho, Affirmative Action’s Affirmative Action: A Reply to Sander, 124 YALE L. J. 2011 (2005); Daniel E. Ho, Why Affirmative Action Does Not Cause Black Students to Fail the Bar, 114 YALE L. J. 1997 (2005); Richard H. Sander, A Reply to Critics, 57 STANFORD L. REV. 1963 (2005); Richard H. Sander, A Systemic Analysis of Affirmative Action in American Law Schools, 57 STAN. L. REV. 367 (2005).
  363. Grutter, 539 U.S. at 337-38.
  364. Schuette v. Coalition to Defend Affirmative Action, 134 S. Ct. 1623, 1629 (2014); see infra notes 440-58 and accompanying text.
  365. Schuette, 134 S. Ct. at 1623.
  366. Id.
  367. See, e.g., Adam Liptak, Court Backs Michigan on Affirmative Action, N.Y. TIMES, (Apr. 22, 2014), (concluding that the decision effectively endorsed bans on affirmative action in States with similar measures to Michigan). States which have banned affirmative action preferences through either referenda or legislative action include: Oklahoma, New Hampshire, Arizona, Nebraska, California, and Washington. Florida banned racial preferences by executive order. See, Affirmative Action: State Action, NATIONAL CONFERENCE OF STATE LEGISLATURES (April 2014),
  368. 438 U.S. 265 (1978).
  369. See id. at 269-76 (describing the admissions program).
  370. Id. at 355-62.
  371. Id. at 359.
  372. Id. at 418 (Stevens, J., concurring in the judgment in part and dissenting in part).
  373. Id. at 291.
  374. Id.
  375. Id. at 311-19.
  376. Grutter, 539 U.S. 306.
  377. Id. at 316-17.
  378. Id. at 317-20.
  379. Id. at 306.
  380. Grutter v. Bolinger, 288 F.3d 732, 780 (2002).
  381. Id. at 747.
  382. See Grutter, 539 U.S. 327-43 (examining the educational benefits of diversity as a compelling interest and applying narrow tailoring prong).
  383. See id. at 341.
  384. Id.
  385. 539 U.S. 244, 251-52 (2003).
  386. See id. at 252-53.
  387. See id. at 271-76.
  388. Id. at 253-54.
  389. Id.
  390. See id. at 275-76.
  391. See id. at 263-67.
  392. Id. at 271-72.
  393. Id. at 275.
  394. See Grutter, 539 U.S. at 378-87 (Rehnquist, C.J. dissenting); see also Croson, 488 U.S. at 526 (“Nothing prevents Richmond from according a contracting preference to identified victims of [purposeful] discrimination.” (Scalia, J.) (emphasis added)); Parents Involved, 551 U.S. at 731-35 (eschewing diversity in educational settings as a compelling interest).
  395. See supra notes 368-75 and accompanying text.
  396. See supra notes 368-75 and accompanying text.
  397. See Grutter, 539 U.S. at 318.
  398. See CHEMERINSKY, supra note 195, at 764.
  399. See id. (analyzing voting patterns at the time Grutter was decided).
  400. Id.
  401. 551 U.S. 701 (2007).
  402. Id. at 701. See supra notes 311-46 and accompanying text.
  403. See supra notes 327-31 and accompanying text.
  404. See Parents Involved, 551 U.S. at 782-98.
  405. Id.
  406. See id. at 792-98.
  407. See supra notes 324-29 and accompanying text (explaining Justice Kennedy’s deference on the compelling interest prong while applying more rigor to the narrow tailoring prong).
  408. 133 S. Ct. 2411 (2013).
  409. Id. at 2416.
  410. See id. The Top Ten Percent Plan was enacted in the wake of Hopwood v. Texas, 78 F.3d 932 (5th Cir. 1996) (holding, among other things, that diversity was not a compelling governmental interest). But see Grutter, 539 U.S. at 328 (2003) (holding that diversity is a compelling governmental interest).
  411. Fisher, 133 S. Ct. at 2416-17 (describing various iterations of the University of Texas undergraduate admission program including the 10% plan).
  412. Id.
  413. Id. at 2415.
  414. See id. at 2415-17.
  415. Id. at 2415.
  416. Id. at 2417. Fisher did not argue that the University had failed to show a compelling interest in the educational benefits of diversity under strict scrutiny analysis, although the Supreme Court had previously held that “statutes are subject to strict scrutiny under the Equal Protection Clause not just when they contain express racial classifications, but also when, though neutral on their face, they are motivated by a racial purpose or object.” Miller v. Johnson, 515 U.S. 900, 913 (1995).
  417. See Fisher, 133 S. Ct. at 2417, 2420-21.
  418. See id. at 2421-22.
  419. Id. at 2419-22.
  420. Id. at 2422 (Scalia, J., concurring); see also id. at 2423-28 (Thomas, J., concurring in judgment).
  421. Id. at 2419-22.
  422. Id. at 2417.
  423. Id. at 2417.
  424. Id. at 2420-21.
  425. Id. at 2422.
  426. Id. at 2422 (Scalia, J., concurring).
  427. Id.
  428. See id. at 2422-32 (Thomas, J., concurring in judgment).
  429. Id. at 2424. (Thomas, J., concurring in judgment).
  430. Id. at 2434 (Ginsburg, J., dissenting) (describing the admissions policy as to race as being only “a factor of a factor of a factor.”).
  431. Id. at 2432-34.
  432. Id. at 2433; cf. Gratz, 539 U.S. at 303 n. 10 (Ginsburg, J. dissenting) (asserting that it is “disingenuous” to characterize percentage plans as “race-neutral”).
  433. Fisher, 133 S. Ct. at 2433 (Ginsburg, J., dissenting).
  434. See Croson, 488 U.S. at 507.
  435. See, Michael E. Rosman, The Quixotic Search for Race-Neutral Alternatives, 47 U. MICH. J.L. REFORM 885, 889 (2014) (dismissing claims of racial neutrality when the very purpose of the criteria selected is to increase the enrollment of racial minorities).
  436. See 2009 Tex. Gen. Laws 4252 (“The purpose of the reforms provided for in this Act is to continue and facilitate progress in general academic institutions in this state with regard to the racial, ethnic, demographic, geographic, and rural/urban diversity of the student bodies of those institutions . . . .”); See generally Fisher v. University of Texas at Austin, 631 F.3d 213, 224 (5th Cir. 2011) (“The Top Ten Percent Law did not by its terms admit students on the basis of race, but underrepresented minorities were its announced target and their admission a large, if not primary, purpose.”), rev’d, 133 S. Ct. 2411 (2013); Brian Fitzpatrick, Strict Scrutiny of Facially Neutral State Action and the Texas Ten Percent Plan, 53 BAYLOR L. REV. 289, 292 (2011). See also Fisher, 133 S. Ct. at 2433 (2013) (Ginsburg, J., dissenting).
  437. The absence of a Top Ten Percent claim must have been strategic since Justice Kennedy would almost certainly have been the one who decided the outcome of the case. Since Kennedy had written at least three times approving the concept of the educational benefits of diversity, it was unlikely he would be receptive at that time to a change of mind.
  438. See Rosman, supra note 435, at 896.
  439. Id. at 885.
  440. 134 S. Ct. at 1624.
  441. Id. at 1629-30; see also supra notes 376-84 and accompanying text.
  442. Schuette, 134 S. Ct. at 1629. That said, an April 2014 survey asked: “In general, do you think affirmative action programs designed to increase the number of black and minority students on college campuses are a good or a bad thing? Overall, 63% thought it was a good thing and 30% a bad thing, and 8% didn’t know. Among whites, 55% gave a positive response while 36% said it was a bad thing with 8% “don’t know” responses. Among blacks, 84% said it was a good thing, while only 8% thought it was a bad thing. Eight percent of this group said they didn’t know. Hispanic responses were 80%, 15% and 5% in each category. Republicans broke down 43% for, 50% against and 8% don’t know. Democrats broke down 78% for, 15% against, and 6% don’t know. See Bruce Drake, Public Strongly Backs Affirmative Action on Campus, PEW RESEARCH CENTER (Apr. 22, 2014),
  443. Coal. to Defend Affirmative Action v. Regents of Univ. of Mich., 701 F.3d 466, 498 (6th Cir. 2012).
  444. See id. at 470.
  445. Id.
  446. Schuette, 134 S. Ct. at 1630.
  447. Kennedy was joined by Justice Alito. Justice Roberts wrote a concurring opinion.
  448. Schuette, 134 S. Ct. at 1624.
  449. See id. at 1634-35.
  450. Id. at 1635.
  451. Id.
  452. Id. at 1638-39.
  453. See id. at 1639-48 (Scalia, J., concurring in the judgment, joined by Thomas, J.).
  454. Id. at 1649-50 (Breyer, J., concurring in judgment).
  455. Id. at 1650.
  456. Id. at 1629-38 (Scalia, J., concurring in the judgment, joined by Thomas, J.).
  457. See id. at 1651-83 (Sotomayor, J., dissenting) (arguing the amendment of Art. I, § 26 to Michigan’s constitution, which prevented preferential treatment in admission to public employment, education, and contracting, violates the Equal Protection Clause as it eliminates affirmative action, thus directly discriminating and creating a more injurious chance of success to minority groups).
  458. Indeed, Justice Sotomayor advanced many arguments in this spirit. She argued that the Equal Protection Clause is typically construed as referring to the treatment of different groups as expressly required under existing laws, as well as to the implementation of many laws that were passed to burden certain groups on the basis of race. See id. at 1651-83. Among other things, she asserted that Proposal 2 had that precise effect. See id.She said that the Amendment had a racial focus and placed a greater burden on minority applicants. See id. Sotomayor argued that the plurality and concurring opinions allow a majority of voters in Michigan to prevent the elected university officials from implementing constitutional race-sensitive admissions policies and therefore they ignore a key purpose of the Equal Protection Clause. See id.
  459. KECK, supra note 8, at 113.
  460. Id.
  461. See LAWRENCE BAUM, THE SUPREME COURT 29-30 (8th ed. 2004).
  462. Id.
  463. Id. at 48. The Senate at this time was controlled by Democrats. Labor and civil rights groups opposed both nominees, which helped to secure their defeat. Id. at 48-49.
  464. KECK, supra note 8, at 113.
  465. Id.
  466. See BAUM, supra note 461, at 47.
  467. Id. at 113-14.
  468. Id. at 47.
  469. Id.
  470. KECK, supra note 8, at 114.
  471. See id.
  472. See id. at 113.
  473. Id. at 158.
  474. Although Reagan faced few political constraints in appointing a successor to Justice Potter Stewart since Republicans controlled the Senate, he followed his campaign pledge to nominate the first female Justice. “Reagan’s selection of O’Connor angered many of his conservative supporters, particularly because her legislative record [in Arizona] showed signs of moderation on the abortion issue . . . . “[I]t is clear that he was downplaying these substantive considerations for the time being.” KECK, supra note 8, at 158. See also O’BRIEN, supra note 4, at 71 (characterizing Reagan’s appointment of O’Connor as more for symbolic reasons than ideological); supra notes 398-99 and accompanying text (describing O’Connor’s majority opinion in Grutter wherein she concluded in a five to four split that the law school’s interest in educational diversity was compelling and the program was narrowly tailored to advance that interest).
  475. One scholar said about the confirmation process: “Bork was defeated mainly because he challenged a style of interpreting the Constitution that has become part of the American political tradition [the right to privacy], and that the public, much to his surprise, largely supports.” RONALD DWORKIN, FREEDOM’S LAW: THE MORAL READING OF THE CONSTITUTION 278, 287 (1996).
  476. O’BRIEN, supra note 4, at 74.
  477. O’BRIEN, supra note 4, at 74.
  478. See id. at 74-75.
  479. See id. at 74-76.
  480. Id. at 75.
  481. O’BRIEN, supra note 4, at 74, 77. (noting allegations that Douglas Ginsburg had smoked marijuana as a law professor forced him to withdraw after New Right senators opposed his nomination on that ground).
  482. See O’BRIEN, supra note 4, at 77 (describing Kennedy during his nomination hearings as claiming “no fixed view” on abortion and distancing himself from some of the Reagan administration’s and Bork’s controversial polices); KECK supra note 8, at 164 (describing Kennedy as a moderately conservative and “politically acceptable judge” at the time of his nomination).
  483. President Franklin Roosevelt too had enormous success in turning a conservative Court into a liberal one and changed the direction of the Court’s policy making. See O’BRIEN, supra note 4, at 65. With respect to civil liberties this was especially true of Roosevelt’s appointees Hugo Black and William Douglas. See BAUM, supra note 470, at 22-23 (commenting that despite the fact that Chief Justice Warren is often credited with the expansion of civil liberties during his stewardship, Justices Black and Douglas played roles equal or greater than Warren’s in that process).
  484. See generally Washington v. Davis, 426 U.S. 229 (1976) (discriminatory impact without discriminatory purpose is insufficient to state an Equal Protection claim); Arlington Heights v. Metro. Hous. Corp.,, 429 U.S. 252 (1977) (explaining ways a discriminatory purpose may be proved); Richmond v. Croson Co., 488 U.S. 469 (strict scrutiny applies to all race-based decision making); Adarand Constructors v. Pena, 515 U.S. 200 (1995) (strict scrutiny applies to all race-based decision making); Personnel Adm’r of Mass. v. Feeney, 442 U.S. 256 (1979) (establishing “because of ” test for proving discriminatory purpose); and Metro Broad., Inc. v. Fed. Commc’ns Comm’n, 497 U.S. 547 (1990) (congressionally approved affirmative action need only meet intermediate scrutiny).
  485. See supra note 7 and accompanying text (discussing the liberal perfectionist vision).
  486. Parents Involved, 551 U.S. at 748.
  489. By “ex ante” we refer to measures of judges’ ideology derived “entirely on pre-appointment information about a judge or Justice.” LEE EPSTEIN, WILLIAM M. LANDES & RICHARD POSNER, THE BEHAVIOR OF FEDERAL JUDGES, A THEORETICAL & EMPIRICAL STUDY OF RATIONAL CHOICE 70 (2013). It is ex ante because it is fixed at the time of appointment. Another ex ante approach is one developed by Jeffrey Segal and Albert Cover (and now Segal with other collaborators). It is based on the editorials published in four newspapers (two conservative and two liberal) about a nominee to the Supreme Court, from the time of the nomination through the Senate’s vote to confirm. See Jeffrey A. Segal and Albert D. Cover, Ideological Values and the Votes of U.S. Supreme Court Justices, 83 AM. POL. SCI. REV. 557 (1989); Segal et al., Ideological Values and the Votes of U.S. Supreme Court Justices Revisited, 57 J. OF POL. 812 (1995). Still another ex ante approach is taken by Cameron and Park. See Charles M. Cameron & Jee-Kwang Park, How Will They Vote? Predicting the Future Behavior of Supreme Court Nominees 1937-2006, 6 J. EMPIRICAL LEGAL STUDIES 485, 510 (2009). We chose the party of the appointing President as our proxy for Justices’ ideology because it is unambiguous, usable for all Article III judges, highly correlated with other measures of judicial ideology (including ex-post measures) and easy to understand. See David Zaring, Reasonable Agencies, 96 VA. L. REV. 135, 139 (2010).
  490. See supra notes 459-72 and accompanying text.
  491. See supra notes 376-94 and accompanying text.
  492. See BAUM, supra note 470, at 42.
  493. See O’BRIEN, supra note 4, at 70-78 (depicting the selection process implemented during the Reagan administration after Sandra Day O’Connor’s selection).
  494. See generally Michael A. Bailey & Forrest Maltzman, Does Legal Doctrine Matter? Unpacking Law and Policy Preferences on the U.S. Supreme Court, 102 AM. POL. SCI. REV. 369, 381-83 (2009) (finding legal factors play an important role in Supreme Court decision making and that the effect of legal factors varies across Justices); Stephen M. Feldman, Supreme Court Alchemy: Turning Law and Politics into Mayonnaise, 12 GEO. J. L. PUB. POL’Y 57,97 (2014)(arguing that law and politics are so inextricably intertwined that they often cannot be differentiated); Mark J. Richards & Herbert M. Kritzer, Jurisprudential Regimes in Supreme Court Decision Making, 96 AM. POL. SCI. REV. 305-20 (2002) (finding evidence that doctrinal changes in First Amendment case law led to measurable changes in determinants of case outcomes). But see generally Segal & Spaeth, supra note 487 (discussing that judges base their decisions on the merits on the facts of the case juxtaposed against their personal policy preferences).
  495. See Gregory C. Sisk & Michael Heise, Ideology “All the Way Down”? An Empirical Study of Establishment Clause Decisions in the Federal Courts, 110 MICH. L. REV. 1201 (2012).
  496. The decisive date is actually set just before Milliken, which took place in of July 1974, so that the votes in Milliken are included with the post-Milliken votes.
  497. See generally ROBERT G. MCCLOSKEY, THE AMERICAN SUPREME COURT 283 (4th ed. 2004) (characterizing Brennan’s appointment as cross-party). See also O’BRIEN, supra note 4, at 50 (citing to Telephone Interview with Attorney General Brownell (September 9, 1956) and explaining Eisenhower wanted “’a very good Catholic, even a conservative Democrat,’ in order to ‘show that we mean our declaration that the Court should be nonpartisan’ ”).
  498. See supra notes 6-11 and accompanying text.
  499. See MCCLOSKEY, supra note 497, at 207 (describing Powell as a “racially moderate but otherwise conservative Southern Democrat”); O’BRIEN, supra note 4, at 47 (asserting Nixon appointed Powell, “a Democratic-Republican from Virginia, for his ‘strict constructionist views’ ”).
  500. See ALAN AGRESTI, CATEGORICAL DATA ANALYSIS 163-206 (3d ed. 2013) (discussing applicability of logistic regression and associated methodology).
  501. Id. at 455.
  502. Id. at 462-73 (discussing GEEs).
  503. In all the statistical models to follow, we specify a binomial distribution for the outcome variable, a logit link function, and requested robust standard errors. The use of GEE entails a “working guess” for the within-group correlation structure. Id. at 462. One statistic that is used as an aid to adjudicate between different correlation structures is known as the Quasi Likelihood under Independence Model Criterion (QIC). See W. Pan, Akaike’s Information Criterion in Generalized Estimating Equations, 57 BIOMETRICS 120 (2001); James Cui & Gouqi Qian, Selection of Working Correlation Structure and Best Model GEE Analyses of Longitudinal Data, 36 COMMUNICATIONS IN STATISTICS: SIMULATION & COMPUTATION 987 (2007). In all the models considered here, an “independent” correlation structure turned out to be best according to the QIC criterion and, as such, this correlation structure was specified.
  504. We compute a McFadden’s R-Squared which increases from 0.13 in the baseline model to 0.25 in a model that accounts for the Milliken legal precedent. However, we recognize and caution that there is no unambiguous interpretation of any RSquared statistic in the case of categorical data analysis unlike in OLS regression. See J. SCOTT LONG, REGRESSION MODELS FOR CATEGORICAL AND LIMITED DEPENDENT VARIABLES 104-06 (1997) (discussing McFadden’s R-Squared).
  505. This is based on comparison of the QIC and QIC_u statistics. See Pan, supra note 503; James Cui & Guoqi Qian, supra note 503. Also, if we compute the McFadden’s R-squared based on the reported “deviance” we obtain a value of 0.33, as opposed to a value of 0.25 for the previous model which did not break down ideology in terms of specific presidential appointment eras.
  506. See supra note 9 and accompanying text.
  507. See supra notes 473-76 and accompanying text (reporting more conservative pre-confirmation scores for Ginsburg than O’Connor). See also Cameron, supra note 489.
  508. See generally LINDA GREENHOUSE, BECOMING JUSTICE BLACKMUN (2005) (depicting Blackmun’s ideological drift to the left during his years on the Court).
  509. See supra note 462-63 and accompanying text.
  510. See supra notes 333-35 and accompanying text (Ginsburg joining in Breyer’s Parents Involved dissent where they bemoan the resegregation of America’s public schools and contend that the holding in Parents Involved “threatened the promise of Brown”).
  511. See Schuette, 134 S. Ct. 1634-36 (enumerating reasons why Michigan’s Proposal 2 fundamentally restructured minority access to government decision makers in violation of the Equal Protection Clause).
  512. See Benjamin Siegel & Jeff Zaleny, When Will Supreme Court’s Aging Liberal Justices Retire?, ( Jun. 26, 2014), One liberal legal scholar, Erwin Chemerinsky, called on Justice Ginsburg to retire at the end of the 2012-13 Term to ensure that her replacement will be made by President Obama and possess Ginsburg’s views and values. See Erwin Chemerinsky, Much Depends on Ginsburg, L.A. TIMES (Mar. 15, 2014), Justice Ginsburg, in turn, has demurred to such suggestions. See Adam Liptak, Court Is ‘One of the Most’ Activist,’ Ginsburg Says, Vowing to Stay, N.Y. TIMES, August 23, 2013, A-1 (reporting that Ginsburg has asserted she will stay on the Court as long as her health and intellect remained strong and the identity of the President who appoints her will not figure into her planning, “even if that new [J]ustice could tilt the balance of the court and overturn some of the landmark women’s rights decisions that are a large part of her legacy.”).
  513. See Noah Feldman, What Republican Senate Means for Supreme Court, BLOOMBERG VIEW (Nov. 5, 2014),
  514. See Adam Liptak, The Polarized Court, N.Y. TIMES (May 10, 2014), (quoting interview with Professor Neal Devins on the Court’s polarization and likely outcome scenarios from congressional and presidential elections).