Supreme Court Opinions on Equal Protection in Education and Their Use of Social Science Research
As part of the series of constitutional amendments adopted in the years immediately following the Civil War, the Fourteenth Amendment to the U.S. constitution, adopted in 1868, was intended to require the states to treat former slaves as full citizens and to provide them with due process and equal protection under the laws.1 This was a dramatic change for individuals who just a few years before had been legally denied the most fundamental human rights and treated as the property of others. But, the meaning of this amendment as applied to various realms of social life changed over time as the U.S. evolved in many areas including its demographic composition as well as its norms and beliefs regarding how members of different groups should be treated and what they are like.
This chapter traces the history of the application in Supreme Court opinions of one of the basic clauses of the Fourteenth Amendment, the Equal Protection Clause, to two issues that caused great controversy in the U.S. in the twentieth century and that are still controversial today—school desegregation and affirmative action in educational institutions. We focus on these two areas because they both, in different ways, deal with the implications of the Equal Protection Clause for the racial/ethnic composition of educational institutions. Also, although they are generally thought of as different areas of litigation, the issues dealt with have become increasingly similar. In addition, we discuss more briefly the extent to which social science has figured in opinions handed down by the Supreme Court in such cases as well as changes over time in the utilization of social science theory and research.
Equal Protection and School Desegregation
Supreme Court cases pertaining to the issue of racial segregation in school systems in which the Equal Protection Clause has figured prominently span more than 100 years. Here, we summarize landmark cases in this area, focusing on the changing conceptions of the requirements of the Equal Protection Clause. Rather ironically the first case we discuss, which in 1896 laid the legal basis for the development of separate school systems serving White and African-American students, apparently had little to do with schools, as it concerned the segregation of railroad passenger cars. Demonstrating the dramatic way that American jurisprudence pertaining to equal protection for members of different racial/ethnic groups has evolved, this section of the chapter ends with a case that concerns whether pre-collegiate educational systems may take account of race in voluntary efforts to increase diversity within their schools.
In the first case considered here, Plessy v. Ferguson,2 the issue at hand was a Louisiana state law that required railroads operating there to provide separate train cars for Whites and African-Americans, or at least to provide sections within train cars separated by a partition.3 The law provided that railroads could be punished for not providing separate facilities and that passengers refusing to sit where instructed by the conductor could be fined or jailed.4 The only exception to the requirement for separate seating was for “nurses attending children of the other race.”5
The Court, in a decision written by Justice Brown, maintained that the requirement for separate seating on trains within Louisiana did not violate the Equal Protection Clause as long as equal facilities were provided to members of both groups.6 Interestingly, the Court noted that the Supreme Court of Massachusetts had earlier let the School Committee in Boston provide separate schools for White and African-American children and that such a practice was quite widespread, being found even in the nation’s capital.7 The basic argument put forward in Plessy v. Ferguson was that “Laws permitting, and even requiring … separation … do not necessarily imply the inferiority of either race.”8 The majority opinion maintained, “The underlying fallacy … consist(s) in the assumption that the enforced separation of the two races stamps the colored race with a badge of inferiority. If this be so, … [it is] solely because the colored race chooses to put that construction upon it.”9 Further, it argued that both groups were treated equally as both were required to sit in the sections set aside for them.10 The Court distinguished between political equality, which it said was required by the Equal Protection Clause, and social equality, which it opined could “neither be accomplished nor promoted by laws.”11 In a spirited dissent, Justice Harlan pointed out that although the law applied equally to members of both groups, “[e]very one knows that the statute in question had its origin in the purpose … to exclude colored people from coaches occupied by … white persons”12 so that the law interfered with the personal freedom of one group of citizens, a situation not consistent with the Equal Protection Clause. He went on to argue that such arbitrary separation of citizens “is a badge of servitude wholly inconsistent with the civil freedom and the equality before the law established by the constitution.”13
The separate but equal doctrine reigned for over 50 years. However, the 1954 Brown v. Board of Education case flatly rejected it.14 In this case, it was decided that denying African-American children admission to White schools under laws requiring or permitting segregation by race was inconsistent with the Equal Protection Clause, even if schools attended by the African-American plaintiffs were equal to those attended by White children with regard to factors such as facilities, curricula, and teachers’ salaries.15 A major rationale presented to justify the about-face was that public education played a dramatically different role in the U.S. at the time of Brown than when the Fourteenth Amendment was passed, or even when Plessy v. Ferguson was decided. Chief Justice Warren wrote for the Court that “Today, education is perhaps the most important function of state and local governments,”16 in contrast to earlier times when state-supported education was neither as widespread nor as necessary for success in life. However, this opinion also picked up on Justice Harlan’s argument in his dissent in Plessy v. Ferguson that separation solely because of group membership “generates a feeling of inferiority.”17 It then argued that such a feeling was likely to affect the motivation of African-American children to learn, thus slowing their development and depriving them of crucially important benefits.18 The opinion succinctly concluded that “in the field of public education the doctrine of ‘separate but equal’ has no place” and that such segregation deprives children of equal protection.19 Given the magnitude of this change and its inconsistency with customs and norms in many parts of the U.S., as well as the great extent to which local districts control education there, it is hardly surprising that in response to tremendous resistance and delay a second Brown decision in 1955 instructed lower courts to “require (a) … prompt and reasonable start toward full compliance” with the earlier ruling.20
The difficulty of gaining timely compliance with the Brown decisions is illustrated by Green v. County School Board of New Kent,21 decided 14 years after Brown I in 1968, and Alexander v. Holmes Country Board of Education,22 decided a year later in 1969. In the former, a county had taken no action for more than a decade after Brown to disestablish a dual school system in which African-American students from all over the district were bused to one elementary school with African-American staff and all the White elementary students were bused to another school with White staff.23 When action was finally taken, a “freedom of choice” plan was developed under which students could apply to attend whichever elementary school they wished.24 But, after three years not a single White child had enrolled in the African-American school, which continued to serve 85 percent of the African-American students.25 Based on Brown II which required dismantling dual systems with “all deliberate speed,”26 this decision said that achieving equal protection by dismantling dual systems was an “affirmative duty”27 and that dual school systems must “take whatever steps might be necessary to convert to a unitary system in which racial discrimination would be eliminated root and branch.”28 Thus, providing the opportunity for African-American students to attend a White school was not enough. The district was responsible for implementing a plan that dismantled its previously existing dual system and eliminated current segregation stemming from earlier segregation. Whereas both the Brown decisions highlighted the need for considering feasibility and local conditions, this decision emphasized a requirement for immediate progress and measurable results. Similarly, the Holmes case called for immediate termination of dual school systems, reflecting the adage that justice delayed is justice denied.29
Delay and opposition to the dismantling of dual school systems led to substantial involvement of the lower courts in deciding whether school systems were complying with the law. Indeed, where systems were obdurate, lower courts often assumed the responsibility of drawing up desegregation plans for them to follow, commonly using court-appointed experts. Both court-ordered and district-developed plans spurred litigation given the controversial nature of the issue. In 1971 in Swann v. Charlotte- Mecklenburg Board of Education, the Supreme Court turned “to the problem of defining with more particularity the responsibilities of school authorities in desegregating a state-enforced dual school system in light of the Equal Protection Clause.”30 For example, it concluded that limited use could be made in planning of desired ratios of students from different backgrounds,31 that the existence of a few one-race schools within a district was not in and of itself indicative of continuation of a dual system,32 and that the pairing of schools that required busing of students from their neighborhood to more distant locations to achieve desegregation was acceptable.33 Charlotte-Mecklenburg, like the school districts in the earlier landmark cases, at one point had a state-sanctioned dual school system, although segregated housing patterns also contributed to the school segregation remaining at the time of this litigation. Interestingly, the majority opinion in this case specifically indicated that it was not ruling on the question of whether state action by authorities other than the school system (such as those concerned with housing issues) that led to segregated schools would suffice to require school system action to desegregate.34
The Keyes v. School District No. 1 case decided in 1973 saw two extremely important developments.35 First, earlier cases had dealt specifically with the rights of White and African-American students, ignoring those from other backgrounds. In Denver, the district involved in the case, roughly 20 percent of the students were Hispanic, 14 percent African-American and 66 percent White.36 Writing for the majority, Justice Brennan concluded that Hispanics constituted an identifiable class for purposes of equal protection, saying “that they suffer from the same educational inequities as Negros and American Indians” and that the combined percentage of African-Americans and Hispanics should be considered in deciding whether schools were segregated.37 This expansion of the groups covered in school desegregation cases was potentially very important from a practical perspective, given the large and growing number of Hispanics in numerous cities across the U.S. during the last half of the twentieth century and at the beginning of the twenty-first century.
Secondly, and also importantly differing from earlier landmark desegregation cases, a formal dual school system like those typically found in Southern states before Brown never existed in Denver.38 Earlier opinions had emphasized the existence of formal dual school systems, creating what was called de jure segregation, as the government action that violated the right to equal protection and hence that had to be remedied by desegregation plans. Yet, the majority opinion in Keyes held that “where plaintiffs prove that the school authorities have carried out a systematic program of segregation affecting a substantial portion of the students … it is only common sense to conclude that there exists a predicate for a finding of the existence of a dual school system.”39 In other words, a formal dual school system did not need to exist to create a duty to desegregate. Rather, showing that school authorities engaged in activities intended to create segregation in substantial parts of the school system would be enough to require desegregation. Perhaps not surprisingly, given the implications of the Keyes decision and the fact that it substantially changed the “rules of the game” regarding which groups could be considered in desegregation litigation and what kinds of actions were sufficient to open districts to the possibility of court-ordered desegregation, the Keyes decision was far from unanimous.
Less than a year later in another important case, Milliken v. Bradley I, the Court decided that a District Court could not order a multi-district solution to desegregating the Detroit city schools, even though that appeared to be the only way to have African-American students in Detroit attend schools with any significant number of White students and the Detroit district had been shown to have discriminated in the past, making it subject to a desegregation order.40 The reasoning underlying this decision was that there was no evidence in the record that the other districts involved in the proposed multi-district desegregation plan had themselves discriminated or that the discrimination that had occurred within Detroit had produced a segregative effect in them.41 Thus, it was seen to be inappropriate to require other districts to participate in a desegregation plan designed to solve Detroit’s problem. Also, in a follow-up decision, Milliken v. Bradley II, the Court determined that the state of Michigan could be required to pay for various educational programs included in a desegregation plan for Detroit developed by the District Court as long as those programs were tailored to remedy the consequences of the constitutional violation, saying “the Constitution is not violated by racial imbalance in the schools, without more.”42 Milliken I and II had huge practical implications, because the demographic composition of many American cities was such that without involving surrounding heavily White suburban districts many cities would have school populations that were heavily minority.
Two major cases in the early 1990s, Board of Education v. Dowell43 and Freeman v. Pitts,44 ushered in a new era in desegregation litigation in which school districts began to be released from desegregation orders. In the former, Justice Rehnquist, writing for the majority, concluded that once equal protection violations had been remedied by a school district’s good faith compliance with a desegregation degree and the vestiges of earlier discrimination had been eliminated to the extent practical, the district should be released from the earlier desegregation order.45 Similarly, in Freeman v. Pitts the Court decided that a Georgia school district that had been determined to have eliminated discrimination in four areas of its operation, but which had not yet eliminated discrimination in faculty assignment or resource allocation, should be released from court supervision in the areas in which it had eliminated discrimination.46 The Court’s opinion in both cases reiterated that racial imbalance between schools in a district in and of itself was not a reason to keep a district under a desegregation order. Even districts with a substantial number of heavily segregated schools could be released partially or completely from court supervision in realms where they no longer discriminated, as long as the remaining imbalances were most likely due to “private” decisions rather than to district action and experience suggested the district had acted in good faith to change discriminatory practices.
These two cases were followed a few years later by one challenging a court-ordered response to prior equal protection violations that went further than many earlier responses, which had focused more on achieving racial balance within schools in a given district through activities such as pairing schools or busing students. Specifically, in Missouri v. Jenkins (1995) the fundamental questions related to equal protection were whether the District Court could order the state of Missouri, which had maintained a dual school system, to pay for efforts to make the formerly segregated Kansas City schools attractive to White students living outside the district as a means of desegregating that school system, and whether it could order state funding for activities such as pay raises for virtually all Kansas City school district employees.47 In addition, there was argument regarding whether that court could order the state to pay for programs designed to increase achievement in the city school district.48 The Court’s opinion on these matters, written by Justice Rehnquist, held that the District Court’s plans to make the Kansas City schools attractive to students from other districts as a way of enhancing racial balance in the city schools was beyond the scope of its authority, since the original violation was an intra-district one and there was no showing that the district’s behavior had directly caused segregation in surrounding districts.49 Furthermore, Missouri was not held responsible for funding the school system.50 Both of these outcomes rested on the principle that desegregation remedies are designed only to restore the victims of discrimination to the place they would have been without the discrimination and that the scope of a remedy must directly reflect the constitutional violation.51
The most recent major Supreme Court case closely related to desegregation at the pre-collegiate level, Parents Involved in Community Schools v. Seattle School District, reflects continuing public skepticism about the value of racial diversity in schools, especially when it comes at the cost of individual preferences in other regards.52 Although this case, decided in 2007, is often thought of as a desegregation case, conceptually it merges with the affirmative action cases to be discussed, because the dispute concerned the legality of districts that were not under desegregation orders taking affirmative steps to foster diversity in their schools. In this case, Seattle, which was never subject to a desegregation order, and Louisville, which had been declared unitary just a year earlier, were sued by citizens who felt that these districts were voluntarily doing too much to racially balance their schools.53 Specifically, in different ways each school district considered the racial balance in a school in deciding whether students from different racial backgrounds who wished to attend it would be admitted there.54 Justice Roberts, writing for the majority, averred that any governmental action that considers individuals’ race in making decisions regarding benefits or burdens requires “strict scrutiny” which means that the action must be narrowly tailored to attain a compelling state interest.55 He argued that the actions of these two districts clearly did not pass that test, distinguishing between the situation in these districts and that in the affirmative action case to be discussed below, Grutter v. Bollinger.56 His opinion also stressed the fact that when these suits were brought, neither city was required to undertake actions to desegregate their schools, thus making their consideration of race particularly problematic.57 The very clear intention of the programs in both cities was to promote racial balance, not to discriminate against minority students. However, his opinion stressed, as had earlier ones, that racial balance in and of itself is not a constitutionally supported goal.58
In striking contrast, Justice Breyer’s extended dissenting opinion in this case, in which three other justices joined, argued that the Equal Protection Clause does not prohibit consideration of race when it is intended to include rather than to exclude and subordinate minority group members.59 He argued that given that the Court decided in Grutter that the University of Michigan law school demonstrated a sufficiently compelling interest in a diverse student body to justify some consideration of race in its admissions process that a compelling interest in the same goal followed a fortiori in this case.60 He also argued that the procedures challenged in the two school districts were, in fact, more narrowly tailored than the plan allowed by the Court in Grutter.61
The preceding discussion reveals a striking evolution in the understanding of the requirements of the Equal Protection Clause with regard to the provision of schooling to students of different backgrounds over about a century. In 1896, the earliest case used that clause to justify segregation in public facilities as long as those facilities were equal. Roughly 50 years later, the “separate but equal” doctrine it propounded was flatly overturned in Brown I. Cases in the following decades explored the limits of the federal courts’ power to prescribe various kinds of remedies for school segregation. In the 1980s and 1990s, cases also began to delineate the conditions under which school systems could be freed from earlier desegregation orders. The Seattle case in 2007 dealt with whether school districts’ voluntary efforts to increase diversity that accomplished this by considering students’ race were constitutional. Thus, it reflected both major changes in the behavior of school districts and continuing public reluctance to embrace the value of diversity when achieving it conflicts with specific individual interests.
The Role of Social Science in Major Supreme Court Desegregation Cases
It is hardly surprising that social science did not figure at all in Plessy v. Ferguson. When that case was heard in 1896, the social sciences were in their infancy. Interestingly, however, statements in that case regarding issues amenable to investigation by social scientists are inconsistent with later social science findings. For example, writing for the majority, Justice Brown asserted “The argument also assumes that social prejudices may be overcome by legislation … We cannot accept this proposition.”62 Yet, there is evidence that, over time, a change in the laws regarding racial segregation can indeed result in changed attitudes as well as changed behavior.63
Indeed, Brown I, roughly half a century later, is often said to be notable as an early example of the deep involvement of social scientists in litigation reaching the Supreme Court.64 According to what Mody characterizes as the “conventional narrative,” the Court not only mentioned this evidence, but relied on it in making this historic decision.65 For example, Justice Breyer, in the years before he joined the Supreme Court, referred to the Court’s “use of social science evidence … to determine a question of law” in Brown I.66 Consistent with this view is the language of part of the Court’s Brown I opinion which made claims that read very much like conclusions from social science:
To separate them (African-American students) from others of similar age and qualification 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. … Segregation … in public schools has a detrimental effect upon the colored children … the policy of separating the races is usually interpreted as denoting the inferiority of the negro group. A sense of inferiority affects the motivation of a child to learn.67
However, another school of thought in legal scholarship contends that claims that Brown I turned on social science evidence misread the situation.68 Indeed, Mody argues that the social science studies were cited in this case primarily to bolster the legitimacy of the Court’s decision, which the justices anticipated would be extremely unpopular, both because it went against existing social customs in many parts of the country and because it suggested a “living Constitution” rather than an “originalist” approach to deciding the case.69
The contentious issue of the role that social science played in Brown I is complex and cannot be resolved in the space available here. However, Mody makes quite a convincing case that social science played a smaller role in the decision than that with which it is usually credited. But, even if its role was primarily a legitimizing one as Mody contends, it was not unimportant. Public acceptance of unpopular decisions is not guaranteed, although it is fundamental to the American system of government. So, even if the social science studies cited only let the justices feel more confident that their decision would ultimately be accepted, such a contribution was noteworthy.
A considerable body of desegregation research conducted by social scientists emerged between the late 1960s and the mid 1970s,70 the period during which school desegregation began on a wide scale.71 Cases decided early in this period, such as Green v. County School Board of New Kent72 and Alexander v. Holmes County Board of Education,73 made essentially no use of such research, although the former did cite some U.S. Civil Rights Commission research on the effectiveness of freedom of choice plans, after saying specifically that it neither adopted nor refused to adopt the views expressed therein. Swann v. Charlotte-Mecklenburg Board of Education cited various facts and figures regarding the demographic composition of the relevant school district, as did many other cases. It also contained information on the amount and distance of busing that occurred in the relevant district with and without racial balance as a goal.74 However, it did not cite social science research, in spite of its suggestion that the age of students was “probably” the pre-eminent factor to be weighed in deciding on whether the time or distance of transportation planned to achieve desegregation was great enough to put children’s health or education at risk, a suggestion on which developmental psychologists might have been able to shed some light.75
The majority opinion in Keyes v. School District No. 1 also made essentially no use of social science data, although a U.S. Commission on Civil Rights report dealing with the situation of Hispanics in the U.S. was cited, as were demographic data on the composition of the Denver school system and particular schools within it.76 However, Justice Powell’s opinion, concurring in part and dissenting in part, quoted the work of sociologist Karl Taeuber to the effect that segregation was a fact of life in virtually all American cities no matter what the local laws or policies.77 Justice Powell used this to support his contention that it was time to end the distinction between de jure segregation, which is that caused by state action intended to segregate, and de facto segregation, which is that caused by the preferences and actions of private individuals. Both the difficulty in determining intent and the disadvantaged position of minority children who happen to have been born in de facto segregated systems rather than in previously de jure segregated systems figured in his argument. He also cited the testimony of a sociologist who served as an expert witness earlier in this case as well as a social scientist’s study of the neighborhood school concept.78
The opinions in Milliken v. Bradley I and II did not explicitly cite social scientific studies, although they did not refrain from making statements on topics about which social science has a lot to say. For example, Justice Marshall’s dissenting opinion in Milliken II made the following statement “Children who have been thus educationally and culturally set apart from the larger community will inevitably acquire habits of speech, conduct and attitudes reflecting their cultural isolation … [this] can affect any children who, as a group, are isolated by force of law from the mainstream.”79 Milliken I also made reference to expert testimony given to the District Court regarding White flight, the departure of Whites from central city areas to surrounding suburban districts.80 The degree to which and the circumstances under which desegregation actually contributes to White flight became a topic of heated controversy and considerable research within the social scientific community shortly after that.81 However, relatively little of that work had been published at the time this opinion was written.
Like most other previous cases, the majority opinion in Board of Education v. Dowell did not explicitly cite social science research.82 However, Justice Marshall’s dissent, in which two other justices joined, referred back extensively to Brown I, including the famous “hearts and minds” quote and asserted that Brown was based on “recognition of the particular social harm that segregated schools inflict” on African-American students.83 It also mentioned two earlier Supreme Court cases asserting a causal connection between school segregation and housing segregation, without citing specific studies supporting that link.84 Finally, in buttressing its argument that stigmatic injury associated with segregation should be considered in cases involving the possible dissolution of desegregation decrees as a vestige of earlier discrimination, it cited two papers suggesting that predominately minority public schools have fewer resources than others.85
The social science-related issue featuring most prominently in Freeman v. Pitts was the relation between school and housing segregation.86 Justice Kennedy’s majority opinion, with which others concurred, cited a study to support the District Court’s conclusion that housing segregation in the relevant district was not caused by school district policies but by private action.87 In his concurring opinion, Justice Scalia emphasized the difficulty of determining whether demographic patterns years or even decades after a desegregation order are a vestige of earlier state action, especially given that the Court had not clearly described how such vestiges could be determined.88 He basically argued that with passing time it becomes untenable to presume that current racial disparities are the consequence of earlier state action.89 Thus, in the near future he would shift the burden of proof to plaintiffs, who presumably would need to present empirical evidence of that connection.90 In sharp contrast, Justice Blackmun’s concurrence suggested that the policies of the DeKalb County School District might have contributed to residential segregation, citing several studies suggesting an interactive effect between school and neighborhood segregation, and pointing out possible avenues of connection between school policy and residential segregation that the District Court neglected to explore before, determining that no part of the current racial imbalance in the schools flowed from earlier de jure discrimination.91
Rather ironically, Missouri v. Jenkins spawned an opinion in this area containing both the most extensive citation of social science studies in desegregation cases since Brown I and the most clear cut rejection of the utility of the social sciences in the Court’s opinions in this area.92 Specifically, in concurring with Justice Rehnquist’s majority opinion, Justice Thomas argued that Brown I was based “on the principle that the government must treat all citizens as individuals, and not as members of racial, ethnic or religious groups.”93 After attacking the validity of the studies cited in Brown I, he cited numerous studies to support his contention that “there simply is no conclusive evidence that desegregation either has sparked a permanent jump in the achievement scores of black children, or has remedied any psychological feelings of inferiority black schoolchildren might have had.”94 Emphasizing his distrust of social science, he opined that the counts should not accept “the unnecessary and misleading assistance” of the field,95 warning lower courts against being swayed by social science’s “easy answers.”96
The most recent major case related to school desegregation, Parents Involved in Community Schools v. Seattle School District, is strikingly different from most preceding ones in that social science figures in it much more prominently, although it got essentially no notice in Justice Roberts’s plurality opinion.97 His only mention of it was to note that the amici briefs, some of which discussed a large number of social science studies, were not in agreement with regard to the possible benefits of racial diversity for academic or social outcomes.98 He also indicated there was no need to deal with this issue because the procedures in question were not constitutional, as they were not “narrowly tailored” to achieve any such possible benefits.99
In contrast, both Justice Thomas’s concurring opinion and Justice Breyer’s dissenting opinion cited numerous studies.100 Justice Thomas again expressed his distrust of social science, citing studies primarily in an effort to illustrate their inconclusiveness and to discredit the way in which they were used in Justice Breyer’s opinion. In addition, he critiqued Justice Breyer’s dissent for maintaining that the social science evidence is strong enough to let a democratically elected school board argue that racial balance is a compelling interest.101 However, he also used the results of research to support a point he wished to make about the possibility of high achievement in all African-American schools.102 Justice Breyer’s opinion is striking in that, unlike most of the other desegregation cases in which social science has been cited, he included tables of data from social science studies and discussed the actual findings of several studies with considerable specificity. For example, when arguing that school districts have a compelling interest in racial diversity he cited roughly a dozen studies, quoting two sentences from one and reporting results from others.103 In fact, some of the opinion reads as if it could have been taken from a social science review of the literature. Justice Breyer acknowledged that there was not complete unanimity in social science findings regarding the benefits of diversity in the schools, but concluded that “[i]f we are to insist upon unanimity in the social science literature before finding a compelling interest, we might never find one.”104