Brown, the Law and Society Movement, and the Quest to Change What Counts As Knowledge

When the NAACP LDF launched its litigation campaign against public school segregation, the place for social science evidence was far from certain. One member of the team, Robert Carter, firmly believed that research could be used to challenge the underlying premises of the “separate but equal” doctrine in education. His goal was to demonstrate the inescapable harms of segregation: Separate could never be equal. The LDF's lead litigator, Thurgood Marshall, had his doubts, but Carter persevered and ultimately prevailed (Reference BeggsBeggs 1995:11–12). Even so, it is not clear that Carter and Marshall shared the same far-ranging vision. Marshall, for example, likened the expert testimony to studies introduced to show damages in a car accident, suggesting that the primary purpose was to describe a particular fact, rather than to disrupt conventional norms (Reference KlugerKluger 1977:316).

At the trial level, the LDF introduced findings by Dr. Kenneth Clark, a psychologist who—along with his wife, Mamie—had been studying the self-image of black and white students living in segregated conditions (Reference KlugerKluger 1977:315–16, 321). As Clark himself later recalled, in pursuing this strategy, “Bob [Carter] was way out on the limb, pretty much by himself. Most of the other lawyers felt this approach was, at best, a luxury and irrelevant” (Reference KlugerKluger 1977:321). Carter had succeeded in persuading Marshall to use social science evidence at trial, but the relevance of the studies remained contentious when the LDF prepared to argue before the Supreme Court. One member of the litigation team's inner circle, James Nabrit, concluded that Clark's research was irrelevant to the constitutional question presented by segregation. A colleague, Columbia law professor Jack Weinstein, “thought it absurd to couch our argument in terms of dubious psychological data.” Rather than rely on “a gimmick,” he believed that the LDF should rest its case on “the general movement of the common law, historical evidence, and the trend of the nation in the wake of the Second World War” (Reference KlugerKluger 1977:555).

Because the issue was so divisive, LDF lawyers drafted legal briefs that did not refer to social science findings. Instead of giving up on this type of evidence altogether, however, the attorneys relied on an approach similar to that used in Sweatt v. Painter (1950), a lawsuit that successfully challenged segregation at the University of Texas Law School. In that case, a group of prominent law professors, at the LDF's request, had signed an amicus brief submitted to the Supreme Court. In Brown, Clark compiled an appendix of research findings on the consequences of segregation for black and white children. LDF attorneys asked leading researchers from Berkeley, Chicago, Columbia, Harvard, Princeton, and Vassar, among other places, to sign the research appendix (Reference KlugerKluger 1977:555–7).

Clark's brief was not the only means by which academics participated in the Brown litigation. The Court specifically asked for clarification of the historical antecedents of the Equal Protection Clause of the Fourteenth Amendment. This provision had been used to justify the “separate but equal” doctrine and was now the LDF's basis for undoing it. The attorneys in Brown had only six months to prepare their responses to the Court's interrogatories. The LDF assembled a team of historians, including Alfred Kelly, who was chair of the history department at Wayne University in Detroit. Kelly was determined not to allow history to be prostituted, even in the service of a cause as noble as the NAACP's quest for racial justice (Reference KlugerKluger 1977:626). Yet during intense preparation of the briefs, he recalled that “I ceased to function as an historian and instead took up the practice of law without a license. The problem we faced was not the historian's discovery of the truth, the whole truth, and nothing but the truth; the problem instead was the formulation of an adequate gloss … sufficient to convince the Court that we had something of an historical case. …” (Reference KlugerKluger 1977:640). Kelly considered it almost fortuitous that he eventually emerged with an interpretation of the Fourteenth Amendment's history that was both accurate and supportive of the NAACP's agenda (Reference KlugerKluger 1977:641).

When the Court struck down state-mandated segregation and cited the LDF's social science evidence in footnote 11 of the Brown opinion, some believed that the case would usher in a new partnership between law and social science. In a 1956 article, Jack Greenberg, an LDF attorney, concluded that “the school segregation cases suggest an entirely different way in which the testimony of social scientists can be made useful to the courts” (Reference GreenbergGreenberg 1956:962). Instead of deploying experts merely to establish relevant facts, research could influence normative judgments as “[a] variety of information is brought to bear along with the court's concepts of justice and welfare” (Reference GreenbergGreenberg 1956:962). In defending the use of social science to adjudicate more than narrow factual disputes, Greenberg rejected any bright-line distinction between facts and norms. Arguing that “moral judgments are generated by awareness of facts,” he concluded that “constitutional interpretation should consider all relevant knowledge” (Reference GreenbergGreenberg 1956:969).

Greenberg predicted that reliance on social science would grow as courts confronted more lawsuits that implicated public law issues. His faith in this evidence was not naïve, however. Because of the “emotional and controversial areas of life” at issue, he recognized that “it may be difficult for the court, and for the social scientists, to separate uncertain controversy from positive fact finding” (Reference GreenbergGreenberg 1956:967). As a result, Greenberg anticipated some judicial distrust of expert testimony, but he was confident that these doubts would dissipate as jurists gradually acquainted themselves with the assumptions and methodologies underlying the research.

Greenberg mostly was preoccupied with how social science could benefit the legal process, but he also saw advantages for social scientists, who would be “afford[ed] … the satisfaction of close participation in the operation of society and the administration of justice” (Reference GreenbergGreenberg 1956:970). Even so, he was not blind to the ways in which involvement in legal controversies might jeopardize a discipline's scholarly standing. Despite the dangers, Greenberg was convinced that social scientists could never enjoy a sheltered existence. He cited Robert S. Lynd's observation that “social science is not a scholarly Arcanum, but an organized part of the culture which exists to help man in continually understanding and building his culture” (quoted in Reference GreenbergGreenberg 1956:970). In that regard, Greenberg remarked, social science had much in common with law. Neither, in truth, was an insular enterprise, and both were highly dependent on cultural forces (Reference GreenbergGreenberg 1956:970).

Despite Greenberg's optimistic predictions, others sounded a cautionary note. Among these was Kelly, who had worked with the LDF on the history of the Fourteenth Amendment. In Brown, he noted, the briefs submitted by each side were “law-office history” at its finest, but nonetheless, they imperiled historical accuracy in the service of a litigation agenda. Although the justices ultimately turned to sociological justifications to overturn the “separate but equal” doctrine, Kelly worried that the Court often manipulated history in the service of “extreme political activism.” In particular, the justices used revisionist history “as a precedent-breaking instrument, by which the Court could purport to return to the aboriginal meaning of the Constitution” (Reference KellyKelly 1965:125). An opinion then could claim that “in breaking with precedent it was really maintaining constitutional continuity” (Reference KellyKelly 1965:125). For Kelly, this perversion of history was a transgression against the truth, whether it served the liberal-minded agenda of the Warren Court or the conservative goals of earlier Courts. Although the justices inevitably had to turn to history in making jurisprudential judgments, he concluded that the Court was most likely to ask “questions of the past that the past cannot answer” when enlisting history to engage in political interventionism (Reference KellyKelly 1965:156; Reference RichardsRichards 1997).

Kelly's critique highlighted the ways in which research could become corrupted when harnessed to the aims of lawyers and judges. Others, however, worried about law's subversion by social science. Edmond Cahn, a professor of jurisprudence, supported the outcome in Brown, but he rejected any approach that would make constitutional principle contingent on empirical findings. He considered “the behavioral sciences” to be “very young, imprecise, and changeful,” leaving the law at the mercy of “tomorrow's new revelation—or new technical fad” (Reference CahnCahn 1955:167). Cahn took issue with the quality of some of the research submitted by the LDF, particularly Clark's psychological studies, insisting that “I would not have the constitutional rights of Negroes—or of other Americans—rest on any such flimsy foundation as some of the scientific demonstrations in these records” (Reference CahnCahn 1955:157–8). Cahn bolstered his position by alluding to the questionable role of science in upholding racist practices in an earlier era, warning that a future generation of researchers might “revert to the ethnic mysticism of the very recent past” and “present us with a collection of racist notions and label them ‘science’” (Reference CahnCahn 1955:167).

Social Darwinism and its cousin, the eugenics movement, had been used to justify presumptions of racial inferiority. The apparatus of science, combined with the authority of law, perpetuated segregation and discrimination in the United States and culminated in the horrors of Nazism abroad (Reference HovenkampHovenkamp 1985:664–72). Indeed, one critic of the Court's decision to uphold a separate-but-equal doctrine in Plessy v. Ferguson (1896) had complained that the case “smuggled Social Darwinism into the Constitution” (Reference KlarmanKlarman 1998; Reference MillerMiller 1966:170). As Cahn's comments suggested, the alliance between law and social science was not new. In fact, the fields “ha[d] been intertwined longer, and in more pernicious ways, than [was] generally credited” (Reference Tomkins and OurslandTomkins & Oursland 1991:114). If anything, then, Brown represented a promise of redemption for law and social science, but the question was whether this promise could be realized.

A number of leading figures in the Law and Society movement were talented individuals with an interest in law who had gravitated toward social science in part due to its prominence in Brown (Reference Garth and SterlingGarth & Sterling 1998:456). Building on the decision's approach, these scholars dedicated themselves to using social science to effect progressive reform through the courts, legislatures, and administrative agencies. Aware of the danger that scientific rigor might be sacrificed to the goal of social transformation, these academics embraced norms of objectivity and neutrality, in part by treating legal institutions as objects of study rather than as forums for advocacy. Law and Society researchers wanted to avoid a perennial problem that previously had hampered social science's influence: This work was systematically subordinated to law's purposes and so was distorted in the process (Reference Garth and SterlingGarth & Sterling 1998; Reference TomlinsTomlins 2000:955–8).

As historian Christopher Tomlins observes, “The superficiality of law's encounters with social science prior to the later 1950s and 1960s can be attributed largely to where these encounters had taken place—in elite centers dominated by the institutional imperative and tradition of leadership in legal training and legal scholarship” (Reference TomlinsTomlins 2000:955). In those encounters, “[s]ocial scientific experimentation was always marginal to the mission of [the elite institutions], except to the extent that it assisted them in their abiding purpose—sustaining the ascendancy of law” (Reference TomlinsTomlins 2000:956). With support from the Russell Sage Foundation, the Law and Society movement sought to disrupt this dynamic, locating research sites outside legal bastions, thereby “creat[ing] a condition for uncertainty over who would have the upper hand [law or social science] in the encounter” (Reference TomlinsTomlins 2000:956).

The impetus for the Law and Society movement also was the source of its most profound dilemma. The Brown decision reinforced scholars' belief that research findings could be a powerful force for reform. In spite of a desire to remain neutral, objective, and somewhat detached from the law, adherents of the Law and Society movement had a deep desire to make a difference—at least indirectly—by changing how knowledge about law was produced. Law would no longer be an insular and autonomous enterprise devoted to the analysis of rules and precedents, but instead would attend to empirical accounts of its own successes and failures. As Trubek writes, “[o]f all the contradictions of the original law and society understanding, the idea that liberal values would be automatically fostered by a neutral science of law in society seems to be the most perplexing” (Reference TrubekTrubek 1990:39). In short, the Law and Society movement was founded on a paradox, one that was especially troubling in the area of race because the urgency of reform placed objectivity and neutrality at particular risk.

That perplexing contradiction was hardly unique to the Law and Society movement. In fact, the Russell Sage Foundation embraced the same paradigm, as did other leading postwar philanthropists. Social scientists were to engage in analysis modeled on the natural and physical sciences and to refrain from direct involvement in reform efforts. Russell Sage officials described this as a move from combating social problems on the front lines to fighting them from the “second trench” (Reference O'ConnorA. O'Connor 2007:85). Under this framework, social scientists would aspire to “better knowledge, rather than direct social betterment” (Reference O'ConnorA. O'Connor 2007:84). In doing so, they would focus pragmatically on “what the sociologist Robert Merton called theories of the middle range: focused, empirically verifiable, applicable to real world situations and problems, and generally applicable within the parameters of existing institutions and political arrangements” (Reference O'ConnorA. O'Connor 2007:85).

The emphasis on “the middle range” ultimately made the Law and Society movement vulnerable to a balance of power that eventually would privilege law over social science. After all, research was designed to facilitate problem-solving without challenging conventional forms of authority. In that sense, expertise was deployed in the service of law and not vice versa. As a result,

By the 1980s, “those who continued to insist too strongly on the importance of social science methods once again found themselves on the margins of the legal world” (Reference Garth and SterlingGarth & Sterling 1998:465).

As law reclaimed its preeminence over social science in the decades following Brown, the decision's egalitarian ethic simultaneously was contained in the political arena with the renaissance of principles of free-market individualism. As a result of this shift, the law and economics movement began to eclipse Law and Society, in part because “[e]conomics seemed to define the problems and the solutions for the 1980s just as sociology did for the 1960s” (Reference Garth and SterlingGarth & Sterling 1998:465). Conservative proponents of law and economics made inroads by questioning the tenets of postwar liberalism, making their market ideology explicit, and issuing sweeping calls for deregulation (Reference O'ConnorA. O'Connor 2007:135–6). At the same time, and not coincidentally, “the relative value of social science expertise declined. A know-how that served to help construct an activist state seemed to lose some of its relevance” (Reference Garth and SterlingGarth & Sterling 1998:465).

The waxing of economics and the waning of other disciplines in turn revealed the limits of a value-free approach to building an intellectual movement. The law and economics movement's bold agenda revealed both the tacit ideological commitments of Law and Society scholars and the limits of a focus on “the middle range.” Like it or not, Law and Society scholarship was indelibly identified with the vision of an activist State that had spawned it. As the tenets of an interventionist government came under attack, the middle lost its footing. This declining influence prompted some scholars to argue that “the time is ripe for a post-‘law and economics’ initiative” that would “renew [the Law & Society Association's] progressive role at the intersection of law and social science” (Reference Garth and SterlingGarth & Sterling 1998:466).

During this time, Law and Society researchers faced challenges from the left as well as the right. The Critical Legal Studies movement questioned the notion that law should be the object of social science inquiry, the dependent variable, rather than an autonomous field with an independent and powerful influence of its own (Reference TomlinsTomlins 2000:960). The Crits, as they were called, pointedly rejected social scientists' claims to objectivity and neutrality, equating their methodology with a positivism that did little more than reinforce the status quo (Reference TomlinsTomlins 2000:961). Nor were the Crits alone in questioning Law and Society's foundational assumptions. Scholars engaged in feminist jurisprudence and critical race theory, for instance, also doubted that law should be understood as a “universal, abstract, objective, and neutral construct created by particular actors” (Reference Menkel-MeadowMenkel-Meadow 1990:107). The result was an uneasy coexistence on the liberal-left:

Attacks from the left and the right revealed the inherent contradictions between a studied neutrality and reformist aspirations. As a result, “[t]he tacit union of objectivist knowledge and progressive politics ha[d] come unstuck” (Reference TrubekTrubek 1990:48). This union in many ways represented the essence of Brown's multidisciplinary legacy. As a leader in bringing social science to bear on legal problems, the Law and Society movement had a special stake in preserving this legacy (Reference HeiseHeise 2005; Reference TanfordTanford 1990). Based on the institutional commitments of the Law and Society movement, no other group of scholars was more invested in assuring that social science evidence continued to play an influential role in judicial decisionmaking. Yet as a review of the Court's jurisprudence on race and education shows, Brown's multidisciplinary legacy is far from secure.

After Brown: Has Social Science Evidence Counted As Knowledge in Desegregation and Affirmative Action Cases?

Law professor Michael Heise has argued that “one of Brown's critical—though underappreciated—indirect effects [is that of] transforming educational opportunity doctrine by casting it empirically” (Reference HeiseHeise 2005:280). In his view, the decision contributed to “law's increasingly multidisciplinary character,” a change that greatly expands what counts as knowledge in the courtroom (Reference HeiseHeise 2005:280). According to Heise, this new role for social science evidence is perhaps Brown's most lasting contribution, a legal innovation on a par with its iconic status, regardless of whether the case achieved lasting gains in school desegregation. Yet even Heise concedes that Brown's multidisciplinary legacy has yielded mixed results. In his view, the decision “narrowed the doctrine, diluted the influence of broader notions of justice, and risked privileging social science evidence over background constitutional values” (Reference HeiseHeise 2002:1311). Moreover, courts and judges were thrust into “relatively unfamiliar intellectual terrain” that revealed their limitations in dealing with expert evidence (Reference HeiseHeise 2002:1312).

Not all scholars share Heise's optimism about the future of law and social science—even with all the qualifications he has attached. Fradella argues that beginning with the Supreme Court's 1987 decision in McCleskey v. Kemp, judges have grown increasingly hostile to social science research (Reference FradellaFradella 2003). In McCleskey, the justices considered statistical evidence on whether trial courts applied the death penalty unequally on the basis of race. The findings by law professor David Baldus and two colleagues showed that defendants were more likely to be sentenced to death for killing white as compared to black victims (Reference BaldusBaldus et al. 1983). A federal district court judge dismissed the Baldus et al. study on methodological grounds but noted that, in any event, the disparities were not proof of intentional discrimination against any particular defendant (McCleskey v. Zant 1984:356–61, 379). A number of experts attested to the reliability of the empirical analysis, and the Supreme Court ultimately presumed that the Baldus et al. study was valid. However, the justices rendered the evidence irrelevant by concluding that an equal protection challenge to the death penalty must focus on specific proof of discrimination in the defendant's case, rather than on aggregate studies of disparate effects (McCleskey v. Kemp 1987:292). Fradella concludes that McCleskey was a turning point and that “[c]ourts appear to have … become more hostile to social science evidence since Brown” (Reference FradellaFradella 2003:114).

Other critics go even further than Fradella, contending that there never was a golden age of law and social science after Brown, which in turn collapsed with the McCleskey decision. Chief Justice Earl Warren himself once remarked that footnote 11 “was only a note, after all” (Reference KlugerKluger 1977:706). Some legal scholars have taken Chief Justice Warren at his word, concluding that the evidence in Brown was mere window dressing, a way to justify a decision that the justices would have reached in any event. According to this view, Brown merely perpetuated disingenuousness where research is concerned. Judges today—as in the past—make limited use of this evidence, primarily as a convenient post hoc justification for the results they desire (Reference ModyMody 2002; Reference RyanRyan 2003).

The contested relationship between law and social science in turn implicates claims that Brown transformed the litigation process, opening it up to a wide range of evidence. According to law professor Abram Chayes, the decision ushered in a “public law litigation” model (Reference ChayesChayes 1976:1284). Earlier lawsuits had focused on intention and fault as the touchstones for ordering private disputes, but Brown effected a transformation that rendered these traditional concepts “mere metaphors” for a broader concern with justice. Judges shifted their attention to collective harms that required prospective relief, and reform-oriented lawsuits assumed an open-ended quality that took on some features of a legislative hearing. Multiple interests were represented, as the courts certified class actions and allowed a range of intervenors to participate. Social science evidence, introduced through direct testimony or amicus briefs, offered a broad perspective on the issues at stake (Reference ChayesChayes 1976:1285–302).

Chayes's model implied a significant role for experts in the trial process as well as on appeal. In fact, experts did play a recurring role in the desegregation cases that followed Brown, prompting Judge John Minor Wisdom to describe the relationship between law and social science as a “love match” (Reference WisdomWisdom 1975:142). Despite this cozy image of a perfect pairing, scholars of education law have remained dubious about the contributions that social scientists made. As Ryan observes, “Brown was the first and only desegregation decision by the Supreme Court that at least appeared to rest on social science evidence regarding the harm that segregated schools inflicted on black students” (Reference RyanRyan 2003:1665). Afterward, this type of proof was deemed unnecessary because the Court refused to “allow the ‘factual’ questions decided in Brown to be reopened” (Reference HashimotoHashimoto 1997:140–2; Reference YudofYudof 1978:70).Footnote ¹

As a result, the partnership that developed between law and social science in de jure school segregation cases largely revolved around the implementation of remedial orders. Here, social scientists arguably played a key role in persuading courts to pursue comprehensive, structural reform and giving them a justification for their mandates. As one judge recalls, “The social science evidence did exactly what I expected it to do. What it did was to educate the parties from the very simplistic approach that both sides had taken in the first hearing” (Reference CheslerChesler et al. 1988:217). The love affair between law and social science began to fade, however, in the mid-1960s, when the Coleman report reshaped the academic debate about the benefits of integration and the harms of segregation (Reference ColemanColeman et al. 1966). The report had been commissioned by the U.S. Office of Education, which anticipated that the results would provide hard evidence of racial isolation's pernicious effects. Instead, the research team offered mixed findings. Black students performed better academically in integrated schools than in segregated ones. But these differences perhaps had more to do with family background and peer influences than with the school itself. Moreover, the gains could be due to socioeconomic rather than racial integration (Reference CheslerChesler et al. 1988:41–2).

This study “broke the nearly united front that social science had presented on school desegregation” (Reference CheslerChesler et al. 1988:43). Nonetheless, school boards had difficulty in obtaining expert testimony. When social scientists participated on behalf of a school district, they felt stymied by evidentiary presumptions about the ongoing harms of de jure segregation, presumptions that left few ways to influence the case (Reference RyanRyan 2003:1666). As one expert witness explained, “The legal doctrine is cast in concrete, and that's been one of my frustrations. It's as though the evidence is really immaterial.” He went on to recall how he was admonished by the court for “questioning the facts of Brown” (Reference CheslerChesler et al. 1988:43). Eventually, however, school boards became increasingly sophisticated in the use of social science evidence. The growing battle of the experts took its toll, as judges faced a bewildering array of conflicting studies and claims, which at times “forced [a court] back to its own common sense approach” (Hobson v. Hansen 1971:859).

This “common sense” approach gradually reinstated a private model of the law governed by relatively manageable concepts like intent and fault, rather than by inquiries into complex, cumulative harms. As education scholar Mark Yudof observed as early as 1978, “Since Brown, my impression is that, with few notable exceptions, there has been a marked decline in the willingness of the Supreme Court to embrace social science evidence as the basis for constitutional decisions.” Instead, he argued, the Court limited itself to occasional references to studies bearing on “factual matters” (Reference YudofYudof 1978:70). Yudof attributed this chilly reception to “a crisis of legitimacy” regarding the objectivity and relevance of research, particularly after “the [perceived] failures of social science-based policy strategies during the War on Poverty of the 1960s” (Reference YudofYudof 1978:71).

The most notable example of the Court's reluctance was its response to the LDF's efforts to declare de facto segregation unconstitutional. De facto segregation arises from housing patterns and other factors not directly attributable to past wrongdoing by school officials. Brown had outlawed only segregation due to official acts of discrimination, but the studies in the Court's famous footnote implied that racial isolation had harmful effects, regardless of its cause. Determined to challenge de facto segregation in the North and West, the LDF launched a litigation campaign that once again turned to social scientists. This time, sociologist Karl Taeuber developed studies on residential segregation, showing that a range of government agencies had acted with segregative intent and that a neighborhood school policy perpetuated the resulting patterns of racial separation (Reference CheslerChesler et al. 1988:50–1). This testimony was designed “to advance a public law view of responsibility and therefore of the issues appropriately in dispute” (Reference CheslerChesler et al. 1988:51–2). Despite Taeuber's evidence, the Court rejected the LDF's claim that de facto segregation was unconstitutional (Milliken v. Bradley 1974). Instead, the justices adopted a standard that “was obviously quite formal and blind to the demographic realities of most metropolitan areas” (Reference RyanRyan 2003:1667).

The Court's standard looked not to neighborhood demographics but to school board autonomy. So long as suburban school districts were separate political entities, they could not be enlisted in a metropolitan remedy unless they had themselves discriminated in ways that led to interdistrict segregation (Milliken v. Bradley 1974:744–5). Under this approach, demographic evidence that public school segregation would become entrenched in predominantly minority inner-city schools surrounded by white suburban districts became irrelevant. Expert testimony on the harms associated with this pattern of racial separation also was immaterial. A return to formal principles that did not turn on social science evidence was a congenial development for school boards that had long resisted reliance on experts. As one school board attorney explained, “I thought we would have been better off if we could agree that they have no sociologists and we would have no sociologists, and we could just present our facts to the Court and let it come up with a decision” (Reference CheslerChesler et al. 1988:54). By resorting to a narrowly doctrinal—some would even say doctrinaire—rule, the Court effectively endorsed its own version of a “no sociologists” approach.

So thoroughgoing was the retreat from social science evidence in desegregation cases that even reliance on experts to implement remedial orders suffered. In the early 1990s, the Supreme Court decided two cases addressing when desegregation decrees should be terminated (Board of Education v. Dowell 1991; Freeman v. Pitts 1992). The Court instructed trial courts to evaluate school districts' “good faith” compliance, a standard that left social scientists with very little role to play (Reference OakesOakes 2008:68; Reference RyanRyan 2003:1668–74). For example, expert testimony on the academic and social benefits of ongoing integration as well as the potential costs of resegregation became irrelevant; the information simply did not bear on whether a district had implemented a remedy in good faith (Reference OakesOakes 2008:88; Reference RyanRyan 2003:1670–1). Taken together, the Court's rulings reverted to a private model of the law, one in which intentionality and fault no longer operated as metaphors for social justice.

Having adopted a narrow remedial approach to desegregation in elementary and secondary schools, the Court eventually had to confront another constitutional dilemma involving race and education. Colleges and universities had been employing affirmative action to desegregate student bodies without any findings of past discrimination in admissions. Disappointed white applicants argued that the programs were an unconstitutional form of reverse discrimination. In Regents of the University of California v. Bakke (1978), Allan Bakke, a disappointed applicant, challenged a set-aside for underrepresented minorities at the University of California-Davis medical school. Four justices categorically rejected the admissions program because it did not rectify intentional wrongdoing, while four others considered it an appropriate means of promoting racial inclusion in an otherwise segregated and stratified society. Social science played a limited role in resolving this stand-off. Justice Lewis Powell cast the deciding vote in Bakke, introducing an entirely new rationale that did not grow out of systematic research findings. Powell believed that college and university administrators could properly promote diversity as a way to expose students to a wide range of backgrounds and perspectives and thus promote the free exchange of ideas. Evidence for this approach was decidedly thin, so Powell instead was forced to rely on broad allusions to tradition and experience (Reference HeiseHeise 2008:876–7).Footnote ² According to Justice Powell's biographer, John Jeffries, the diversity rationale was not the product of pedagogical expertise but a plausible justification for the results that Justice Powell instinctively wanted. Justice Powell thought that the law should permit race to be considered but that there should not be “carte blanche for racial preferences” and the programs should be temporary (Reference JeffriesJeffries 1994:469).

Thirty years later, the Court again confronted the constitutionality of affirmative action in university admissions in Grutter v. Bollinger (2003). With the programs in operation for decades, one might have expected social science studies on the pedagogical and social impact to be well-developed and influential. In fact, in responding to charges of reverse discrimination, the University of Michigan had to generate a great deal of expert testimony on the benefits of a diverse student body. Even so, it is unclear what impact, if any, this evidence had. For one thing, the research was hotly contested. A professor of psychology at Michigan, Dr. Patricia Gurin, found positive correlations between diversity in the classroom and interracial socializing on the one hand and intellectual engagement and active learning, particularly for white students, on the other (Reference Moran, Moran and CarbadoMoran 2008:464). Because these findings were directly relevant to Justice Powell's rationale, the survey research came under vigorous attack from experts on the other side. Critics assailed Gurin's methodology and questioned the relevance of mere correlations, when the law required proof of a causal relationship between admissions practices and improved pedagogical outcomes (Reference Moran, Moran and CarbadoMoran 2008:465–6).

Faced with a battle of the experts, the trial court concluded that Michigan's researchers had demonstrated that diversity could yield educational benefits, but this proved something of a Pyrrhic victory. The trial court went on to find that the evidence, though credible, was irrelevant because a majority of the Supreme Court had never recognized the diversity rationale. According to the district court judge, Justice Powell wrote only for himself and so his opinion did not constitute binding precedent (Grutter v. Bollinger 2001:847–9; Reference Moran, Moran and CarbadoMoran 2008:471). On appeal, the Sixth Circuit sidestepped the social science evidence, in part because Chief Judge Boyce Martin, who wrote the majority opinion, had serious reservations about the reliability and integrity of the research. So he simply assumed that diversity was a compelling constitutional interest and refused to treat the matter as open to empirical dispute (Grutter v. Bollinger 2002:738–44; Reference Moran, Moran and CarbadoMoran 2008:476; Reference StohrStohr 2004:206).

When the case reached the Supreme Court, a number of amicus briefs were submitted, including some from leading social scientists. Ultimately, a majority of the Court concluded that diversity could generate educational benefits and referred to “expert studies and reports entered into evidence at trial” as well as amicus briefs and other studies. The Court went on to add that: “These benefits are not theoretical but real,” based on submissions by a group of retired military generals and by chief executive officers of Fortune 500 companies (Grutter v. Bollinger 2003:331). The amicus briefs from military and business leaders played an integral role in Justice Sandra Day O'Connor's majority opinion, arguably overshadowing the social science evidence (Reference Moran, Moran and CarbadoMoran 2008:479, 490). Ironically, research methodologies designed to document the benefits of diversity were shunted aside as “theoretical,” while anecdotal evidence became the basis for a grounded description of reality. No one cited the methodological limitations of these first-person accounts; on the contrary, the Court revealed its preference for what it has termed the “pages of human experience” (Parham v. J. R. 1979:602; see also Reference Bersoff and GlassBersoff & Glass 1995; Reference HashimotoHashimoto 1997).

The uncertain place of social science in cases on race and education was brought home in the Court's recent decision in Parents Involved in Community Schools v. Seattle School District (2007). School boards in Seattle, Washington, and Louisville, Kentucky, defended their voluntary integration plans by arguing that the diversity rationale applies to elementary and secondary education. The plurality opinion by Chief Justice John Roberts concluded that academic freedom, and hence diversity, was a tradition unique to colleges and universities (Parents Involved in Community Schools v. Seattle School District 2007:724–5). Consequently, his opinion made no effort to resolve social science debates about the benefits of diverse classrooms for primary, middle, and high school students (Reference HeiseHeise 2008:880–1). Justice Clarence Thomas, by contrast, went out of his way to point out that available research was inconclusive and so could not form an appropriate basis for constitutional decisionmaking (Parents Involved in Community Schools v. Seattle School District 2007:761–6). In dissent, Justice Stephen Breyer agreed that the evidence was in dispute but argued that the Court should defer to local boards' conclusions about potential gains from diversity (Reference HeiseHeise 2008:881–3; Parents Involved in Community Schools v. Seattle School District 2007:839–45). Justice Anthony M. Kennedy, the crucial swing vote, did not rely explicitly on social science evidence in finding that diversity could be a compelling interest in elementary and secondary schools (Parents Involved in Community Schools v. Seattle School District 2007:789). As the Parents Involved opinion made plain, the alliance between law and social science has been badly fractured, and there is no clear consensus about the relevance of research in resolving legal questions. Brown's multidisciplinary legacy can no longer be taken for granted (Reference Frankenberg and GarcesFrankenberg & Garces 2008:746).

Social Science Enters the Courtroom: Does This Evidence Generally Count As Knowledge?

The gap between Brown's multidisciplinary aspirations and today's jurisprudential realities derives at least in part from inherent tensions between the epistemologies of law and social science, tensions that were not fully addressed in the flush of a landmark school desegregation victory. As Susan Haack, a professor of law and philosophy, explains:

Haack contends that the divergence of law from science calls into question the very legitimacy of the adversarial process as a truth-finding device—at least when “key factual questions can be answered only with the help of scientific work beyond the comprehension of anyone not trained in the relevant discipline” (Reference HaackHaack 2003:63).

The clash of epistemologies that Haack describes has grown even more fraught due to an information explosion that makes quality control urgent yet extremely difficult to achieve. According to law professor Elizabeth Warren, because research can play a strategic role in calls for reform, markets for data have arisen that distort the neutrality and objectivity of expertise. The problem is especially pressing when the flow of information is unregulated, for example, in the political process. As she writes,

Warren worries that assurances of quality, particularly those associated with an academic reputation for independence and integrity, have been seriously degraded. Increasingly, scholars must seek outside funding to support their work. As government grants shrink, there is increasing pressure to undertake research for hire. According to Warren, “For anyone who does independent academic research, who has little to trade in but her independence and reputation, the idea that the market for data has devalued the premier signal for independence and quality—university affiliation—is deeply discouraging” (Reference WarrenWarren 2002:30).

Though Warren focuses primarily on the troubled relationship between law and social science in the legislative realm, courts have not been exempt from these perils. Judges have worried about whether “hired guns” distort the pursuit of the truth in the adversarial process. Justice David Souter sparked a firestorm of controversy when he announced in Exxon Shipping Co. v. Baker (2008) that the Court “decline[d] to rely” on research that ran counter to anecdotal reports on the unpredictability of punitive damage awards. Justice Souter dismissed the studies because they had been funded by Exxon, which sought to limit its liability after a major oil spill in Alaska (Reference LiptakLiptak 2008). Justice Souter's comments were especially hard-hitting because some of the rejected work was prepared by prominent scholars and published in prestigious law journals.Footnote ³ Though directed at Exxon's efforts to manipulate the academic debate, Justice Souter's skepticism clearly had ramifications for the credibility of social science evidence more generally (Reference LiptakLiptak 2008; Reference WeinsteinWeinstein 1994). Indeed, to identify work potentially tainted by bias, the Court now requires friends of the court to disclose any source of compensation that could distort their submissions (Sup. Ct. R. 37(6); see also Reference GarciaGarcia 2008:351–2; Reference SimardSimard 2008:700). Despite efforts to root out misleading evidence, one sociologist involved in the Exxon Shipping litigation has concluded that “The legal system and the scientific method co-exist in a way that is really hard on truth” (Reference LiptakLiptak 2008:16).

Battles over the meaning of racial equality are not apt to attract the sort of big-money players involved in corporate and business disputes like the Exxon Shipping case. Nonetheless, high-profile public interest litigation is extremely polarized and inevitably triggers an arms race for amicus briefs. In the late 1940s and early 1950s, amicus briefs were filed in 23 percent of cases litigated before the Supreme Court. With the advent of a public law litigation model, the number rose to 85 percent in the late 1980s and early 1990s. The mean number of briefs filed in each case also went up, especially when a prominent social controversy was at issue (Reference Kearney and MerrillKearney & Merrill 2000:752–4). Cases with more than 20 amicus briefs first appeared on the Court's docket in the 1970s. Regents of the University of California v. Bakke (1978) held the record with 57 briefs until it was toppled by Webster v. Reproductive Health Services (1989), an abortion case with 78 briefs (Reference O'Connor and EpsteinK. O'Connor & Epstein 1982:317; Reference SimardSimard 2008:672, n. 10). More recently, the affirmative action lawsuits against the University of Michigan attracted record numbers of amici (Reference Alger and KrislovAlger & Krislov 2004:506, n. 25). Of course, not all these briefs were filed by social scientists, but typically, at least some of them were (Reference MickelsonMickelson 2008:1175–8; Reference OakesOakes 2008:83–7, 91–2).

Confronted with an onslaught of information, the Court has struggled to regulate access to the adversarial process in a meaningful way. In the early 1990s, the justices revisited the standards for admitting expert testimony in Daubert v. Merrill Dow Pharmaceuticals (1993). Previously, the Frye test had looked to whether findings were generally accepted in the scientific community to decide whether research was reliable (Frye v. United States 1923). The Daubert decision added several additional criteria, including the falsifiability of the findings, the known or potential error rate for the methodology, and peer review and publication of the results (Daubert v. Merrill Dow Pharmaceuticals 1993:592–6). The changes were made in response to fears that “junk science” was entering the courtroom.

This new approach empowered courts to second-guess the experts' conventional wisdom, given increasing doubts about the integrity of the partisan evidence being introduced (Reference ChesebroChesebro 1993; Reference HuberHuber 1991). Daubert, however, created problems of its own. For one thing, it was not clear that judges were competent to make independent assessments of scientific reliability (Reference MnookinMnookin 2008:1019). For another, a significant body of evidence did not conform to Daubert's model of scientific inquiry, which was based on traditions in the natural and physical sciences (Reference BrodinBrodin 2005:869–70, 876–7). As a result, trial courts had to treat some expert testimony as “other specialized knowledge,” which further muddied the standards for admissibility (Reference RenakerRenaker 1996:1664–5, 1673–84).

Most of the controversy surrounding Daubert has ignored the way in which it targeted adjudicative facts, specific factual questions that arise in applying a doctrinal principle. The standard did not reach legislative facts, which inform courts in making normative judgments about relevant policy concerns (Reference AnchetaAncheta 2008:1121; Reference HashimotoHashimoto 1997:111–13, 126–7). So, for example, in Brown itself, data on the equalization of teachers and facilities in segregated schools related to adjudicative facts. The findings pertained to whether conditions in each school district satisfied the “separate but equal” doctrine (Reference HashimotoHashimoto 1997:118). Today, in an affirmative action lawsuit, data on the weight given to race in the admissions process also would be an adjudicative fact. These studies evaluate whether race is so influential that it operates as an impermissible quota rather than a constitutionally acceptable plus (Reference AnchetaAncheta 2008:112, n. 22). Expert testimony on adjudicative facts like these is carefully scrutinized for reliability under Daubert.

Brown's brave new vision, however, was focused on social science's role in effecting transformation in the law, not merely in resolving narrow factual questions under existing doctrine. So, in Brown, research on the inescapable harms of segregation, even in dual school systems that had equalized, was a legislative fact. It bore on the normative question at the heart of the Court's constitutional dilemma: Could separate ever be equal (Reference HashimotoHashimoto 1997:118)? In Parents Involved in Community Schools v. Seattle School District (2007), studies on the benefits of diversity in elementary and secondary schools played an analogous role. This research was deployed to support a normative commitment to color consciousness, not just as a remedy for past discrimination but as a bridge to a multiracial future (Reference AnchetaAncheta 2008:1143–9; Reference MickelsonMickelson 2008:1178–9).

Daubert does not reach evidence on legislative facts, which judges are free to admit at their discretion. For that reason the Court has been able to adopt a liberal, open-door policy on amicus briefs. Though a formal rule mandates that briefs be submitted only when they provide new factual or legal information, in practice the Court grants nearly every application to file (Reference GarciaGarcia 2008:321; Reference HarringtonHarrington 2005:675; Reference Kearney and MerrillKearney & Merrill 2000:761–6). This open-door policy is important to Brown's multidisciplinary legacy because amicus briefs can address legislative facts that counsel may not address due to procedural and evidentiary constraints (Reference RoeschRoesch et al. 1991; Reference SimardSimard 2008:674–5). Yet merely submitting evidence is not the same as wielding influence, particularly when there are few safeguards to assure reliability and relevance.

The real business of sifting through amicus briefs occurs when the justices and their clerks decide which ones are worth reading. These choices are made behind closed doors, leaving the impact of any submission hard to gauge. There are anecdotal accounts that amicus briefs often become part of a vast unread literature (Reference MauroMauro 2005). Yet some do feature in the Court's opinions. Indeed, one common way of measuring the briefs' influence is to count how many times they are actually cited (Reference Kearney and MerrillKearney & Merrill 2000:811).

This approach yields mixed results. The Court has grown increasingly willing to refer to amicus briefs in decisions. “From 1946 to 1955, 17.60 percent of all opinions cited an amicus brief, a figure that steadily grew in the ensuing decades to 27.57 percent in 1976–1985 and then to 36.97 percent in 1986–1995 (Reference Kearney and MerrillKearney & Merrill 2000:758). Nonetheless, the likelihood that any particular brief will make its way into an opinion remains quite low. For instance, just 3 percent of all amicus briefs filed were actually cited between 1946 and 1995 (Reference Kearney and MerrillKearney & Merrill 2000:759–60). Given these long odds, only a few repeat players like the Solicitor General could feel relatively confident of getting the justices' attention (Reference GarciaGarcia 2008; Reference Kearney and MerrillKearney & Merrill 2000:760–1; Reference McLauchlanMcLauchlan 2005; Reference O'ConnorK. O'Connor & Epstein 1983).

The analysis of citation rates has been supplemented by an examination of success rates for amicus brief filers—that is, how often they prevail relative to an overall rate of winning outcomes. Once again, the results are not straightforward. Amici generally do not bolster the chances that a petitioner who has successfully obtained a grant of certiorari will win; however, amicus briefs do systematically enhance a respondent's chances of prevailing. A notable exception is the Solicitor General, who enjoys an extraordinary rate of success, whether supporting the petitioner or the respondent (Reference Kearney and MerrillKearney & Merrill 2000:789, 792, 803). Again, the status of being a respected repeat player seems key where influence is concerned.

These studies do not address the impact of briefs filed by social scientists in particular. This is an issue worthy of further exploration, but there are reasons to doubt that the submissions exert any special pull on the Court. Scholars, who typically participate in cases on an ad hoc basis, are unlikely to enjoy the substantial reputational advantages that come with being a repeat player. To mitigate this disadvantage, briefs can boast multiple signatories, preferably from prestigious institutions, to enhance credibility and clout (Reference OakesOakes 2008:83–4, 87). Moreover, enlisting organizational support, for example, from professional associations, may help as well (Reference ColkerColker 2007:540–1; Reference LynchLynch 2004:50–1, 66–7). Social science briefs, to the extent that they are filed in controversial cases, face special obstacles because of the large numbers of submissions on each side. In this sort of arms race, amici are not likely to affect outcomes unless they offer original arguments (Reference Kearney and MerrillKearney & Merrill 2000:814; Reference LynchLynch 2004:45). Yet very few amicus briefs are cited for their substantive propositions, and these typically are filed by repeat players, most notably the Solicitor General (Reference GarciaGarcia 2008:324; Reference LynchLynch 2004:43–5; Reference SimardSimard 2008:688).

Even with a dramatic expansion in amicus participation, there is no decisive support for the argument that briefs filed by social scientists shape the judicial decisionmaking process. The unique impact of repeat players in the Supreme Court bar, especially the Solicitor General, probably has little to do with Brown's multidisciplinary turn. If anything, the justices are likely swayed by the trustworthiness of legal interpretations offered by these experienced practitioners, not by their artful use of nonlaw experts. Until further study is done, the impact of social science evidence—at least when introduced in amicus briefs—will remain an open question (Reference GarciaGarcia 2008:352). Available research on citation counts and success rates could be usefully supplemented by efforts to gauge the impact of expert testimony in other settings, such as lower court proceedings and the Supreme Court's grant of petitions for certiorari (Reference Caldeira and WrightCaldeira & Wright 1988:1119, 1122; Reference HarringtonHarrington 2005; Reference SimardSimard 2008).

Inquiries like these shed some light on whether a proliferation of amicus briefs amounts to nothing more than a thin veneer of constitutional empiricism. According to legal scholar Timothy Zick, the Court cites findings selectively, deploying social science information in ways that blunt its actual impact on outcomes (Reference ZickZick 2003). If he is correct, then social science evidence is mere window dressing in constitutional disputes. The haunting possibility therefore remains that an arms race in amicus briefs, including those filed by social scientists, has not changed in any definitive way what counts as knowledge in the courts.

The Closing of the American Judicial Mind: How Has the Court Redefined What Counts As Knowledge?

In discussing the role of experts, Reference HaackHaack (2003) describes irreducible tensions between law and social science as distinct ways of understanding the world. What Brown augured, however, was a change in the courts' epistemological universe, one that would reconcile these different ways of knowing. Sociologists Philippe Nonet and Philip Selznick describe Brown as the triumph of “responsive law,” which requires legal institutions to “give up the insular safety of autonomous law and become more dynamic instruments of social ordering and social change. In that reconstruction, activism, openness, and cognitive competence … combine as basic motifs” (Reference Nonet and SelznickNonet & Selznick 2001:74). Responsiveness should make law look more like science. To use Haack's words, legal analysis becomes “investigative, speculative, generalizing, and thoroughly fallibilist,” and like most scientific conjectures, “even the best-warranted claims are subject to revision if new evidence demands it, and progress is ragged and uneven” (Reference HaackHaack 2003:57).

With this flexibility and openness, a responsive model of the law can be generous—even bold—in using social science evidence to reconsider fundamental normative commitments, much as the Brown Court was. Social pressures then become “sources of knowledge and opportunities for self-correction” (Reference HaackHaack 2003:77). This approach is not without risk, however. As courts become increasingly receptive to alternative sources of knowledge, the adjudicative process loses its claim to a unique authority. This loss of authoritativeness in turn jeopardizes integrity, although Nonet and Selznick ultimately conclude that the gains justify the costs. In particular, other forms of knowledge, including social science, enable courts to distill the meaning of the public good in ways that transcend a purely self-interested use of political power (Reference Nonet and SelznickNonet & Selznick 2001).

Today, the Supreme Court is awash in information, a phenomenon that might appear to vindicate responsive law's possibilities (Reference ZickZick 2003:120, 195–6). Yet as already noted, bombarding the justices with briefs does not necessarily mean that social science becomes a source of knowledge for self-correction. Writing about the Rehnquist Court, Zick contends that constitutional empiricism often has served as a smokescreen to reinstate a formalistic approach to the law (Reference ZickZick 2003:221). In his view, the Rehnquist Court was able to manipulate research because there were no clear benchmarks for interpreting the findings. Without a “way to distinguish ‘good’ and ‘bad’ empirical results,” he asserts, “courts [were] not using data to falsify their own notions of what the law should be, but to support their claims of what the law is. …” (Reference ZickZick 2003:211). Contrary to appearances, the Rehnquist Court's epistemological universe did not expand, and the divide between law and social science evidence remained wide. According to Zick, research remained subordinated to legal verities, always confirming rather than testing them.

Under Chief Justice William Rehnquist's successor, Chief Justice Roberts, the Court now includes a plurality of justices who embrace formalism. They do not indulge in any pretense of constitutional empiricism and so largely exclude social science evidence as a way of knowing. Chief Justice Roberts and his colleagues, Justices Antonin Scalia, Clarence Thomas, and Samuel Alito share a belief that law is abstract and universal; it can be discerned from legal texts and, for some of the justices, from legal history (Reference RossumRossum 2006:27–44; Reference ScaliaScalia 1989:1184; Reference Scalia1997:16–18, 23–5, 29–37). This jurisprudential philosophy has significant epistemological consequences. According to Gilmore,

Under this closed system of American law, “[s]tare decisis [with its assumption of limited change in the law has] reigned supreme” (Reference GilmoreGilmore 1977:63).

Because legal interpretation does not require attention to context or changing conditions, formalism maximizes the tensions between law and science as ways of knowing. A formalist approach requires courts to look to their judicial predecessors, not contemporary social scientists, to determine what the law should look like. Haack's dichotomy reemerges with a vengeance: Law is immutable; science is tentative; law is certain; science is speculative. At most, then, social science can speak to adjudicative facts, but it cannot offer up legislative facts that serve as the motive force in a public law litigation model. For instance, in the challenge to affirmative action at the University of Michigan, the constitutionality of race-conscious admissions policies would depend entirely on the text of the Fourteenth Amendment and perhaps its history, but certainly not on social science research on the benefits of diversity that Michigan had amassed.

Given Brown's precepts, the decision has been something of a thorn in the formalists' side. After all, the reasoning in Chief Justice Warren's opinion, including footnote 11, bears little resemblance to the closed epistemological universe that Chief Justice Roberts and his colleagues envision. In a recent debate over constitutional philosophy, Justice Breyer asked how Brown's result could be squared with Justice Scalia's commitment to strict reliance on constitutional text. Justice Scalia did not answer the question, but he has called the tactic “waving the bloody shirt of Brown” (Reference LiptakLiptak 2009:14). In truth, dramatic changes in American race relations, catalyzed in part by the Court's constitutional leadership, pose a seemingly insurmountable challenge to the static system of jurisprudence that formalists endorse.

Perhaps reacting prudentially to the “bloody shirt,” other members of the Court have declined to adopt a formalist philosophy. In Grutter, for example, Justice O'Connor penned the majority opinion, which clearly rejected a textualist claim that the Constitution is color-blind based on race-neutral language that “no State shall deny to any person within its jurisdiction the equal protection of the laws” (U.S. Const., am. 14). Instead, her decision rested on the diversity rationale without reaching more profound questions of social justice. According to law professor Cass Sunstein, Justice O'Connor was practicing the virtues of “judicial minimalism” (Reference SunsteinSunstein 1999:9). Minimalism, as Sunstein defines it, is a far cry from responsive law. A minimalist judge strives, to the extent possible, to reserve legislating for legislators. Courts therefore dispose of cases on grounds that “leave open the most fundamental and difficult constitutional questions” (Reference SunsteinSunstein 1996:7). In doing so, judges allow the democratic process to resolve complex questions that provoke deep and divided views among the citizenry.

In the area of affirmative action, for example, a minimalist adopts neither a strictly color-blind approach that bans race-conscious admissions policies, nor a theory of justice that would legitimate quotas and set-asides. As Sunstein says, members of the Court who adhere to minimalism have “endorsed no rule and no theory” in this hotly contested area (Reference SunsteinSunstein 1999:135). Their stance “has, however, attempted to help trigger public debate, with, perhaps, an understanding on the part of some of the justices that until recently, the debate was neither broadly inclusive nor properly deliberative—and that it did not honestly reflect people's underlying concerns” (Reference SunsteinSunstein 1999:135). Because this jurisprudential strategy has been democracy-promoting and keeps the discussion of affirmative action alive, Sunstein concludes that it is “possible to celebrate what many have seen as the Court's indefensible course of rule-free judgment” (Reference SunsteinSunstein 1999:136).

Race presents some unique problems for Sunstein's celebratory account of minimalism, particularly insofar as Brown itself “appears to be the strongest example against the claim that [Sunstein means] to defend” (Reference SunsteinSunstein 1999:37). In an attempt to reclaim the jurisprudential high ground, Sunstein argues that Brown“was far less maximalist than it might seem; it can even be taken as a form of democracy-promoting minimalism” (Reference SunsteinSunstein 1999:38–9). To justify this rather improbable statement, he relies on the fact that the landmark decision was the culmination of a litigation campaign that involved incremental victories. Moreover, in Brown II (Brown v. Board of Education 1955), when the Court addressed implementation of its pathbreaking school desegregation decision, the justices relied on a gradualist approach. The decision to integrate “with all deliberate speed” allowed the Court to wait until the political branches signaled their support before moving aggressively to enforce the integration mandate (Brown v. Board of Education 1955:301).

Sunstein's account of Brown is not wholly satisfactory. Efforts to integrate higher education in the years before Brown were arguably maximalist in their way. Certainly, images of George McLaurin, an African American graduate student, sitting in roped-off sections of the classroom, cafeteria, and library at the University of Oklahoma suggest that democratic deliberation was, standing alone, unlikely to carry the day (Reference KlarmanKlarman 2004:208–11). Nor do memories of forcible integration, for example, when President Dwight D. Eisenhower sent federal troops to Little Rock, Arkansas, indicate that dialogue and reason were the spur to meaningful implementation of the Court's mandate (Reference KlarmanKlarman 2004:326–9). If minimalism prizes judicial humility and deference to the political process, Chief Justice Warren and his opinion in Brown seem unlikely candidates for accolades. After all, President Eisenhower, in the wake of the desegregation decision, described “[t]he appointment of that S.O.B. Earl Warren” to the Supreme Court as the worst mistake of his Presidency (Reference AmbroseAmbrose 1981:30; Reference SchwartzSchwartz 1997:477).

If anything, Sunstein's account suggests the limits of responsive law, the political perils that come with judicial engagement in broad social controversies. These dangers in turn explain the Court's gradual retreat from the bold innovations of the Warren Court. Through the judicial appointments process, Congress has steadily populated the Court with justices who—at least during the nomination hearings—expressly disavow any desire to make rather than apply the law. Confirmation proceedings have served as a vehicle to discredit responsive law by treating it as the province of wayward judicial activists (Reference EisgruberEisgruber 2007; Reference Epstein and SegalEpstein & Segal 2005). Law professor Stephen Carter attributes the shift directly to the Supreme Court's stand on school desegregation:

According to Carter, today's hearings “follow the same model that they did half a century ago when the Dixiecrats invented them” (Reference CarterCarter 2009:9). This screening process, then as now, is designed to limit the prospects for responsive law, including its openness to social science evidence as a source of normative guidance.

Moving Forward: Can the Dialogue Over What Counts As Knowledge Be Transformed?

The Law and Society movement faces several challenges in ensuring that Brown's legacy with regard to law and social science evidence remains robust. One very significant difficulty stems from an explosion of information, much of it of questionable quality. A critical problem is how to sift through the authentic and inauthentic, the significant and insignificant, the productive and counterproductive. This is a task that applies as much to social science evidence as to other forms of information. Social scientists have an obvious edge, of course, insofar as their disciplines already impose standards to ensure accuracy and rigor. But the turn in Brown was multidisciplinary, and the very breadth of methodological approaches can create problems as scholars try to position themselves as guardians of quality.

Consider, for example, the Empirical Legal Studies (ELS) movement's efforts to safeguard the integrity and credibility of research. Law professor Elizabeth Chambliss describes ELS as an attempt to create “an ‘empirical’ brand” committed to “quantitative, statistical, and experimental methods” (Reference ChamblissChambliss 2008:31). In fact, the movement's slogan is “Bringing Methods to Our Madness” (Reference ChamblissChambliss 2008:32). To that end, ELS calls for disclosure of methods and seems to prefer those that produce measurable results that can be replicated (Reference MitchellG. Mitchell 2004:197–204). According to Chambliss, ELS's methodological commitments have allied it with positivist social science, research that investigates questions posed by legal doctrine without necessarily interrogating underlying normative assumptions. These are the problems of the “middle range” that Merton described, except that the objects of legal study are now “corporate law, political theory, research methodology, and courts” (Reference ChamblissChambliss 2008:33). Work in the middle range turns on a sense that legal frameworks are stable and solid. In fact, the rise of ELS in the corporate field may be a tribute to the success of law and economics in securing doctrinal precepts that now enjoy widespread acceptance. Moreover, research on business-related matters positions ELS to participate in the high-powered markets for data that Elizabeth Warren has described as seriously in need of quality control. This need grows even more urgent as business and corporate disputes occupy an increasing proportion of the Supreme Court's otherwise shrinking docket (Reference LazarusLazarus 2008).

So far, Chambliss asserts, ELS has spent relatively little time on issues related to race (Reference ChamblissChambliss 2008:33). The Court's fractured jurisprudence in this area probably seems like particularly treacherous terrain for positivist social science. Split decisions, with majorities and dissents sometimes openly sniping at one another, make normative uncertainty palpable and legal standards patently insecure. These battles in turn can frustrate efforts to focus narrowly on the middle range. Shifting paradigms of law can eclipse even high-quality social science. For instance, in the McCleskey v. Kemp (1987) case, death penalty opponents developed statistical evidence on inequitable sentencing patterns based on race. Although the analysis was deemed reliable, the Court rebuffed efforts to bring it to bear on the constitutionality of capital punishment. Instead, the work was marginalized by a jurisprudential philosophy that privileged discriminatory intent and disregarded data on disparate impact.

Even in areas of race and the law that would seem to invite quantitative evidence of the type that ELS prizes, doctrinal uncertainties can have a chilling effect. Consider, for instance, developments under the Voting Rights Act. First-generation enforcement efforts targeted denials of the franchise through literacy tests and other practices that interfered with individual registration and voting. Once these problems were addressed, second-generation litigation emphasized fairness in aggregating votes so that minorities would enjoy not only formal access but also meaningful representation. To address problems of aggregation, courts required data on racial polarization in the voting process. This information was critical to understanding how choices about the boundaries of electoral districts helped or hindered minority representation (Reference GuinierGuinier 1991:1093–7).

Under this framework, researchers should have been central to voting rights litigation. As legal scholar Richard H. Pildes observes, “Law and social science are perhaps nowhere more mutually dependent than in the voting-rights field” because “the critical elements of the cause of action that the Voting Rights Act … creates are defined in terms of legal concepts that necessarily must be given content through the kind of data that social-scientific analysis makes available” (Reference PildesPildes 2002:1518). Even here, however, doctrinal flux has plagued the alliance between law and social science. The difficulties are evident in the Court's recent decision in Bartlett v. Strickland (2009). There, the justices heard a challenge to a North Carolina redistricting plan that eliminated a district with a majority of minority voters and substituted several districts with substantial pluralities of these voters. By establishing a safe district that guaranteed minority representation, the original plan enhanced minority political power, but it also reduced the possibility for coalitional politics that allowed minority voters to wield influence by building cross-racial alliances. If, as studies have shown, white voters are increasingly willing to support nonwhite candidates, coalitional districts actually could enhance minority representation (Reference PildesPildes 2002:1518, 1529, 1567–8).

With consistent evidence that crossover voting has been a robust phenomenon, advocates urged the Court to permit coalitional districts to supplant safe ones, so long as the shift enhanced overall minority influence (Bartlett v. Strickland 2009:1243–4; Reference PildesPildes 2002:1534, 1539). The justices rejected this plea, and in doing so, they sent a cautionary message about the role of social science evidence. Justice Kennedy cited the need for “workable standards” and “clear lines.” He deemed inappropriate any approach that “would place courts in the untenable position of predicting many political variables and tying them to race-based assumptions” (Bartlett v. Strickland 2009:1249). Justice Kennedy considered predictions about crossover voting to be “speculative, and the answers (if they could be supposed) … elusive” (Bartlett v. Strickland 2009:1245). His doubts about speculative social science evidence, his distaste for race consciousness even in data collection, and his desire for bright-line rules—all led him to treat demographic research as a diversion from the fundamental legal questions before the Court. Even with the data-driven demands of the Voting Rights Act, the tensions between law and science that Reference HaackHaack (2003) describes were evident.

For ELS, this kind of jurisprudential ambiguity, so typical of race cases, disrupts the underlying model of how social science relates to law. Law provides doctrinal principles, which in turn raise factual questions that social scientists can usefully answer. Cases like McCleskey v. Kemp (1987) and Bartlett v. Strickland (2009) are a sobering reminder that research, even of high quality, can be dismissed as irrelevant in cases characterized by deep conflicts over constitutional values. Because this kind of marginalization hardly boosts ELS's authority and credibility, the Court's fractured decisions on race are something of a red flag. ELS therefore appears unlikely to resurrect Brown's promise broadly understood, even if data-driven research becomes prominent in addressing problems of the middle range in relatively well-settled areas of law.

The New Legal Realism (NLR) movement has adopted a very different approach to reclaiming the partnership between law and social science. Like the Law and Society movement, NLR aspires to include a range of methodologies, both quantitative and qualitative (Reference Mitchell and MertzT. Mitchell & Mertz 2006:4). This openness to various methods is important to Brown's vision because diverse types of research can reveal new ways of framing social controversies and constitutional disputes. Consistent with an emphasis on lived experience, NLR encourages bottom-up inquiries that explore how law affects people's day-to-day existence. This research in turn can supplement large-scale surveys that support broad generalizations but may obscure the nuances of individual difference (Reference ErlangerErlanger et al. 2005).

NLR is explicitly concerned with power and hierarchy in the study of law. Adherents acknowledge the politics of knowledge, calling into doubt whether any social science study can be truly neutral and objective (Reference ErlangerErlanger et al. 2005:339–43). In this, NLR appears to be somewhat at odds with ELS, which emphasizes neutrality and objectivity as hallmarks of quality and trustworthiness. NLR is also relatively forthcoming about its reformist aspirations. Proponents expressly endorse “legal optimism,” a belief that law sometimes can effect positive social change (Reference ErlangerErlanger et al. 2005:345). The methodological openness of NLR, its willingness to question the status quo, and its express faith in law's transformative possibilities make it a hospitable venue for research on race (Reference ParksParks 2008:712). In fact, a substantial number of contributions in the field have addressed these concerns.

Yet the commodious intellectual tenets and reformist ambitions of NLR raise the specter of marginalization. For social science inquiries to do more than reinforce the status quo, Law and Society scholar Stewart Macaulay says, scholars must be willing to ask hard questions and adopt unorthodox approaches when interpreting their findings. In his view, “the hard part is to get such research funded or to have it count toward tenure in universities that are more and more pushed to please the powerful as the schools struggle for funds” (Reference MacaulayMacaulay 2005:395). In the market for data, there may be few eager consumers of work that upends prevailing conventions, particularly in ways that empower the disadvantaged. Even if researchers succeed in conducting these studies, Macaulay points out, “people are well armed with defenses to ward off offensive or inconvenient knowledge” (Reference MacaulayMacaulay 2005:396).

Barriers to unconventional inquiry can be especially troublesome in the area of race, particularly if this research regularly leads to uncomfortable findings. For example, recent experiments on unconscious bias challenge an antidiscrimination framework that presumes that color blindness is the rule while prejudice is the exception. This research has potentially dramatic consequences, for example, in the area of employment discrimination law (Reference Bielby and CoukosBielby & Coukos 2007:1582–3). Currently, under Title VII of the Civil Rights Act, an employee can allege either disparate impact or disparate treatment (Ricci v. DeStefano 2009:2672–3). Disparate impact claims do not require proof of intent (Ricci v. DeStefano 2009:2678; Watson v. Fort Worth Bank & Trust 1988:992), but they have been steadily declining in importance in part because of the courts' increasing resistance to statistical evidence. Even an express congressional endorsement of this cause of action in an amended Civil Rights Act (1991) did not alter the trend (Reference Bielby and CoukosBielby & Coukos 2007:1584–7). With the decline in disparate impact claims, disparate treatment actions have grown in significance. Disparate treatment requires evidence that an employer's action was motivated by race, color, religion, sex, or national origin (Ricci v. DeStefano 2009:2672). Traditionally, intent has been quite difficult to prove, particularly in complex organizations with multiple actors. Proof becomes even more subtle and elusive as old-style racism disappears and is replaced by subtler forms of bias (Reference Bielby and CoukosBielby & Coukos 2007:1567–8, 1580–2; Reference CarbadoCarbado et al. 2008:84–5).

To amplify the reach of disparate treatment claims, law professor Linda Hamilton Krieger and cognitive psychologist Susan T. Fiske have argued that judges' commonsense beliefs about intent are flawed. Courts wrongly assume that disparate treatment results only from conscious prejudice, but in fact, a substantial body of psychological research shows that much of the time people act automatically and unconsciously when they discriminate. This kind of discrimination can occur even when individuals consciously reject negative stereotypes (Reference Krieger and FiskeKrieger & Fiske 2006:1004, 1027–8, 1032–3). Krieger and Fiske call on judges to adopt “behavioral realism [which] stands for the proposition that judicial models—of what discrimination is, what causes it to occur, how it can be prevented, and how its presence or absence can best be discerned in particular cases—should be periodically revisited and adjusted so as to remain continuous with progress in psychological science” (Reference Krieger and FiskeKrieger & Fiske 2006:1001). Though acknowledging that “law … is not epistemology,” the two scholars urge judges to “take reasonable steps, whether through the solicitation of expert testimony, amicus participation, or otherwise, to make sure they have the science right” (Reference Krieger and FiskeKrieger & Fiske 2006:1002).

Krieger and Fiske's approach would make evidence of divergent patterns in the treatment of whites and nonwhites far more probative of illicit intent. If bias is not a rare and conscious act of animus but a commonplace and unconscious cognitive habit, then systematic differences in the way racial groups are treated could support a prima facie case of disparate treatment. In practice, this approach would revive a significant role for statistics in employment discrimination lawsuits. Should the strategy succeed, it could lead to a central role for expert evidence and culminate in substantial awards for plaintiffs alleging unconscious bias (Reference ParloffParloff 2007). In fact, law professor Gregory Mitchell and business professor Philip E. Tetlock characterize Krieger and Fiske's proposal as “but a small part of an ambitious project to use implicit prejudice research to remake the law” (Reference Mitchell and TetlockG. Mitchell & Tetlock 2006:1027–8). They question the propriety of relying on this work to effect fundamental change:

Mitchell and Tetlock insist that the courts apply a higher level of scrutiny to evidence on unconscious bias lest “an epistemic disaster of minor-epic proportions” ensue (Reference Mitchell and TetlockG. Mitchell & Tetlock 2006:1118). This polarizing discourse in turn demonstrates how counterintuitive empirical findings can be thoroughly denounced when they challenge conventional wisdom and prevailing legal paradigms.

Though race poses special difficulties for the partnership between law and social science, Brown also creates unique opportunities insofar as it shows that social justice cannot be captured in any simple, straightforward formula. In recent years, law and economics has overshadowed the Law and Society movement because “microeconomics is based on a model of behavior that can be readily applied to legal issues” (Reference ErlangerErlanger et al. 2005:342). The rational actor is a concept that neatly coexists with doctrinal images of the reasonable person or the arm's-length bargainer. As a result, the insularity and supremacy of law are not threatened. Yet as constitutional theorist Bruce Ackerman points out, Brown poses a daunting challenge to any easy partnership that obscures fundamental value judgments:

The “bloody shirt” of Brown serves as a constant reminder that abstraction cannot shield the courts from profound moral dilemmas. Recently, racial polemics have obscured the role that social science evidence can play in debates over equal protection law. An axiomatic insistence on color blindness has hardened the discourse. With a focus on legal history and text, formalists have been impervious to data on the ongoing realities of racial stratification. The Critical Race Theory (CRT) movement in turn has adopted an explicitly oppositionalist stance, treating pervasive racism and intractable racial self-interest as foundational assumptions. CRT treats these postulates as givens, in part because of its self-consciously political project to resist racism (Reference ParksParks 2008:7067). Rather than subject these claims to empirical verification, critical race scholars have made heavy use of narrative, a technique that relies on first-person accounts to reveal the victim's perspective. By deploying these epistemological moves, CRT seeks to destabilize hierarchies of knowledge, making empathy and transformative change possible.

This dialectic on race reduces the relevance of empirical inquiry by invoking absolutes on each side. Recently, Gregory Parks, a scholar of race, law, and social science, has called for “critical race realism,” an initiative that would build a bridge between critical race scholars and social science researchers. Despite this plea, Law and Society scholars so far have had only episodic contact with CRT. Parks acknowledges that a rapprochement could be anathema to those who believe that courts use facts as mere pretexts for decisions; that social science can be neither neutral nor objective; that research typically reinforces the status quo; and that narrative is a superior way to disrupt racial hegemony (Reference ParksParks 2008). In a no-holds-barred ideological conflict over the role of color blindness and color consciousness, social science evidence may be seen as a hindrance rather than a help, precisely because empirical uncertainty complicates the clean lines of moral outrage.

Today, the desire for axiomatic truths about race threatens Brown's multidisciplinary legacy. Ironically, both formalists and oppositionalist CRT scholars have concluded that constitutional norms about race are too significant to turn on the vagaries of a contingent and contested social science. So, for those like Parks who call for critical race realism, the challenge is to break through the epistemological gridlock that can arise in the shadow of a polarized politics. In fact, Brown's visionary aspirations for law and social science may hang in the balance.

Conclusion

The Law and Society movement was rooted in a paradox, the conviction that value-free research would naturally lead to progressive change. That optimistic assumption has since been attacked from the left and the right, and a new paradox has emerged. A value-laden, highly polarized politics of race has led to a loss of faith in social science as a source of knowledge for self-correction. As a result, the contemporary dialogue about race is framed in absolutes that appear impervious to data. It seems doubtful that normative battles over race can be resolved in an epistemological vacuum, but it seems certain that the struggle for Brown's multidisciplinary legacy can be lost in one.