Precedent, Parity, and Racial Discrimination: A Federal/State Comparison of the Impact of Brown v. Board of Education

Abstract

Questions regarding Brown v. Board of Education's short-term effect remain unanswered, particularly its comparative impact on federal district courts and state supreme courts. We test this through an analysis of racial discrimination cases in those venues in the twenty-year period bifurcated by the decision in May 1954. Our findings suggest that while federal district courts and state courts were similarly unresponsive to discrimination claims before that date, Brown exerted a significant impact on district court decisions but had little influence at the state level. Furthermore, a third pattern was found in federal appellate courts, where discrimination claims had a high likelihood of pro-minority decisions even before the Supreme Court directive.

Introduction

As the fiftieth anniversary of the 1954 Brown v. Board of Education (347 U.S. 483) decision approaches, considerations of this precedent's long-term consequences are sure to proliferate, since debate on this point persists.¹ But scholars should keep in mind that some of the ostensibly easier questions concerning the impact of the Supreme Court proclamation declaring de jure segregation unconstitutional remain unresolved. In particular, a comparative assessment of the immediate, shorter-term effect of this opinion on federal district courts and state supreme courts has not been delineated. While evidence suggests that Brown exerted significant influence on district court decisionmaking, the next logical step should be to explore whether state courts displayed the same fidelity to the mandate and if state and federal venues differed in their willingness to protect minorities prior to the precedent. The primary focus of this investigation then is a comparison of racial discrimination cases in state supreme courts and federal district courts in the twenty-year period (1944–64) bifurcated by announcement of the initial Brown v. Board of Education opinion in May 1954.

Our findings suggest two clear conclusions. First, in the years preceding Brown, federal district courts and state supreme courts displayed a similar lack of responsiveness to discrimination claims. Once the opinion was announced, however, state supreme courts lagged seriously behind in heeding the Supreme Court mandate. In short, Brown made a significant difference in federal district court decisions, shifting overall outcomes from generally unsympathetic to strongly in support of minority claims, regardless of region. But, in this narrow time frame at least, it had no such impact at the state level.

In addition, we extend these conclusions with the introduction of appellate-level data. These findings establish that federal circuit courts displayed a strong record of minority protection prior to 1954, thus complementing the historical value of our study with a deeper exploration of federal/state parity. In this vein, our findings highlight the potential divergence of outcomes within the federal system as well as the discrete nature of two sometimes conflated aspects of the supposed bulwark tendencies of federal courts: protection of minority rights and fidelity to Supreme Court precedent.

Literature Review

At its core, then, our analysis taps into and contributes to two broad and long-standing issues concerning judicial policymaking: whether federal courts, even in the absence of strong precedent, are more deferential to minority rights than are state courts, and whether lower federal courts respond more obediently to Supreme Court doctrine than do state courts. In this section, after using these general controversies to frame our inquiry and highlight its significance, we focus on the literature related to the specific historical questions our investigation targets.

The first relevant aspect of the existing scholarship concerns the general effects of rules and structure in fostering a federal judiciary that is more protective than state courts of the constitutional rights of groups potentially handicapped in the democratic process.² Supporters of the view that federal courts alone are capable of fulfilling this bulwark role rely on factors such as stronger allegiance to the federal system and the Constitution, and the life tenure–induced freedom of federal judges from majoritarian pressures. In the abstract, proponents of this view range from the Founding Fathers (Alexander Hamilton outlined this scenario in Federalist 78; Hamilton, Madison, & Jay 1961) to contemporary scholars (a modern classic is Neuborne 1977). On the other side of the debate, defenders of the parity argument (Chemerinsky 1991) find the assertion that federal courts are inherently more protective simply unpersuasive.

Empirical studies have generally reported little variation in targeted comparisons of federal and state outcomes (Beiser 1968; Gerry 1999; Grunbaum & Wenner 1980; Pinello 2001; Solimine & Walker 1999). In fact, both Pinello (2001) and Galie (1982) suggest that in some legal environments, state courts may even provide greater protection, a possibility Emmert (1992) attributes to the broad rights increasingly covered by state constitutions. But Haas's (1982) study of prisoner rights claims cautions scholars to restrict comparisons to similar court levels (appellate should be assessed against appellate, and trial against trial). Utilizing that strategy, he reports significant federal/state differences.

Haas's concern is grounded in a significant complication of the bulwark theory—the possibility that all federal courts are not alike and that the appellate level will display greater independence from majoritarian pressures than will the district level. This springs from the argument that, regardless of the freedom from reelection or reappointment that all federal jurists share, district court personnel are more closely tied to the preferences of a given region, whether because federal districts are geographically smaller and more homogeneous than federal circuits, or because U.S. senators, reflecting local preferences, play a significant role in the selection of district, but not appellate, judges.

Somewhat relatedly, the second relevant branch of judicial scholarship concerns varying responsiveness to Supreme Court doctrine. Generally, this literature provides not so much a “yes” or “no” answer as a spectrum in which greater adherence by any court level is dependent upon particular characteristics of the precedent itself and the environment in which it is received (Benesh & Reddick 2002; Caldeira 1985; Canon 1973; Glick & Vines 1973; Johnson 1987; Kilwein & Brisbin 1997; Tarr 1977). Still, a blunt assumption that Supreme Court opinions (particularly controversial ones) will be more faithfully and obediently received by federal than state courts is widely if not universally accepted (Baum 1997; Haas 1982; Wasby 1970). It is possible, however, that federal courts differ in their responsiveness in this dimension as well. Again, as a result of local pressures, district court judges might display a weaker commitment.³

In summary, significant federal/state divergence in regard to protection of minority rights and adherence to Supreme Court precedent is indicated but remains controversial. Furthermore, the more subtle question of whether federal courts differ from each other in regard to these tendencies is largely unsettled. It would appear, however, that despite this general lack of agreement, there is consensus that in regard to racial discrimination claims in the 1950s and 1960s, federal courts were both more protective of minority claims and more receptive of Brown (Shapiro 1995; Solimine & Walker 1999).

References to this scenario abound, but the best evidence for the ubiquity of these assumptions is their acceptance even by scholars who otherwise reject the notion of federal/state differences. For example, in his argument emphasizing the tendency for state courts to uphold the U.S. Constitution just as ardently as federal courts, Bator observes that the persuasive power of the alternate, bulwark argument largely rests on “a special historical experience, involving the division of the country on the issue of racial segregation” (1981:631). Similarly, Chemerinsky (1991), another advocate of state/federal parity, also acknowledges the handling of racial discrimination cases in those years as the exception to his thesis.

Yet the inference that federal courts were more faithful adherents of the precedent and thus more protective of minority rights in the years immediately following the decision has been grounded largely in anecdotal or limited empirical evidence, some of which offers contradictory conclusions. Older accounts of federal district courts in the post-Brown period provide contrasting assessments, some stressing fidelity to the new precedent (Woodward 1974) and others highlighting recalcitrance (Murphy 1959a; Peltason 1961; Rodgers & Bullock 1972; Steamer 1960). This discrepancy results from the nature of the studies, many of which can be characterized as useful but impressionistic narratives of the conflicting pressures on federal district judges in the South that lack firm data on outcomes. The empirical efforts were either published too soon to observe more than a few years of the post-Brown era or focused to a greater extent on variables that fostered district court adherence than on the overall district court record (Giles & Walker 1975; Vines 1964).

Sanders's (1995) account largely settles the narrow question of federal district court fealty, establishing that the Brown opinion moved these tribunals from a 27 to 83% rate of support for minority discrimination claims. Still, there is very little evidence on how state supreme courts reacted to minority claims in this time frame and thus little basis for federal/state comparisons either before or after the decision. The most useful analysis is Vines's 1965 study of state supreme courts, which incorporated comparative data from his earlier (1964) district court analysis. Oddly, though, given modern assumptions, his findings do not suggest significant federal/state variation. While Vines reports a pro-minority decision rate of 33% in the states versus 51% in the district courts, he downplays the difference, noting that once criminal cases (usually, challenges to convictions based partly on the claim of nonintegrated juries) are eliminated the records are virtually the same. In addition, he notes the high rate of 90% at which state supreme courts overturned lower-level dismissals of discrimination claims.

However, there are limitations to Vines's findings. While his time frame runs from 1954 to 1962 (federal cases) or 1963 (state cases), there is no survey of the pre-Brown era, so the impact the precedent had on each level is unclear. Furthermore, even the post-Brown comparison is hampered by the lack of a systematic method of gathering relevant data. For example, many of the cases Vines examines are procedural, repeat cases, or tangential to Brown (e.g., criminal procedure cases). Thus, while a valuable record, these early investigative efforts can and should be expanded and refined. In short, a reconciliation of conventional wisdom with the factual record is due and will contribute not only to a historical clarification but to a more nuanced, comparative understanding of federal and state courts.

Design and Methods

The primary question is whether federal district courts and state supreme courts differed in their responses to minority claims of constitutionally based racial discrimination either before or after announcement of the Brown v. Board of Education decision. We chose the period 1944–64 because it represents a relatively short time frame bisected by this opinion. The endpoint is crucial, as it represents the year the U.S. Congress passed a major Civil Rights Act that would replace Brown as the legal basis for many such lawsuits.⁴

Data collection was facilitated by use of West's Decennial Digest, which indexes cases by legal claim. The authors read and coded all federal district court and state supreme court cases that were listed under various subheadings of the Constitutional Law heading/Equal Protection subheading.⁵ As the unit of analysis, each case represents a minority plaintiff challenging some sort of state-sponsored discrimination as a violation of their federal constitutional rights. We believe that the decision to include only these cases is justified because both the Brown opinion and the bulwark thesis are grounded in constitutional reasoning.

Murphy clarifies this minority discrimination/constitutional nexus, commenting that “lack of social standing, minority status, and claims of discrimination by public officials are often valuable assets in a judicial system which operates under a constitutional command of equal protection of the law” (1959b:372). In short, it would make little sense to test the impact of either the precedent or the more protective nature of federal courts on cases in which the plaintiffs themselves did not raise constitutional issues or relied solely on constitutional provisions other than equal protection (e.g., fair trial guarantees). At the same time, this means it will be imperative to keep our findings in perspective—any conclusions in regard to state/federal comparisons will apply solely to the realm of constitutionally based claims.

We collected information for each case by court level (state or federal, which was also broken down by single judge or three-judge panel), exact date of the decision (whether before or after the announcement of Brown on May 17, 1954), and outcome (for or against the minority claim). The final data set contained 161 cases, with 118 at the district court level and 43 at the state supreme court level. While our overall N may seem small, that is a function of our focus on constitutional claims. We are secure that we largely captured the universe of relevant cases.⁶ It is likely that the smaller number of cases heard in the state courts reflects the common perception that plaintiffs fared better at the federal level. More concretely, the National Association for the Advancement of Colored People (NAACP), an important legal resource for discrimination plaintiffs, was less likely to sponsor cases in state courts, due to local laws forbidding its participation (Murphy 1959b) or because it simply chose not to (Castillo 2001).⁷

We hasten to add here that we have not forgotten Haas's (1982) caveat against comparing federal trial-level to state appellate-level courts. However, given the assumptions outlined above as to this specific legal arena and time frame, our strategy actually gives the weaker, or at least less popular, hypothesis (i.e., that federal courts are no more protective or faithful to Supreme Court precedent than state courts) the best chance of succeeding. As Neuborne notes, “[i]f a competence gap exists at all, it is very slight and may, indeed, favor state appellate judges” (1977:1116). More important, we respond to Haas, and more broadly to the general question of whether federal courts differ from each other, by expanding our analysis in two ways to consider the role of federal appellate courts. First, in the district court equation, we controlled for the presence of three-judge panels, in recognition of the possibility that district court decisions might be driven by the impact of an appellate-level judge. Second, relying upon all district court cases in our data set subsequently heard by circuit courts, we ran a separate equation to assess the extent to which Brown influenced federal appellate decisions, as a comparison to the district-level record.⁸

We also controlled for three additional, potentially important factors. First, since the data set comprised challenges to alleged de jure discrimination in a variety of settings, we controlled for the particular type of case. We acknowledge the possibility of puzzlement over why, in a test of the public school–based Brown v. Board of Education decision, we look at claims not only regarding public grammar and high schools, but also higher education, public facilities, and so on since, as Johnson (1987) asserts, decisions are likely to have greatest subsequent impact on factually similar cases. But the question of factual similarity is not so simple here, for, on the same day Brown was decided, the Supreme Court also remanded two public facilities cases to lower courts for reconsideration in light of the new decision. Therefore, we coded for and controlled for type of case, dichotomizing all decisions as to whether they concerned education (at any level) or another type of discrimination (e.g., public facilities/accommodations, public housing, and private behavior such as marriage and adoption). Through this control, we respond to Beiser's (1968) proposal that future studies on Brown's impact focus on this issue of factual similarity. To the extent that these distinctions prove significant, they could highlight limits on the opinion's clarity and breadth of impact.⁹

Another control is the nature of the discrimination, for just as our data set included de jure discrimination in numerous settings, it was also marked by variation in the extent of alleged government implication. Broadly speaking, all of these cases involved challenges to state-sponsored discrimination, but there exist potentially salient dimensions of official involvement. In some cases, blatant laws or actions were challenged (e.g., the operation of segregated school systems, public golf courses, and airport waiting rooms), while in others the concern was with a more indirect link between government and the act of discrimination (e.g., a segregated, privately owned coffee shop operated in a city-owned building, or the common practice by white-only lunch counter owners to rely on local police to enforce their “private” policy). Again, it is important to control for these differences, since the extent to which Brown or the U.S. Constitution barred these more indirect actions was unsettled at the time (Kelly & Harbison 1970). Thus, we divided all cases into direct and indirect de jure discrimination categories, as indicated by West's description of plaintiff's claim. To the degree that these variations in official involvement prove significant, they may also suggest limits on the impact of Brown.

Finally, and most critically, we controlled for general political/social environment, choosing region as the obvious fit for the particular legal issue of racial discrimination, in which the most crucial conceptual differentiation is between the South and other areas. As Shapiro comments, “[s]urely the institutionalization of racial segregation reached levels in the Southern states never even imagined in the North and West” (1995:54). A number of studies have stressed the role played by region either in general minority protection or specifically in racial discrimination cases at both the state supreme court (Canon 1991; Kilwein & Brisbin 1997; Murphy 1959a; Vines 1964) and district court (Carp & Rowland 1983) levels. At least one version of the lore of Brown's immediate aftermath is that southern federal judges largely ignored the directive, indicating a greater allegiance to their home region than to the Supreme Court. It is crucial to test this possibility, as well as the analogous potential that state supreme courts outside of the South did respond to the precedent. In other words, a simple regional distinction must be assessed as a reasonable alternative to the federal/state variation scenario.

However, it is also important to avoid bluntness on this point, for there is evidence that, within the South, certain subregions were more moderate. Older commentary (Steamer 1960) and more recent scholarship (Black 1976; Black & Black 1987; Shapiro 1995) both reflect some movement toward acceptance of integration in parts of the South. As described by Bartley, referring to the year immediately following Brown, “[t]he Deep South sank deeper into hysterical reaction, while border states cemented their psychological identification with the nation, and the peripheral South, like an unstable planet, swayed between the magnetic attraction of North and South.” (1969:81). Therefore, utilizing the classic works of Black and Black (1987), Key (1949), and Woodward (1974) for substantive coding guidance, we grouped all cases into non-South, border state, peripheral South, and deep South categories. To the extent that pro-minority decisions are less likely in these respective categories regardless of federal or state distinction, the limitations of both precedent and the federal system as bulwark will be emphasized.¹⁰

We used two methods to determine whether chance could be eliminated as the explanation for the observed differences: difference in means testing and difference in coefficient estimates from multivariate logistic regression analysis of state and federal discrimination decisions. Given below are the equations for (1) state supreme court and (2) federal district court decisions in discrimination cases. The “test” for differential response rates was a comparison of the Brown coefficients for the two equations.¹¹

(1) $\begin{array}{cc}P\left(Pro\text{-}MinorityDecision|X\right)& ={\beta }_0+{\beta }_1\left(Brown\right)+{\beta }_2\left(Education\right)\\ & +{\beta }_3\left(Direct\right)+{\beta }_4\left(Border\right)\\ & +{\beta }_5\left(Peripheral\right)+{\beta }_6\left(Deep\right)\end{array}$

(2) $\begin{array}{cc}P\left(Pro\text{-}MinorityDecision|X\right)& ={\beta }_0+{\beta }_1\left(Brown\right)+{\beta }_2\left(Education\right)\\ & +{\beta }_3\left(Direct\right)+{\beta }_4\left(Border\right)\\ & +{\beta }_5\left(Peripheral\right)\\ & +{\beta }_6\left(Deep\right)+{\beta }_7\left(Panel\right)\end{array}$

Because the dependent variable in these analyses is whether the court made a pro-minority decision (1=yes, 0=no), Equations (1) and (2) were modeled using logistic regression.¹² The test variable, Brown, divides the cases into those decided before and after May 17, 1954 (1=after, 0=before). If Brown exerted a liberalizing effect on decisionmaking, then this coefficient should be positively signed and statistically significant. Serving as controls are the variables Panel (capturing whether federal cases were heard by a three-judge panel; 1=yes, 0=no), Education (capturing whether the case involved discrimination in education; 1=yes, 0=no), and Direct (capturing whether the case involved direct official involvement; 1=yes, 0=no). We also included three region dummies: Border (Md., Ky., Mo., and W.Va.), Peripheral South (N.C., Va., Ark., Tenn., Fla., and Tex.), and Deep South (Miss., S.C., La., Ga., and Ala.). Non-Southern states served as the comparison group.¹³

Findings

We begin with an examination of the percentage of pro-minority decisions before and after May 17, 1954, at the state and federal levels. Table 1 gives these results. We note first that, prior to Brown, pro-minority decisions were uncommon at either the state or federal level. And although minority plaintiffs fared slightly better in the 1944–54 period in federal courts, a simple difference in means test reveals that the nine-point difference between the levels' pre-Brown minority support was statistically insignificant (t=0.63, ρ=0.54). After Brown, dramatic changes occurred, but only at the federal level. There, pro-minority decisions increased by thirty-six percentage points, as compared to only two percentage points in the state courts.

Table 1. Percentage of Pro-Minority Decisions in State Supreme and Federal District Courts Before and After the Brown Decision

NOTE: Number in parentheses represents the total number of cases in cell.

Although the results in Table 1 are persuasive, it is possible that factors other than Brown were at play. To obtain a more complete explanation of discrimination case outcomes, in Table 2 we offer the results of our multivariate analysis of federal and state court decisionmaking. With a slight enhancement, the story remains the same. In this full equation, the only factor significantly influencing federal district court decisions was the introduction of the precedent. The impact of this variable should put to rest the argument that any change in district court outcomes was attributable not solely to Brown, but to the broader collection of decisions prior to 1954 that hinted at a doctrinal evolution. These earlier cases may have helped pave the way, but the results in Table 2 clearly document Brown's influence.

Table 2. The Influence of the Brown Decision on Federal District and State Supreme Court Pro-Minority Decisions in Discrimination Cases

^** Significance level: <0.01

Furthermore, two controls—the setting of discrimination (Education) and the nature of state involvement (Direct)—failed to influence decisions, suggesting that the Brown doctrine was read broadly, at least by federal district judges. Once it appeared, many challenges to de jure acts were upheld, no matter the setting (schools versus public facilities) or degree of state involvement (direct or indirect). (The impact of three-judge panels on district court decisions is discussed below.)

Again, state supreme courts are shown to be unaffected, with Brown increasing only slightly, and insignificantly, the likelihood of a pro-minority decision. Although falling short of traditional levels of statistical significance, region appeared to have influenced state supreme court decisions to a degree, particularly in the peripheral and deep South, where pro-minority decisions were substantially less likely than in non-Southern states.¹⁴ No similar regional effect appeared at the federal level. Still, an elaboration of regional influence on state and federal decisions in the pre- and post-Brown eras is called for.

Table 3 presents this analysis. These results mirror the findings in Table 2 that federal district courts, but not state supreme courts, were responsive to Brown. In addition, they highlight the comparatively greater degree of regional variation among the state tribunals. Before and after the Brown decision, non-Southern state supreme courts were about as likely to support civil rights as not, while minority claims had at best a one in four chance of prevailing in Southern state supreme courts. However, regional variation within the federal district courts was almost nonexistent, with both non-Southern and Southern venues displaying remarkable similarities within the pre- and post-Brown periods.

Table 3. Region's Influence on Pro-Minority Decisions in Civil Rights Cases

In short, the analysis tells a simple story that clarifies the historical record. The Brown opinion mattered a great deal, but only at the federal district court level. Within these federal courts, it even appeared to neutralize the potential influence of region. The independence from majoritarian prejudice and recalcitrance that southern district courts mustered in light of the new Supreme Court doctrine is aptly reflected in the opinion in Hall v. St. Helena Parish School Board (197 F.Supp. 649, 1961). In that challenge to an official effort in Louisiana to avoid the strictures of Brown by transforming public schools into private, the court cautioned that “this is not the moment in history for a state to experiment with ignorance” (659).

Although we have certainly seen that the federal district courts responded differently than state supreme courts, and that the trigger to this was most definitely the Supreme Court decision, some of the more subtle questions associated with the bulwark thesis may also be considered here through one final test: an analysis of federal appellate court actions on this same set of cases. This may help settle whether the theory's expectations should be consistently applied to all federal court levels.

In this vein, we will first discuss in more detail the impact of the three-judge panel variable in Table 2. Although falling short of statistical significance, this control is too close to dismiss out of hand. Clearly, the inclusion of circuit court personnel exerted a liberalizing influence on district court outcomes. But, notably, the inclusion of this variable fails to mitigate the strength of the precedent itself, thus signaling that district court judges did not need appellate judges to spur their adherence to Brown. While we do not engage in an individual-level analysis, it is useful to point out here that dissenting votes on these three-judge panels were rare (two dissents out of twenty-four panels), thus precluding the possibility that the high levels of district court support for minority claims post-Brown masked significant levels of individual-level recalcitrance.¹⁵

More important, we also investigated the outcome of all federal district court cases in this data set that were subsequently considered at the appellate level. Our purpose here was to examine the extent to which Brown's effect on these courts resembled the effect on the district courts. Let us consider the potential outcomes here to clarify our curiosity on this point. Certainly, no reasonable observer of history (or our findings to this point) would suggest that the federal appellate courts were likely to have been resistant to the 1954 mandate. But our question concerns a subtle issue involving the nature of this legal arena and the basic precepts of the bulwark theory: to what extent can we separate a bulwark protection of minority constitutional rights from the more prosaic tendency to heed Supreme Court precedent? The difficulty of assessing reaction to Brown in this regard is that it contains both elements—it is a precedent that supports minority rights. Through a pre- and post-Brown test, however, our primary state supreme court/district court comparison made a key clarification: minority rights were no better protected in federal district courts until the Supreme Court spoke on this matter. By extending this analysis to the federal appellate level, we can ascertain whether these courts similarly required the Supreme Court stimulus.

We conducted a simple test of the impact of Brown on federal appellate court decisions on cases from our district court data set. The results are illuminating. Before Brown, appellate-level support for minority claims was 77%, as compared to 38% for district courts in that same time frame. Still, Brown exerted a significant impact on these already strongly pro-minority tendencies; a difference in means test reveals that the precedent increased the appellate courts' already high pro-minority tendency to a certainty: ρ=1.0 (t=2.59, significance=0.01).¹⁶ The suggested scenario then is clear. Of the three courts examined here, only the federal appellate level demonstrated the bulwark's theorized commitment to protection of the constitutional rights of minorities in the absence of Supreme Court precedent.

Conclusions

These findings should first of all settle the short-term, historical question of federal district versus state supreme court reception of the Brown precedent and allow scholars to consider subsequent years with a more solid grounding in the immediate impact of this decision. We have provided empirical support for the conventional (albeit not universal) wisdom that the federal district courts more faithfully adhered to the mandate, while putting to rest the alternative assumption that all southern courts, whether federal or state, resisted. At the same time, our findings indicate that both federal district and state supreme courts were dismal venues for the remediation of minority discrimination before May 17, 1954.

Limitations apply to these conclusions, of course. Given the emphasis on constitutionally based challenges, the outcome for a broader universe of claims may differ, although it is hard to imagine why the patterns we uncover would differ in regard to nonconstitutional claims. A more important caveat is to keep the conclusions in historical perspective. This study represents a narrow test of the Brown precedent and leaves open the question of the impact of the expanding weight of numerous Supreme Court precedents after 1954. As Giles and Walker state, “[m]uch of what we know about the first decade post-Brown may not be applicable in the second decade following that decision” (1975:922). In later years, a variety of factors likely altered the outcomes.¹⁷

The second contribution of this inquiry is its emphasis of a significant but subtle aspect of the bulwark theory hinted at but not fully embraced by existing literature. In short, blunt assertions of broad variation between state and federal courts are too simplistic. While our analysis did identify an apparent difference between federal district courts and state supreme courts in regard to precedent, a difference that dwarfed the impact of region, our findings also highlighted a divergence between the federal district and appellate levels.

This suggests two considerations that should be considered in future studies of federal/state parity. First, conclusions about one level of federal courts cannot be assumed to extend to other levels. In this regard, recall Haas's (1982) caution that researchers reporting no differences between federal trial-level and state appellate-level courts may be missing significant variation that becomes apparent only when similar levels are compared. Our findings contribute an additional layer to this warning. While we did uncover a significant divergence in federal trial-level versus state appellate-level outcomes, it would have been a mistake to leave the comparison there, since the inclusion of circuit court data showed the federal appellate level to differ even more. In short, we found three distinct patterns of outcomes for the three courts we studied.

Furthermore, it is essential to avoid conflating two potentially distinct proclivities: protection of minority rights and adherence to Supreme Court precedent. This study of the impact of a minority protecting precedent indicates notable dissimilarities. State courts declined to protect minorities even after precedent mandated it, federal district courts protected minorities only after precedent mandated it, and federal appellate courts were protecting minorities before precedent mandated it.