The Resegregation of Public Schools ? Examining Parents Involved in Practice

This study investigated the efficacy of race-neutral student assignment policies following the 2007 Supreme Court decision in Parents Involved. Highlighting one urban school district—Chicago Public Schools—we examined differences in racial composition at their elite, “selective enrollment” high schools before and after voluntary race -based policies became unconstitutional. Using repeated measures ANOVA, we found the transition from racial to socioeconomic criteria have resegregated these schools— significantly reducing Africanand Asian-American enrollment. We argue the Supreme Court’s decision to strike down voluntary race-based student assignment policies has contributed to local policy changes for urban districts like Chicago, reducing minority access and opportunity.

Unlike most districts, CPS has a contentious history of racial and academic inequality (Luppescu et al., 2011).On one hand, the district was previously under court-ordered desegregation-from 1980-2009 (United States v. Board of Education of the City of Chicago, 1980).On the other hand, former Education Secretary William Bennett once called the district the "worst in the nation" (Rury, 1999, The New York Times, 1987).These issues gave birth to CPS' choice reform movement in the late 1990s and early 2000s-i.e.charter schools and elite, "selective enrollment" high schools (Spirou & Judd, 2016).Together, they were thought to address inequalities, especially for minority students who made up 90% of the student population (CPS, 2018a).Nearly two decades later, Chicago's charters remain on par with their traditional counterparts (Booker, Gill, Zimmer, & Sass, 2008) while selective enrollment schools have become the district's and state's premier academic performers (Quiroz & Lindsay, 2015).Post-secondary matriculation (CPS, 2018a;Quiroz & Lindsay, 2015) and degree completion (CPS, 2018a) are significantly higher in selective enrollment schools compared to CPS' neighborhood schools.Using repeated measures ANOVA to compare percent student of color matriculation in selective enrollment vs. open enrollment high schools, however, we demonstrate their resegregation.Since PICS, African-and Asian-American enrollment has significantly reduced, illustrating its consequences for minority student access and opportunity.
We first unpack the case's arguments and rulings.Second, we examine the subsequent precedent's impact on K-12 school districts in practice, highlighting CPS as a case.Third, we articulate a conceptual framework focusing on: a) why PICS marked a distinct shift in K-12 desegregation jurisprudence since Brown-from "race-based" to "race-neutral"; b) how this shift has affected K-12 affirmative student assignment policies; and c) why such policy tools in American education remain a compelling government interest.Fourth, seeking to empirically examine this framework in practice, we detail our study design, methods, and data sources.Fifth, we demonstrate our analytical findings, their implications, and conclusions.

Setting the Stage: Parents Involved v. Seattle
While the dismantling of K-12 desegregation law had its beginnings in the early 1990s (Armor, 1995;Lutz, 2011;Orfield & Eaton, 1996), PICS became the official "pivot point" in practice (Fischbach et al., 2008).Indeed, student assignment policies have historically served as tools for addressing racial segregation (Frankenberg, 2017).For example, Brown had translated into many school districts using race as a criterion for such policies (Frankenberg, 2017;Glenn, 2015).But in PICS, the question became whether school districts not under court-imposed desegregation orders were violating the Equal Protection Clause.The plaintiff argued Seattle's use of a "racial tiebreaker" for balancing ethnic diversity was unconstitutional.Conversely, the defendant argued such measures equitably diversified their student body.The Supreme Court thus had to decide whether local school boards could lawfully: 1) use an individual student's race in determining public school assignments, and 2) voluntarily integrate districts on such bases.
Casting the tiebreaker vote, Chief Justice Roberts argued while "racial diversity" is a compelling government interest, "racial balancing" as exemplified by Seattle violated the Equal Protection Clause (PICS, 2007).He stated: "The way to stop discrimination on the basis of race is to stop discriminating on the basis of race" (PICS, 2007).However, we must point out Justice Kennedy said schools could "consider the impact a given approach might have on students of different races" (PICS, 2007) so long as they did not "involve decision-making on the basis of any individual student's race" (U.S.Department of Education, 2011, p. 6).Such legal approaches included: 1) students' socioeconomic status; 2) parental education; 3) students' household status; 4) neighborhood socioeconomic status; 5) geography; or 6) composition of area housing (U.S.Department of Education, 2011).In this way, K-12 desegregation law and policy under the Court evolved (Tushnet, 2016)-specifically the criteria by which such efforts could be achieved (Brown & Yi, 2008).Without evidence of de jure segregation (i.e.court-ordered), race-based student assignment was no longer constitutional (Browne & Yi, 2008).This decision and related guidance therefore had significant consequences "for the schools, for the democratic process, and for America's efforts to create, out of its diversity, one Nation" (Justice Breyer's dissenting opinion, PICS, 2007).We further explain these consequences in practice, highlighting Chicago Public Schools as a case.

Examining Parents Involved in Practice: Chicago Public Schools
Similar to other school districts nationwide, considerable confusion existed as to what strategies were legally permissible (Frankenberg, 2018; US Department of Justice & Department of Education, 2011).Coupled with the financial pressures brought on by the Great Recession (Coffee & Frankenberg, 2009), McDermott and colleagues (2012) argue PICS resulted in "boundary setting effects".That is, school districts without new or ongoing litigation rolled back race-based student assignment policies altogether, implementing "the easiest and safest response" (The Civil Rights Project, 2008, par. 7).Thus, the Court became a powerful mechanism for influencing such policies employed by local districts and schools generally (Superfine, 2010(Superfine, , 2013)).
To empirically illustrate these local policy responses in practice, we highlight one urban school district-Chicago Public Schools (CPS).CPS was chosen as a critical case for several reasons.First and foremost, Chicago has a long-standing history of racial segregation-both residentially and educationally (Lipman, 2008;Lipman & Haines, 2007;Rury, 1999).Prior to PICS, the district was under a court-ordered desegregation decree .This included their selective enrollment high schools (United States v. Board of Education of the City of Chicago, 2004) whereby no more than 35% of the student body could be White.Post-PICS, however, CPS started modifying its affirmative student assignment policies, reflecting the Court's race-neutral approach to diversity.Seeking to be a policy tool innovator for its selective enrollment high schools, the district produced what colloquially has been referred to as the "Tier system" (see next section for details).Towards these ends, CPS represents both the voluntary and involuntary retreat of race-based student assignment polices in practice.
Second, these admissions-based high schools (N = 11) are open to all Chicago students (providing a representative sample) and uniformly considered the best the city has to offer.US News & World Report (2017) ranks 6 of them in Illinois' "top 10," whereas students achieve, on average, 35 percentile points higher on the ACT nationally then those enrolled in neighborhood CPS schools (Chicago Public Schools, 2018a).And given more than 90% of these students (regardless of race) attend college, they also "serve as [a] gateway to college matriculation" (Quiroz & Lindsay, 2015, p. 379).
Third, Chicago reflects current policy and political conditions that research shows are contributing to K-12 resegregation post-PICS.The district has adopted socioeconomic status as a proxy for promoting racial diversity (Century Foundation, 2016;Frankenberg, 2018).Yet scholars show the relationship between race and income varies within communities, limiting their intended efficacy (e.g.Ellison & Pathak, 2016;Frankenberg, 2018;Harwell, 2018).At the same time, the city has recently been plagued by district corruption (Perez, 2017a), police brutality towards people of color (Madhani, 2017), and inequitable education funding (Perez, 2017b).Diem (2017) argues such political variables present contextual challenges when voluntarily designing, implementing, and accepting race-neutral student assignment policies under a post-PICS landscape (also see Diem, Cleary, Ali, & Frankenberg, 2014).
Taken together, we empirically show how Chief Justice's Roberts' decision to strike down voluntary, race-based student assignment policies has contributed to the resegregation of CPS' selective enrollment high schools.Employing repeated measures ANOVA, we ask the following research questions: 1) To what extent have racial demographics in Chicago's selective enrollment high schools changed in comparison to its open enrollment schools since the introduction of socioeconomic status as a proxy for individual student race in admissions policy?2) To what extent have racial demographics changed in the highest performing selective enrollment high schools compared to lower performing selective enrollment high schools throughout the implementation of socioeconomic status as a proxy for student race in admissions policy?
We will further unpack these research questions in the methods section.Below is the conceptual framework we use to examine them.

Conceptual Framework
At its nexus, Parents Involved became a symbol for pushing back against race-conscious diversity policies (McDermott et al., 2012).The legal rationale that government has a compelling interest in maintaining non-discrimination has thus largely replaced the legacy of Brown and its goal of eliminating racial segregation (Frankenberg, 2017;McDermott, 2001).According to Bonilla-Silva (2004, 2010), this socio-political shift has become the "new racism," whereby race-neutral policies are resegregating populations of color in favor of concentrating White populations.This shift is also true for elite schools (Henig, Hula, Orr, & Pedescleaux, 1999;Orfield, 2018;Orfield & Ayscue, 2018).We argue such critical perspectives can be traced to Supreme Court jurisprudence in K-12 desegregation law.
Early on, the Court demonstrated "command-and-control" (Superfine, 2010).No ruling exemplifies such a position more than Chief Justice Earl Warren's order to desegregate "with all deliberate speed" (Brown v. Board of Education II, 1955).However, over the next few decades, this authoritative approach softened, "limiting the scope of student assignment remedies" (Superfine, 2010, p. 116).This is particularly true since the early 1990s, beginning with Board of Education v. Dowell (1991) and continuing under PICS.Consequently, Thro and Russo (2009) contend a fundamental change in the Court has transpired-the core means by which "diversity" is achieved.
Indeed, local districts and schools can no longer voluntarily use individual race as a means to promote diversity; rather, they must implement "proxy" student assignment policies like assigning students by SES.Equally, enforcement of Consent Degrees is diminishing-or being lifted entirely-in an effort to promote Chief Justice Roberts' call for "non-discrimination" (McDermott et al., 2012).Accordingly, PICS becomes a constitutional "pivot point" in practice, highlighting the "schizophrenic identity" of the Supreme Court over time (Fischbach et al., 2008) and its influence on K-12 resegregation (Siegel-Hawley et al., 2018).
To conceptually unpack this evolution in Court desegregation law/policy and its influence on K-12 resegregation, we utilize policy regimes theory.Rooted in political science epistemology, policy regimes theory illustrates how, over time, ideas, interests, and institutions diverge, contributing to changes in policy paradigms and subsequent tools (Cohen-Vogel & McLendon, 2009;Esping-Andersen, 1987;Mabbett & Bolderson, 1999;May, Sapotichne, & Workman, 2009;McGuinn, 2006;Wilson, 2000).Researchers have used it in many areas of domestic policy, including immigration (Wright & Bloemraad, 2012), environmental protection (Young, 1999), and education (Cohen-Vogel & McLendon, 2009;Nelson & Tyler, 2016;Superfine & Thompson, 2016).Unlike other frameworks, policy regimes theory argues change happens gradually (Superfine & Thompson, 2016).As gaps emerge between old regimes and new ideas, new policy paradigms may form.We therefore argue policy regimes theory appropriately reflects this progressive transition in Court jurisprudence, its influence on K-12 desegregation policy over time, and the increasing resegregation of public schools.
Drawing upon key Supreme Court cases, we will first show how regimes over time have combined to shift the diversity narrative from "race-based" to "race-neutral".Second-highlighting Chicago as a case-we will historically explain how this shift in turn has affected affirmative student assignment policies, increasingly resegregating local school districts.Third, we end this section discussing why using individual student race as a policy tool to promote K-12 diversity remains a compelling government interest.

From "Race-based" to "Race-neutral" in the Court: Moving Towards Resegregation
Although K-12 desegregation law essentially began with Brown, many of the policy tools arising from earlier regimes directly influenced the political conditions surrounding the landmark case.In 1865, slavery became illegal in the US.A year later, the Civil Rights Act of 1866 granted citizenship to all persons born in the country while the Fourteenth Amendment expanded the "definition of citizenship, equal protection under the law, and due process" (Conneeley, 2008, p. 98).Despite both laws seeking to address past injustices for African Americans, these new rights did not extend to public education (Conneeley, 2008).Policymakers, uneasy with these newfound freedoms bestowed on former slaves, thus began legislating "Jim Crow" laws.These laws specifically sought to separate White and Black citizens.To comply with the Equal Protection Clause, however, they mandated "separate but equal" treatment.
Under this doctrine, the Supreme Court accepted such policy tools.For example, Plessy v. Ferguson (1896) found separate facilities for Blacks and Whites were constitutional in so far as they were "equal".Yet by 1950, these tensions had resurfaced.Given segregated Black schools were found to be inferior to White schools in terms of "teacher quality, teacher pay, funding, and resources" (Conneely, 2008), the lawfulness of Jim Crow laws came into question.Under these circumstances, the Supreme Court agreed to hear Brown v. Board of Education (Brown I, 1954) and whether state-imposed segregation in K-12 public schools violated the Fourteenth Amendment.
Finding "separate educational facilities [were] inherently unequal" (Brown I, 1954), the Court unequivocally overturned the century-long "separate but equal" doctrine.In doing so, a new policy regime took shape, whereby the Court intervened in state/local affairs for the purposes of maintaining racial equality in public schooling (Superfine, 2010).School boards were now obligated to consider individual race when assigning students to schools (Owen, 2007).
By the 1970s, however, the Supreme Court showed signs of exhaustion concerning K-12 desegregation.Vietnam, the Watergate scandal, and President Nixon's subsequent resignation had left the American public stunned.Further, four new justices had been appointed during his tenure-Warren Burger (Chief), Harry Blackmun, Lewis Powell, and William Rehnquist.Collectively, the regime once again began to shift.Court decisions no longer garnered unanimous support, often with contentious dissenting opinions.No case illustrated this fact more so than Milliken v. Bradley (Milliken I, 1974).
In a divided opinion, the majority (led by Chief Justice Burger) found remedies extending across district lines must each demonstrate de jure segregation.As such, Detroit's multi-district desegregation plan was ruled unconstitutional.This decision broadly centered on the Court's position towards local control (Superfine, 2010) and the policy tools they implement (Armor, 1995).So, whereas the prior policy regime emphasized "command-and-control"-as evidenced by Green, Swann, and Keyes (Superfine, 2010)-Milliken I underscored a retreat by the Court.Integrationminded school officials were thus somewhat constrained when enacting race-based student assignment policies (Nelson & Tyler, 2016).These circumstances were echoed in Justice Thurgood Marshall's dissenting opinion, declaring Milliken I "a giant step backwards" concerning the effort to integrate America's public schools.
This paradigmatic shift under the Court continued in the early 2000s.For example, they refused to hear such cases as Belk v. Charlotte-Mecklenburg Board of Education (2001) in which the lower courts invalidated the use of busing to rectify segregated schooling.Yet the 2007 ruling in PICS was the true "pivot point" (Fischbach et al., 2008) for K-12 desegregation law in practice.On one hand, the decision demonstrated the Court's shifting position from "race-based" to "race-neutral".On the other hand, the decision signified a move towards K-12 resegregation.Taken together, Court jurisprudence from the early 1970s through 2000s simply cultivated a landscape by which such a shift could transpire.
As Superfine (2013) argues, Court rulings do not take place in a vacuum; they are subject to the greater socio-political landscape, influencing the kinds of laws and tools a given regime subscribes to.From this perspective, the Courts' shedding of "command-and-control" (Superfine, 2010), the personal political preferences of judges appointed since the 1970s (Pullin, 2007), and a return to local control since the Reagan Administration (Superfine & De Voto, 2016) had combined to change the narrative surrounding desegregation in America's schools.Policy regimes theory thus illustrates this gradual epistemological change in the means by which racial equality and diversity are viewed under the Equal Protection Clause (from Brown to Parents Involved); whereas Brown promoted race-based student assignment, PICS now stresses a race-neutral approach.
At the same time, research is beginning to show the Court's limitations on race-based student assignment policies are resegregating local districts (e.g.Biegel, 2008;Diem et al., 2014;Frankenberg, 2018;Orfield et al., 2014;Siegel-Hawley et al., 2018), including their elite, admissionbased schools (Orfield, 2018;Orfield & Ayscue, 2018).Still, some scholars argue such resegregation is not tied to law or policy, but rather uneven distribution of students between school districts in the same area (e.g.An & Gamoran, 2009;Bischoff, 2008;Fiel, 2013Fiel, , 2015)).Towards these ends, we highlight Chicago Public Schools (CPS) as a critical case.The district represents 11 selective enrollment high schools (open to all 77 neighborhoods), was previously under court-order (pre-PICS), and has a host of racially segregated, proximal neighborhood schools from which to empirically examine such scholarly disagreements.Below, we detail a history of CPS's desegregation efforts and their student assignment policies before and after PICS.

History of Desegregation in Chicago Public Schools Before and After PICS
Like many urban metropolises, Chicago has historically experienced neighborhood segregation (Lipman & Haines, 2007;Mirel, 1993;Rury, 1999).Broadly, these circumstances are the outcome of past discriminatory housing policies (Chicago Board of Education, 1999;Fron, Poulos, & Schnieder, 2013).In turn, CPS has remained quite segregated (Jankof & Caref, 2017).For example, in 1956, 71% of their high schools were made up of a single race (Chicago Board of Education, 1999).Despite recognizing such issues-and the subsequent constitutional violations underscored by Brown-superintendents were largely unable to address them (Rury, 1999).As a result, many minority students remained in highly segregated schools with access to limited resources and opportunities (Kirby, Harris, Crain, & Rossell, 1973).
By the spring of 1979, however, the federal government decided to step in.Accusing CPS of continually supporting segregation in its schools, they demanded the district submit a comprehensive desegregation plan or face legal action.When their plan was deemed unsatisfactory in April of 1980, the Department of Justice found CPS' Board of Education guilty of unlawful segregation (see United States v. Board of Education of the City of Chicago, 1980).Seeking to avoid further legal action, the district entered into a Consent Decree.Under this Decree, they were required to implement a desegregation plan.With support from the Court, this plan mandated the district to: 1) desegregate its schools; 2) provide compensatory programming for any schools remaining segregated; 3) maximize integration across all racial or ethnic groups; and 4) make sure "the burdens of desegregation [were] not imposed arbitrarily" (United States v. Board of Education of the City of Chicago, 1983).Such measures became compulsory until they met "unitary status" (see Green v. County School Board, 1968).
That said, after two decades, the Court acknowledged CPS had not yet reached compliance in several areas-particularly diverse student enrollment in their magnet and selective enrollment high schools (Quick, 2016).As a result, the city entered into a new, Modified Consent Decree (see United States v. Board of Education of the City of Chicago, 2004).Building on the former plan, it mandated the classification of students as either "White" or "Minority" (making no distinctions within minority category).Further, it required selective enrollment high schools award no more than 35% of their seats to White students.The belief was this modified plan ensured a more equitable admissions process for students of color across many of these elite, "college-gateway" high schools.
Indeed, where the original Consent Decree had failed, the Modified plan was successful.During this time, none of these high schools enrolled more than 35% White students (CPS, 2018a).Further, "minority" enrollment-i.e.African American, Hispanic, Asian, and Native Americanreached all-time highs (CPS, 2018a).Just a year later, however, the decision in PICS underscored a marked shift in the use of such race-based policy tools-both for districts under court-order and those voluntarily doing so.
Almost immediately, CPS asked the Court to lift its Modified Consent Decree, which was granted (see United States v. Board of Education of the City of Chicago, 2009).Subsequently, the district needed to formulate an admissions policy for its selective enrollment high schools that did not "not raise similar constitutional concerns" (Hopwood v. Texas, 1996, p. 946).Like many local school districts, Chicago defaulted to socioeconomic status (SES) as a proxy policy tool (Conneely, 2008).At the time, 64% of Black children and 44% of Hispanic children were living in poverty, compared with 13% of White children in the US (U.S. Census Bureau, 2010).Beginning in 2009, incoming ninth graders were thereby subject to an admissions system utilizing SES instead of individual race (Chicago Board of Education, 2017).
However, it is important to note CPS wanted to be innovators of such proxy policy tools.In the process, the district convened a "Blue Ribbon Committee" (Chicago Board of Education, 2017) to evaluate the prior year's results and hear concerns by parents, educators, and community members.Its recommendations were to: 1) not reinstate the 5% principal discretion policy; 2) move to a multi-year policy; and 3) institute a class-based "Tier process" (Chicago Public Schools, n. d.), which the Board accepted (Chicago Board of Education, 2017).Because SES varies greatly in Chicago's often-segregated neighborhoods, the district believed such a system would "ensure equal access and equity in the provision of magnet and selective enrollment schools" (Chicago Public Schools, 2018b).
These "Tiers" were specifically tied to US Census data (Quiroz & Lindsay, 2015) and designed to mimic the Modified Consent Decree.The Chicago Board of Education designated five socioeconomic characteristics of a given neighborhood: 1) median family income; 2) percentage of single-parent households; 3) percentage of households where English is not the first language; 4) percentage of homes occupied by the homeowner; and 5) level of adult education attainment (CPS, 2018b).A composite score was then formulated for each of the city's 800-plus census tracts, separated into four distinct groups-or "Tiers".Each Tier consisted of approximately the same number of school-age children.70% of seats are now determined via these Tiers.
Under this method, students living in "Tier 1" neighborhoods have a decided advantage over those living in "Tier 4" neighborhoods as they are not required to earn the same admissions score (CPS, 2018b).Based on a 900-point scale, the district combines an individual students' selective enrollment test scores, annual standardized tests scores, and academic grades, adjusting the "cut-off" scores across each Tier.Students from lower socioeconomic Tier designations are thereby given priority over those from more affluent neighborhoods.For a neighborhood map of these Tiers across the city, please refer to Appendix A.
At the same time, in order to "raise academic standards" across such elite schools (Chicago Board of Education, 2017, p. 1), the Blue Ribbon Committee also recommended that 30% of seats be allocated to students with the highest academic performance citywide-regardless of SES or Tier.Initially, this became a point of contention for families of color and students residing in "lessaffluent" Tiers (Quick, 2016;Siegel-Hawley, 2010;Yednack & Little, 2009).Mayor Rahm Emanuel had recently closed 49 schools in predominately minority-populated areas (Belsha, 2017;Lee & Lubienski, 2016;Weber, Farmer, & Donoghue, 2016).Equally, the most popular option among White students was to attend a selective enrollment high school (Barrow & Sartain, 2017).Given these 800-plus census tracks were consolidated across Chicago's 77 neighborhoods, students from wealthier "pockets" in gentrifying neighborhoods could thus be grouped into the same Tier as students from high poverty communities (Golab & Rossi, 2010).As a result, high-SES and low-SES students often compete for a seat since wealthier students tended to perform better on both the selective enrollment test and annual standardized test (CPS, 2018a).
In sum, Chicago represents the influence of Court desegregation jurisprudence on student assignment policy over time.Under the auspices of Brown, CPS was required to desegregate its schools.During this period, their elite, selective enrollment high schools became more racially diverse.However, now operating under a race-neutral regime (post-PICS), we will show that the district's "Tier system" has not been as efficacious towards such ends.Accordingly, we contend policies subscribing to individual student race as a means to promote ethnic diversity remain a compelling government interest.
Coupled with minority student enrollment being three times what it was a half-century ago (Orfield, 2001; U.S. Census Bureau, 2010), improving the racial diversity of public schools thus becomes a compelling government interest (Kurlaender & Yun, 2001).But the US Commission on Civil Rights (2018) finds "the reality of American schooling is fundamentally inconsistent with the American ideal of public education operating as a means to equalize life opportunity, regardless of zip code, race, economic status, or life circumstance" (p.10).Moreover, they find "low-income students and students of color are often relegated to low-quality school facilities" (p.10).So, while elite schools afford such equitable opportunities (Finn & Hockett, 2012), the current policy regime constrains their access.Towards these ends, this paper will empirically investigate such issuesparticularly for urban districts (Ravitch, 2011)-and the extent to which PICS may have played a role.It will also challenge Reardon and Owens' (2014) claim that "the last 25 years [of]...school racial segregation has changed little" (p.213), posing "convincing evidence about the importance of racial integration in K-12 schools" (Frankenberg & Garces, 2007, p. 24) in a post-PICS, race-neutral landscape.

Sample
As discussed, CPS was selected as a single case for studying the efficacy of race-neutral student assignment policies following PICS.Specifically, we used critical case sampling (Patton, 2002), examining the district's selective enrollment high schools (N = 11).Patton (2002) argues while studying one case does not yield broad generalizations, it does, however, afford logical generalizations from the weight of evidence produced.According to overall academic achievement indicators (CPS, 2018a), these schools are widely considered the best public high schools in the district (Lindsay & Quiroz, 2015) and state (U.S. News & World Report, 2018).CPS has also sought to be a policy innovator, producing one of the most comprehensive race-neutral student assignment strategies in the nation (Ellison & Pathak, 2016).Finally, because these schools are admissions-based secondary options, all incoming ninth graders across Chicago's 77 neighborhoods can apply, providing a representative sample from which to examine the potential impacts of such shifting policy regimes on K-12 resegregation (Patton, 2002).
On the other hand, to help account for changes in overall community demographics that dissenting scholars (e.g.An & Gamoran, 2009;Fiel, 2013Fiel, , 2015;;Reardon & Owens, 2014) contend are responsible for K-12 resegregation, we also selected a comparison group of 19 open enrollment neighborhood CPS schools.These open enrollment schools were selected based on geographic proximity (as determined by postal code) to one or more of the selective enrollment high schools included in our study.Geographic proximity was used to select control schools for two different reasons.In CPS, student assignment to open enrollment schools is the default strategy if they are not enrolled in a choice school and is based on geographic proximity using a neighborhood boundary system (see Chicago Public Schools, 2018c).Thus, open enrollment schools have student racial/ethnic compositions that are similar to neighborhood demographic composition.In the case of CPS selective school demographic enrollment patterns, such neighborhood demographics are an important consideration due to transportation constraints on parent/guardian choice.CPS does not provide transportation to its selective schools, and transportation and its associated time and cost have previously been shown to be an important constraint on parent/guardian school choice decisions (Bell, 2007(Bell, , 2009;;Blank, Levine, & Steel, 1996;Frankenburg, Lee, & Orfield, 2003).Purposively selecting control, open enrollment schools that are geographically proximal to the selective schools therefore was a way of accounting for transportation constraints on parent/guardian choice.These schools were determined using the "school locator tool" found on the CPS website (Chicago Public Schools, 2018d).

Data Sources
We obtained selective high school enrollment by student race data from CPS' Department of School Quality Measurement database (CPS, 2018a).We also obtained such data for our 19 open enrollment CPS schools using the National Center for Education Statistics (NCES) Elementary and Secondary Information System (ELSi).NCES public school data is taken from the Common Core of Data (CCD) surveys that are conducted annually and collect general descriptive information, student and staff characteristics data, and fiscal data from all public schools in the US.This information can be accessed via the ELSi platform from 1986-1987 to the most recent CCD administration (see National Center for Education Statistics, ELSi, 2018).The selective school enrollment was also contained in the ELSi reports because the selective schools are classified as public schools in the CCD.ELSi data were thereby used to confirm the CPS Department of School Quality Measurement database, showing an exact correlation with the CPS reported enrollment data for the selective schools in the study.However, had a discrepancy been present, ELSi data for selective schools would have been substituted for CPS data since the CCD is a consistent form of federal reporting.

Analysis
To examine changes in student enrollment by race between CPS selective enrollment high schools and neighborhood open enrollment schools pre-and post-PICS, we conducted a two-way repeated measures ANOVA.Several key time-points were chosen: 2006 (i.e.just prior to the decision), 2012 (i.e. the first policy implementation year of CPS' Tier system), and 2016 (i.e. the most current year for which NCES demographic data was available).Our inclusion of 2016 in the analysis was important for two reasons.First, a four-year span between policy articulation and implementation allows for a lag time between policy implementation and policy effects on enrollment.Second, CPS' selective high schools serve grades 9-12.Therefore, a four-year passage of time between demographic measures would minimize the possibility of students who were grandfathered under the previous selection policy at the time of measurement.On the other hand, comparing enrollment from 2006 to 2012 and 2012 to 2016 allowed us to track the changes in student enrollment between pre-policy and policy initiation and policy initiation and a full four-year, or whole school implementation, change in enrollment.
Our analysis was carried out using the General Linear Model repeated measures function in SPSS.Five separate analyses were conducted using percentage African American enrollment, percentage Asian enrollment, percentage Hispanic enrollment, percentage Native American enrollment, and percentage White enrollment as dependent variables, with selective or open enrollment schools used as the between-schools factor.
In addition to comparing student enrollment by race between CPS' selective high schools and open enrollment schools, we also compared these selective schools separately (see research question 2).This supplementary analysis was conducted because of differences in their academic performance (CPS, 2018a) and popularity across SES Tiers (Barrow, Sartain, & de la Torre, 2018).Specifically, we used multiple performance indicators, including: 1) CPS school quality ratings; 2) percent acceptance rates; 3) percent college enrollment; and 4) U.S. News and World Report's national school ranking data (CPS, 2018a; U.S. News & World Report, 2018).Based on these criteria, we then identified two distinct groups of selective enrollment high schools: "Upper-tier" performers and "Lower-tier" performers.Whitney M. Young Magnet, Northside College Prep, Walter Payton College Prep, and Jones College Prep high schools were classified as Upper-tier performers, whereas Lindblom Math and Science Academy, Hancock, Martin Luther King, and Westinghouse high schools were classified as Lower-tier performers.Several sample schools were excluded, however, because they were either not open across all measured time-points or were too similar as defined by the criteria above (thereby skewing the results).Due to the further reduced sample size (and power), this comparison was made using descriptive statistics and visual inspection of longitudinal graphs.

Research Question 1
The overall ANOVA comparing White student enrollment between selective and open enrollment schools was not significant.However, given our sample size was small due to examining a specific case of race-neutral student assignment policies, individual mean differences in enrollment were still examined where the overall ANOVA was not significant.At all three time-points, White student enrollment was significantly higher at selective schools compared to open enrollment schools, but there was no significant change in either school type across time.
The results for enrollment of students of color in selective schools compared to open enrollment schools from 2006-2016 can be compared to these patterns of change for White students (see Figure 1).The overall ANOVA comparing African American student enrollment in selective schools and open enrollment schools from 2006 to 2016 was not significant (F (2, 26) = 2.00, p = 0.16, partial   , observed power = 0.376), and there were no significant mean differences in African American student enrollment between selective and open enrollment schools at any single time-point.In contrast to White student enrollment, however, we found African American student enrollment at selective schools significantly decreased from 2012 to 2016 (Mean Difference = 4.036, p < .05).There was no significant change in African American student enrollment from 2006-2012, so this change is temporally related to the district's race-neutral "Tier system" policy.Lastly, there was a near significant decrease in African American student enrollment in selective schools from 2006-2016 (Mean Difference = -7.880,p < .10),but this overall decrease is driven by the decrease found during 2012-2016.
The overall ANOVA comparing Asian student enrollment between selective schools and open enrollment schools from 2006 to 2016 was significant (F (2, 26) = 4.571, p < .05,partial   , observed power = 0.725).However, there were no significant differences between selective and open enrollment schools at any single time-point.There was a significant decrease in Asian student enrollment in selective schools from 2006-2012 (Mean Difference = -7.780,p < .001)and a significant decrease in Asian student enrollment in selective schools from 2006-2016 (Mean Difference = -8.200,p < .001).There was no significant change in Asian student enrollment at CPS' selective high schools just following "Tier system" implementation in 2011-2012.There were also no significant changes in Asian student enrollment across all time-points in open enrollment schools.
The overall ANOVA comparing Hispanic student enrollment between selective schools and open enrollment schools from 2006-2016 was not significant.At all three time-points, there was no significant difference in enrollment of Hispanic students between selective and open enrollment schools.There were near significant increases in Hispanic student enrollment in selective schools from 2006-2012 (Mean Difference = 4.937, p < .10)and between 2006-2016 (Mean Difference = 6.890, p < .10).However, there were no significant differences in Hispanic student enrollment at selective schools between 2012-2016 (i.e. during "Tier system" policy implementation).There were also no significant differences in Hispanic student enrollment in open enrollment schools between any time-points.
The overall ANOVA comparing Native American student enrollment between selective schools and open enrollment schools from 2006-2016 was significant (F (2, 26) = 7.465, p < .01,  , observed power = 0.914).There were significant differences between Native student enrollment between selective and open enrollment schools in 2006, with selective schools enrolling more Native American students than open enrollment schools (Mean Difference = 0.265, p < .01).There were also significant differences in Native American student enrollment in 2016, with selective schools enrolling fewer Native American students compared to open enrollment schools (Mean Difference = -0.198,p < .05).Conversely, there was a significant decrease in Native American student enrollment from 2006-2012 in selective Schools (Mean Difference = -0.330,p < .01).However, there was no significant difference in Native American student enrollment at selective schools from 2012 to 2016 (i.e. during "Tier system" policy implementation).Taken together, this result indicates there was an existing Native American enrollment disparity in CPS between selective and open enrollment schools, but that disparity was not significantly exacerbated by the "Tier system" policy.See Table 1 for numerical changes in enrollment followed by Figure 1 for an illustrative version: Note: The race-neutral, "Tier system" policy was fully implemented in the 2011-2012 academic year.
Research Question 2 In addition to comparing student enrollment by race between selective and open enrollment schools, we compared Upper-and Lower-tier performing selective schools.Both Upper-tier and Lower-tier selective schools had nearly the same decrease in African American student enrollment from 2012 to 2016 (~4-5%).Lower-tier performers, however, had a larger decrease compared to Upper-tier performers from 2006 to 2012 (~9% compared to ~2%).Lower-tier selective schools also had a larger increase in Hispanic student enrollment compared to Upper-tier selective schools during this same time period (~11% compared to ~2%).Conversely, Upper-tier selective schools had a larger increase in White student enrollment from 2012 to 2016, whereas Lower-tier selective schools exhibited a decrease during this same time period.Taken together, we found Upper-tier selective schools enrolled more White students and fewer students of color compared to Lower-tier schools across all years, particularly African-and Asian-Americans post-PICS (see Figure 2   Note: The race-neutral, "Tier system" policy was fully implemented in the 2011-2012 academic year.

Discussion
Although Chicago's selective enrollment schools were established in part to address minority student access and opportunity, our findings show they have resegregated since Parents Involved.On one hand, African American enrollment has significantly decreased.These findings support research showing African Americans typically attend the lowest-performing K-12 public schools (Henricks, A. E. Lewis, Arenas, & D. G. Lewis, 2017;Rothwell, 2012).On the other hand, descriptive statistics (see Figure 2) show Asian American enrollment at Upper-tier selective high schools has also significantly decreased.This may be considered surprising given Asian students academically outperform other minorities (Joo, Reeves, & Rodrigue, 2016) and attend better schools generally (Rothwell, 2012).But further examination of Asian American student ethnic groups in CPS shows a troubling bimodal achievement pattern.While the overall percentage of Asian American students in CPS achieving a college degree is on par with White students, 13% of all Asian American Chicagoans do not attain a high school diploma-a rate similar to African American Chicagoans and twice as high as White Chicagoans.Chinese, Vietnamese, and Cambodian Chicagoans are particularly at risk for not attaining a high school diploma (Scarborough, Lewis, & Arenas, 2018).This study therefore highlights the need to consider some Asian ethnic groups in Chicago may already be impacted by an educational opportunity gap similar to African Americans (Henricks et al., 2017), potentially exacerbated by the race-neutral, "Tier system" policy CPS has adopted.
Taken together, we find these admission-based schools have become enclaves for Whites.Demographically making up just 10.2% in the entire district, Whites now make up for as much as 43.0%across them post-PICS.Indeed, this is a 10% increase from the number of seats accounted for during CPS' Modified Consent Decree-era (ending seven years ago).As discussed, these schools provide the best public education in the city and (for many) state.Post-secondary matriculation (CPS, 2018a;Quiroz & Lindsay, 2015) and degree completion (CPS, 2018a) remain significantly higher than open enrollment schools.The same is true for their funding (Ali & Cherone, 2016) despite research showing a significant and positive relationship between higher funded Illinois schools and postsecondary success (Houston, 2018).Accordingly, we present a case challenging Reardon and Owen's (2014) assertion the past several decades of K-12 racial segregation has changed little.Our findings are also congruent with other current policy work conducted in CPS showing White students are increasingly overrepresented across the district's most advantageous educational contexts-including private and selective public schools (Henricks et al., 2017).
At the same time, we acknowledge Chicago is just one critical case with a smaller sample of 11 selective enrollment high schools.Still, these findings are particularly concerning as it relates to minority access and opportunity.Broadly, they reflect the changing demographics of elite schools nationally (Finn & Hockett, 2012;Lawyers' Committee for Civil Rights and Economic Justice, 2017;Orfield, 2018;Orfield & Ayscue, 2018;Pathak & Sönmez, 2013;Shakarian, 2014); 90% of CPS students are minority (CPS, 2018a), most of which attend neighborhood schools with a high percentage of low-income students (69%; Bechteler & Kane-Willis, 2017).Yet like Whites, they tend to favor these selective schools (Barrow et al., 2018).In the process, only 14% earn a four-year college degree within 10 years of beginning high school (Healey, Nagaoka, & Michelman, 2014)many graduating from these selective enrollment high schools (CPS, 2018a).We correspondingly contend the Supreme Court striking down race-based student assignment policies has had a disparate impact on students of color across the city-particularly African-and Asian-Americans.These findings are supported by other anecdotal (Catalyst Chicago, 2013;Cox, 2014;Karp, 2010Karp, , 2012;;Loury, 2017;Novak & Fusco, 2014) and empirical sources (Ellison & Pathak, 2016;Jankof & Caref, 2017;Metzger & Lutton, 2014;Quiroz & Lindsay, 2015).
While Frankenberg (2017) demonstrates one urban school district (Jefferson County, Kentucky) successfully implementing a proxy student assignment policy, for most, this transition has been difficult-mainly due to politics, contextual factors, and historical issues (Diem, 2017;Diem et al., 2014;Frankenberg et al., 2015).For Chicago, we find such issues have resulted in design and implementation challenges.Likely, the "30% rule"-whereby 30% of selective enrollment seats are allocated strictly based on assessment performance (regardless of SES or Tier)-has played a key role, given Whites traditionally do better on such measures (CPS, 2018a).So in spite of their desire to be a policy innovator post-PICS, spending nearly $500,000 to create a progressive SES-based "Tier system," access to its selective enrollment high schools for students of color has declined.School closures (Belsha, 2017;Lee & Lubienski, 2016), gentrification (Siegel-Hawley, 2010), tierfraud (FitzPatrick, 2016;Matthews, 2016), and selective enrollment test preparation companies promising higher scores (Chanen, 2017) have further complicated these efforts.
However, we must affirm this study is not an indictment on Chicago Public Schools.Working within the legal precedent handed down by the Court, they sought guidance from the federal government, a consulting firm, and the "Blue Ribbon Committee" to pioneer a viable substitute policy tool.Indeed, research shows the "Tier system" is better than weighing academic performance alone (see Barrow et al., 2018).Though, as many scholars point out, there is no ideal proxy policy tool for promoting racial diversity in K-12 public schools (Ellison & Pathak, 2016;Linn & Welner, 2007;Pathak & Sönmez, 2013;Reardon, Baker, Kasman, Klasik, & Townsend, 2015); the relationship between race and income varies within communities, limiting their efficacy (Frankenberg, 2018).Consequently, we also argue evolving Court jurisprudence has played an important role in the bifurcation of educational equality and racial diversity in public educationfrom race-based to race-neutral-resegregating urban settings in the process.

Limitations and Future Research
Our study provides insights into the changes in enrollment patterns based on race and ethnicity that have occurred in CPS against the backdrop of this bifurcation.Yet we acknowledge this study has limitations with regards to directly linking changes in enrollment of students by race and ethnicity at CPS' selective high schools to the race-neutral, "Tier system" policy post-PICS.Toward these ends, as discussed, inclusion of open enrollment public schools as a comparison group strengthens examination of this phenomenon across time by comparing changes in selective school enrollment to changes in the enrollment of other schools within the same geographic areas (Wong, Cook, & Steiner, 2015).However, because all of CPS' selective enrollment high schools were subject to the "Tier system" policy, a comparison group controlling for other potentially confounding variables (e.g.changes in overall application numbers, changes in transportation availability, and changes in SES classifications for different race/ethnicity students) would also be useful.In this way, our study should be viewed as a starting point for further empirical work that includes a wider variety of external covariates and compares race-neutral policies in varying state and local contexts throughout the US.It can also serve as a point of hypothesis generation with regards to how changes in open enrollment schools are related to changes in selective schools within a race-neutral policy landscape-an empirical approach that has not been conducted prior.

Conclusion
In sum, we find Chicago's selective enrollment high schools are resegregating.While some scholars contend residential factors are largely responsible for K-12 resegregation, we further demonstrate the importance Parents Involved has played towards such ends.On one hand, we show how the decision has constrained the policy tools and behaviors local districts and schools can use in promoting racial diversity.These changes are furthering negative conditions for minority students in urban settings-e.g."exposure to more low-income peers, more crime, fewer positive role models… fewer resources, etc." (Reardon, Kalogrides, & Shores, 2017, p. 38)-setting up distinct disadvantages as they enter the workforce.They are also reinforcing what it means to be an urban minority-i.e.those who cannot escape their plight amid school access issues and institutionalized poverty (Friend, Caruthers, & McCarther, 2009).On the other hand, we show the decision's impact on defining and promoting racial diversity within K-12 public schools.Whereas Brown ruled "separate is unequal," Parents Involved has ruled just the opposite: "equal is equal."Voluntary, racebased policy tools have thereby been replaced with race-neutral policy tools.Amid context and design challenges, we find this has resulted in resegregation for many urban districts, including their elite schools.
Put simply, "diversity" as praxis has been reimagined."We are now as far from Brown as Brown was from… Plessy v. Ferguson" (Orfield et al., 2014 p. 20).And while Justice Kennedy said "school boards may pursue the goal of bringing together students of diverse backgrounds and races through other means," (PICS, 2007) our research along with others highlight the access issues that have transpired for students of color when individual race as a factor is locally constrained.Therefore, affirmative student assignment policies remain a compelling government interest as a means to promote diversity.
That said, given the Trump Administration recently rescinded the 2011 federal guidance (Walsh, 2018), there is little indication of another policy regime shift in the near future.Hence, more empirical research is needed to shed light on the resegregation of public schools and its impact on minority student access and opportunity.Despite our significant findings, Chicago is one critical case, characterized by political and contextual limitations.Equally, US Supreme Court decisions often have effects that are delayed, inconsistent, or both (McDermott et al., 2012).Researchers therefore need to play a continuing role in understanding this intersection between K-12 desegregation law and local policy. below).

Figure 2 :
Figure 2: Mean Percentage Enrollment by Student Race/Ethnicity in Upper-Tier Selective Schools Compared to Lower-Tier Selective Schools in Chicago Public Schools from 2006 to 2016.

Table 1
Percent Enrollment by Race/Ethnicity in Chicago Selective Enrollment Schools Compared to Open Enrollment Mean Percentage Enrollment by Student Race/Ethnicity in Selective and Open Enrollment Schools in Chicago Public Schools from 2006 to 2016.