Sherice Sargent v. School District of Philadelphia

PRECEDENTIAL

UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT ___________

No. 24-3112 ___________

SHERICE SARGENT, individually, as next friend of her minor child, and on behalf of those similarly situated; FALLON GIRINI, individually, as next friend of her minor child, and on behalf of those similarly situated; MICHELE SHERIDAN, individually, as next friend of her minor child, and on behalf of those similarly situated; JOSHUA MEYER, individually, as next friend of his minor child, and on behalf of those similarly situated

v.

SCHOOL DISTRICT OF PHILADELPHIA; SUPERINTENDENT, SCHOOL DISTRICT OF PHILADELPHIA; BOARD OF EDUCATION, THE SCHOOL DISTRICT OF PHILADELPHIA; JOYCE WILKERSON; LETICIA EGEA HINTON; JULIA DANZY; MALLORY FIX LOPEZ; MARIA MCCOLGAN; LISA SALLEY; REGINALD STREATER; CECILIA THOMPSON, each in their official capacities as members of the Board of Education of the School District of Philadelphia; SABRIYA JUBILEE, in her official capacity as Director of Diversity, Equity, and Inclusion for the School District of Philadelphia; KARYN LYNCH, in her official capacity as Chief of Student Support Services for the School District of

Philadelphia

Sherice Sargent, Michele Sheridan, Joshua Meyer, Appellants

On Appeal from the United States District Court for the Eastern District of Pennsylvania (D.C. No. 2:22-cv-01509) District Judge: Hon. Chad F. Kenney

Argued September 9, 2025

Before: HARDIMAN, KRAUSE, and FREEMAN, Circuit Judges.

(Filed: February 2, 2026)

____________

OPINION OF THE COURT ____________

HARDIMAN, Circuit Judge.

Three parents of Philadelphia students appeal the District Court’s summary judgment rejecting their Equal Protection challenge to the School District’s 2022 Admissions Policy for four selective high schools. Citing the Supreme Court’s decision in Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252

(1977), the District Court held that no reasonable factfinder could find that the Admissions Policy had either a racially discriminatory purpose or impact. So it applied rational basis review and upheld the Policy. Viewing the record in the light most favorable to the parents, there is sufficient evidence for a reasonable factfinder to conclude that the Admissions Policy had a discriminatory purpose and impact, requiring strict scrutiny. We will therefore vacate and remand for a factfinder to assess discriminatory purpose and impact.

I

The School District of Philadelphia has three types of high schools: (1) neighborhood schools; (2) citywide schools; and (3) criteria-based schools. Criteria-based schools are selective and require city residents to apply for admission. The four most competitive criteria-based schools are the Academy at Palumbo, George Washington Carver High School of Engineering and Science, Central High School, and Julia R. Masterman High School. This appeal involves the Admissions Policy’s effect on applicants to those four schools.

A

Before the School District implemented the 2022 Admissions Policy, each criteria-based school had individualized admissions teams and standards tied to grades, punctuality, and—depending on the school—additional requirements like letters of recommendation, writing samples, or interviews. The schools also required applicants to take the Pennsylvania System of School Assessment (PSSA), but that standardized test was suspended by the Commonwealth of Pennsylvania during the COVID-19 pandemic. Applicants needed A’s and B’s “with the possible exception of one C in

[a] major subject on report cards,” and three of the four schools required “[e]xemplary” or “[e]xcellent attendance and punctuality.”1 Sargent v. Sch. Dist. of Phila., 2024 WL 4476555, at *5 (E.D. Pa. Oct. 11, 2024). Masterman High School also required algebra and foreign language classes. Final admissions decisions were typically made in a decentralized manner by the admissions team or principal of each school. Students who did not meet the minimum qualifications sometimes were admitted because of “[o]ther factors” like legacy admissions, App. 230, or patronage practices based on “who you know and who you don’t know,” App. 311.

In 2017, the Pew Charitable Trusts released a report analyzing the School District’s criteria-based school system and its admissions process. According to the School District, the Pew Report identified flaws in its admissions policy. One of those flaws was that certain geographic areas were underrepresented at the criteria-based schools.2 But the School

1 Applicants to the fourth school, Carver, needed to meet a 95% attendance requirement. 2 Long after the 2017 Pew Report—and after the Admissions Policy had been implemented—the School District’s Office of Research and Evaluation expressed similar findings in a February 2022 report. That report concluded that some zip codes sent “very large percentages” of students to Palumbo, Carver, Central, and Masterman while others sent only “very small percentages.” App. 1433. It added that the “discrepancy in geographic access may be an indicator of inequities in student access to the opportunities offered by these schools.” Id.

District took no action on the Pew Report for three years.

On June 15, 2020—three weeks after the death of George Floyd—the School District issued an “Anti-Racism Declaration.” Superintendent Dr. William Hite wrote:

[I]t is imperative that we take a laser focus on acknowledging and dismantling systems of racial inequity. For us, this goes deeper and far beyond focusing on individual acts of prejudice and discrimination, but refers to uprooting policies, deconstructing processes, and eradicating practices that create systems of privilege and power for one racial group over another . . . . We must be bold and courageous, willing to do the necessary work to acknowledge and disrupt racist ideologies and behaviors within our own lives in an effort to dismantle racism within our school system . . . . [R]ace is the social construction that set the foundation and built the infrastructure for the United States we know today. Racism is the root of all other forms of injustice and provides the nourishment needed for other systems of oppression to thrive . . . . As we move forward with this charge, we will do so together, with intention and deep purpose, centering our work through the lens of racial equity . . . . No justice, no peace!

App. 679 (emphases added).

Karyn Lynch, the School District’s Chief of Student Support Services, confirmed that the School District responded to the death of George Floyd by examining all processes,

procedures, and practices to ensure “equity.” App. 309. This examination was called an “equity lens review.” App. 310. The equity lens was “a tool” that the Chief of Equity for the School District, Dr. Sabriya Jubilee, used to assess “to what degree a particular policy or action [raised] equit[y] issues or concerns.” App. 502. Dr. Jubilee and her office reviewed all School District policies, including the admissions processes for the four schools at issue, through this “equity lens.”

The School District’s and its officials’ references to equity provide context for how that term was used. For example, the School District’s “Living Glossary” defines “educational equity” as “raising the achievement of all students, while eliminating the racial predictability and disproportionality of which student groups occupy the highest and lowest achievement categories.” App. 680. And when asked whether there is a difference between viewing something through the lens of “equity” and “racial equity,” Dr. Hite responded, “I don’t believe so, no.” App. 418. Likewise, Board of Education Member Mallory Fix-Lopez, confirmed that “equity does include race.” App. 1180.

Less than six months after the Anti-Racism Declaration, on December 10, 2020, the Board of Education released its “Goals & Guardrails,” a policy document describing the principles that guide the Board’s educational mission. “Guardrail 4” declared that, going forward, “students’ potential will not be limited by practices that perpetuate systemic racism and hinder student achievement.” App. 1421. And under Guardrail 4, Indicator 4.1 announced:

Among 8th grade students who are qualified to attend Special Admission High [S]chools, the percentage who are Black/African American or

Hispanic/Latinx will grow from 33.8% in August 2020 to at least 52.0% (making progress towards being proportional to [the] population as a whole) by August 2026.

Id. (emphasis added).

Prepared remarks for the January 14, 2021 Board of Education meeting provide additional context for the School District’s Goals and Guardrails. It was “typical” for the School District to provide and/or collaborate with certain board members on scripts for presentations given at such meetings. App. 1178. At the outset, the January 14th remarks “acknowledge[d] that systemic racism is real and alive in [the School] District.” App. 759. To that end, the portion of the remarks ascribed to Board Member Mallory Fix-Lopez read: “We must work to dismantle the practices that result in different outcomes for our students. We must focus on equity in everything we do.” Id. The portion attributed to Board Member Leticia Egea-Hinton promised that the Goals and Guardrails would “serve as a framework for the work of the Superintendent and the District.” App. 760. That portion of the remarks also promised that in 2021 the public would “see the District begin to implement plans to reach [those] goals.” App. 762.

The School District kept that promise. In October 2021, the day before the admissions window opened for the 2022 to 2023 school year, it announced its new Admissions Policy, which brought significant changes to the application process for Palumbo, Carver, Central, and Masterman. The School District’s announcement read:

This past year, in alignment with our commitment toward antiracism and equity, as outlined in the Board of Education’s Goals and Guardrails, the school selection process underwent an initial equity review . . . . As a result of the initial review, the District will be implementing several important changes to this year’s process.

App. 1313.

The Admissions Policy scrapped the prior approach in favor of a centralized process with new admissions criteria for the four schools. Those changes included: (1) continuing the pandemic-induced suspension of standardized testing; (2) eliminating recommendation letters and interviews; and (3) introducing a 90-minute mandatory essay exam (MI Write) scored by a computer program (thereby eliminating all previous school-specific writing sample requirements). At the same time, the Admissions Policy changed the standards for grades (now requiring all A’s and B’s instead of allowing exceptions for a stray C) and punctuality (replacing the “excellent” or “exemplary” attendance standards previously used at three of the schools with a 95% attendance rate to qualify for all four schools).

Most significantly for this appeal, the Admissions Policy introduced six preferred zip codes. Students who resided in those six zip codes did not receive a mere thumb on the scale in their favor; they received a “golden ticket” of sorts, with “qualified applicants”—i.e., those who met the School District’s new criteria—admitted automatically to any of the four most coveted schools to which they applied. Meanwhile, students who met the new criteria but did not reside in the six

zip codes were subject to a lottery and thus had to hope for the best for the remaining available seats. The new system also effectively abolished a previous practice of the School District: no longer could students who had attended some of the schools as seventh or eighth graders continue there as ninth graders if they maintained satisfactory grades. Those students needed to reapply.

Five of the six zip codes—19121, 19132, 19133, 19134, and 19140—had a majority Black and Hispanic population. Their Black and Hispanic populations were 78.7%, 95.1%, 94.4%, 64.7%, and 92.3%, respectively. Meanwhile, their Asian populations were 3.5%, 1.15%, 3.5%, 1.6%, and 1.7%. And their white populations were 15.4%, 2.5%, 3.6%, 30.3%, and 4.3%. The sixth zip code, 19135, had a plurality of white residents (44.7%) and the remainder of its population was 20.9% Black, 24.2% Hispanic, and 5.7% Asian.

B

The School District moved for summary judgment in the District Court. In support of its motion, the School District submitted the declaration of Dr. Tonya Wolford, the Chief of District Evaluation, Research, and Accountability. Dr. Wolford described her “heav[y] involve[ment] in the creation” of the Goals and Guardrails document, which, according to her, “was not intended to and did not [] impact” the admissions process. App. 127. Her office also determined which zip codes would receive preferred status under the 2022 Admissions Policy. Dr. Wolford said her office “examined which zip codes had the lowest representation of [ninth] grade students at Palumbo, Carver, Central, and Masterman.” App. 128. She averred that she “did not consider” the racial demographics of the zip codes or “any factor other than the number of students

enrolled at [the] four schools in prior years.” Id. According to Dr. Wolford, “race was not a factor and was not considered when recommending which zip codes were to receive priority during the school selection process.” Id.

In deposition testimony, other officials painted a different picture of the School District’s intent in implementing the new Admissions Policy. While Dr. Wolford spoke about how her office selected the specific zip codes that received preference under the Policy, these other officials discussed why the School District pursued a zip code preference in the first place. Karyn Lynch, the Chief of Student Support Services, asserted that the zip code preference was added to the Admissions Policy after the School District concluded that the prior system was “bias[ed].” App. 230. And the Chief of Equity, Dr. Jubilee, stated that being “colorblind” is “not a good thing,” App. 617, and that “[r]ace is a factor, but it is not an isolated factor” in decisions “toward equity,” App. 653. She acknowledged that race could be considered in admissions if “the goal is to be racially equitable.” App. 650. When asked if her office became aware of the racial makeup of the six zip codes at some point, Dr. Jubilee responded, “[r]ace is always a part of the conversation.” App. 642.

Dr. Jubilee reaffirmed that sentiment, stating: “I [have already] said that any time we’re talking about any measure of th[e] effort [to update admissions criteria], race is always a consideration.” Id. When asked if she had ever become aware that there were a disproportionately high number of Asian students at one of the schools, Dr. Jubilee replied: “As we were going through my involvement in [the] school selection process, at some point I had an understanding of different demographics across a number of continuums.” App. 520.

For his part, Dr. Hite confirmed that, as Superintendent, he was aware of “systems” within the School District that led to racial inequity and that one such system was “disproportionality.” App. 393. According to Dr. Hite, disproportionality occurs when a racial group’s percentage as part of a subgroup is not proportional to that group’s percentage in the overall population. Dr. Hite also opined that “white students maintain[] privilege over” Black and Latino students within the School District. App. 413. Dr. Jubilee likewise confirmed that “[a]ny time there’s disproportionality, equity is a concern.” App. 525.

This deposition testimony of Ms. Lynch, Dr. Jubilee, and Dr. Hite also echoed earlier public statements by School District officials. At a December 2021 meeting of the Council of the City of Philadelphia Committee on Education, Ms. Lynch and Dr. Jubilee spoke in favor of the Admissions Policy. Dr. Jubilee declared: “Through the school selection process, we have the opportunity to redesign a process that from inception to current practice has only truly benefited a small group of stakeholders, many of whom do not reflect the majority demographic of our School District or City.” App. 743.

Dr. Jubilee also opined that “[e]quity and equality are not synonymous.” App. 744. “Equity,” she said, “is about fairness and the mission to achieve balance.” Id. Ms. Lynch added that, “[a]s a result” of the national conversation about racism, “we committed to examining all of our processes, procedures[,] and practices to ensure equity.” App. 1531. “For more than a year,” she continued, “the School District obtained feedback about the school selection process and listened to individuals and representatives of groups to inform the improvements to the school selection process.” Id.

At that same December 2021 Committee meeting, Appellant Sherice Sargent spoke against the Admissions Policy. She explained her frustration with how quickly the changes unfolded, blindsiding parents. She pointed out that, because the changes were adopted wholesale in the middle of the school year, parents had to make quick decisions based on the new policy while managing COVID-19-related disruptions to their children’s education. Sargent criticized the School District for overhauling the process all at once (rather than adopting a phased implementation) and not permitting exceptions for eighth graders (like her daughter) who had relied on the old admissions process in planning for their futures.

Councilmember David Oh also criticized the School District: “[T]his [policy] was announced the day before a seven-week window open[ed] for applications to these criteria- based admissions schools. And people were reeling from the fact that an entirely new process had been put in without their knowledge, without their opportunity to discuss or to have experts weigh in on it.” App. 1446. Councilmember Helen Gym added: “[The announcement of the Admissions Policy] created chaos and uncertainty, fear and suspicion . . . . I think that there was no question that there was no significant public input.” App. 1451.

Finally, Dr. Joshua Wilson, an expert in the use of automated scoring and automated writing evaluation tools, expressed his disagreement with the Admissions Policy’s reliance on the MI Write tool. He commended the School District for its efforts to use a more objective tool to score writing samples, but he opined that the MI Write tool was inappropriate for high-stakes admissions decisions. That tool, according to Dr. Wilson, was “deliberately designed not to

evaluate content.” App. 1479. Instead, it “only evaluate[d] the quality of the writing.” Id. Nonetheless, to qualify for Central or Masterman, students needed to score 22 out of 30. For Palumbo and Carver, the qualifying score was 17 out of 30. Dr. Wilson “[knew] of no research that validate[d]” whether these scores were “predictive of future success in those respective schools.” App. 1481–82. At the conclusion of the meeting, the Committee took no action.

In the meantime, as the Admissions Policy had taken effect in October 2021, the School District began applying it to applications submitted for the upcoming 2022 to 2023 school year.

II

Appellants Sherice Sargent, Michele Sheridan, and Joshua Meyer are parents of children who participated in the 2022 Admissions Process for the criteria-based schools. Each of their children resided outside the six preferred zip codes and met the new criteria but were not admitted to their first-choice schools.

The parents, in their own rights, as next friends of their minor children, and on behalf of those similarly situated, sued the School District, the Board of Education, and related individuals in their official capacities.3 They alleged that the School District “chang[ed] its selection process for criteria-

3 The individual Defendants include: William Hite, Joyce Wilkerson, Leticia Egea-Hinton, Julia Danzy, Mallory Fix- Lopez, Maria McColgan, Lisa Salley, Reginald Streater, Cecilia Thompson, Sabriya Jubilee, and Karyn Lynch.

based schools from a race-neutral process to a racially discriminatory process.” App. 1786. And they claimed that the School District did so “by moving from a highly individualized and criteria-based process to a gerrymandered lottery system where [B]lack and Latino students were given preferential treatment.” App. 1787.

The amended complaint included claims under 42 U.S.C. § 1983 for violations of Title VI of the Civil Rights Act of 1964, 42 U.S.C. § 2000d, and the Equal Protection Clause of the Fourteenth Amendment, U.S. Const. amend. XIV, § 1, and state-law claims for violations of Article I, Sections 26 and 29 of the Pennsylvania Constitution.4 Though they initially

4 Title VI’s prohibition against discrimination is coextensive with the Equal Protection Clause. See Doe ex rel. Doe v. Lower Merion Sch. Dist., 665 F.3d 524, 557 (3d Cir. 2011). For our purposes, the state constitutional provisions are, too. See, e.g., Commonwealth v. Albert, 758 A.2d 1149, 1151 (Pa. 2000) (“[T]he equal protection provisions of the Pennsylvania Constitution are analyzed under the same standards” as the federal Equal Protection Clause) (citation modified)). But see Allegheny Reprod. Health Ctr. v. Pa. Dep’t of Hum. Servs., 309 A.3d 808, 945 (Pa. 2024) (holding that “Section 26 of our Charter affords broader protections than the federal Equal Protection Clause”).

Appellants have standing. An injury resulting from governmental racial discrimination provides a basis for standing “only to those persons who are [] denied equal treatment” by the challenged conduct. Lower Merion, 665 F.3d at 542 (citation modified). One form of injury is being forced to compete in a race-based system that may disadvantage the plaintiff. See id. (quoting Parents Involved in Cmty. Schs. v.

sought an injunction, they now seek only damages.

At the close of discovery, the District Court granted the School District’s summary judgment motion. Sargent, 2024 WL 4476555, at *19. Because Plaintiffs were challenging “a facially neutral law or policy that is applied evenhandedly,” id. at *9 (citation omitted), for strict scrutiny to apply, they needed to show that the policy had a racially discriminatory purpose and impact, Doe ex rel. Doe v. Lower Merion Sch. Dist., 665 F.3d 524, 543 (3d Cir. 2011); see also Antonelli v. New Jersey, 419 F.3d 267, 274 (3d Cir. 2005). The Court held that no reasonable factfinder could find either discriminatory purpose or impact. Sargent, 2024 WL 4476555, at *10, *16. It therefore applied rational basis review and held that the Admissions Policy was rationally related to the legitimate interest in “increasing access for all qualified children to some of the City’s best schools.” Id. at *19.

In determining that no reasonable jury could find that the School District acted with a discriminatory purpose, the District Court made four conclusions central to its holding. Id. at *14. First, the Court concluded that the statements made by Dr. Hite, Dr. Jubilee, and Ms. Lynch neither revealed a discriminatory purpose nor were they sufficiently linked to the Admissions Policy. Id. at *16. Second, the Court held that nothing in the Anti-Racism Declaration could lead one to reasonably conclude that the School District implemented the changes for a discriminatory purpose. Id. Third, the Court

Seattle Sch. Dist. No. 1., 551 U.S. 701, 719 (2007)). Sargent, Sheridan, and Meyer allege that the Policy set up a race-based system that harmed their children by depriving them of admission to their preferred schools.

noted that statements at the City Council meeting by Dr. Jubilee and Ms. Lynch revealed only “a desire to correct racial biases that manifested in the previous school selection process.” Id. at *17. Lastly, the Court dismissed the idea that the zip codes were a proxy for race, citing, inter alia, Dr. Hite’s and Dr. Jubilee’s denials that they considered race in selecting the zip codes. See id. at *16. The Court determined instead that the zip codes were selected because those geographic areas had sent the lowest percentage of students to the schools in the past. Id. at *18.

The District Court also held that the Admissions Policy did not have a discriminatory impact. In doing so, the Court relied heavily on a recent decision of the United States Court of Appeals for the Fourth Circuit, Coalition for TJ v. Fairfax County School Board, 68 F.4th 864 (4th Cir. 2023), cert. denied, --- S. Ct. ---, 2024 WL 674659 (Feb. 20, 2024). Consistent with that case, the District Court explained that Plaintiffs’ data showed only the difference in the student body’s racial composition “before-and-after” the changed policy without accounting for how successful each racial group was in receiving admissions offers. Sargent, 2024 WL 4476555, at *13. A “simplistic” before-and-after approach, according to the District Court, would create a de facto “immutable quota,” whereby any change from a prior year’s demographics would raise constitutional alarm bells. Id. (quoting Coal. for TJ, 68 F.4th at 881). The Court also noted that Plaintiffs’ data were incomplete because they failed to address how many students of each race applied to the schools during that application cycle and whether the success rates of applicants by race changed after the implementation of the

Policy. Id. at *13–14.

Sargent, Sheridan, and Meyer timely appealed.5

III

The Fourteenth Amendment provides that no State shall “deny to any person within its jurisdiction the equal protection of the laws.” U.S. Const. amend. XIV, § 1. The “core purpose” of the Equal Protection Clause is to eliminate “all governmentally imposed discrimination based on race.” Students for Fair Admissions, Inc. v. President & Fellows of Harvard Coll., 600 U.S. 181, 206 (2023) (SFFA). “Eliminating racial discrimination means eliminating all of it.” Id.

There are three categories of racially discriminatory state action: (1) facially discriminatory policies containing racial classifications, see id.; (2) facially neutral policies that

5 The District Court had jurisdiction under 28 U.S.C. § 1331. We have jurisdiction under 28 U.S.C. § 1291. Our review is de novo. Faush v. Tuesday Morning, Inc., 808 F.3d 208, 215 (3d Cir. 2015). To prevail on summary judgment, the moving party must show that “there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). All facts must be viewed in the “light most favorable to the non-moving party” with “all reasonable inferences drawn in that party’s favor.” McKinney v. Univ. of Pittsburgh, 915 F.3d 956, 960 (3d Cir. 2019) (citation modified). We do not “weigh the evidence or assess its truth but simply determine[] whether or not there is a genuine issue for trial.” Antonelli, 419 F.3d at 272 (citing Anderson v. Liberty Lobby, 477 U.S. 242, 249 (1986)).

are applied differently on the basis of race, see Yick Wo v. Hopkins, 118 U.S. 356, 373 (1886); and (3) facially neutral policies that are applied evenhandedly but have a racially discriminatory purpose and impact, see Lower Merion, 665 F.3d at 543; Antonelli, 419 F.3d at 274. We agree with the parties and the District Court that the challenged Admissions Policy implicates only the third category.

In Arlington Heights, the Supreme Court clarified that state action cannot be unconstitutional solely because of its disproportionate impact on members of different racial groups. 429 U.S. at 264–65. The Court explained that “[p]roof of racially discriminatory intent or purpose is required to show a violation of the Equal Protection Clause.” Id. at 265 (emphasis added). But it also observed that a policy’s disproportionate or discriminatory impact might provide circumstantial evidence of state actors’ discriminatory purpose, along with other indicators like the policy’s historical background and the legislative or administrative history. Id. at 266–68.

Our Court, like some other Courts of Appeals, has read Arlington Heights’s discussion of purpose and impact as a conjunctive test requiring challengers to prove both discriminatory purpose and impact to trigger strict scrutiny6 of

6 We read Lower Merion to foreclose us from considering whether evidence of discriminatory purpose, by itself, can trigger strict scrutiny. But we acknowledge a circuit split on this issue and add two observations.

First, we note that the Supreme Court’s focus in this context has been on intent. In a recent dissent from the denial of certiorari, Justice Alito, joined by Justice Thomas, criticized lower courts for “mistakenly treat[ing] evidence of disparate

impact as a necessary element of an equal-protection claim” even though the Supreme Court had “never said as much.” Bos. Parent Coal. for Acad. Excellence Corp. v. Sch. Comm. for Bos., 604 U.S. ---, 145 S. Ct. 15, 17 (2024). In Arlington Heights, the Court made clear that impact “may provide an important starting point” to assess whether there is a hidden discriminatory purpose behind a facially neutral policy. 429 U.S. at 266. But it discussed disparate impact as one potential indicator of discriminatory purpose. See Arlington Heights, 429 U.S. at 268 (“The foregoing summary [of indicia (including disparate impact)] identifies, without purporting to be exhaustive, subjects of proper inquiry in determining whether racially discriminatory intent existed.”). And in the decades since Arlington Heights, including quite recently, the Supreme Court has repeatedly described that inquiry as one about purpose. See, e.g., United States v. Skrmetti, 605 U.S. 495, 516 (2025) (“But where a law’s classifications are neither covertly nor overtly based on sex . . . we do not subject the law to heightened review unless it was motivated by an invidious discriminatory purpose.”); Miller v. Johnson, 515 U.S. 900, 913 (1995) (“[S]tatutes are subject to strict scrutiny . . . when, though race neutral on their face, they are motivated by a racial purpose.”).

Second, these pronouncements of the high court echo our dictum in Lower Merion: “‘When there is a proof that a discriminatory purpose has been a motivating factor in the decision’ . . . courts should apply strict scrutiny.” 665 F.3d at 551 (quoting Arlington Heights, 429 U.S. at 265–66). See also Pryor v. Nat’l Coll. Athletic Ass’n, 288 F.3d 548, 562 (3d Cir. 2002) (“Once a plaintiff establishes a discriminatory purpose based on race, the decisionmaker must come forward and try

a policy that is facially neutral and applied evenhandedly. See, e.g., Lower Merion, 665 F.3d at 549; Antonelli, 419 F.3d at 274; Coal. for TJ, 68 F.4th at 879; Bos. Parent Coal. for Acad. Excellence Corp. v. Sch. Comm. for Bos., 89 F.4th 46, 56–57 (1st Cir. 2023); Chinese Am. Citizens All. of Greater N.Y. v. Adams, 116 F.4th 161, 165 (2d Cir. 2024); Lewis v. Ascension Par. Sch. Bd., 806 F.3d 344, 359 (5th Cir. 2015). But not all have. See, e.g., United States v. Viveros-Chavez, 114 F.4th 618, 622 (7th Cir. 2024) (“[A] facially neutral law fails constitutional muster ‘if there is proof that a discriminatory purpose has been a motivating factor’ in its enactment.” (quoting Arlington Heights, 429 U.S. at 267)); Stevenson v. Blytheville Sch. Dist. #5, 800 F.3d 955, 970 (8th Cir. 2015) (“An allegation of disproportionate impact is only relevant to the extent that it reflects a discriminatory purpose.” (citation modified)).

In its pathmarking 2023 decision in Students for Fair Admissions, Inc. v. President & Fellows of Harvard College, the Supreme Court held two universities’ race-conscious admissions policies unconstitutional. 600 U.S. at 231. One school used race as a “determinative tip,” id. at 195, and the other considered it a “plus” factor, id. at 196. Applying strict scrutiny to the racially discriminatory policy,7 the Court

to show that the policy or rule at issue survives strict scrutiny.”). 7 In SFFA, the Supreme Court noted that in only two other contexts had it found a race-based interest sufficiently compelling to satisfy strict scrutiny. 600 U.S. at 207. First, “remediating specific, identified instances of past discrimination that violated the Constitution or a statute,” and

rejected interests such as “better educating [] students through diversity” and “enhancing appreciation, respect, and empathy, cross-racial understanding, and breaking down stereotypes” as insufficiently compelling. Id. at 214. The Court also held that the universities had failed to “articulate a meaningful connection between the means they employ[ed] and the goals they pursue[d]” so the policies were not narrowly tailored. Id. at 215. In the zero-sum competitive admissions processes at issue in SFFA, amorphous “diversity” goals did not pass constitutional muster. Id. at 214, 218–19.

The Court’s opinion in SFFA was an extension of its prior statements that it is “patently unconstitutional” for a public school to seek “some specified percentage of a particular group merely because of its race or ethnic origin,” Grutter v. Bollinger, 539 U.S. 306, 329–30 (2003) (citation modified), and that racial balancing, such as aiming to “attain[] a level of diversity within the schools that approximates the district’s overall demographics” is an illegitimate objective, Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 727 (2007) (plurality opinion) (citation modified). Though the Court had tolerated limited race-conscious admissions policies in the past, see, e.g., Regents of Univ. of Cal. v. Bakke, 438 U.S. 265, 320 (1978) (opinion of Powell, J.); Grutter, 539 U.S. at 343, no longer was it “willing to dispense temporarily with the Constitution’s unambiguous guarantee of equal protection.” SFFA, 600 U.S. at 212. The Court expressly rejected the goal of proportionality and cautioned that, going forward, schools “may not simply establish through application essays or other means the regime

second, “avoiding imminent and serious risks to human safety in prisons.” Id.

we hold unlawful today.” Id. at 230. Because “[w]hat cannot be done directly cannot be done indirectly,” id. (quoting Cummings v. Missouri, 71 U.S. 277, 325 (1867)), when a discriminatory purpose has been a motivating factor in a state actor’s decision to enact a facially neutral policy that produces a discriminatory impact, judicial deference is no longer justified. See Arlington Heights, 429 U.S. at 265–66. “‘The Constitution deals with substance, not shadows,’ and the prohibition against racial discrimination is ‘levelled at the thing, not the name.’” SFFA, 600 U.S. at 230 (quoting Cummings, 71 U.S. at 325).

With these principles to guide our assessment of the zero-sum, competitive Admissions Policy at issue in this appeal, we hold that a reasonable factfinder could decide that: (1) the School District acted with a discriminatory purpose in adopting the Policy; and (2) the Policy had a discriminatory impact. Because there are genuine disputes of material fact on each issue, we will vacate the District Court’s order granting summary judgment.

A

Start with the purpose prong. State action is taken with “a discriminatory purpose” when the decisionmaker adopted the challenged action at least partially to benefit or burden an identifiable group. Lower Merion, 665 F.3d at 548. The decisionmaker must have “selected or reaffirmed a particular course of action at least in part ‘because of,’ not merely ‘in spite of,’ the action’s beneficial or adverse effects” on certain groups. Id. at 551 (quoting Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 279 (1979)). State actors’ “mere awareness” of the racial effects of a policy is not actionable. Lower Merion, 665 F.3d at 548. But when the evidence shows that a discriminatory

purpose has been a motivating factor in the decision and there is also a discriminatory impact, as required by our precedents, judicial deference is no longer justified. Id. at 549, 551 & n.42.8

To determine whether a discriminatory purpose was a motivating factor, we undertake a “sensitive inquiry” into all “circumstantial and direct evidence of intent.” Arlington Heights, 429 U.S. at 266. Direct evidence is rare. Circumstantial evidence is more common and might include: (1) whether the official action has a racially disproportionate impact, including “whether it bears more heavily on one race than another”; (2) the historical background of the decision; and (3) the legislative or administrative history of the decision. Id. at 266–68 (quoting Washington v. Davis, 426 U.S. 229, 242 (1976)).

Appellants argue that summary judgment was improper because, when viewing the record in the light most favorable to them, a reasonable factfinder could find that the School District implemented the Admissions Policy to alter the racial composition of the student body at the four schools. They cite:

8 A defendant may avoid liability for the action if it proves “that the same decision would have resulted even had the impermissible purpose not been considered.” Arlington Heights, 429 U.S. at 270 n.21. In such cases, the plaintiffs can “no longer fairly [] attribute the injury complained of to improper consideration of a discriminatory purpose.” Id. The School District does not raise this defense here, so our discussion at this stage of the litigation concerns only whether there is sufficient evidence in the record for a reasonable factfinder to determine that a discriminatory purpose was at least a motivating factor in the adoption of the Admissions Policy.

(1) the Anti-Racism Declaration announcing the School District’s intent to “uproot[]” its policies and “center[] [its] work through the lens of racial equity,” App. 679; (2) the School District’s Goals and Guardrails document, including the goal in Indicator 4.1 that the percentage of qualified Black or Hispanic students “will grow” from 33.8 percent to “at least” 52 percent, “making progress towards being proportional to [the] population as a whole,” App. 1421; and (3) the school officials’ declarations, deposition testimony, and statements at the December 2021 City Council Committee Hearing.

The School District counters that the “evidence, including testimony and [School] District documents, uniformly showed that the 2022 Admissions Process was race- neutral and motivated by legitimate goals, such as increasing objectivity and improving access for qualified students from underrepresented geographic areas.” School District Br. 25. The School District maintains that the Admissions Policy was adopted to promote standardization in the processes, enhance objectivity and fairness, and increase access to education. And it cites the Pew Report as evidence that the preferred zip codes were selected merely because of “historically low enrollment rates” in those areas. School District Br. 7.

As we have explained, the historical context in which the policy was adopted along with the legislative or administrative history of its adoption can provide circumstantial evidence of discriminatory intent. Lower Merion, 665 F.3d at 552. The “specific sequence of events leading up to the challenged decision also may shed some light on the decisionmaker’s purposes.” Arlington Heights, 429 U.S. at 267. In particular, departures from the normal process of adopting and implementing a policy might provide evidence that the actions were taken for improper purposes. See id.

Reports or contemporary statements by members of the decisionmaking body are particularly relevant. Id. at 268.

The School District’s actions in this case did not occur in a vacuum. Just three weeks after George Floyd’s death sparked a national conversation about racism, the School District shared its Anti-Racism Declaration, which vowed to scrutinize all its work “through the lens of racial equity.” App. 679. Less than six months later, the Board released the Goals and Guardrails document, announcing its intention to increase the percentage of Black or Hispanic applicants who met admissions criteria from 33.8 percent to at least 52 percent of the overall total by August 2026. That growth, the document stated, would “mak[e] progress” towards achieving a pool of “qualified” students that is “proportional to [the] population as a whole.” App. 1421. And in early 2021, a month after the release of its Goals and Guardrails document, Board members assured the public it would see new policies implemented that year to further those goals. The School District followed through on that commitment when it announced the new Admissions Policy in October 2021. In its announcement, the School District explained that it had adopted the Admissions Policy “[a]s a result of” the equity lens review of all its policies. App. 1313. And in a possible departure from normal processes, the Admissions Policy was announced the day before the application opened, and it took effect immediately, with “little input” from the public. App. 1700.

These statements and actions, taken together in context, could support a finding that the School District adopted the Admissions Policy to achieve racial proportionality. The record also could support the conclusion that the zip code preference was a proxy for race because it benefitted members of certain racial groups more than others. Dr. Jubilee herself

confirmed that “the [z]ip code preference was equity.” App. 616. Five of the six zip codes have a majority Black and Hispanic population with disproportionately low Asian and (in some zip codes) disproportionately low white populations. So the “golden ticket” of automatic admission was more likely to be awarded to Black and Hispanic students who lived in the preferred zip codes. Cf. Rice v. Cayetano, 528 U.S. 495, 514 (2000) (invalidating a state constitutional amendment that used ancestry as a proxy for race to limit voting rights); Griggs v. Duke Power Co., 401 U.S. 424, 431–32 (1971) (holding that a high school diploma requirement was an “artificial, arbitrary, and unnecessary barrier[] to employment” that “operate[d] invidiously to discriminate on the basis of [race]”).

In sum, viewing the evidence and all reasonable inferences therefrom in the light most favorable to Appellants, a reasonable factfinder could find that the School District adopted the Admissions Policy to change the racial makeup of the four most selective schools. Altering the schools’ racial makeup would increase the representation of Black and Hispanic students while decreasing white and Asian students’ representation in a zero-sum admissions game. See SFFA, 600 U.S. at 272 (Thomas, J., concurring) (highlighting the zero- sum nature of the admissions process).

No doubt, some evidence—such as the Pew Report and the Wolford Declaration—supports the School District’s stated rationales of increasing geographic diversity and decreasing subjectivity in the admissions process. But the District Court’s conclusion that the “evidence uniformly demonstrate[d] that the changes to the admissions process were not motivated by a racially discriminatory purpose,” Sargent, 2024 WL 4476555, at *15 (emphasis added), is belied by the record. In focusing on the evidence of how the School District went about

implementing the Admissions Policy, the District Court did not adequately consider the evidence of why the School District implemented the Policy in the first place, including the School District’s stated goals, the historical context behind the “equity” aims, and statements made by School District officials. So while a factfinder could agree with the District Court’s reading of the record and find a discriminatory purpose lacking, we cannot determine as a matter of law that no reasonable factfinder could infer a discriminatory purpose. The District Court erred in holding otherwise.

B

Having determined that one could reasonably find a discriminatory purpose, we turn next to consider impact. “[D]iscriminatory impact must be shown to establish an equal protection violation because ‘plaintiffs must show that they have been injured as a result’ of the governmental action to ensure that courts ‘can impose a meaningful remedy.’” Lower Merion, 665 F.3d at 549–50 (quoting Garza v. Cnty. of Los Angeles, 918 F.2d 763, 771 (9th Cir. 1990)). Here, a reasonable factfinder could conclude that the Admissions Policy had a discriminatory impact.

Appellants provide data on the raw number of students accepted by race (without accounting for the number of students by race who applied) before and after the implementation of the Admissions Policy. Their data show that admissions offers to Asian students decreased at three of the four schools and offers to white students decreased at all four. Meanwhile, admissions offers to Black and Hispanic students increased significantly at three of the four schools.

The School District counters that the data “merely lists

the year-to-year racial composition of students receiving offers the year before and the year of the 2022 Admissions Process” without comparing “the success rates of students” who were deemed “qualified” under the new admissions criteria “by race and ethnicity.” School District Br. 34. To show the race-neutral impact of its Admissions Policy, the School District has provided its own data. Citing Dr. Wolford’s declaration, the School District notes that the success rates by race were “nearly identical” because the proportion of admitted students by race was approximately equivalent to the proportion of “qualified applicants” by race, showing a neutral effect across racial groups. School District Br. 41. For example, for the 2021 to 2022 school year, 38 percent of “qualified applicants” to Palumbo were Asian and 39 percent of admitted students were Asian. App. 145. So the School District contends there was no discriminatory impact because the student body composition (by race) tracked the percentage of “qualified applicants” (by race).

The School District’s presentation of the success rates lacks context. The Admissions Policy also changed the relevant qualifications, so a “qualified applicant”9 did not have

9 In one respect, the School District raised standards. The Admissions Policy required students to have all A’s and B’s when a stray C had previously been acceptable. But the standards declined in other ways, including the lack of standardized testing and the introduction of an automated writing sample tool that did not assess content. And the School District did away with letters of recommendation, traditional writing samples, and interview requirements. So to qualify under the Admissions Policy, a student in a preferred zip code only had to pass the automated writing assessment and meet

the same meaning in 2021 and 2022. Moreover, the percentage of “qualified applicants” (by race) in the pool changed significantly from 2021 to 2022, and qualifying for admission alone resulted in automatic admission in a preferred zip code. The data in the Wolford Declaration show that the Admissions Policy yielded some striking differences in the “qualified applicant” pools for each of the schools between 2021 and 2022:

• • Palumbo. The percentage of qualified applicants who were Black more than doubled from 11 percent to 24 percent and the percentage of qualified applicants who were Hispanic did likewise from 6 percent to 15 percent. The percentage of qualified applicants who were Asian or white declined from 38 percent to 28 percent and from 40 percent to 26 percent, respectively.
• • Carver. The percentage of qualified applicants who were Black increased from 16 percent to 29 percent and the percentage of qualified applicants who were Hispanic doubled from 6 percent to 12 percent. The percentage of qualified applicants who were Asian and white decreased from 35 percent to 31 percent and 36 percent to 21 percent, respectively.
• • Central. The percentage of qualified applicants who were Black doubled from 11 percent to 22 percent and the percentage of qualified applicants who were Hispanic increased from 6 percent to 13 percent. The percentage of qualified applicants who were Asian and white decreased from 38 percent to 32 percent and from

attendance and grade requirements like those that had been required before.

39 percent to 26 percent, respectively.

• • Masterman. The percentage of qualified applicants who were Black increased slightly from 11 percent to 12 percent and the percentage of qualified applicants who were Hispanic increased from 5 percent to 8 percent. Qualified applicants who were Asian also saw a mild increase from 37 percent to 39 percent. The percentage of qualified applicants who were white declined from 39 percent to 35 percent.

See App. 145.

The notable changes in the “qualified applicant” pool discussed above track the School District’s stated aims in its Goals and Guardrails document. Recall the School District announced in Indicator 4.1 that the percentage of eighth grade students qualified for the criteria-based schools who are “Black/African American or Hispanic/Latinx will grow from 33.8% in August 2020 to at least 52.0% (making progress towards being proportional to [the] population as a whole by August 2026).” App. 1421. In the year at issue, the School District increased the percentage of “qualified applicants” who were Black or Hispanic at each of the four schools.

Based on these increases and the fact that being “qualified” resulted in automatic admission for students in preferred zip codes, a reasonable factfinder could conclude that the School District was well on its way to achieving racially proportional representation in the four most competitive public high schools. And that it did so by both changing the admissions criteria and giving preference to students in zip codes where Asian Americans were “woefully underrepresented,” meaning that the benefits of the Policy—

i.e., automatic admission—were almost certain to redound to the benefit of students who were not Asian American. Reply Br. 2. The School District does not dispute that the zip code preference accomplished its stated goal of increasing the number of enrolled students from those areas.

Yet the School District argues, and the District Court held, that Appellants’ data do not suffice at the summary judgment stage under either our precedents or two of our sister courts’ analytical framework. We disagree. Our precedents, despite involving facially neutral but evenhandedly applied policies like this one, are otherwise factually inapposite. And we cannot abide the First and Fourth Circuit’s flawed disparate impact analyses.

Consider first our precedents. In Antonelli, we evaluated a facially neutral fire department entry exam adopted after New Jersey was subject to a consent decree requiring “affirmative action to increase the proportion of African-American and Hispanic personnel.” 419 F.3d at 270–71. A group of mostly white applicants who failed the exam alleged that it was racially discriminatory in violation of the Equal Protection Clause. Id. at 272. But the appellants there “ha[d] not provided any evidence that [the relevant portion of the exam] had a discriminatory impact on white candidates.” Id. at 276. And the mean scores and passing rates were “remarkably similar for African-American, Hispanic[,] and white applicants.” Id. Because of that lack of evidence, we held that there was no discriminatory impact. Id.

Next, in Lower Merion, we assessed the constitutionality of a public-school redistricting plan intended

to equalize student enrollment at two high schools. 665 F.3d at 539. The district court entered judgment for the school district after a bench trial, and we affirmed, holding that the plan comported with the Equal Protection Clause. Id. at 557. To show discriminatory impact, the challengers had to “show that similarly situated individuals of a different race were treated differently.” Id. at 550. And because all students within the rezoned area were treated the same, there was no discriminatory impact. Id. That is, all students within the rezoned area were assigned to the same high school as each other, regardless of race, and all students outside the rezoned area were assigned to the same high school as each other, regardless of race. Id. Based on the district court’s factual findings after a bench trial, we concluded there was no discriminatory impact and applied rational basis review to the policy. See id. at 539–40, 550.

The School District contends that here, like in Lower Merion, all students within the preferred zip codes received preferential treatment, regardless of race, while all students outside the preferred zip codes were also treated equally in the lottery. That analogy misses the mark for several reasons. For starters, the Lower Merion case went to trial and the trier of fact made its own factual findings on discriminatory purpose and impact. 665 F.3d at 539. Our remand order today requires factfinding on discriminatory purpose and impact. 665 F.3d at 539. Second, the school admissions context—a zero-sum game in which the admission of one student means the rejection of another—is fundamentally different from redistricting a public-school zone. Criteria-based high school admissions, especially for schools like Palumbo, Carver, Central, and Masterman, resemble the competitive college admissions processes at issue in SFFA. For every student admitted because

of a potentially racially discriminatory policy, another is rejected. Third, in Lower Merion, the school district had decided to redistrict for nondiscriminatory reasons before it evaluated proposed plans and adopted one with only incidental racial effects. 665 F.3d at 530. Here, conversely, Appellants alleged the Admissions Policy was adopted because the School District wanted to alter the racial makeup of the schools. Finally, it is unclear whether Lower Merion’s permissive attitude towards race-conscious policies and diversity goals survived SFFA. Compare Lower Merion, 665 F.3d at 547–48 (explaining that the district court had wrongly “conflated [use of] race as a factor with [a] discriminatory purpose” (citation modified)), with SFFA, 600 U.S. at 202–03, 231 (holding that using race as a factor in admissions is racially discriminatory in violation of the Equal Protection Clause).

Contrary to the School District’s assertion, the correct inquiry is not whether everyone inside the preferred zip code is treated equally. With a facially neutral and evenly applied policy like this one, all students residing within the zip codes will receive the same treatment. The correct inquiry is whether the antecedent decision to implement the Admissions Policy itself—including the zip code preference—had a discriminatory effect on applicants. We ask whether the Policy “bears more heavily on one race than another.” Arlington Heights, 429 U.S. at 266. For the reasons we have explained, a reasonable factfinder could conclude that it does. As we have noted, the Asian American representation in each of the preferred zip codes was disproportionately low—ranging from only 1.15 percent to 5.7 percent—in contrast to 10 percent in the School District as a whole.

In sum, this case does not involve a remedial consent decree as in Antonelli or a rezoning plan that treated all

students equally as in Lower Merion. Instead, if the factfinder determines that it was adopted for a discriminatory purpose, this Policy would be more like the policies the Supreme Court addressed in SFFA (and its prior cases Grutter and Bakke) that intentionally gave preferences to some racial groups over others in a competitive admissions process, all in pursuit of a student body that matched administrators’ notions of diversity, proportionality, or proper racial balance. See, e.g., SFFA, 600 U.S. at 214; Grutter, 539 U.S. at 330; Bakke; 438 U.S. at 311; see also Parents Involved, 551 U.S. at 726 (plurality opinion) (“In design and operation, the plans [that had assigned K-12 students to public schools, in part on the basis of race] are directed only to racial balance, pure and simple, an objective this Court has repeatedly condemned as illegitimate.”). The only legally significant difference we can identify between this case and SFFA is that the admissions policies at issue in SFFA were not facially neutral. SFFA, 600 U.S. at 195–97. The policies there explicitly stated the race-based motivations of the defendants whereas here, under Arlington Heights, the court must conduct a more searching inquiry. If the factfinder concludes that the School District’s facially neutral policy here was enacted for the same, race-based reasons the Supreme Court rejected in SFFA, the court must also look to that line of cases for guidance, as we do, because “what cannot be done directly cannot be done indirectly.” SFFA, 600 U.S. at 230 (citation modified).

Our sister courts have recently addressed cases involving similar competitive high school admissions policies. See Coal. for TJ, 68 F.4th at 872–75; Bos. Parent Coal., 89 F.4th at 51–55; Chinese Am. Citizens All., 116 F.4th at 164–68. For example, the admissions policy at issue in Coalition for TJ,

on which the District Court most heavily relied, allocated seats at one of the nation’s best public schools—Thomas Jefferson High School for Science and Technology in Alexandria, Virginia—to students from each middle school in the school district. 68 F.4th at 875. Despite that facially neutral classification, Asian American challengers alleged that the new policy was intended to, and did, hinder their chances for admission. Id. at 871. The Fourth Circuit rejected the challengers’ attempt to compare the number and proportion of Asian American students admitted to Thomas Jefferson before and after the policy was implemented. Id. at 880–81. It held that courts could not rely on that kind of data. Id. Doing so would “turn the previous status quo into an immutable quota,” such that schools could never change their policies without fear of “constitutional attack.” Id. at 881 (citation modified).

In directing entry of summary judgment in the school board’s favor, the Fourth Circuit said the district court should have instead compared the “success rate” of Asian Americans against other groups. Id. at 881. And because Asian American students still had “greater success” in securing admission to Thomas Jefferson than other racial groups—despite receiving fewer offers than the year before—the court found no discriminatory impact. Id. at 887. Even in the face of evidence of discriminatory intent and a measurable effect on enrollment, so long as the allegedly disfavored racial group remained more successful than other racial groups, the Fourth Circuit held that it need only apply rational basis review. Id.

We are unpersuaded by the Fourth Circuit’s reasoning in Coalition for TJ. In dissent, Judge Rushing rejected the majority’s conclusion that using “before-and-after” data would turn “the previous status quo into an immutable quota.” Id. at 905 (Rushing, J., dissenting) (citation modified). To her it was

obvious—as it is to us—that these concerns were unfounded because challengers must always prove racially discriminatory intent, not just impact. Id. No facially neutral policy will be held unconstitutional solely because of its disproportionate impact, as the Supreme Court first made clear in 1976. See Davis, 426 U.S. at 242 (“Disproportionate impact is not irrelevant, but it is not the sole touchstone of an invidious racial discrimination forbidden by the Constitution.”). Judge Rushing saw no reason why evidence of a measurable change derived from a straightforward “before-and-after” analysis of enrollment rates by race would not suffice to show discriminatory impact. Coal. for TJ, 68 F.4th at 904–05. Neither do we.10

In Boston Parent Coalition, the First Circuit echoed the Fourth Circuit’s reasoning in Coalition for TJ. 89 F.4th at 57– 58. There, a coalition acting on behalf of parents of Asian and white students challenged the Boston school system’s admissions policy that gave priority enrollment to certain zip

10 On the other hand, simple “before-and-after” data may not suffice to satisfy a factfinder of discriminatory impact in every case, particularly if that data shows only a de minimis or statistically insignificant change. While we recognize that such questions are usually ones for the factfinder, not every “before- and-after” change, no matter how small, will necessarily establish a discriminatory impact. If discriminatory impact were always satisfied, it would compress our Lower Merion framework into a de facto one-part test about discriminatory purpose alone—and would thus contravene our binding precedent. See supra n.6 (discussing Lower Merion’s holding that both purpose and impact are required to trigger strict scrutiny).

codes. Bos. Parent Coal., 89 F.4th at 53. The court acknowledged that the school district “expected” its new admissions policy to reduce the percentage of white and Asian students in certain selective schools. Id. at 57. But it held that the policy did not have a discriminatory impact because white and Asian students remained “stark[ly] over-represent[ed]”— i.e., disproportionately successful in receiving admission, even if less so than before—after its implementation. Id. at 58.

Our disagreement with the First and Fourth Circuits is not without support. The Supreme Court itself looked to a form of before-and-after data when it struck down the universities’ race conscious admissions programs in SFFA. There, “Harvard’s consideration of race ha[d] led to an 11.1% decrease in the number of Asian Americans admitted” and its “policy of considering applicants’ race . . . overall result[ed] in fewer Asian American and white students being admitted.” SFFA, 600 U.S. at 218.11 Neither the Equal Protection Clause nor our precedent requires more complex inquiries.

Though the Supreme Court declined to hear both Coalition for TJ and Boston Parent Coalition, Justice Alito, joined by Justice Thomas, dissented from the denial of certiorari in each case. Coal. for TJ v. Fairfax Cnty. Sch. Bd., --- S. Ct. ---, 2024 WL 674659, at *5 (Feb. 20, 2024) (Alito, J., dissenting from denial of certiorari); Bos. Parent Coal. for Acad. Excellence Corp. v. Sch. Comm. for Bos., 604 U.S. ---,

11 Although the Supreme Court’s discussion of similar data in SFFA was in service of its conclusion that the admissions policies there failed strict scrutiny, 600 U.S. at 213, 218, we see no reason why data showing that a race-based policy fails strict scrutiny could not also help show that the policy is race- based in the first place.

145 S. Ct. 15, 18 (2024) (Alito, J., dissenting from denial of certiorari). And in a separate statement respecting the denial of certiorari in Boston Parent Coalition, Justice Gorsuch explained that he shared many of Justice Alito’s concerns. Bos. Parent Coal., 145 S. Ct. at 15 (Gorsuch, J., statement respecting the denial of certiorari).

Calling the Fourth Circuit’s reasoning “indefensible,” Justice Alito described its holding as “in essence, [] that intentional racial discrimination is constitutional so long as it is not too severe.” Coal. for TJ, 2024 WL 674659, at *1. He deemed the Fourth Circuit’s analysis a “patently incorrect and dangerous understanding” of disparate impact. Id. Justice Alito maintained that all a party must show for disparate impact “is that an admission[s] policy reduced one racial group’s chance of admission and increased another racial group’s chance of admission.” Bos. Parent Coal., 145 S. Ct. at 17. He also expressed concern that the admissions policy at Thomas Jefferson had “been trumpeted to potential replicators as a blueprint for evading SFFA,” Coal. for TJ, 2024 WL 674659, at *5, and that the Fourth Circuit’s “reasoning offer[ed] a roadmap for other federal courts to provide cover,” id. at *5 n.9.

We share Justice Alito’s concerns. In SFFA, the Supreme Court clarified that race-based government action is tolerable only in extraordinary cases. 600 U.S. at 207–08. And in just the past two years, three Justices have expressed concerns about using convoluted disparate impact analyses to dilute the Equal Protection Clause’s force in cases like this one. For these reasons, we respectfully disagree with the First and Fourth Circuits’ approaches to the discriminatory impact inquiry and reject their framework. We will not sanction new exceptions to the Equal Protection Clause’s mandate simply

because a court determines one racial group is, in its view, sufficiently successful. See SFFA, 600 U.S. at 229 (2023) (rejecting the notion of “a judiciary that picks winners and losers based on the color of their skin”); Feeney, 442 U.S. at 277 (“Invidious discrimination does not become less so because the discrimination accomplished is of a lesser magnitude.”).

In our view, the Second Circuit has taken a sounder approach than the First and Fourth Circuits. In Chinese American Citizens Alliance, New York City had changed the admissions policy for its most competitive high schools, reserving a greater proportion of seats for students based “on the economic status of the student applicant’s community as a whole, rather than on an individual basis.” 116 F.4th at 164. The court held that the plaintiffs did not need to show aggregate discriminatory impact against any one racial group; instead, “a valid equal protection claim can be based on a showing that any individual has been negatively affected or harmed by that discriminatory law or policy based on race, even if there is no disparate impact to members of that racial class in the aggregate.” Id. at 173 (emphasis added). The Second Circuit explained that the “discriminatory impact” prong of the analysis could be satisfied by an individualized inquiry into whether the plaintiffs had been harmed by a policy that was enacted with a discriminatory purpose. Id.

To be sure, aggregate disparate impact might still be strong evidence of a constitutional violation, but the Second Circuit explained it was not a prerequisite. Id. at 171–72. “It is axiomatic that ‘[a]t the heart of the Constitution’s guarantee of equal protection lies the simple command that the Government must treat citizens as individuals, not as simply components of a racial, religious, sexual or national class.’” Id. at 173 (quoting

Miller v. Johnson, 515 U.S. 900, 911 (1995)) (citation modified).

We join the Second Circuit in holding that, while evidence of an aggregate discriminatory impact may satisfy the “discriminatory impact” inquiry, it is not the only permissible form of proof. Appellants must show that “similarly situated individuals of a different race were treated differently.” Lower Merion, 665 F.3d at 550. To determine whether they have met this burden, the District Court may look to the aggregate disparate impact data the parties have provided. Statistically significant changes in so-called “before-and-after” data are relevant, including, but not limited to, year-over-year changes to enrollment by race, “qualified applicants” by race, and success rates by race. But on remand, the District Court at trial may also look to whether any plaintiff has shown that he or she “has been negatively affected or harmed by [the] discriminatory law or policy based on race.” Chinese Am. Citizens All., 116 F.4th at 173; see also Cooper v. Harris, 581 U.S. 285, 319 (2017) (“[I]n no area of our equal protection law have we forced plaintiffs to submit one particular form of proof to prevail.”); Gratz v. Bollinger, 539 U.S. 244, 262 (2003) (describing the injury there as “the denial of equal treatment resulting from the imposition of the barrier, not the ultimate ability to obtain the benefit.”). While the form of proof necessary to show that the Admissions Policy caused a reduced likelihood of admission on an individual or aggregate basis may vary, the District Court must ensure that the discriminatory impact is real and identifiable because discriminatory purpose without an actual injury is insufficient to trigger strict scrutiny under our caselaw. Cf. Chinese Am. Citizens All., 116 F.4th at 176–77 (explaining that the “unequal treatment” of Asian Americans students by excluding them

from reserved seats was the injury demonstrating discriminatory impact).

***

Based on the data in the record—including admissions offers year-over-year, the changes in the “qualified applicant” pool resulting from the new criteria, and the demographics of the chosen zip codes—a reasonable factfinder could infer that the 2022 Admissions Policy increased Black and Hispanic students’ chances of admission to Palumbo, Carver, Central, and Masterman while decreasing Asian and white students’ chances. We therefore hold that the record supports, though it does not compel, a finding that the Admissions Policy had both the purpose and effect of discriminating on the basis of race in violation of the Equal Protection Clause of the Fourteenth Amendment to the Constitution.12

IV

The Supreme Court recently clarified that the core purpose of the Equal Protection Clause is “doing away with all governmentally imposed discrimination based on race.” SFFA, 600 U.S. at 206 (citation modified). And “[e]liminating racial discrimination means eliminating all of it.” Id. “[T]he guarantee of equal protection cannot mean one thing when applied to one individual and something else when applied to a person of another color.” Id. (quoting Bakke, 438 U.S. at 289–

12 That is, a factfinder may ultimately determine that the Policy did not have a discriminatory purpose or impact and thus that strict scrutiny does not apply. We also express no opinion on whether the Policy would survive strict scrutiny. See SFFA, 600 U.S. at 206–07; see also supra n.7.

90 (opinion of Powell, J.)).

The Philadelphia School District’s Anti-Racism Declaration aspired to “center[] [its] work through the lens of racial equity.” App. 679. To that end, it tried to “mak[e] progress towards” achieving a student body racial composition proportional to the broader population. App. 1421. After conducting an “equity lens review” of its policies, it adopted the 2022 Admissions Policy. App. 1313. That Policy changed the admissions qualifications and gave preference to six zip codes with disproportionately high Black and Hispanic populations relative to their white and/or Asian American populations. School District officials made public and private statements—both before and after the enactment of the Admissions Policy—that could support a finding that the Policy was intended to alter (and did alter) the racial makeup of the schools. So a reasonable factfinder could conclude that the School District acted with a discriminatory purpose.

A reasonable factfinder could also determine that the Admissions Policy had a discriminatory impact. The data here—though imperfect—reveal potentially significant changes in the racial composition of the “qualified applicant” pool and the student body from 2021 to 2022. And the record shows that the benefits of the zip code policy were more likely to flow to individuals of certain races than others, consistent with the School District’s push for proportional representation at its most competitive high schools.

For those reasons, we will vacate and remand for further proceedings consistent with this opinion.

Jonathan F. Mitchell [Argued] Mitchell Law

Walter S. Zimolong, III James J. Fitzpatrick, III Zimolong

Reed D. Rubinstein Ian Prior America First Legal Foundation

Counsel for Appellants

William K. Kennedy, II Montgomery McCracken Walker & Rhoads

John W. Borkowski [Argued]

Aleksandra O. Rushing Husch Blackwell

Hannah M. Girer-Rosenkrantz Lynn R. Rauch School District of Philadelphia Office of General Counsel

Renee N. Smith Jackson Lewis

Counsel for Appellees

Allison Scharfstein [Argued] Molly M. Cain

Donya Khadem Michaele N. Turnage Young NAACP Legal Defense & Educational Fund

Counsel for Amici Curiae American Civil Liberties Union Foundation, American Civil Liberties Union of Pennsylvania, Advocates for Trans Equality, Asian Americans Advancing Justice-AAJC, Asian American Legal Defense & Educational Fund; Equal Justice Society, LatinoJustice PRLDEF, NAACP Legal Defense & Educational Fund, Inc., National Women’s Law Center, and National Partnership for Women & Families, in support of Appellees