Christa McAuliffe Intermediate School PTO, Inc. v. De Blasio

UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF NEW YORK CHRISTA McAULIFFE INTERMEDIATE SCHOOL PTO, INC.; CHINESE AMERICAN CITIZENS ALLIANCE OF GREATER NEW YORK; ASIAN AMERICAN COALITION FOR EDUCATION; PHILLIP YAN HING WONG; YI FANG CHEN; and OPINION AND ORDER CHI WANG, 18 Civ. 11657 (ER) Plaintiffs, - against - BILL de BLASIO, in his official capacity as Mayor of New York; and RICHARD A. CARRANZA, in his official capacity as Chancellor of the New York City Department of Education, Defendants. Ramos, D.J.: Plaintiffs bring this action against Bill De Blasio, former Mayor of New York, and Richard A. Carranza, former Chancellor of the New York City Department of Education (“DOE”), claiming that the Mayor and Chancellor’s changes to the admissions process for the eight specialized New York City public high schools violate the Equal Protection Clause of the Fourteenth Amendment of the United States Constitution by discriminating against Asian American students. Plaintiffs are three organizations (Christa McAuliffe Intermediate School PTO, Inc., Chinese American Citizens Alliance of Greater New York, and Asian American Coalition for Education), and three individuals who are parents of students in New York City public schools (Phillip Yan Hing Wong, Yi Fang Chen, and Chi Wang). Before the Court is Defendants’ motion for summary judgement. For the reasons set forth below, the motion is GRANTED. I. Factual Background The Court assumes familiarity with the facts and procedural posture of this action, previously set forth in its March 4, 2019 Order, see Christa McAuliffe Intermediate School PTO, Inc. v. de Blasio (“McAuliffe I”), 364 F. Supp. 3d 253 (S.D.N.Y. 2019), Doc. 69, and March 24, 2020 Order, see Christa McAuliffe Intermediate School PTO, Inc. v. de Blasio (“McAuliffe II”, No. 18 Civ. 11657 (ER), 2020 WL 1432213 (S.D.N.Y. Mar. 24. 2020), Doc. 124. As such, the Court will summarize only the facts necessary to resolve this motion. These facts are taken from this Court’s previous decisions and from parties’ Rule 56.1 statements. DOE operates eight highly prestigious1 specialized high schools—Bronx High School of Science (“Bronx Science”); Stuyvesant High School (“Stuyvesant”); Brooklyn Technical High School; Brooklyn Latin School; High School for Mathematics, Science and Engineering at City College of New York; High School of American Studies at Lehman College; Staten Island Technical High School; and Queens High School for the Sciences at York College (collectively, the “Schools”). Historically, Black and Latino students have been underrepresented at the Schools compared to the City’s public school system overall. See McAuliffe I, 364 F. Supp. 3d at 266. As noted in the Court’s March 4, 2019 Order, the racial makeup of New York City’s public high schools at that time was 40% Hispanic, 26% Black, 16.1% Asian American, and 15% white. Id. In sharp contrast, the racial makeup of Stuyvesant, the second largest of the Schools, was 1 The public generally regards these schools are “elite,” “exclusive,” and “among the best . . . in the country.” McAuliffe I, 364 F. Supp. 3d at 265 (citing The Exclusive Eight, N.Y. Times (Oct. 16, 2012), available at https://www.nytimes.com/2012/10/17/opinion/the-exclusive-eight- highschools.html?rref=collection%2Ftimestopic%2FStuyvesant%20High%20School&action=click&contentCollecti on=timestopics&region=stream&module=stream_unit&version=latest&contentPlacement=27&pgtype=collection; Laura Meckler, NYC plan to diversify elite high schools challenged in court, Wash. Post (Dec. 13, 2018), available at https://www.washingtonpost.com/local/education/nyc-plan-to-diversify-elite-high-schools-challenged-in- court/2018/12/13/37810eb6-ff20-11e8-862a-b6a6f3ce8199_story.html?utm_term=.b68752394a36). 73.5% Asian American, 17.8% white, 2.8% Hispanic, and 0.7% Black. Id. The other Schools were more representative, but none come close to proportionate representation. Id. While Black and Hispanic students made up 66% of New York City public high schools, they only made up 13.5% of Brooklyn Tech, 8.7% of Bronx Science, 3.5% of Staten Island Tech, 23.9% of Brooklyn Latin, 25.2% of the High School for Math, Science & Engineering, 8.4% of Queens High School for the Sciences, and 15% of the High School of American Studies. Id. Asian American students made up 61.3% of Brooklyn Tech, 65.6% of Bronx Science, 48.4% of Staten Island Tech, 51.5% of Brooklyn Latin, 36.2% of the High School for Math, Science & Engineering, 81% of the Queens High School for the Sciences, and 22% of the High School of American Studies. Id. For decades, the demographically skewed student populations of the Schools have attracted scrutiny from civil rights groups and government agencies. Id. Despite the DOE’s multi-faceted efforts over the years, the problem of Black and Latino underrepresentation in the Schools has, if anything, seemed to worsen. Id. The Schools admit applicants solely on the basis of an academic exam, as required by the Hecht-Calandra Act (“the Act”), see N.Y. Educ. L. § 2590-g(12)(b) (1997). See McAuliffe II, 2020 WL 1432213 at *1; McAuliffe I, 364 F. Supp. 3d at 264. The Act specifies that that admission to the Schools “shall be [determined] solely and exclusively by taking a competitive, objective and scholastic achievement examination.” N.Y. Educ. L. § 2590-g(12)(b) (1997). To apply, students order their preference for the Schools and then take the Specialized High School Admissions Test (“SHSAT”). McAuliffe I, 364 F. Supp. 3d at 264. The state then scores the exams and ranks them from highest to lowest. Id. The student with the highest score is offered a seat at his first-choice school. Id. The student with the next highest score is then offered a seat in his first-choice school, and so on, until all the seats in a student's first-choice school have been filled. In that case, the student is offered a seat in his second-choice school. Id. If all the seats in the second-choice school have been filled, the student is placed in his third-choice school, and so on. Id. This process continues until all the seats at the Schools have been filled. Because of this system, after each admissions cycle, each School has a cut-off score for admission: the SHSAT score of the last student offered admission. Id. The Act provides only one other method of admission: the Discovery Program (the “Program” or “Discovery”). Id. To be eligible for Discovery, a student must: (1) be disadvantaged; (2) be certified by his current school as being “high potential”; (3) score just below the lowest overall score of all admitted students; and (4) pass a summer preparatory program demonstrating his ability to “cope with the special high school program.” McAuliffe I, 364 F. Supp. 3d at 265. Importantly, the Act neither defines “disadvantaged” nor prescribes the number of students that may be admitted through the Program. McAuliffe II, 2020 WL 1432213 at *1. The Act leaves those determinations to the discretion of the Chancellor. Id. In spring 2018, a DOE working group recommended to Chancellor Carranza that he change the Program’s eligibility criteria to increase the racial, ethnic, geographic, and socio- economic diversity of the Schools. Id. at *2. Specifically, the group recommended that Discovery be expanded to 20% of the seats at each School over a two-year period and that the definition of “disadvantaged” be modified. See Doc. 156 ¶¶ 12–13, 16. On June 3, 2018, Chancellor Carranza accepted the recommendations, announcing a 20% expansion of the Program over two years,2 along with a redefinition of “disadvantaged,” which went into effect in September 2019. Id. ¶¶ 11–12. Plaintiffs contest both of these changes. Id. ¶ 11. 2 In the summer of 2019, 528 seats were to be allocated to the Program, which is approximately 13% of the total seats at the eight Schools. See Doc. 156 ¶ 14. And in the summer of 2020, approximately 800 seats were to be allocated for Discovery, approximately 20% of the total seats at the eight Schools. Id. Prior to the reforms, a student was considered disadvantaged if the student’s family met one of the following criteria: (1) the student’s family income qualified the student for free lunch or reduced price lunch based on the student’s meal form; (2) the student’s family received assistance from the Human Resources Administration;3 (3) the student was in foster care, a ward of the state, or was a Student in Temporary Housing; or (4) the student was an English Language Learner or was an English Language Learner within the previous 2 school years, and enrolled in a DOE school for the first time within the last four years. Id. ¶ 16. Though the Chancellor left those four criteria mostly intact,4 he added a new requirement: that a DOE student attended a school with an Economic Need Index (“ENI”) of 0.6 or above.5 A middle school’s ENI is determined by averaging the Economic Need Value (“ENV”) of each student. Id. at 2. ENV is a measure created by DOE to assess the percentage of students in each intermediate school facing economic hardship.6 Id. ¶ 19, 21. A student’s ENV is 1.0 if: (1) the student lives in a household that is eligible for public assistance from the New York City Human Resources Administration; (2) the student lived in temporary housing in the past four 3 According to information taken from the publicly available Demographic Snapshot, of which this Court has taken judicial notice, see McAuliffe I, 364 F. Supp. 3d at 253, 262, eligibility for public assistance is defined as eligibility for any of the following programs: the Supplemental Nutrition Assistance Program, 7 U.S.C. § 2011, et seq.; the Temporary Assistance for Needy Families Program, 42 U.S.C. § 601, et seq.; Medicaid, 42 U.S.C. 1396, et seq.; and the Special Supplemental Nutrition Program for Women, Infants, And Children, 42 U.S.C. § 1786. Doc. 156 at ¶ 25, n.2. 4 Criterion 1 now states: the students’ family income qualifies the student for free or reduced-price lunch. Criterion 2 now states: the student’s family receives assistance from the New York City Human Resources Administration. See id. ¶ 17. 5 Modeling conducted by the DOE working group projected that adding the ENI requirement would only slightly change the racial makeup of the Program. Id. Specifically, the projections showed a slight decline of Asian American and white enrollment and a slight increase in Black and Hispanic enrollment. See McAuliffe I, 364 F. Supp. 3d at 268. DOE policymakers had access to these projections while considering and designing the changes to Discovery. Id. 6 Many social science research studies support the idea that students attending high-poverty schools are more disadvantaged and face more academic challenges than similarly situated students attending lower-poverty schools. Doc. 156 ¶ 17. years; or (3) has a home language other than English and enrolled in a DOE school for the first time within the past four years. Id.7 The Plaintiffs note, and Defendants do not dispute, that other methods exist to measure poverty level, one of which—the Demographic Snapshot—shows a 67% poverty level at Christa McAuliffe. ¶ 74. The ENI formula does not rely entirely on individual circumstances, as some students receive an ENV based on the census tract in which they live, see supra n.7. Id. ¶ 18. As a result of the revised criteria, only students who attend an intermediate school with a relatively low-income student body are eligible for Discovery. McAuliffe I, 364 F. Supp. 3d at 268. Thus, if a student is very low-income but attends an intermediate school with an ENI below 60%, the student is ineligible for Discovery, even though that student would have been eligible for the Program under the prior criteria. Id. About half of all New York intermediate schools have an ENI below 60%. Id.8 After adopting the changes to Discovery, Mayor de Blasio and Chancellor Carranza publicly praised the reforms, saying that the new plan would increase racial diversity at the Schools. Id. at 269. The Mayor issued a press release stating that the changes would support 7 If a student does not meet any of the above criteria, then the student’s ENV is calculated based on the percentage of families (with school-age children) in the student’s census track whose income is below the poverty income level, as estimated by the United States Census Bureau’s American Community Survey Five-Year estimate. The student’s ENV equals this percentage divided by 100. If the student attends a non-DOE school, such as a charter or private school, or is home-schooled, then the criterion used for purposes of defining disadvantage is whether the student resides in a census tract where 60% or more of the families with school-age children living in the census tract have an income below the poverty line, as estimated by the U.S. Census Bureau’s American Community Survey Five- Year estimate. Doc. 156 ¶¶ 26–27. 8 According to documents submitted after the complaint was filed, before students apply, DOE sends invitations to parents of students who may be eligible for the Program. DOE decides to whom to send these invitations based on middle schools’ recommendations; middle schools base their recommendations on applications made by parents. For many of its students, DOE has records that show if a student is disadvantaged, but for some students DOE lacks such records. DOE also has no records that show whether a non-DOE student attending a private or charter school meets the individual disadvantaged criteria. Rather than eliminate any potentially eligible students, DOE sends invitation letters to all students who are recommended and meet the other requirements for eligibility: SHSAT score and attending a school with an ENI of at least 60%. See Doc. 148 ¶ 40–47; see also Doc. 149 at 17. greater geographic, racial, and socioeconomic diversity at the Schools. Id. (quoting Press Release, Mayor de Blasio and Chancellor Carranza Announce Plan to Improve Diversity at Specialized High Schools, Office of the Mayor of New York City (June 3, 2018)). The press release also stated that with the new changes, “[b]ased on modeling of current offer patterns, an estimated 16% of offers would go to Black and Latino students, compared to 9% currently.” Id. That same day, Mayor de Blasio posted a string of messages on his official Twitter account, stating: Stuyvesant High School just admitted almost a thousand students, but only ten of those students were African American and less than thirty were Latino. In a city that is majority African American and Latino. These schools are the proving grounds for future leaders, and unless we believe our leaders should only come from certain communities, we cannot have our most prestigious schools available to only some. Our first reforms will commit 20% of the seats to kids from disadvantaged communities. And we will work with Albany to eliminate a system where one broken test dictates a child’s future. Id. at 263.9 Plaintiffs acknowledge that the proposed changes to the Program are facially race-neutral but brought the instant action, alleging that that the changes would have a disparate impact on Asian American students, and that Defendants intended this effect. See generally Doc. 1. Under the reforms—and after the complaint was filed—the percentage of offers received by Asian American students who took the SHSAT increased in 2019 and 2020 more than they would have if offers had been made using the admissions criteria for the 2018 entering class. See Doc. 156 ¶ 29. Representation of Asian American students at the Schools also increased in 2019 and 2020. Id. ¶ 28. Additionally, in 2019, Asian American students constituted 30.7% of the applicant pool and received 52.5% of the total offers extended; in 2020, Asian American 9 For the reasons set forth in this Court’s March 4, 2019 Order, see McAuliffe I, 364 F. Supp. 3d at 261–64, the Court takes judicial notice of these June 3, 2018 tweets. students constituted 31.4% of the applicant pool and received 54.8 % of the total offers extended. Moreover, in 2019, 33.2% of Asian American students who applied received offers, and in 2020 31.9% of Asian Americans who applied received offers. See Doc. 149 at 1–2. A higher percentage of Asian American test takers received offers in 2019 and 2020, than any of the other major racial or ethnic groups.10 However, for two of the Schools, Stuyvesant and Bronx Science, the reforms resulted in a decrease in the representation of Asian American students in 2020 compared to the prior year, id; see also Doc. 154-1, Expert Witness Report of Jacob L. Vigdor, Ph.D., but only slightly so.11 II. Procedural History Plaintiffs filed their complaint and a motion for preliminary injunction on December 13, 2018, alleging that the revised Program violated the Equal Protection Clause of the Fourteenth Amendment. Docs. 1, 10. On January 10, 2019, Defendants answered the complaint, Doc. 42, and on January 16, 2019, they requested an expedited decision on the preliminary injunction motion, citing administrative concerns, Doc. 43. On February 25, 2019, the Court denied Plaintiffs’ preliminary injunction motion, finding that the revised Program would likely survive both rational basis and strict scrutiny review. Docs. 66, 69; see McAuliffe I, 364 F. Supp. 3d at 273–84. The Court also found that only the three organizations and one of the individual plaintiffs had standing. See McAuliffe I, 364 F. Supp. 3d at 271–76. Plaintiffs appealed. Doc. 68. On December 20, 2019, the Second Circuit affirmed the decision but found that, on the 10 The second highest performing group was white students, with 28.2% receiving offers in 2019 and 23.7% receiving offers in 2020. See Doc. 149 at 2. 11 Asian American students received 66.9% of the offers to attend Stuyvesant but would have received 67.6% offers if the pre-implementation 2018 criteria were used. As Defendants note, this is a difference of .7% or 9 students. See Doc. 159 at 10. Applying the same framework to Bronx Science shows that Asian American students received 55.8% of the offers in 2020 “but would have received 57.2% if the 2018 criteria were used—a difference of 1.4% or 13 students.” Id. record before it, the organizational plaintiffs also lacked standing, though it suggested that this defect might be cured. Doc. 119. On March 24, 2020, the Court granted a motion to intervene filed by a group of organizations and children (i.e., Teens Take Charge, Desis Rising Up and Moving, Hispanic Federation, and Coalition for Asian American Children and Families; as well as O.R., by and through his mother, Elizabeth Pierret; A.S., by and through his father, Odunlami Showa; C.M., by and through his mother, Rosa Velasquez; K.B. by and through her mother Tiffany M. Bond; and N.D.F. and N.E.F., by and through their mother Lauren R. Mahoney), seeking to defend the reforms to Discovery. See generally McAuliffe I, 2020 WL 1432213. On October 1, 2021, Defendant moved for summary judgment. Doc. 144. III. Legal Standard a. FRCP 56(a) Motion for Summary Judgment Standard Summary judgment is appropriate where “the movant shows that there is no genuine dispute as to any material fact.” Fed. R. Civ. P. 56(a). “An issue of fact is ‘genuine’ if the evidence is such that a reasonable jury could return a verdict for the non-moving party.” Senno v. Elmsford Union Free Sch. Dist., 812 F. Supp. 2d 454, 467 (S.D.N.Y. 2011) (quoting SCR Joint Venture L.P. v. Warshawsky, 559 F.3d 133, 137 (2d Cir. 2009)). A fact is “material” if it might affect the outcome of the litigation under the governing law. Id. The party moving for summary judgment is first responsible for demonstrating the absence of any genuine issue of material fact. Celotex Corp. v. Catrett, 477 U.S. 317, 323 (1986). If the moving party meets its burden, “the nonmoving party must come forward with admissible evidence sufficient to raise a genuine issue of fact for trial in order to avoid summary judgment.” Saenger v. Montefiore Med. Ctr., 706 F. Supp. 2d 494, 504 (S.D.N.Y. 2010) (internal quotation marks omitted) (citing Jaramillo v. Weyerhaeuser Co., 536 F.3d 140, 145 (2d Cir. 2008)). In deciding a motion for summary judgment, the Court must “‘construe the facts in the light most favorable to the non-moving party and must resolve all ambiguities and draw all reasonable inferences against the movant.’” Brod v. Omya, Inc., 653 F.3d 156, 164 (2d Cir. 2011) (quoting Williams v. R.H. Donnelley, Corp., 368 F.3d 123, 126 (2d Cir. 2004)). However, in opposing a motion for summary judgment, the non-moving party may not rely on unsupported assertions, conjecture, or surmise. Goenaga v. March of Dimes Birth Defects Found., 51 F.3d 14, 18 (2d Cir. 1995). To defeat a motion for summary judgment, “the non-moving party must set forth significant, probative evidence on which a reasonable fact-finder could decide in its favor.” Senno, 812 F. Supp. 2d at 467–68 (citing Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 256–57 (1986)). IV. Discussion Defendants contend that Plaintiffs cannot prevail on their Equal Protection Claim because the challenged policies have not had a discriminatory effect on Asian American students. See Doc. 149 at 10. For the reasons explained below, the Court agrees. a. The Equal Protection Clause i. Discriminatory Effect & Discriminatory Intent The Equal Protection Clause of the Fourteenth Amendment to the United States Constitution provides that “[n]o State shall make or enforce any law which . . . deny to any person within its jurisdiction equal protection of the laws.” U.S.C.A. Amend. 14. “The central purpose of the Equal Protection Clause . . . is the prevention of official conduct discriminating on the basis of race.” Washington v. Davis, 426 U.S. 229, 239 (1976). The government discriminates on the basis of race when: (1) a law or statute expressly classifies individuals by race, see Adarand Construtors, Inc. v. Pena, 515 U.S. 200, 213 (1995); or (2) the government enforces a facially neutral statute in a racially discriminatory way, see Yick Wo v. Hopkins, 118 U.S. 356, 373–74 (1886). Here, however, the challenged reforms to the Program are facially neutral because they do not classify students by race, and Plaintiffs do not allege that they are being administered in a racially discriminatory way. See generally Doc. 1. Nonetheless, a facially neutral statute or policy that is not enforced in a discriminatory manner can still violate equal protection if: (1) its application results in a discriminatory effect, and (2) it is motivated by a discriminatory purpose. See Davis, 426 U.S. at 42; see also Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 264–65 (1977). Caselaw makes clear that absent extreme discriminatory effect—of the kind present, for example, in Gomillion v. Lightfoot, 364 U.S. 339, 81 (1960) or Yick Wo v. Hopkins, 118 U.S. 356 (1886)—both discriminatory purpose and discriminatory effect are required to prevail on an equal protection claim against such a policy. In Davis, the Supreme Court addressed whether a facially neutral policy could violate the Equal Protection Clause based purely on its disproportionate impact. See 426 U.S. 229. The court held that it could not: that racially discriminatory purpose must also be shown. Id. at 249. Nowhere in its opinion did the court regard proof of discriminatory effect unnecessary to prevail on an equal protection claim. Instead, the Court adopted “the basic equal protection principle” it had previously advanced in its school desegregation cases:12 “The invidious quality of a law claimed to be racially discriminatory must ultimately be traced to a racially discriminatory purpose.” Id. at 240. Put differently, a law shown to have discriminatory effect must also have been motivated by discriminatory intent. One without the other will not suffice. The court in Arlington Heights applied that rule, finding a purportedly discriminatory ordinance constitutional 12 See, e.g., Keyes v. School Dist. No. 1, Denver, Colo., 413 U.S. 189, 208, (1973) absent proof that “discriminatory purpose was a motivating factor in the Village’s decision.” See 492 U.S. at 266, 270 (“Absent a pattern as stark as that in Gomillion[, 364 U.S. 339,] or Yick Wo[, 118 U.S. 356], impact alone is not determinative.”) As has the Second Circuit. See, e.g., United States v. City of Yonkers, 96 F.3d 600, 611 (2d Cir. 1996) (“A plaintiff is required to show not only that the state action complained of had a disproportionate or discriminatory impact but also that the action was taken with intent to discriminate.”). In Hayden v. County of Nassau, 180 F.3d 42, 48 (2d Cir. 1999), for example, the Second Circuit upheld a decision dismissing the claims of white and Latino police department applicants that the department’s entrance exam—which was designed to reduce discriminatory impact on minority candidates—violated the Equal Protection Clause. In doing so, the Court explained that “a facially neutral statute violates equal protection if it was motivated by discriminatory animus and its application results in a discriminatory effect.” Id. at 48 (emphasis added). Because plaintiffs “fail[ed] to set forth allegations which would support a claim that they were adversely impacted by the redesign of the . . . entrance exam,” the action could not be sustained. Id. at 51– 52; accord Atkins v. Westchester Cty. Dep’t of Soc. Serv., 31 F. App’x 52 (2d Cir. 2002) (denying the plaintiffs’ equal protection claim against a purportedly discriminatory entrance exam because they did not make a “prima facie showing that the . . . exam . . . exerted a disparate impact on African-American test-takers”).13 13 Additionally, in the instant case, the Second Circuit affirmed the Court’s denial of Plaintiffs’ motion for preliminary injunction wherein this Court explained that “a law or policy that is facially neutral discriminates on the basis of race if it is motived by a discriminatory purpose and its application results in a discriminatory effect.” See McAuliffe I, 364 F. Supp. 3d at 276 (S.D.N.Y. 2019) (emphasis added) (citing Village of Arlington Heights, 429 U.S. at 264–65); accord Lewis v. Ascension Par. Sch. Bd., 806 F.3d 344, 358–59 (5th Cir. 2015); Doe v. Lower Merion Sch. Dist., 665 F.3d 524, 549–50 (3d Cir. 2011); Garza v. Cty. of Los Angeles, 918 F.2d 763, 771 (9th Cir. 1990). ii. Defining Discriminatory Effect To determine whether a policy has produced a racially discriminatory effect, courts— consistent with earlier decisions construing Title VII of the Civil Rights Act of 1964, see, e.g., Griggs v. Duke Power Co., 401 U.S. 424 (1971)—look to whether the policy has disproportionately affected a racial group in the aggregate.14 In Davis, for example, the court described disproportionate effect to mean: “affect[ing] a greater portion of one race than another” or, said differently, “bear[ing] more heavily on one race than another.” 426 U.S. at 242. Similarly, in Arlington Heights, the court left intact the Seventh Circuit’s approach of examining a policy’s “ultimate effect” on a racial group to discern discriminatory impact. 429 U.S. at 557, 566. The Second Circuit has adopted the same construction. In Hayden, for example, the court rejected plaintiffs’ equal protection claim because they could not show that the exam “injured or disadvantaged [them] in some way.” 180 F.3d at 52–53. Although the court noted that plaintiffs “suffered no discriminatory impact in the administration or scoring of the [] exam,” (i.e., that all test-takers were treated equally, regardless of race), the court rested its holding that the new test had no discriminatory effect on the fact that plaintiffs “continued to score higher than [B]lack candidates, on average.” Id. In Atkins, the Second Circuit construed discriminatory effect in a like manner, looking to the net effect of a purportedly discriminatory employment exam against Black employees. See 31 F. App’x at 53 (finding no disparate impact where “plaintiffs [could] offer no statistical evidence to suggest that the difference between the score distributions for whites and African-Americans [was] statistically significant and not 14 Courts regularly use the terms “disproportionate impact,” see Davis, 426 U.S. at 265, “disproportionate effect[],” see Hayden, 180 F.3d at 54, and “disparate impact,” see Atkins, 31 Fed.Appx.at 53, interchangeably with “discriminatory effect.” attributable to chance.”). Both in Hayden and Atkins, the Second Circuit centered its focus on the aggregate impact the policy had on specific racial groups—not on whether individuals who belonged to a given racial group received disproportionately unequal treatment, alone. b. Plaintiffs’ Construction of Discriminatory Effect With respect to the question of discriminatory effect, Plaintiffs contend that they need not show that the Discovery reforms have disproportionately harmed Asian American students in the aggregate. They key inquiry, they say, “is whether there is equal treatment.” See Doc. 153 at 10. According to Plaintiffs “[i]f the challenged policy treats all applicants equally, it is unlikely to violate the Equal Protection Clause. But where the government shifts to a policy that treats applicants differently based on a factor that is designed to operate as a proxy for race, it becomes suspect.” Id. at 10–11. In Plaintiffs’ view, neither Davis nor Arlington Heights requires them to prove that the reforms to Discovery disproportionately impact Asian American students. Davis, they say, addressed solely “whether a facially neutral policy could violate the Equal Protection Clause based on disproportionate impact alone.” Id. at 8. And Arlington Heights merely provides a method for courts to check for discriminatory purpose. Id. (citing 429 U.S. at 265–66). The Court in neither case, Plaintiffs assert, expressly required a showing of discriminatory impact in the aggregate. Id. at 8–9 (citing Arlington Heights, 429 U.S. at 266 (describing aggregate impact analysis as not more than “an important starting point”)). Similarly, Plaintiffs claim that in Hayden, the equal protection argument failed “not because more white and Hispanic applicants passed the test in the aggregate, but because the applicants were neither ‘excluded from full consideration because of their race, nor were they disadvantaged because of their race.’” Id. at 10 (emphasis added) (citing Hayden, 180 F.3d at 52). The Second Circuit’s analysis turned on the question of equal treatment—not the net effect of the test on white and Hispanic applicants’ employment. Id. In Hayden, test-takers were treated equally, regardless of race. But here, Plaintiffs argue, Asian American students are not. Plaintiffs contend that the reforms treat Asian Americans unequally in two ways. First, the reforms prevent economically disadvantaged students attending ENI schools of 0.6 or less from applying to Discovery. See Doc. 153 at 11. And second, they make it harder for students at Discovery-ineligible schools to gain admission by raising the “exam score cutoff . . . higher than it . . . would have been [before] Discovery expansion.” Id. “The result is that two individually disadvantaged applicants who are in all other ways similarly situated . . . will be treated differently in the admissions process based on the school they attend in eighth grade.” Id. Plaintiffs allege that although the reforms are facially neutral, DOE uses ENI as proxy for race, and thereby engages in unconstitutional racial balancing. Id. at 10 (citing Ass’n for Educ. Fairness v. Montgomery Cty. Bd. of Educ., 560 F. Supp. 3d 929, 952–56 (D. Md. 2021) (finding plausible that a county’s use of “local norming” and “peer grouping” in admissions to prestigious magnet middle schools had a disparate impact against Asian Americans)). Plaintiffs’ argument is without merit for two reasons. First and foremost, the idea that discriminatory effect can be shown by way of “unequal treatment”—without any evidence of aggregate harm—conflicts with Supreme Court precedent and lacks any support. As this Court has already explained, to prove discriminatory effect, a plaintiff must show that a policy has disparately impacted a particular racial group. Plaintiffs have not identified any case in which a court has deemed proof of disparate impact unnecessary to support an equal protection claim against a facially neutral policy enforced in a non-discriminatory manner.15 Nor do Plaintiffs identify a single case that has held that discriminatory impact can be proved by showing “unequal treatment” when the policy has not resulted in disproportionate harm to a particular group. As noted, Davis, Arlington, and their progeny consider such an effect necessary to sustain an equal protection violation. Additionally, Plaintiffs’ reliance on Montgomery for the proposition that “where the government [adopts] a policy that treats applicants differently based on a factor that is designed to operate as a proxy for race, it becomes suspect” is misplaced. Doc. 153 at 10. In Montgomery, the proxy actually resulted in harm to Asian American students. See 560 F. Supp. 3d at 952 (“As to disparate impact, no real dispute exists that the [new admissions] criteria disproportionately affected Asian American students.”). The court did not obviate the requirement to prove disparate impact. In fact, the court in Montgomery adopted a reading of Arlington Heights in conflict with Plaintiffs’ reading of the opinion. See id. (noting that “[p]laintiff alleges intentional discrimination and so must make plausible that the field test disparately impacted Asian American students[.]”) (emphasis added) (citing Arlington Heights, 429 U.S. at 265; Lower Merion, 665 F.3d at 549–50)). Second, even assuming, arguendo, that discriminatory effect can be proved by showing unequal treatment, the reforms do not treat students differently on the basis of race. Disadvantaged students attending schools with an ENI of 0.6 or more have the same eligibility to 15 Indeed, the only other case cited by Plaintiffs is Ass’n for Educ. Fairness v. Montgomery Cty. Bd. of Educ., 560 F. Supp. 3d 929 (D. Md. 2021). There, a Maryland court denied defendants’ motion to dismiss an equal protection claim against a county’s changes to the admissions process of magnet middle schools. This case is both non- controlling and unhelpful to Plaintiffs’ point. The contested program in Montgomery reduced the percentage of offers received by Asian American students (the purportedly targeted group) and increased the percentage of offers received by Latino and white students. See id. at 25. So unlike here—as discussed more fully below—there was evidence of an adverse effect on Asian Americans in Montgomery. participate in the Program. And all disadvantaged students attending schools with an ENI of less than 0.6 are equally prohibited from participating in Discovery, regardless of their race. ENI assesses a school’s poverty level by averaging the poverty of each student. Even if some students’ ENV is determined by geography, ENI, at its core, is a measure of economic disadvantage—not of a school’s racial composition. As Defendants note, scientific studies show that “low-income students attending high-poverty schools are more disadvantaged and face more academic challenges than similarly-situated students attending lower-poverty schools.” Doc. 159 at 11 (citing, e.g., Annotated Bibliography: The Impact of School-Based Poverty Concentration on Academic Achievement & Student Outcomes, Poverty & Race Research Action Council). Plaintiffs offer no evidence to the contrary. Racial balancing, furthermore, refers to a process structured to cause “some specified percentage of a particular group merely because of race or ethnic origin.” Fisher v. Univ. of Texas at Austin, 570 U.S. 297, 311 (2013) (internal quotation marks and citation omitted). The reforms to Discovery challenged here designate no such specific percentage. The Court therefore finds that the fact that a very low-income student attending an intermediate school with an ENI below 60% (e.g., a student attending McAuliffe) is ineligible for Discovery does not render the changes to the Program racially unequal, even if that student would have been eligible for Discovery under the prior criteria. c. Application To sustain their claim therefore, Plaintiffs must show that the reforms: (1) have resulted in a discriminatory effect, and (2) were motivated by a discriminatory purpose. See Davis, 426 U.S. at 42. i. Discriminatory Effect16 The reforms have not resulted in a discriminatory effect upon Asian American students. Defendants ask the Court to apply two well-established methods for analyzing disparate impact. Plaintiffs, meanwhile, ask the Court to adopt another, lesser used—and recently criticized— method. In any event, under all three methods, there is no evidence that the reforms to Discovery have adversely affected Asian American students. According to Defendants, to discern whether a facially neutral policy has a discriminatory effect, courts should consider: (i) the racial/ethnic composition of the passing pool (here, students who received offers) to that of the applicant pool (here, students who took the test for admissions), or (ii) the comparative passing rate of each racial/ethnic group who receive offers (i.e., compare[] the percentage of Asian test-takers who receive 16 As a preliminary matter, the Court finds that Plaintiffs have abandoned their argument that the announcement of the reforms to Discovery have had a discouraging effect on Asian American students, resulting in detrimental behavioral changes. According to Plaintiffs’ expert, Dr. Jacob Vigdor, when Defendants announced the changes to the Program, they sent an “implicit message [to Asian Americans] that their group was overrepresented[.]” See Doc. 154-1 at 21. Dr. Vigdor posits that the announcement could have thereby discouraged Asian American students from undertaking certain beneficial admissions practices (e.g., listing Stuyvesant as a top choice) and reduced Asian American students’ chances of receiving an offer. See id. at 21–25. In Defendants’ memorandum in support of their motion, they call Dr. Vigdor’s claim a “completely unprecedented basis for a finding of disparate impact,” drawing particular attention to Dr. Vigdor’s failure to undertake any qualitative research. See Doc. 149 at 30. Plaintiffs, however, fail to address any of Defendants’ arguments relating to this issue in their opposition. See generally Doc. 153. As the Second Circuit has made clear, “[p]leadings often are designed to include all possible claims. Moreover, preparation of a response to a motion for summary judgment is a particularly appropriate time for a non-movant party to decide whether to pursue or abandon some claims or defenses. Indeed, Rule 56 is known as a highly useful method of narrowing the issues for trial.” Jackson v. Fed. Express, 766 F.3d 189, 196 (2d Cir. 2014). Therefore, in deciding “whether summary judgment is legally and factually appropriate . . . a court may . . . infer from a party’s partial opposition that relevant claims or defenses that are not defended have been abandoned.” Id. at 197–98. Here, Plaintiffs fail to respond to any of Defendants’ critiques of Dr. Vigdor’s testimony that the announcement caused Asian American to change their behavior. Accordingly, the Court finds this argument abandoned. offers to the percentages of Black, Hispanic and white test-takers who receive offers). Doc. 149 at 7. Put differently, Defendants ask the Court to “compare the applicant pool and offer pool data,” by comparing either: (1) “the percentage of the plaintiff group in the applicant pool to the percentage of offers received by that group,” or (2) “the passing or failing rates of the plaintiff group to other groups.” See Doc. 159 at 9. As Defendants make clear, courts have routinely adopted these two approaches to check for disparate impact in the educational context. See Davis, 426 U.S. at 237 (noting that four times as many Black applicants failed the test than did white applicants); Paradise v. Prescott, 580 F. Supp. 171, 172 (M.D. Ala. 1983), aff’d, United States v. Paradise, 480 U.S. 149, 186 (1987) (measuring adverse impact according to the “four-fifths rule,” which provides that the passing rate of any group must be at least four fifths of the rate of the group with the highest passing rate); Hayden, 180 F.3d 42 at 52 (failing to see how plaintiffs could establish a claim for prejudice after “conced[ing] that, on average, they scored higher than [B]lack applicants on the” exam); United States v City of N.Y., 637 F. Supp. 2d 225, 261 (E.D.N.Y. 2009) (finding that the much lower passing rate of Black candidates on a firefighter’s entrance exam compared to other groups’ passing rates satisfied the discriminatory impact prong of the equal protection claim). Applying Defendants’ first method, the Court finds no evidence of disparate impact. In 2019, Asian American students were 30.7% of the applicant pool and received 52.5% of the offers to attend one of the Schools; in 2020, Asian American students were 31.4% of the applicant pool and received 54.8 % of the offers. See Doc. 149 at 7–8. Applying the second method, the Court also sees no evidence of disparate impact. A higher percentage of Asian American test takers received offers in 2019 and 2020 than any other major ethnic or racial group. Id. at 8. In 2019, 33.2% of Asian American students who applied received offers; the second-highest performing group was white students, with 28.2% receiving offers. Id. at 7–8. In 2020, 31.9% of Asian Americans who applied received offers; the second highest performing group was, again, white students, with 23.7% receiving offers. Id. at 8. Plaintiffs’ complaint, motion for preliminary injunction, and opposition to Defendants’ motion for summary judgment assert that the proper way to check for disparate impact is to compare Asian American students’ performance before and after the contested reforms to the Program. See Doc. 1 ¶¶ 40, 48; Doc. 11 at 10–13; Doc. 153 at 15. The First Circuit, however, recently criticized this method of determining disparate impact. See Boston Parent Coal. For Acad. Excellence Corp. v. Sch. Comm. Of City of Boston, 996 F.3d 37, 45–46 (1st Cir. 2021) (noting that plaintiff “offers no analysis or argument for why” the court should adopt the approach of “comparing the projected admissions under [a contested admissions plan] to prior admissions under the predecessor plan,” and calling plaintiff’s reliance on that framework for analyzing disparate impact a “weakness” that “certainly cut[s] against finding that the degree of disproportionate effect contributes to plaintiff’s likelihood of success on the merits.”). And what’s more, compared to the breadth of caselaw that supports Defendants’ preferred methods, Plaintiffs point only to Montgomery as an example of a case that has adopted this approach. See 560 F. Supp. 3d at 952. Nonetheless, even under their favored framework, Plaintiffs still would not prevail. Compared to 2018, when the reforms to Discovery were not yet in effect, both in 2019 and 2020, the percentage of offers received by Asian American students increased, the offer-rate among Asian Americans increased, and the spread between Asian American students’ share of offers and share of test-takers increased. See Doc. 149 at 8. This data demonstrates that Asian American students have fared better after the reforms than before them. See Boston Parent Coal. For Acad. Excellence Corp., 996 F.3d at 46. For these reasons, the Court finds that the changes to Discovery—irrespective of the motive behind their passage—have not resulted in a discriminatory effect on Asian Americans. A. The Two-Year Period The Court rejects Plaintiffs’ contention that this action should go forward because a two- year period does not provide enough time to observe whether the reforms disproportionately impact Asian Americans.17 Courts often decide equal protection cases after only a single selection cycle. See, e.g., Davis, 426 U.S. at 229. Plaintiffs have not identified any cases where a court has found no adverse impact but nonetheless decided that the case should remain open to see if adverse impact would develop. Analyzing for discriminatory effect is not a wait-and-see exercise. “If there is adverse impact in the future, Plaintiffs may bring a new case” at that time. Doc. 159 at 6. 17 In support of this proposition, Plaintiffs point to the Fourth Circuit’s decision in N.C. State Conference of the NAACP v. McCrory, 831 F.3d 204 (4th Cir. 2016). But that case is not analogous. There, the Fourth Circuit reversed a district court’s ruling that a state election reform law—which imposed voter-identification requirements, reduced the number of early voting days, eliminated-same day registration, and prohibited counting out-of-precinct provisional ballots—did not disproportionately harm Black Americans since voter turnout among Black Americans increased 1.8% despite the new voting restrictions. The Fourth Circuit explained that the district court’s finding “that African Americans disproportionately used each of the removed mechanisms, as well as disproportionately lacked . . . photo ID” was “sufficient [to show] disparate impact [under] Arlington Heights analysis.” Id. at 231. In that context, the appellate court criticized the district court for awarding “almost dispositive weight” to the 1.8% increase in Black voter turnout, noting that “courts should not place much evidentiary weight on any one election.” Id. at 232. Nowhere in McCrory did the court hold that two cycles worth of data is insufficient to discern whether a policy has produced disparate impact. The court simply deemed it improper to rely almost exclusively on a 1.8% uptick in Black voters to rule out disparate impact without first considering whether—irrespective of turnout—the law placed any disproportionate burden on Black Americans. That the data came from only one year and not more is a helpful, but not dispositive, point. Separately, McCrory is an election case—not a selection case. In criticizing the district court’s over-reliance on the 1.8% figure, the appeals court explained that voting patterns differ between midterm elections and general elections; that could explain why there were 1.8% more Black North Carolina voters in the 2016 general election than the 2014 midterm election. Analogous distinctions do not exist here between the Schools’ application seasons. B. Stuyvesant and Bronx Science As a last resort, Defendants attempt to show disproportionate impact on Asian American students by pointing to data compiled by their expert, Dr. Vigdor. See Doc. 154-1. Plaintiffs argue that Dr. Vigdor’s study shows that the reforms have disparately impacted Asian American students for the Fall 2020 entering class at Stuyvesant and Bronx Science. In conducting his analysis, Dr. Vigdor used invitation-to-the-Discovery-Program data rather than the approved- offer-to-the-Discovery-Program data. See supra note 8 at 6. Defendants argue that because “invitations” are given to some students who ultimately would not be eligible for the Program, this evidence is irrelevant. See Doc. 149 at 23. But even setting aside Defendants’ contention, Dr. Vigdor’s data still does not substantiate Plaintiffs’ disparate impact allegation. Dr. Vigdor’s report shows that Asian American students received 66.9% of the invitations to attend Stuyvesant but would have received 67.6% of the invitations if the pre- implementation 2018 criteria were used. As Defendants note, this is a difference of .7% or 9 students. See Doc. 159 at 14. Applying the same framework to Bronx Science shows that Asian American students received 55.8% of the invitations in 2020 but would have received 57.2% if the 2018 criteria were used—a difference of 1.4% or 13 students. Id. There is no showing that these minor differences are significant. Additionally, this Court does not accept that trends in two of the eight Schools can sustain Plaintiffs’ disparate impact claim. See Sharif v. N.Y. State Educ. Dep’t, 709 F. Supp. 345, 362 (S.D.N.Y. 1989) (“[In a case alleging statewide discrimination, such a focus [individual schools and counties] does not rebut plaintiffs’ statewide prima facie case.”). For these reasons, this data does not reasonably evidence disproportionate impact on Asian Americans. ii. Discriminatory Purpose Because the Court concludes that there is no discriminatory effect, and because a plaintiff must evidence discriminatory effect to challenge a facially neutral law enforced in a non- discriminatory matter, see supra section IV(a)(1) at 10-12, the Court need not reach any conclusion with respect to whether Defendants enacted the reforms with discriminatory intent. V. Conclusion For the foregoing reasons, the Defendants’ motion to for summary judgement is GRANTED. The Clerk of the Court is respectfully directed to terminate the motion, Doc. 144, and close the case. It is SO ORDERED. Dated: September 7, 2022 ft } a New York, New York a bt \a_—. Edgardo Ramos, U.S.D.J. 23