Students for Fair Admissions, Inc. v. President and Fellows of Harvard College et al

Filing 421

DECLARATION re 412 MOTION for Summary Judgment by Students for Fair Admissions, Inc.. (Attachments: # 1 Exhibit 1, # 2 Exhibit 2, # 3 Exhibit 3, # 4 Exhibit 4, # 5 Exhibit 5, # 6 Exhibit 6, # 7 Exhibit 7, # 8 Exhibit 8, # 9 Exhibit 9, # 10 Exhibit 10, # 11 Exhibit 11, # 12 Exhibit 12, # 13 Exhibit 13, # 14 Exhibit 14, # 15 Exhibit 15, # 16 Exhibit 16, # 17 Exhibit 17, # 18 Exhibit 18, # 19 Exhibit 19, # 20 Exhibit 20, # 21 Exhibit 21, # 22 Exhibit 22, # 23 Exhibit 23, # 24 Exhibit 24, # 25 Exhibit 25, # 26 Exhibit 26, # 27 Exhibit 27, # 28 Exhibit 28, # 29 Exhibit 29, # 30 Exhibit 30, # 31 Exhibit 31, # 32 Exhibit 32, # 33 Exhibit 33, # 34 Exhibit 34, # 35 Exhibit 35, # 36 Exhibit 36, # 37 Exhibit 37, # 38 Exhibit 38, # 39 Exhibit 39, # 40 Exhibit 40, # 41 Exhibit 41, # 42 Exhibit 42, # 43 Exhibit 43, # 44 Exhibit 44, # 45 Exhibit 45, # 46 Exhibit 46, # 47 Exhibit 47, # 48 Exhibit 48, # 49 Exhibit 49, # 50 Exhibit 50, # 51 Exhibit 51, # 52 Exhibit 52, # 53 Exhibit 53, # 54 Exhibit 54, # 55 Exhibit 55, # 56 Exhibit 56, # 57 Exhibit 57, # 58 Exhibit 58, # 59 Exhibit 59, # 60 Exhibit 60, # 61 Exhibit 61, # 62 Exhibit 62, # 63 Exhibit 63, # 64 Exhibit 64, # 65 Exhibit 65, # 66 Exhibit 66, # 67 Exhibit 67, # 68 Exhibit 68, # 69 Exhibit 69, # 70 Exhibit 70, # 71 Exhibit 71, # 72 Exhibit 72, # 73 Exhibit 73, # 74 Exhibit 74, # 75 Exhibit 75, # 76 Exhibit 76, # 77 Exhibit 77, # 78 Exhibit 78, # 79 Exhibit 79, # 80 Exhibit 80, # 81 Exhibit 81, # 82 Exhibit 82, # 83 Exhibit 83, # 84 Exhibit 84, # 85 Exhibit 85, # 86 Exhibit 86, # 87 Exhibit 87, # 88 Exhibit 88, # 89 Exhibit 89, # 90 Exhibit 90, # 91 Exhibit 91, # 92 Exhibit 92, # 93 Exhibit 93, # 94 Exhibit 94, # 95 Exhibit 95, # 96 Exhibit 96, # 97 Exhibit 97, # 98 Exhibit 98, # 99 Exhibit 99, # 100 Exhibit 100, # 101 Exhibit 101, # 102 Exhibit 102, # 103 Exhibit 103, # 104 Exhibit 104, # 105 Exhibit 105, # 106 Exhibit 106, # 107 Exhibit 107, # 108 Exhibit 108, # 109 Exhibit 109, # 110 Exhibit 110, # 111 Exhibit 111, # 112 Exhibit 112, # 113 Exhibit 113, # 114 Exhibit 114, # 115 Exhibit 115, # 116 Exhibit 116, # 117 Exhibit 117, # 118 Exhibit 118, # 119 Exhibit 119, # 120 Exhibit 120, # 121 Exhibit 121, # 122 Exhibit 122, # 123 Exhibit 123, # 124 Exhibit 124, # 125 Exhibit 125, # 126 Exhibit 126, # 127 Exhibit 127, # 128 Exhibit 128, # 129 Exhibit 129, # 130 Exhibit 130, # 131 Exhibit 131, # 132 Exhibit 132, # 133 Exhibit 133, # 134 Exhibit 134, # 135 Exhibit 135, # 136 Exhibit 136, # 137 Exhibit 137, # 138 Exhibit 138, # 139 Exhibit 139, # 140 Exhibit 140, # 141 Exhibit 141, # 142 Exhibit 142, # 143 Exhibit 143, # 144 Exhibit 144, # 145 Exhibit 145, # 146 Exhibit 146, # 147 Exhibit 147, # 148 Exhibit 148, # 149 Exhibit 149, # 150 Exhibit 150, # 151 Exhibit 151, # 152 Exhibit 152, # 153 Exhibit 153, # 154 Exhibit 154, # 155 Exhibit 155, # 156 Exhibit 156, # 157 Exhibit 157, # 158 Exhibit 158, # 159 Exhibit 159, # 160 Exhibit 160, # 161 Exhibit 161, # 162 Exhibit 162, # 163 Exhibit 163, # 164 Exhibit 164, # 165 Exhibit 165, # 166 Exhibit 166, # 167 Exhibit 167, # 168 Exhibit 168, # 169 Exhibit 169, # 170 Exhibit 170, # 171 Exhibit 171, # 172 Exhibit 172, # 173 Exhibit 173, # 174 Exhibit 174, # 175 Exhibit 175, # 176 Exhibit 176, # 177 Exhibit 177, # 178 Exhibit 178, # 179 Exhibit 179, # 180 Exhibit 180, # 181 Exhibit 181, # 182 Exhibit 182, # 183 Exhibit 183, # 184 Exhibit 184, # 185 Exhibit 185, # 186 Exhibit 186, # 187 Exhibit 187, # 188 Exhibit 188, # 189 Exhibit 189, # 190 Exhibit 190, # 191 Exhibit 191, # 192 Exhibit 192, # 193 Exhibit 193, # 194 Exhibit 194, # 195 Exhibit 195, # 196 Exhibit 196, # 197 Exhibit 197, # 198 Exhibit 198, # 199 Exhibit 199, # 200 Exhibit 200, # 201 Exhibit 201, # 202 Exhibit 202, # 203 Exhibit 203, # 204 Exhibit 204, # 205 Exhibit 205, # 206 Exhibit 206, # 207 Exhibit 207, # 208 Exhibit 208, # 209 Exhibit 209, # 210 Exhibit 210, # 211 Exhibit 211, # 212 Exhibit 212, # 213 Exhibit 213, # 214 Exhibit 214, # 215 Exhibit 215, # 216 Exhibit 216, # 217 Exhibit 217, # 218 Exhibit 218, # 219 Exhibit 219, # 220 Exhibit 220, # 221 Exhibit 221, # 222 Exhibit 222, # 223 Exhibit 223, # 224 Exhibit 224, # 225 Exhibit 225, # 226 Exhibit 226, # 227 Exhibit 227, # 228 Exhibit 228, # 229 Exhibit 229, # 230 Exhibit 230, # 231 Exhibit 231, # 232 Exhibit 232, # 233 Exhibit 233, # 234 Exhibit 234, # 235 Exhibit 235, # 236 Exhibit 236, # 237 Exhibit 237, # 238 Exhibit 238, # 239 Exhibit 239, # 240 Exhibit 240, # 241 Exhibit 241, # 242 Exhibit 242, # 243 Exhibit 243, # 244 Exhibit 244, # 245 Exhibit 245, # 246 Exhibit 246, # 247 Exhibit 247, # 248 Exhibit 248, # 249 Exhibit 249, # 250 Exhibit 250, # 251 Exhibit 251, # 252 Exhibit 252, # 253 Exhibit 253, # 254 Exhibit 254, # 255 Exhibit 255, # 256 Exhibit 256, # 257 Exhibit 257, # 258 Exhibit 258, # 259 Exhibit 259, # 260 Exhibit 260, # 261 Exhibit 261)(Consovoy, William) (Additional attachment(s) added on 6/18/2018: # 262 Unredacted version of Declaration, # 263 Exhibit 1 (filed under seal), # 264 Exhibit 2 (filed under seal), # 265 Exhibit 5 (filed under seal), # 266 Exhibit 6 (filed under seal), # 267 Exhibit 7 (filed under seal), # 268 Exhibit 8 (filed under seal), # 269 Exhibit 9 (filed under seal), # 270 Exhibit 10 (filed under seal)) (Montes, Mariliz). (Additional attachment(s) added on 6/18/2018: # 271 Exhibit 11 (filed under seal), # 272 Exhibit 12(filed under seal), # 273 Exhibit 13 (filed under seal), # 274 Exhibit 14 (filed under seal), # 275 Exhibit 16 (filed under seal), # 276 Exhibit 17(filed under seal), # 277 Exhibit 18(filed under seal), # 278 Exhibit 19 (filed under seal), # 279 Exhibit 20 (filed under seal), # 280 Exhibit 22 (filed under seal), # 281 Exhibit 23 (filed under seal), # 282 Exhibit 24 (filed under seal), # 283 Exhibit 25(filed under seal), # 284 Exhibit 26 (filed under seal), # 285 Exhibit 28 (filed under seal), # 286 Exhibit 29 (filed under seal), # 287 Exhibit 31 (filed under seal), # 288 Exhibit 32 (filed under seal), # 289 Exhibit 33 (filed under seal), # 290 Exhibit 35 (filed under seal), # 291 Exhibit 36 (filed under seal), # 292 Exhibit 37 (filed under seal), # 293 Exhibit 38(filed under seal), # 294 Exhibit 39 (filed under seal), # 295 Exhibit 40 (filed under seal), # 296 Exhibit 41, # 297 Exhibit 42 (filed under seal), # 298 Exhibit 43 (filed under seal), # 299 Exhibit 44(filed under seal), # 300 Exhibit 45 (filed under seal), # 301 Exhibit 46 (filed under seal), # 302 Exhibit 47 (filed under seal), # 303 Exhibit 48 (filed under seal), # 304 Exhibit 51 (filed under seal)) (Montes, Mariliz).

EXHIBIT 200

Title: Quantifying the Asian Penalty in College Admissions
By:
Class: Statistics

SHS 000005

Abstract
In today's educational climate, it is almost taken for granted that race matters in college
admissions. Institutionally, African Americans, Hispanics, and other minority groups are favored
by affirmative action policies. Such policies are justified on the basis of promoting diversity,
which is considered to be a "compelling state interest)." Students, academics, and teachers,
however, suspect that colleges practice overt discrimination against Asian students —
systemically disadvantaging them over peers of all races. Though widely accepted to be true,
such claims have seldom been empirically tested and never been quantified from the perspective
of an applicant. Meanwhile, colleges continue to deny the existence of an "Asian penalty." This
paper addresses this dearth of data. Using multivariate regression, we provide compelling
evidence that the Asian penalty not only exists, but is highly significant. Holding SAT scores,
GPA, gender, socio-economic status, and rigor of curriculum constant, we find that, on average,
an Asian student will be admitted to a school that is 8.4 rankings lower than the school to which
s/he would have been admitted had s/he not been Asian.

Introduction:
Despite being frequently covered in mainstream periodicals, the empirical literature on Asian
discrimination during the college admissions process is sparse. The most well-cited study (and
indeed, the only major one) on the topic was conducted by Thomas J. Espenshade, a Professor of
Sociology at Princeton University. By analyzing data from the National Study of College
Experience (NSCE) survey, which included hundreds of thousands of students who applied to 10
selective colleges between the early 1980s and late 1990s, he found that, on average, Asian

1 Regents of the University

of California vs. Bakke. Supreme Court. 26 June 1978. Print.

SHS 000006

students accepted to these universities scored 140 points higher on their SATs than did their
white peers and 450 points higher than African American students.2 Espenshade's mean-driven
analysis was enough to raise red flags regarding Asian discrimination, because it proved that
"something was up," but did not causally link self-reporting an Asian ethnicity to college
admission decisions. Espenshade himself cautions against drawing the conclusion of overt
discrimination, given the fact that his study does not control for outside variables, including
gender, socio-economic status, GPA, extra-curricular participation, rigor of high school
curriculum, among other factors. A study conducted by the Center for Equal Opportunity at the
University of Wisconsin, Madison3 found a similar, but less extreme result: Asians at the
university had median math and reading SAT scores of 1370 out of 1600, compared with 1340
for white students, 1250 for students of Hispanic descent and 1190 for black students. A number
of authors have suggested that Asian American students may have higher mean SAT scores
because, as a group, they perform more weakly when it comes to other admission criteria.4

Critics of the Asian penalty are also likely to cite a 17% "Asian quota." Ron Unz for the New
York Times explains that the percentage of Asians among the student bodies of Ivy League
schools has remained at 17 percent for the past 20 years, despite "a huge rise in the number of
Asians winning top academic awards in our high schools or being named National Merit

2 Espenshade, Thomas J., Alexandria Walton Radford, and Chang Young. Chung. No Longer
Separate, Not Yet Equal: Race and Class in Elite College Admission and Campus Life.
Princeton, NJ: Princeton UP, 2009. Print.
3 Slotnik, Daniel. "Do Asian-Americans Face Bias in Admissions at Elite Colleges?" New York
Times, 8 Feb. 2012. Web. 16 June 2014.
4 Ibid

SH S 000007

Scholarship semifinalists.s" While this statistical trend is suspicious, it does not conclusively
demonstrate that an overt anti-Asian bias exists. Indeed, Ivy League schools vehemently deny
the existence of such quotas. Jeff Neal, director of university communications for Harvard
recently asserted that, "the admissions committee does not use quotas of any kind."

This study builds on the available literature in three major ways. First, we will control for
exogenous variables through multivariate regression, allowing us to draw causal and decisive
conclusions. Second, we use college rank as our dependent variable in order to effectively
measure the height of the "bamboo ceiling.6" Finally, our study uses data from 2007, which is
significantly more current than either Espenshade or the Center for Equal Opportunity, both of
whom rely on decades old data. As a result, our study has stronger statistical robustness than
does the current literature and presents novel findings.

Methods
Stuyvesant High School is an elite New York City high school whose students compete for spots
at America's top universities. Its student body is comprised of almost exclusively of Caucasian
and Asian students, with Hispanics and African Americans representing less than 5% of the
school's population. Comparing college admissions results between Asian students and their
non-Asian peers allows us to quantify the effect of the purported Asian penalty. We used the
official college admissions data for the Class of 2007 to model the level of Asian discrimination
at Stuyvesant. This cluster sample is representative of the greater Stuyvesant population (it was
5 Utz, Run. " Statistics Indicate an Ivy League Asian Quota." New York Times, 3 Dec. 2013.
Web. 16 June 2014.
6 Chen, Carolyn. "Affirmative Action and Breaking the 'Bamboo Ceiling' for Asian
Americans." Los Angeles Times. 13 June 2013. Web. 16 June 2014.

SH S 000008

selected at random from a dataset that included all students who have graduated since 2003).
Stuyvesant High School is representative of a highly competitive ivy-league feeder high school.
Its high-proportion Asian/White population makes it especially effective at modeling the
differences in college admission between Asians and Caucasians.

This study relied on a multivariate regression model with the following independent variables:
1. GPA: Grade point average, measured on a 100-point scale. GPA is often considered the
single-most important factor in college admissions
2. SAT Score: Total score on the SAT out of 2400. Another important metric used by
colleges in their holistic admissions process.
3. Meal Code.• Stuyvesant's students are all classified as receiving free, reduced or full price
lunches, as a function of their family's socioeconomic status. These meal codes were
used to model the effect of income on college admissions, which, although not statedly an
important factor in college decisions, is widely considered to have an impact. Socioeconomic status also generally predicts students' legacy status. (0 = free lunch, 1
reduced lunch, 2 = full-paid lunch).
4. # of AP Classes: Level of academic rigor throughout one's high school career is heavily
weighted by admissions officers when making their decisions.
5. Ethnicity (binary): The independent variable of interest in this study. Represents the
unique effect of being Asian on the dependent variable (0 = Asian, 1 = Non-Asian)
6. Gender (binary): It has been proposed that female students are held to a higher standard
than their male peers, given that young women already make up the majority of the

SHS 000009

population at many elite universities. To control for this possible effect, we have included
gender as an independent variable (0 = Female, 1 = Male).

A significant variable not accounted for in this multivariate regression model is level of
participation in extracurricular activities. However, a simple mean-driven analysis of a 2014
survey conducted by the Stuyvesant Spectator, Stuyvesant's official school newspaper, showed
no systemic difference in extracurricular participation
between Asian and non-Asian students, indicating that
this is not a confounding variable. (Omitted variable
biases can only effect multivariate regression when the
omitted variable shares variance with another
independent variable being analyzed.) Quality of
college essays may be a lurking variable and was not
inches in this analysis.

Figure 1: Omission of extracurricular participation
will not confound the multivariate regression
because there is no systemic relationship between
ethnicity and extracurricular participation.

The dependent variable used in this study was the highest ranked college to which a student was
accepted. We chose to use the rank of the highest ranked college to which a student was
accepted, as opposed to using the school the student chose to attend in order to control for a
student selection bias. We did not run the regression on every school to which the student was
accepted, because this would have over-weighted students who chose to apply to a greater
number of safety schools (or just a greater number of schools, period). Rankings were
determined using the Forbes America's Top Colleges ranking scale!

7

Forbes. "America's Top Colleges" Forbes Magazine, July 2013. Web. 16 June 2014.

Si-IS 000010

a: .05
Null hypothesis: Oethnic => 0.
Alternate hypothesis: ficd„,i, <0
Model: Rank = + )61 * GPA + )62 * gender + fi.thme * ethnic + 134 * SAT Total + 135* mealcode
4116 * AP Classes

See Appendix 1 for all of the data.

Results
Model: Rank = 941.1495 - 8.1152 * GPA + 0.2958 * gender - 8.4068 * ethnic - 0.0556 *
SAT Total - 0.2044 * mealcode - 1.5678 * AP Classes

The multivariate regression yielded a correlation coefficient of -8.40676, for the ethnicity
variable, with a p-value of 2.68611E-2. This is a statistically significant finding, with p less than
alpha. All else being equal, non-Asian students will be admitted to schools that are ranked 8,4
rankings better (lower, meaning closer to 1) than they would have had they been Asian (see
Appendix 2 for full breakdown of the multivariate regression results).

Though not relevant in the context of our hypothesis, we should note that the correlation
coefficients for GPA, SAT Score, and # APs taken all came out as significant (which was
expected), with p-values of 0.E+0, 1.39712E-3 and 1.6641E-2, respectively. Mealcode and
gender were not significant with p-values of 4.70845E-1 and 4.71586E-1, respectively. These

SHS 000011

findings would indicate the colleges are indeed gender/need blind when making their admissions
decisions, and that academic performance in heavily weighted in the college decision process.

Discussion
Our finding is highly statistically significant and indicates that colleges systemically discriminate
against Asian students, with the Asian penalty at Stuyvesant High School being equivalent to S
rankings on the Forbes America's Top Colleges ranking scale. Discrimination against Asians is
not minor, nor is it indirect. The race-conscious process by which students are selected by elite
universities is nothing less than racism.

SHS 000012