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

Filing 415

DECLARATION re 412 MOTION for Summary Judgment by Students for Fair Admissions, Inc.. (Attachments: # 1 Exhibit Expert Report, # 2 Exhibit Rebuttal Expert Report, # 3 Errata Errata)(Consovoy, William) (Additional attachment(s) added on 6/18/2018: # 4 Unredacted version of Declaration of P. Arcidiacono, # 5 Exhibit A- unredacted version, # 6 Exhibit B-unredacted version) (Montes, Mariliz).

Download PDF
EXHIBIT B REBUTTAL EXPERT REPORT OF PETER S. ARCIDIACONO Students for Fair Admissions, Inc. v. Harvard No. 14-cv-14176-ADB (D. Mass) TABLE OF CONTENTS 1 Executive Summary ....................................................................................................1 2 Professor Card and I agree on many aspects of my methodology, analysis, and conclusions.................................................................................................................11 2.1 Professor Card and I largely agree on the relevant dataset..........................11 2.2 Professor Card does not challenge my descriptive analysis about the relative qualifications of students by race/ethnicity. .....................................................12 2.3 Professor Card generally agrees with my methodological approach to modeling Harvard’s admissions decisions ........................................................16 3 There are several key flaws in Professor Card’s modeling choices that drive his conclusions about the size of the Asian-American penalty. ....................................17 3.1 Professor Card’s results are skewed by his decision to include in the analysis many applicants who are unaffected by racial penalties and preferences......17 3.1.1 Professor Card misleadingly includes non-competitive applicants in his models, which tends to obscure the racial penalties and preferences Harvard employs in its admissions process.............................................17 3.1.2 Professor Card errs by including in all of his models those applicants who are members of Harvard’s special recruiting categories .................19 3.2 Professor Card errs in failing to include interaction terms. ..........................19 3.3 Professor Card’s models ignore the fact that Asian-American applicants face a penalty in the personal rating........................................................................22 3.3.1 According to standard statistical practice, my model is considered an excellent fit of the personal rating ...........................................................23 3.3.2 Because there is compelling evidence that racial preferences and ii penalties affect the personal rating, this rating should not be included in the analysis. ..........................................................................................25 3.4 Professor Card’s argument that Asian-American applicants are worse on nonacademic measures is misleading. ..............................................................27 3.5 Parental occupation varies in highly unusual and unexplained ways over time, undermining its reliability as a variable and its usefulness as a control.................................................................................................................31 4 Professor Card’s models, once corrected of their key flaws, show that Harvard imposes a penalty against Asian-American applicants. ..........................................33 4.1 Professor Card’s preferred yearly model is less sound than a pooled approach. ............................................................................................................34 4.2 Professor Card’s pooled results, with small corrections, show that Harvard imposes a penalty against Asian-American applicants. ..................................36 4.3 Professor Card’s yearly models, with small corrections, confirm that Harvard imposes a penalty against Asian-American applicants. ..................................39 4.4 Professor Card’s analysis of applicants whose race is missing further confirms the existence of an Asian-American penalty ....................................................44 5 Professor Card’s Analysis Actually Demonstrates That Race Is a Determinative Factor in Harvard’s Admissions Decisions. .............................................................45 5.1 A model with race as the only control would be expected to perform poorly relative to other factors. ....................................................................................48 5.2 Professor Card’s argument that racial preferences are not relevant for most African-American and Hispanic applicants misleadingly focuses on uncompetitive applicants. .................................................................................49 5.3 Professor Card’s method of calculating the importance of unobserved factors is incorrect and substantially overstates their importance. ............................51 iii 6 Professor Card Fails to Refute the Overwhelming Statistical Evidence of a Floor for African-American Admissions. ...........................................................................54 6.1 Professor Card’s speculation that Harvard would not want to use a floor based on a non-public admissions rate misses the point. ................................56 6.2 Contrary to Professor Card’s arguments, there is additional evidence that Harvard began implementing the floor in 2017. ..............................................57 6.3 Professor Card’s analysis of other data does nothing to undermine my claim that Harvard maintained a floor on the admission rate for single-race African-American applicants. ...........................................................................58 6.4 Differences in the characteristics of admitted single-race African Americans after 2016 further support evidence of a floor. .................................................59 7 A Number of the Other Variables Added by Professor Card Are of Questionable Reliability and Undermine the Confidence of His Conclusions. .............................61 7.1 Intended career varies in highly unusual and unexplained ways over time, undermining its reliability as a variable and its usefulness as a control. ......62 7.2 Professor Card’s approach to using the rating variables suffers from a smallpopulation problem and masks racial preferences, which undermine its reliability............................................................................................................63 7.3 Staff interviews are selectively given and thus should not be used as a control.................................................................................................................66 8 Incorporating Most of Professor Card’s Variables Into My Preferred Model Confirms My Findings Regarding the Effect of Harvard’s Racial Penalties and Preferences. ...............................................................................................................67 8.1 Changes advocated by Professor Card that I incorporate in my updated model ..................................................................................................................68 8.2 The results of the updated preferred model confirm my previous findings and iv conclusions .........................................................................................................70 8.3 Even incorporating many of Professor Card’s manifestly unsound modeling choices does not alter the result of my model. ..................................................71 9 My Updated Preferred Model Yields Additional Reasons to Doubt Professor Card’s Approach....................................................................................................................73 9.1 The penalties Asian-American applicants face are substantial ....................74 9.2 Estimates of my admissions and personal ratings models show that AsianAmerican applicants are strong on non-academic measures. ..........................75 9.3 Dockets with high shares of Asian-American applicants are penalized .......77 v 1 Executive Summary In my opening report, I explained my professional and academic background in econometrics and my prior scholarly work about the use of race/ethnicity in the admissions processes of colleges and universities. My report explained how, using my experience and expertise, I reviewed and analyzed six-years of admissions data obtained from Harvard College and built a model to test the effect that race/ethnicity has in the admissions process. The model included and controlled for more than 200 variables from Harvard’s admissions data, and was constructed using standard techniques used in my field for statistical modeling. It produced a number of reliable conclusions about the way an applicant’s race/ethnicity affects his or her admissions prospects at Harvard. Most importantly, I found: • Asian-American applicants as a whole are stronger on many objective measures than any other racial/ethnic group, including test scores, academic achievement, and extracurricular activities. • Harvard penalizes Asian-American applicants (relative to white applicants) in the scoring of applicants for admission, particularly in the personal and overall ratings assigned by Harvard’s admissions officers. • Harvard also penalizes Asian-American applicants (again, relative to white applicants) in the selection of applicants for admission. • Race/ethnicity plays a significant role in admissions decisions. In addition to the racial penalties that Harvard imposes on Asian-American applicants, Harvard affords substantial racial preferences to Hispanic and AfricanAmerican applicants. The combined effect of the penalties and preferences is of such great magnitude that, for example, a male non-disadvantaged AsianAmerican student with characteristics that would suggest a 25% probability of admission would see those chances rise to 95% if he were treated as an African American. • Since the admissions cycle for the class of 2017, the admit rate for those applicants who identify as African American using the federal IPEDS (Integrated Postsecondary Education Data System) methodology, i.e., single- 1 race African Americans, is almost identical to the admit rate of all other domestic applicants. The probability of this occurring without direct manipulation is less than 0.2%. • Many of my conclusions are consistent with analyses performed by Harvard’s own Office of Institutional Research (OIR), including my conclusions about (1) the relative strength of Asian-American applicants, (2) Harvard’s discrimination against Asian Americans in the personal ratings, (3) the penalty Harvard imposes on Asian-American applicants, and (4) the disproportionate role race plays with respect to Hispanic and AfricanAmerican applicants. In his report on behalf of Harvard, Professor David Card generally agrees that the logit model I used is an appropriate way to analyze the effect of race/ethnicity in Harvard’s admissions process. He does not dispute the objective indicia of quality regarding the strong qualifications of Asian-American applicants, particularly in academic achievements. Nor does he dispute that my findings are consistent with those of OIR. Professor Card, however, makes a number of modeling choices that lead him to reach different conclusions than mine: in particular, he contends that the evidence of a penalty against Asian-American applicants is not compelling. Professor Card argues that the effect of race on admissions is smaller than I report—although he still concedes that the use of race substantially increases the admissions prospects of Hispanic and African-American applicants. Notably, he never challenges the overwhelming statistical evidence that Harvard has imposed a minimum floor for the admission of African-American applicants. Instead, Professor Card simply speculates that there is no reason for Harvard to do so. None of Professor Card’s arguments are persuasive. His modeling choices are inconsistent with standard econometric practices and appear designed to understate the effect of race in the admissions process generally, and on Asian-American applicants specifically. Moreover, his modeling is not robust—with small adjustments to his models to correct his methodological flaws, his models actually 2 confirm my findings and bolster the conclusion that Harvard imposes penalties on Asian-American applicants. Among the key flaws in Professor Card’s approach: Professor Card’s models are distorted by his inclusion of applicants for whom there is no reason to believe race plays any role. As my opening report noted, there are several categories of applicants to whom Harvard extends preferences for reasons other than race: recruited athletes, children of faculty and staff, those who are on the Dean’s List or Director’s List ( ), legacies, and those who apply for early admission.1 Because of the significant advantage that each of these categories confers on applicants, my report analyzed the effect of race on an applicant pool without these special categories of applicants (the baseline dataset), which allowed me to test for the effect of race on the bulk of the applicant pool that did not fall into one of these categories.2 Professor Card, however, includes all of these applicants in his model, taking the remarkable position that there is no penalty against Asian-American applicants unless Harvard imposes a penalty on every Asian-American applicant. But this is an untenable position. I do not assert that Harvard uses race to penalize AsianAmerican applicants who are recruited athletes, children of donors (or others identified on the Dean’s List), legacies, or other preferred categories. By including these special recruiting categories in his models, Professor Card obscures the extent to which race is affecting admissions decisions for all other applicants. Giving preferences for early action is consistent with the yield rate being higher for early action applicants. However, unlike the other special recruiting categories, the penalty against Asian-American applicants who apply early action is similar to the penalty for those who apply regular action. 1 I also analyzed a dataset that included the special categories of applicants (the expanded dataset). I included in this dataset interactions for some of the special categories and race, allowing for the possibility that racial preferences may operate differently for these special recruiting categories. 2 3 Professor Card further exacerbates this problem by including in his calculations the large majority of applicants whose characteristics guarantee rejection regardless of their race. Harvard admits a tiny fraction of applicants – only five or six percent in recent years. This means that a huge proportion of applicants have no realistic chance of admission. If an applicant has no chance of admission, regardless of his race, then Harvard obviously does not “discriminate” based on race in rejecting that applicant. Professor Card uses this obvious fact to assert that Harvard does not consider race at all in most of its admissions decisions. Further, he constructs his models in ways that give great weight to these applicants, again watering down the effect of race in Harvard’s decisions where it clearly does matter. (To put it in simple terms, it is akin to reducing the value of a fraction by substantially increasing the size of its denominator.) Professor Card removes interaction terms, which has the effect of understating the penalty Harvard imposes on Asian-American applicants. As Professor Card notes, his model differs from mine in that he removes the interaction terms. An interaction term allows the effects of a particular factor to vary with another distinct factor. In the context of racial discrimination, interaction terms are especially helpful (and often necessary) in revealing where certain factors operate differently for subgroups within a particular racial or ethnic group. For example, if a law firm singled out African-American women for discriminatory treatment but treated African-American males and other women fairly, a regression model would probably not pick up the discrimination unless it included an interaction between African-American and female. Professor Card rightly recognizes that interaction terms should be included in a model when there is evidence that racial preferences operate differently for particular groups of applicants; yet he nonetheless removes interaction terms for variables that satisfy this condition. The most egregious instance of this is Professor Card’s decision not to interact race with disadvantaged status—even though the data clearly indicate that Harvard treats disadvantaged students differently by race. 4 Harvard gives a preference to disadvantaged applicants. But as I demonstrated in my opening report, the preference Harvard gives African-American and Hispanic applicants for disadvantaged status is much smaller than that given to AsianAmerican and white applicants (Hispanic applicants receive a modest preference for disadvantaged status, and African-American applicants receive no preference for disadvantaged status in the selection of applicants for admission). Arcidiacono Report 8, 64. The interaction term for race and disadvantage allows one to capture those distinctions. Without it, the size of the preference Harvard gives to disadvantaged Asian-American and white applicants is muted by the inclusion of African-American and Hispanic applicants. Since Asian-American applicants are more likely to be disadvantaged than white applicants, the practical implication of this is an understatement of the Asian-American penalty. Professor Card includes the personal rating in many of his analyses, despite clear evidence that this rating is affected by racial preferences. Professor Card includes Harvard’s personal rating in his models—notwithstanding the clear finding yielded by my analysis (and that of OIR) that this rating shows strong evidence of racial bias. Professor Card contends that my model showing racial bias in the personal rating is a poor statistical fit, but that is demonstrably wrong. According to academic works discussing this measure of fit, my model achieves an “excellent” fit. And Professor Card ignores other indicators of racial penalties and preferences in the personal rating (such as the substantial preferences given to African-American and Hispanic applicants), instead assuming—against the evidence and the uniform testimony of Harvard’s admissions officers—that Asian-American applicants as a group are weaker on unobserved personal qualities. Professor Card commits other analytical errors that raise doubts about the reliability of his results. • Professor Card claims that Asian-American applicants are weaker on nonacademic measures. In an attempt to support this claim, however, he distorts the data in two ways. First, he includes legacies, recruited athletes, children of faculty and staff, and those on the Dean’s/Director’s List in his analysis, essentially crediting these applicants as having non-academic achievements. 5 But their higher admission rates are because they are members of these specially recruited groups, not because of their non-academic qualifications. Of course, the reason Professor Card includes them in his analysis is because Asian-American applicants are underrepresented in these categories, as compared to the overall applicant pool. This distorts the analysis in a way that allows Professor Card to make the non-academic qualifications of AsianAmerican applicants appear lower than they actually are. His inclusion of the personal rating further distorts his results; the racial bias in this rating artificially holds down the non-academic qualifications of Asian-American applicants and, at the same time, artificially boosts the non-academic qualifications of African-American and Hispanic applicants. Removing the personal rating from Professor Card’s model shows (as does my model) that Asian-American applicants are at least as strong as white applicants on nonacademic measures, and much stronger on academic measures. • Professor Card’s results are heavily influenced by his inclusion of “parental occupation” (i.e., the occupations of an applicant’s parents) as a control variable. First, the data produced by Harvard for this field oscillates wildly from year-to-year, rendering the data unreliable and any results using it suspect. Second, there is no evidence in the records that Harvard’s admissions office considers parental occupation important aside from its value as a measure of SES, which I already control for in my models. • Professor Card also uses intended career as a control, even though this variable suffers from the same kind of inaccuracies as parental occupation. • Professor Card also includes the staff interview rating variable. But staff interviews are offered only to a very small portion of the pool (2.2% of Professor Card’s dataset); they disproportionately include applicants who fall within the special recruiting categories (recruited athlete, legacy, etc.); and those who receive an interview are admitted at a very high rate (roughly 50%). Moreover, the probability of getting a staff interview is much lower for Asian-American applicants than others, in part because these interviews are disproportionately given to recruited athletes and legacies. Because staff 6 interviews appear to be given on the basis of these other preferences, it is inappropriate as a control variable. • In his yearly analysis, Professor Card also adds controls for reported extracurricular activities in a way seemingly designed to distort the extracurricular variable and thereby disfavor Asian-American applicants. He combines 29 reported categories of extracurricular activities into 12, in a somewhat arbitrary fashion. He then adds a control for the number of hours an applicant spends on “work” (i.e., a job). This choice, which ignores the bulk of the data provided by applicants on the hours they spend on non-academic activities, seems calculated to disfavor Asian-American applicants. Although “work” is only the eighth-most popular non-academic activity listed by white applicants, it is one of the few activities for which they report higher average hours than Asian-American applicants. Making small corrections to Professor Card’s own models results in the finding of a penalty against Asian-American applicants. Professor Card’s models show significant penalties against Asian-American applicants once corrective adjustments are made to remedy his various errors. As stated above, there is no evidence of a penalty against Asian-American applicants who are in one of the special recruiting categories so I remove applicants in those categories from Professor Card’s model. For Professor Card’s pooled analysis, making this one correction plus implementing any one of the following changes results in a statistically significant penalty against Asian-American applicants: (i) recognizing the fact that preferences for disadvantaged status vary with race and therefore interacting race with disadvantaged status; (ii) recognizing that the personal quality measure includes racial preference and therefore should not be included in the model; or (iii) recognizing that the parental occupation variables are unreliable and removing them from the analysis. Professor Card’s yearly models also show significant Asian-American penalties when small corrections are made. Once special recruiting categories are removed from his models, either removing the personal rating or the parental occupation variables yields evidence of an Asian-American penalty. When the extracurricular 7 measures are also corrected, then interacting race with disadvantaged status is enough to reveal a statistically significant penalty.3 Even Professor Card’s flawed models suggest substantial racial preferences for Hispanics and African-Americans—preferences that increase once corrective adjustments are made. Even Professor Card’s analysis—with all of its flaws—confirms my opening report’s finding that race plays a “significant role in admissions decisions at Harvard.” Arcidiacono Report 7-8. Indeed, without making any adjustments to his approach, his models show that racial preferences are responsible for tripling the number of African-American admits and doubling the number of Hispanic admits. Professor Card attempts to explain away these effects, but these efforts can be easily shown to be both incorrect and very misleading; indeed, his arguments often prove the exact opposite of his conclusions. Adopting many of Professor Card’s variables into my models further confirms my initial findings. As I have explained, making small corrective adjustments to Professor Card’s methodology yields results that actually confirm my findings and bolster the conclusion that Harvard applies racial penalties against Asian-American applicants and affords large racial preferences to Hispanic and African-American applicants. On top of supporting my case, this proves the fragility of Professor Card’s models. For the reasons I’ve described, Professor Card’s approach is flawed among many dimensions, and appears designed, in many ways, to conceal the effect of Harvard’s admissions process on Asian-American applicants. My models, on the other hand, are robust. Indeed, adding many of the new variables suggested by Professor Card does not materially change my results. My updated models find that the size of the penalty on Asian-American applicants, and the size When I refer to statistical significance, I am referring to whether we can be 95% certain that the measured effect is different from zero. Even without the corrections to the extracurricular activities, the estimated penalty is statistically significant at the 90% level in Professor Card’s yearly models when special recruiting categories are removed and race is interacted with disadvantaged status. 3 8 of the preferences for African-American and Hispanic applicants, are just as high if not higher. Professor Card offers no analysis to contradict my finding that Harvard has imposed a floor for admissions of those identifying as African-American via IPEDS. In my opening report, I showed that Harvard maintained a floor on the admission rate for single-race African Americans (as identified by IPEDS) in the classes of 2017, 2018, and 2019. In each of these years, the admit rate of single-race African Americans was virtually identical to the admit rate of all other domestic applicants. The chance of this match occurring in three consecutive years (without direct manipulation) is less than two-tenths of one percent—making it a near certainty that Harvard was purposely setting a floor on the admission rate of those applicants. Professor Card does not challenge that finding. Instead, he speculates that Harvard had no reason to use a non-public admission rate as a floor, no reason to institute the floor beginning with the class of 2017, and that Harvard has not set a floor under other metrics. None of these responses is persuasive. Why Harvard chose to set a floor and why it did so in 2017 are not at all relevant to my analysis. But even if they were, there are a number of reasons in the record that would explain why Harvard would want to use the IPEDS metric as a floor, and why it did so beginning in 2017. Indeed, numerous pieces of evidence confirm that Harvard was very concerned about criticisms tied to its IPEDS data at the precise time the first evidence of the floor appears in the data. And the fact that Harvard chose to implement this floor, and not a floor based on another metric, does not change anything. What is certain—and undisputed—is that Harvard was purposely taking steps to ensure that the admission rate of single-race African-American applicants approximated or exceeded the overall admission rate of all other domestic applicants. * * 9 * Professor Card’s report changes none of my conclusions; to the contrary, given how easy it is to alter the results of his models and that my own models report the same results even incorporating a number of his controls, my opinions in this case have only been strengthened: Harvard penalizes Asian-American applicants; Harvard imposes heavy racial preferences in favor of Hispanic and African-American applicants; and Harvard has been manipulating its admission of single-race African-American applicants to ensure their admission rate approximates or exceeds the overall admission rate. Professor Card has demonstrated that it is possible to mask the true effects of race in Harvard’s admission process by changing the scope of the analysis in incorrect ways and choosing inappropriate combinations of control variables. But Professor Card cannot reach these results by applying accepted statistical methods and treating the data fairly. 10 2 Professor Card and I agree on many aspects of my methodology, analysis, and conclusions Although the bulk of this report will respond to and rebut criticisms of my work that Professor Card sets forth in his report, it is useful to note that there are substantial areas of agreement between the two of us.4 2.1 Professor Card and I largely agree on the relevant dataset As discussed in my opening report, I reached my conclusions using a dataset containing Harvard admissions data for the 2014 through 2019 admission cycles. I then performed two general categories of analysis: (1) descriptive analysis, in which I drew conclusions based on simple calculations from my dataset; and (2) regression analysis, in which I used statistical models to estimate how various factors influence Harvard’s admissions decisions and rating of the applicants. Professor Card’s analysis modifies my dataset to create a dataset he calls “Augmented Arcidiacono Data.” Specifically, he creates his dataset by adding additional control variables to my dataset and then performing what he describes as “technical corrections” and fixing what he describes as “technical errors.” Card Report 47-51. Several of the additional variables that Professor Card adds are problematic in terms of relevance and reliability, as I explain infra, at 3.5, 7. Beyond that, except for one “technical error” with which I agree,5 the rest of Card’s modifications are not “errors” or “corrections,” but merely judgment calls. Because his modifications are so minor, I have accepted the majority of them in the interest of avoiding unnecessary disputes. In formulating my rebuttal report, I have not relied upon any data or material other than the material produced with Professor Card’s report, the material cited in this report, and the data and materials identified in my opening report. 4 When the SAT score is not present but an ACT score is present, I now use the ACT science section in my conversions in the same manner as Professor Card. 5 11 2.2 Professor Card does not challenge my descriptive analysis about the relative qualifications of students by race/ethnicity. My descriptive analysis is contained primarily in Section 3.1.1 through Section 3.5.4 of my opening report. See Arcidiacono Report 24-53. Professor Card does not challenge the accuracy of any of this analysis. This is not surprising, because he has no substantive concerns with my dataset and the descriptive analysis involves straightforward assessments of the relevant data. Among the specific findings from the descriptive analysis that Professor Card does not dispute: 2.2.1 Asian-American applicants are, on average, significantly stronger academically than all other racial groups. In terms of academic performance, Asian-American applicants are significantly stronger than all other racial groups. Asian-American applicants have (1) the highest test scores; (2) the highest high school GPAs; (3) taken more AP exams; and (4) scored higher on those AP exams than any other racial group. Arcidiacono Report 33. Asian-American applicants also are rated higher on Harvard’s metrics for assessing academic performance than all other racial groups. In particular, Asian-American applicants’ academic indexes and academic ratings are higher than all other racial groups.6 For example, in the baseline dataset,7 58.6% of Asian-American applicants have academic ratings of 3+ or higher, compared with 44.7% of whites, 14.7% of Hispanics, and 7.3% of African Americans. Arcidiacono Report 33, 36-37. The “academic index” is a score derived from a formula combining standardized testing and high-school performance. The “academic rating” is a rating assigned by Harvard readers. 6 The “baseline” dataset includes all domestic applicants minus certain applicants whose characteristics were associated with a preference (e.g., legacy, athlete), and the “expanded” dataset include all domestic applicants. For both datasets, I removed a small number of applicants who were missing certain information from their application (e.g., test scores). Arcidiacono Report 2. 7 12 2.2.2 Asian-American applicants are so strong academically that their admission rates would more than double in the baseline dataset if based on academics alone. If a random lottery were conducted conditional on being in the top N academic index deciles, the share of Asian-American admits would rise significantly. For example, randomly drawing from all those in the top nine academic index deciles would increase the share of Asian-American admits from 24.9% to 30.4% in the baseline dataset, a jump of more than 22%. More dramatically, randomly drawing from the top academic index decile (in the baseline dataset) would cause Asian-American admits to more than double—resulting in more than 51% of the admitted class being Asian American. Arcidiacono Report 41-42, 44-45. But even if the number of admits from all other groups besides whites and Asian Americans were held fixed and admits for whites and Asian Americans were randomly drawn from the top decile, the share of the class that was Asian American would still substantially increase, resulting in an Asian-American admitted share of 36.5%, a 47% increase. Arcidiacono Report 45. This occurs because Asian-American applicants dominate white applicants in their respective shares of the top academic decile. 2.2.3 Asian-American applicants are strong in non-academic categories. Asian-American applicants excel in more than academics. They also have higher extracurricular ratings and overall alumni ratings than any other racial group. Asian-American applicants likewise are stronger than African-American and Hispanic applicants on counselor ratings, teacher 1 ratings, teacher 2 ratings, and alumni personal ratings, and have similar or slightly lower ratings than whites in these categories. Arcidiacono Report 37. 2.2.4 Despite their high academic and non-academic ratings, AsianAmerican applicants have lower scores in the subjective personal rating than all other racial groups. Despite their superiority on more objective factors, Asian-American applicants have the lowest scores of the four major racial groups on Harvard’s personal rating—the 13 most subjective of all the ratings.8 These low scores on the personal rating are outliers in several respects. First, they differ significantly from the scores Asian-American applicants receive from other individuals, including the ratings from alumni interviewers, teachers, and counselors. For example, alumni interviewers score Asian-American applicants higher on the personal rating than African-American and Hispanic applicants and only slightly lower than white applicants. Arcidiacono Report 37-38. Second, the low scores Asian-American applicants receive on Harvard’s personal rating do not square with the higher scores Asian-American applicants receive on other ratings. As I have shown, higher academic indexes are associated with higher academic ratings, higher extracurricular scores, and higher personal scores. Yet even though Asian-American applicants have the highest academic and extracurricular scores, they are ranked substantially lower in the personal category than the other groups in the same academic index decile. For example, AsianAmerican applicants receive a 2 or better on the personal score more than 20% of the time only in the top academic index decile. By contrast, white applicants receive a 2 or better on the personal score more than 20% of the time in the top six deciles, Hispanics receive such personal scores more than 20% of the time in the top seven deciles, and African Americans receive such scores more than 20% of the time in the top eight deciles. Arcidiacono Report 48-50 & Table 5.6. See, e.g., Chen Depo. at 72 (“Personal quality is one of the categories admissions readers are asked to assess. It is a subjective determination of a combination of many, many factors.”); Walsh Depo. at 60-61 (The personal rating involves “[w]hether that student would contribute to the class, classroom, roommate group, to the class as a whole, their human qualities…. It is a little hard to talk about in general but sort of add it all up and get a feeling”); McGrath I Depo at 171 (The reading guidelines for rating the personal category are “not terribly helpful” and “readers will construe [it] in different ways”); McGrath II Depo at 360 (The personal rating “includes perhaps likability, also character traits, such as integrity, helpfulness, courage, kindness”). 8 14 2.2.5 Despite their high scores on academic and non-academic ratings, Asian-American applicants are admitted at lower rates than all other racial groups. The Asian-American admit rate was below the total admit rate every year from the Class of 2000 through the Class of 2019. Asian-American applicants had this low admit rate despite the fact that during this 20-year span they had higher test scores than all other racial groups in every year. Indeed, Asian-American applicants as a whole had higher test scores than both African-American and Hispanic admits. Arcidiacono Report 24-27 & Figure 1.2. 2.2.6 Among applicants with the same overall rating, Asian-American applicants are less likely to be admitted than all other racial groups. Among those applicants with the same overall rating, Asian-American applicants are less likely to be admitted than any other racial group. For example, in the baseline dataset, 81.4% of African-American applicants with an overall rating of 2+, 2, or 2- were admitted; 76.0% of Hispanic applicants with this overall rating were admitted; 61.0% of white applicants with this overall rating were admitted; and only 59.4% of Asian-American applicants with this overall rating were admitted. The gap between white and Asian-American applicants is even larger in the expanded dataset. Arcidiacono Report 39. Similarly, higher academic index deciles are associated with higher overall ratings by both Harvard readers and alumni interviewers. Asian-American applicants receive overall ratings similar to whites who are one decile lower in terms of their academic indexes. In the top three deciles, Hispanic applicants are between 2.5 and 4.5 times more likely to receive a 2 or better on the overall rating than AsianAmerican applicants, and African-American applicants are between 4.4 and 9.9 times more likely to receive such a score. Arcidiacono Report 50-52. 2.2.7 Higher academic indexes are associated with higher admit rates and higher reader ratings. Yet regular-decision Asian-American admit rates lag behind all other racial groups. Higher academic index deciles are associated with higher admit rates and AsianAmerican applicants have the highest academic indexes. Yet regular-decision 15 Asian-American admit rates are lower than all other racial groups. Asian-American admit rates in any academic-index decile are roughly equivalent to white admit rates one academic index decile lower, Hispanic admit rates three deciles lower, and African-American admit rates five deciles lower. Arcidiacono Report 42-44. 2.2.8 My results are consistent with Harvard’s own findings, as performed by the Office of Institutional Research (OIR). Using data over ten years, OIR found that Harvard’s admissions officers assigned substantially lower personal ratings to Asian-American applicants versus white applicants, especially when compared to the ratings assigned by teachers, counselors, and alumni interviewers. Arcidiacono Report 38. OIR also found that had the academic index and academic rating been used to evaluate the applicants, Asian Americans would have been 43% of the admitted class. These findings are consistent with my findings. In both my analysis and OIR’s analysis, the number of Asian-American admits would more than double if admissions were based on these two criteria. Arcidiacono Report 45-46. 2.3 Professor Card generally agrees with my methodological approach to modeling Harvard’s admissions decisions In my opening expert report, I used regression analysis, and in particular logit models, to draw various conclusions about Harvard’s admissions process and the way in which admissions decisions are affected by an applicant’s race. My report describes the basic methodology and approach, as well as the supporting statistical equations. See Arcidiacono Report 17-23, Appendix A. Professor Card “agrees with [my] general approach” because “[m]ultivariate regression analysis is a widely accepted and common statistical technique in both academia and litigation.” In particular, Card concludes that a logit model like mine “is appropriate where, as here, the outcome of interest—in this case admission to Harvard—is binary, taking values of either zero (not admitted) or one (admitted).” Card Report 47. Card instead disagrees with my specific “modeling decisions.” Card Report 47. The nature of that disagreement, and why my analysis remains more appropriate and reliable, are described in the rest of this report. 16 3 There are several key flaws in Professor Card’s modeling choices that drive his conclusions about the size of the Asian-American penalty. Professor Card makes several fundamental errors in his approach that bear directly on his claims that (1) there is no statistically significant penalty against AsianAmericans, and (2) race plays a lesser role in Harvard’s admissions decisions than I demonstrated in my opening report. These errors explain the difference in our conclusions. 3.1 Professor Card’s results are skewed by his decision to include in the analysis many applicants who are unaffected by racial penalties and preferences. In my opening report, I employed accepted statistical methods to demonstrate that Harvard applies racial penalties and preferences to various racial/ethnic groups. More particularly, I demonstrated that Harvard applies these penalties and preferences where they matter—within the band of applicants who are competitive for admission. Professor Card’s models operate to conceal these racial penalties and preferences by diminishing their magnitude. One of the principal ways Professor Card’s models do so is by his inclusion of applicants who are not impacted by Harvard’s racial penalties and preferences. The inclusion of such applicants has the practical effect of making these penalties and preferences appear to be of smaller magnitude than they actually are. (To put it in simple terms, it is akin to reducing the value of a fraction by substantially increasing the size of its denominator.) 3.1.1 Professor Card misleadingly includes non-competitive applicants in his models, which tends to obscure the racial penalties and preferences Harvard employs in its admissions process. Harvard is a highly selective school. More than 90% of all domestic applicants were rejected over this period, and a substantial number of them are not at all competitive for admission. Those that are affected by racial preferences are competitive applicants. In my report, I showed that Asian-American applicants who had particular characteristics would see substantially higher probabilities of admission were it not for their race. For example, I showed that a male AsianAmerican applicant who was not disadvantaged with observed characteristics that 17 would dictate a 25% probability of admission would see his probability of admission rise to over 36% if treated as a white applicant and to over 95% if treated as an African-American applicant. Professor Card argues that I am distorting the picture by examining the effects for competitive applicants. Professor Card’s approach, however, seeks to dilute the estimates of preferences by including many applicants whose characteristics are such that rejection is assured. Both Professor Card’s and my models show that there is a set of observed characteristics that guarantee rejection; the models perfectly predict rejection without the use of race. While arguments can be made regarding the scope over the set of applicants where one should test for racial penalties and preference, it should be quite clear that this set should not include those who are sufficiently below the bar that race could not possibly enter into consideration. By including applicants who are perfect rejects in his models, Professor Card is able to artificially hold down the average marginal effect of race with respect to any particular racial group. Professor Card’s insistence on including perfect predictions in his model implies that he believes Harvard’s discrimination against certain racial groups and in favor of others is of no consequence unless Harvard actually discriminates against or in favor of every applicant within the affected racial/ethnic groups. This is an absurd proposition. It is a given that Harvard’s low admittance rate means a large number of applicants will be denied without their race ever becoming a factor. But that does not exonerate Harvard for its use of race among the competitive pool. “We don’t always engage in racial discrimination” is not a defense. A conservative position would be to focus the testing for racial preferences or penalties on all of those applicants who are not immediately ruled out—which would mean removing perfect predictions. Under this approach, there will be many applicants who will be included in the analysis even though their admission chances are miniscule, and for whom any effect of racial preferences and penalties will necessarily be small. Throughout my response to Professor Card’s points, I take this conservative approach, showing the average marginal effects of race for all those who are not perfectly predicted to be rejected. 18 3.1.2 Professor Card errs by including in all of his models those applicants who are members of Harvard’s special recruiting categories Professor Card makes a similar modeling error by always including recruited athletes, children of faculty and staff, applicants who are on the Dean’s List or Director’s List, and legacies in his models. Harvard acknowledges that it affords significant preferences to applicants in these special recruiting categories. By including these special recruiting categories in his models, Professor Card is able to obscure the extent to which race is affecting admissions decisions for those not fortunate enough to belong to one of these groups. The inclusion of applicants in these special categories specifically tends to obscure the penalty Harvard imposes on Asian-American applicants. Professor Card’s inclusion of these applicants reflects his position that there is no penalty against Asian-American applicants unless Harvard imposes a penalty on every AsianAmerican applicant. But I am not claiming, for example, that Harvard penalizes recruited athletes who are Asian-American because of their race. My claim is that the effects of Harvard’s use of race occur outside these special categories. There is no reason for their inclusion in his models (at least without interactions with race) other than to conceal the extent to which Harvard penalizes Asian-American applicants in the admissions process. 3.2 Professor Card errs in failing to include interaction terms. In Section 5.1.1 of his report, Professor Card argues against the inclusion of interactions in my models. In discussing the interaction terms between race and disadvantaged status, Professor Card writes: The typical approach in a model trying to isolate the effect of AsianAmerican ethnicity on admissions outcomes would be to include an interaction between race and disadvantaged status only if the effect of being disadvantaged is different for Asian-American and white applicants (or, equivalently, if the effect of race is different for disadvantaged and non-disadvantaged applicants). Prof. Arcidiacono’s results, however, show that is not the case. 19 Card Report 49. As I discuss in sections 8.1 and 8.3, I believe the various interaction terms I include are all appropriate. But the one that deserves special attention here—because it is key to Professor Card’s finding of no Asian-American penalty—is his removal of the interaction terms between race and disadvantaged status. Here, I show why it is inappropriate to exclude these interaction terms; in later sections, I show how their inclusion undermines the reliability of Professor Card’s findings. First, Professor Card is correct that an interaction between race and disadvantaged status makes sense when disadvantage has a different effect for different races. But his analysis becomes misleading when he suggests that the relevant races are only whites and Asian Americans. Understanding Harvard’s use of race in evaluating domestic applicants involves distinctions drawn across all four major racial groups in the applicant pool: Asian Americans, whites, African Americans, and Hispanics. Indeed, Professor Card does not exclude these groups from his models. One of the major findings in my report is that although Harvard gives African-American applicants a large preference, it does not give disadvantaged African-American students any preference for being disadvantaged. Thus, the effect of being disadvantaged is different across racial lines—precisely the condition that Professor Card acknowledges would warrant inclusion of the race/disadvantage interaction terms. So long as African Americans are used in the estimation of the model, the model requires these interaction terms. Yet Professor Card does not include them in his model. This is a relatively basic point; it is odd that Professor Card misses it. But perhaps the effect of his excluding these interaction terms from his models explains this. By excluding the interaction terms between race and disadvantaged status but keeping African-American applicants in the model, Professor Card significantly weakens the effect of disadvantage as an explanatory term. His regression model is essentially finding that disadvantaged status is a fuzzier phenomenon than it actually is and thus downgrades its role in the admissions process. And, because more AsianAmerican applicants than white applicants are disadvantaged, the weaker effect of disadvantaged status in his model in turn weakens the distinctions between white 20 and Asian-American applicants, thus tending to conceal the magnitude of discrimination against Asian-American applicants. It follows that if one compares only Asian-American and white applicants—and excludes the other races from the analysis entirely—then whether the disadvantage/race interaction is included is less important. As I illustrate in section 4.2, estimating Professor Card’s models using only white and Asian-American applicants yields similar penalties against Asian Americans as a model that includes all races, but interacts race with disadvantaged status. And in both cases, the penalties are substantially larger than when the same model is estimated using all races but the interaction terms are removed. Relatedly, this is why I include interactions of race and disadvantaged status in my models of Harvard’s ratings. If interactions are important for one racial group, then they need to be included any time that racial group is included in the analysis. These interaction terms are helpful in diagnosing the extent to which racial preferences affect Harvard’s ratings. Professor Card concedes that the overall rating (and the ultimate admissions decision itself) are affected by race/ethnicity for African-American and Hispanic applicants. Card Report 51, 81. Both the overall rating and the admissions decision show substantial preferences for AfricanAmerican applicants, and smaller preferences for disadvantaged status. But both of these measures also show that African-American applicants either receive a diminished preference for being disadvantaged (in the ratings) or no preference at all (in the admissions outcome).9 Similarly, including interaction terms between race and disadvantaged status in the model more accurately captures the extent to which Harvard’s personal ratings are affected by racial bias. As I described in my opening report, African-American applicants receive a larger preference through the personal rating and a smaller preference for disadvantaged status than other racial groups. That the same Statistically, this is demonstrated by the fact that the coefficient on the interaction term between African American and disadvantaged is negative, and either of the same magnitude or slightly smaller than the positive coefficient on disadvantage itself. 9 21 pattern occurs for the overall rating—and is not present in the other ratings models—provides additional evidence that racial preferences impact the personal rating. Because racial preferences impact the personal rating, that rating should not be used in the analysis—a point that Professor Card must concede, given his own exclusion of the overall rating because it is impacted by racial preferences. 3.3 Professor Card’s models ignore the fact that Asian-American applicants face a penalty in the personal rating. My opening report described how the personal rating assigned to applicants by Harvard’s admissions officers showed clear evidence of racial preferences. Despite their general strength overall, Asian-American applicants have the lowest share of 1s or 2s (the best ratings) on the personal scores. And while academic qualifications are generally correlated with higher personal ratings, Asian-American applicants received lower personal ratings than white applicants despite having better academic and extracurricular ratings. And similarly situated African-American applicants receive much higher personal ratings than their Asian-American counterparts. African-American applicants in the third-worst decile receive higher personal ratings than Asian-American applicants in the top decile. See Arcidiacono Report 5-6, 53-61, Table 5.6. In his report, Professor Card objects to my model that demonstrates a penalty against Asian-American applicants (compared to whites) and a preference in favor of African-American and Hispanic applicants. Specifically, he claims that my model of the personal rating fits the data poorly. Professor Card’s criticisms are misplaced; the personal rating model I rely upon fits the data quite well and I show it is within the range of what is considered to be an “excellent fit.” And the model’s conclusion of a penalty against Asian-American applicants is unmistakable. Moreover, this conclusion is consistent with similar findings by Harvard’s own Office of Institutional Research (which Professor Card does not address). See HARV00065745. And Professor Card estimates no ratings models of his own to counter my findings and conclusions on this point. Finally, as I show in section 8.3, adding his additional variables has no effect on my findings of racial preferences and penalties in the personal rating. 22 3.3.1 According to standard statistical practice, my model is considered an excellent fit of the personal rating The classic citation for what is considered an “excellent fit” based on the Pseudo Rsquare is McFadden (1979) page 307: Those unfamiliar with the ρ2 index should be forewarned that its values tend to be considerably lower than those of the R2 index and should not be judged by the standards for a ‘good fit’ in ordinary regression analysis. For example, values of 0.2 to 0.4 for ρ2 represent an excellent fit. D. McFadden, “Quantitative Methods for Analysing Travel Behavior: Some Recent Developments,” Chapter 13 in Behavioral Travel Modeling, D.A. Hensher and P.R. Stopher, editors, Croom Helm Ltd., 1979. The ρ2 referred to above later became known as McFadden’s R-Square, or the Pseudo R-square that I use in my analysis. Note that the value Professor Card criticizes as “unreliable”—0.28—is within the range characterizing an “excellent” fit.10 Professor Card further attempts to characterize my model as having a “poor” fit by using the probability in my model to predict applicant’s personal ratings. He assigns each applicant the rating that has the highest probability for that applicant, and then assesses the percent correctly predicted for those who actually received a 1. Applying this “percent correctly predicted” method to my model assigns zero applicants a rating of 1. Professor Card criticizes my model for failing to predict a 1 for any of the 47 applicants out of 150,643 (or 0.03%) who actually received a 1 on the personal rating. This attack is nonsensical. It is absurd for Professor Card to claim that a failure to predict the correct personal rating for 0.03% of applicants is evidence of a “poor” fit. Professor Card makes a similar criticism of my model of the overall rating—where the fit is even better and thus (again) well within the range understood to be an “excellent fit.” Card Report 154; Arcidiacono Report Table B.6.8. 10 23 My models assign higher probabilities to ratings that occur more often in the data. When only 0.03% of applicants receive a 1 on the personal rating, the chances of that rating having the highest probability for any applicant is miniscule.11 Professor Card’s use of the percentage correctly predicted method just naturally assigns zero applicants a personal rating of 1. To further illustrate the absurdity of Card’s standard for considering whether a model is a poor fit, consider my model of the academic rating, which he refers to as “more reliable” because of its higher Pseudo R-square. Card Report 70. There are 674 (out of 150,643) applicants who received a 1 on the academic rating, over 13 times the number of applicants who received a 1 on the personal rating. Yet Professor Card’s method of using my model to assign ratings to individual applicants would result in zero applicants being assigned a 1. Indeed, Professor Card’s focus on the model’s ability to correctly predict individual outcomes is a common error. The classic textbook on discrete choice model is by Professor Kenneth Train of the University of California, Berkeley. In discussing the inappropriateness of using the percent correctly predicted, Professor Train writes: Another goodness-of-fit statistic that is sometimes used, but should actually be avoided, is the ‘percent correctly predicted’…. Suppose an estimated model predicts choice probabilities of .75 and .25 in a two-alternative situation. Those probabilities mean that if 100 people faced the representative utilities that gave these probabilities (or one person faced these representative utilities 100 times), the researcher’s best prediction of how many people would choose each alternative are 75 and 25. However, the ‘percent correctly predicted’ statistic is based on the notion that the best prediction for each person is the alternative with the highest probability. This notion would predict that one alternative would be chosen by all 100 As another example, suppose data were available on the height of males in one of five bins, and the last bin was six feet nine inches and higher, something true for about 0.03% of men. Given virtually any set of observed characteristics outside of height itself, the probability associated with this bin will never be the highest. 11 24 people while the other alternative would never be chosen. The procedure misses the point of probabilities, gives obviously inaccurate market shares, and seems to imply that the researcher has perfect information. Kenneth E. Train, Discrete Choice Methods with Simulation, 69 (2d ed. 2009). This issue is of course compounded when the events are extremely rare. 3.3.2 Because there is compelling evidence that racial preferences and penalties affect the personal rating, this rating should not be included in the analysis. As explained in my original report, there is strong evidence that race affects the personal rating, penalizing applicants who are Asian-American and favoring African Americans and Hispanics. Professor Card does not dispute that Asian Americans receive disproportionately lower personal ratings. But he argues that Asian-American applicants have lower scores on the personal ratings because they have weaker average unobserved characteristics than white applicants.12 Crucially, Professor Card ignores the clear evidence of bias in the personal ratings in favor of African Americans and Hispanics. For example, Table 6.1 in my previous report showed that if African-American applicants were treated as whites their average probability of receiving a 2 or better would fall by 22%, and would fall by 35% if they were treated as Asian-American applicants. And here, I can readily It is worth noting that no one in Harvard’s admissions office has advanced Professor Card’s arguments that Asian-American applicants, as a general matter, have some unobserved qualities that explain lower personal ratings. Indeed, numerous admissions officers—including Dean Fitzsimmons himself, who has worked in the admissions office for more than 30 years and reads files to this day—denied that there was any reason to believe that Asian-American applicants were less qualified on the “personal” metric than any other applicant. See, e.g., Fitzsimmons Depo. at 347-348; Ray Depo. at 22; Yong Depo at 234-235; Hansen Depo. at 110-111. Harvard’s own materials likewise leave it to the subjective judgment of the reader as to how the score should be assigned. See HARV00021322 (instructing readers to assign the personal rating on the following scale: “1. Outstanding. 2. Very strong. 3. Generally positive. 4. Bland or somewhat negative or immature. 5. Questionable personal qualities. 6. Worrisome personal qualities”). If Professor Card has any support for why Asian-American applicants have weaker “personal qualities” than other racial groups, he does not provide it. 12 25 show that the observed characteristics of African-American and Hispanic applicants predict much weaker—not stronger—ratings vis-à-vis Asian-American and white applicants. By Professor Card’s reasoning, this should demonstrate that the personal rating incorporates racial preferences. If racial preferences are operating through the personal rating for these groups, then the personal rating is suspect in the same way that the overall rating is suspect—indeed, Professor Card concedes that the overall rating is suspect, and thus excludes from his own analysis. Because the data make clear that racial preferences do, in fact, affect the personal rating, it is unreasonable for Professor Card to conclude that the estimated negative effect for Asian-American applicants is not the result of racial penalties against AsianAmerican applicants.13 Finally, Professor Card makes a number of misleading arguments about what my ratings models show. First, Professor Card states that my finding of a positive and significant relationship between Asian-American applicants and academic and extracurricular activities, even after adding controls, somehow suggests that Harvard cannot be discriminating against Asian-American applicants on the personal and overall ratings. Card Report 71. But Asian-American applicants are stronger than any other racial group on the observed characteristics associated with higher scores on both these ratings. We would therefore also expect them to be stronger on unobservable characteristics, providing an explanation for why there is a statistically significant effect of being Asian American on both these activities. As explained in my opening report, the case for discrimination is very strong when a group of applicants is strong on the observed characteristics associated with a particular rating, yet faces a penalty. And if a group of applicants is weak on the observed characteristics associated with a particular rating, yet receive a higher than expected rating, it further supports the conclusion that racial preferences And here too the interaction terms make the case that racial preferences are affecting the personal rating. That is, there is a substantial preference for being disadvantaged in the personal rating that is significantly diminished for African Americans, mirroring the pattern seen for both the overall rating and for admissions itself. 13 26 affect this rating. That is what the data show: Asian-American applicants have observed characteristics associated with higher personal ratings yet receive a penalty in their personal ratings, and African-American and Hispanic applicants have observed characteristics associated with lower personal ratings yet receive a preference in their personal ratings. Second, Professor Card argues that as more controls are added, the penalty Asian Americans face on the personal rating is diminished and, therefore, if even more controls were added, the effect may go away. But it is not universally true that adding controls leads to lower estimated penalties for Asian-American applicants. Indeed, in my previous report, adding all the controls basically resulted in the same penalty for Asian-American applicants as in the model with no controls, implying that the order in which the controls are added matters. So the inclusion of over 200 controls as a whole does nothing to reduce the Asian-American penalty. Why would we expect that the next set of controls would lead to different results? 3.4 Professor Card’s argument that Asian-American applicants are worse on nonacademic measures is misleading. In my opening report, I noted that my findings of penalties against Asian-American applicants were particularly striking because these applicants are the strongest on observable measures. In particular, they have the highest academic ratings, their ratings on extracurricular activities were better than white applicants, and they generally received higher ratings on other dimensions with the exception of the athletic rating and personal rating. See supra, Section 2. In response, Professor Card contends that Asian-American applicants are weaker than white applicants on nonacademic measures. Card Report 39, Exhibit 10. To arrive at this conclusion, Professor Card uses my estimated model of admissions to form an admissions index for how strong each applicant is based on observed characteristics. Professor Card then removes from this admissions index the variables associated with academics, forming a “non-academic” index. Professor Card finds that Asian-American applicants are generally worse than white applicants on this metric; more specifically, he finds that Asian-American applicants have the lowest share of the four major racial/ethnic groups in the top 27 decile of non-academic achievements. The problem is that Professor Card’s methodology is seriously flawed in two respects. First, Professor Card errs in using the expanded dataset. That model includes athletes, legacies, the children of faculty and staff, and the applicants on the Dean’s and Director’s Lists. This means that every student who receives a preference in one of these special recruiting categories is given a boost in Professor Card’s measure of his “non-academic” achievements. But this makes no sense. White applicants are not stronger than Asian-American applicants on “nonacademic” characteristics because they are more likely to be legacies and therefore treated preferentially in the admissions process. Since Asian-American applicants are substantially less likely to be in these special recruiting categories, Professor Card’s classification works to their detriment. See Arcidiacono Report, Table B.3.2. As I will show, focusing on those who are not in one of these special categories (which again is where the Asian-American penalty is implemented) paints a markedly different picture. In similar fashion, Professor Card includes the personal rating in his measure of “non-academic achievement.” But as I have shown in Section 3.3.2, the personal rating incorporates preferences for African-American and Hispanic applicants and penalties against Asian-American applicants. 14 Using the personal rating as a marker for non-academic achievement is thus highly misleading. In Table 3.1N, I show how each of these features results in a distorted picture of the strength of Asian-American applicants on non-academic measures. 14 See also Arcidiacono Report 37-38. 48-50. 28 Table 3.1N: Asian-American applicants are strong on non-academic measures besides the personal rating Panels 2-4 remove special recruiting categories. The non-academic ratings in panel 4 are: extracurricular, athletic, counselor, teacher1, teacher2, alumni personal, and alumni overall. The first panel of Table 3.1N shows Professor Card’s results from Exhibit 10. The second panel reflects Professor Card’s Exhibit 10, while also removing those who are in one of the special recruiting categories. The third panel does the same and also removes the personal rating. The fourth panel looks only at the non-academic ratings assigned by Harvard’s admissions officers or alumni interviewers. 29 The third panel shows that Asian-American applicants are just as strong as white applicants on non-academic measures once the personal rating and special recruiting categories are removed, and substantially stronger than AfricanAmerican applicants.15 Yet when the personal rating is in the model (second panel), the share of Asian-American applicants in the top decile of the admissions index is similar to that of African-American applicants and much lower than white applicants. This further illustrates that racial preferences influence the personal rating. Including it significantly improves the relative position of African-American applicants in Professor Card’s non-academic index. And including the personal rating substantially weakens the relative position of Asian-American applicants in Professor Card’s non-academic index, despite the fact that they have higher academic ratings and similar non-academic qualifications to whites when the personal rating is not included. But even this third panel incorporates various forms of preferences. Some of these will favor Asian-American applicants relative to whites (such as disadvantaged status) and some will not (such as geography). In the fourth panel, I use the portion of the admissions index that comes from Harvard’s ratings that are not inherently academic in nature with the exception of the personal rating (having already shown that this rating is biased).16 These ratings include the following: extracurricular, athletic, teacher1, teacher2, counselor, alumni personal, and alumni overall. As panel 4 shows, Asian-American applicants are just as strong as white applicants on these non-academic ratings. Note that the findings in the fourth panel also speak directly to Professor Card’s selective comparisons of white and Asian-American ratings in section 4.2 of his report. Here he gives equal weight to Harvard’s four profile ratings and ignores the other ratings measures (school support and the alumni ratings). First, this Note that Hispanics do well on this measure, at least in the top admissions decile, as the remaining “non-academic” factors are also affected by preferences, for example, for disadvantaged status and geography. 15 These ratings include the following: extracurricular, athletic, teacher1, teacher2, counselor, alumni personal, and alumni overall. 16 30 overweights the athletic rating; in practice, the athletic rating is not as important to the admissions decision as the other ratings once recruited athletes are removed. Second, it includes the personal rating, which is affected by racial preferences. In the fourth panel, the weights associated with each rating measure are determined by how Harvard values them in the admissions process. Here Asian-American applicants are just as strong as white applicants on the non-academic measures and, as shown in the previous report, substantially stronger on the academic measures. 3.5 Parental occupation varies in highly unusual and unexplained ways over time, undermining its reliability as a variable and its usefulness as a control. There also are substantive issues with Professor Card’s additional variables. His finding of no statistically significant discrimination against Asian Americans hinges in part on adding these as controls. See Card Report 62-75. But the unusual variations among some of these variables raise serious doubts about their reliability. Parental occupation is one example. Professor Card aggregates mother’s and father’s occupations into 24 categories. Table 3.2N shows the number of mothers and fathers in each of these categories for five of the occupations conditional on being in Professor Card’s pooled dataset; the rest of the occupations are shown in Appendix Tables B.3.1N, B.3.2N. 31 Table 3.2N: Mother’s and father’s occupations vary in non-credible ways over time The yearly variations among these five occupations point toward these variables either being the result of incorrect mappings across years, or being recorded incorrectly. No mother or father is listed as self-employed in 2014, yet over 900 mothers and over 2,100 fathers are listed as self-employed in each of the other years. Over 1,000 of mothers and fathers each are listed as low-skilled in 2014, but in every other year, no more than 50 mothers and fathers were recorded as lowskilled.17 The problems with this variable are not confined to inconsistencies between 2014 and the other years of data. Consider the unemployed category. In 2018 and 2019 there are 10 or fewer unemployed mothers and fathers in each of the years. From 2015 to 2017, however, the number of unemployed mothers was always above 2,200 and the number of unemployed fathers was at least 1,300. These inconsistencies raise doubts about the reliability of the field and its A review of Professor Card’s analysis shows an incorrect translation between some of the coded occupations and how Professor Card aggregates occupations. For example, on the summary sheet of applicant 1759088, both the applicant’s mother and father were listed as “Laborer, unskilled.” Handwritten notes show the occupations as “caregiver” and “newspaper deliveryman”. Yet Professor Card’s classification scheme results in this applicant being coded as “Skilled Trades Incl. Construction.” The error for this occupation appears to be in the mapping Professor Card provides in ca_occupation_to_bls_minorg.xlsx. 17 32 usefulness as a control. If there is little reason to trust the accuracy of a factor, incorporating it into a model will not inform the resulting estimates. Professor Card nowhere offers an explanation for why these data would vary so wildly across these years. Nor does he provide a particularly compelling explanation for how parental occupation categories influence admissions decisions. To the extent there is testimony about this topic in the record, it suggests that parental occupation is useful mainly to help identify disadvantaged students 18 —and the model I use already accounts for any applicants that Harvard identified as disadvantaged. As a result, I see no reason why the parental occupation Professor Card uses would increase our understanding about the admissions process at Harvard—let alone serve as a firm basis for opining that there is no significant discrimination against Asian-American applicants in Harvard’s admissions process 4 Professor Card’s models, once corrected of their key flaws, show that Harvard imposes a penalty against Asian-American applicants. In Sections 3.6 and 3.7 of my opening report, I employed statistical methods to demonstrate that Harvard imposes a penalty against Asian-American applicants in both the scoring and selection of applicants for admission. In Section 5 of his report, Professor Card contends that his analysis reveals no significant penalty against Asian-American applicants. Card Report 46-80. In analyzing this question, Professor Card employs two versions of his models—one that pools the six-years of data and analyzes it as a whole (my preferred approach) and the other separately analyzes each year (Professor Card’s preferred approach). 33 This is one of the most important modeling differences between us, so I will begin by explaining why the pooled model is the better approach, before demonstrating that under either approach, making some basic corrections to the models confirms that there is a penalty against Asian-American applicants. 4.1 Professor Card’s preferred yearly model is less sound than a pooled approach. Professor Card contends that regression models of discrimination in Harvard admissions should focus on individual years, rather than pooling the six years of data Harvard has disclosed. This is his key rationale: First, the admissions process at Harvard is, by its nature, an annual process. Each applicant is compared to other applicants who applied in that year. A pooled analysis does not reflect how the process actually works, because it effectively compares applicants from different years to each other. Card Report 51. Professor Card is wrong for two principal reasons. First, he is wrong that all applicants each year are compared to all other applicants. That is certainly not true with respect to recruited athletes, who are compared only to other athletes, and it is largely untrue for legacies. Many of the early action applicants are not compared to those in the regular admission pool. And, as I have noted earlier, a large proportion of applicants do not meet minimal Harvard admissions criteria and are thus eliminated from consideration at an early stage (e.g., their applications do not receive a second read). Second, the main effect of using a yearly instead of pooled analysis is that it reduces the statistical power of the sample. Statistics is largely driven by the law of large numbers; in any quantitative analysis, the ease of distinguishing random variations from systematic factors rises in proportion to the square root of the sample size. In any analysis of discrimination, it is logical and important to use the largest sample that is relevant to the comparisons involved. 34 To see this, consider the example of a large firm that discriminates against women in making promotions to partnership. Suppose that data disclosed by the firm show that over a six-year period, women with similar evaluation scores to men, with similar billings, and so on, are promoted at a substantially lower rate. Suppose there is substantial corroborating evidence of discrimination in evaluations and work assignments. Suppose then that the firm’s response is that it is inappropriate to attack it for a pattern manifest over a six-year period. “Decisions to promote associates to partnership are made on an annual basis,” the firm’s expert says. “One must analyze one year at a time.” Doing so would, of course, reduce the statistical significance of findings of discrimination, but it would not make any sense. Finally, Professor Card contends that a yearly model is more appropriate because admit rates for the same rating profile are different across years. Professor Card gives as an example that those who receive a 2 on all four profile ratings (academic, personal, extracurricular, and athletic) have admission rates that vary between 61% and 77% across admissions cycles. Card Report 54. This is misleading: just because admission rates vary across years for this rating combination does not mean a pooled model should be ruled out. In fact, my pooled model actually does an excellent job in predicting these exact fluctuations. Specifically, for those who receive this rating combination, the correlation between my model’s prediction of the yearly admission rates and the actual yearly admission rates is extremely high (0.91). My model is able to explain the differences across years through a variety of channels, including that admissions are becoming more competitive over time (as captured by year effects) and in how the other characteristics vary across years for those who received this particular rating combination (e.g., racial composition, disadvantaged status). By using a yearly model, Professor Card achieves results that weaken the effect of race in Harvard’s admissions process by adding noise to the estimated racial preferences and penalties. But making even a few adjustments to both his pooled and yearly models result in significant findings of an Asian-American penalty, and even more substantial racial preferences than even Professor Card finds to exist for Hispanic and African-American applicants. 35 4.2 Professor Card’s pooled results, with small corrections, show that Harvard imposes a penalty against Asian-American applicants. Professor Card first pools the six years of admissions data and runs logit models similar to mine. But he chooses a different set of controls and, importantly, substantially restricts how race affects the admissions process. Professor Card’s pooled results have a number of flaws, many of which have been discussed above, and all of which have the effect of concealing the extent of Harvard’s discrimination against Asian-American applicants. First, Professor Card improperly relies upon the assumption that Harvard’s discrimination against regular Asian-American applicants is irrelevant unless the same level of discrimination is present with respect to Asian-American applicants who are athletes, legacies, and/or members of other special (preferred) recruiting categories. Including these variables in the analysis—and also ignoring how race interacts with these variables—serves only to conceal the impact of the penalty Harvard imposes on Asian-American applicants. It is thus essential to either (1) remove these applicants from the analysis; or (2) allow for the possibility that the effect of race is different for these applicants (i.e., interacting these variables with race).19 By failing to do either, Professor Card makes it impossible to fairly consider the effects of the racial preferences and penalties Harvard employs in its admissions process. Second, Professor Card errs in assuming that racial preferences operate the same way for disadvantaged students as they do for advantaged students. As noted in section 3.7 of my opening report, African-American and Hispanic applicants receive a smaller preference for disadvantaged status than Asian-American and white applicants, and including them in the analysis without interacting race weakens the distinctions between white and Asian-American applicants. Third, Professor Card errs in including Harvard’s personal rating in his models. See Card Report 69-74. As discussed above, it is clear that racial preferences affect the Note that removing these observations is equivalent to allowing the coefficients of the model to be fully interacted with special recruiting status. 19 36 personal rating. By ignoring this evidence, Professor Card’s regressions simply spread Harvard’s discrimination across multiple variables, making the main effect (Asian-American discrimination) smaller and thus harder to measure as statistically significant. Fourth, Professor Card errs by including data for parental occupation that, as shown in section 3.5, is unreliable, given the wide variation in yearly patterns. See Card Report 43-45. Table 4.1N shows that, after removing those in the special recruiting categories, corrective adjustments to Professor Card’s models that account for any of these issues results in significant estimates of discrimination against Asian-American applicants. Table 4.1N: Small corrective adjustments to Professor Card’s model show penalties against Asian-American applicants *=statistically different from zero at the 95% level. Marginal effects are calculated without perfect predictions. The first row of Table 4.1N reports the marginal effect for Professor Card’s pooled model. This is the average marginal effect of Asian-American status itself for all Asian-American applicants who have a non-zero probability of admission.20 The Throughout Professor Card’s report when he calculates marginal effects he includes those whose characteristics are so bad that rejection is guaranteed regardless of their race. This serves to lessen the actual penalty or preference by averaging in zeros from those who are clearly not competitive. I have removed those applicants in this analysis to get a more accurate measure of the marginal effects. 20 37 average effect in this first row is small and statistically insignificant. The second row removes those in the special recruiting categories. The average marginal effect increases by 61%, but the coefficient remains insignificant—meaning that we cannot rule out that the effect is different from zero at the 95% level. The next three rows illustrate how Professor Card’s decision to drop the interaction of disadvantaged status with race affects the magnitude and statistical significance of the penalty Asian-American applicants face. Row 3 adds interactions between race and disadvantaged status, allowing the preference for disadvantaged applicants to vary by race. Once that is done, it reveals a statistically significant penalty for Asian-American applicants. In other words, Professor Card’s conclusion that there is no Asian-American penalty hinges on his error in failing to include interactions between race and disadvantaged status. Rows 4 and 5 in Table 4.1N underscore how Professor Card’s model ignores the way that Harvard’s racial preferences operate in practice. If I estimate the model only on white and Asian-American applicants (row 4), it shows a statistically significant penalty against Asian-American applicants. Note how close the estimated marginal effect is to the one where race is interacted with disadvantaged status. This confirms that Professor Card’s model downplays the effect of race by ignoring Harvard’s differential treatment of disadvantaged students. Row 5 further illustrates this point: by estimating only on students who are not disadvantaged— including African Americans and Hispanics—Professor Card’s model once again shows a statistically significant penalty against Asian-American applicants. All of this demonstrates the fragility of Professor Card’s models, exposing the unreliability of his finding that there is no Asian-American penalty. The same effect occurs if we discard Professor Card’s use of the personal rating, which shows compelling evidence of bias on the part of Harvard’s admission officers. See Section 3.3.2, supra. Row 6 of Table 4.1N retains Professor Card’s flawed application of disadvantaged status, but removes the personal rating. Once again, there is a statistically significant penalty against Asian Americans. Professor Card’s use of unreliable data on parental occupation likewise skews his results. In row 7 of Table 4.1N, I keep Professor Card’s flawed application of 38 disadvantaged status and the personal rating, but instead remove the parental occupation variables. Making this change alone once again reveals a statistically significant penalty against Asian-American applicants. This exercise reveals the extent to which small adjustments to Professor Card’s pooled model expose his flawed conclusions. It underscores that his model is not robust, and that his analysis appears carefully constructed to downplay the extent to which Harvard’s process penalizes Asian Americans. The next section reveals the same shortcomings in Professor Card’s yearly model. 4.3 Professor Card’s yearly models, with small corrections, confirm that Harvard imposes a penalty against Asian-American applicants. In my report, I analyzed Harvard’s admissions process over the full six-year period. I chose to pool the data because Harvard’s admissions process (and its use of race) underwent no material changes during this time, and the six-year period ensures a larger overall sample size, increasing the confidence in the results. Professor Card claims that it is inappropriate to analyze the results of Harvard’s application process by pooling six years of data, and that instead every year should be treated independently. See Card Report 51-54. I disagree, in part because treating each year separately decreases the sample size and thus makes it more difficult to measure the effects of race in Harvard’s admissions decisions. But even using Professor Card’s yearly approach confirms that Harvard imposes a penalty against Asian-American applicants, once the key flaws in his model are corrected. To demonstrate this, I estimated slightly modified versions of Professor Card’s yearly models, correcting his models for the key flaws I have identified and explained elsewhere. As before, I removed those applicants in special recruiting categories, and otherwise adopted the alternative versions of Professor Card’s models described in the pooled analysis (interacting race with disadvantaged status; discarding the biased personal rating; 21 and discarding the unreliable parental Professor Card does do some limited analysis with the personal rating included in his yearly models. But he presents his results in a misleading way. Namely, his results with the personal rating in the model) show a statistically significant penalty against Asian 21 39 occupation variables). The first column of Table 4.2N shows the results of these slight corrective adjustments to Professor Card’s yearly model. Any one of these adjustments reveals that the estimated penalty against Asian-American applicants is substantially higher than Professor Card suggests. The models in rows 2 through 7 show a statistically significant penalty against Asian-American applicants at the 90% level, with the last three also statistically significant at the 95% level. Table 4.2N: Small corrective adjustments to Professor Card’s yearly model show penalties against Asian-American applicants *=statistically different from zero at the 95% level. Marginal effects are calculated without perfect predictions. Again, the models I report take the additional variables that Professor Card has added at face value, ignoring that these additional variables actually distort the analysis in ways that tend to conceal the discrimination against Asian-American applicants. In particular, these models include Professor Card’s questionable use of the data on extracurricular activities. Americans—even with all their other flaws. Professor Card claims that this evidence is nonetheless weak because in only one of the individual years is it statistically significant. But because Professor Card’s model is estimated at the yearly level, the imprecision of the estimates becomes much larger. This makes it impossible to rule out the possibility of even very large penalties in any given year. Professor Card also argues that once 2018—the year that preceded the SFFA lawsuit—is removed, the average over the remaining five years is no longer significant. But it is arbitrary to remove 2018 in this manner. One could just as easily remove 2019, and Card’s model shows significant estimates across the remaining years. The upshot is that Professor Card’s models that include the personal rating show a penalty against Asian-American applicants even with no other adjustments to the model. 40 Data on extracurricular activities come from applicants listing (1) each activity they participated in, (2) the years in which they participated in this activity, (3) the hours per week and weeks per year they participated in the activity, and (4) whether their participation was during the school year or outside the school year. Each of the activities is assigned to one of 29 categories (e.g., work, academics, musical instruments). In his analysis of these activities, Professor Card considers the first two activities listed, aggregating the listed activities into one of twelve groups in a somewhat arbitrary manner. For example, Card aggregates some large categories like religious and volunteer activities, groups some categories like “school spirit” and “LGBT” into an “other” category, and leaves “Junior ROTC”—one of the smallest categories—by itself. More importantly, the level of participation of the activity is done only for the work category, where Professor Card calculates the total hours in work activities over the course of the applicant’s high school career. This distorts the analysis in two ways. First, it overemphasizes the weight that work is given in the process, as work activities are only the eighth most popular activity listed for whites. See Card Report 180, Exhibit 66. Second, white applicants work significantly more hours than Asian-American applicants. Yet there are many activities where AsianAmerican applicants invest substantially more hours than white applicants. 22 Professor Card provides no explanation in his report for this idiosyncratic approach to extracurricular activities. I make the following adjustments to Professor Card’s extracurricular activities in order to more accurately account for their effect on admissions decisions: • Rather than use Professor Card’s groupings when constructing indicators for each of the first two listed activities, I use the original 29 activity categories. This is presumably among the reasons why Asian-American applicants tend to have higher scores than white applicants on Harvard’s extracurricular rating. 22 41 • Rather than use the total hours of work over the course of the applicant’s high school career, I consider broader groupings of categories and measure participation both by (1) counting the number of grades in which the applicant participated in each activity and (2) indicating whether the applicant’s total accumulated hours in a category was above the median for those who had any positive hours in the category. Making these adjustments more precisely accounts for the impact of extracurricular activities on admissions decisions. This more accurate picture of extracurricular activities reveals that the penalty against Asian-American applicants is higher than Professor Card suggests. The results of the models using additional measures of extracurricular involvement are given in the second column of Table 4.2N. For each model, the Asian-American penalty is larger in column 2 than in column 1. Further, all of the deviations that are in rows 2-7 are statistically significant at the 95% level. All Professor Card demonstrates is that one can selectively choose and count extracurricular activities in a way that disadvantages Asian-American applicants— and thereby conceal the discriminatory nature of Harvard’s admissions process. But accounting for the full distribution of activities shows that my finding of an AsianAmerican penalty is robust. Finally, column 3 of Table 4.2N shows the results using the various corrective adjustments to Professor Card’s models to the six-year pool as a whole, rather than year-by-year. Examining the marginal effects shows that, if anything, the effects on Asian-American applicants are more often larger in the year-by-year model. I suspect that Professor Card prefers the yearly model for the same reason he prefers to add other irrelevant or unreliable information to his model: it introduces more noise into the estimates by adding many more variables, all of which tend to conceal the degree to which Harvard discriminates against Asian-American applicants. Consider the estimated penalties (marginal effects) in column 1 for the model using only white and Asian-American applicants. The estimated penalty is bigger than the corresponding penalty from the pooled model in column 3 (-0.37% versus -0.34%), yet is not statistically significant at the 95% level. Professor Card actually shows many results of the marginal effect of being Asian 42 American for his yearly models where the effects are rarely significant at the 95% level (see Exhibits 19, 21, 22, 23, and 25). But because Professor Card has effectively introduced a lot of noise into his models, it is also not possible to rule out large penalties against Asian-American applicants. In all the years except for 2019, the 95% confidence interval in Professor Card’s yearly model with just the special recruiting categories removed contains an Asian-American penalty of 0.9 percentage points. This is a large change given the admit rate is 5.1% for Asian-American applicants who are not within the special recruiting categories. In fact, 2019 consistently shows the smallest Asian-American penalty of all the yearly models (or, in some cases, no penalty at all). It also is the first (and only) admissions cycle after the SFFA lawsuit. In the first column of Table 4.3N, I show the marginal effects of being Asian American (the Asian-American penalty) by year for each of the models. Table 4.3N: The year-over-year evidence of an Asian-American penalty *=statistically different from zero at the 95% level. Marginal effects are calculated without perfect predictions. Note that this is the specification that uses Professor Card’s extracurricular measures. In every specification that does not include the special recruiting categories, the marginal effect is negative in all years but 2019. In 2019, the estimate is positive for all specifications, except in the last column that implements all three corrections: interacting disadvantage with race, removing parental occupations, and removing the personal rating. The last row of Table 4.3N shows the average marginal effect excluding 2019. Now the marginal effects are significant at the 95% level for specifications (2) through 43 (4)—even when Professor Card’s extracurricular controls are used. Professor Card also argues that the evidence is even weaker for discrimination against Asian-American applicants when considering particular subgroups, namely those on California dockets and females. 23 But there are problems with his subgroup analyses. First, the same criticisms of Professor Card’s yearly and pooled analyses apply here. It is incorrect to include special recruiting categories; it is incorrect to ignore racial interactions with disadvantaged status; the parental occupation variable is unreliable; and the personal rating is biased. Second, Professor Card uses his yearly model to generate his findings. These yearly models have very large standard errors that increase when significantly less data are used in the analysis. What Professor Card has failed to show is whether any of his subgroup analyses yield results that are statistically different from his other findings. Third, Professor Card controls for each unique rating combination, aggregating combinations with less than 100 applicants by how similar they are in their admission rates. This too serves to hide racial preferences and penalties, an issue that becomes more salient for smaller estimation samples (which is the case in his subgroup analysis). As I show in section 7.2, these aggregations in the yearly models are inappropriate but are surely worse when these aggregations are done at the yearly subgroup level.24 4.4 Professor Card’s analysis of applicants whose race is missing further confirms the existence of an Asian-American penalty Professor Card makes another argument that inadvertently shows a penalty against Asian-American applicants. In my opening report, I noted that the impact of racial preferences on Asian-American applicants is likely understated due to some Asian Americans choosing not to report their race. Without racial preferences, some of those applicants would see their probabilities of admission rise, as would be the case for all applicants who are not underrepresented minorities. As I show in section 9.3, there is actually evidence of discrimination against dockets that have a higher share of Asian-American applicants. 23 The impact of these aggregations on the magnitude of racial preferences in my pooled model is substantial. See Table 8.2N. 24 44 In an attempt to undermine my argument, Professor Card shows that, using other sources, it is possible to identify the race for some of the applicants who choose not to report. Card Report 55. Professor Card notes that when he uses this information to classify many of those who do not report a race to particular racial groups and estimates his model, the estimated penalty for Asian-American applicants goes down. But Card’s analysis does not undermine my argument at all—it actually shows that Harvard does not impose a racial penalty on those Asian-American applicants who do not identify their race.25 That the penalty against Asian-American applicants falls when some of this group is included as Asian American in the analysis necessarily means that this group is actually treated better than those who report their race as Asian-American. 5 Professor Card’s Analysis Actually Demonstrates That Race Is a Determinative Factor in Harvard’s Admissions Decisions. I previously demonstrated that race plays a “significant role in admissions decisions at Harvard.” Arcidiacono Report 7-8. Professor Card does not disagree. See Card Report 10, 81, 93. Professor Card instead claims that race is not a “determinative factor” in admissions decisions. He attempts to support this claim by showing the average marginal effect of race by year for Asian Americans, African Americans, Hispanics, and applicants who do not identify their race using his preferred yearly model. See Card Report 81 & Exhibit 26. But in doing so, Professor Card actually demonstrates that race is in fact a determinative factor in admissions decisions. Average marginal effects of race show how, on average, admission probabilities change as a result of the applicant’s race/ethnicity. This is what Professor Card shows in his Exhibit 26. But what Professor Card leaves out is that average marginal effects must be interpreted relative to the baseline probability of Indeed, this coincides with anecdotal evidence that some Asian-American applicants hide their race on college applications to avoid discrimination. See, e.g., Fearing discrimination, Asian college applicants don’t always declare ethnicity, Associated Press (Dec. 3, 2011), www.nydailynews.com/ news/national/fearing-discrimination-asian-college-applicants-dondeclare-ethnicity-article-1.986416. 25 45 admission. To illustrate, consider the case of a relatively moderately selective college, where the average admit rate for a particular racial group is fifty percent in the absence of racial preferences. If the average marginal effect of race for that group were six percentage points, then the average admission probability with racial preferences would be fifty-six percent—i.e., the effect of racial preferences would amount to a 12% increase in the number of admitted students in this racial group. But the impact of racial preferences resulting in a six-percentage-point effect is much greater at a highly selective school—where the baseline probability of admission is much lower. For example, if the average admit rate for a specific racial group were three percent in the absence of racial preferences and the average marginal effect of race were (again) six percent, that would mean that the average admission probability with racial preferences would be nine percent. In this scenario, the effect of racial preferences would be massive—a tripling of the admit rate and thus the predicted number of individuals admitted for that specific racial group. Professor Card’s analysis of his own preferred yearly model shows this very scenario at Harvard. Table 5.1N below replicates Professor Card’s results for African Americans and Hispanics in Exhibit 26, but includes the average probability of admission for these groups—both with and without racial preferences—in order to illustrate the effect of racial preferences as compared against the baseline. Professor Card’s own models show that racial preferences are responsible for tripling the number of African-American admits and doubling the number of Hispanic admits.26 For purposes of this analysis, and consistent with Professor Card’s approach, I am including within the Hispanic category those applicants whose race/ethnicity is identified as Native American, Hawaiian/Pacific, or “Other.” 26 46 Table 5.1N: Admission probabilities and marginal effects for African Americans and Hispanics in Professor Card’s yearly models *indicates statistically significant at the 95% level. To illustrate, consider domestic applicants for the class of 2014 (in the first row of Table 5.1N). Absent racial preferences, African-American applicants would be treated as white applicants. Professor Card’s models predict that if racial preferences were removed, the average admit rate for African-American applicants would be 3.38%. This is the baseline (i.e. the starting point absent racial preferences). Compared against this baseline, Professor Card’s average marginal effect of race for African-American applicants (7.43%) would increase the admit rate for African-American applicants to 10.81%, more than tripling the admit rate for African-American applicants in 2014. And this is not an outlier. Professor Card’s overall average marginal admit rate for the entire six-year period is 6.12%. Compared with a baseline of 2.79% (the average admit rate absent racial preferences for the six-year period), the tripling effect exists for the entire period. Professor Card’s own preferred model and analysis thus reveal that racial preferences more than triple the admit rate for African-American applicants. Professor Card is undoubtedly aware of the multiplying effect that racial preferences have on African-American and Hispanic admit rates. The fact that he fails to address them is revealing. His analysis—which demonstrates that racial preferences alone are responsible for doubling and tripling the admit rates for African-American and Hispanic applicants—quite obviously demonstrates that race is a determinative factor in admissions decisions at Harvard. 47 So how does Professor Card come to the conclusion that race is not a determinative factor in admissions decisions at Harvard? Professor Card resorts to three misleading and/or plainly incorrect arguments, claiming that: • A model with race as the only control does a poorer job of explaining admissions decisions than other sets of controls (e.g. profile ratings, dockets) See Card Report 83, Exhibit 27. • For most African-American and Hispanic applicants, the average marginal effect of racial preferences is small. See Card Report 84, Exhibit 28. • Unmeasured factors are more important than racial preferences. See Card Report 86-87, Exhibits 29 and 30. Below I show that each of these arguments is incorrect or misleading. In doing so, I rely only on Professor Card’s models. As I show in Section 8.2, my preferred model shows even larger estimates of racial preferences. 5.1 A model with race as the only control would be expected to perform poorly relative to other factors. Professor Card’s first argument can be ruled out almost immediately. As I showed in my opening report, there are vast differences in academic preparation across racial/ethnic groups. For example, in the expanded data set, over 37% of AfricanAmerican applicants are in the bottom decile of the academic index compared to fewer than 4% of Asian-American applicants. And less than 1% of African-American applicants are in the top decile of the academic index compared to almost 18% of Asian-American applicants See Arcidiacono Report, at Table B.5.1. Because race is generally correlated with academic preparation, one would expect that race would have at least some explanatory power with respect to the admissions results. Professor Card’s findings in Exhibit 27—that race alone explains little variation in admissions—actually suggests, as a statistical matter, that racial preferences are quite large. To more clearly see this, suppose Harvard had a strict quota system, accepting the 48 best 10% of each racial group. Estimating a model of admission where the only control was race would have zero explanatory power even though an explicit quota was in place. In order to properly evaluate the role of race in the admissions process, it is paramount that one controls for the relevant factors in the admissions decision. Professor Card’s analysis in Exhibit 27 does nothing to support his claims. 5.2 Professor Card’s argument that racial preferences are not relevant for most African-American and Hispanic applicants misleadingly focuses on uncompetitive applicants. Professor Card’s second argument is that, for the majority of Harvard applicants, race is not relevant to the admissions decision. This argument is a dodge. Of course race is not relevant for a large number of applicants. No one would claim otherwise, given that Harvard is a highly selective school where more than 90% of all applicants are rejected. See Section 3.1, supra. A further example may help illustrate the point. Suppose Harvard sent automatic rejection letters to the 80 percent of its applicant pool with the lowest standardized test scores. Further assume that, among the remaining twenty percent, half (ten percent of the pool) were admitted. Of those who were admitted, suppose Harvard did so using a specific quota for each racial group, and admitted the other half based on purely non-racial factors, such as academics, extra-curricular achievement, and so on. It would still be true that race did not affect most application decisions; indeed, race would only affect 5% of Harvard’s decisions. But this would be no defense to the way race was used in admitting the competitive applicants. The fact that the majority of all applicants are rejected regardless of their race tells us nothing about the effect race has among those applicants who are seriously considered for admission to Harvard. This fallacy can be seen in Exhibit 28 of Professor Card’s report (which is reproduced in the first two columns of Table 5.2N, below). Here, Professor Card ranks applicants according to their admissions index which, given the estimates of his yearly models, describes the strength of applicants based on how the applicant’s observed characteristics translate into admissions. Professor Card does this ranking 49 separately by race, implying that 10% of each racial group is in each decile.27 Table 5.2N: Average marginal effects of race by within race and across race admission index deciles using Professor Card’s yearly models The first five rows of column 1 show no effect of race for African-American applicants in the bottom five deciles. The reason there is no effect in the bottom five deciles is that Professor Card’s yearly models predict that more than 50% of African-American applicants have other observed characteristics (combinations of test scores, Harvard ratings, etc.) where everyone who has these characteristics is rejected. And I agree that racial preferences are not relevant for uncompetitive applicants. I further agree with Professor Card that racial preferences are most salient for the competitive applicants. But Professor Card makes a mistake when he describes who is affected by racial preferences: [T]he applicants with the largest estimated positive effect of race on their likelihood of admission are the strongest applicants—i.e., those Part of the note to Exhibit 30 reads “Deciles are constructed by race based on the predicted probabilities of admission when the race factor is turned off.” Note that whether the race factor is turned off or not has zero relevance as to who is assigned to what decile, when the deciles themselves are constructed by race. 27 50 whose estimated likelihood of admission is in the top 10% of the applicant pool absent consideration of race. Card Report 84. But Professor Card’s calculations are not for those who are in the top 10% of the applicant pool, but instead for the top 10% of African-American applicants. The third and fourth columns of Table 5.2N show the average marginal effects by admissions index decile where the deciles are constructed across all racial groups and where racial preferences are turned off.28 Racial preferences are relevant for the top 10% of African-American applicants (column 1), who are distributed across the top 30% of the applicant pool (column 3). Given that the admission rate across all racial groups over this period is slightly over 7%, it is not surprising to find smaller effects of racial preferences for those in the bottom 70% of applicants. To further illustrate this point, the last four columns show the share of AfricanAmerican and Hispanic admits in each of the academic index deciles. Using the within-race deciles, over 82% of African-American admits are in the top decile (the top 10% of African-American applicants). But this is exactly where the marginal effects of race are enormous: Professor Card estimates the marginal effect for this group to be over 47%, as shown in column 1. 5.3 Professor Card’s method of calculating the importance of unobserved factors is incorrect and substantially overstates their importance. Professor Card next claims that unobserved characteristics are more important than race, again suggesting that race is not a determinative factor. Card Report 8586. Professor Card reaches this conclusion using erroneous methods. Properly accounting for the role of unobserved characteristics shows that Professor Card vastly overstates the importance of unobserved characteristics relative to race for Professor Card defines his admissions indexes for this table without accounting for differences in admission rates by year. Hence, the same applicant would have a higher index in 2014 than in 2019 as admission rates as a whole were higher in 2014. To form the across-race deciles, I remove the effects of year by creating the deciles at the year level; each decile has 10% of each year’s applications. While my method is the correct one, this has little effect on the patterns shown in Table 5.2N. 28 51 African-American and Hispanic applicants. Indeed, I will show that unobserved characteristics are decidedly less important than race for these two groups. Professor Card attributes the share unobserved characteristics play in the admissions decision as the absolute value of the difference between the predicted probability of admission and the actual admit decision. But this does not equate to the share of the admissions decisions explained by unobserved characteristics. What it does give—at least for admits—is whether the applicant has unobservable characteristics above a particular percentile. Knowing that the unobservable is above a particular percentile is useful, but not in the way Professor Card uses it. The predicted probability of admission indicates how often we would expect an applicant to be admitted given a random draw from the distribution of unobserved characteristics. Some of those random draws would result in rejection, others in acceptance. When an applicant is rejected, this tells us that the set of possible unobserved characteristics had to lie in some range, but not the exact value of the unobserved characteristic. For example, suppose an applicant has characteristics associated with a 90% chance of admission and that applicant was admitted. This means that the applicant’s unobserved characteristics were above the 10th percentile. Professor Card’s method, however, would imply that unobserved characteristics explains 10% of the admissions decision, which is simply false. Knowing that an applicant was admitted and his or her predicted probability of admission tells us the range of possible values for the applicant’s unobserved characteristic. Because both Professor Card and I use logit models to estimate the admissions decisions, our models assume that the unobserved characteristic comes from a particular distribution. I can use these three pieces of information—the distribution of the unobserved characteristic, the predicted probability of admission, and the actual admission decisions—to show:29 • How often the expected value of the unobserved characteristic is larger than the preferences for a particular racial/ethnic group; and 29 The derivations of the formulas are given in Appendix A.1. 52 • The probability of the unobserved characteristic being larger than the preferences for a particular racial/ethnic group. Table 5.3N shows both Professor Card’s incorrect method, as well as my calculations. Table 5.3N: Racial preferences are more important than unobserved factors for African-American and Hispanic admits using Professor Card’s models The first column replicates Professor Card’s result in the last column of Exhibit 29 for Hispanics and African Americans. Here it is important to note that Professor Card does two misleading things in reporting his results. First, he includes perfect predictions in his estimates as though (tautologically) their unobserved characteristics were at least as important as race to their admissions chances. But these observations provide no information on whether race is more or less important than unobserved characteristics. With over 50% of African Americans having observed characteristics that result in a 100% chance of rejection, Professor Card’s inclusion of these applicants in his calculations substantially overstates his actual findings, even aside from his incorrect method. Second, Professor Card does not break out the results by admitted and rejected applicants. Column 2 uses Professor Card’s method but reports the results only for admits. The differences are striking: Professor Card claims that for 94% of AfricanAmerican applicants, unobserved characteristics are more important than race. But, even under Professor Card’s own model, unobserved characteristics are more important than race for only 30% of African-American admits. Moreover, even this 30% figure is a gross overstatement. In column 3, I show how 53 often the expected value of the unobserved characteristic for each admit is bigger than the estimated racial preference. Using this measure, the unobserved characteristic is bigger than the racial preference only 4% of the time for AfricanAmerican admits. In column 4, I instead show how often admitted applicants could expect to draw an unobserved characteristic that was bigger than their racial preference. The average probability of an African-American admit drawing an unobserved characteristic that was bigger than their racial preference is 10%. The corresponding shares are larger for Hispanic admits as racial preferences for Hispanics are weaker than those for African Americans. Nonetheless, both measures show that racial preferences are more important than unobserved characteristics more than 65% of the time. Clearly, then, unobserved characteristics are substantially less important than racial preferences for these two groups. 6 Professor Card Fails to Refute the Overwhelming Statistical Evidence of a Floor for African-American Admissions. In my report, I showed that Harvard maintained a floor on the admission rate for single-race African-Americans in the classes of 2017, 2018, and 2019. In each of these years, the admit rate for single-race African Americans (as identified under the federal Integrated Postsecondary Education Data System (IPEDS)) 30 was virtually identical to the admit rate for all other domestic applicants, as reflected in Table 6.1N below: Table 6.1N: The admit rate for single-race African Americans is implausibly close to the admit rate for other domestic applicants IPEDS counts an individual as African American only if the individual marks “Not Hispanic; Black or African American only.” If the individual marks, for example, Hispanic and African American, the individual is counted as Hispanic. And if the individual marks White and African American, the individual is reported as “two or more races.” See Collecting Race and Ethnicity Data from Students and Staff Using the New Categories, National Center for Educational Statistics, https://nces.ed.gov/ipeds/Section/collecting_re. 30 54 The difference in these two sets of rates is never larger than .00064—about as small a divergence as is possible, especially given the size of the applicant pool. I found that the chance of this match occurring in three consecutive years (without direct manipulation) is less than 0.2%, and arguably much smaller. See Arcidiacono Report 29. In response to my analysis, Professor Card has three responses, arguing that: 1) It makes no sense that Harvard would impose a floor on the AfricanAmerican admit rate based on IPEDS metrics, because the admission rate by race is never publicly reported. See Card Report 88-89. 2) Because Harvard began reporting its results using the federal IPEDS method before 2017, there is no reason why it would impose a floor during that year. See Card Report 88-89. 3) Under a variety of alternative measures of race and alternative places where a floor could be implemented, there is no evidence of a floor. See Card Report 89-93. None of these responses is persuasive. Indeed, none of them even address the compelling statistical evidence I present. To begin, Professor Card makes no attempt to contest the near mathematical certainty that Harvard is, in fact, manipulating the admissions rate for single-race African Americans to match its overall rate. That is my primary claim based on the statistical evidence. Instead, Professor Card provides irrelevant responses, focusing on other admissions statistics and racial categories that say nothing about what Harvard was doing with IPEDS admissions rates and single-race African-American applicants beginning with the Class of 2017. Moreover, further examination of the characteristics of single-race, African-American admitted applicants confirms that a change occurred with the 2017 cycle, and further bolsters my conclusion that Harvard was in fact taking steps to ensure its admission rate for these applicants was at least as high as the overall admission rate. 55 6.1 Professor Card’s speculation that Harvard would not want to use a floor based on a non-public admissions rate misses the point. Professor Card notes that under the IPEDS reporting process, admissions rates by racial group are not publicly reported by the federal government, and Harvard has declined to make the data public on its own. He assumes that Harvard would have no reason for imposing a floor that the public would never see. To begin, there are undoubtedly many ways Harvard could impose racial floors. They could impose a floor based on the expected number of admits, the share of admits of a particular race, or the relative acceptance rates of particular races. Alternatively, Harvard could impose a floor based on the expected number of enrollees of a particular race. Furthermore, Harvard could do this using a variety of different measures of race. Invariably, each of these different ways and different measures would result in different patterns in the data. However, my claim is that the data show that Harvard implemented a particular kind of floor using a particular definition of race. Why Harvard chose this particular way of imposing a floor is irrelevant as a statistical matter. That said, there are several reasons why Harvard might use a floor that is tied to a metric not publicly reported: • Because rigid floors and other racial quotas are plainly illegal, it seems logical that if Harvard were attempting to ensure a minimum level of admissions for a particular race, it would want to use a metric that was not publicly available, lest its unlawful conduct be detected. • Although the IPEDS admissions rates are not publicly reported, they are tracked within the admissions office and could be used internally—for example, to rebut any allegations that the admissions office was not admitting African Americans at a sufficient rate. Indeed, there is evidence that Harvard was very concerned about the way its IPEDS enrollment numbers were being perceived by the public in early 2013 (during the consideration of applications for the class of 2017).31 See HARV00023588 (Feb. 6, 2013 email to Dean Michael Smith) (“[This] is a piece that explains how we collect and report demographic data, as per federal guidelines.”); 31 56 • Harvard shares IPEDS admissions data by race with other institutions. For example, Harvard shares annual admissions data by racial group—including admissions rates—on an annual basis through the Consortium on Financing Higher Education (COFHE), a voluntary association of 35 selective, private liberal arts colleges and universities. See, e.g., HARV00004736-38 (setting forth Harvard admissions rates by racial group and overall admission rates under IPEDS method); HARV00009158-59 (describing COFHE’s use of IPEDS data). • Likewise, admissions officers from Harvard attend semi-annual meetings of the Association of Black Admissions and Financial Aid Officers of the Ivy League and Sister Schools (ABAFAOILSS), at which Harvard officers bring data on admissions rates—including IPEDS data—and other institutions appear to share IPEDS admission rates by racial group during the admissions cycle. See HARV00014684-868; HARV00067679. Ultimately, Harvard’s reason (or combination of reasons) for establishing a floor for single-race African-American admissions based on IPEDS metrics is outside both my (and Professor Card’s) expertise. The data demonstrate that this racial floor exists. Professor Card does not and cannot dispute that Harvard maintained a floor on the admission rate for single-race African-Americans in the classes of 2017, 2018, and 2019. 6.2 Contrary to Professor Card’s arguments, there is additional evidence that Harvard began implementing the floor in 2017. Professor Card notes that Harvard changed their reporting of race to the federal government, using the IPEDS method, before the 2017 cycle. Again, this is irrelevant to the fact that Harvard maintained a floor on the admission rate for single-race African Americans in the classes of 2017, 2018, and 2019. There is evidence that the IPEDS numbers became salient to the admissions office during the 2017 cycle. For example, Harvard has produced numerous examples of “one-pagers”—statistical summaries of the applicant pool and admitted class that are provided on a regular basis to the leadership of the admissions office—from the HARV00023594 (“[T]he IPEDS reporting system leads to significantly underreported percentages for all ethnicities except Hispanic Americans. The method used by Harvard and many peer institutions gives a more complete report of the way many students, especially those of mixed heritage, actually view their racial and ethnic identities.”). 57 2017 and 2018 admissions cycles. Early versions of the one-pagers during the 2017 cycle lack any IPEDS data. See, e.g., HARV00014628 (one-pager from January 7, 2013). It appears that IPEDS numbers are reported on a one-pager for the first time on or about January 12, 2013. See HARV00019910 (one-pager dated January 12, 2013, comparing early admits for 2016 with early admits for 2017).32 And again, this change coincides precisely with evidence reflecting increased concern within the admissions office about IPEDS reporting and the admission of students by race.33 6.3 Professor Card’s analysis of other data does nothing to undermine my claim that Harvard maintained a floor on the admission rate for single-race African-American applicants. In Exhibits 31 through 34, Professor Card shows changes in the fraction of admitted students by race/ethnicity over time. He concludes that because these numbers vary over time, there must not be a floor. But all of Professor Card’s exhibits use a measure of race that is not the one I claim that Harvard used in imposing a floor on the admission rate of single-race African-American applicants. Further, it uses an outcome measure—the fraction of admitted students of a particular race/ethnicity— that is unrelated to my claim. None of these exhibits have anything to do with my claim. To repeat, my claim is that there was a floor on the admit rate of single-race African-American applicants for the classes of 2017 to 2019. Professor Card also argues that a post-2016 floor cannot be occurring because the estimated marginal effect of race on African-American admissions (including both single-race and multi-race African Americans) is smaller in the period between 2017-2019 than in the period 2014-2016. Professor Card’s argument on this point is misleading, for several reasons: Further evidence of the then-emerging salience of the IPEDS measure of race is that the variables used to construct the IPEDS measure were not included in the main data file Harvard produced for years prior to 2017. After reviewing Professor Card’s report, I discovered that IPEDS numbers for the pre-2017 years were located in other spreadsheets provided by Harvard. Its absence from the main data file further indicates that the admissions office changed its tracking of these data in 2017. 32 33 See HARV00026562; HARV00030511; HARV00023613. 58 • First, and most importantly, the race measure Professor Card uses is (again) not the one upon which my observation of a floor is based. • Second, Professor Card overstates the difference in the marginal effects. The marginal effects he estimates across years are similar in magnitude and not statistically different from one another. A similar pattern emerges if the marginal effects are averaged over the two periods. The difference in the average marginal effects is small and statistically indistinguishable from zero. • Third, because overall admit rates have been falling over time, it is no surprise that the marginal effects would be slightly smaller in later years (though, as noted above, not significantly different). • Finally, as explained above, see supra Section 5, how meaningful marginal effects are necessarily depends on how competitive the pool is: a difference of six percentage points is much more meaningful when the baseline admit rate is 5% than when it is 10%. 6.4 Differences in the characteristics of admitted single-race African Americans after 2016 further support evidence of a floor. While Professor Card’s response does not address my key claims, there is another way to test whether Harvard changed its admissions practices with respect to single-race African Americans in 2017: compare the difference in characteristics between single-race African-American admits and multi-race African-American admits in the admitted classes of 2014-16 and the 2017-19 cycles. Because my claim is that there was a shift in focus towards the admit rate of single-race African Americans, I would expect to see a change in the strength of admitted single-race African Americans relative to their multi-race counterparts. I focus on the academic index as a measure of applicant strength because it is a continuous measure with a well-defined formula. 59 Table 6.2N: Mean academic index for admitted single-race and multi-race African Americans by class *=statistically significant at 95% level. Academic index is in standard deviation units. Difference refers to the single-race academic index minus the multi-race academic index. Table 6.2N shows the average academic index, in standard deviation units, for admitted single-race and multi-race African Americans by year. The difference between the average academic index for single-race African-American admits and multi-race African-American admits is presented in the third column. There is no significant difference between the academic index of single-race and multi-race African American admits in any of the pre-2017 cycles and, as shown in the last row of the first panel, aggregating across the three pre-2017 cycles shows no significant differences. But the results for the post-2016 cycles, shown in the bottom panel, indicate a markedly different pattern. In each case the difference is negative, and more negative than any of the differences in the pre-2017 cycles. This gap is significant for 2019 as well as for the period as a whole. These differences show that Harvard was admitting single-race African Americans with significantly lower academic indexes than their multi-race counterparts beginning in the post-2016 period. This is striking because it is precisely what would be expected if Harvard began imposing a floor on single-race African-American admit rates after 2016. Examining the admit rates of single-race and multi-race African Americans in the different admission cycles further confirms that Harvard changed its practices in 2017. These admit rates are shown in Table 6.3N. 60 Table 6.3N: Admit rates for single-race and multi-race African Americans by class *=statistically significant at 95%level. Difference refers to the single-race admit rate minus the multi-race admit rate. Ratio refers to the multi-race admit rate divided by the single-race admit rate. Single-race African-American admit rates are 3.2 percentage points lower than multi-race African-American admit rates in the pre-2017 period. After 2016, however, the difference narrows substantially to 1.6 percentage points. Put another way, the average admit rate for multi-race African Americans is 50% higher than the single-race African American admit rate in the pre-2017 period, but only 26% higher in the post-2016 period. This additional statistical evidence further confirms that Harvard changed its admissions practices in 2017 in a manner consistent with the existence of a floor on admission rates of single-race African Americans such that it was equivalent to the admission rates for all other domestic applicants. Professor Card has not challenged the statistical evidence I used in my opening report, instead choosing to focus on data and racial categories that are irrelevant to the question at hand. The evidence on this point is thus both statistically compelling and unrebutted. 7 A Number of the Other Variables Added by Professor Card Are of Questionable Reliability and Undermine the Confidence of His Conclusions. Thus far, I have highlighted the numerous errors and questionable modeling choices that undermine Professor Card’s analysis of the racial penalty Harvard imposes on 61 Asian-American applicants and the racial preferences Harvard affords to AfricanAmerican and Hispanic applicants. Professor Card’s findings and conclusions are further undermined by his inclusion of several variables of questionable reliability. In his report, Professor Card argues that increasing the number of variables analyzed in a model necessarily yields more complete results. See Card Report 4050. But that is true only if the variables are (1) relevant to the analysis, (2) correctly specified (i.e., accurate), and (3) not themselves influenced by racial preferences. Some of the variables that Professor Card uses violate at least one of these criteria. One of them is parental occupation, as explained above in Section 3.5. Other variables that are less important to Professor Card’s result, but still too questionable to rely on, include intended career and staff interviews. The weaknesses in these variables are what led me to exclude them from my original analysis. Further, I disagree with Professor Card’s approach to the ratings data, believing it introduces unnecessary noise into the model and disguises racial preferences. In this section, I describe those variables that (in addition to parental occupation) I choose not to incorporate in my preferred analysis—though in section 8, I show that even including these faulty measures do not affect my findings. 7.1 Intended career varies in highly unusual and unexplained ways over time, undermining its reliability as a variable and its usefulness as a control. Like parental occupation, the applicant’s intended career also varies in ways that are inconsistent over time, casting doubts upon the reliability of this metric and further undermining Professor Card’s models. There are fourteen intended career categories in the Harvard database for these admissions cycles. Table 7.1N shows the number in each of these categories for five of the intended careers; the full set of intended careers are shown in the Appendix, see Table B.4.1N. 62 Table 7.1N: Intended career varies in non-credible ways over time The differences across years are enormous for the same intended career. For academics, the number of applicants who are listed ranges from 13 to 2,247; for law it ranges from 708 to 2,093. Medicine varies from a low of 3 in 2018 to a high of 6,254 in 2014. Health varies from a low of 85 in 2016 to 4,944 in 2018. Again, Professor Card provides no explanation as to why he would be confident about the accuracy of this information or why it varies so widely. 7.2 Professor Card’s approach to using the rating variables suffers from a small-population problem and masks racial preferences, which undermine its reliability. In my original report, I included indicators for each of Harvard’s profile ratings. Professor Card argues instead that all combinations of the profile ratings should be included. In Professor Card’s pooled dataset, there are 287 combinations of athletic, personal, extracurricular, and academic ratings. Of the 287 combinations, 26 of these combinations yield a perfect prediction of admission—meaning every applicant who receives these combinations of scores is admitted. Another 153 combinations yield a perfect prediction of rejection—all of the applicants with these combinations are rejected. This is in part mechanical: of the 179 combinations that perfectly predict rejection, 53 of the combinations contain only one applicant, and the median number of applicants in a combination is 13. Professor Card then pools rating combinations based on their admission rates when the number of applicants in that category is less than 100. There are a number of problems with this approach. The first problem is that aggregating combinations with very small populations leads to admissions patterns that are inconsistent with Harvard’s ratings. For example, consider the admit 63 profile 4341, where the first number is the athletic rating, the second the personal rating, the third the extracurricular rating, and the fourth the academic rating. Two individuals were assigned this rating over this period; both were white, and both were admitted. Yet there are ratings that are objectively higher on Harvard’s scale that nonetheless have substantially lower admit rates: • 4321 is a rating profile that is two points higher on the extracurricular rating and identical on the other three ratings. The 65 applicants that received this rating profile had an average admit rate of 73.0%;34 • 4331 has a rating profile that is one point higher on the extracurricular and identical on the other three ratings. The 126 applicants that received this rating profile had an admit rate of 42.1%; • 3331 has a rating profile that is one point higher on both the athletic and extracurricular ratings. The 96 applicants that received this ranking had an admit rate of 37.5%. The second problem with this approach is that racial preferences are embedded in the ratings aggregation. To see this, suppose a particular ratings combination had more African-American applicants than another ratings combination, but the admit rates for the two combinations were the same. The average admit rates for the two groups are in part due to the strength of the rating profiles, but also in part due to the share of African-American applicants in the two groups. In this example, the rating profile associated with the second group is actually the better profile as the admit rate for the first rating profile was more affected by racial preferences. Professor Card’s aggregation method, then, works to conceal the true effect of racial preferences.35 These issues are compounded in the yearly analysis, where there are even fewer numbers in each of the ratings combinations. Across the six admission cycles, 244 of 34 Throughout this section, when I refer to the average admit rate, I am referring to the average admit rate for the category to which this rating combination was assigned. I show how Professor Card’s rating scheme conceals racial preferences in Table 8.2N and Section 8.3. 35 64 the 287 categories show up as perfect predictions in at least one of the years. Further underscoring the small-sample problem, 15 of the categories perfectly predict admission in one of the years and perfectly predict rejection in at least one of the other years. Constructing ratings groupings at the yearly level results in dramatic fluctuations in the year-by-year admit rates for the same rating, and again in ways that are inconsistent with higher ratings being associated with higher admit rates. This is illustrated in Table 7.2N, which shows by year the average admit rates for those applicants who received each of the following four ratings combinations: 4311, 3321, 4312, and 3312. It also shows the number of observations in each year for that category. Table 7.2N: Using Professor Card’s rating combinations for his yearly regressions leads to inconsistent patterns The admission rates for the same rating combination fluctuate substantially across years. The admit rates for 4321 range from 29.7% to 100%; the admit rates for 4312 range from 17.2% to 63.6%. These large fluctuations result because of sampling variability: using such few observations leads to large sampling error. Comparing the top two rows to one another as well as the bottom two rows to one another shows the inconsistent patterns in how the ratings profiles translate into admission rates. We would expect those who receive a 3321 to be admitted at a higher rate than those who receive a 4321; it is by all accounts a better score. Yet in half of the years, this is not the case. And in two of the four years, admission rates are higher for 4312 than for 3312. The inconsistency itself raises red flags about 65 using the ratings data in this way; the fact that it also reflects Harvard’s racial preferences (as described above) further shows the inappropriateness of Professor Card’s approach to using the ratings data. 7.3 Staff interviews are selectively given and thus should not be used as a control. 36 Who are these fortunate few who receive staff interviews? Table 7.3N shows the number and fraction of each of the four main racial/ethnic groups who receive a staff interview by whether or not they were in one of Harvard’s special recruiting categories.37 Table 7.3N: Staff interviews vary substantially by race and special recruiting status There is an error in how Professor Card codes the scoring of the staff interviews in his pooled analysis. Namely, Professor Card creates a flag for whether someone received some combination of 1’s and 2’s on the staff interviewers overall rating and personal rating, another flag for a combination of 2 on one and a 3 on the other, and finally a flag for two 3’s. Those who receive one 1 and one 3 are then effectively treated as though they had no staff interview. This error, however, has virtually no effect on the results due to the small number of applicants in this category. 36 Special categories are athletes, legacies, faculty or staff children, and Dean’s/Director’s List selections. 37 66 Twenty percent of those who fall into any one of Harvard’s special recruiting categories receive staff interviews. Those who are in these special recruiting categories are disproportionately white. 38 For applicants not in one of these categories, the probability of receiving a staff interview is less than 1.3%. AsianAmerican applicants are least likely to receive a staff interview, both overall and conditional on the special recruitment status. Because these interviews clearly depend on preferences, I do not include them in my analysis. 8 Incorporating Most of Professor Card’s Variables Into My Preferred Model Confirms My Findings Regarding the Effect of Harvard’s Racial Penalties and Preferences. To recap, we can divide Professor Card’s analysis of my report into two broad parts. In one part, he constructs a model to show that Harvard does not discriminate against Asian-American applicants vis-à-vis white applicants. As I have shown, this model is dependent upon many inaccurate assumptions and poor modeling choices. Moreover, it is not robust: if I change just one or two of these assumptions and choices, Professor Card’s model no longer supports his findings and conclusions; in particular, his model confirms the penalty against Asian-American applicants. In the other part, Professor Card tries to show that Harvard’s racial preferences in favor of African-American and Hispanic applicants are not substantial or pervasive. But as I have shown, this analysis is exceedingly weak; even when we use Professor Card’s own model results, they show that Harvard gives African-American and Hispanic applicants heavy racial preferences. For some of the arguments and model specifications Professor Card uses, there is simply no sound justification for the choices, and it is hard to imagine any reason for their use other than to intentionally conceal the effects of race in Harvard admissions. Other adjustments suggested by Professor Card are reasonable, and for still others there is at least a weak case for inclusion. The question is, are my 8.0% of white applicants are in one of these categories, compared to 2.7% of African Americans, 2.2% of Hispanics, and 2.0% of Asian Americans. 38 67 results as fragile as Professor Card’s? Or are they robust and consistent when we incorporate specific changes suggested by him? In this section, I examine the robustness of my model in the context of Professor Card’s analysis. 8.1 Changes advocated by Professor Card that I incorporate in my updated model In my updated model, I adopted six general types of revisions that reflect unobjectionable choices made by Professor Card. 1. Modifying variables. Professor Card codes several variables in a different way than my original model. My update incorporates these changes: • I treat profile ratings of 7, 8, and 9 as missing values; • I include blank teacher ratings as a missing category; • When the SAT score is not present but an ACT score is present, I use the ACT science section in my conversions the same way Professor Card does; • I no longer remove from the analysis those who are missing the overall rating. 2. Adding variables. I incorporate dozens of additional variables that Professor Card uses in his analysis, so long as they meet three conditions: • They must not be themselves measures contaminated by apparent racial or other preferences (e.g., I exclude the personal rating and staff interviews); • They must display consistent patterns over time, thus demonstrating reliability; • They must be present in each year of the data, so that they can be included in the pooled analysis. Many of the additional variables used by Professor Card meet all these restrictions, and I thus incorporate them in my updated model. These include Professor Card’s College Board variables on the characteristics of applicant high schools and home neighborhoods; whether the mother or father is deceased; whether a parent attended an Ivy League university (other than Harvard); whether a parent attended graduate school at Harvard; and the type of high school the applicant attended. 68 As an alternative to Professor Card’s flawed use of the various combinations of ratings data to capture the multidimensionality of the applicant, I create indicators for whether the applicant had each possible combination of a two or better on Harvard’s four profile ratings, indicators for whether the applicant had two or three 2’s or better on their school support measures, and an indicator for whether the applicant had 2’s or better on both of the alumni ratings.39 3. Including early admission applicants in my baseline model. In my original report, my baseline model excluded recruited athletes, legacies, faculty and staff children, those on the Dean’s/Director’s list, and applicants for early admission, in order to focus on the part of the admissions process where anti-Asian discrimination was concentrated, and not on applicants who were subject to special admissions procedures. Professor Card’s model includes all applicants in a single model. As I have pointed out, this produces misleading results because Harvard does not discriminate against Asian-American applicants who are in the special recruiting categories. But I do not have similar objections about including early applicants. Although the early admissions process necessarily involves different considerations than the bulk of the application process, Harvard’s racial penalties and preferences largely apply in this process the same way they do in the regular admissions process. I thus include early admission applicants in my updated baseline model, which is intended to include all applicants whom I believe are at risk of discrimination. 4. Racial definitions. Professor Card collapses racial categories in a different manner than I did in my original report. In my updated model, I use his definitions, which place Native Americans and Hawaiian/Pacific Islanders into the “Hispanic” category. 5. Interactions with year. Professor Card claims that a yearly model is appropriate in part because the composition of the pool changes from year-to-year, and Harvard may pay attention to this. Indeed, we know exactly how Harvard pays Note that these are in addition to indicators for each possible value of the individual ratings (e.g. 2 on the academic rating) that were in my original model. 39 69 attention to it through their “one-pagers,” which provides admissions officials with a snapshot of the current admissions process compared to the prior year. Although Professor Card errs in employing a yearly model, I can account for year-to-year changes in my pooled model by including in my model interactions with year and the characteristics listed on these one-pagers: female, disadvantaged status, intended major, dockets, and, in some specifications, race. 6. Reporting results. Professor Card emphasizes the marginal effects of race in discussing results—in other words, how many percentage points does membership in a particular race increase or decrease one’s admissions rate? The numbers below adopt this approach, reporting these marginal effects, but only for those whose characteristics are such that rejection is not guaranteed (i.e., the perfect predictions are removed). 8.2 The results of the updated preferred model confirm my previous findings and conclusions Table 8.1N, below, shows the marginal effects of race in my original model and my updated model for my baseline dataset that includes early action applicants.40 Table 8.1N: Basic racial penalties and preferences under my original and revised model *=statistically significant at the 95% level. Marginal effects calculated without perfect predictions. As Table 8.1N shows, the numbers in the updated model are slightly different than in the original model, but the story is unchanged. African-American applicants receive extremely large preferences, on average 7.29% off a base of 2.25%; more 40In order to calculate the marginal effects from my original model, I use the results from the original report that included both the special recruiting categories and early action applicants. I then remove the special recruiting categories to calculate the marginal effects. 70 than quadrupling their chances of admission. Hispanic applicants experience large preferences, 4.17% on average off a base of 2.97%, increasing their chances of admission by 2.4 times. Asian-American applicants experience a substantial admissions penalty that lowers their chances of admission by a full percentage point; Asian-American admission rates would be 19% higher if they were treated as white applicants. 8.3 Even incorporating many of Professor Card’s manifestly unsound modeling choices does not alter the result of my model. I now address the modeling choices that Professor Card made and which I find to be unsound or indefensible. I have dissected many of these choices earlier in the report. My goal in this section is to explain the degree to which I think Professor Card’s choices would substantively change the results of my basic analyses, and thereby make clear which assumptions really matter. In the first column of the first panel of Table 8.2N, I show marginal effects for each racial group in the baseline dataset. The rest of the entries show that my findings are robust along a number of dimensions. • Including the personal rating. As noted in earlier sections, Harvard’s personal rating of applicants is severely contaminated with racial bias; ratings are inflated for preferred racial groups, and penalized for Asian Americans. They therefore cannot be included in any sound model of Harvard admissions that is trying to separate out discriminatory effects. Nevertheless, as I show in the bottom panel of Table 8.2N, including the personal rating makes the discriminatory effects in my model smaller (as one would expect), but it does not make them statistically insignificant, or change their basic pattern. • Including questionable variables. I now show that my model is robust to the inclusion of the parental occupation and intended career variables, despite their flaws. As the results in the second column of Table 8.2N show, including these questionable variables does not materially alter my key results. 71 • Interaction terms. As noted earlier, Professor Card excludes several interaction terms used in my model. The most important of these is the interaction of race and disadvantage status. As I have explained previously, see supra Section 3.2, this interaction requires inclusion because Harvard takes a student’s disadvantaged status into account differently for applicants of different races. Throughout, I keep these interactions in my model.41 My original and updated models also include interactions for gender and race, and gender and intended major. Including these interactions matters less for my basic results (as shown in column 3 of Table 8.2N), but they are an important part of the evidence along other dimensions. They show, for example, that Harvard significantly penalizes African-American women relative to African-American men in the personal rating, perhaps because Harvard wishes to balance out the gender disparity among African-American applicants (female African-American applicants substantially outnumber male African-American applicants). • Interacting ratings variables. My updated model also declines to follow Professor Card’s methodology for interacting various ratings combinations. As shown in section 7.2, the groupings Professor Card uses are too fine and are based on the false premise that small sets of ratings that have similar admit rates should be pooled together. They should not be pooled, because the small sample sizes produce a phenomenon known as “over-fitting”—with many combinations guaranteed to either be rejected or admitted—and because their associated admit rates depend on other characteristics of the applicants. For example, if a particular rating group has a disproportionate number of African-American applicants, and African Americans receive large 41 I also continue to include interactions between missing SAT2 and race and missing alumni interview and race. I do this because the missing indicators effectively assign the same score or rating for all those who are missing. The interactions allow the data to assign different values based on the race of the applicant. 72 preferences, then pooling that rating group will both distort the effect of that rating and will disguise the effect of race. As shown in column 4 of Table 8.2N, significant penalties and preferences are still present when I use Professor Card’s ratings variables from his pooled analysis. But, consistent with my criticism, the effects of race are attenuated, and this is especially true for African-American and Hispanic applicants. Given that adding controls virtually always leads to an increase in the estimated preferences for African-American applicants, this suggests that Professor Card’s use of the ratings masks racial preferences. Finally, the last column of Table 8.2N shows that even if all four changes are implemented in my preferred model—including the personal rating, controlling for Professor Card’s suspect variables, removing interactions between gender and race and gender and major, and using Professor Card’s rating controls—it still results in substantial racial preferences for African-American and Hispanic applicants and significant penalties for Asian-American applicants. Table 8.2N: The racial penalties and preferences I estimate for admissions are robust to Professor Card’s key changes *=statistically significant at the 95% level. Marginal effects calculated without perfect predictions. 9 My Updated Preferred Model Yields Additional Reasons to Doubt Professor Card’s Approach In this final section, I show three additional results from my updated preferred model that underscore the weaknesses of Professor Card’s approach and demonstrate that his findings and conclusions are untenable. First, I show how the penalties against Asian-American applicants vary with how competitive the 73 applicant is. Second, I show that, contrary to Professor Card’s claims, AsianAmerican applicants are as strong as white applicants on non-academic measures. Third, there is evidence that Asian-American applicants are hurt by other preferences that Harvard employs. Namely, I show that dockets that have a higher share of Asian-American applicants have lower admit rates.42 9.1 The penalties Asian-American applicants face are substantial The estimated effects of the Asian-American penalty depend on the strength of the applicant. As I have already noted, some applicants are rated in such a way that no matter their race or their unobserved characteristics, they will be rejected. Hence for certain applicants, there is no penalty or—in the case of African-American and Hispanic applicants—no preference. Similarly, Asian-American applicants that only have very small probabilities of being admitted will see their admissions chances only slightly improve if Asian-American penalties are removed. Table 9.1 shows how the Asian-American penalty differs depending on the strength of the observed characteristics of the applicants. In particular, I use the baseline dataset to calculate deciles of the Asian-American admissions index for those who have positive predicted probabilities of admission. These are the applicants affected by the Asian-American penalty. The deciles are calculated such that 10% of these Asian-American applicants are in each decile. The first set of columns shows the results for my preferred model; the second set shows the results for my preferred model with the personal rating also included. A more detailed discussion of my updated preferred model is included in Section 3 of Appendix A 42 74 Table 9.1: The effects of the Asian-American penalty at different admissions deciles Columns 4 and 6 show the percentage increase in admissions chances if the penalty were removed. The average penalty faced across all deciles in my preferred model is almost one percentage point. Because the overall Asian-American admit rate is 5.2%, removing the penalty would increase the Asian-American admit rate by 19.2%. As would be expected, the effect varies substantially across the admissions index deciles. The biggest percentage point increases are for the most competitive applicants; these applicants see a 6.2 percentage point increase in their admissions probabilities, a 14.8% increase. The percentage point increases are smaller in the lower deciles, but as their base probability of admission is smaller, the percentage increases are higher: those in the bottom deciles only see a 0.02 percentage point penalty, but removing this penalty would increase their admission rate by 40%. 9.2 Estimates of my admissions and personal ratings models show that Asian-American applicants are strong on non-academic measures. Throughout his report, Professor Card claims that the Asian-American penalties found in my models of both admissions and the personal rating can be explained by Asian-American applicants being weaker on non-academic dimensions. As I showed using a corrected version of Professor Card’s Exhibit 10, this is not supported by the data in my original model. See Table 3.1N. It is also not supported in my updated model. As before, I construct an admissions index which assesses applicants’ strengths based on how their observed characteristics translate into a probability of 75 admission, after removing race and year effects. I then construct deciles of the admissions index, with higher deciles associated with stronger observed characteristics. Based on the admissions deciles for my updated baseline dataset and my preferred model, Asian-American applicants are the strongest group overall, with 13.1% of Asian-American applicants in the top decile. As shown in the first panel of Appendix C, Table 7.3R, this is higher than the respective numbers of white (10.5%), Hispanic/Other (5.7%), and African-American applicants (4.1%). The second panel shows that even with the personal rating included, Asian-American applicants are the strongest group. But to further test Professor Card’s claim that Asian-American applicants are actually weak on non-academic characteristics (which he claims are more likely to be in the unobservables), I create a non-academic index following Professor Card’s approach in Exhibit 10, removing those variables that are explicitly academic in nature (e.g., test scores, grades, academic ratings).43 Results from my preferred model are shown in the first panel of Appendix C, Table 7.4R. Asian-American and white applicants have the same share in the top decile (11.3%); Asian-American applicants have a greater share of the following decile, and have smaller shares in the bottom deciles. It is thus clear that on non-academic measures other than the personal rating, Asian-American applicants are at least as strong as white applicants. The second panel illustrates the bias when the personal rating is included—only then do Asian-American applicants fall behind white applicants on non-academic measures. The same point can be illustrated through the effects of Harvard’s other ratings besides personal and academic. These include the following ratings: extracurricular, athletic, teacher1, teacher2, counselor, and both of the alumni ratings. Creating the admissions index using these variables alone shows the same pattern as seen in Appendix C, Table 7.5R. Asian Americans have greater representation in the top deciles than white applicants as long as the personal rating is not included; when it Since I am using the baseline dataset, I am not vulnerable to the mistake Professor Card makes by including preferences for special recruiting categories as part of his non-academic index. 43 76 is included, they have the lowest share in the top decile. This is yet another indication that the personal rating exhibits bias. The same point is demonstrated when I examine the non-race variables that affect the personal rating. I create personal indexes in a similar manner to the admission indexes described above; higher indexes are associated with higher probabilities of receiving a high rating on the personal quality measure. In Appendix D, Table B.6.13R, I show the representation of each racial group in the resulting deciles (using the baseline dataset without academic factors.) 44 Under this approach, Asian-American applicants are actually 0.07 standard deviations stronger than white applicants—even without considering any academic factors. Looking instead at deciles of the non-academic rating components, Asian-American applicants are slightly worse than whites, but the margin is less than -0.02 standard deviations. These results make clear that any differences between white and Asian-American applicants on non-academic ratings is quite small, and—contrary to Professor Card—could not possibly explain the substantial differences in their personal ratings. 9.3 Dockets with penalized. high shares of Asian-American applicants are Harvard could also impose racial preferences or penalties through indirect channels, such as geographic preferences based on the demographics of the targeted areas. If certain dockets have high shares of Asian American applicants, and Harvard want to disfavor them, it could simply penalize these dockets. To investigate this possibility, I examined the relationship between the estimated docket by year fixed effects and the Asian-American share of domestic applicants from each docket-year combination. Using the docket-by-year effects from my preferred model and expanded dataset, I find that the larger the share of AsianAmerican applicants in that docket-year combination, the more negative the Results with the academic factors included are shown in Appendix D, Table B.6.11R. Here, too, Asian-American applicants are at least as strong as white applicants on the observed characteristics associated with higher personal ratings. 44 77 estimated docket-year fixed effect will be. More precisely, a one-standard deviation in the share of Asian-American applicants in a docket-year leads to a reduction in the admissions index of 0.14.45 This reflects the penalty Harvard imposes on dockets with a high share of Asian-American applicants, and it is more than one-third the magnitude of the Asian-American penalty Harvard already imposes on AsianAmerican applicants.46 Put differently, Harvard imposes a penalty on applicants from any docket with a high share of Asian-American applicants, and that penalty is more than a third of the direct penalty Harvard imposes on Asian-American applicants generally. To be clear, this penalty is not imposed solely on Asian-American applicants; its effects extend to any applicant from that docket regardless of race. But because this penalty is imposed only on those applicants in dockets with a high share of AsianAmerican applicants, this strongly indicates that its real target is Asian-American applicants themselves. There are two important points here: first, it appears that Harvard penalizes Asian-American applicants in indirect ways on top of the already substantial penalties it imposes on Asian-American applicants; second, my findings thus tend to understate the true magnitude of the penalties Harvard imposes on Asian-American applicants. This is the penalty using the expanded dataset; the penalty is larger in the baseline dataset. 45 46The coefficient on Asian American is -0.39 using my preferred model with the expanded dataset. Note that this is the coefficient for male Asian-American applicants who are not disadvantaged. 78 Dated: January 29, 2018 s/ Peter S. Arcidiacono Peter S. Arcidiacono 79 APPENDIX A 1 Expected errors conditional on choices In Appendix A of my original report, I defined a latent index πi where i indexes individuals and where πi = Xi γ + εi (1) The university accepts individual i if πi > 0. In the above equation, Xi represents attributes about candidate i that I observe in the data. The εi represents the unobserved characteristic of the individual. To characterize the role of unobservables, I need to be able to calculate the expected value of εi conditional on the admission decision. A mathematically equivalent model–one that leads to same the estimation procedure and model predictions– would be instead to define the payoff the university receives from accepting the applicant and rejecting the applicant respectively as u1i and u0i and where: u1i = Xi γ + u0i = 1i 0i The university admits the applicants when u1i − u0i > 0. Note that ε in equation (1) is then identical to 1i E( 1 − − 0i . 0 |y I want to recover E(ε|y = 1) where y indicates admission. This is the same as recovering = 1) but this second way is mathematically easier to derive the expectation. Under this second way of expressing the logit model, 1 and 0 are distributed Type 1 extreme value. This error distribution has the following property: E( 0 ) = γ = P r(y = 0)E( 0 |y = 0) + P r(y = 1)E( 0 |y = 1) (2) where γ is Euler’s constant. Rearranging terms yields: E( 0 |y = 1) = γ − P r(y = 0)E( 0 |y = 0) P r(y = 1) (3) The previous literature has shown that E( 0 |y = 0) can be expressed as:1 E( 0 |y = 0) = γ − ln(P r(y = 0)) (4) Substituting (4) into (3) yields: γ − P r(y = 0)[γ − ln(P r(y = 0))] P r(y = 1) P r(y = 0) ln(P r(y = 0)) = γ+ P r(y = 1) E( 0 |y = 1) = (5) (6) Recognizing that: E( 1 |y = 1) = γ − ln(P r(y = 1)) 1 (7) See, for example, V.J. Hotz and R.A. Miller “Conditional Choice Probabilities and the Estimation of Dynamic Models”, Review of Economic Studies, Vol. 60, No.3, July 1993., page 504. 1 we can form E( 1 − 0 |y E( 1 = 1) using: − 0 |y = 1) = − ln(P r(y = 1)) − P r(y = 0) ln(P r(y = 0)) P r(y = 1) (8) The individual probabilities of admission (P r(y = 1)) and rejection (P r(y = 0)) then translate directly into how strong we expect the applicant to be on unobserved characteristics conditional on being admitted. This can then be compared to the estimated admissions preference which I label µ (this is the coefficient on race in the logit model) to see how often the expected unobserved characteristic is bigger than the racial preference. 2 Probability of unobserved draws The previous section showed how to calculate the expected value of the unobservable characteristic conditional on three pieces of information: (i) the distribution of the unobserved characteristic, (ii) the probability the individual was admitted, and (iii) whether the individual was actually admitted. These three pieces of information can also be used to calculate the probability the unobserved characteristic is bigger than the racial preference for each applicant. The probability of the unobserved factor being greater than µ is given by one minus the logistic cumulative distribution function. P r( > µ) = 1 − 1 1 + exp(−µ) (9) The probability of the unobserved factor being greater than µ conditional on being admitted given observed characteristics x, where these observed characteristics x translate into an admit probability of P r(y = 1), is given by: 1− P r( > µ|y = 1, x) = min 1 1+exp(−µ) P r(y = 1) ,1 (10) The reason for the min operator is that some individuals would have x’s such that is assured that their unobservable characteristics had to be bigger than µ. 2 3. In this Appendix, I document the results of my analysis from moving to my updated models. While my results are generally robust to the changes I have made, there are three important points to keep in mind regarding the samples under consideration. First, because I have adopted Professor Card’s controls from the College Board data, information for the U and V dockets are dropped from the analysis. U and V dockets contain data on applicants who are living abroad, and the College Board data only contain information on those attending high school domestically. This works to lower the sample sizes (as well as the number of seats in the counterfactuals). Second, I now include early action applicants in my baseline dataset. This increases the number of observations in the baseline dataset, and thus correspondingly increases the number of seats in the counterfactuals. It also changes some of the descriptive statistics, for two reasons. First, early action applicants tend to be stronger than regular decision applicants, which is one reason they have higher admit rates. Second, Asian-American applicants are more likely to apply early action (once special recruiting categories are removed). This changes the baseline descriptive tables, showing that when early action applicants are included in the baseline dataset, admit rates for Asian-American applicants are sometimes higher than the admit rates for white applicants. Third, I no longer include athletes in my expanded model. Athletes by far have the highest admit rates and it is clear that the admission process for this group is very different.1 Additional time and analysis has underscored the extent to which recruited athletes are truly outliers, even within the special recruiting categories. For example, the probability of getting admitted with an academic rating of 4 is minuscule for non-athletes (.076%) and nearly a thousand times greater for athletes (70.46%). One in seven admitted athletes have an academic rating of 4 or worse; the rate for non-athletes is one in every 600. Recruited athletes also make up a much 1 3 In the next three sections I summarize the findings of my updated models, and how they relate to my original conclusions. 3.1 The role of race in the scoring of applicants for admission In my opening report, I showed that there is a significant penalty against Asian-American applicants in the scoring of applicants for admission despite the fact that Asian-American applicants are stronger on the observed characteristics than all the other races/ethnicities. I also showed that there is a significant preference given to African-American and Hispanic applicants in both the personal and overall ratings. These findings are unaltered when I use the updated sample and employ additional control variables. Tables B.6.1R through B.6.8R in Appendix D present a series of ordered logit estimates of the probability of receiving a particular rating on one of Harvard’s components. 2 For ease of tracking multiple variables, the ratings have been recoded so that higher values are associated with better ratings. Consistent with my original report, my revised regressions indicate that the personal and overall ratings are biased against Asian-American applicants, and not the product of having better unobserved characteristics (as Professor Card contends). They further show that preferences are given to AfricanAmerican and Hispanic applicants in these ratings: • As objective controls are added to the models for the academic and extracurricular ratings, the race coefficients tend to become smaller in magnitude. This suggests that if more observables were added, the effect of race would continue to diminish. • A very different pattern emerges for more subjective ratings, such as the personal and overall ratings. As additional controls are added to smaller portion of the applicant pool than legacies or early action applicants. 2Moving across the columns within a particular Harvard component rating (academic, for example) shows how the results change as more controls are added. 4 the model, the race coefficients tend to increase in magnitude, with the racial preference for African-American and Hispanic applicants getting larger and the racial penalty for Asian-American applicants becoming stronger. This pattern is inconsistent with the notion that racial preferences simply reflect the impact of unobserved characteristics. • Appendix C Table 6.1R shows how the probability of receiving a personal rating of two or better would change for each race/ethnicity if they were treated like each of the other races/ethnicities. Had Asian American applicants been treated as white applicants, the probability of receiving a two or better personal rating would increase by 4 percentage points, reflecting a 21% increase. If treated like Hispanic applicants, their probability of receiving a two or better would rise by 38%, and if treated like African-Americans, it would rise by 58%. Similar patterns exist for the overall rating. While the racial penalty Harvard imposes on Asian-American applicants is especially stark for the personal and overall rating, there is some evidence that Harvard penalizes Asian-American applicants in Harvard’s scoring of the teacher and counselor reports. In each of these models, Asian-American applicants are subjected to a penalty despite being stronger than all other racial groups on the observed characteristics associated with high ratings. See Appendix D, Tables B.6.11R and B.6.12R. My updated models thus confirm that racial preferences work throughout the admissions process, not simply at the final decision point. Professor Card concedes that the overall rating contains racial preferences; Tables B.6.3R and B.6.4R shows that the pattern of racial preferences/penalties is extremely similar for the overall and personal ratings models. This is why it is improper for Professor Card to control for the personal rating in the admissions model. 2.2 The role of race in the selection of applicants for admission My opening report showed that Harvard imposes a penalty on Asian- 5 American applicants in the selection of applicants for admission—a penalty that is separate and apart from the penalty that Harvard imposes against them in the scoring of applicants for admission. This Asian-American penalty in the selection of applicants is unaltered by the changes in the sample and control variables discussed previously. Appendix D Table B.7.1R and Table B.7.2R display estimates of a series of logit models of admission for the updated baseline and expanded dataset, respectively. Model 5 is the preferred specification, as it includes all controls other than the personal rating. The changes in the race coefficients as additional controls are included mimic the patterns seen in the personal rating. The race coefficients for African-American and Hispanic applicants become larger and positive as additional applicant characteristics are included. This occurs because African-American and Hispanic applicants are weaker on the observed characteristics that predict admission, meaning that the racial preference has to grow to explain the admissions decisions. Table 7.2R in Appendix C puts the admissions penalty against AsianAmerican applicants in context. It shows that Asian-American admit rates would increase by 19% if Asian Americans were treated as whites in the preferred model. The preferences for African-American and Hispanic applicants are even larger in magnitude than the Asian-American penalty. In the preferred model, admit rates for Asian American applicants in the baseline dataset would increase almost three-fold if they were treated like Hispanic applicants, and over five-fold if they were treated like AfricanAmerican applicants. Similar to the ratings models, my updated models assess whether the penalties Asian Americans suffer could reasonably be attributed to unobservable characteristics. Indexes can be constructed net of year and race that give the strength of the applicant based on the controls, effectively aggregating all the measures Harvard uses and weighting them the same way the data indicates that Harvard weighs them in their admissions decisions. These indexes are not well defined for those who have 6 characteristics that perfectly predict rejection and admission, so I focus on deciles of the admissions indexes where those who have characteristics that guaranteed rejection were assigned to the bottom decile, and characteristics that guaranteed admission to the top decile.3 These deciles then give the strength of the application based on how the characteristics of the applicant translate into admissions probabilities net of race/ethnicity. Appendix C Table 7.3R shows the share of each racial/ethnic group that is in each of the deciles for the preferred model, as well as a variation that includes the overall and personal ratings for the baseline and expanded models, respectively. These deciles show that, based on observables, AsianAmerican applicants are substantially less likely to be in the bottom five deciles and are substantially more likely to be in the top deciles. For the preferred model, the share of Asian-American applicants rises steadily with every decile; the opposite trend occurs for African-American applicants. And even when the personal rating is added, Asian Americans are still overrepresented at the top of the distribution. Selection on unobservables would have to be working in the opposite direction of selection on observables to explain the negative Asian-American coefficient. 2.3 How the removal of preferences would impact the admitted class As in my opening report, I evaluate how the removal of penalties and preferences for particular racial groups would affect admissions rates, fixing the overall admissions rate in a particular year for a particular dataset (baseline or expanded) to match with the data. For example, turning off the penalty against Asian-American applicants would increase the number of Asian Americans admitted. If no other adjustments were made, then Harvard’s admitted class would be larger than Harvard intended. The constant term in the logit admissions models is thus lowered for all groups until the model-predicted overall probability of admission is the same as the Note that I include perfect predictions here to show the total strength of the applicant pool by race. Including the perfect predictions in this instance is appropriate because we are looking at the full distribution of the effects. 3 7 probability of admission in the data.4 Results for these models are given in Appendix C, Tables 8.1R and 8.2R. Similar to my previous report, I find that the removal of racial penalties and preferences has a profound impact on the admitted class: • Using the updated baseline dataset and my preferred model, removing the Asian-American penalty in admissions results in increased AsianAmerican admits in all years. The model predicts 261 more AsianAmerican admits over this six-year period, a 13% increase. • Removing preferences for African-American and Hispanic applicants (but keeping the penalty against Asian American applicants) results in even larger gains, with 537 more Asian-American admits over the period, an increase of more than 26%. And removing all racial preferences and penalties—treating everyone as though they were white—raises the number of Asian Americans by 799, a 40% increase. The second panel of Table 8.1R in Appendix C looks at the share of the admitted class by race/ethnicity under the different policies. Again, the results are striking: • In the preferred model, removing the penalty against Asian Americans increases their share of the admitted class by at least 2.3 percentage points in all years, with the largest change in 2018 of 4.7 percentage points. • Treating all applicants in a manner similar to whites has dramatic effects: the share of admits who are Asian American increases by more than 10 percentage points (a 40% increase in share), while the share of admits who are African American falls by over 11 percentage points (a 72% decrease in share). • The fact that the racial composition of Harvard’s admitted class depends so strongly on racial preferences indicates that race is a determinative factor in admissions decisions. My updated model again finds that the effect of removing racial preferences To perform this exercise, I re-estimate the preferred model (Model 5) and the model that includes the personal rating (Model 6) but now allowing for race times year effects. Including these interactions ensures that, in each year, the admissions rate for each racial/ethnic group matches the actual admit rate for that group. 4 8 on African Americans and Hispanics admit rates depends on disadvantaged status. The estimates show that Harvard has a preference for disadvantaged applicants, but that preference is smaller for Hispanics, who already receive a large preference, and nonexistent for African Americans. With the removal of racial preferences, disadvantaged African Americans and Hispanics receive the same preference as other disadvantaged applicants. As shown in Appendix C Table 8.3R (using Models 5 and 6), this preference is smaller than the preference with racial preferences, but nonetheless substantial: • Disadvantaged African-American applicants see a 52% fall in the number of admitted students in the preferred model. • For non-disadvantaged African-American applicants, the decline is much larger at 80%. This occurs because the added boost that nondisadvantaged African-American applicants receive because of their race is significantly smaller than the added boost disadvantaged African-American applicants receive because of their race. As a result, the share of African-American admits who are disadvantaged shifts from 29% to 50%. • Similar patterns, though not quite as stark, occur for Hispanic applicants: the drop in admits is 60% for non-disadvantaged students and below 36% for disadvantaged students. Using the expanded dataset and incorporating athletes5 brings additional insight into how all of the preferences Harvard employs (race, legacy, athlete, etc.) work against Asian-American applicants. Appendix C Table 8.2R shows how the admitted class would change as racial and other preferences are eliminated. Focusing on the preferred model and the scenario where race, legacy, and athlete preferences are eliminated: 5 Recall that my updated expanded model no longer includes athletes. Hence all counterfactuals that do not involve athletes treat the admissions decisions for athletes as unchanged. When I do counterfactuals with athletes included, I replace their athletic rating and extracurricular rating with 2s and then use the model to predict their admissions probabilities. 9 • The total number of Asian-American admits would rise by over 1,200 over the six-year period, or more than 50%. • In contrast, African-American admits would fall by 939 over the same six-year window, a decline of over 68%.6 6 The number of white admits would increase by only 3% in this scenario. This occurs because while the removal of racial preferences tends to favor white applicants, the removal of legacy and athlete preferences harms white applicants. 10 APPENDIX B Table B.3.1N: Mother Occupations by Class Year Table B.3.2N Father Occupations by Class Year Table B.4.1N: Intended Career by Class Year APPENDIX C Table 1.1R: Single-race African American admit rates and all other domestic admit rates by admissions cycle 2019 African American Non-African American Difference IPEDS Admit 0.06059 0.06084 -0.00025 2018 African American Non-African American Difference 0.06585 0.06521 0.00064 2017 African American Non-African American Difference 0.06399 0.06424 -0.00025 Table 2.1R Admission Decisions for the Baseline and Expanded Samples by Race/Ethnicity Race/Ethnicity Rejected Panel 1: Baseline Sample White 84.5 African American 87.6 Hispanic 86.7 Asian 83.5 Total 84.8 Admission Status Waitlist Rejected Admit 10.6 4.8 7.2 11.4 9.7 Observations 4.9 7.6* 6.2* 5.1 5.5 57,582 15,664 17,970 40,415 142,728 Panel 2: Baseline Sample without Early Applicants White 85.0 10.8 African American 88.6 5.0 Hispanic 87.4 7.4 Asian 84.8 11.3 Total 85.7 9.8 4.2 6.5* 5.2* 3.9 4.5 52,370 14,230 16,468 36,705 129,646 Panel 3: Expanded Sample White 81.1 African American 87.1 Hispanic 85.7 Asian 82.7 Total 82.8 6.8* 7.9* 6.7* 5.7 6.6 61,657 15,959 18,322 41,142 148,769 12.1 5 7.5 11.6 10.6 A * indicates statistically different at the 5% level * Constructed using results from basicFreqs.do Table 3.1R: Application summary statistics by race, baseline sample Admitted Disadvantaged SAT1 math (z-score) SAT1 verbal (z-score) SAT2 avg (z-score) High school GPA (z-score) Academic index (z-score) Number of AP tests taken Average score of AP tests N Reject 0.00 5.94 0.12 (0.82) 0.31 (0.76) -0.01 (0.86) 0.17 (0.86) 0.16 (0.80) 4.08 (3.91) 4.39 (0.59) 54,768 *Constructed using results from sumStatsTablesPoolRej.do *Subset of the results in Table B.3.1 White Admit 100.00 14.61 0.56 (0.50) 0.72 (0.43) 0.58 (0.50) 0.50 (0.52) 0.76 (0.38) 5.91 (3.85) 4.74 (0.34) 2,814 Total 4.89 6.36 0.15 (0.81) 0.33 (0.75) 0.03 (0.85) 0.18 (0.85) 0.19 (0.79) 4.16 (3.93) 4.41 (0.58) 57,582 Reject 0.00 29.27 -1.17 (1.07) -0.77 (1.07) -1.24 (1.13) -0.51 (1.18) -1.23 (1.12) 2.11 (3.14) 3.78 (0.78) 14,477 African American Admit 100.00 28.48 0.14 (0.67) 0.41 (0.56) 0.15 (0.62) 0.31 (0.76) 0.33 (0.52) 5.06 (3.85) 4.51 (0.42) 1,187 Total 7.58 29.21 -1.07 (1.10) -0.68 (1.08) -1.10 (1.17) -0.45 (1.17) -1.11 (1.16) 2.33 (3.29) 3.88 (0.78) 15,664 Reject 0.00 23.47 -0.69 (1.05) -0.45 (1.05) -0.61 (1.04) -0.08 (0.97) -0.63 (1.01) 3.52 (3.82) 3.96 (0.75) 16,863 Hispanic Admit 100.00 37.40 0.28 (0.64) 0.44 (0.59) 0.41 (0.54) 0.45 (0.62) 0.50 (0.46) 6.20 (3.83) 4.55 (0.47) 1,107 Total 6.16 24.33 -0.63 (1.05) -0.39 (1.05) -0.53 (1.04) -0.04 (0.97) -0.56 (1.02) 3.68 (3.87) 4.02 (0.75) 17,970 Reject 0.00 10.26 0.41 (0.73) 0.31 (0.80) 0.32 (0.82) 0.21 (0.82) 0.39 (0.78) 5.60 (4.07) 4.48 (0.56) 38,343 Asian American Admit Total 100.00 5.13 21.86 10.85 0.77 0.43 (0.37) (0.72) 0.74 0.33 (0.41) (0.79) 0.81 0.35 (0.38) (0.81) 0.52 0.22 (0.47) (0.81) 0.91 0.42 (0.32) (0.77) 7.50 5.68 (3.38) (4.06) 4.82 4.50 (0.28) (0.55) 2,072 40,415 Table 3.2R: Application summary statistics for special treatment by race, expanded sample Admitted Early action applicant Legacy Faculty child Staff child Dean / Director's List N Reject 0.00 8.94 3.44 0.03 0.13 1.60 57,481 White Admit 100.00 31.59 24.50 0.79 1.03 15.92 4,176 *Constructed using results from sumStatsTablesPoolRej.do *Subset of the results from Table B.3.2 Total 6.77 10.47 4.87 0.08 0.19 2.57 61,657 Reject 0.00 8.13 1.12 0.00 0.05 0.38 14,691 African American Admit Total 100.00 7.95 24.76 9.45 5.21 1.45 0.00 0.00 0.16 0.06 2.13 0.52 1,268 15,959 Reject 0.00 7.65 0.93 0.01 0.05 0.46 17,093 Hispanic Admit 100.00 25.79 7.16 0.16 0.49 4.56 1,229 Total 6.71 8.87 1.35 0.02 0.08 0.73 18,322 Reject 0.00 8.22 0.77 0.00 0.11 0.37 38,800 Asian American Admit Total 100.00 5.69 33.60 9.67 6.92 1.12 0.56 0.03 1.11 0.17 5.64 0.67 2,342 41,142 Table 4.1R: Application rating summary statistics by race, baseline sample Reject Academic rating <3=3-, 3, or 3+ >3+ Extracurricular rating <3=3-, 3, or 3+ >3+ Athletic rating <3=3-, 3, or 3+ >3+ Personal rating <3=3-, 3, or 3+ >3+ Teacher 1 rating <3=3-, 3, or 3+ >3+ Teacher 2 rating <3=3-, 3, or 3+ >3+ School counselor rating <3=3-, 3, or 3+ >3+ Alumni Personal rating <3=3-, 3, or 3+ >3+ Alumni Overall rating <3=3-, 3, or 3+ >3+ N White Admit Total Reject Reject Hispanic Admit Total Reject Asian American Admit Total 10.39 46.56 43.05 0.04 11.19 88.77 9.88 44.83 45.29 54.91 40.02 5.07 0.08 40.52 59.39 50.75 40.06 9.19 37.77 48.68 13.55 0.00 34.60 65.40 35.44 47.81 16.74 8.42 33.22 58.36 0.00 5.60 94.40 7.99 31.80 60.21 3.74 74.39 21.87 0.71 26.18 73.11 3.59 72.03 24.38 8.02 79.40 12.58 0.76 47.22 52.02 7.47 76.96 15.57 5.98 79.79 14.23 1.26 42.10 56.64 5.69 77.46 16.85 2.02 72.43 25.54 0.19 21.53 78.28 1.93 69.82 28.25 33.58 53.93 12.49 33.00 45.64 21.37 33.56 53.52 12.92 43.65 50.04 6.30 36.49 49.10 14.41 43.11 49.97 6.92 43.36 49.57 7.07 40.96 43.53 15.52 43.22 49.20 7.59 46.98 48.31 4.71 48.22 44.50 7.28 47.04 48.11 4.84 0.45 81.49 18.06 0.00 16.24 83.76 0.43 78.30 21.27 0.50 85.02 14.47 0.00 25.61 74.39 0.47 80.52 19.01 0.51 84.70 14.79 0.00 22.13 77.87 0.48 80.85 18.68 0.51 84.86 14.63 0.00 26.74 73.26 0.48 81.88 17.64 0.57 70.57 28.86 0.00 22.60 77.40 0.54 68.16 31.31 1.11 83.60 15.28 0.00 40.44 59.56 1.02 79.97 19.01 0.90 78.61 20.49 0.00 36.50 63.50 0.84 75.80 23.35 0.51 70.39 29.10 0.00 25.35 74.65 0.48 68.03 31.49 0.48 69.24 30.28 0.00 22.97 77.03 0.45 66.57 32.98 0.80 82.42 16.78 0.00 41.76 58.24 0.72 78.23 21.06 0.82 77.56 21.61 0.00 33.05 66.95 0.76 74.05 25.20 0.51 69.99 29.50 0.05 24.59 75.36 0.48 67.36 32.16 0.63 75.44 23.93 0.00 23.25 76.75 0.60 72.78 26.62 1.99 86.42 11.60 0.00 41.73 58.27 1.82 82.65 15.54 1.28 83.59 15.13 0.00 40.73 59.27 1.20 80.74 18.07 0.64 75.85 23.51 0.00 26.30 73.70 0.61 73.22 26.17 7.40 31.43 61.17 0.51 5.89 93.60 6.98 29.89 63.13 10.55 35.82 53.62 1.12 9.13 89.75 9.62 33.19 57.19 10.24 35.55 54.21 0.28 6.54 93.19 9.41 33.11 57.48 8.26 31.55 60.19 0.34 6.38 93.28 7.77 29.98 62.25 18.60 37.47 43.92 54,768 0.84 10.45 88.71 2,814 17.52 35.83 46.65 57,582 41.29 35.45 23.26 14,477 2.25 22.39 75.37 1,187 37.34 34.13 28.53 15,664 34.14 36.81 29.05 16,863 1.84 16.41 81.75 1,107 31.37 35.06 33.57 17,970 16.90 34.70 48.39 38,343 0.44 7.57 91.99 2,072 15.87 32.99 51.14 40,415 *Constructed using results from sumStatsSubRatTablesPoolRej.do African American Admit Total Table 4.2R: Admission shares by race and overall rating, baseline sample White Admit Score Share Panel 1: Baseline Sample <3 0.02 3 1.94 3+ 8.89 2 65.48 1 100.00 Panel 2: Expanded Sample <3 0.10 3 2.94 3+ 11.32 2 71.34 1 100.00 Pop. Share African American Admit Pop. Share Share Hispanic Admit Pop. Share Share Asian American Admit Pop. Share Share 42.62 39.52 13.42 4.39 0.04 0.02 5.75 22.50 84.63 100.00 65.07 21.41 8.23 5.23 0.05 0.01 4.06 19.14 79.48 100.00 57.43 28.78 9.92 3.85 0.03 0.01 1.79 7.96 65.51 100.00 37.82 42.77 14.58 4.76 0.08 40.93 39.60 13.91 5.51 0.06 0.02 6.01 23.15 84.82 100.00 64.65 21.49 8.31 5.49 0.06 0.02 4.35 20.41 80.64 100.00 56.70 28.98 10.16 4.12 0.04 0.01 2.05 8.64 67.65 100.00 37.48 42.63 14.69 5.12 0.08 Table 5.1R: Number and Share of Applicants by Race/Ethnicity and Academic Index Decile, Baseline Sample Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Total Number of Applicants in Each Decile African Whites American Hispanic 2,822 5,921 3,583 4,404 3,600 3,755 6,073 2,291 2,926 6,359 1,285 2,182 7,658 897 1,719 5,924 508 1,077 7,053 445 949 6,478 326 820 5,717 196 539 4,963 132 380 57,451 15,601 17,930 Asian American 1,511 2,045 2,644 3,020 3,874 3,614 4,527 5,316 6,532 7,225 Total 14,593 14,658 15,014 13,865 15,426 12,110 14,145 14,253 14,303 13,989 40,308 142,356 Share of Applicants in each Decile African Whites American Hispanic 4.91 37.95 19.98 7.67 23.08 20.94 10.57 14.68 16.32 11.07 8.24 12.17 13.33 5.75 9.59 10.31 3.26 6.01 12.28 2.85 5.29 11.28 2.09 4.57 9.95 1.26 3.01 8.64 0.85 2.12 Asian American 3.75 5.07 6.56 7.49 9.61 8.97 11.23 13.19 16.21 17.92 Total 10.25 10.3 10.55 9.74 10.84 8.51 9.94 10.01 10.05 9.83 Table 5.2R: Admit Rates by Race/Ethnicity and Academic Index Decile, Baseline Sample Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Average Whites 0.00% 0.39% 0.56% 1.82% 2.57% 4.20% 4.79% 7.53% 10.77% 15.27% African American 0.03% 1.03% 5.19% 12.76% 22.41% 29.72% 41.12% 44.48% 54.59% 56.06% Hispanic 0.00% 0.32% 1.95% 5.50% 9.13% 13.65% 17.28% 22.93% 26.16% 31.32% Asian American 0.00% 0.20% 0.64% 0.86% 1.86% 2.49% 3.98% 5.12% 7.55% 12.69% Total 0.01% 0.53% 1.65% 3.29% 4.40% 5.64% 6.61% 8.22% 10.40% 14.58% 4.90% 7.58% 6.16% 5.14% 5.46% Table 5.3R: Share of admits of each race/ethnicity if equally drawn from different academic index deciles African Asian Whites American Hispanic American Actual Share of Admitted Class 37.61 15.81 14.90 24.86 Randomly sampling from: Top 9 deciles Top 8 deciles Top 7 deciles Top 6 deciles Top 5 deciles Top 4 deciles Top 3 deciles Top 2 deciles Top decile 42.76 44.41 45.01 44.87 43.80 42.71 40.33 37.75 35.48 7.58 5.38 3.86 2.97 2.34 1.94 1.54 1.16 0.94 11.23 9.36 7.82 6.51 5.47 4.74 4.09 3.25 2.72 30.37 32.49 34.77 36.91 39.56 41.63 44.83 48.63 51.65 Table 5.4R: Share Receiving a Two or Higher on the Academic and Extracurricular Ratings by Race/Ethnicity and Academic Index Decile, Baseline Sample Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Average Academic Rating African Whites American 0.11% 0.02% 0.41% 0.08% 1.91% 0.96% 9.14% 6.07% 26.26% 23.08% 50.19% 48.43% 68.37% 68.54% 82.73% 80.37% 93.30% 93.37% 97.16% 94.70% 45.32% 9.18% Hispanic 0.03% 0.05% 0.68% 4.45% 17.04% 43.83% 64.28% 79.63% 91.47% 95.26% Asian American 0.00% 0.54% 1.36% 7.98% 26.36% 51.08% 71.46% 86.16% 95.12% 98.08% Total 0.05% 0.25% 1.44% 7.88% 25.24% 49.98% 69.39% 83.98% 94.20% 97.63% 16.75% 60.21% 42.30% Extracurricular Rating African Whites American 11.41% 9.02% 16.35% 13.75% 20.14% 18.86% 22.02% 23.27% 23.83% 22.85% 25.08% 26.38% 26.64% 27.42% 27.31% 27.91% 30.45% 32.65% 33.04% 38.64% 24.38% 15.56% Hispanic 9.27% 12.73% 15.86% 18.74% 20.65% 23.31% 27.61% 24.63% 28.94% 29.21% Asian American 12.97% 15.99% 18.57% 21.59% 23.67% 25.51% 28.34% 29.78% 34.92% 37.98% Total 10.09% 14.75% 18.92% 21.61% 23.61% 25.36% 27.39% 28.11% 32.76% 35.96% 16.84% 28.27% 23.73% Table 5.5R: Share Receiving a Two or Higher on School Support Measures by Race/Ethnicity and Academic Index Decile, Baseline Sample Teacher 1 Academic Index Decile Teacher 2 1 2 3 4 5 6 7 8 9 10 Whites 7.76% 13.42% 19.00% 23.87% 26.39% 32.41% 34.64% 39.72% 44.92% 50.17% African American 7.75% 13.97% 19.38% 25.06% 29.65% 36.42% 40.22% 46.63% 47.45% 55.30% Average 30.46% 17.15% Hispanic 8.85% 13.87% 20.03% 23.60% 30.19% 31.94% 35.62% 37.68% 43.60% 49.47% Asian American 7.41% 14.18% 16.98% 21.03% 23.00% 26.59% 30.22% 33.09% 39.73% 46.64% Total 8.06% 13.69% 18.92% 23.27% 26.20% 30.65% 33.37% 36.72% 42.24% 48.02% 21.60% 30.84% 27.90% Counselor Whites 6.20% 10.24% 15.46% 21.21% 23.31% 27.53% 31.04% 36.66% 41.47% 47.11% African American 5.46% 11.50% 16.98% 22.41% 31.55% 35.43% 35.06% 39.88% 42.86% 50.76% 27.16% 14.83% Hispanic 6.42% 11.00% 17.77% 20.81% 25.54% 28.97% 32.77% 37.32% 38.59% 49.74% Asian American 6.55% 11.69% 13.80% 18.01% 20.26% 24.29% 26.18% 29.67% 36.15% 41.90% Total 6.02% 11.00% 15.95% 20.54% 23.30% 27.09% 29.49% 33.61% 38.66% 44.10% Whites 4.64% 8.99% 14.49% 18.49% 22.06% 25.59% 29.24% 34.39% 39.16% 44.63% African American 4.88% 10.86% 16.72% 20.31% 26.42% 32.87% 35.73% 38.04% 43.88% 49.24% Hispanic 5.72% 10.15% 14.83% 17.32% 21.06% 25.26% 30.35% 34.15% 34.32% 45.00% Asian American 5.76% 9.19% 12.25% 14.93% 17.84% 22.61% 24.96% 27.69% 33.88% 38.34% Total 5.20% 9.84% 14.46% 17.69% 21.12% 24.89% 28.01% 31.40% 36.32% 40.87% 18.86% 27.44% 24.77% 25.29% 13.86% 16.49% 25.16% 22.79% Table 5.6R: Share Receiving a Two or Higher on the Personal Rating and Alumni Interview Personal Rating by Race/Ethnicity and Academic Index Decile, Baseline Sample Personal Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Whites 8.11% 12.58% 16.25% 18.62% 20.40% 22.72% 22.59% 26.10% 28.23% 29.62% African American 9.49% 15.75% 23.35% 28.95% 33.89% 35.04% 40.00% 39.57% 40.31% 46.97% Average 21.29% 19.01% Hispanic 8.48% 13.16% 17.77% 20.39% 25.60% 28.41% 30.03% 32.20% 30.24% 34.21% Asian American 8.01% 12.91% 13.46% 14.24% 15.69% 16.46% 18.11% 17.93% 20.87% 22.20% Total 8.81% 13.57% 17.16% 18.91% 20.56% 21.69% 22.01% 23.20% 24.74% 25.46% 18.69% 17.65% 19.52% Alumni Personal African Whites American 26.33% 30.96% 33.72% 39.83% 39.77% 46.84% 44.27% 55.56% 48.43% 59.98% 51.84% 62.20% 54.08% 69.89% 58.20% 67.48% 62.20% 70.92% 64.98% 73.48% 49.79% 42.79% Hispanic 26.29% 33.42% 38.59% 43.86% 50.32% 54.50% 56.90% 62.44% 62.89% 71.05% Asian American 28.13% 32.03% 36.35% 40.66% 44.24% 46.96% 51.93% 53.78% 57.46% 63.61% Total 28.25% 34.89% 40.06% 44.64% 48.28% 51.18% 54.07% 56.77% 60.18% 64.49% 41.25% 50.21% 48.09% *Note that those who do not have an alumni interview are coded as not having received a 2 or higher on the alumni overall rating Table 5.7R: Share Receving a Two or Higher on Overall Rating and Alumni Interviewer Overall Rating by Race/Ethnicity and Academic Index Decile, Baseline Sample Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Average Final Reader Overall Rating African Whites American Hispanic 0.00% 0.00% 0.00% 0.18% 0.47% 0.08% 0.26% 2.05% 0.68% 0.69% 7.16% 2.15% 1.57% 15.50% 4.54% 2.95% 23.82% 8.82% 4.15% 30.79% 12.01% 7.63% 37.42% 15.61% 10.97% 45.41% 19.67% 15.64% 46.97% 27.37% 4.44% 5.29% 3.88% Asian American 0.00% 0.15% 0.23% 0.40% 1.34% 1.80% 3.11% 4.61% 7.59% 12.93% Total 0.00% 0.22% 0.63% 1.54% 2.71% 4.05% 5.22% 7.39% 10.12% 14.70% 4.85% 4.60% Alumini Interviewer Overall Rating African Whites American Hispanic 7.26% 7.31% 7.14% 13.35% 14.89% 11.82% 19.38% 23.88% 19.07% 26.23% 33.46% 24.84% 32.68% 43.48% 35.08% 38.07% 51.38% 39.65% 42.62% 56.18% 44.89% 49.26% 59.20% 50.98% 56.69% 61.22% 59.18% 63.13% 66.67% 64.47% 36.50% 20.82% Asian American 7.35% 12.22% 17.51% 23.11% 29.71% 36.08% 42.37% 46.95% 54.16% 63.10% Total 7.34% 13.37% 19.86% 26.12% 32.87% 38.34% 43.30% 48.50% 55.91% 63.28% 40.91% 34.59% 23.64% *Note that those who do not have an alumni interview are coded as not having received a 2 or higher on the alumni overall rating Table 6.1R: Probability of Receiving a 2 or Higher on Personal Rating for own race/ethnicity and counterfactual race/ethnicity, preferred model Race/Ethnicity Own Race if White if African American if Hispanic if Asian American Panel 1: Baseline sample White African American Hispanic Asian American 0.214 0.193 0.192 0.178 0.283 0.152 0.168 0.216 0.217 0.283 0.245 0.174 0.176 0.126 0.138 0.247 Panel 2: Expanded sample White African American Hispanic Asian American 0.227 0.197 0.196 0.181 *created using ologitpersonal.do 0.299 0.155 0.172 0.219 0.223 0.286 0.257 0.177 0.250 0.190 0.129 0.143 Table 6.2R: Probability of receiving each overall rating for own race/ethnicity and counterfactual race/ethnicity, preferred model, baseline sample White African American Hispanic Asian American <3 3 3+ >3+ <3 3 3+ >3+ <3 3 3+ >3+ <3 3 3+ >3+ Own Race 0.425 0.392 0.136 0.047 0.649 0.210 0.087 0.054 0.565 0.289 0.103 0.043 0.375 0.425 0.148 0.052 if White *calculated using gologitComponentsExpIndices.do 0.750 0.188 0.049 0.013 0.658 0.254 0.069 0.019 0.366 0.418 0.161 0.056 if African American 0.265 0.346 0.206 0.183 0.526 0.274 0.122 0.079 0.221 0.338 0.231 0.210 if Hispanic 0.310 0.396 0.190 0.105 0.678 0.218 0.074 0.030 0.259 0.400 0.219 0.122 if Asian American 0.434 0.397 0.125 0.043 0.755 0.188 0.045 0.012 0.664 0.255 0.063 0.017 Table 7.1R: Probability of admission for an Asian American if treated like other races/ethnicities when base probability is 0.25 Probability of admission Baseline sample Expanded Sample Counterfactual group Preferred Model +Overall and Personal Preferred Model +Overall and Personal Asian/male/no disadvantage Black 0.958 0.957 0.947 0.944 Hispanic 0.790 0.779 0.768 0.754 White 0.347 0.317 0.330 0.301 Asian/female/no disadvantage Black 0.943 0.944 0.931 0.927 Hispanic 0.771 0.761 0.751 0.734 White 0.297 0.275 0.293 0.261 Asian/male/disadvantaged Black 0.805 0.805 0.759 0.759 Hispanic 0.646 0.629 0.604 0.596 White 0.315 0.286 0.296 0.271 Asian/female/disadvantaged Black 0.748 0.756 0.703 0.702 Hispanic 0.620 0.605 0.581 0.571 White 0.268 0.247 0.261 0.234 Asian/male/no disadvantage White legacy 0.807 0.816 White double legacy 0.893 0.903 Table 7.2R: Average Probability of admission for Asian American applicants if treated like other races/ethnicities Probability of admission Baseline sample Preferred Model +Personal Data Model White Preferences African American Preferences Hispanic Preferences 0.052 0.062 0.271 0.152 0.052 0.057 0.243 0.135 Expanded Sample Preferred Model 0.052 0.061 0.267 0.150 +Personal 0.052 0.057 0.238 0.133 Table 7.3R: Share of each race/ethnicity in each admissions index decile, baseline sample Preferred Model Admissions Decile 5 or lower 6 7 8 9 10 White 0.458 0.111 0.112 0.107 0.107 0.105 African American 0.786 0.051 0.042 0.039 0.041 0.041 Hispanic 0.695 0.069 0.060 0.059 0.059 0.057 Asian American 0.381 0.113 0.120 0.128 0.128 0.131 Hispanic 0.697 0.070 0.058 0.055 0.056 0.065 Asian American 0.377 0.116 0.123 0.132 0.127 0.124 +Personal Ratings Admissions Decile 5 or lower 6 7 8 9 10 White 0.458 0.109 0.112 0.106 0.107 0.107 * created using admissionsLogitsIndices.do. African American 0.786 0.047 0.038 0.039 0.044 0.046 Table 7.4R: Share of each race/ethnicity in each admissions, non-academic ratings component score index decile, baseline sample Admissions Decile 5 or lower 6 7 8 9 10 White 0.457 0.105 0.109 0.108 0.108 0.113 Preferred Model African American 0.684 0.072 0.073 0.065 0.057 0.050 Hispanic 0.642 0.074 0.080 0.073 0.068 0.063 Asian American 0.439 0.113 0.106 0.111 0.117 0.113 Admissions Decile 5 or lower 6 7 8 9 10 White 0.460 0.105 0.107 0.106 0.107 0.114 +Personal Ratings African American 0.666 0.072 0.063 0.062 0.068 0.068 Hispanic 0.627 0.078 0.072 0.071 0.075 0.077 Asian American 0.447 0.111 0.115 0.116 0.110 0.101 * created using admissionsLogitsIndicesRatings.do. Table 7.5R: Share of each race/ethnicity in each admissions non-academic component score index decile, baseline sample Admissions Decile 5 or lower 6 7 8 9 10 White 0.480 0.106 0.104 0.104 0.103 0.103 Preferred Model African American 0.599 0.080 0.082 0.080 0.075 0.084 Hispanic 0.569 0.088 0.085 0.081 0.084 0.094 Asian American 0.466 0.104 0.107 0.109 0.110 0.104 Admissions Decile 5 or lower 6 7 8 9 10 White 0.479 0.104 0.104 0.104 0.105 0.105 +Personal Ratings African American 0.584 0.076 0.080 0.079 0.086 0.095 Hispanic 0.555 0.086 0.085 0.085 0.087 0.101 Asian American 0.477 0.109 0.108 0.108 0.103 0.095 * created using admissionsLogitsIndices.do. Table 8.1R: Admissions levels and shares by race/ethnicity under different admissions policies, baseline sample 2014 2015 2016 Preferred Model 2017 2018 2019 Total 2014 2015 2016 Panel 1: Changes in Admissions Levels Model Asian No Asian-American penalty American No African American or Hispanic preferences No racial preferences 369 421 455 513 356 405 458 508 346 390 418 457 315 350 399 430 300 355 399 458 327 354 422 445 2013 2274 2550 2812 369 403 452 489 356 383 451 480 346 364 415 431 315 335 394 413 Model No Asian-American penalty No African American or Hispanic preferences No racial preferences 198 190 58 55 219 211 66 61 178 172 62 57 180 175 56 52 195 185 52 47 193 188 55 52 1163 1121 349 324 198 193 63 60 219 215 74 71 178 176 64 62 Model No Asian-American penalty No African American or Hispanic preferences No racial preferences 189 180 104 98 215 206 113 105 182 174 105 97 194 187 96 90 217 205 105 96 191 185 101 95 1188 1137 624 581 189 183 108 103 215 210 122 118 Model No Asian-American penalty No African American or Hispanic preferences No racial preferences 572 538 703 660 526 497 670 624 460 436 558 517 401 383 512 477 390 362 525 479 355 342 464 438 2704 2557 3433 3195 572 550 699 671 Panel 2: Changes in Admission Shares Model Asian No Asian-American penalty American No African American or Hispanic preferences No racial preferences 0.271 0.309 0.334 0.376 0.264 0.300 0.340 0.378 0.272 0.306 0.329 0.359 0.265 0.295 0.335 0.362 0.256 0.302 0.340 0.390 0.284 0.307 0.366 0.387 0.269 0.303 0.340 0.375 Model No Asian-American penalty No African American or Hispanic preferences No racial preferences 0.145 0.140 0.043 0.040 0.163 0.157 0.049 0.045 0.140 0.135 0.049 0.045 0.151 0.147 0.047 0.044 0.166 0.158 0.044 0.040 0.168 0.163 0.048 0.045 Model No Asian-American penalty No African American or Hispanic preferences No racial preferences 0.139 0.132 0.077 0.072 0.160 0.153 0.084 0.078 0.143 0.137 0.083 0.076 0.163 0.157 0.081 0.076 0.185 0.174 0.089 0.082 Model No Asian-American penalty No African American or Hispanic preferences No racial preferences 0.420 0.395 0.516 0.484 0.390 0.369 0.497 0.463 0.362 0.343 0.439 0.406 0.337 0.322 0.430 0.401 0.332 0.308 0.448 0.408 African American Hispanic White African American Hispanic White Include Personal Rating 2017 2018 2019 Total 300 332 392 428 327 337 418 426 2013 2154 2522 2666 180 177 62 60 195 189 58 55 193 191 58 57 1163 1141 379 365 182 179 109 105 194 190 99 95 217 210 116 110 191 189 107 104 1188 1161 660 636 526 510 659 635 460 450 556 539 401 391 507 489 390 373 515 489 355 350 459 448 2704 2624 3397 3271 0.271 0.296 0.331 0.359 0.264 0.285 0.335 0.356 0.272 0.286 0.326 0.339 0.265 0.281 0.331 0.347 0.256 0.283 0.334 0.364 0.284 0.292 0.363 0.370 0.269 0.287 0.336 0.356 0.155 0.150 0.047 0.043 0.145 0.142 0.046 0.044 0.163 0.159 0.055 0.053 0.140 0.138 0.050 0.049 0.151 0.149 0.053 0.051 0.166 0.161 0.050 0.047 0.168 0.166 0.050 0.049 0.155 0.152 0.051 0.049 0.166 0.161 0.087 0.083 0.158 0.152 0.083 0.078 0.139 0.134 0.079 0.076 0.160 0.156 0.091 0.087 0.143 0.141 0.085 0.083 0.163 0.160 0.083 0.080 0.185 0.179 0.099 0.094 0.166 0.164 0.093 0.090 0.158 0.155 0.088 0.085 0.308 0.297 0.403 0.380 0.361 0.341 0.458 0.426 0.420 0.404 0.513 0.492 0.390 0.378 0.489 0.471 0.362 0.354 0.437 0.423 0.337 0.329 0.427 0.412 0.332 0.318 0.439 0.417 0.308 0.304 0.398 0.389 0.361 0.350 0.453 0.436 Table 8.2R: Admissions levels and shares by race/ethnicity under different admissions policies, expanded sample 2014 2015 2016 Preferred Model 2017 2018 2019 Total 2014 2015 2016 2019 Total Panel 1: Changes in Admissions Levels Model Asian No Asian-American penalty American No African and Hispanic preference No racial preferences No legacy preferences No athlete preferences No race/legacy/athlete 403 457 483 543 426 445 634 404 448 502 550 427 443 642 406 442 472 503 421 435 565 379 402 462 481 395 411 549 364 419 457 517 378 394 589 402 422 496 514 411 441 586 2358 2591 2870 3108 2458 2569 3564 403 439 479 520 425 444 606 404 426 496 520 428 443 607 406 417 468 478 420 434 536 379 388 457 465 394 410 527 364 396 450 486 377 394 552 402 403 493 492 412 441 560 2358 2470 2843 2961 2456 2566 3389 Model No Asian-American penalty No African and Hispanic preference No racial preferences No legacy preferences No athlete preferences No race/legacy/athlete 231 224 84 81 242 232 74 246 239 86 82 256 251 78 201 196 82 79 210 201 73 228 225 93 91 236 219 71 226 217 77 73 234 229 62 235 232 88 85 245 233 70 1367 1332 511 491 1423 1366 428 231 226 89 86 242 233 80 246 243 94 92 257 251 89 201 200 85 84 210 201 78 228 227 101 100 235 219 82 226 221 86 83 234 228 73 235 235 90 90 245 233 76 1367 1351 544 534 1423 1365 478 Model No Asian-American penalty No African and Hispanic preference No racial preferences No legacy preferences No athlete preferences No race/legacy/athlete 212 203 128 122 223 226 131 246 238 136 130 259 258 136 207 201 134 127 217 220 135 222 218 118 114 232 239 123 245 234 128 120 256 270 136 233 229 132 128 242 248 132 1365 1323 776 741 1428 1462 792 212 206 132 127 222 227 138 246 242 144 141 259 258 150 207 205 137 135 216 220 144 222 220 120 118 231 240 128 245 239 138 133 255 270 151 233 233 138 137 242 247 143 1365 1345 808 791 1424 1461 853 Model No Asian-American penalty No African and Hispanic preference No racial preferences No legacy preferences No athlete preferences No race/legacy/athlete 945 909 1088 1043 903 895 960 865 838 1025 985 821 818 902 805 785 908 872 774 757 799 748 735 874 849 716 704 784 760 729 911 865 728 700 784 679 668 806 783 655 625 718 4802 4664 5613 5397 4598 4499 4947 945 921 1084 1054 904 895 972 865 851 1016 997 820 818 915 805 798 906 894 775 758 825 748 743 870 861 719 704 800 760 742 900 874 731 700 796 679 678 801 796 655 625 732 4802 4734 5577 5476 4605 4500 5041 Panel 2: Changes in Admission Shares Model Asian No Asian-American penalty American No African and Hispanic preference No racial preferences No legacy preferences No athlete preferences No race/legacy/athlete 0.217 0.247 0.260 0.293 0.230 0.240 0.342 0.220 0.244 0.274 0.300 0.233 0.242 0.350 0.229 0.249 0.266 0.284 0.237 0.245 0.319 0.219 0.232 0.267 0.278 0.228 0.238 0.317 0.213 0.246 0.268 0.303 0.221 0.231 0.345 0.238 0.251 0.294 0.305 0.244 0.261 0.347 0.223 0.245 0.271 0.294 0.232 0.243 0.337 0.217 0.237 0.259 0.280 0.229 0.240 0.327 0.220 0.232 0.270 0.284 0.233 0.242 0.331 0.229 0.236 0.264 0.270 0.237 0.245 0.302 0.219 0.224 0.264 0.269 0.228 0.237 0.305 0.213 0.232 0.264 0.285 0.221 0.231 0.324 0.238 0.239 0.292 0.292 0.244 0.261 0.332 0.223 0.233 0.269 0.280 0.232 0.243 0.320 Model No Asian-American penalty No African and Hispanic preference No racial preferences No legacy preferences No athlete preferences No race/legacy/athlete 0.125 0.121 0.046 0.044 0.130 0.125 0.040 0.134 0.131 0.047 0.045 0.140 0.137 0.042 0.113 0.111 0.047 0.045 0.118 0.114 0.041 0.132 0.130 0.054 0.052 0.136 0.127 0.041 0.133 0.127 0.045 0.043 0.138 0.134 0.036 0.139 0.137 0.052 0.051 0.145 0.138 0.042 0.129 0.126 0.048 0.046 0.134 0.129 0.040 0.125 0.122 0.048 0.046 0.130 0.125 0.043 0.134 0.132 0.051 0.050 0.140 0.137 0.049 0.113 0.113 0.048 0.047 0.118 0.114 0.044 0.132 0.131 0.058 0.058 0.136 0.127 0.047 0.133 0.130 0.050 0.049 0.137 0.134 0.043 0.139 0.139 0.054 0.053 0.145 0.138 0.045 0.129 0.128 0.051 0.051 0.134 0.129 0.045 Model No Asian-American penalty No African and Hispanic preference No racial preferences No legacy preferences No athlete preferences No race/legacy/athlete 0.114 0.110 0.069 0.066 0.120 0.122 0.071 0.134 0.130 0.074 0.071 0.141 0.141 0.074 0.117 0.113 0.075 0.072 0.122 0.124 0.076 0.128 0.126 0.068 0.066 0.134 0.138 0.071 0.144 0.137 0.075 0.070 0.150 0.159 0.079 0.138 0.136 0.078 0.076 0.144 0.147 0.078 0.129 0.125 0.073 0.070 0.135 0.138 0.075 0.114 0.111 0.071 0.069 0.120 0.122 0.074 0.134 0.132 0.079 0.077 0.141 0.141 0.082 0.117 0.116 0.077 0.076 0.122 0.124 0.081 0.128 0.127 0.069 0.068 0.134 0.139 0.074 0.144 0.140 0.081 0.078 0.149 0.158 0.088 0.138 0.138 0.082 0.081 0.143 0.147 0.085 0.129 0.127 0.076 0.075 0.135 0.138 0.081 Model No Asian-American penalty No African and Hispanic preference No racial preferences No legacy preferences No athlete preferences No race/legacy/athlete 0.510 0.490 0.587 0.563 0.487 0.483 0.518 0.472 0.457 0.559 0.537 0.448 0.446 0.492 0.454 0.443 0.512 0.492 0.437 0.427 0.451 0.432 0.425 0.505 0.491 0.414 0.407 0.453 0.446 0.428 0.534 0.507 0.427 0.410 0.460 0.403 0.396 0.478 0.465 0.388 0.371 0.426 0.454 0.441 0.531 0.510 0.435 0.425 0.468 0.510 0.497 0.585 0.568 0.488 0.483 0.524 0.472 0.464 0.554 0.544 0.447 0.446 0.499 0.454 0.451 0.511 0.504 0.437 0.428 0.466 0.432 0.430 0.503 0.498 0.416 0.407 0.463 0.446 0.435 0.528 0.513 0.429 0.411 0.467 0.403 0.402 0.475 0.472 0.389 0.371 0.434 0.454 0.447 0.527 0.518 0.435 0.425 0.476 African American Hispanic White African American Hispanic White Include Personal Rating 2017 2018 Table 8.3R: The Effects of Removing Racial/Ethnic Preferences and Penalties by Race/Ethnicity and Disadvantaged Status, baseline sample Advantaged admits Asian Americans Model Removal Racial Preferences African Americans Model Removal Racial Preferences Hispanics Model Removal Racial Preferences Whites Model Removal Racial Preferences Preferred Model (Model 5) Disadvantaged Share Admits Disadvantaged Add Personal Rating (Model 6) Advantaged Disadvantaged Share admits Admits Disadvantaged 1564 449 0.223 1564 449 0.223 2255 556 0.198 2130 536 0.201 829 334 0.287 829 334 0.287 163 161 0.497 194 171 0.469 754 434 0.365 754 434 0.365 302 280 0.481 341 295 0.464 2303 401 0.148 2303 401 0.148 2742 453 0.142 2809 462 0.141 APPENDIX D Table A.1: Coding decisions made for irregular ratings and their frequencies in the expanded sample Imputed Final Original Rating Reader Score Frequency 122 1 2 212 2 1 213 2 1 222 2 70 223 2 35 232 2225 233 2179 253 21 322 3+ 180 323 3+ 427 332 3 35 333 3 73 334 3 3 342 3 1 343 3 8 433 4 1 554 5 1 604 4 2 622 2 1 623 26 632 3+ 8 633 3 210 634 33 643 352 644 4 45 645 5 1 653 33 654 4 2 655 4 4 Observations 1580 Table A.2R: Applicants and Admit Rate by Preferred Group Not Athlete Athlete Number of Applicants Admit Rate 165,353 0.060 1374 0.860 Not Legacy Legacy 162,083 4644 0.059 0.336 Not Child of Faculty or Staff Child of Faculty or Staff 166,406 321 0.066 0.467 Not Dean and Director's Interest List Dean and Director's Interest List 164,226 2501 0.061 0.422 *created using actionpools2.do 2014 2015 2016 2017 2018 2019 Applicants 24,376 28,260 25,696 23,604 23,390 24,757 Table A.3R: Applicants, Admits, and Admit Rate by Year and Regular vs. Early Regular Action Early Action Admits Admit rate Applicants Admits 1,986 0.081 0 0 1,923 0.068 0 0 1,012 0.039 3,582 825 870 0.037 4,111 947 817 0.035 3,958 971 790 0.032 4,993 991 Admit Rate 0.230 0.230 0.245 0.198 2014 2015 2016 2017 2018 2019 Table A.4R: Applicants, Admits, and Admit Rate by Year, Regular vs. Early, and Special Circumstances Regular Action Early Action Regular Applicant Special Circumstances Regular Applicant Special Circumstances Applicants Admits Admit Rate Applicants Admits Admit Rate Applicants Admits Admit Rate Applicants Admits Admit Rate 23,176 1,471 0.063 1,200 515 0.429 0 0 0 0 27,016 1,408 0.052 1,244 515 0.414 0 0 0 0 24,968 857 0.034 728 155 0.213 2,982 458 0.154 600 367 0.612 22,963 754 0.033 641 116 0.181 3,448 487 0.141 663 460 0.694 22,799 709 0.031 591 108 0.183 3,272 520 0.159 686 451 0.657 24,134 690 0.029 623 100 0.161 4,238 524 0.124 755 467 0.619 *Sample excludes foreign applicants and transfers. Applications Harvard labels as withdrawals, incompletes, or departed are excluded. Ony first time applications are included. *Results based on actionPools.do Table A.5R: Sample Cuts Non-transfer, non-foreign sample size Withdraws, Incompletes, Departed Repeat Applicant Missing a Rating Profile (AEAP) Academic or Personal Rating>5 SAT Math or SAT Verbal Missing Academic Index Missing Athlete Additional Baseline Cuts Legacy Staff or Faculty Child Dean/Director Preference *Results based on sampeCuts.do Admits Removed 0 0 0 64 0 5 59 1,179 Applicants Removed 0 4,512 601 3,636 192 7,079 5,708 1,343 Remaining Obs. 171,840 167,328 166,727 163,091 162,899 155,820 150,112 148,769 1,479 113 449 4,371 248 1,422 144,398 144,150 142,728 Table A.6: Harvard's Assignment of Race/Ethnicity under the Old Methodology Race/Ethnicity Member in Which Group A A,B A,B,P A,B,P,W A,B,W A,P A,P,W A,W B B,P B,P,W B,W N N,A N,A,B N,A,B,P N,A,B,P,W N,A,B,W N,A,P N,A,P,W N,A,W N,B N,B,P N,B,P,W N,B,W N,P N,P,W N,W P P,W W Total White 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 75,492 75,493 African American 3 526 6 5 139 0 0 0 19,378 33 12 1,685 0 0 24 5 2 33 0 0 0 486 5 1 369 0 0 0 0 0 2 22,714 Hispanic 1 0 0 0 0 0 0 0 0 0 0 0 492 0 0 0 0 0 0 0 0 0 0 0 0 0 0 429 0 0 13,331 14,253 Asian American Native American 55,331 0 0 0 0 160 106 5,446 0 0 0 0 0 32 0 0 0 0 4 7 133 0 0 0 0 0 0 0 0 1 2 61,222 0 0 0 0 0 0 0 0 0 0 0 0 620 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1,108 0 0 5 1,735 Hawaiian 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 4 0 244 132 1 384 Missing 1 0 0 0 0 0 0 3 3 0 0 2 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 4 0 0 5 20 Total 55,336 526 6 5 139 160 106 5,449 19,381 33 12 1,687 1,112 33 24 5 2 33 4 7 134 488 5 1 369 3 4 1,542 244 133 88,838 175,821 Table A.7R: Descriptive Statistics by Admit Status for Baseline and Expanded Sample Admitted Female Disadvantaged First-generation college Early action applicant Legacy Faculty child Staff child Dean's / Director's List Mother highest ed: no college Mother highest ed: BA degree Mother highest ed: MA degree Mother highest ed: PhD/JD/MD degree Mother highest ed: Missing Father highest ed: no college Father highest ed: BA degree Father highest ed: MA degree Father highest ed: PhD/JD/MD degree Father highest ed: Missing Application read by 3rd reader Missing alumni rating Waiver Financial Aid SAT1 math (z-score) SAT1 verbal (z-score) SAT2 avg (z-score) Never took SAT2 Standardized high school GPA (z-score) Academic index (z-score) Academic index percentile Number of AP tests taken Average score of AP tests N Reject 0.00 49.23 12.03 8.77 Baseline Sample Admit Total 100.00 5.45 48.73 49.21 22.11 12.58 8.54 8.76 29.42 32.86 24.29 10.07 0.03 27.52 24.09 24.93 19.36 0.04 11.84 23.00 16.93 78.42 -0.03 (1.00) 0.09 (0.94) -0.08 (1.01) 12.62 0.07 (0.93) -0.02 (1.01) 0.49 (0.29) 4.28 (4.02) 4.34 (0.65) 134,944 *Constructed using results from sumStatsTablesPoolRej.do 25.86 28.75 28.49 13.73 0.03 26.23 19.84 25.41 24.74 0.04 95.57 2.12 19.75 79.07 0.51 (0.57) 0.64 (0.50) 0.55 (0.55) 1.63 0.47 (0.57) 0.69 (0.46) 0.73 (0.21) 6.27 (3.81) 4.70 (0.38) 7,784 29.23 32.63 24.52 10.27 0.03 27.45 23.86 24.96 19.66 0.04 16.41 21.86 17.08 78.45 0.00 (0.99) 0.12 (0.93) -0.04 (1.00) 12.02 0.09 (0.92) 0.01 (1.00) 0.50 (0.29) 4.37 (4.04) 4.36 (0.64) 142,728 Reject 0.00 49.20 11.74 8.56 8.60 2.08 0.01 0.11 0.95 28.84 32.74 24.48 10.62 0.03 26.90 23.96 25.03 20.07 0.04 12.88 22.62 16.53 77.41 -0.03 (1.00) 0.10 (0.93) -0.07 (1.00) 12.42 0.06 (0.94) 0.48 (0.29) -0.02 (1.00) 4.25 (4.02) 4.34 (0.64) 138,944 Expanded Sample Admit Total 100.00 6.60 48.66 49.17 17.82 12.14 7.01 8.46 31.07 10.09 15.05 2.94 0.61 0.05 0.87 0.16 10.08 1.56 21.48 28.35 27.89 32.42 29.20 24.79 18.58 11.14 0.03 0.03 21.47 26.54 18.95 23.63 27.02 25.16 29.31 20.68 0.03 0.04 94.36 18.26 2.33 21.28 15.93 16.49 67.99 76.78 0.50 0.00 (0.57) (0.98) 0.63 0.13 (0.49) (0.92) 0.54 -0.03 (0.55) (0.99) 1.42 11.69 0.39 0.08 (0.62) (0.92) 0.71 0.50 (0.22) (0.29) 0.66 0.02 (0.47) (0.99) 5.89 4.35 (3.84) (4.03) 4.71 4.37 (0.37) (0.63) 9,825 148,769 Table A.8R: Harvard Rankings by Admit Status for Baseline and Expanded Sample Reject Academic rating <3=3-, 3, or 3+ >3+ Extracurricular rating <3=3-, 3, or 3+ >3+ Athletic rating <3=3-, 3, or 3+ >3+ Personal rating <3=3-, 3, or 3+ >3+ Teacher 1 rating <3=3-, 3, or 3+ >3+ Teacher 2 rating <3=3-, 3, or 3+ >3+ School counselor rating <3=3-, 3, or 3+ >3+ Alumni Personal rating <3=3-, 3, or 3+ >3+ Alumni Overall rating <3=3-, 3, or 3+ >3+ N Baseline Sample Admit Total Reject Expanded Sample Admit Total 18.16 41.86 39.98 0.04 17.86 82.10 17.17 40.55 42.28 17.93 42.09 39.98 0.16 18.65 81.19 16.76 40.54 42.70 3.96 74.93 21.11 0.67 30.30 69.04 3.78 72.50 23.72 3.92 74.84 21.24 0.64 33.08 66.28 3.70 72.08 24.21 40.17 51.12 8.71 39.07 45.61 15.32 40.11 50.82 9.07 39.80 51.23 8.97 36.21 46.18 17.61 39.56 50.90 9.54 0.49 83.36 16.14 0.00 22.11 77.89 0.46 80.02 19.51 0.49 83.10 16.41 0.02 24.14 75.84 0.46 79.21 20.33 0.64 72.86 26.49 0.00 28.68 71.32 0.61 70.45 28.93 0.64 72.85 26.51 0.00 30.86 69.14 0.60 70.08 29.33 0.55 71.77 27.68 0.01 28.20 71.79 0.52 69.39 30.09 0.55 71.72 27.73 0.03 30.54 69.43 0.52 69.00 30.48 0.86 77.75 21.39 0.00 29.93 70.07 0.81 75.14 24.04 0.84 77.69 21.47 0.00 31.52 68.48 0.78 74.64 24.57 8.34 32.39 59.28 0.50 6.78 92.73 7.91 30.99 61.10 8.26 32.25 59.49 0.67 8.17 91.16 7.76 30.66 61.58 22.08 36.25 41.67 134,944 1.08 12.35 86.57 7,784 20.93 34.95 44.12 142,728 21.81 36.28 41.91 138,944 1.40 14.00 84.60 9,825 20.46 34.81 44.73 148,769 * Constructed using results from sumStatsSubRatTablesPoolRej.do Table B.1.1R: Single-race African American admit rates and all other domestic admit rates by admissions cycle IPEDS 2019 Non-African American African American Difference Admit 0.06084 0.06059 0.00025 Admit Total 1,677 176 1,853 2018 Non-African American African American Difference 0.06521 0.06585 -0.00064 1,657 177 1,834 2017 Non-African American African American Difference 0.06424 0.06399 0.00025 1,665 172 1,837 2016 Non-African American African American Difference 0.06763 0.05556 0.01207 1,713 147 1,860 2015 Non-African American African American Difference 0.06832 0.06519 0.00313 1,780 188 1,968 2014 Non-African American African American Difference 0.07937 0.07439 0.00498 1,839 172 2,011 Table B.1.2R: Admit rates for single-race African Americans and other domestic applicants by day, 2017 Date 3/1/13 3/2/13 3/3/13 3/4/13 3/5/13 3/6/13 3/7/13 3/8/13 3/9/13 3/10/13 3/11/13 3/12/13 3/13/13 3/15/13 3/16/13 3/17/13 3/18/13 3/19/13 3/20/13 3/21/13 3/22/13 3/23/13 3/24/13 3/25/13 3/26/13 3/27/13 3/28/13 3/29/13 3/30/13 3/31/13 4/1/13 4/2/13 4/3/13 4/4/13 4/5/13 4/6/13 4/7/13 4/8/13 4/9/13 4/10/13 4/11/13 4/12/13 4/13/13 4/14/13 4/15/13 4/16/13 4/17/13 4/18/13 4/19/13 4/20/13 4/21/13 4/22/13 4/23/13 4/24/13 4/25/13 4/26/13 4/27/13 4/28/13 4/29/13 4/30/13 5/1/13 5/2/13 5/3/13 5/6/13 5/7/13 5/8/13 5/9/13 5/10/13 5/13/13 5/14/13 5/15/13 5/16/13 5/17/13 5/20/13 5/21/13 5/22/13 5/23/13 5/24/13 5/27/13 5/28/13 5/29/13 5/30/13 5/31/13 6/2/13 6/3/13 6/4/13 6/5/13 6/6/13 6/7/13 6/10/13 6/11/13 6/12/13 6/13/13 6/14/13 6/16/13 6/17/13 6/18/13 6/20/13 6/24/13 6/25/13 6/26/13 6/27/13 6/28/13 6/29/13 6/30/13 7/1/13 7/3/13 7/5/13 7/6/13 7/8/13 7/9/13 7/11/13 7/22/13 7/24/13 7/25/13 7/26/13 7/30/13 8/5/13 8/6/13 8/16/13 8/19/13 8/26/13 8/29/13 Single-race Single-race African American All other African American admits domestic admits applicants 136 845 1552 137 2688 1556 137 2688 1556 140 2688 1597 143 2688 1618 146 2688 1653 151 2688 1679 161 2688 1719 162 2688 1727 162 2688 1727 166 2688 1758 173 2688 1779 177 2688 1785 171 2688 1702 171 2688 1702 171 2688 1702 167 2688 1586 172 2688 1650 172 2688 1649 172 2688 1650 172 2688 1650 172 2688 1650 172 2688 1650 172 2688 1650 172 2688 1649 172 2688 1649 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1648 172 2688 1649 172 2688 1650 172 2688 1650 172 2688 1650 172 2688 1650 172 2688 1652 172 2688 1652 172 2688 1654 172 2688 1657 172 2688 1658 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1649 172 2688 1662 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1659 172 2688 1662 172 2688 1662 172 2688 1662 172 2688 1662 172 2688 1663 172 2688 1663 172 2688 1663 172 2688 1663 172 2688 1663 172 2688 1663 172 2688 1663 172 2688 1663 172 2688 1666 172 2688 1666 172 2688 1666 172 2688 1666 172 2688 1666 172 2688 1666 172 2688 1666 172 2688 1666 172 2688 1666 172 2688 1665 All other domestic applicants 8774 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 25918 Single-race African American Single-race admit rate-Other African American Other domestic domestic admit admit rate admit rate rate 0.16095 0.17689 -0.0159395 0.05097 0.06004 -0.0090682 0.05097 0.06004 -0.0090682 0.05208 0.06162 -0.0095341 0.05320 0.06243 -0.0092283 0.05432 0.06378 -0.0094626 0.05618 0.06478 -0.0086056 0.05990 0.06632 -0.0064287 0.06027 0.06663 -0.0063654 0.06027 0.06663 -0.0063654 0.06176 0.06783 -0.0060734 0.06436 0.06864 -0.0042794 0.06585 0.06887 -0.0030228 0.06362 0.06567 -0.0020526 0.06362 0.06567 -0.0020526 0.06362 0.06567 -0.0020526 0.06213 0.06119 0.0009350 0.06399 0.06366 0.0003258 0.06399 0.06362 0.0003644 0.06399 0.06366 0.0003258 0.06399 0.06366 0.0003258 0.06399 0.06366 0.0003258 0.06399 0.06366 0.0003258 0.06399 0.06366 0.0003258 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06359 0.0004029 0.06399 0.06362 0.0003644 0.06399 0.06366 0.0003258 0.06399 0.06366 0.0003258 0.06399 0.06366 0.0003258 0.06399 0.06366 0.0003258 0.06399 0.06374 0.0002486 0.06399 0.06374 0.0002486 0.06399 0.06382 0.0001714 0.06399 0.06393 0.0000557 0.06399 0.06397 0.0000171 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06362 0.0003644 0.06399 0.06413 -0.0001372 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06401 -0.0000215 0.06399 0.06413 -0.0001372 0.06399 0.06413 -0.0001372 0.06399 0.06413 -0.0001372 0.06399 0.06413 -0.0001372 0.06399 0.06416 -0.0001758 0.06399 0.06416 -0.0001758 0.06399 0.06416 -0.0001758 0.06399 0.06416 -0.0001758 0.06399 0.06416 -0.0001758 0.06399 0.06416 -0.0001758 0.06399 0.06416 -0.0001758 0.06399 0.06416 -0.0001758 0.06399 0.06428 -0.0002916 0.06399 0.06428 -0.0002916 0.06399 0.06428 -0.0002916 0.06399 0.06428 -0.0002916 0.06399 0.06428 -0.0002916 0.06399 0.06428 -0.0002916 0.06399 0.06428 -0.0002916 0.06399 0.06428 -0.0002916 0.06399 0.06428 -0.0002916 0.06399 0.06424 -0.0002530 *Bolded rows indicate the difference between the two admit rates is minimized given the number of applicants of each race and the total number of admits Table B.1.3R: Admit rates for single-race African Americans and other domestic applicants by day, 2018 Date 3/1/14 3/2/14 3/3/14 3/4/14 3/5/14 3/6/14 3/7/14 3/8/14 3/9/14 3/10/14 3/11/14 3/12/14 3/13/14 3/14/14 3/15/14 3/16/14 3/17/14 3/18/14 3/19/14 3/20/14 3/21/14 3/22/14 3/23/14 3/24/14 3/25/14 3/26/14 3/27/14 3/28/14 3/29/14 3/30/14 3/31/14 4/1/14 4/2/14 4/3/14 4/4/14 4/5/14 4/6/14 4/7/14 4/8/14 4/9/14 4/10/14 4/11/14 4/12/14 4/13/14 4/14/14 4/15/14 4/16/14 4/17/14 4/18/14 4/19/14 4/20/14 4/21/14 4/22/14 4/23/14 4/24/14 4/25/14 4/26/14 4/27/14 4/28/14 4/29/14 4/30/14 5/1/14 5/2/14 5/3/14 5/4/14 5/5/14 5/6/14 5/7/14 5/8/14 5/9/14 5/12/14 5/13/14 5/14/14 5/15/14 5/16/14 5/17/14 5/18/14 5/19/14 5/20/14 5/21/14 5/22/14 5/23/14 5/24/14 5/26/14 5/27/14 5/28/14 5/29/14 5/30/14 6/2/14 6/3/14 6/4/14 6/5/14 6/9/14 6/10/14 6/11/14 6/12/14 6/13/14 6/15/14 6/16/14 6/17/14 6/18/14 6/19/14 6/20/14 6/21/14 6/22/14 6/23/14 6/24/14 6/25/14 6/26/14 6/29/14 7/1/14 7/2/14 7/3/14 7/4/14 7/7/14 7/8/14 7/9/14 7/11/14 7/17/14 7/23/14 7/28/14 8/15/14 8/24/14 8/25/14 8/26/14 8/27/14 Single-race Single-race African American All other African American admits domestic admits applicants 159 1600 826 159 1600 2688 160 1608 2688 164 1632 2688 168 1658 2688 170 1684 2688 172 1707 2688 172 1711 2688 172 1711 2688 177 1745 2688 187 1792 2688 188 1808 2688 188 1824 2688 190 1826 2688 190 1826 2688 190 1826 2688 179 1659 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1633 2688 178 1632 2688 178 1631 2688 178 1631 2688 178 1634 2688 178 1640 2688 178 1643 2688 178 1644 2688 178 1644 2688 178 1644 2688 178 1644 2688 178 1644 2688 178 1644 2688 178 1644 2688 178 1644 2688 178 1644 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1654 2688 178 1658 2688 178 1658 2688 178 1658 2688 178 1658 2688 178 1658 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1660 2688 178 1660 2688 178 1660 2688 178 1660 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1659 2688 178 1658 2688 178 1658 2688 178 1658 2688 178 1658 2688 178 1657 2688 178 1657 2688 178 1657 2688 177 1657 2688 All other domestic applicants 7931 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 25411 Single-race African American Single-race admit rate-Other African American Other domestic domestic admit admit rate admit rate rate 0.19249 0.20174 -0.0092461 0.05915 0.06296 -0.0038131 0.05952 0.06328 -0.0037559 0.06101 0.06422 -0.0032123 0.06250 0.06525 -0.0027473 0.06324 0.06627 -0.0030265 0.06399 0.06718 -0.0031875 0.06399 0.06733 -0.0033449 0.06399 0.06733 -0.0033449 0.06585 0.06867 -0.0028228 0.06957 0.07052 -0.0009522 0.06994 0.07115 -0.0012098 0.06994 0.07178 -0.0018395 0.07068 0.07186 -0.0011741 0.07068 0.07186 -0.0011741 0.07068 0.07186 -0.0011741 0.06659 0.06529 0.0013056 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06426 0.0019567 0.06622 0.06422 0.0019961 0.06622 0.06418 0.0020354 0.06622 0.06418 0.0020354 0.06622 0.06430 0.0019174 0.06622 0.06454 0.0016813 0.06622 0.06466 0.0015632 0.06622 0.06470 0.0015238 0.06622 0.06470 0.0015238 0.06622 0.06470 0.0015238 0.06622 0.06470 0.0015238 0.06622 0.06470 0.0015238 0.06622 0.06470 0.0015238 0.06622 0.06470 0.0015238 0.06622 0.06470 0.0015238 0.06622 0.06470 0.0015238 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06509 0.0011303 0.06622 0.06525 0.0009729 0.06622 0.06525 0.0009729 0.06622 0.06525 0.0009729 0.06622 0.06525 0.0009729 0.06622 0.06525 0.0009729 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06533 0.0008942 0.06622 0.06533 0.0008942 0.06622 0.06533 0.0008942 0.06622 0.06533 0.0008942 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06529 0.0009336 0.06622 0.06525 0.0009729 0.06622 0.06525 0.0009729 0.06622 0.06525 0.0009729 0.06622 0.06525 0.0009729 0.06622 0.06521 0.0010123 0.06622 0.06521 0.0010123 0.06622 0.06521 0.0010123 0.06585 0.06521 0.0006402 Table B.1.4R: Admit rates for single-race African Americans and other domestic applicants by day, 2019 Date 3/1/15 3/2/15 3/4/15 3/6/15 3/8/15 3/9/15 3/10/15 3/11/15 3/12/15 3/13/15 3/14/15 3/16/15 3/17/15 3/18/15 3/19/15 3/20/15 3/23/15 3/24/15 3/25/15 3/26/15 3/30/15 3/31/15 4/1/15 4/2/15 4/3/15 4/5/15 4/6/15 4/7/15 4/8/15 4/9/15 4/10/15 4/12/15 4/13/15 4/14/15 4/15/15 4/16/15 4/17/15 4/20/15 4/21/15 4/22/15 4/24/15 4/27/15 4/28/15 4/29/15 4/30/15 5/1/15 5/2/15 5/4/15 5/5/15 5/7/15 5/11/15 5/12/15 5/14/15 5/15/15 5/18/15 5/19/15 5/20/15 5/21/15 5/22/15 5/26/15 5/27/15 5/28/15 6/1/15 6/2/15 6/3/15 6/4/15 6/5/15 6/8/15 6/9/15 6/10/15 6/11/15 6/15/15 6/16/15 6/17/15 6/19/15 6/22/15 6/23/15 6/30/15 7/2/15 7/6/15 7/7/15 7/8/15 7/10/15 7/13/15 8/3/15 8/17/15 8/19/15 8/24/15 Single-race Single-race African American All other African American admits domestic admits applicants 153 1521 2899 153 1520 2899 153 1519 2899 153 1519 2899 153 1519 2899 153 1529 2899 153 1530 2899 153 1530 2899 153 1530 2899 153 1531 2899 192 1785 2904 192 1784 2904 192 1784 2904 177 1651 2905 171 1581 2905 176 1600 2905 176 1600 2905 176 1600 2905 176 1600 2905 176 1600 2905 176 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1600 2905 177 1599 2905 177 1599 2905 177 1599 2905 177 1599 2905 176 1597 2905 176 1597 2905 176 1597 2905 176 1597 2905 176 1597 2905 176 1597 2905 176 1597 2905 176 1597 2905 176 1597 2905 176 1597 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1645 2905 176 1663 2905 176 1662 2905 176 1662 2905 176 1662 2905 176 1667 2905 176 1668 2905 176 1668 2905 176 1668 2905 176 1668 2905 176 1668 2905 176 1668 2905 176 1668 2905 176 1668 2905 176 1678 2905 176 1678 2905 176 1678 2905 176 1678 2905 176 1678 2905 176 1678 2905 176 1678 2905 176 1678 2905 176 1678 2905 176 1678 2905 176 1678 2905 All other domestic applicants 27520 27520 27520 27520 27520 27530 27531 27531 27531 27532 27556 27556 27556 27565 27565 27565 27565 27565 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 27566 Single-race African American Single-race admit rate-Other African American Other domestic domestic admit admit rate admit rate rate 0.05278 0.05527 -0.0024921 0.05278 0.05523 -0.0024557 0.05278 0.05520 -0.0024194 0.05278 0.05520 -0.0024194 0.05278 0.05520 -0.0024194 0.05278 0.05554 -0.0027626 0.05278 0.05557 -0.0027969 0.05278 0.05557 -0.0027969 0.05278 0.05557 -0.0027969 0.05278 0.05561 -0.0028312 0.06612 0.06478 0.0013385 0.06612 0.06474 0.0013748 0.06612 0.06474 0.0013748 0.06093 0.05989 0.0010346 0.05886 0.05736 0.0015087 0.06059 0.05804 0.0025406 0.06059 0.05804 0.0025406 0.06059 0.05804 0.0025406 0.06059 0.05804 0.0025427 0.06059 0.05804 0.0025427 0.06059 0.05804 0.0025427 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05804 0.0028869 0.06093 0.05801 0.0029232 0.06093 0.05801 0.0029232 0.06093 0.05801 0.0029232 0.06093 0.05801 0.0029232 0.06059 0.05793 0.0026515 0.06059 0.05793 0.0026515 0.06059 0.05793 0.0026515 0.06059 0.05793 0.0026515 0.06059 0.05793 0.0026515 0.06059 0.05793 0.0026515 0.06059 0.05793 0.0026515 0.06059 0.05793 0.0026515 0.06059 0.05793 0.0026515 0.06059 0.05793 0.0026515 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.05967 0.0009102 0.06059 0.06033 0.0002573 0.06059 0.06029 0.0002935 0.06059 0.06029 0.0002935 0.06059 0.06029 0.0002935 0.06059 0.06047 0.0001121 0.06059 0.06051 0.0000759 0.06059 0.06051 0.0000759 0.06059 0.06051 0.0000759 0.06059 0.06051 0.0000759 0.06059 0.06051 0.0000759 0.06059 0.06051 0.0000759 0.06059 0.06051 0.0000759 0.06059 0.06051 0.0000759 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 0.06059 0.06087 -0.0002869 Table B.1.5R: Admit rates for single-race African Americans and other domestic applicants by day, 2014 (pre-IPEDS) Date 3/1/10 3/2/10 3/3/10 3/4/10 3/5/10 3/6/10 3/8/10 3/9/10 3/10/10 3/11/10 3/12/10 3/13/10 3/15/10 3/16/10 3/17/10 3/18/10 3/19/10 3/20/10 3/21/10 3/22/10 3/23/10 3/24/10 3/25/10 3/26/10 3/29/10 3/30/10 3/31/10 4/1/10 4/6/10 4/7/10 4/12/10 4/14/10 4/15/10 4/28/10 4/29/10 4/30/10 5/3/10 5/4/10 5/5/10 5/6/10 5/7/10 5/10/10 5/11/10 5/12/10 5/13/10 5/14/10 5/17/10 5/18/10 5/19/10 5/26/10 6/1/10 6/2/10 6/3/10 6/4/10 6/8/10 6/18/10 6/25/10 6/28/10 6/29/10 7/1/10 7/22/10 7/30/10 8/2/10 8/4/10 8/9/10 8/11/10 8/17/10 Single-race Single-race African American All other African American admits domestic admits applicants 126 1526 2209 127 1597 2311 129 1618 2311 129 1637 2311 143 1680 2311 143 1681 2311 153 1761 2311 156 1800 2311 158 1812 2311 171 1855 2311 180 1874 2311 184 1894 2311 183 1910 2311 183 1911 2311 183 1926 2312 168 1803 2312 164 1718 2312 169 1755 2312 169 1755 2312 169 1755 2312 170 1755 2312 170 1755 2312 170 1755 2312 170 1755 2312 170 1755 2312 170 1753 2312 170 1753 2312 170 1753 2312 170 1753 2312 170 1754 2312 170 1754 2312 170 1754 2312 170 1754 2312 170 1754 2312 170 1754 2312 170 1754 2312 170 1754 2312 170 1756 2312 170 1754 2312 170 1754 2312 170 1754 2312 170 1754 2312 170 1773 2312 171 1801 2312 171 1801 2312 171 1801 2312 171 1801 2312 171 1801 2312 171 1801 2312 171 1802 2312 171 1803 2312 171 1803 2312 171 1821 2312 171 1821 2312 171 1821 2312 171 1821 2312 172 1836 2312 172 1835 2312 172 1835 2312 172 1835 2312 172 1837 2312 172 1838 2312 172 1839 2312 172 1838 2312 172 1840 2312 172 1840 2312 172 1839 2312 All other domestic applicants 22509 23169 23169 23169 23169 23169 23169 23169 23169 23169 23169 23169 23169 23169 23168 23168 23168 23168 23168 23168 23168 23168 23168 23168 23168 23167 23167 23168 23169 23170 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23170 23170 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 23171 Single-race African American Single-race admit rate-Other African American Other domestic domestic admit admit rate admit rate rate 0.05704 0.06780 -0.01076 0.05495 0.06893 -0.01397 0.05582 0.06983 -0.01401 0.05582 0.07065 -0.01483 0.06188 0.07251 -0.01063 0.06188 0.07255 -0.01068 0.06621 0.07601 -0.00980 0.06750 0.07769 -0.01019 0.06837 0.07821 -0.00984 0.07399 0.08006 -0.00607 0.07789 0.08088 -0.00300 0.07962 0.08175 -0.00213 0.07919 0.08244 -0.00325 0.07919 0.08248 -0.00329 0.07915 0.08313 -0.00398 0.07266 0.07782 -0.00516 0.07093 0.07415 -0.00322 0.07310 0.07575 -0.00265 0.07310 0.07575 -0.00265 0.07310 0.07575 -0.00265 0.07353 0.07575 -0.00222 0.07353 0.07575 -0.00222 0.07353 0.07575 -0.00222 0.07353 0.07575 -0.00222 0.07353 0.07575 -0.00222 0.07353 0.07567 -0.00214 0.07353 0.07567 -0.00214 0.07353 0.07566 -0.00214 0.07353 0.07566 -0.00213 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07578 -0.00225 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07570 -0.00217 0.07353 0.07652 -0.00299 0.07396 0.07773 -0.00376 0.07396 0.07773 -0.00376 0.07396 0.07773 -0.00377 0.07396 0.07773 -0.00377 0.07396 0.07773 -0.00376 0.07396 0.07773 -0.00376 0.07396 0.07777 -0.00381 0.07396 0.07781 -0.00385 0.07396 0.07781 -0.00385 0.07396 0.07859 -0.00463 0.07396 0.07859 -0.00463 0.07396 0.07859 -0.00463 0.07396 0.07859 -0.00463 0.07439 0.07924 -0.00484 0.07439 0.07919 -0.00480 0.07439 0.07919 -0.00480 0.07439 0.07919 -0.00480 0.07439 0.07928 -0.00489 0.07439 0.07932 -0.00493 0.07439 0.07937 -0.00497 0.07439 0.07932 -0.00493 0.07439 0.07941 -0.00502 0.07439 0.07941 -0.00502 0.07439 0.07937 -0.00497 Table B.1.6R: Admit rates for single-race African Americans and other domestic applicants by day, 2015 (pre-IPEDs) Date 3/2/11 3/3/11 3/4/11 3/5/11 3/6/11 3/7/11 3/8/11 3/9/11 3/10/11 3/11/11 3/12/11 3/14/11 3/15/11 3/16/11 3/17/11 3/18/11 3/19/11 3/20/11 3/21/11 3/22/11 3/23/11 3/24/11 3/25/11 3/28/11 3/29/11 3/30/11 4/8/11 4/28/11 5/4/11 5/5/11 5/6/11 5/9/11 5/10/11 5/11/11 5/12/11 5/13/11 5/16/11 5/17/11 5/19/11 5/31/11 6/1/11 6/2/11 6/3/11 6/6/11 6/14/11 6/16/11 6/17/11 6/20/11 6/21/11 6/22/11 6/23/11 6/24/11 6/25/11 6/26/11 6/27/11 6/28/11 6/29/11 6/30/11 7/1/11 7/2/11 7/5/11 7/6/11 7/8/11 7/18/11 7/22/11 8/5/11 8/15/11 8/18/11 8/29/11 Single-race Single-race African American All other African American admits domestic admits applicants 177 1612 2884 175 1613 2884 177 1677 2884 176 1683 2884 176 1683 2884 182 1731 2884 191 1795 2884 201 1847 2884 200 1881 2884 201 1943 2884 205 1965 2884 207 1989 2884 209 2004 2884 210 2010 2884 196 1875 2884 186 1748 2884 188 1747 2884 188 1747 2884 188 1747 2884 188 1748 2884 188 1748 2884 188 1750 2884 188 1751 2884 188 1751 2884 188 1750 2884 188 1751 2884 188 1751 2884 188 1749 2884 188 1755 2884 188 1757 2884 188 1761 2884 188 1765 2884 188 1765 2884 188 1769 2884 188 1760 2884 188 1760 2884 188 1760 2884 188 1760 2884 188 1760 2884 188 1769 2884 188 1768 2884 188 1768 2884 188 1768 2884 188 1768 2884 188 1778 2884 188 1775 2884 188 1775 2884 188 1775 2884 188 1775 2884 188 1775 2884 188 1775 2884 188 1775 2884 188 1775 2884 188 1775 2884 188 1775 2884 188 1779 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 188 1780 2884 All other domestic applicants 26048 26048 26048 26048 26048 26048 26048 26048 26048 26048 26048 26048 26048 26048 26048 26049 26049 26049 26049 26049 26050 26050 26050 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 26052 Single-race African American Single-race admit rate-Other African American Other domestic domestic admit admit rate admit rate rate 0.06137 0.06189 -0.00051 0.06068 0.06192 -0.00124 0.06137 0.06438 -0.00301 0.06103 0.06461 -0.00359 0.06103 0.06461 -0.00359 0.06311 0.06645 -0.00335 0.06623 0.06891 -0.00268 0.06969 0.07091 -0.00121 0.06935 0.07221 -0.00286 0.06969 0.07459 -0.00490 0.07108 0.07544 -0.00436 0.07178 0.07636 -0.00458 0.07247 0.07693 -0.00447 0.07282 0.07717 -0.00435 0.06796 0.07198 -0.00402 0.06449 0.06710 -0.00261 0.06519 0.06707 -0.00188 0.06519 0.06707 -0.00188 0.06519 0.06707 -0.00188 0.06519 0.06710 -0.00192 0.06519 0.06710 -0.00191 0.06519 0.06718 -0.00199 0.06519 0.06722 -0.00203 0.06519 0.06721 -0.00202 0.06519 0.06717 -0.00199 0.06519 0.06721 -0.00202 0.06519 0.06721 -0.00202 0.06519 0.06713 -0.00195 0.06519 0.06737 -0.00218 0.06519 0.06744 -0.00225 0.06519 0.06760 -0.00241 0.06519 0.06775 -0.00256 0.06519 0.06775 -0.00256 0.06519 0.06790 -0.00272 0.06519 0.06756 -0.00237 0.06519 0.06756 -0.00237 0.06519 0.06756 -0.00237 0.06519 0.06756 -0.00237 0.06519 0.06756 -0.00237 0.06519 0.06790 -0.00272 0.06519 0.06786 -0.00268 0.06519 0.06786 -0.00268 0.06519 0.06786 -0.00268 0.06519 0.06786 -0.00268 0.06519 0.06825 -0.00306 0.06519 0.06813 -0.00295 0.06519 0.06813 -0.00295 0.06519 0.06813 -0.00295 0.06519 0.06813 -0.00295 0.06519 0.06813 -0.00295 0.06519 0.06813 -0.00295 0.06519 0.06813 -0.00295 0.06519 0.06813 -0.00295 0.06519 0.06813 -0.00295 0.06519 0.06813 -0.00295 0.06519 0.06829 -0.00310 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 0.06519 0.06832 -0.00314 Table B.1.7R: Admit rates for single-race African Americans and other domestic applicants by day, 2016 (pre-IPEDS) Date 3/1/12 3/2/12 3/3/12 3/4/12 3/5/12 3/6/12 3/7/12 3/8/12 3/9/12 3/10/12 3/11/12 3/12/12 3/13/12 3/14/12 3/15/12 3/16/12 3/17/12 3/18/12 3/19/12 3/20/12 3/21/12 3/22/12 3/23/12 3/24/12 3/25/12 3/26/12 3/27/12 3/28/12 3/29/12 3/30/12 3/31/12 4/1/12 4/2/12 4/3/12 4/4/12 4/5/12 4/6/12 4/7/12 4/8/12 4/9/12 4/10/12 4/11/12 4/12/12 4/13/12 4/14/12 4/15/12 4/16/12 4/17/12 4/18/12 4/19/12 4/20/12 4/21/12 4/22/12 4/23/12 4/24/12 4/25/12 4/26/12 4/27/12 4/28/12 4/29/12 4/30/12 5/1/12 5/2/12 5/3/12 5/4/12 5/7/12 5/8/12 5/9/12 5/10/12 5/11/12 5/14/12 5/15/12 5/16/12 5/17/12 5/18/12 5/19/12 5/20/12 5/21/12 5/22/12 5/23/12 5/24/12 5/25/12 5/26/12 5/28/12 5/29/12 5/30/12 5/31/12 6/1/12 6/2/12 6/4/12 6/5/12 6/6/12 6/7/12 6/8/12 6/9/12 6/11/12 6/12/12 6/13/12 6/15/12 6/16/12 6/17/12 6/18/12 6/19/12 6/20/12 6/21/12 6/22/12 6/24/12 6/25/12 6/26/12 6/28/12 6/29/12 6/30/12 7/2/12 7/3/12 7/5/12 7/9/12 7/11/12 7/13/12 8/2/12 8/10/12 8/17/12 8/20/12 Single-race Single-race African American All other African American admits domestic admits applicants 142 1600 2567 142 1594 2567 140 1569 2567 140 1569 2567 141 1611 2645 144 1641 2645 144 1657 2645 147 1677 2645 150 1713 2645 152 1753 2645 152 1753 2645 153 1775 2645 157 1802 2645 162 1807 2645 166 1844 2645 168 1860 2645 163 1779 2645 163 1779 2645 148 1661 2645 147 1675 2645 147 1673 2645 147 1673 2645 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1673 2646 147 1672 2646 147 1672 2646 147 1671 2646 147 1671 2646 147 1671 2646 147 1671 2646 147 1671 2646 147 1688 2646 147 1701 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1692 2646 147 1708 2646 147 1702 2646 147 1702 2646 147 1702 2646 147 1702 2646 147 1702 2646 147 1702 2646 147 1702 2646 147 1702 2646 147 1702 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1707 2646 147 1711 2646 147 1711 2646 147 1711 2646 147 1711 2646 147 1711 2646 147 1713 2646 147 1713 2646 147 1713 2646 147 1713 2646 147 1713 2646 147 1713 2646 147 1713 2646 147 1713 2646 147 1713 2646 All other domestic applicants 24877 24881 24881 24882 25329 25329 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25328 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 25329 Single-race African American Single-race admit rate-Other African American Other domestic domestic admit admit rate admit rate rate 0.05532 0.06432 -0.00900 0.05532 0.06406 -0.00875 0.05454 0.06306 -0.00852 0.05454 0.06306 -0.00852 0.05331 0.06360 -0.01029 0.05444 0.06479 -0.01035 0.05444 0.06542 -0.01098 0.05558 0.06621 -0.01063 0.05671 0.06763 -0.01092 0.05747 0.06921 -0.01175 0.05747 0.06921 -0.01175 0.05784 0.07008 -0.01224 0.05936 0.07115 -0.01179 0.06125 0.07134 -0.01010 0.06276 0.07280 -0.01004 0.06352 0.07344 -0.00992 0.06163 0.07024 -0.00861 0.06163 0.07024 -0.00861 0.05595 0.06558 -0.00962 0.05558 0.06613 -0.01056 0.05558 0.06605 -0.01048 0.05558 0.06605 -0.01048 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06605 -0.01050 0.05556 0.06601 -0.01046 0.05556 0.06601 -0.01046 0.05556 0.06597 -0.01042 0.05556 0.06597 -0.01042 0.05556 0.06597 -0.01042 0.05556 0.06597 -0.01042 0.05556 0.06597 -0.01042 0.05556 0.06664 -0.01109 0.05556 0.06716 -0.01160 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06680 -0.01125 0.05556 0.06743 -0.01188 0.05556 0.06720 -0.01164 0.05556 0.06720 -0.01164 0.05556 0.06720 -0.01164 0.05556 0.06720 -0.01164 0.05556 0.06720 -0.01164 0.05556 0.06720 -0.01164 0.05556 0.06720 -0.01164 0.05556 0.06720 -0.01164 0.05556 0.06720 -0.01164 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06739 -0.01184 0.05556 0.06755 -0.01200 0.05556 0.06755 -0.01200 0.05556 0.06755 -0.01200 0.05556 0.06755 -0.01200 0.05556 0.06755 -0.01200 0.05556 0.06763 -0.01207 0.05556 0.06763 -0.01207 0.05556 0.06763 -0.01207 0.05556 0.06763 -0.01207 0.05556 0.06763 -0.01207 0.05556 0.06763 -0.01207 0.05556 0.06763 -0.01207 0.05556 0.06763 -0.01207 0.05556 0.06763 -0.01207 Table B.2.1R Admission Decisions by Race/Ethnicity and Year for the Baseline Sample Admission Status Race Rejected Waitlist Rejected Admit Observations 2014 White African American Hispanic Asian American 81.7 86.1 85.8 81.8 12.0 4.6 7.4 12.1 6.3 9.4* 6.7 6.2 9,371 2,158 2,459 6,186 83.5 87.0 86.4 83.6 11.2 5.1 7.0 11.3 5.3 7.9* 6.6* 5.1 10,416 2,810 3,139 7,184 84.4 87.6 86.7 83.1 10.5 5.4 7.6 11.4 5.1 7.0* 5.6 5.5 9,417 2,604 2,904 6,436 86.5 90.0 87.9 85.0 9.0 3.0 5.9 10.0 4.5 7.0* 6.2* 5.0 9,315 2,615 2,911 6,430 85.1 87.4 85.2 84.1 10.6 4.9 8.2 11.5 4.2 7.7* 6.6* 4.4 9,488 2,651 3,086 7,095 85.8 87.6 87.8 83.4 10.3 5.5 6.8 11.8 3.9* 6.9* 5.4 4.8 9,575 2,826 3,471 7,084 2015 White African American Hispanic Asian American 2016 White African American Hispanic Asian American 2017 White African American Hispanic Asian American 2018 White African American Hispanic Asian American 2019 White African American Hispanic Asian American A * indicates statistically different at the 5% level * Constructed using results from basicFreqs.do Table B.2.2R Admission Decisions by Race/Ethnicity and Year for the Expanded Sample Admission Status Race Rejected Waitlist Rejected Admit Observations 2014 White African American Hispanic Asian American 78.5 85.3 85.1 81.2 13.2 4.9 7.8 12.3 8.2* 9.8* 7.1 6.5 10,055 2,204 2,496 6,288 80.3 86.6 85.6 82.9 12.9 5.2 7.4 11.6 6.8* 8.1* 7.0* 5.5 11,101 2,858 3,191 7,284 80.8 87.3 85.7 82.1 12.2 5.6 8.2 11.8 7.0* 7.1 6.1 6.1 10,111 2,653 2,970 6,559 82.9 89.3 87.1 84.0 10.5 3.1 6.2 10.3 6.6* 7.6* 6.7 5.7 9,979 2,666 2,963 6,553 81.7 86.8 84.1 83.2 12.0 5.2 8.6 11.8 6.3* 7.9* 7.3* 5.0 10,174 2,697 3,162 7,213 82.5 86.9 86.8 82.4 11.8 5.6 7.2 12.0 5.7 7.5* 6.1 5.6 10,237 2,881 3,540 7,245 2015 White African American Hispanic Asian American 2016 White African American Hispanic Asian American 2017 White African American Hispanic Asian American 2018 White African American Hispanic Asian American 2019 White African American Hispanic Asian American A * indicates statistically different at the 5% level * Constructed using results from basicFreqs.do Table B.2.3R Admission Decisions by Race/Ethnicity and Year for the Baseline Sample without Early Applicants Admission Status Race Rejected Waitlist Rejected Admit Observations 2014 White African American Hispanic Asian American 81.7 86.1 85.8 81.8 12.0 4.6 7.4 12.1 6.3 9.4* 6.7 6.2 9,371 2,158 2,459 6,186 83.5 87.0 86.4 83.6 11.2 5.1 7.0 11.3 5.3 7.9* 6.6* 5.1 10,416 2,810 3,139 7,184 85.7 89.3 88.2 85.4 10.4 5.7 7.5 10.8 3.9 5.0* 4.3 3.8 8,253 2,291 2,585 5,626 87.6 91.4 89.2 87.4 9.1 2.8 5.9 9.6 3.2 5.8* 4.9* 3.0 8,058 2,271 2,575 5,541 86.1 89.1 86.2 86.0 11.0 5.5 8.9 11.6 2.9 5.5* 4.9* 2.4 8,225 2,306 2,738 6,178 86.6 88.9 88.6 85.2 10.8 6.0 7.4 12.0 2.6 5.1* 4.0* 2.8 8,047 2,394 2,972 5,990 2015 White African American Hispanic Asian American 2016 White African American Hispanic Asian American 2017 White African American Hispanic Asian American 2018 White African American Hispanic Asian American 2019 White African American Hispanic Asian American A * indicates statistically different at the 5% level * Constructed using results from basicFreqs.do Table B.3.1R: Application summary statistics by race, baseline sample Admitted Female Disadvantaged First-generation college Mother highest ed: no college Mother highest ed: BA degree Mother highest ed: MA degree Mother highest ed: PhD/JD/MD degree Mother highest ed: Missing Father highest ed: no college Father highest ed: BA degree Father highest ed: MA degree Father highest ed: PhD/JD/MD degree Father highest ed: Missing Application read by 3rd reader Missing alumni rating Applied for fee waiver Financial Aid SAT1 math (z-score) SAT1 verbal (z-score) SAT2 avg (z-score) Never took SAT2 Standardized high school GPA (z-score) Academic index (z-score) Academic index percentile Number of AP tests taken Average score of AP tests N Reject 0.00 45.75 5.94 4.29 21.98 37.92 25.74 12.12 0.02 21.29 29.79 24.71 21.65 0.03 10.94 22.10 8.00 73.83 0.12 (0.82) 0.31 (0.76) -0.01 (0.86) 12.35 0.17 (0.86) 0.16 (0.80) 0.52 (0.26) 4.08 (3.91) 4.39 (0.59) 54,768 *Constructed using results from sumStatsTablesPoolRej.do White Admit 100.00 43.14 14.61 4.05 18.27 33.65 29.00 17.24 0.02 20.22 25.27 26.26 26.12 0.02 94.78 2.38 12.15 72.17 0.56 (0.50) 0.72 (0.43) 0.58 (0.50) 1.74 0.50 (0.52) 0.76 (0.38) 0.75 (0.19) 5.91 (3.85) 4.74 (0.34) 2,814 Total 4.89 45.62 6.36 4.28 21.80 37.71 25.90 12.37 0.02 21.23 29.57 24.79 21.86 0.03 15.03 21.14 8.20 73.75 0.15 (0.81) 0.33 (0.75) 0.03 (0.85) 11.83 0.18 (0.85) 0.19 (0.79) 0.53 (0.26) 4.16 (3.93) 4.41 (0.58) 57,582 Reject 0.00 59.98 29.27 14.41 44.47 27.33 18.73 6.82 0.03 51.02 20.68 15.94 9.17 0.03 11.69 26.87 43.81 93.68 -1.17 (1.07) -0.77 (1.07) -1.24 (1.13) 28.28 -0.51 (1.18) -1.23 (1.12) 0.19 (0.18) 2.11 (3.14) 3.78 (0.78) 14,477 African American Admit 100.00 55.01 28.48 7.67 29.06 27.46 27.30 13.65 0.03 32.69 20.22 22.49 20.22 0.04 95.11 2.44 28.14 90.73 0.14 (0.67) 0.41 (0.56) 0.15 (0.62) 2.44 0.31 (0.76) 0.33 (0.52) 0.56 (0.21) 5.06 (3.85) 4.51 (0.42) 1,187 Total 7.58 59.61 29.21 13.90 43.30 27.34 19.38 7.34 0.03 49.63 20.65 16.43 10.00 0.03 18.02 25.02 42.63 93.46 -1.07 (1.10) -0.68 (1.08) -1.10 (1.17) 26.32 -0.45 (1.17) -1.11 (1.16) 0.21 (0.21) 2.33 (3.29) 3.88 (0.78) 15,664 Reject 0.00 50.70 23.47 22.03 52.01 25.70 14.51 5.83 0.02 51.29 20.71 15.03 10.50 0.02 13.76 29.98 35.57 88.54 -0.69 (1.05) -0.45 (1.05) -0.61 (1.04) 17.87 -0.08 (0.97) -0.63 (1.01) 0.30 (0.23) 3.52 (3.82) 3.96 (0.75) 16,863 Hispanic Admit 100.00 45.98 37.40 19.96 43.99 26.29 18.61 8.94 0.02 46.43 19.33 17.80 14.27 0.02 97.11 1.99 35.59 88.98 0.28 (0.64) 0.44 (0.59) 0.41 (0.54) 2.35 0.45 (0.62) 0.50 (0.46) 0.63 (0.21) 6.20 (3.83) 4.55 (0.47) 1,107 Total 6.16 50.41 24.33 21.90 51.51 25.73 14.76 6.02 0.02 50.99 20.62 15.20 10.73 0.02 18.90 28.25 35.58 88.56 -0.63 (1.05) -0.39 (1.05) -0.53 (1.04) 16.91 -0.04 (0.97) -0.56 (1.02) 0.32 (0.24) 3.68 (3.87) 4.02 (0.75) 17,970 Reject 0.00 49.12 10.26 7.98 25.93 30.87 27.80 9.83 0.06 18.98 19.10 31.86 23.03 0.07 12.18 20.25 12.88 76.37 0.41 (0.73) 0.31 (0.80) 0.32 (0.82) 5.16 0.21 (0.82) 0.39 (0.78) (0.78) (0.27) 5.60 (4.07) 4.48 (0.56) 38,343 Asian American Admit Total 100.00 5.13 52.65 49.30 21.86 10.85 9.65 8.07 26.06 25.93 23.46 30.49 33.20 28.08 11.58 9.92 0.06 0.06 21.38 19.10 12.69 18.77 29.34 31.73 30.26 23.40 0.06 0.07 96.14 16.48 1.79 19.30 18.39 13.16 77.27 76.41 0.77 0.43 (0.37) (0.72) 0.74 0.33 (0.41) (0.79) 0.81 0.35 (0.38) (0.81) 0.34 4.91 0.52 0.22 (0.47) (0.81) 0.91 0.42 (0.32) (0.77) 0.83 0.63 (0.16) (0.27) 7.50 5.68 (3.38) (4.06) 4.82 4.50 (0.28) (0.55) 2,072 40,415 Reject 0.00 49.23 12.03 8.77 29.42 32.86 24.29 10.07 0.03 27.52 24.09 24.93 19.36 0.04 11.84 23.00 16.93 78.42 -0.03 (1.00) 0.09 (0.94) -0.08 (1.01) 12.62 0.07 (0.93) -0.02 (1.01) 0.49 (0.29) 4.28 (4.02) 4.34 (0.65) 134,944 Total Admit 100.00 48.73 22.11 8.54 25.86 28.75 28.49 13.73 0.03 26.23 19.84 25.41 24.74 0.04 95.57 2.12 19.75 79.07 0.51 (0.57) 0.64 (0.50) 0.55 (0.55) 1.63 0.47 (0.57) 0.69 (0.46) 0.73 (0.21) 6.27 (3.81) 4.70 (0.38) 7,784 Total 5.45 49.21 12.58 8.76 29.23 32.63 24.52 10.27 0.03 27.45 23.86 24.96 19.66 0.04 16.41 21.86 17.08 78.45 0.00 (0.99) 0.12 (0.93) -0.04 (1.00) 12.02 0.09 (0.92) 0.01 (1.00) 0.50 (0.29) 4.37 (4.04) 4.36 (0.64) 142,728 Table B.3.2R: Application summary statistics by race, expanded sample Reject Admitted 0.00 Female 45.76 Disadvantaged 5.69 First-generation college 4.14 Early action applicant 8.94 Legacy 3.44 Faculty child 0.03 Staff child 0.13 Dean / Director's List 1.60 Mother highest ed: no college 21.27 Mother highest ed: BA degree 37.58 Mother highest ed: MA degree 26.00 Mother highest ed: PhD/JD/MD degree 12.95 Mother highest ed: Missing 0.02 Father highest ed: no college 20.49 Father highest ed: BA degree 29.37 Father highest ed: MA degree 24.90 Father highest ed: PhD/JD/MD degree 22.73 Father highest ed: Missing 0.03 Application read by 3rd reader 12.72 Applied for fee waiver 7.71 Financial Aid 72.27 Missing alumni rating 21.49 SAT1 math (z-score) 0.12 (0.81) SAT1 verbal (z-score) 0.31 (0.75) SAT2 avg (z-score) 0.00 (0.85) Never took SAT2 12.00 Standardized high school GPA (z-score) 0.15 (0.87) Academic index (z-score) 0.52 (0.26) Academic index percentile 0.15 (0.79) Number of AP tests taken 4.05 (3.90) Average score of AP tests 4.40 (0.58) N 57,481 * Constructed using results from sumStatsTablesPoolRej.do *Taken from Table 6 White Admit 100.00 44.37 10.03 3.11 31.59 24.50 0.79 1.03 15.92 13.58 31.59 29.86 23.32 0.02 14.46 22.82 28.26 32.78 0.02 93.39 8.48 56.54 2.71 0.52 (0.52) 0.69 (0.44) 0.55 (0.52) 1.39 0.36 (0.61) 0.72 (0.20) 0.69 (0.44) 5.44 (3.83) 4.75 (0.34) 4,176 Total 6.77 45.67 5.99 4.07 10.47 4.87 0.08 0.19 2.57 20.75 37.18 26.26 13.65 0.02 20.08 28.93 25.13 23.41 0.02 18.18 7.76 71.20 20.22 0.15 (0.80) 0.34 (0.74) 0.04 (0.85) 11.28 0.16 (0.85) 0.53 (0.26) 0.19 (0.79) 4.13 (3.91) 4.42 (0.58) 61,657 Reject 0.00 59.87 28.98 14.27 8.13 1.12 0.00 0.05 0.38 44.14 27.21 18.79 7.24 0.03 50.54 20.65 16.07 9.56 0.03 12.07 43.41 93.32 26.70 -1.17 (1.07) -0.77 (1.07) -1.23 (1.13) 28.14 -0.51 (1.18) 0.19 (0.18) -1.23 (1.12) 2.10 (3.13) 3.78 (0.78) 14,691 African American Admit 100.00 55.60 26.89 7.18 24.76 5.21 0.00 0.16 2.13 27.52 26.89 27.21 15.93 0.02 31.07 19.79 22.63 22.32 0.04 94.40 26.74 88.01 2.37 0.13 (0.67) 0.41 (0.57) 0.14 (0.62) 2.29 0.28 (0.76) 0.55 (0.21) 0.32 (0.52) 4.85 (3.88) 4.51 (0.42) 1,268 Total 7.95 59.53 28.82 13.70 9.45 1.45 0.00 0.06 0.52 42.82 27.19 19.46 7.93 0.03 48.99 20.58 16.59 10.58 0.03 18.61 42.09 92.90 24.77 -1.06 (1.10) -0.67 (1.08) -1.09 (1.17) 26.09 -0.45 (1.17) 0.22 (0.21) -1.10 (1.16) 2.31 (3.28) 3.89 (0.78) 15,959 Reject 0.00 50.71 23.23 21.79 7.65 0.93 0.01 0.05 0.46 51.54 25.68 14.68 6.11 0.02 50.76 20.69 15.14 10.88 0.03 14.23 35.21 88.03 29.72 -0.69 (1.05) -0.44 (1.05) -0.60 (1.04) 17.73 -0.08 (0.97) 0.30 (0.23) -0.62 (1.01) 3.52 (3.81) 3.97 (0.75) 17,093 Hispanic Admit 100.00 46.14 34.01 18.06 25.79 7.16 0.16 0.49 4.56 40.44 25.39 19.93 12.04 0.02 42.23 18.23 19.45 17.90 0.02 96.42 32.22 82.83 2.12 0.28 (0.64) 0.45 (0.58) 0.41 (0.54) 2.12 0.42 (0.63) 0.63 (0.21) 0.50 (0.46) 6.04 (3.85) 4.57 (0.46) 1,229 Total 6.71 50.40 23.95 21.54 8.87 1.35 0.02 0.08 0.73 50.79 25.66 15.04 6.51 0.02 50.19 20.53 15.43 11.35 0.02 19.74 35.01 87.68 27.87 -0.62 (1.05) -0.38 (1.05) -0.52 (1.04) 16.68 -0.05 (0.96) 0.32 (0.24) -0.55 (1.02) 3.68 (3.87) 4.03 (0.75) 18,322 Reject 0.00 49.13 10.16 7.91 8.22 0.77 0.00 0.11 0.37 25.73 30.79 27.85 10.07 0.06 18.80 19.04 31.80 23.34 0.07 12.60 12.77 75.94 20.11 0.41 (0.73) 0.31 (0.80) 0.32 (0.82) 5.13 0.20 (0.83) 0.62 (0.27) 0.39 (0.78) 5.58 (4.07) 4.48 (0.56) 38,800 Asian American Admit Total 100.00 5.69 51.84 49.29 19.73 10.71 8.67 7.95 33.60 9.67 6.92 1.12 0.56 0.03 1.11 0.17 5.64 0.67 23.70 25.62 23.19 30.36 33.30 28.16 14.26 10.31 0.06 0.06 19.39 18.83 12.47 18.67 30.49 31.72 31.77 23.82 0.06 0.07 95.35 17.31 16.52 12.98 71.56 75.69 1.84 19.07 0.75 0.43 (0.39) (0.72) 0.73 0.33 (0.42) (0.79) 0.79 0.35 (0.40) (0.81) 0.34 4.86 0.48 0.22 (0.51) (0.81) 0.82 0.63 (0.17) (0.27) 0.89 0.41 (0.35) (0.77) 7.18 5.66 (3.53) (4.06) 4.81 4.50 (0.28) (0.55) 2,342 41,142 Reject 0.00 49.20 11.74 8.56 8.60 2.08 0.01 0.11 0.95 28.84 32.74 24.48 10.62 0.03 26.90 23.96 25.03 20.07 0.04 12.88 16.53 77.41 22.62 -0.03 (1.00) 0.10 (0.93) -0.07 (1.00) 12.42 0.06 (0.94) 0.48 (0.29) -0.02 (1.00) 4.25 (4.02) 4.34 (0.64) 138,944 Total Admit 100.00 48.66 17.82 7.01 31.07 15.05 0.61 0.87 10.08 21.48 27.89 29.20 18.58 0.03 21.47 18.95 27.02 29.31 0.03 94.36 15.93 67.99 2.33 0.50 (0.57) 0.63 (0.49) 0.54 (0.55) 1.42 0.39 (0.62) 0.71 (0.22) 0.66 (0.47) 5.89 (3.84) 4.71 (0.37) 9,825 Total 6.60 49.17 12.14 8.46 10.09 2.94 0.05 0.16 1.56 28.35 32.42 24.79 11.14 0.03 26.54 23.63 25.16 20.68 0.04 18.26 16.49 76.78 21.28 0.00 (0.98) 0.13 (0.92) -0.03 (0.99) 11.69 0.08 (0.92) 0.50 (0.29) 0.02 (0.99) 4.35 (4.03) 4.37 (0.63) 148,769 Table B.4.1R: Application rating summary statistics by race, expanded sample Reject Academic rating <3=3-, 3, or 3+ >3+ Extracurricular rating <3=3-, 3, or 3+ >3+ Athletic rating <3=3-, 3, or 3+ >3+ Personal rating <3=3-, 3, or 3+ >3+ Teacher 1 rating <3=3-, 3, or 3+ >3+ Teacher 2 rating <3=3-, 3, or 3+ >3+ School counselor rating <3=3-, 3, or 3+ >3+ Alumni Personal rating <3=3-, 3, or 3+ >3+ Alumni Overall rating <3=3-, 3, or 3+ >3+ N White Admit Total Reject Reject Hispanic Admit Total Reject Asian American Admit 10.24 46.73 43.02 0.26 14.37 85.37 9.57 44.54 45.89 54.78 40.15 5.06 0.08 41.56 58.36 Total 50.44 40.27 9.30 37.49 48.80 13.71 0.00 34.09 65.91 34.97 47.81 17.21 8.42 33.32 58.26 0.00 6.62 93.38 7.94 31.80 60.26 3.68 74.23 22.09 0.62 31.92 67.46 3.47 71.37 25.16 7.95 79.34 12.71 0.79 46.96 52.25 7.38 76.77 15.85 5.96 79.70 14.33 1.30 42.72 55.98 5.65 77.22 17.13 2.02 72.45 25.53 0.21 23.57 76.22 1.92 69.66 28.42 33.21 53.99 12.80 29.83 46.57 23.60 32.98 53.49 13.53 43.44 50.19 6.38 35.99 48.68 15.32 42.84 50.07 7.09 43.16 49.66 7.18 39.32 44.29 16.39 42.90 49.30 7.80 46.82 48.37 4.81 46.15 45.29 8.56 46.78 48.20 5.02 0.45 81.11 18.44 0.05 20.62 79.33 0.42 77.01 22.57 0.51 84.85 14.63 0.00 25.24 74.76 0.47 80.12 19.41 0.51 84.55 14.94 0.00 23.52 76.48 0.48 80.46 19.06 0.51 84.79 14.71 0.00 28.22 71.78 0.48 81.57 17.95 0.57 70.61 28.83 0.00 27.59 72.41 0.53 67.61 31.86 1.11 83.61 15.28 0.00 40.06 59.94 1.01 79.77 19.21 0.90 78.55 20.55 0.00 38.24 61.76 0.83 75.63 23.54 0.51 70.43 29.06 0.00 27.04 72.96 0.48 67.90 31.61 0.47 69.24 30.28 0.05 28.09 71.86 0.44 66.00 33.56 0.82 82.43 16.75 0.00 41.66 58.34 0.73 78.07 21.21 0.81 77.53 21.66 0.00 34.15 65.85 0.74 73.81 25.45 0.50 69.97 29.53 0.04 26.57 73.39 0.47 67.19 32.34 0.62 75.39 23.99 0.00 27.88 72.12 0.58 72.04 27.39 1.97 86.42 11.61 0.00 41.70 58.30 1.80 82.47 15.74 1.27 83.45 15.28 0.00 40.59 59.41 1.18 80.35 18.47 0.63 75.87 23.49 0.00 27.44 72.56 0.60 73.02 26.38 7.30 31.22 61.48 0.74 8.51 90.75 6.76 29.34 63.90 10.49 35.79 53.72 1.13 9.67 89.20 9.52 33.10 57.38 10.16 35.46 54.38 0.42 6.48 93.11 9.28 32.83 57.90 8.26 31.49 60.24 0.56 6.90 92.53 7.73 29.79 62.47 18.30 37.39 44.31 57,481 1.33 13.85 84.82 4,176 16.89 35.43 47.68 61,657 41.04 35.52 23.44 14,691 2.43 22.96 74.62 1,268 36.95 34.19 28.85 15,959 33.83 36.89 29.28 17,093 1.75 16.38 81.88 1,229 30.85 34.99 34.16 18,322 16.90 34.73 48.37 38,800 0.83 8.44 90.74 2,342 15.78 32.90 51.33 41,142 * Constructed using results from sumStatsSubRatTablesPoolRej.do African American Admit Total Table B.5.1R: Number and Share of Applicants by Race/Ethnicity and Academic Index Decile, Expanded Sample Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Total Numer of Applicants in Each Decile African Asian Whites American Hispanic American Total 2,948 5,997 3,600 1,528 14,847 4,690 3,659 3,799 2,087 15,130 6,475 2,335 2,974 2,696 15,632 6,832 1,315 2,228 3,087 14,556 8,196 944 1,764 3,968 16,224 6,421 519 1,118 3,698 12,818 7,556 462 984 4,613 14,843 6,940 330 850 5,406 14,891 6,160 197 564 6,631 14,939 5,303 138 401 7,321 14,512 61,521 15,896 18,282 41,035 148,392 Share of Applicants in each Decile African Asian Whites American Hispanic American Total 4.79 37.73 19.69 3.72 10.01 7.62 23.02 20.78 5.09 10.20 10.52 14.69 16.27 6.57 10.53 11.11 8.27 12.19 7.52 9.81 13.32 5.94 9.65 9.67 10.93 10.44 3.26 6.12 9.01 8.64 12.28 2.91 5.38 11.24 10.00 11.28 2.08 4.65 13.17 10.03 10.01 1.24 3.09 16.16 10.07 8.62 0.87 2.19 17.84 9.78 Table B.5.2R: Admit Rates by Race/Ethnicity and Academic Index Decile, Expanded Sample Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Whites 0.27% 1.11% 1.56% 3.26% 4.12% 6.34% 6.87% 10.19% 14.09% 17.95% Average 6.79% African Asian American Hispanic American Total 0.07% 0.00% 0.07% 0.10% 1.12% 0.45% 0.34% 0.84% 5.57% 2.05% 0.85% 2.20% 13.38% 6.01% 1.36% 4.26% 23.73% 9.98% 2.32% 5.58% 30.25% 14.67% 3.11% 7.19% 42.64% 18.39% 4.66% 8.11% 44.85% 23.88% 5.92% 9.91% 54.82% 26.95% 8.28% 12.37% 57.25% 34.66% 13.35% 16.24% 7.95% 6.71% 5.70% 6.61% Table B.5.3R: Share Receiving a Two or Higher on the Academic and Extracurricular Ratings by Race/Ethnicity and Academic Index Decile, Expanded Sample Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Average Academic Rating African Asian Whites American Hispanic American Total 0.10% 0.02% 0.03% 0.00% 0.05% 0.41% 0.08% 0.05% 0.53% 0.24% 1.92% 0.94% 0.71% 1.37% 1.47% 9.66% 6.24% 4.53% 8.07% 8.25% 27.16% 23.09% 17.74% 26.66% 25.91% 51.55% 48.75% 43.92% 51.46% 50.88% 68.99% 68.61% 64.94% 71.84% 69.88% 83.23% 80.61% 80.12% 86.31% 84.27% 93.62% 93.40% 91.49% 95.16% 94.34% 97.28% 94.93% 95.51% 98.09% 97.69% 45.93% 9.29% 17.22% 60.26% 42.72% Extracurricular Rating African Asian Whites American Hispanic American Total 11.64% 9.09% 9.33% 13.09% 10.21% 16.84% 13.94% 12.87% 15.91% 15.02% 20.69% 18.97% 15.97% 18.81% 19.30% 22.44% 23.80% 18.81% 21.77% 21.96% 24.34% 24.15% 20.98% 23.61% 24.01% 26.04% 26.40% 23.97% 25.77% 25.97% 27.81% 28.14% 27.95% 28.51% 28.14% 28.29% 28.18% 24.82% 30.10% 28.69% 31.48% 32.49% 29.43% 35.15% 33.33% 34.04% 39.86% 30.42% 38.21% 36.47% 25.16% 15.84% 17.13% 28.44% 24.22% Table B.5.4R: Share Receiving a Two or Higher on School Support Measures by Race/Ethnicity and Academic Index Decile, Expanded Sample Teacher 1 Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Whites 7.73% 13.41% 19.17% 24.30% 26.88% 32.89% 35.49% 40.76% 45.78% 50.74% Average 31.04% Teacher 2 African Asian American Hispanic American Total 7.80% 8.86% 7.40% 8.10% 13.94% 13.90% 13.99% 13.65% 19.53% 19.94% 17.25% 19.06% 25.78% 23.79% 21.15% 23.64% 29.87% 30.22% 22.88% 26.46% 36.42% 31.84% 26.66% 31.04% 40.91% 35.77% 30.48% 34.00% 46.97% 38.00% 33.44% 37.47% 47.72% 43.97% 40.05% 42.89% 56.52% 50.37% 46.71% 48.44% 17.36% 21.79% 30.97% 28.32% Whites 6.24% 10.51% 15.69% 21.66% 23.88% 28.27% 31.55% 37.41% 42.60% 47.82% 27.77% Counselor African Asian American Hispanic American Total 5.47% 6.42% 6.61% 6.05% 11.48% 11.13% 11.69% 11.10% 17.13% 17.82% 13.91% 16.11% 22.74% 21.01% 18.33% 20.98% 31.78% 25.74% 20.19% 23.63% 35.65% 28.71% 24.36% 27.47% 35.28% 33.13% 26.51% 29.99% 40.30% 37.65% 30.00% 34.22% 42.64% 39.36% 36.54% 39.51% 50.72% 50.37% 42.04% 44.57% 14.98% 19.13% 27.62% 25.22% African Asian Whites American Hispanic American Total 4.68% 4.94% 5.75% 5.76% 5.23% 9.34% 10.85% 10.27% 9.20% 9.95% 14.69% 16.96% 14.93% 12.50% 14.64% 19.25% 20.53% 17.64% 15.06% 18.19% 22.63% 27.01% 21.43% 17.84% 21.57% 26.51% 32.56% 25.40% 22.88% 25.54% 30.21% 36.58% 31.10% 25.19% 28.73% 35.36% 38.48% 34.71% 28.12% 32.16% 40.29% 44.16% 35.46% 34.17% 37.12% 45.75% 50.00% 47.38% 38.61% 41.67% 26.08% 14.06% 16.89% 25.37% 23.34% Table B.5.5R: Share Receiving a Two or Higher on the Personal Raing and Alumni Interview Personal Rating by Race/Ethnicity and Academic Index Decile, Expanded Sample Personal Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Whites 8.45% 13.41% 17.10% 20.16% 21.80% 24.22% 24.11% 27.54% 29.76% 30.61% Average 22.58% African Asian American Hispanic American Total 9.70% 8.50% 8.18% 9.02% 15.96% 13.40% 12.84% 13.94% 23.68% 17.82% 13.69% 17.67% 29.43% 20.92% 14.58% 19.90% 34.96% 26.08% 15.83% 21.53% 35.26% 28.62% 16.98% 22.79% 40.69% 30.79% 18.40% 23.05% 40.00% 32.35% 18.41% 24.20% 40.61% 30.67% 21.31% 25.74% 48.55% 36.41% 22.50% 26.22% 19.41% 19.08% 17.97% 20.35% Alumni Personal African Asian Whites American Hispanic American Total 26.80% 31.17% 26.33% 28.53% 28.48% 34.65% 39.96% 33.59% 32.20% 35.29% 40.82% 47.07% 38.97% 36.57% 40.72% 45.67% 55.89% 44.39% 41.08% 45.56% 49.78% 60.28% 51.02% 44.66% 49.27% 53.20% 62.62% 55.37% 47.57% 52.22% 55.43% 69.91% 57.52% 52.27% 54.95% 59.37% 67.58% 62.71% 54.25% 57.59% 63.07% 71.07% 63.48% 57.64% 60.76% 65.87% 74.64% 71.82% 63.88% 65.07% 50.97% 43.08% 41.77% *Note that those who do not have an alumni interview are coded as not having received a 2 or higher on the alumni overall rating 50.53% 48.86% Table B.5.6R: Share Receving a Two or Higher on Overall Rating and Alumni Interviewer Overall Rating by Race/Ethnicity and Academic Index Decile, Expanded Sample Academic Index Decile 1 2 3 4 5 6 7 8 9 10 Average Final Reader Overall Rating African Asian Whites American Hispanic American Total 0.03% 0.00% 0.00% 0.00% 0.01% 0.21% 0.49% 0.08% 0.14% 0.23% 0.45% 2.27% 0.71% 0.33% 0.77% 1.29% 7.60% 2.29% 0.42% 1.90% 2.26% 16.31% 4.82% 1.39% 3.20% 4.11% 24.47% 9.30% 2.30% 4.94% 5.65% 32.47% 12.50% 3.64% 6.32% 9.54% 37.88% 16.24% 5.27% 8.64% 13.31% 45.69% 20.74% 8.14% 11.58% 17.86% 48.55% 29.43% 13.41% 16.01% 5.58% 5.56% 4.16% 5.22% 5.30% Alumni Interviewer Overall Rating African Asian Whites American Hispanic American Total 7.60% 7.44% 7.14% 7.46% 7.48% 13.90% 15.00% 12.00% 12.31% 13.65% 20.19% 24.33% 19.50% 17.77% 20.41% 27.52% 33.76% 25.13% 23.39% 26.95% 34.11% 43.86% 35.88% 29.96% 33.82% 39.78% 51.64% 40.25% 36.61% 39.50% 44.00% 56.71% 45.63% 42.81% 44.26% 50.76% 59.39% 51.41% 47.50% 49.51% 57.73% 61.42% 59.57% 54.34% 56.54% 64.17% 66.67% 65.84% 63.30% 63.88% 37.74% 21.14% 24.19% *Note that those who do not have an alumni interview are coded as not having received a 2 or higher on the alumni overall rating 41.17% 35.38% Table B.5.7R: Number and Share of Applicants by Race/Ethnicity, Year, and Academic Index Decile, Baseline Sample Number of Applicants in Each Decile Academic African Asian Index Decile Whites American Hispanic American Total 2014 1 375 841 483 206 2,017 2 723 517 534 304 2,195 3 968 311 417 420 2,232 4 1,087 161 299 514 2,180 5 1,274 116 234 605 2,324 6 1,042 69 159 605 1,950 7 1,120 70 119 709 2,086 8 1,040 32 102 821 2,054 9 950 27 60 984 2,074 10 783 8 51 1,011 1,893 2015 1 464 1,153 638 257 2,642 2 784 666 699 350 2,641 3 1,180 377 525 460 2,732 4 1,192 216 356 584 2,498 5 1,456 169 278 767 2,841 6 1,145 74 185 734 2,257 7 1,268 71 175 833 2,462 8 1,124 44 133 994 2,398 9 1,001 21 84 1,144 2,327 10 798 16 65 1,058 1,991 2016 1 505 1,094 636 219 2,594 2 785 569 613 340 2,475 3 1,063 387 454 415 2,543 4 1,154 185 366 547 2,497 5 1,270 128 276 608 2,560 6 965 76 166 574 2,025 7 1,131 69 148 760 2,363 8 1,017 42 131 842 2,330 9 816 31 71 1,036 2,249 10 708 19 41 1,091 2,101 2017 1 462 968 548 249 2,357 2 737 597 623 348 2,464 3 964 401 459 459 2,508 4 958 229 354 430 2,198 5 1,230 128 289 611 2,531 6 926 89 180 598 2,033 7 1,156 69 152 658 2,318 8 1,007 56 131 815 2,354 9 1,012 30 102 1,064 2,563 10 804 32 58 1,153 2,411 2018 1 467 903 536 279 2,278 2 686 608 601 356 2,369 3 939 384 524 442 2,425 4 986 232 389 500 2,224 5 1,225 175 323 646 2,575 6 937 105 191 579 1,939 7 1,193 72 167 780 2,393 8 1,157 82 156 934 2,542 9 1,018 42 108 1,194 2,593 10 847 27 77 1,346 2,526 2019 1 549 962 742 301 2,705 2 689 643 685 347 2,514 3 959 431 547 448 2,574 4 982 262 418 445 2,268 5 1,203 181 319 637 2,595 6 909 95 196 524 1,906 7 1,185 94 188 787 2,523 8 1,133 70 167 910 2,575 9 920 45 114 1,110 2,497 10 1,023 30 88 1,566 3,067 Share of Applicants in Each Decile African Whites American Hispanic Asian American Total 4.01 7.72 10.34 11.61 13.61 11.13 11.96 11.11 10.15 8.36 39.08 24.02 14.45 7.48 5.39 3.21 3.25 1.49 1.25 0.37 19.65 21.72 16.97 12.16 9.52 6.47 4.84 4.15 2.44 2.07 3.33 4.92 6.8 8.32 9.79 9.79 11.47 13.29 15.92 16.36 9.6 10.45 10.63 10.38 11.06 9.28 9.93 9.78 9.87 9.01 4.46 7.53 11.33 11.45 13.98 11 12.18 10.8 9.61 7.66 41.08 23.73 13.43 7.7 6.02 2.64 2.53 1.57 0.75 0.57 20.33 22.28 16.73 11.34 8.86 5.9 5.58 4.24 2.68 2.07 3.58 4.87 6.41 8.13 10.68 10.22 11.6 13.84 15.93 14.73 10.66 10.65 11.02 10.08 11.46 9.1 9.93 9.67 9.39 8.03 5.36 8.34 11.29 12.26 13.49 10.25 12.01 10.8 8.67 7.52 42.08 21.88 14.88 7.12 4.92 2.92 2.65 1.62 1.19 0.73 21.92 21.12 15.64 12.61 9.51 5.72 5.1 4.51 2.45 1.41 3.4 5.29 6.45 8.5 9.45 8.92 11.82 13.09 16.11 16.96 10.93 10.43 10.71 10.52 10.78 8.53 9.95 9.82 9.47 8.85 4.99 7.96 10.41 10.35 13.29 10 12.49 10.88 10.93 8.69 37.25 22.97 15.43 8.81 4.92 3.42 2.65 2.15 1.15 1.23 18.92 21.51 15.85 12.22 9.98 6.22 5.25 4.52 3.52 2 3.9 5.45 7.19 6.73 9.57 9.37 10.31 12.76 16.66 18.06 9.93 10.38 10.57 9.26 10.66 8.56 9.77 9.92 10.8 10.16 4.94 7.26 9.93 10.43 12.96 9.91 12.62 12.24 10.77 8.96 34.33 23.12 14.6 8.82 6.65 3.99 2.74 3.12 1.6 1.03 17.45 19.56 17.06 12.66 10.51 6.22 5.44 5.08 3.52 2.51 3.95 5.05 6.26 7.09 9.16 8.21 11.05 13.24 16.92 19.08 9.55 9.93 10.16 9.32 10.79 8.13 10.03 10.65 10.87 10.58 5.75 7.21 10.04 10.28 12.59 9.52 12.41 11.86 9.63 10.71 34.2 22.86 15.32 9.31 6.43 3.38 3.34 2.49 1.6 1.07 21.42 19.77 15.79 12.07 9.21 5.66 5.43 4.82 3.29 2.54 4.25 4.9 6.33 6.29 9 7.41 11.12 12.86 15.69 22.13 10.72 9.97 10.2 8.99 10.29 7.56 10 10.21 9.9 12.16 Table B.5.8R: Number and Share of Applicants by Race/Ethnicity, Year, and Academic Index Decile, Expanded Sample Number of Applicants in Each Decile Academic African Asian Index Decile Whites American Hispanic American Total 2014 1 399 855 484 211 2,063 2 767 526 539 313 2,266 3 1,057 315 423 427 2,347 4 1,169 166 303 527 2,292 5 1,365 123 237 623 2,449 6 1,136 72 167 621 2,076 7 1,206 74 119 721 2,192 8 1,102 32 107 832 2,137 9 1,024 27 63 992 2,161 10 821 8 53 1,014 1,940 2015 1 490 1,166 639 259 2,688 2 835 678 707 357 2,726 3 1,247 386 533 472 2,840 4 1,268 222 364 591 2,606 5 1,550 175 287 779 2,972 6 1,239 75 188 742 2,372 7 1,351 72 177 839 2,555 8 1,202 44 138 1,013 2,505 9 1,068 21 89 1,160 2,423 10 847 16 68 1,069 2,062 2016 1 533 1,107 639 221 2,645 2 841 579 620 346 2,562 3 1,126 395 465 429 2,653 4 1,254 190 378 556 2,642 5 1,353 136 281 627 2,698 6 1,050 78 173 591 2,156 7 1,208 71 160 771 2,479 8 1,098 42 135 852 2,433 9 880 32 74 1,056 2,353 10 765 19 43 1,106 2,183 2017 1 475 979 551 254 2,392 2 780 608 631 356 2,539 3 1,019 408 465 462 2,599 4 1,041 232 361 443 2,315 5 1,309 138 292 627 2,655 6 1,004 90 185 612 2,146 7 1,236 74 160 676 2,443 8 1,099 57 137 827 2,478 9 1,092 30 107 1,076 2,674 10 861 34 59 1,175 2,508 2018 1 487 911 544 282 2,317 2 728 616 609 361 2,443 3 1,007 395 531 449 2,527 4 1,051 240 395 513 2,328 5 1,328 180 338 656 2,713 6 1,005 106 198 591 2,033 7 1,287 73 174 799 2,525 8 1,226 84 162 944 2,639 9 1,108 42 114 1,215 2,719 10 913 29 83 1,364 2,630 2019 1 564 979 743 301 2,742 2 739 652 693 354 2,594 3 1,019 436 557 457 2,666 4 1,049 265 427 457 2,373 5 1,291 192 329 656 2,737 6 987 98 207 541 2,035 7 1,268 98 194 807 2,649 8 1,213 71 171 938 2,699 9 988 45 117 1,132 2,609 10 1,096 32 95 1,593 3,189 Share of Applicants in Each Decile African Whites American Hispanic Asian American Total 3.97 7.63 10.52 11.64 13.59 11.31 12.00 10.97 10.19 8.17 38.90 23.93 14.33 7.55 5.60 3.28 3.37 1.46 1.23 0.36 19.40 21.60 16.95 12.14 9.50 6.69 4.77 4.29 2.53 2.12 3.36 4.98 6.80 8.39 9.92 9.89 11.48 13.25 15.79 16.14 9.41 10.34 10.71 10.45 11.17 9.47 10.00 9.75 9.86 8.85 4.42 7.52 11.24 11.43 13.97 11.17 12.17 10.83 9.62 7.63 40.84 23.75 13.52 7.78 6.13 2.63 2.52 1.54 0.74 0.56 20.03 22.16 16.71 11.41 9.00 5.89 5.55 4.33 2.79 2.13 3.56 4.90 6.48 8.12 10.70 10.19 11.52 13.91 15.93 14.68 10.44 10.59 11.03 10.12 11.54 9.21 9.92 9.73 9.41 8.01 5.27 8.32 11.14 12.41 13.39 10.39 11.95 10.86 8.71 7.57 41.79 21.86 14.91 7.17 5.13 2.94 2.68 1.59 1.21 0.72 21.53 20.89 15.67 12.74 9.47 5.83 5.39 4.55 2.49 1.45 3.37 5.28 6.54 8.48 9.57 9.02 11.76 13.00 16.11 16.87 10.66 10.33 10.70 10.65 10.88 8.69 9.99 9.81 9.49 8.80 4.79 7.87 10.28 10.50 13.20 10.13 12.46 11.08 11.01 8.68 36.94 22.94 15.40 8.75 5.21 3.40 2.79 2.15 1.13 1.28 18.69 21.40 15.77 12.25 9.91 6.28 5.43 4.65 3.63 2.00 3.90 5.47 7.10 6.81 9.63 9.40 10.39 12.71 16.53 18.05 9.67 10.26 10.50 9.35 10.73 8.67 9.87 10.01 10.80 10.13 4.80 7.18 9.93 10.36 13.10 9.91 12.69 12.09 10.93 9.00 34.04 23.02 14.76 8.97 6.73 3.96 2.73 3.14 1.57 1.08 17.28 19.35 16.87 12.55 10.74 6.29 5.53 5.15 3.62 2.64 3.93 5.03 6.26 7.15 9.14 8.24 11.14 13.16 16.94 19.01 9.31 9.82 10.16 9.36 10.91 8.17 10.15 10.61 10.93 10.57 5.52 7.24 9.98 10.27 12.64 9.66 12.41 11.88 9.67 10.73 34.14 22.73 15.20 9.24 6.69 3.42 3.42 2.48 1.57 1.12 21.03 19.62 15.77 12.09 9.31 5.86 5.49 4.84 3.31 2.69 4.16 4.89 6.32 6.32 9.07 7.48 11.15 12.96 15.64 22.01 10.43 9.87 10.14 9.03 10.41 7.74 10.07 10.27 9.92 12.13 Table B.5.9R: Admit Rates by Race/Ethnicity and Academic Index Decile Baseline Sample Academic African Asian Index Decile Whites American Hispanic American 2014 1 0.00% 0.00% 0.00% 0.00% 2 0.41% 1.35% 0.56% 0.33% 3 1.34% 7.40% 2.64% 0.24% 4 1.66% 19.88% 5.35% 1.75% 5 3.61% 31.03% 12.82% 1.49% 6 5.09% 46.38% 13.84% 2.64% 7 6.61% 51.43% 21.85% 4.80% 8 8.75% 53.13% 24.51% 7.19% 9 15.05% 48.15% 20.00% 10.77% 10 19.28% 75.00% 39.22% 14.54% 2015 1 0.00% 0.00% 0.00% 0.00% 2 0.38% 1.20% 0.14% 0.29% 3 0.17% 7.16% 0.95% 1.09% 4 2.85% 15.28% 8.15% 0.51% 5 2.20% 27.81% 10.43% 2.22% 6 4.10% 32.43% 13.51% 3.00% 7 4.73% 43.66% 21.71% 4.08% 8 8.45% 52.27% 24.81% 4.83% 9 12.29% 71.43% 33.33% 8.48% 10 19.05% 75.00% 30.77% 12.85% 2016 1 0.00% 0.00% 0.00% 0.00% 2 0.38% 1.76% 0.33% 0.29% 3 0.47% 5.68% 2.20% 0.48% 4 1.82% 10.81% 5.74% 0.73% 5 2.60% 16.41% 8.33% 1.64% 6 4.35% 31.58% 10.84% 2.26% 7 5.22% 40.58% 15.54% 4.47% 8 8.75% 54.76% 26.72% 4.63% 9 11.40% 64.52% 29.58% 8.59% 10 18.79% 63.16% 24.39% 14.76% 2017 1 0.00% 0.00% 0.00% 0.00% 2 0.27% 0.50% 0.16% 0.29% 3 0.52% 3.99% 3.70% 1.09% 4 1.04% 13.10% 5.08% 1.16% 5 3.01% 25.00% 9.34% 1.64% 6 4.32% 26.97% 12.78% 2.34% 7 4.24% 39.13% 18.42% 4.41% 8 7.05% 32.14% 21.37% 5.52% 9 9.29% 46.67% 21.57% 6.20% 10 13.81% 59.38% 27.59% 12.92% 2018 1 0.00% 0.22% 0.00% 0.00% 2 0.29% 0.49% 0.00% 0.00% 3 0.53% 3.39% 1.15% 0.23% 4 1.62% 9.48% 4.11% 0.60% 5 1.47% 18.29% 8.05% 1.70% 6 2.88% 27.62% 16.23% 2.42% 7 4.44% 38.89% 15.57% 2.82% 8 7.17% 46.34% 24.36% 4.18% 9 8.84% 57.14% 28.70% 5.61% 10 12.75% 44.44% 38.96% 11.44% 2019 1 0.00% 0.00% 0.00% 0.00% 2 0.58% 0.93% 0.73% 0.00% 3 0.42% 4.18% 1.46% 0.67% 4 1.73% 10.31% 4.78% 0.45% 5 2.58% 18.23% 6.90% 2.35% 6 4.40% 18.95% 14.29% 2.10% 7 3.63% 35.11% 12.23% 3.43% 8 5.21% 37.14% 17.37% 4.62% 9 7.93% 46.67% 23.68% 6.13% 10 10.07% 43.33% 26.14% 10.86% Expanded Sample African Whites American Hispanic Asian American 0.25% 1.30% 2.65% 3.17% 5.42% 7.48% 8.87% 11.07% 18.46% 21.32% 0.12% 1.52% 7.62% 21.08% 32.52% 47.22% 51.35% 53.13% 48.15% 75.00% 0.00% 0.93% 2.60% 5.94% 13.50% 15.57% 21.85% 24.30% 19.05% 41.51% 0.47% 0.64% 0.23% 2.28% 1.93% 3.22% 4.85% 7.93% 11.09% 14.60% 0.61% 0.84% 0.88% 4.18% 3.42% 6.21% 6.66% 10.32% 15.17% 20.90% 0.00% 1.33% 7.77% 15.77% 28.00% 33.33% 44.44% 52.27% 71.43% 75.00% 0.00% 0.14% 0.94% 8.52% 11.15% 14.36% 22.60% 26.09% 32.58% 33.82% 0.00% 0.28% 1.48% 0.85% 2.57% 3.37% 4.29% 5.63% 8.97% 13.28% 0.75% 1.19% 1.15% 3.43% 4.58% 6.76% 6.62% 11.75% 15.11% 21.44% 0.00% 1.73% 5.82% 11.58% 16.18% 30.77% 42.25% 54.76% 65.63% 63.16% 0.00% 0.48% 2.37% 6.08% 8.90% 11.56% 17.50% 27.41% 29.73% 25.58% 0.00% 0.58% 0.93% 1.26% 2.55% 3.05% 5.06% 5.40% 9.28% 15.10% 0.00% 0.90% 1.47% 3.27% 4.35% 6.57% 6.31% 10.46% 12.73% 16.96% 0.00% 0.66% 4.66% 13.79% 26.81% 27.78% 43.24% 33.33% 46.67% 61.76% 0.00% 0.16% 3.87% 4.99% 9.93% 14.05% 20.63% 24.09% 23.36% 28.81% 0.00% 0.28% 1.52% 1.81% 1.91% 3.10% 5.33% 6.53% 6.88% 14.04% 0.00% 1.10% 1.89% 2.47% 3.01% 4.38% 7.46% 9.87% 12.73% 16.32% 0.33% 0.65% 3.54% 10.00% 19.44% 27.36% 39.73% 46.43% 57.14% 44.83% 0.00% 0.33% 1.13% 4.81% 9.47% 16.16% 16.67% 24.69% 30.70% 43.37% 0.00% 0.00% 0.22% 1.17% 1.83% 3.05% 4.01% 4.66% 6.75% 12.10% 0.00% 1.35% 1.47% 2.86% 4.03% 6.48% 5.36% 7.91% 10.53% 12.86% 0.00% 0.92% 4.59% 10.57% 21.35% 20.41% 36.73% 38.03% 46.67% 46.88% 0.00% 0.72% 1.80% 5.85% 7.90% 15.94% 12.89% 18.13% 24.79% 31.58% 0.00% 0.28% 0.66% 0.88% 3.05% 2.77% 4.58% 5.65% 7.16% 11.93% Table B.6.1R: Ordered logit estimates of Harvard's Academic and Extracurricular Ratings, baseline sample Model 1 African American Hispanic Asian American Missing Female Disadvantaged First generation Early Decision Waiver Applied for Financial Aid Humanities Biology Physical Sciences Engineering Mathematics Computer Science Unspecified Academic index Model 2 Academic Model 3 Model 4 Model 5 Model 6 Model 1 Model 2 Extracurricular Model 3 Model 4 Model 5 Model 6 -1.685 (0.019) -0.944 (0.017) 0.614 (0.014) 0.318 (0.023) -0.272 (0.011) 0.131 (0.020) -0.209 (0.022) 0.437 (0.020) -0.713 (0.018) -0.128 (0.014) 0.048 (0.018) 0.099 (0.015) 0.727 (0.024) 0.386 (0.018) 0.499 (0.026) 0.438 (0.030) -0.504 (0.029) 0.085 (0.027) -0.188 (0.024) 0.017 (0.019) 0.058 (0.032) 0.109 (0.014) 0.025 (0.024) -0.020 (0.027) 0.173 (0.025) -0.060 (0.022) -0.087 (0.019) 0.035 (0.023) 0.020 (0.019) 0.151 (0.032) -0.026 (0.023) 0.038 (0.034) -0.060 (0.039) -0.092 (0.037) 3.842 (0.047) 0.038 (0.039) -0.146 (0.033) 0.026 (0.027) 0.075 (0.046) 0.164 (0.033) 0.117 (0.044) -0.016 (0.027) 0.029 (0.038) -0.061 (0.022) -0.089 (0.019) 0.073 (0.036) 0.038 (0.028) 0.177 (0.042) -0.027 (0.030) 0.121 (0.045) -0.049 (0.046) -0.055 (0.053) 3.845 (0.047) -0.006 (0.043) -0.112 (0.037) 0.136 (0.031) 0.082 (0.051) 0.116 (0.034) 0.048 (0.046) -0.016 (0.028) -0.037 (0.040) -0.054 (0.024) -0.010 (0.021) 0.032 (0.038) 0.100 (0.030) 0.223 (0.044) 0.065 (0.032) 0.134 (0.047) 0.030 (0.049) 0.027 (0.055) 3.746 (0.049) -0.001 (0.044) -0.109 (0.037) 0.132 (0.031) 0.075 (0.051) 0.117 (0.034) 0.055 (0.046) -0.015 (0.028) -0.035 (0.040) -0.054 (0.024) -0.011 (0.021) 0.032 (0.038) 0.099 (0.030) 0.222 (0.044) 0.064 (0.032) 0.133 (0.047) 0.026 (0.049) 0.026 (0.055) 3.746 (0.049) -0.503 (0.023) -0.302 (0.021) 0.246 (0.015) 0.133 (0.025) 0.207 (0.012) 0.372 (0.024) -0.013 (0.027) 0.458 (0.021) -0.250 (0.022) -0.063 (0.015) 0.076 (0.020) -0.493 (0.017) -0.436 (0.027) -0.553 (0.020) -0.516 (0.029) -0.568 (0.034) -0.649 (0.035) 0.034 (0.027) -0.102 (0.023) 0.120 (0.016) 0.075 (0.026) 0.261 (0.013) 0.343 (0.024) 0.049 (0.027) 0.357 (0.021) -0.046 (0.022) -0.069 (0.015) 0.065 (0.020) -0.527 (0.017) -0.594 (0.027) -0.656 (0.021) -0.627 (0.030) -0.685 (0.034) -0.504 (0.035) 0.467 (0.039) -0.066 (0.040) -0.146 (0.032) 0.103 (0.023) 0.041 (0.038) 0.151 (0.029) 0.350 (0.042) 0.051 (0.027) 0.275 (0.032) -0.046 (0.022) -0.064 (0.015) 0.111 (0.031) -0.573 (0.025) -0.668 (0.035) -0.763 (0.027) -0.725 (0.039) -0.751 (0.041) -0.537 (0.051) 0.468 (0.039) -0.217 (0.044) -0.146 (0.036) 0.171 (0.026) 0.077 (0.043) 0.021 (0.031) 0.202 (0.045) 0.025 (0.029) 0.188 (0.035) -0.102 (0.024) 0.002 (0.018) 0.024 (0.033) -0.540 (0.027) -0.708 (0.037) -0.688 (0.029) -0.765 (0.041) -0.759 (0.043) -0.464 (0.053) 0.081 (0.044) -0.291 (0.045) -0.177 (0.036) 0.198 (0.026) 0.099 (0.043) 0.004 (0.031) 0.132 (0.045) 0.023 (0.029) 0.163 (0.035) -0.104 (0.024) -0.003 (0.018) 0.032 (0.034) -0.524 (0.027) -0.679 (0.038) -0.664 (0.029) -0.727 (0.041) -0.718 (0.043) -0.459 (0.054) 0.085 (0.044) 0.047 (0.041) -0.103 (0.035) 0.090 (0.029) 0.062 (0.048) 0.150 (0.034) 0.147 (0.045) -0.009 (0.028) 0.037 (0.039) -0.069 (0.024) -0.012 (0.020) 0.054 (0.037) 0.059 (0.029) 0.188 (0.043) -0.015 (0.031) 0.106 (0.046) -0.021 (0.048) -0.060 (0.055) 3.882 (0.049) -0.069 (0.042) -0.136 (0.034) 0.172 (0.024) 0.075 (0.040) 0.135 (0.030) 0.378 (0.043) 0.036 (0.028) 0.271 (0.033) -0.082 (0.024) 0.012 (0.017) 0.085 (0.032) -0.579 (0.026) -0.683 (0.036) -0.779 (0.028) -0.744 (0.040) -0.752 (0.042) -0.568 (0.052) 0.455 (0.040) Table B.6.1R Continued: Ordered logit estimates of Harvard's Academic and Extracurricular Ratings, baseline sample Model 1 AI Sq. X (AI>0) Model 2 1.200 (0.044) 0.428 (0.009) AI Sq. X (AI<0) Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspec Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Observations Pseudo R Sq. 142728 0.161 142728 0.547 Academic Model 3 Model 4 Model 5 Model 6 1.198 (0.044) 0.428 (0.009) -0.067 (0.047) -0.035 (0.038) -0.054 (0.064) 0.023 (0.047) -0.189 (0.067) -0.009 (0.089) -0.067 (0.073) 0.051 (0.043) -0.065 (0.040) -0.063 (0.035) -0.057 (0.060) -0.083 (0.058) -0.207 (0.058) -0.060 (0.059) -0.104 (0.106) 0.198 (0.075) 0.266 (0.073) 0.284 (0.065) 0.188 (0.103) 142728 0.547 1.136 (0.046) 0.430 (0.009) -0.036 (0.049) -0.048 (0.040) -0.068 (0.067) -0.047 (0.049) -0.146 (0.070) -0.071 (0.093) -0.080 (0.076) 0.097 (0.045) -0.051 (0.042) -0.068 (0.037) -0.066 (0.063) -0.120 (0.061) -0.262 (0.060) -0.092 (0.061) -0.008 (0.111) 0.182 (0.079) 0.238 (0.077) 0.218 (0.067) 0.177 (0.108) 136208 0.565 1.132 (0.046) 0.430 (0.009) -0.036 (0.049) -0.050 (0.040) -0.067 (0.067) -0.049 (0.049) -0.148 (0.070) -0.072 (0.093) -0.080 (0.076) 0.095 (0.045) -0.051 (0.042) -0.068 (0.037) -0.065 (0.063) -0.127 (0.061) -0.266 (0.060) -0.091 (0.061) -0.010 (0.111) 0.187 (0.079) 0.239 (0.077) 0.216 (0.067) 0.180 (0.108) 136208 0.565 1.187 (0.045) 0.441 (0.009) -0.059 (0.048) -0.031 (0.039) -0.061 (0.066) 0.019 (0.048) -0.149 (0.069) -0.059 (0.091) -0.056 (0.075) 0.064 (0.044) -0.070 (0.041) -0.063 (0.036) -0.049 (0.062) -0.121 (0.060) -0.226 (0.059) -0.060 (0.060) -0.046 (0.110) 0.183 (0.078) 0.258 (0.075) 0.255 (0.066) 0.187 (0.106) 136208 0.551 Model 1 Model 2 0.173 (0.025) 0.011 (0.007) 142728 0.041 142728 0.066 Extracurricular Model 3 Model 4 Model 5 Model 6 0.165 (0.025) 0.010 (0.007) -0.060 (0.041) 0.088 (0.034) 0.164 (0.055) 0.256 (0.041) 0.218 (0.059) 0.176 (0.076) 0.062 (0.070) 0.150 (0.044) 0.048 (0.039) 0.012 (0.029) 0.064 (0.050) 0.032 (0.059) 0.088 (0.057) -0.104 (0.054) -0.112 (0.097) 0.022 (0.073) -0.026 (0.068) 0.237 (0.048) 0.099 (0.076) 142728 0.066 0.064 (0.030) -0.020 (0.008) -0.006 (0.044) 0.095 (0.036) 0.199 (0.058) 0.234 (0.044) 0.299 (0.062) 0.203 (0.081) 0.081 (0.074) 0.216 (0.046) 0.079 (0.042) 0.002 (0.031) 0.010 (0.053) 0.106 (0.062) 0.076 (0.060) -0.073 (0.057) 0.020 (0.103) -0.011 (0.077) -0.065 (0.072) 0.158 (0.051) 0.111 (0.082) 136208 0.128 0.096 (0.030) -0.021 (0.008) -0.004 (0.044) 0.099 (0.036) 0.203 (0.058) 0.231 (0.044) 0.285 (0.062) 0.194 (0.081) 0.102 (0.074) 0.235 (0.047) 0.076 (0.042) 0.002 (0.031) 0.008 (0.054) 0.129 (0.062) 0.067 (0.061) -0.064 (0.057) 0.037 (0.104) -0.029 (0.077) -0.065 (0.073) 0.168 (0.052) 0.113 (0.082) 136208 0.138 0.181 (0.026) 0.006 (0.007) -0.036 (0.042) 0.100 (0.035) 0.179 (0.056) 0.278 (0.042) 0.237 (0.060) 0.182 (0.078) 0.099 (0.072) 0.178 (0.045) 0.042 (0.041) 0.006 (0.030) 0.036 (0.051) 0.014 (0.060) 0.062 (0.059) -0.098 (0.055) -0.048 (0.100) 0.034 (0.075) 0.003 (0.070) 0.233 (0.049) 0.125 (0.078) 136208 0.071 Table B.6.2R: Ordered logit estimates of Harvard's School Support Measures, baseline sample African American Hispanic Asian American Missing Female Disadvantaged First generation Early Decision Waiver Applied for Financial Aid Humanities Biology Physical Sciences Engineering Mathematics Computer Science Unspecified Academic index Model 1 -0.606 (0.024) -0.289 (0.021) -0.048 (0.015) -0.015 (0.025) -0.001 (0.012) 0.430 (0.024) 0.028 (0.028) 0.486 (0.021) -0.194 (0.022) -0.017 (0.015) 0.094 (0.021) -0.022 (0.017) 0.253 (0.026) 0.015 (0.020) 0.176 (0.028) 0.032 (0.033) -0.311 (0.038) Model 2 0.049 (0.028) -0.006 (0.023) -0.256 (0.016) -0.128 (0.026) 0.062 (0.013) 0.420 (0.025) 0.090 (0.028) 0.361 (0.021) 0.034 (0.023) -0.027 (0.016) 0.087 (0.021) -0.076 (0.018) 0.038 (0.027) -0.123 (0.021) 0.033 (0.029) -0.122 (0.034) -0.174 (0.039) 0.502 (0.043) Teacher 1 Model 3 Model 4 0.086 0.120 (0.043) (0.045) -0.030 -0.028 (0.033) (0.034) -0.280 -0.188 (0.022) (0.024) -0.149 -0.105 (0.038) (0.040) 0.117 0.152 (0.030) (0.031) 0.358 0.386 (0.042) (0.043) 0.084 0.029 (0.028) (0.029) 0.297 0.296 (0.033) (0.033) 0.035 -0.096 (0.023) (0.025) -0.026 -0.054 (0.016) (0.017) 0.161 0.167 (0.032) (0.034) -0.052 -0.066 (0.026) (0.027) 0.095 0.055 (0.034) (0.035) -0.122 -0.157 (0.028) (0.028) 0.116 0.069 (0.038) (0.039) -0.106 -0.109 (0.041) (0.042) -0.154 -0.167 (0.055) (0.056) 0.503 0.571 (0.043) (0.044) Model 5 0.012 (0.048) -0.023 (0.037) -0.159 (0.026) -0.080 (0.043) 0.093 (0.032) 0.188 (0.045) 0.017 (0.030) 0.153 (0.035) -0.076 (0.026) -0.041 (0.018) 0.134 (0.035) 0.023 (0.028) 0.124 (0.037) -0.038 (0.029) 0.124 (0.040) 0.016 (0.043) -0.071 (0.058) 0.151 (0.048) Model 6 -0.105 (0.049) -0.074 (0.037) -0.104 (0.026) -0.031 (0.044) 0.065 (0.033) 0.087 (0.045) 0.015 (0.031) 0.100 (0.035) -0.083 (0.026) -0.045 (0.018) 0.141 (0.035) 0.043 (0.028) 0.168 (0.037) -0.003 (0.030) 0.181 (0.041) 0.086 (0.044) -0.063 (0.058) 0.162 (0.049) Model 1 -0.551 (0.026) -0.256 (0.023) -0.086 (0.016) -0.063 (0.026) -0.027 (0.013) 0.453 (0.026) 0.000 (0.030) 0.531 (0.022) -0.197 (0.024) 0.000 (0.016) 0.077 (0.022) -0.058 (0.018) 0.244 (0.027) 0.010 (0.022) 0.178 (0.030) 0.016 (0.035) -0.230 (0.040) Model 2 0.082 (0.030) 0.011 (0.025) -0.289 (0.017) -0.169 (0.028) 0.041 (0.014) 0.438 (0.027) 0.070 (0.031) 0.400 (0.023) 0.033 (0.025) -0.009 (0.017) 0.062 (0.023) -0.105 (0.019) 0.035 (0.028) -0.119 (0.023) 0.043 (0.031) -0.136 (0.036) -0.094 (0.041) 0.500 (0.047) Teacher 2 Model 3 Model 4 0.173 0.211 (0.046) (0.048) 0.021 0.008 (0.035) (0.037) -0.328 -0.239 (0.024) (0.026) -0.180 -0.143 (0.040) (0.042) 0.124 0.142 (0.032) (0.033) 0.429 0.458 (0.045) (0.046) 0.064 0.012 (0.031) (0.033) 0.355 0.340 (0.035) (0.036) 0.035 -0.087 (0.026) (0.027) -0.009 -0.027 (0.017) (0.019) 0.175 0.173 (0.035) (0.036) -0.059 -0.081 (0.028) (0.029) 0.098 0.061 (0.036) (0.038) -0.112 -0.141 (0.030) (0.030) 0.145 0.085 (0.040) (0.041) -0.086 -0.098 (0.043) (0.044) -0.011 -0.042 (0.057) (0.059) 0.501 0.568 (0.047) (0.049) Model 5 0.104 (0.051) 0.024 (0.039) -0.203 (0.028) -0.115 (0.046) 0.085 (0.035) 0.278 (0.048) 0.003 (0.034) 0.182 (0.037) -0.069 (0.028) -0.010 (0.019) 0.139 (0.037) -0.001 (0.030) 0.126 (0.039) -0.018 (0.032) 0.132 (0.043) 0.024 (0.046) 0.073 (0.061) 0.149 (0.053) Model 6 -0.009 (0.052) -0.027 (0.040) -0.150 (0.028) -0.069 (0.046) 0.060 (0.035) 0.180 (0.049) 0.003 (0.034) 0.134 (0.038) -0.082 (0.028) -0.012 (0.019) 0.152 (0.038) 0.025 (0.030) 0.168 (0.039) 0.023 (0.032) 0.190 (0.043) 0.101 (0.046) 0.086 (0.062) 0.163 (0.053) Model 1 -0.577 (0.026) -0.289 (0.023) -0.054 (0.016) -0.048 (0.027) 0.032 (0.013) 0.451 (0.026) 0.027 (0.031) 0.623 (0.022) -0.177 (0.025) -0.107 (0.016) 0.048 (0.022) -0.046 (0.018) 0.195 (0.028) 0.033 (0.022) 0.180 (0.030) -0.059 (0.036) -0.316 (0.041) Model 2 0.187 (0.030) 0.034 (0.025) -0.263 (0.017) -0.159 (0.028) 0.105 (0.014) 0.429 (0.027) 0.107 (0.031) 0.488 (0.022) 0.103 (0.026) -0.128 (0.017) 0.047 (0.023) -0.117 (0.019) -0.048 (0.029) -0.134 (0.023) 0.000 (0.031) -0.244 (0.037) -0.143 (0.042) 0.545 (0.046) Counselor Model 3 Model 4 0.186 0.250 (0.046) (0.049) -0.007 0.021 (0.036) (0.037) -0.297 -0.155 (0.024) (0.026) -0.198 -0.136 (0.041) (0.043) 0.067 0.103 (0.032) (0.034) 0.341 0.400 (0.045) (0.047) 0.095 0.072 (0.032) (0.033) 0.389 0.386 (0.034) (0.035) 0.103 -0.011 (0.026) (0.027) -0.126 -0.068 (0.017) (0.018) 0.085 0.068 (0.035) (0.036) -0.144 -0.148 (0.028) (0.029) -0.055 -0.079 (0.037) (0.038) -0.195 -0.223 (0.030) (0.031) 0.005 -0.045 (0.040) (0.042) -0.266 -0.244 (0.044) (0.046) -0.162 -0.201 (0.058) (0.060) 0.546 0.583 (0.046) (0.048) Model 5 0.164 (0.052) 0.017 (0.040) -0.095 (0.028) -0.116 (0.047) 0.034 (0.035) 0.168 (0.049) 0.068 (0.034) 0.222 (0.037) 0.038 (0.029) -0.054 (0.019) 0.015 (0.038) -0.042 (0.030) 0.003 (0.040) -0.073 (0.032) 0.033 (0.044) -0.086 (0.048) -0.085 (0.063) 0.000 (0.053) Model 6 0.002 (0.054) -0.050 (0.042) -0.022 (0.029) -0.057 (0.048) -0.004 (0.036) 0.024 (0.050) 0.065 (0.035) 0.154 (0.038) 0.026 (0.029) -0.060 (0.020) 0.020 (0.039) -0.013 (0.031) 0.061 (0.041) -0.025 (0.033) 0.106 (0.045) 0.014 (0.049) -0.081 (0.064) 0.026 (0.054) Table B.6.2R Continued: Ordered logit estimates of Harvard's School Support Measures, baseline sample Model 1 Model 2 0.340 (0.027) 0.014 (0.010) 136958 0.03 136958 0.078 AI Sq. X (AI>0) AI Sq. X (AI<0) Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspec Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Observations Pseudo R Sq. Teacher 1 Model 3 Model 4 0.335 0.401 (0.027) (0.028) 0.014 0.006 (0.010) (0.010) -0.130 -0.134 (0.043) (0.044) -0.048 -0.044 (0.035) (0.036) -0.135 -0.122 (0.054) (0.056) 0.016 0.024 (0.042) (0.043) -0.196 -0.207 (0.058) (0.060) -0.005 -0.023 (0.077) (0.079) -0.042 -0.037 (0.077) (0.079) -0.095 -0.099 (0.048) (0.050) -0.039 -0.037 (0.041) (0.042) 0.023 0.021 (0.030) (0.030) 0.031 0.023 (0.050) (0.052) 0.093 0.044 (0.061) (0.064) 0.195 0.080 (0.057) (0.059) 0.015 0.008 (0.054) (0.055) -0.095 -0.002 (0.098) (0.100) 0.041 0.062 (0.078) (0.080) 0.060 0.126 (0.069) (0.071) 0.141 0.126 (0.049) (0.050) 0.112 0.086 (0.076) (0.078) 136958 130733 0.078 0.088 Model 5 0.215 (0.031) -0.010 (0.011) -0.093 (0.046) -0.045 (0.038) -0.132 (0.058) -0.023 (0.045) -0.177 (0.062) -0.016 (0.082) -0.029 (0.081) -0.068 (0.051) 0.007 (0.044) 0.033 (0.031) 0.006 (0.054) 0.122 (0.066) 0.093 (0.061) 0.017 (0.057) 0.041 (0.104) 0.008 (0.084) 0.085 (0.074) 0.031 (0.052) 0.072 (0.082) 130733 0.142 Model 6 0.247 (0.031) -0.012 (0.012) -0.090 (0.047) -0.042 (0.038) -0.132 (0.059) -0.032 (0.045) -0.199 (0.063) -0.040 (0.083) -0.010 (0.082) -0.030 (0.052) 0.010 (0.044) 0.027 (0.032) -0.004 (0.055) 0.180 (0.067) 0.105 (0.062) 0.022 (0.058) 0.057 (0.106) -0.009 (0.085) 0.090 (0.075) 0.055 (0.053) 0.071 (0.083) 130733 0.163 Model 1 Model 2 0.346 (0.029) 0.016 (0.012) 115618 0.029 115618 0.074 Teacher 2 Model 3 Model 4 0.345 0.400 (0.030) (0.031) 0.017 0.008 (0.012) (0.012) -0.201 -0.195 (0.046) (0.048) -0.091 -0.076 (0.038) (0.039) -0.139 -0.122 (0.058) (0.060) 0.009 0.025 (0.045) (0.046) -0.237 -0.221 (0.063) (0.064) -0.108 -0.127 (0.082) (0.085) -0.169 -0.122 (0.082) (0.084) -0.127 -0.126 (0.053) (0.054) -0.074 -0.080 (0.044) (0.046) 0.048 0.037 (0.032) (0.033) 0.061 0.083 (0.053) (0.055) -0.032 -0.096 (0.067) (0.070) 0.086 -0.014 (0.062) (0.064) 0.005 0.007 (0.057) (0.059) -0.147 -0.113 (0.105) (0.108) -0.062 -0.055 (0.086) (0.089) 0.064 0.138 (0.075) (0.078) 0.156 0.149 (0.052) (0.054) -0.070 -0.091 (0.083) (0.085) 115618 110195 0.075 0.083 Model 5 0.228 (0.033) -0.007 (0.014) -0.170 (0.049) -0.080 (0.040) -0.135 (0.062) -0.025 (0.048) -0.175 (0.067) -0.141 (0.088) -0.120 (0.087) -0.096 (0.056) -0.053 (0.047) 0.056 (0.034) 0.062 (0.057) -0.053 (0.072) -0.033 (0.066) 0.002 (0.061) -0.075 (0.112) -0.102 (0.092) 0.108 (0.081) 0.045 (0.056) -0.145 (0.089) 110195 0.137 Model 6 0.256 (0.034) -0.007 (0.014) -0.167 (0.050) -0.081 (0.041) -0.128 (0.062) -0.037 (0.048) -0.195 (0.067) -0.171 (0.089) -0.101 (0.088) -0.071 (0.057) -0.053 (0.048) 0.050 (0.034) 0.052 (0.058) 0.004 (0.073) -0.028 (0.067) 0.006 (0.062) -0.059 (0.113) -0.115 (0.094) 0.105 (0.082) 0.067 (0.057) -0.146 (0.090) 110195 0.156 Model 1 Model 2 0.307 (0.028) -0.018 (0.009) 134341 0.046 134341 0.103 Counselor Model 3 Model 4 0.303 0.362 (0.028) (0.030) -0.017 -0.022 (0.009) (0.010) -0.057 -0.070 (0.046) (0.048) 0.052 0.053 (0.038) (0.039) 0.010 0.021 (0.058) (0.060) 0.155 0.161 (0.045) (0.047) -0.020 -0.026 (0.062) (0.064) 0.061 0.036 (0.084) (0.086) 0.034 0.067 (0.082) (0.085) -0.031 -0.057 (0.052) (0.054) -0.017 -0.042 (0.045) (0.046) 0.007 -0.006 (0.032) (0.033) 0.097 0.099 (0.054) (0.056) 0.067 -0.046 (0.067) (0.069) 0.270 0.154 (0.062) (0.065) 0.096 0.108 (0.058) (0.060) -0.208 -0.145 (0.108) (0.111) 0.195 0.229 (0.082) (0.084) 0.001 0.022 (0.075) (0.078) 0.248 0.235 (0.051) (0.052) 0.064 0.038 (0.081) (0.083) 134341 128288 0.103 0.116 Model 5 0.211 (0.033) -0.072 (0.011) -0.016 (0.050) 0.062 (0.041) 0.038 (0.063) 0.123 (0.049) 0.041 (0.067) 0.042 (0.090) 0.113 (0.088) -0.012 (0.056) 0.003 (0.048) 0.000 (0.034) 0.098 (0.059) 0.009 (0.072) 0.186 (0.068) 0.126 (0.062) -0.128 (0.116) 0.213 (0.088) -0.003 (0.082) 0.171 (0.055) 0.025 (0.088) 128288 0.185 Model 6 0.249 (0.034) -0.071 (0.011) -0.005 (0.052) 0.066 (0.042) 0.046 (0.064) 0.113 (0.050) 0.013 (0.069) 0.001 (0.092) 0.149 (0.091) 0.038 (0.058) 0.004 (0.050) -0.008 (0.035) 0.091 (0.060) 0.101 (0.074) 0.205 (0.070) 0.141 (0.064) -0.116 (0.119) 0.192 (0.091) -0.005 (0.084) 0.201 (0.056) 0.045 (0.090) 128288 0.219 Table B.6.3R: Ordered logit estimates of Harvard's Personal Rating and Alumni Personal Rating, baseline sample African American Hispanic Asian American Missing Female Disadvantaged First generation Early Decision Waiver Applied for Financial Aid Humanities Biology Physical Sciences Engineering Mathematics Computer Science Unspecified Academic index Model 1 -0.108 (0.025) -0.075 (0.023) -0.346 (0.018) -0.237 (0.029) 0.170 (0.014) 0.754 (0.026) 0.010 (0.030) 0.648 (0.022) -0.182 (0.025) -0.129 (0.017) 0.010 (0.023) -0.230 (0.019) -0.207 (0.030) -0.233 (0.023) -0.262 (0.033) -0.523 (0.041) -0.414 (0.041) Personal Rating Model 2 Model 3 Model 4 0.481 0.693 0.706 (0.029) (0.043) (0.045) 0.174 0.188 0.213 (0.025) (0.036) (0.037) -0.486 -0.559 -0.482 (0.019) (0.027) (0.029) -0.324 -0.396 -0.369 (0.031) (0.047) (0.049) 0.224 0.207 0.201 (0.015) (0.033) (0.034) 0.744 0.745 0.791 (0.026) (0.044) (0.045) 0.074 0.059 0.034 (0.031) (0.031) (0.032) 0.557 0.505 0.504 (0.023) (0.035) (0.036) 0.019 0.021 -0.020 (0.025) (0.025) (0.027) -0.137 -0.140 -0.020 (0.018) (0.018) (0.020) 0.002 0.060 0.014 (0.023) (0.036) (0.037) -0.274 -0.281 -0.264 (0.020) (0.029) (0.030) -0.370 -0.378 -0.368 (0.031) (0.040) (0.042) -0.345 -0.409 -0.412 (0.024) (0.032) (0.032) -0.382 -0.420 -0.421 (0.034) (0.045) (0.046) -0.644 -0.717 -0.685 (0.042) (0.051) (0.052) -0.278 -0.189 -0.198 (0.041) (0.058) (0.059) 0.403 0.403 0.379 (0.047) (0.047) (0.048) Model 5 0.682 (0.053) 0.279 (0.044) -0.398 (0.034) -0.347 (0.056) 0.161 (0.039) 0.553 (0.052) 0.013 (0.036) 0.287 (0.041) 0.028 (0.031) 0.025 (0.023) -0.066 (0.043) -0.140 (0.034) -0.294 (0.047) -0.240 (0.037) -0.365 (0.052) -0.517 (0.058) -0.027 (0.067) -0.154 (0.057) Model 1 -0.132 (0.021) -0.111 (0.019) -0.010 (0.014) 0.019 (0.023) 0.177 (0.011) 0.172 (0.022) 0.048 (0.025) 0.251 (0.019) -0.032 (0.020) -0.056 (0.014) 0.001 (0.019) -0.169 (0.016) -0.149 (0.025) -0.194 (0.019) -0.230 (0.026) -0.295 (0.031) -0.432 (0.033) Model 2 0.300 (0.024) 0.073 (0.021) -0.140 (0.015) -0.057 (0.024) 0.237 (0.012) 0.146 (0.022) 0.098 (0.025) 0.177 (0.019) 0.133 (0.020) -0.051 (0.014) -0.010 (0.019) -0.196 (0.016) -0.289 (0.025) -0.276 (0.019) -0.334 (0.027) -0.403 (0.031) -0.313 (0.033) 0.442 (0.036) Alumni Personal Rating Model 3 Model 4 Model 5 0.421 0.463 0.236 (0.035) (0.037) (0.041) 0.071 0.073 0.062 (0.029) (0.030) (0.034) -0.166 -0.100 -0.181 (0.021) (0.022) (0.025) -0.057 -0.040 -0.129 (0.036) (0.036) (0.041) 0.203 0.205 0.240 (0.027) (0.028) (0.032) 0.105 0.113 -0.075 (0.039) (0.039) (0.044) 0.091 0.074 0.036 (0.025) (0.025) (0.028) 0.166 0.165 0.120 (0.029) (0.030) (0.034) 0.135 0.087 0.061 (0.020) (0.022) (0.024) -0.049 -0.006 0.006 (0.014) (0.016) (0.018) 0.022 0.009 0.000 (0.030) (0.030) (0.034) -0.223 -0.221 -0.140 (0.024) (0.024) (0.027) -0.331 -0.337 -0.345 (0.032) (0.033) (0.037) -0.334 -0.333 -0.224 (0.025) (0.025) (0.029) -0.365 -0.373 -0.387 (0.035) (0.036) (0.041) -0.483 -0.478 -0.460 (0.037) (0.038) (0.042) -0.275 -0.271 -0.006 (0.047) (0.048) (0.053) 0.442 0.427 -0.395 (0.036) (0.037) (0.044) Model 6 0.207 (0.041) 0.051 (0.034) -0.165 (0.025) -0.116 (0.041) 0.232 (0.032) -0.105 (0.044) 0.034 (0.028) 0.102 (0.034) 0.060 (0.024) 0.005 (0.018) 0.000 (0.034) -0.135 (0.027) -0.332 (0.037) -0.215 (0.029) -0.371 (0.041) -0.439 (0.042) -0.006 (0.053) -0.391 (0.044) Table B.6.3R Continued: Ordered logit estimates of Harvard's Personal Rating and Alumni Personal Rating, baseline sample Model 1 AI Sq. X (AI>0) AI Sq. X (AI<0) Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspec Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Observations Pseudo R Sq. 142728 0.06 Personal Rating Model 2 Model 3 Model 4 0.007 0.018 0.014 (0.030) (0.030) (0.031) 0.014 0.016 0.015 (0.010) (0.010) (0.010) -0.091 -0.070 (0.047) (0.048) 0.015 0.009 (0.039) (0.040) 0.018 0.002 (0.063) (0.064) 0.154 0.160 (0.047) (0.048) 0.093 0.088 (0.067) (0.069) 0.262 0.220 (0.090) (0.092) -0.178 -0.166 (0.082) (0.084) -0.258 -0.265 (0.049) (0.050) -0.074 -0.077 (0.044) (0.045) 0.084 0.085 (0.035) (0.036) 0.130 0.133 (0.059) (0.061) -0.231 -0.292 (0.061) (0.063) 0.096 0.027 (0.059) (0.060) 0.118 0.101 (0.057) (0.058) -0.089 -0.032 (0.104) (0.106) 0.126 0.103 (0.075) (0.078) -0.007 0.007 (0.071) (0.073) 0.114 0.090 (0.054) (0.055) 0.061 0.024 (0.084) (0.087) 142728 142728 136208 0.088 0.089 0.097 Model 5 -0.189 (0.038) 0.000 (0.012) 0.003 (0.056) -0.004 (0.046) -0.002 (0.073) 0.072 (0.055) 0.141 (0.078) 0.245 (0.104) -0.184 (0.095) -0.239 (0.057) -0.015 (0.051) 0.095 (0.040) 0.118 (0.069) -0.324 (0.073) -0.048 (0.070) 0.058 (0.067) 0.068 (0.123) 0.050 (0.091) -0.071 (0.085) -0.059 (0.064) 0.029 (0.101) 136208 0.289 Model 1 111524 0.012 Alumni Personal Rating Model 3 Model 4 Model 5 0.184 0.208 -0.183 (0.024) (0.024) (0.031) 0.021 0.019 -0.018 (0.007) (0.007) (0.009) -0.044 -0.036 -0.019 (0.039) (0.040) (0.045) 0.053 0.058 0.001 (0.032) (0.032) (0.037) 0.098 0.102 0.095 (0.050) (0.051) (0.058) 0.136 0.136 -0.008 (0.038) (0.039) (0.044) 0.068 0.072 0.167 (0.054) (0.055) (0.062) 0.282 0.270 0.261 (0.071) (0.072) (0.081) -0.079 -0.089 -0.190 (0.065) (0.066) (0.074) -0.172 -0.165 -0.066 (0.040) (0.040) (0.045) -0.034 -0.038 -0.021 (0.036) (0.037) (0.041) 0.023 0.014 0.053 (0.027) (0.028) (0.031) 0.028 0.040 0.034 (0.046) (0.047) (0.054) -0.002 0.004 0.101 (0.053) (0.054) (0.061) 0.139 0.124 0.174 (0.052) (0.053) (0.060) 0.033 0.029 0.087 (0.049) (0.050) (0.056) -0.010 0.005 0.078 (0.089) (0.090) (0.101) -0.092 -0.078 -0.057 (0.063) (0.064) (0.073) -0.070 -0.052 -0.128 (0.060) (0.062) (0.070) 0.121 0.105 0.015 (0.044) (0.045) (0.052) -0.100 -0.093 -0.127 (0.070) (0.072) (0.082) 111524 111524 108054 108054 0.027 0.028 0.031 0.341 Model 2 0.183 (0.024) 0.021 (0.007) Model 6 -0.169 (0.031) -0.019 (0.009) -0.018 (0.045) 0.002 (0.037) 0.096 (0.058) -0.010 (0.044) 0.163 (0.062) 0.254 (0.081) -0.181 (0.074) -0.056 (0.046) -0.022 (0.041) 0.050 (0.031) 0.031 (0.054) 0.122 (0.061) 0.181 (0.060) 0.089 (0.056) 0.077 (0.101) -0.053 (0.073) -0.123 (0.070) 0.023 (0.052) -0.125 (0.082) 108054 0.342 Table B.6.4R: Ordered logit estimates of Harvard's Overall Rating and Alumni Overall Rating, baseline sample African American Hispanic Asian American Missing Female Disadvantaged First generation Early Decision Waiver Applied for Financial Aid Humanities Biology Physical Sciences Engineering Mathematics Computer Science Unspecified Academic index Model 1 -0.821 (0.019) -0.237 (0.016) 0.160 (0.012) 0.095 (0.020) -0.017 (0.010) 0.603 (0.019) -0.170 (0.021) 0.683 (0.018) -0.533 (0.017) -0.090 (0.013) 0.075 (0.017) -0.078 (0.014) 0.266 (0.021) 0.104 (0.016) 0.125 (0.023) 0.019 (0.027) -0.529 (0.027) Model 2 0.950 (0.022) 0.549 (0.019) -0.238 (0.014) -0.117 (0.022) 0.196 (0.011) 0.650 (0.020) 0.016 (0.022) 0.564 (0.018) 0.034 (0.018) -0.091 (0.013) 0.061 (0.018) -0.205 (0.014) -0.174 (0.022) -0.218 (0.017) -0.228 (0.024) -0.340 (0.028) -0.216 (0.029) 1.605 (0.032) Overall Rating Model 3 Model 4 1.137 1.212 (0.032) (0.034) 0.621 0.683 (0.026) (0.027) -0.290 -0.179 (0.019) (0.020) -0.133 -0.098 (0.032) (0.033) 0.180 0.176 (0.025) (0.025) 0.852 0.900 (0.035) (0.036) 0.009 -0.004 (0.022) (0.023) 0.466 0.470 (0.027) (0.028) 0.032 -0.010 (0.018) (0.019) -0.099 -0.013 (0.013) (0.014) 0.085 0.051 (0.027) (0.028) -0.208 -0.207 (0.021) (0.022) -0.210 -0.221 (0.029) (0.030) -0.275 -0.281 (0.022) (0.023) -0.208 -0.231 (0.032) (0.033) -0.373 -0.361 (0.033) (0.034) -0.219 -0.226 (0.040) (0.041) 1.615 1.609 (0.032) (0.033) Model 5 1.503 (0.038) 0.922 (0.030) -0.136 (0.022) -0.086 (0.036) 0.117 (0.027) 0.743 (0.038) 0.004 (0.024) 0.291 (0.029) 0.079 (0.020) 0.013 (0.015) 0.005 (0.029) -0.073 (0.023) -0.069 (0.031) -0.064 (0.024) -0.109 (0.034) -0.109 (0.035) -0.076 (0.043) 0.497 (0.037) Model 6 1.451 (0.038) 0.906 (0.030) -0.081 (0.022) -0.035 (0.036) 0.092 (0.027) 0.671 (0.038) 0.008 (0.024) 0.249 (0.029) 0.078 (0.021) 0.012 (0.015) 0.014 (0.029) -0.055 (0.023) -0.026 (0.031) -0.035 (0.024) -0.058 (0.034) -0.043 (0.035) -0.073 (0.044) 0.514 (0.037) Model 1 -0.664 (0.020) -0.358 (0.019) 0.232 (0.014) 0.187 (0.023) -0.027 (0.011) 0.191 (0.021) -0.023 (0.024) 0.291 (0.019) -0.237 (0.020) -0.073 (0.014) 0.021 (0.019) -0.066 (0.016) 0.183 (0.024) 0.013 (0.018) 0.077 (0.026) 0.052 (0.031) -0.536 (0.032) Model 2 0.254 (0.024) 0.030 (0.021) -0.042 (0.015) 0.039 (0.024) 0.133 (0.012) 0.143 (0.022) 0.091 (0.025) 0.153 (0.019) 0.121 (0.020) -0.050 (0.014) 0.010 (0.019) -0.127 (0.016) -0.124 (0.025) -0.184 (0.019) -0.165 (0.027) -0.182 (0.031) -0.328 (0.033) 0.935 (0.036) Alumni Overall Rating Model 3 Model 4 0.373 0.412 (0.035) (0.037) 0.037 0.048 (0.029) (0.030) -0.045 0.025 (0.021) (0.022) 0.042 0.070 (0.035) (0.036) 0.119 0.112 (0.027) (0.028) 0.159 0.165 (0.038) (0.039) 0.087 0.078 (0.025) (0.025) 0.112 0.114 (0.029) (0.030) 0.123 0.071 (0.020) (0.021) -0.050 -0.005 (0.014) (0.016) 0.046 0.027 (0.029) (0.030) -0.160 -0.159 (0.023) (0.024) -0.139 -0.151 (0.032) (0.033) -0.256 -0.254 (0.025) (0.025) -0.146 -0.155 (0.035) (0.036) -0.227 -0.226 (0.037) (0.038) -0.351 -0.352 (0.046) (0.047) 0.934 0.931 (0.036) (0.037) Model 5 0.126 (0.040) 0.001 (0.033) 0.160 (0.024) 0.165 (0.040) -0.094 (0.031) 0.068 (0.043) 0.030 (0.028) -0.047 (0.033) 0.031 (0.024) 0.003 (0.017) -0.003 (0.033) 0.035 (0.026) 0.168 (0.036) 0.041 (0.028) 0.188 (0.040) 0.228 (0.042) -0.207 (0.052) 0.750 (0.043) Model 6 0.126 (0.040) 0.000 (0.033) 0.159 (0.024) 0.165 (0.040) -0.094 (0.031) 0.068 (0.043) 0.030 (0.028) -0.047 (0.033) 0.031 (0.024) 0.003 (0.017) -0.002 (0.033) 0.035 (0.026) 0.168 (0.036) 0.041 (0.028) 0.188 (0.040) 0.228 (0.042) -0.207 (0.052) 0.750 (0.043) Table B.6.4R Continued: Ordered logit estimates of Harvard's Overall Rating and Alumni Overall Rating, baseline sample Model 1 Model 2 -0.177 (0.022) 0.079 (0.007) 142701 0.059 142701 0.195 AI Sq. X (AI>0) AI Sq. X (AI<0) Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspec Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Observations Pseudo R Sq. Overall Rating Model 3 Model 4 -0.167 -0.136 (0.022) (0.023) 0.084 0.088 (0.007) (0.007) -0.035 -0.020 (0.036) (0.037) 0.010 0.020 (0.029) (0.029) 0.087 0.101 (0.046) (0.047) 0.150 0.160 (0.034) (0.035) -0.055 -0.033 (0.049) (0.050) 0.120 0.077 (0.063) (0.064) 0.009 0.014 (0.056) (0.058) -0.140 -0.155 (0.036) (0.037) -0.083 -0.093 (0.032) (0.033) 0.030 0.026 (0.025) (0.025) 0.041 0.023 (0.042) (0.043) -0.637 -0.670 (0.049) (0.050) -0.303 -0.356 (0.047) (0.048) 0.083 0.059 (0.045) (0.046) -0.252 -0.218 (0.082) (0.084) 0.249 0.242 (0.061) (0.062) 0.098 0.091 (0.056) (0.057) 0.180 0.157 (0.042) (0.043) 0.069 0.071 (0.066) (0.068) 142701 136183 0.196 0.201 Model 5 -0.136 (0.026) 0.082 (0.008) 0.033 (0.038) 0.002 (0.031) 0.097 (0.049) 0.063 (0.037) -0.030 (0.053) 0.064 (0.067) 0.065 (0.061) -0.163 (0.040) -0.013 (0.035) 0.040 (0.026) 0.011 (0.045) -0.684 (0.053) -0.353 (0.051) 0.100 (0.048) -0.155 (0.088) 0.200 (0.065) 0.007 (0.060) -0.006 (0.044) 0.056 (0.071) 136183 0.331 Model 6 -0.084 (0.026) 0.083 (0.008) 0.036 (0.039) 0.006 (0.031) 0.102 (0.049) 0.065 (0.037) -0.043 (0.053) 0.034 (0.068) 0.086 (0.061) -0.130 (0.040) -0.008 (0.035) 0.034 (0.027) 0.001 (0.045) -0.646 (0.053) -0.344 (0.051) 0.105 (0.049) -0.161 (0.088) 0.196 (0.066) 0.015 (0.060) 0.015 (0.045) 0.063 (0.072) 136183 0.346 Model 1 111524 0.035 Alumni Overall Rating Model 3 Model 4 0.360 0.388 (0.024) (0.025) 0.021 0.019 (0.007) (0.007) -0.054 -0.044 (0.039) (0.040) 0.062 0.069 (0.032) (0.032) 0.028 0.040 (0.050) (0.051) 0.179 0.180 (0.038) (0.038) -0.058 -0.051 (0.054) (0.054) 0.140 0.119 (0.070) (0.071) 0.040 0.034 (0.064) (0.066) -0.170 -0.162 (0.039) (0.040) -0.043 -0.053 (0.036) (0.037) -0.023 -0.029 (0.027) (0.028) 0.006 0.002 (0.046) (0.047) -0.070 -0.050 (0.052) (0.053) 0.046 0.031 (0.051) (0.053) -0.036 -0.040 (0.049) (0.049) -0.056 -0.021 (0.087) (0.089) -0.048 -0.043 (0.063) (0.064) -0.007 0.022 (0.060) (0.061) 0.156 0.132 (0.044) (0.045) -0.019 -0.016 (0.070) (0.071) 111524 111524 108054 0.095 0.096 0.1 Model 2 0.362 (0.024) 0.020 (0.007) Model 5 0.339 (0.030) -0.016 (0.009) 0.004 (0.044) 0.045 (0.035) -0.025 (0.056) 0.112 (0.042) -0.137 (0.060) -0.110 (0.079) 0.144 (0.072) -0.085 (0.044) -0.014 (0.040) -0.062 (0.030) -0.041 (0.052) -0.066 (0.059) -0.077 (0.058) -0.060 (0.054) -0.071 (0.098) 0.006 (0.071) 0.090 (0.067) 0.038 (0.050) 0.047 (0.080) 108054 0.375 Model 6 0.339 (0.030) -0.016 (0.009) 0.004 (0.044) 0.044 (0.035) -0.026 (0.056) 0.111 (0.042) -0.139 (0.060) -0.109 (0.079) 0.144 (0.072) -0.085 (0.044) -0.014 (0.040) -0.062 (0.030) -0.041 (0.052) -0.066 (0.059) -0.076 (0.058) -0.059 (0.054) -0.070 (0.098) 0.005 (0.071) 0.090 (0.067) 0.038 (0.050) 0.048 (0.080) 108054 0.375 Table B.6.5R: Ordered logit estimates of Harvard's Academic and Extracurricular Ratings, expanded sample African American Hispanic Asian American Missing Female Disadvantaged First generation Early Decision Legacy Double Legacy Faculty or Staff Child Dean's/Director's List Waiver Applied for Financial Aid Humanities Biology Physical Sciences Engineering Mathematics Computer Science Unspecified Academic index AI Sq. X (AI>0) AI Sq. X (AI<0) Model 1 -1.693 (0.019) -0.941 (0.017) 0.607 (0.014) 0.303 (0.022) -0.268 (0.011) 0.132 (0.019) -0.209 (0.022) 0.478 (0.019) -0.232 (0.035) 0.358 (0.082) 0.369 (0.124) 0.029 (0.044) -0.714 (0.018) -0.136 (0.014) 0.047 (0.017) 0.109 (0.014) 0.743 (0.024) 0.398 (0.017) 0.517 (0.025) 0.447 (0.030) -0.482 (0.028) Model 2 0.084 (0.027) -0.190 (0.024) 0.013 (0.019) 0.051 (0.031) 0.108 (0.014) 0.029 (0.024) -0.019 (0.027) 0.187 (0.024) 0.016 (0.046) 0.111 (0.105) 0.360 (0.158) 0.179 (0.058) -0.061 (0.022) -0.095 (0.018) 0.042 (0.022) 0.028 (0.019) 0.162 (0.031) -0.019 (0.023) 0.046 (0.033) -0.063 (0.039) -0.091 (0.036) 3.856 (0.046) 1.197 (0.043) 0.432 (0.009) Academic Model 3 Model 4 0.043 0.049 (0.039) (0.041) -0.140 -0.098 (0.033) (0.034) 0.025 0.086 (0.027) (0.029) 0.068 0.055 (0.045) (0.047) 0.170 0.154 (0.032) (0.033) 0.123 0.156 (0.044) (0.045) -0.015 -0.008 (0.027) (0.028) 0.060 0.069 (0.036) (0.037) 0.045 -0.126 (0.055) (0.067) 0.117 0.117 (0.105) (0.109) 0.450 0.481 (0.289) (0.291) 0.295 0.268 (0.339) (0.342) -0.063 -0.071 (0.022) (0.024) -0.096 -0.020 (0.018) (0.020) 0.084 0.066 (0.035) (0.036) 0.049 0.070 (0.028) (0.029) 0.192 0.207 (0.041) (0.042) -0.020 -0.008 (0.030) (0.031) 0.140 0.126 (0.045) (0.046) -0.050 -0.022 (0.046) (0.047) -0.053 -0.057 (0.052) (0.053) 3.860 3.895 (0.046) (0.048) 1.195 1.183 (0.043) (0.045) 0.431 0.445 (0.009) (0.009) Model 5 -0.006 (0.043) -0.109 (0.037) 0.131 (0.031) 0.075 (0.051) 0.118 (0.034) 0.056 (0.046) -0.015 (0.028) -0.009 (0.038) -0.160 (0.068) 0.113 (0.111) 0.470 (0.295) 0.175 (0.347) -0.055 (0.024) -0.016 (0.020) 0.039 (0.036) 0.110 (0.029) 0.242 (0.043) 0.072 (0.031) 0.153 (0.046) 0.030 (0.048) 0.027 (0.054) 3.757 (0.049) 1.127 (0.046) 0.434 (0.009) Model 6 0.000 (0.043) -0.105 (0.037) 0.128 (0.031) 0.069 (0.051) 0.119 (0.034) 0.063 (0.046) -0.014 (0.028) -0.005 (0.038) -0.154 (0.068) 0.115 (0.111) 0.461 (0.295) 0.178 (0.346) -0.055 (0.024) -0.017 (0.020) 0.039 (0.036) 0.109 (0.029) 0.240 (0.043) 0.071 (0.031) 0.152 (0.046) 0.026 (0.048) 0.025 (0.054) 3.757 (0.049) 1.122 (0.046) 0.435 (0.009) Model 1 -0.499 (0.023) -0.303 (0.021) 0.241 (0.015) 0.123 (0.024) 0.206 (0.012) 0.372 (0.024) -0.013 (0.027) 0.462 (0.019) 0.072 (0.038) 0.045 (0.082) 0.005 (0.123) 0.235 (0.046) -0.250 (0.022) -0.063 (0.015) 0.079 (0.019) -0.497 (0.017) -0.436 (0.026) -0.558 (0.020) -0.514 (0.028) -0.569 (0.033) -0.640 (0.034) Model 2 0.037 (0.026) -0.104 (0.023) 0.115 (0.016) 0.064 (0.025) 0.257 (0.013) 0.344 (0.024) 0.049 (0.027) 0.353 (0.020) 0.146 (0.038) -0.024 (0.083) -0.020 (0.124) 0.275 (0.047) -0.047 (0.022) -0.068 (0.015) 0.068 (0.019) -0.533 (0.017) -0.597 (0.026) -0.662 (0.020) -0.628 (0.029) -0.688 (0.033) -0.505 (0.034) 0.485 (0.039) 0.180 (0.025) 0.011 (0.007) Extracurricular Model 3 Model 4 -0.059 -0.059 (0.040) (0.042) -0.137 -0.126 (0.032) (0.033) 0.106 0.178 (0.022) (0.024) 0.038 0.075 (0.038) (0.039) 0.155 0.140 (0.028) (0.029) 0.346 0.372 (0.042) (0.043) 0.050 0.035 (0.027) (0.028) 0.290 0.287 (0.030) (0.030) 0.205 0.148 (0.045) (0.053) -0.020 -0.078 (0.083) (0.086) -0.185 -0.160 (0.225) (0.227) 0.085 0.070 (0.264) (0.266) -0.047 -0.079 (0.022) (0.023) -0.065 0.008 (0.015) (0.017) 0.113 0.089 (0.030) (0.031) -0.574 -0.580 (0.025) (0.026) -0.676 -0.692 (0.034) (0.035) -0.767 -0.785 (0.027) (0.028) -0.727 -0.744 (0.038) (0.039) -0.752 -0.753 (0.040) (0.041) -0.542 -0.574 (0.049) (0.050) 0.486 0.475 (0.039) (0.040) 0.172 0.189 (0.025) (0.026) 0.010 0.007 (0.007) (0.007) Model 5 -0.208 (0.044) -0.136 (0.036) 0.175 (0.026) 0.076 (0.042) 0.024 (0.030) 0.196 (0.044) 0.025 (0.029) 0.204 (0.032) 0.112 (0.056) -0.024 (0.090) -0.071 (0.238) 0.086 (0.279) -0.099 (0.024) -0.002 (0.017) 0.021 (0.032) -0.544 (0.026) -0.718 (0.036) -0.695 (0.028) -0.768 (0.040) -0.760 (0.042) -0.477 (0.052) 0.095 (0.043) 0.065 (0.029) -0.019 (0.008) Model 6 -0.283 (0.044) -0.167 (0.036) 0.200 (0.026) 0.098 (0.042) 0.005 (0.030) 0.126 (0.045) 0.023 (0.029) 0.181 (0.032) 0.087 (0.056) -0.052 (0.090) -0.061 (0.239) 0.055 (0.280) -0.101 (0.024) -0.006 (0.017) 0.029 (0.032) -0.527 (0.027) -0.689 (0.037) -0.670 (0.029) -0.731 (0.040) -0.718 (0.042) -0.469 (0.052) 0.096 (0.044) 0.096 (0.029) -0.020 (0.008) Table B.6.5R Continued: Ordered logit estimates of Harvard's Academic and Extracurricular Ratings, expanded sample Model 1 Model 2 148769 0.161 148769 0.547 Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspec Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Legacy X African American Legacy X Hispanic Legacy X Asian American Legacy X Missing Special X African American Special X Hispanic Special X Asian American Special X Missing Observations Pseudo R Sq. Academic Model 3 Model 4 -0.075 -0.068 (0.045) (0.047) -0.041 -0.038 (0.037) (0.038) -0.067 -0.076 (0.063) (0.065) 0.024 0.020 (0.046) (0.047) -0.214 -0.181 (0.066) (0.068) -0.011 -0.064 (0.088) (0.090) -0.071 -0.062 (0.071) (0.073) 0.053 0.068 (0.043) (0.044) -0.075 -0.078 (0.039) (0.041) -0.067 -0.065 (0.035) (0.036) -0.053 -0.044 (0.058) (0.060) -0.086 -0.128 (0.058) (0.060) -0.212 -0.234 (0.057) (0.059) -0.061 -0.064 (0.058) (0.060) -0.098 -0.044 (0.106) (0.109) 0.175 0.159 (0.073) (0.075) 0.238 0.230 (0.070) (0.073) 0.255 0.231 (0.062) (0.064) 0.191 0.171 (0.096) (0.099) -0.213 -0.193 (0.165) (0.170) -0.128 -0.150 (0.164) (0.170) 0.035 0.020 (0.141) (0.149) 0.007 0.060 (0.153) (0.160) -0.275 -0.268 (0.257) (0.267) 0.057 -0.066 (0.212) (0.222) 0.017 0.051 (0.172) (0.187) -0.183 -0.225 (0.193) (0.202) 148769 141852 0.547 0.552 Model 5 -0.042 (0.047) -0.054 (0.039) -0.085 (0.066) -0.046 (0.048) -0.176 (0.069) -0.081 (0.091) -0.085 (0.074) 0.102 (0.045) -0.057 (0.041) -0.067 (0.036) -0.062 (0.061) -0.127 (0.061) -0.270 (0.060) -0.097 (0.061) -0.008 (0.111) 0.159 (0.077) 0.220 (0.074) 0.198 (0.065) 0.168 (0.101) -0.206 (0.172) -0.154 (0.172) 0.026 (0.153) 0.112 (0.163) -0.359 (0.271) -0.072 (0.226) 0.031 (0.190) -0.251 (0.205) 141852 0.565 Model 6 -0.042 (0.048) -0.057 (0.039) -0.085 (0.066) -0.048 (0.048) -0.178 (0.069) -0.081 (0.091) -0.085 (0.074) 0.100 (0.045) -0.058 (0.041) -0.067 (0.036) -0.061 (0.061) -0.135 (0.061) -0.275 (0.060) -0.096 (0.061) -0.010 (0.111) 0.164 (0.077) 0.219 (0.074) 0.194 (0.065) 0.167 (0.101) -0.200 (0.172) -0.153 (0.172) 0.030 (0.153) 0.104 (0.164) -0.359 (0.271) -0.071 (0.226) 0.045 (0.190) -0.241 (0.205) 141852 0.566 Model 1 Model 2 148769 0.042 148769 0.067 Extracurricular Model 3 Model 4 -0.058 -0.036 (0.039) (0.040) 0.081 0.092 (0.033) (0.034) 0.183 0.198 (0.053) (0.054) 0.253 0.276 (0.040) (0.041) 0.219 0.232 (0.058) (0.059) 0.176 0.179 (0.074) (0.076) 0.073 0.109 (0.068) (0.070) 0.134 0.163 (0.044) (0.045) 0.034 0.027 (0.039) (0.040) 0.005 -0.001 (0.029) (0.029) 0.063 0.036 (0.048) (0.050) 0.037 0.021 (0.058) (0.060) 0.093 0.071 (0.057) (0.059) -0.099 -0.091 (0.054) (0.055) -0.087 -0.023 (0.096) (0.099) 0.021 0.033 (0.070) (0.072) -0.054 -0.030 (0.065) (0.067) 0.203 0.196 (0.046) (0.047) 0.068 0.086 (0.071) (0.073) 0.160 0.145 (0.155) (0.160) -0.097 -0.108 (0.144) (0.150) -0.283 -0.328 (0.110) (0.116) -0.244 -0.207 (0.123) (0.128) 0.205 0.085 (0.243) (0.253) 0.120 0.123 (0.184) (0.193) 0.200 0.244 (0.130) (0.139) 0.121 0.112 (0.153) (0.159) 148769 141852 0.068 0.073 Model 5 -0.003 (0.042) 0.089 (0.035) 0.212 (0.056) 0.233 (0.043) 0.295 (0.061) 0.191 (0.079) 0.094 (0.072) 0.201 (0.046) 0.062 (0.041) -0.003 (0.030) 0.009 (0.052) 0.114 (0.062) 0.086 (0.060) -0.066 (0.057) 0.044 (0.102) -0.017 (0.074) -0.094 (0.069) 0.124 (0.049) 0.077 (0.077) 0.198 (0.167) -0.122 (0.155) -0.305 (0.121) -0.185 (0.134) -0.023 (0.262) 0.026 (0.198) 0.148 (0.145) 0.021 (0.166) 141852 0.131 Model 6 -0.001 (0.042) 0.094 (0.036) 0.217 (0.057) 0.231 (0.043) 0.282 (0.061) 0.183 (0.080) 0.112 (0.072) 0.223 (0.046) 0.062 (0.041) -0.003 (0.031) 0.005 (0.052) 0.137 (0.062) 0.078 (0.061) -0.057 (0.057) 0.053 (0.103) -0.039 (0.075) -0.094 (0.070) 0.133 (0.049) 0.082 (0.077) 0.137 (0.169) -0.104 (0.156) -0.304 (0.122) -0.174 (0.134) 0.001 (0.265) 0.054 (0.199) 0.159 (0.145) -0.009 (0.167) 141852 0.14 Table B.6.6R: Ordered logit estimates of Harvard's School Support Measures, expanded sample African American Hispanic Asian American Missing Female Disadvantaged First generation Early Decision Legacy Double Legacy Faculty or Staff Child Dean's/Director's List Waiver Applied for Financial Aid Humanities Biology Physical Sciences Engineering Mathematics Computer Science Unspecified Academic index AI Sq. X (AI>0) AI Sq. X (AI<0) Model 1 -0.602 (0.024) -0.289 (0.021) -0.049 (0.015) -0.006 (0.024) 0.003 (0.012) 0.433 (0.024) 0.028 (0.027) 0.511 (0.019) -0.026 (0.037) 0.112 (0.081) 0.109 (0.119) 0.132 (0.046) -0.195 (0.022) -0.019 (0.015) 0.106 (0.020) -0.012 (0.017) 0.263 (0.025) 0.021 (0.020) 0.190 (0.027) 0.031 (0.032) -0.289 (0.036) Model 2 0.059 (0.028) -0.008 (0.023) -0.255 (0.016) -0.114 (0.025) 0.064 (0.013) 0.424 (0.025) 0.090 (0.028) 0.373 (0.020) 0.104 (0.039) 0.002 (0.083) 0.089 (0.123) 0.225 (0.047) 0.033 (0.023) -0.030 (0.015) 0.102 (0.021) -0.067 (0.017) 0.044 (0.026) -0.118 (0.021) 0.042 (0.028) -0.127 (0.034) -0.161 (0.037) 0.523 (0.042) 0.334 (0.027) 0.018 (0.010) Teacher 1 Model 3 Model 4 0.093 0.127 (0.042) (0.044) -0.021 -0.021 (0.032) (0.034) -0.275 -0.182 (0.022) (0.024) -0.149 -0.104 (0.038) (0.039) 0.120 0.152 (0.029) (0.030) 0.364 0.394 (0.041) (0.043) 0.083 0.028 (0.028) (0.029) 0.336 0.334 (0.030) (0.031) 0.074 -0.016 (0.045) (0.054) 0.008 -0.061 (0.083) (0.087) 0.029 0.054 (0.220) (0.222) 0.120 0.144 (0.259) (0.262) 0.035 -0.099 (0.023) (0.025) -0.028 -0.054 (0.015) (0.017) 0.175 0.180 (0.031) (0.032) -0.044 -0.058 (0.025) (0.026) 0.095 0.056 (0.033) (0.034) -0.120 -0.156 (0.027) (0.028) 0.123 0.074 (0.037) (0.038) -0.117 -0.122 (0.040) (0.041) -0.143 -0.156 (0.053) (0.054) 0.524 0.588 (0.042) (0.044) 0.329 0.394 (0.027) (0.028) 0.019 0.010 (0.010) (0.010) Model 5 0.015 (0.047) -0.018 (0.036) -0.152 (0.026) -0.078 (0.042) 0.092 (0.031) 0.197 (0.044) 0.015 (0.030) 0.185 (0.032) -0.068 (0.056) -0.048 (0.090) 0.033 (0.230) 0.083 (0.272) -0.079 (0.026) -0.037 (0.017) 0.143 (0.033) 0.030 (0.027) 0.126 (0.036) -0.036 (0.029) 0.129 (0.039) 0.005 (0.043) -0.060 (0.056) 0.162 (0.047) 0.202 (0.030) -0.008 (0.011) Model 6 -0.103 (0.048) -0.069 (0.037) -0.098 (0.026) -0.030 (0.043) 0.064 (0.032) 0.095 (0.045) 0.013 (0.031) 0.136 (0.032) -0.121 (0.057) -0.075 (0.091) 0.021 (0.234) 0.030 (0.275) -0.086 (0.026) -0.041 (0.018) 0.151 (0.034) 0.050 (0.028) 0.171 (0.036) 0.000 (0.029) 0.187 (0.040) 0.078 (0.043) -0.047 (0.056) 0.170 (0.048) 0.235 (0.031) -0.011 (0.011) Model 1 -0.551 (0.026) -0.256 (0.022) -0.084 (0.016) -0.055 (0.025) -0.024 (0.013) 0.450 (0.026) 0.004 (0.030) 0.546 (0.021) -0.039 (0.039) 0.083 (0.085) 0.125 (0.124) 0.218 (0.047) -0.196 (0.024) -0.005 (0.016) 0.086 (0.021) -0.052 (0.018) 0.252 (0.027) 0.020 (0.021) 0.189 (0.029) 0.024 (0.034) -0.213 (0.039) Model 2 0.089 (0.030) 0.013 (0.024) -0.284 (0.017) -0.155 (0.027) 0.043 (0.014) 0.435 (0.027) 0.073 (0.031) 0.404 (0.022) 0.090 (0.041) -0.026 (0.087) 0.093 (0.128) 0.328 (0.048) 0.035 (0.025) -0.014 (0.016) 0.074 (0.022) -0.101 (0.019) 0.038 (0.028) -0.112 (0.022) 0.047 (0.030) -0.131 (0.036) -0.087 (0.040) 0.500 (0.046) 0.357 (0.029) 0.017 (0.012) Teacher 2 Model 3 Model 4 0.175 0.212 (0.045) (0.048) 0.026 0.012 (0.035) (0.036) -0.320 -0.231 (0.024) (0.025) -0.172 -0.137 (0.040) (0.041) 0.125 0.141 (0.031) (0.032) 0.422 0.450 (0.045) (0.046) 0.067 0.016 (0.031) (0.032) 0.367 0.352 (0.032) (0.033) 0.097 0.012 (0.047) (0.057) -0.026 -0.058 (0.087) (0.091) 0.163 0.227 (0.218) (0.220) 0.488 0.495 (0.263) (0.266) 0.037 -0.086 (0.025) (0.027) -0.014 -0.032 (0.016) (0.018) 0.181 0.174 (0.033) (0.034) -0.058 -0.083 (0.027) (0.028) 0.085 0.047 (0.036) (0.037) -0.109 -0.143 (0.029) (0.030) 0.139 0.074 (0.039) (0.040) -0.090 -0.106 (0.042) (0.043) -0.012 -0.042 (0.055) (0.057) 0.501 0.564 (0.046) (0.048) 0.357 0.413 (0.029) (0.030) 0.018 0.009 (0.012) (0.012) Model 5 0.100 (0.051) 0.026 (0.039) -0.197 (0.027) -0.106 (0.045) 0.079 (0.033) 0.270 (0.047) 0.006 (0.033) 0.180 (0.034) -0.009 (0.059) -0.038 (0.094) 0.253 (0.228) 0.499 (0.275) -0.070 (0.028) -0.013 (0.019) 0.134 (0.036) -0.006 (0.029) 0.108 (0.038) -0.020 (0.031) 0.115 (0.042) 0.020 (0.045) 0.067 (0.059) 0.138 (0.052) 0.244 (0.033) -0.006 (0.014) Model 6 -0.014 (0.052) -0.026 (0.040) -0.144 (0.028) -0.061 (0.045) 0.054 (0.034) 0.171 (0.048) 0.007 (0.034) 0.135 (0.035) -0.063 (0.060) -0.059 (0.096) 0.230 (0.231) 0.434 (0.278) -0.082 (0.028) -0.015 (0.019) 0.146 (0.036) 0.020 (0.030) 0.151 (0.038) 0.021 (0.031) 0.174 (0.042) 0.100 (0.045) 0.084 (0.059) 0.148 (0.052) 0.273 (0.033) -0.007 (0.014) Model 1 -0.573 (0.026) -0.280 (0.022) -0.054 (0.016) -0.048 (0.026) 0.032 (0.013) 0.448 (0.026) 0.029 (0.031) 0.640 (0.021) -0.059 (0.039) 0.112 (0.083) 0.114 (0.121) 0.298 (0.046) -0.178 (0.025) -0.115 (0.016) 0.055 (0.021) -0.037 (0.018) 0.216 (0.027) 0.039 (0.021) 0.189 (0.029) -0.067 (0.035) -0.300 (0.039) Model 2 0.198 (0.030) 0.044 (0.025) -0.261 (0.017) -0.151 (0.027) 0.103 (0.014) 0.426 (0.027) 0.108 (0.031) 0.492 (0.021) 0.092 (0.040) -0.006 (0.086) 0.093 (0.126) 0.422 (0.048) 0.102 (0.026) -0.137 (0.016) 0.057 (0.022) -0.110 (0.019) -0.031 (0.028) -0.131 (0.022) 0.003 (0.030) -0.257 (0.037) -0.142 (0.040) 0.549 (0.045) 0.314 (0.028) -0.016 (0.009) Counselor Model 3 Model 4 0.195 0.261 (0.046) (0.048) -0.001 0.027 (0.035) (0.037) -0.295 -0.152 (0.024) (0.026) -0.199 -0.135 (0.041) (0.042) 0.075 0.110 (0.031) (0.032) 0.339 0.400 (0.045) (0.046) 0.097 0.071 (0.031) (0.033) 0.406 0.399 (0.031) (0.032) 0.082 -0.002 (0.047) (0.056) -0.007 -0.050 (0.086) (0.089) -0.166 -0.027 (0.223) (0.226) 0.085 0.043 (0.263) (0.266) 0.104 -0.012 (0.026) (0.027) -0.135 -0.072 (0.016) (0.018) 0.105 0.085 (0.033) (0.034) -0.132 -0.137 (0.027) (0.028) -0.042 -0.070 (0.036) (0.037) -0.191 -0.222 (0.029) (0.030) 0.014 -0.037 (0.039) (0.041) -0.279 -0.258 (0.043) (0.045) -0.149 -0.188 (0.056) (0.058) 0.550 0.589 (0.045) (0.047) 0.311 0.372 (0.028) (0.029) -0.016 -0.021 (0.009) (0.010) Model 5 0.176 (0.051) 0.021 (0.040) -0.094 (0.028) -0.112 (0.046) 0.043 (0.034) 0.170 (0.049) 0.066 (0.034) 0.219 (0.034) -0.067 (0.059) -0.027 (0.094) -0.101 (0.235) -0.110 (0.277) 0.036 (0.028) -0.054 (0.019) 0.028 (0.036) -0.031 (0.030) 0.017 (0.039) -0.070 (0.031) 0.044 (0.043) -0.094 (0.047) -0.073 (0.060) -0.001 (0.052) 0.220 (0.032) -0.073 (0.011) Model 6 0.012 (0.053) -0.048 (0.041) -0.022 (0.028) -0.053 (0.047) 0.004 (0.035) 0.025 (0.050) 0.063 (0.035) 0.155 (0.035) -0.136 (0.060) -0.065 (0.096) -0.130 (0.240) -0.207 (0.283) 0.026 (0.029) -0.059 (0.019) 0.034 (0.037) -0.002 (0.030) 0.077 (0.040) -0.022 (0.032) 0.118 (0.044) 0.011 (0.048) -0.061 (0.062) 0.017 (0.053) 0.260 (0.033) -0.073 (0.011) Table B.6.6R Continued: Ordered logit estimates of Harvard's School Support Measures, expanded sample Model 1 Model 2 142945 0.031 142945 0.079 Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspec Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Legacy X African American Legacy X Hispanic Legacy X Asian American Legacy X Missing Special X African American Special X Hispanic Special X Asian American Special X Missing Observations Pseudo R Sq. Teacher 1 Model 3 Model 4 -0.131 -0.137 (0.041) (0.043) -0.048 -0.045 (0.035) (0.036) -0.123 -0.112 (0.053) (0.054) 0.019 0.025 (0.041) (0.043) -0.194 -0.206 (0.057) (0.059) 0.014 0.000 (0.075) (0.078) -0.040 -0.039 (0.074) (0.076) -0.096 -0.100 (0.048) (0.049) -0.040 -0.036 (0.040) (0.041) 0.016 0.015 (0.029) (0.030) 0.036 0.030 (0.049) (0.051) 0.085 0.036 (0.061) (0.063) 0.194 0.076 (0.057) (0.059) 0.013 0.004 (0.053) (0.055) -0.082 0.004 (0.097) (0.099) 0.028 0.048 (0.075) (0.077) -0.002 0.060 (0.066) (0.068) 0.100 0.084 (0.046) (0.047) 0.076 0.051 (0.071) (0.073) 0.066 0.060 (0.169) (0.175) -0.091 -0.094 (0.148) (0.153) 0.147 0.076 (0.111) (0.117) 0.181 0.131 (0.122) (0.128) 0.547 0.662 (0.242) (0.252) -0.250 -0.197 (0.189) (0.198) -0.122 -0.076 (0.133) (0.142) 0.128 0.123 (0.151) (0.158) 142945 136326 0.08 0.09 Model 5 -0.092 (0.044) -0.045 (0.037) -0.132 (0.056) -0.023 (0.044) -0.175 (0.061) 0.006 (0.080) -0.031 (0.079) -0.067 (0.051) 0.009 (0.043) 0.030 (0.031) 0.012 (0.052) 0.112 (0.065) 0.089 (0.061) 0.012 (0.057) 0.042 (0.103) -0.008 (0.080) 0.025 (0.071) -0.003 (0.049) 0.047 (0.076) 0.106 (0.180) -0.059 (0.159) 0.121 (0.122) 0.165 (0.133) 0.786 (0.261) -0.298 (0.204) -0.191 (0.147) 0.050 (0.165) 136326 0.144 Model 6 -0.088 (0.045) -0.041 (0.037) -0.133 (0.057) -0.030 (0.045) -0.196 (0.062) -0.018 (0.081) -0.015 (0.080) -0.025 (0.052) 0.015 (0.044) 0.024 (0.031) -0.001 (0.053) 0.171 (0.067) 0.102 (0.062) 0.018 (0.058) 0.048 (0.105) -0.035 (0.082) 0.027 (0.072) 0.016 (0.050) 0.052 (0.078) 0.070 (0.184) -0.012 (0.161) 0.104 (0.124) 0.173 (0.135) 0.816 (0.267) -0.265 (0.207) -0.217 (0.150) 0.032 (0.166) 136326 0.165 Model 1 Model 2 120968 0.03 120968 0.076 Teacher 2 Model 3 Model 4 -0.193 -0.181 (0.044) (0.046) -0.085 -0.070 (0.037) (0.038) -0.104 -0.085 (0.056) (0.058) 0.015 0.035 (0.044) (0.045) -0.215 -0.197 (0.061) (0.063) -0.081 -0.094 (0.080) (0.083) -0.155 -0.112 (0.079) (0.081) -0.134 -0.135 (0.052) (0.054) -0.077 -0.083 (0.043) (0.045) 0.030 0.018 (0.031) (0.032) 0.052 0.078 (0.052) (0.054) -0.024 -0.087 (0.067) (0.069) 0.094 -0.006 (0.061) (0.064) 0.007 0.010 (0.057) (0.059) -0.127 -0.097 (0.104) (0.107) -0.022 -0.015 (0.082) (0.085) 0.035 0.103 (0.072) (0.074) 0.149 0.144 (0.050) (0.051) -0.108 -0.127 (0.077) (0.080) 0.097 0.132 (0.179) (0.184) -0.252 -0.266 (0.156) (0.162) 0.030 -0.010 (0.116) (0.123) 0.093 0.021 (0.129) (0.136) -0.125 -0.143 (0.288) (0.304) 0.230 0.251 (0.186) (0.196) 0.215 0.286 (0.133) (0.143) 0.315 0.266 (0.157) (0.165) 120968 115189 0.076 0.085 Model 5 -0.151 (0.047) -0.071 (0.040) -0.097 (0.060) -0.013 (0.047) -0.143 (0.065) -0.112 (0.086) -0.100 (0.084) -0.100 (0.055) -0.054 (0.046) 0.040 (0.033) 0.059 (0.055) -0.045 (0.071) -0.027 (0.066) 0.004 (0.061) -0.065 (0.111) -0.047 (0.088) 0.088 (0.077) 0.061 (0.053) -0.173 (0.083) 0.167 (0.190) -0.273 (0.168) -0.004 (0.128) 0.027 (0.141) -0.255 (0.315) 0.252 (0.201) 0.239 (0.149) 0.218 (0.171) 115189 0.138 Model 6 -0.146 (0.048) -0.070 (0.040) -0.093 (0.061) -0.023 (0.047) -0.164 (0.066) -0.142 (0.087) -0.085 (0.085) -0.073 (0.056) -0.051 (0.047) 0.035 (0.033) 0.047 (0.056) 0.014 (0.073) -0.019 (0.067) 0.008 (0.062) -0.057 (0.112) -0.065 (0.090) 0.082 (0.078) 0.079 (0.054) -0.166 (0.084) 0.136 (0.193) -0.229 (0.170) -0.002 (0.130) 0.051 (0.143) -0.232 (0.317) 0.284 (0.206) 0.210 (0.150) 0.204 (0.172) 115189 0.158 Model 1 Model 2 140271 0.048 140271 0.106 Counselor Model 3 Model 4 -0.076 -0.090 (0.044) (0.046) 0.041 0.043 (0.037) (0.038) 0.023 0.033 (0.056) (0.058) 0.151 0.160 (0.044) (0.046) -0.034 -0.043 (0.060) (0.062) 0.070 0.049 (0.082) (0.085) 0.007 0.036 (0.079) (0.082) -0.036 -0.065 (0.051) (0.053) -0.022 -0.050 (0.044) (0.045) 0.003 -0.011 (0.031) (0.032) 0.093 0.095 (0.052) (0.054) 0.067 -0.042 (0.066) (0.069) 0.269 0.154 (0.062) (0.064) 0.099 0.113 (0.058) (0.059) -0.219 -0.155 (0.107) (0.110) 0.192 0.227 (0.078) (0.081) 0.013 0.034 (0.071) (0.074) 0.214 0.204 (0.048) (0.050) 0.068 0.048 (0.075) (0.078) 0.015 0.004 (0.176) (0.182) 0.146 0.088 (0.150) (0.158) 0.080 0.051 (0.116) (0.122) -0.026 -0.084 (0.129) (0.135) 0.056 0.050 (0.267) (0.281) 0.120 0.117 (0.185) (0.195) 0.081 0.042 (0.132) (0.143) 0.248 0.148 (0.155) (0.163) 140271 133831 0.107 0.119 Model 5 -0.037 (0.048) 0.049 (0.040) 0.039 (0.061) 0.117 (0.048) 0.016 (0.065) 0.044 (0.089) 0.075 (0.085) -0.016 (0.055) -0.003 (0.047) 0.000 (0.034) 0.091 (0.057) 0.010 (0.072) 0.183 (0.067) 0.130 (0.062) -0.153 (0.115) 0.213 (0.085) 0.040 (0.078) 0.156 (0.052) 0.055 (0.082) 0.040 (0.190) 0.156 (0.165) 0.067 (0.129) -0.069 (0.142) -0.010 (0.308) 0.121 (0.202) -0.054 (0.149) 0.088 (0.171) 133831 0.189 Model 6 -0.025 (0.049) 0.056 (0.041) 0.045 (0.062) 0.112 (0.049) -0.010 (0.067) 0.005 (0.091) 0.104 (0.088) 0.037 (0.057) 0.002 (0.049) -0.007 (0.034) 0.080 (0.058) 0.104 (0.074) 0.205 (0.070) 0.146 (0.064) -0.150 (0.119) 0.182 (0.088) 0.033 (0.080) 0.181 (0.054) 0.083 (0.084) -0.003 (0.194) 0.221 (0.169) 0.046 (0.132) -0.043 (0.145) 0.006 (0.314) 0.173 (0.209) -0.091 (0.153) 0.080 (0.173) 133831 0.223 Table B.6.7R: Ordered logit estimates of Harvard's Personal Rating and Alumni Personal Rating, expanded sample Model 1 African American -0.103 (0.024) Hispanic -0.074 (0.022) Asian American -0.343 (0.017) Missing -0.231 (0.028) Female 0.168 (0.014) Disadvantaged 0.754 (0.026) First generation 0.014 (0.030) Early Decision 0.652 (0.021) Legacy 0.350 (0.039) Double Legacy 0.183 (0.084) Faculty or Staff Child 0.320 (0.124) Dean's/Director's List 0.688 (0.047) Waiver -0.185 (0.024) Applied for Financial Aid -0.135 (0.017) Humanities 0.005 (0.022) Biology -0.228 (0.019) Physical Sciences -0.200 (0.029) Engineering -0.239 (0.022) Mathematics -0.266 (0.032) Computer Science -0.544 (0.040) Unspecified -0.402 (0.039) Academic index AI Sq. X (AI>0) AI Sq. X (AI<0) Personal Rating Model 2 Model 3 Model 4 0.482 0.693 0.714 (0.028) (0.042) (0.045) 0.170 0.198 0.220 (0.025) (0.035) (0.037) -0.477 -0.547 -0.472 (0.019) (0.027) (0.029) -0.312 -0.390 -0.366 (0.030) (0.046) (0.047) 0.218 0.213 0.203 (0.014) (0.032) (0.032) 0.745 0.746 0.794 (0.026) (0.043) (0.044) 0.076 0.061 0.035 (0.031) (0.031) (0.032) 0.553 0.512 0.504 (0.022) (0.032) (0.032) 0.442 0.421 0.259 (0.040) (0.047) (0.056) 0.111 0.108 0.064 (0.086) (0.086) (0.089) 0.311 0.049 0.064 (0.127) (0.221) (0.224) 0.750 0.395 0.384 (0.048) (0.262) (0.264) 0.013 0.016 -0.025 (0.025) (0.025) (0.027) -0.144 -0.147 -0.028 (0.017) (0.017) (0.019) 0.000 0.060 0.019 (0.022) (0.034) (0.035) -0.274 -0.274 -0.260 (0.019) (0.029) (0.029) -0.363 -0.377 -0.373 (0.030) (0.039) (0.040) -0.352 -0.412 -0.417 (0.023) (0.031) (0.032) -0.388 -0.427 -0.427 (0.033) (0.044) (0.045) -0.665 -0.737 -0.709 (0.041) (0.049) (0.050) -0.278 -0.195 -0.215 (0.040) (0.056) (0.057) 0.419 0.418 0.396 (0.046) (0.046) (0.047) 0.002 0.015 0.011 (0.029) (0.030) (0.030) 0.015 0.018 0.018 (0.010) (0.010) (0.010) Model 5 0.691 (0.053) 0.286 (0.043) -0.391 (0.033) -0.343 (0.055) 0.162 (0.037) 0.560 (0.051) 0.014 (0.036) 0.268 (0.038) 0.255 (0.065) 0.171 (0.102) 0.064 (0.253) 0.325 (0.303) 0.024 (0.030) 0.019 (0.022) -0.069 (0.040) -0.140 (0.033) -0.299 (0.046) -0.247 (0.036) -0.373 (0.051) -0.532 (0.056) -0.062 (0.065) -0.132 (0.056) -0.198 (0.037) 0.005 (0.012) Model 1 -0.136 (0.020) -0.106 (0.019) -0.012 (0.014) 0.021 (0.023) 0.180 (0.011) 0.172 (0.022) 0.052 (0.024) 0.261 (0.018) 0.109 (0.034) 0.129 (0.075) -0.046 (0.108) 0.307 (0.043) -0.032 (0.020) -0.053 (0.014) -0.001 (0.019) -0.161 (0.015) -0.152 (0.024) -0.192 (0.018) -0.223 (0.026) -0.289 (0.030) -0.401 (0.032) Model 2 0.294 (0.024) 0.077 (0.020) -0.138 (0.015) -0.050 (0.024) 0.237 (0.012) 0.147 (0.022) 0.101 (0.025) 0.180 (0.018) 0.172 (0.034) 0.071 (0.075) -0.073 (0.109) 0.334 (0.043) 0.133 (0.020) -0.049 (0.014) -0.012 (0.019) -0.189 (0.016) -0.293 (0.024) -0.276 (0.019) -0.328 (0.026) -0.397 (0.031) -0.288 (0.032) 0.455 (0.036) 0.177 (0.023) 0.022 (0.007) Alumni Personal Rating Model 3 Model 4 Model 5 0.418 0.462 0.234 (0.035) (0.036) (0.041) 0.068 0.072 0.066 (0.029) (0.030) (0.033) -0.165 -0.102 -0.179 (0.021) (0.022) (0.025) -0.068 -0.052 -0.141 (0.035) (0.036) (0.040) 0.206 0.208 0.242 (0.026) (0.027) (0.031) 0.103 0.109 -0.074 (0.039) (0.039) (0.044) 0.095 0.077 0.034 (0.025) (0.025) (0.028) 0.174 0.178 0.125 (0.027) (0.028) (0.031) 0.175 0.105 -0.055 (0.040) (0.048) (0.055) 0.068 -0.014 -0.088 (0.075) (0.078) (0.089) -0.318 -0.279 -0.171 (0.194) (0.195) (0.231) 0.024 0.021 -0.079 (0.229) (0.230) (0.268) 0.135 0.087 0.063 (0.020) (0.021) (0.024) -0.047 -0.008 0.008 (0.014) (0.015) (0.017) 0.025 0.014 -0.010 (0.028) (0.029) (0.033) -0.217 -0.215 -0.136 (0.023) (0.023) (0.026) -0.337 -0.342 -0.351 (0.031) (0.032) (0.036) -0.331 -0.333 -0.225 (0.024) (0.025) (0.028) -0.362 -0.372 -0.397 (0.034) (0.035) (0.040) -0.474 -0.469 -0.470 (0.036) (0.037) (0.041) -0.245 -0.246 0.020 (0.045) (0.046) (0.051) 0.454 0.441 -0.380 (0.036) (0.037) (0.043) 0.178 0.204 -0.187 (0.023) (0.024) (0.030) 0.022 0.021 -0.017 (0.007) (0.007) (0.008) Model 6 0.204 (0.041) 0.055 (0.033) -0.164 (0.025) -0.128 (0.040) 0.233 (0.031) -0.105 (0.044) 0.032 (0.028) 0.107 (0.031) -0.069 (0.055) -0.101 (0.089) -0.178 (0.230) -0.101 (0.268) 0.062 (0.024) 0.007 (0.017) -0.009 (0.033) -0.130 (0.026) -0.337 (0.036) -0.215 (0.028) -0.380 (0.040) -0.447 (0.041) 0.022 (0.051) -0.377 (0.043) -0.173 (0.030) -0.018 (0.008) Table B.6.7R Continued: Ordered logit estimates of Harvard's Personal Rating and Alumni Personal Rating, expanded sample Model 1 Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspec Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Legacy X African American Legacy X Hispanic Legacy X Asian American Legacy X Missing Special X African American Special X Hispanic Special X Asian American Special X Missing Observations Pseudo R Sq. 148769 0.068 Personal Rating Model 2 Model 3 Model 4 -0.097 -0.081 (0.044) (0.046) 0.003 0.000 (0.038) (0.039) 0.033 0.022 (0.060) (0.062) 0.146 0.155 (0.046) (0.047) 0.094 0.094 (0.065) (0.067) 0.265 0.233 (0.088) (0.090) -0.170 -0.149 (0.079) (0.081) -0.274 -0.282 (0.048) (0.049) -0.090 -0.090 (0.043) (0.044) 0.070 0.072 (0.034) (0.035) 0.142 0.146 (0.057) (0.059) -0.232 -0.293 (0.061) (0.063) 0.091 0.021 (0.058) (0.060) 0.117 0.101 (0.056) (0.058) -0.054 0.014 (0.102) (0.104) 0.150 0.131 (0.072) (0.074) -0.021 -0.003 (0.068) (0.070) 0.099 0.086 (0.051) (0.052) -0.006 -0.040 (0.078) (0.081) 0.186 0.229 (0.158) (0.163) -0.021 -0.064 (0.146) (0.152) 0.155 0.150 (0.117) (0.123) -0.038 -0.036 (0.131) (0.136) -0.098 -0.183 (0.255) (0.267) -0.176 -0.170 (0.185) (0.194) 0.052 0.193 (0.135) (0.144) 0.173 0.177 (0.156) (0.163) 148769 148769 141852 0.095 0.096 0.104 Model 5 -0.005 (0.053) -0.013 (0.045) 0.005 (0.071) 0.062 (0.054) 0.143 (0.076) 0.236 (0.102) -0.151 (0.092) -0.257 (0.056) -0.029 (0.050) 0.087 (0.039) 0.132 (0.067) -0.327 (0.073) -0.060 (0.070) 0.053 (0.067) 0.111 (0.120) 0.087 (0.087) -0.046 (0.081) -0.035 (0.060) -0.031 (0.094) 0.222 (0.187) -0.082 (0.176) 0.146 (0.143) -0.017 (0.157) -0.272 (0.307) -0.329 (0.223) 0.170 (0.166) 0.059 (0.190) 141852 0.294 Model 1 117109 0.013 Alumni Personal Rating Model 3 Model 4 Model 5 -0.052 -0.050 -0.017 (0.038) (0.038) (0.043) 0.055 0.058 0.004 (0.031) (0.032) (0.036) 0.103 0.103 0.099 (0.049) (0.050) (0.056) 0.129 0.130 -0.010 (0.037) (0.038) (0.043) 0.076 0.079 0.185 (0.053) (0.053) (0.060) 0.277 0.265 0.283 (0.069) (0.071) (0.080) -0.090 -0.098 -0.202 (0.063) (0.064) (0.072) -0.173 -0.164 -0.065 (0.039) (0.040) (0.045) -0.035 -0.038 -0.029 (0.035) (0.036) (0.041) 0.020 0.012 0.048 (0.027) (0.027) (0.031) 0.048 0.062 0.048 (0.045) (0.046) (0.052) 0.001 0.008 0.108 (0.053) (0.054) (0.060) 0.141 0.127 0.174 (0.052) (0.053) (0.059) 0.037 0.032 0.086 (0.049) (0.050) (0.056) 0.000 0.017 0.071 (0.088) (0.090) (0.100) -0.099 -0.089 -0.050 (0.061) (0.062) (0.071) -0.077 -0.064 -0.126 (0.058) (0.059) (0.067) 0.106 0.087 -0.004 (0.042) (0.043) (0.049) -0.102 -0.103 -0.124 (0.066) (0.067) (0.077) -0.291 -0.339 -0.201 (0.136) (0.138) (0.158) 0.113 0.086 -0.025 (0.127) (0.131) (0.151) 0.061 0.058 0.246 (0.101) (0.106) (0.120) -0.003 -0.021 0.235 (0.112) (0.116) (0.131) 0.271 0.247 -0.017 (0.226) (0.231) (0.259) 0.470 0.378 0.300 (0.170) (0.176) (0.201) 0.123 0.209 -0.098 (0.119) (0.127) (0.145) 0.279 0.287 0.267 (0.141) (0.146) (0.167) 117109 117109 113323 113323 0.028 0.028 0.032 0.341 Model 2 Model 6 -0.016 (0.043) 0.005 (0.036) 0.099 (0.056) -0.011 (0.043) 0.181 (0.060) 0.275 (0.080) -0.194 (0.072) -0.055 (0.045) -0.029 (0.041) 0.045 (0.031) 0.045 (0.052) 0.130 (0.060) 0.182 (0.059) 0.089 (0.056) 0.068 (0.100) -0.049 (0.071) -0.123 (0.067) 0.004 (0.049) -0.120 (0.077) -0.213 (0.158) -0.019 (0.151) 0.241 (0.120) 0.238 (0.131) -0.005 (0.258) 0.318 (0.202) -0.104 (0.145) 0.265 (0.168) 113323 0.342 Table B.6.8R: Ordered logit estimates of Harvard's Overall Rating and Alumni Overall Rating, expanded sample Model 1 African American -0.822 (0.018) Hispanic -0.235 (0.016) Asian American 0.156 (0.012) Missing 0.090 (0.020) Female -0.014 (0.010) Disadvantaged 0.600 (0.019) First generation -0.165 (0.021) Early Decision 0.731 (0.017) Legacy 0.620 (0.033) Double Legacy 0.440 (0.073) Faculty or Staff Child 0.920 (0.106) Dean's/Director's List 0.606 (0.040) Waiver -0.530 (0.017) Applied for Financial Aid -0.105 (0.012) Humanities 0.071 (0.016) Biology -0.075 (0.013) Physical Sciences 0.276 (0.021) Engineering 0.108 (0.016) Mathematics 0.138 (0.023) Computer Science 0.020 (0.026) Unspecified -0.512 (0.027) Academic index AI Sq. X (AI>0) AI Sq. X (AI<0) Model 2 0.944 (0.022) 0.541 (0.018) -0.240 (0.013) -0.116 (0.021) 0.196 (0.010) 0.649 (0.020) 0.021 (0.022) 0.586 (0.017) 0.964 (0.034) 0.257 (0.074) 0.827 (0.107) 0.781 (0.041) 0.035 (0.018) -0.108 (0.013) 0.063 (0.017) -0.206 (0.014) -0.171 (0.022) -0.219 (0.017) -0.225 (0.024) -0.347 (0.027) -0.220 (0.028) 1.590 (0.032) -0.156 (0.022) 0.076 (0.007) Overall Rating Model 3 Model 4 1.132 1.211 (0.032) (0.034) 0.628 0.689 (0.025) (0.027) -0.287 -0.175 (0.019) (0.020) -0.134 -0.098 (0.031) (0.032) 0.182 0.177 (0.024) (0.025) 0.848 0.899 (0.035) (0.036) 0.014 0.001 (0.022) (0.023) 0.513 0.514 (0.025) (0.026) 0.998 0.834 (0.039) (0.047) 0.265 0.151 (0.074) (0.077) 0.935 0.989 (0.195) (0.195) 0.945 0.917 (0.228) (0.228) 0.033 -0.008 (0.018) (0.019) -0.115 -0.024 (0.013) (0.014) 0.088 0.056 (0.026) (0.027) -0.208 -0.207 (0.021) (0.021) -0.217 -0.226 (0.028) (0.029) -0.277 -0.282 (0.022) (0.022) -0.206 -0.227 (0.031) (0.032) -0.379 -0.366 (0.032) (0.033) -0.213 -0.223 (0.039) (0.040) 1.600 1.592 (0.032) (0.033) -0.146 -0.116 (0.022) (0.023) 0.082 0.087 (0.007) (0.007) Model 5 1.500 (0.037) 0.928 (0.030) -0.135 (0.022) -0.083 (0.035) 0.114 (0.026) 0.743 (0.037) 0.010 (0.024) 0.313 (0.027) 0.977 (0.049) 0.256 (0.080) 1.203 (0.203) 0.980 (0.239) 0.083 (0.020) 0.004 (0.015) 0.000 (0.028) -0.075 (0.022) -0.075 (0.030) -0.064 (0.023) -0.108 (0.033) -0.105 (0.034) -0.076 (0.042) 0.475 (0.036) -0.131 (0.025) 0.081 (0.007) Model 6 1.444 (0.037) 0.911 (0.030) -0.081 (0.022) -0.031 (0.036) 0.088 (0.026) 0.671 (0.038) 0.014 (0.024) 0.273 (0.027) 0.953 (0.050) 0.251 (0.081) 1.254 (0.205) 0.946 (0.241) 0.083 (0.020) 0.004 (0.015) 0.011 (0.028) -0.056 (0.022) -0.030 (0.031) -0.033 (0.024) -0.053 (0.034) -0.035 (0.035) -0.065 (0.042) 0.488 (0.036) -0.076 (0.025) 0.081 (0.007) Model 1 -0.669 (0.020) -0.355 (0.018) 0.226 (0.014) 0.177 (0.022) -0.024 (0.011) 0.189 (0.021) -0.017 (0.024) 0.310 (0.018) 0.105 (0.033) 0.236 (0.073) 0.002 (0.106) 0.266 (0.041) -0.239 (0.020) -0.075 (0.014) 0.024 (0.018) -0.061 (0.015) 0.183 (0.024) 0.015 (0.018) 0.090 (0.025) 0.065 (0.030) -0.513 (0.031) Model 2 0.247 (0.023) 0.031 (0.020) -0.043 (0.015) 0.038 (0.023) 0.133 (0.011) 0.142 (0.022) 0.096 (0.025) 0.158 (0.018) 0.231 (0.034) 0.131 (0.074) -0.053 (0.107) 0.322 (0.042) 0.119 (0.020) -0.051 (0.014) 0.014 (0.018) -0.123 (0.015) -0.126 (0.024) -0.183 (0.018) -0.155 (0.026) -0.170 (0.030) -0.313 (0.032) 0.943 (0.036) 0.362 (0.024) 0.020 (0.007) Alumni Overall Rating Model 3 Model 4 0.371 0.414 (0.035) (0.036) 0.030 0.043 (0.029) (0.030) -0.047 0.022 (0.020) (0.022) 0.033 0.062 (0.035) (0.035) 0.123 0.116 (0.026) (0.027) 0.156 0.160 (0.038) (0.039) 0.093 0.083 (0.025) (0.025) 0.124 0.127 (0.027) (0.027) 0.269 0.185 (0.040) (0.048) 0.129 0.050 (0.074) (0.076) -0.234 -0.175 (0.190) (0.192) 0.122 0.117 (0.224) (0.226) 0.121 0.070 (0.020) (0.021) -0.051 -0.010 (0.014) (0.015) 0.060 0.041 (0.028) (0.029) -0.155 -0.155 (0.023) (0.023) -0.139 -0.151 (0.031) (0.032) -0.253 -0.252 (0.024) (0.025) -0.132 -0.143 (0.034) (0.035) -0.207 -0.205 (0.036) (0.037) -0.330 -0.335 (0.045) (0.045) 0.943 0.941 (0.036) (0.037) 0.360 0.388 (0.024) (0.024) 0.021 0.020 (0.007) (0.007) Model 5 0.129 (0.040) -0.007 (0.032) 0.155 (0.024) 0.168 (0.039) -0.092 (0.030) 0.065 (0.042) 0.037 (0.028) -0.047 (0.030) 0.166 (0.053) 0.094 (0.086) -0.088 (0.219) 0.076 (0.257) 0.029 (0.023) -0.001 (0.017) 0.015 (0.032) 0.035 (0.026) 0.176 (0.035) 0.044 (0.027) 0.207 (0.039) 0.253 (0.041) -0.211 (0.050) 0.748 (0.042) 0.344 (0.029) -0.016 (0.009) Model 6 0.129 (0.040) -0.008 (0.032) 0.155 (0.024) 0.167 (0.039) -0.091 (0.030) 0.065 (0.042) 0.036 (0.028) -0.047 (0.030) 0.166 (0.053) 0.094 (0.086) -0.089 (0.219) 0.075 (0.257) 0.029 (0.023) -0.001 (0.017) 0.015 (0.032) 0.035 (0.026) 0.176 (0.035) 0.044 (0.027) 0.207 (0.039) 0.253 (0.041) -0.212 (0.050) 0.748 (0.042) 0.344 (0.029) -0.016 (0.009) Table B.6.8R Continued: Ordered logit estimates of Harvard's Overall Rating and Alumni Overall Rating, expanded sample Model 1 Model 2 Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspec Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Legacy X African American Legacy X Hispanic Legacy X Asian American Legacy X Missing Special X African American Special X Hispanic Special X Asian American Special X Missing Observations Pseudo R Sq. 148742 0.065 148742 0.199 Overall Rating Model 3 Model 4 -0.038 -0.027 (0.034) (0.036) 0.007 0.016 (0.028) (0.029) 0.108 0.119 (0.045) (0.046) 0.151 0.162 (0.034) (0.034) -0.054 -0.038 (0.048) (0.049) 0.116 0.074 (0.062) (0.063) -0.012 -0.005 (0.055) (0.056) -0.129 -0.144 (0.036) (0.037) -0.089 -0.097 (0.031) (0.032) 0.025 0.022 (0.024) (0.025) 0.048 0.031 (0.041) (0.042) -0.636 -0.667 (0.049) (0.050) -0.300 -0.352 (0.047) (0.048) 0.085 0.059 (0.045) (0.046) -0.234 -0.192 (0.081) (0.083) 0.245 0.237 (0.059) (0.060) 0.064 0.054 (0.054) (0.055) 0.134 0.116 (0.040) (0.041) 0.028 0.024 (0.062) (0.064) -0.297 -0.371 (0.139) (0.142) -0.321 -0.355 (0.122) (0.126) 0.109 0.086 (0.097) (0.102) -0.028 -0.025 (0.106) (0.110) -0.342 -0.354 (0.215) (0.223) -0.280 -0.300 (0.160) (0.167) 0.074 0.097 (0.114) (0.122) 0.053 0.025 (0.134) (0.140) 148742 141827 0.2 0.206 Model 5 0.034 (0.037) 0.002 (0.030) 0.108 (0.048) 0.065 (0.036) -0.030 (0.051) 0.053 (0.066) 0.046 (0.059) -0.145 (0.039) -0.014 (0.034) 0.040 (0.026) 0.015 (0.044) -0.684 (0.053) -0.351 (0.051) 0.095 (0.048) -0.141 (0.087) 0.200 (0.063) -0.001 (0.058) -0.015 (0.042) 0.019 (0.067) -0.527 (0.150) -0.402 (0.132) 0.204 (0.107) -0.010 (0.115) -0.367 (0.231) -0.376 (0.173) -0.024 (0.129) -0.020 (0.146) 141827 0.335 Model 6 0.038 (0.037) 0.007 (0.030) 0.112 (0.048) 0.069 (0.036) -0.045 (0.052) 0.023 (0.067) 0.063 (0.060) -0.108 (0.039) -0.007 (0.034) 0.036 (0.026) 0.002 (0.044) -0.646 (0.053) -0.342 (0.051) 0.100 (0.048) -0.157 (0.087) 0.192 (0.064) 0.005 (0.058) 0.005 (0.043) 0.036 (0.067) -0.570 (0.151) -0.399 (0.133) 0.181 (0.108) 0.016 (0.116) -0.362 (0.234) -0.330 (0.173) -0.079 (0.131) -0.013 (0.147) 141827 0.349 Model 1 117109 0.035 Alumni Overall Rating Model 3 Model 4 -0.070 -0.064 (0.037) (0.038) 0.059 0.065 (0.031) (0.031) 0.026 0.034 (0.049) (0.050) 0.175 0.175 (0.037) (0.038) -0.067 -0.061 (0.052) (0.053) 0.118 0.095 (0.069) (0.070) 0.029 0.023 (0.062) (0.064) -0.172 -0.163 (0.039) (0.039) -0.038 -0.047 (0.035) (0.036) -0.022 -0.027 (0.027) (0.027) 0.022 0.020 (0.044) (0.046) -0.071 -0.049 (0.052) (0.053) 0.048 0.034 (0.051) (0.053) -0.032 -0.035 (0.048) (0.049) -0.032 -0.001 (0.086) (0.088) -0.065 -0.063 (0.061) (0.062) -0.015 0.007 (0.058) (0.059) 0.149 0.125 (0.042) (0.043) -0.022 -0.026 (0.065) (0.067) -0.193 -0.227 (0.134) (0.136) 0.175 0.127 (0.123) (0.128) -0.182 -0.161 (0.099) (0.104) -0.160 -0.192 (0.109) (0.112) 0.280 0.286 (0.221) (0.227) 0.186 0.223 (0.164) (0.170) 0.265 0.341 (0.119) (0.126) 0.191 0.181 (0.137) (0.143) 117109 117109 113323 0.096 0.096 0.101 Model 2 Model 5 -0.013 (0.042) 0.039 (0.035) -0.040 (0.055) 0.109 (0.041) -0.159 (0.059) -0.145 (0.077) 0.142 (0.070) -0.086 (0.044) -0.004 (0.040) -0.055 (0.030) -0.038 (0.050) -0.071 (0.059) -0.074 (0.058) -0.054 (0.054) -0.053 (0.097) -0.014 (0.069) 0.089 (0.065) 0.055 (0.048) 0.042 (0.075) -0.050 (0.153) 0.135 (0.144) -0.277 (0.115) -0.315 (0.126) 0.270 (0.251) -0.059 (0.190) 0.309 (0.141) -0.070 (0.160) 113323 0.375 Model 6 -0.013 (0.042) 0.039 (0.035) -0.040 (0.055) 0.108 (0.041) -0.160 (0.059) -0.143 (0.077) 0.142 (0.070) -0.086 (0.044) -0.004 (0.040) -0.055 (0.030) -0.038 (0.050) -0.071 (0.059) -0.074 (0.058) -0.054 (0.054) -0.052 (0.097) -0.015 (0.069) 0.089 (0.065) 0.054 (0.048) 0.042 (0.075) -0.051 (0.153) 0.135 (0.144) -0.277 (0.115) -0.316 (0.126) 0.272 (0.251) -0.060 (0.190) 0.310 (0.141) -0.069 (0.160) 113323 0.375 Table B.6.9R: Generalized Ordered Logit Model of Harvard's Overall Rating African American aditional advantage at 3/3+ cutoff addiitional advantage at 3+/2 cutoff Hispanic/Other aditional advantage at 3/3+ cutoff addiitional advantage at 3+/2 cutoff Asian American aditional disadvantage at 3/3+ cutoff addiitional disadvantage at 3+/2 cutoff Disadvantaged DisadvantagedXAfrican American DisadvantagedXHispanic DisadvantagedXAsian American First generation Waiver Applied for financial aid Humanities Biological Sciences Physical Science Engineering Mathematics Computer Science Female FemaleXHumanities FemaleXBiological Sciences FemaleXPhysical Science FemaleXEngineering FemaleXMathematics FemaleXComputer Science FemaleXUnspecified FemaleXAfrican American FemaleXHispanic FemaleXAsian American Early action Early actionXAfrican American Early ActionXHispanic Early ActionXAsian American Legacy Double legacy LegacyXAfrican American LegacyXHispanic LegacyXAsian American Faculty or Staff child Dean's/Director's List Special recruiting Special recruitingXAfrican American Special recruitingXHispanic American Special recruitingXAsian American Observations Pseudo R-sq Baseline Model 5 Model 6 Coeff. Std. Error Coeff. Std. Error 1.370 (0.049) 1.320 (0.050) 0.521 (0.042) 0.518 (0.043) 1.043 (0.057) 1.109 (0.060) 0.911 (0.038) 0.903 (0.038) 0.178 (0.037) 0.174 (0.038) 0.398 (0.056) 0.432 (0.058) -0.096 (0.026) -0.054 (0.027) -0.101 (0.026) -0.064 (0.027) -0.055 (0.041) 0.005 (0.042) 0.826 (0.043) 0.745 (0.043) -0.672 (0.065) -0.638 (0.065) -0.377 (0.060) -0.371 (0.060) 0.058 (0.055) 0.062 (0.055) 0.070 (0.030) 0.075 (0.030) 0.172 (0.025) 0.174 (0.025) 0.003 (0.016) 0.001 (0.017) 0.028 (0.033) 0.040 (0.033) -0.066 (0.026) -0.046 (0.026) -0.070 (0.034) -0.023 (0.035) -0.061 (0.027) -0.027 (0.027) -0.099 (0.038) -0.041 (0.038) -0.104 (0.040) -0.027 (0.040) 0.129 (0.030) 0.104 (0.030) 0.000 (0.043) 0.006 (0.044) -0.036 (0.035) -0.032 (0.035) 0.097 (0.054) 0.102 (0.055) 0.070 (0.041) 0.071 (0.041) -0.021 (0.059) -0.036 (0.059) 0.056 (0.075) 0.027 (0.076) 0.081 (0.074) 0.095 (0.074) -0.080 (0.049) -0.052 (0.049) 0.027 (0.041) 0.023 (0.041) 0.043 (0.029) 0.035 (0.029) 0.353 (0.032) 0.310 (0.032) 0.080 (0.077) 0.068 (0.078) -0.012 (0.068) 0.002 (0.069) 0.048 (0.049) 0.067 (0.049) 136183 0.360 136183 0.378 Expanded Model 5 Model 6 Coeff. Std. Error Coeff. Std. Error 1.368 (0.048) 1.313 (0.049) 0.544 (0.041) 0.545 (0.043) 0.990 (0.056) 1.054 (0.058) 0.920 (0.037) 0.910 (0.038) 0.186 (0.036) 0.186 (0.038) 0.324 (0.054) 0.356 (0.056) -0.097 (0.026) -0.057 (0.026) -0.076 (0.026) -0.036 (0.026) -0.088 (0.039) -0.030 (0.040) 0.826 (0.043) 0.745 (0.043) -0.676 (0.065) -0.641 (0.065) -0.373 (0.059) -0.365 (0.060) 0.054 (0.054) 0.059 (0.055) 0.076 (0.030) 0.080 (0.030) 0.174 (0.024) 0.178 (0.025) -0.005 (0.016) -0.006 (0.016) 0.023 (0.031) 0.036 (0.032) -0.069 (0.025) -0.047 (0.025) -0.073 (0.034) -0.024 (0.034) -0.059 (0.026) -0.023 (0.027) -0.096 (0.037) -0.036 (0.037) -0.099 (0.039) -0.019 (0.039) 0.124 (0.029) 0.099 (0.029) -0.001 (0.042) 0.005 (0.042) -0.035 (0.034) -0.029 (0.035) 0.107 (0.053) 0.111 (0.053) 0.070 (0.040) 0.072 (0.041) -0.021 (0.057) -0.038 (0.058) 0.044 (0.074) 0.016 (0.075) 0.060 (0.071) 0.068 (0.072) -0.060 (0.048) -0.026 (0.049) 0.029 (0.040) 0.029 (0.040) 0.043 (0.029) 0.037 (0.029) 0.401 (0.030) 0.362 (0.030) 0.045 (0.074) 0.026 (0.075) -0.035 (0.065) -0.025 (0.066) 0.018 (0.046) 0.034 (0.047) 1.010 (0.052) 0.998 (0.053) 0.247 (0.086) 0.259 (0.087) -0.555 (0.166) -0.609 (0.169) -0.518 (0.141) -0.531 (0.143) 0.320 (0.117) 0.294 (0.118) 1.608 (0.236) 1.722 (0.241) 1.227 (0.274) 1.241 (0.280) -0.642 (0.282) -0.751 (0.287) -0.477 (0.281) -0.435 (0.286) -0.418 (0.187) -0.387 (0.189) 0.084 (0.139) 0.010 (0.142) 141827 141827 0.366 0.385 Table B.6.10R: Probability of receiving each overall rating for own race/ethnicity and counterfactual race/ethnicity if African if Asian Own Race if White American if Hispanic American Panel 1: Baseline sample, including personal rating White <3 0.425 0.270 0.312 0.429 3 0.392 0.357 0.401 0.395 3+ 0.136 0.203 0.186 0.129 >3+ 0.047 0.169 0.100 0.047 African American <3 0.649 0.745 0.676 0.746 3 0.210 0.189 0.218 0.190 3+ 0.087 0.052 0.075 0.050 >3+ 0.054 0.014 0.031 0.014 Hispanic <3 0.565 0.656 0.529 0.658 3 0.289 0.253 0.276 0.255 3+ 0.103 0.072 0.119 0.068 >3+ 0.043 0.020 0.076 0.020 Asian American <3 0.375 0.371 0.229 0.265 3 0.425 0.421 0.356 0.412 3+ 0.148 0.155 0.229 0.213 >3+ 0.052 0.052 0.187 0.111 Panel 2: Expanded sample, preferred model White <3 0.409 3 0.392 3+ 0.140 >3+ 0.059 African American <3 0.645 3 0.211 3+ 0.088 >3+ 0.056 Hispanic <3 0.558 3 0.291 3+ 0.105 >3+ 0.046 Asian American <3 0.372 3 0.423 3+ 0.149 >3+ 0.056 0.261 0.351 0.205 0.183 0.741 0.191 0.052 0.016 0.648 0.256 0.073 0.023 0.368 0.422 0.153 0.057 Panel 3: Expanded sample, including personal rating White <3 0.408 3 0.392 3+ 0.141 >3+ 0.059 African American <3 0.645 0.746 3 0.211 0.190 3+ 0.088 0.049 >3+ 0.056 0.015 Hispanic <3 0.558 0.651 3 0.291 0.257 3+ 0.106 0.070 >3+ 0.046 0.022 Asian American <3 0.372 0.363 3 0.423 0.418 3+ 0.150 0.158 >3+ 0.056 0.061 *calculated using gologitComponentsExpIndices.do 0.523 0.276 0.121 0.080 0.226 0.354 0.228 0.191 0.255 0.340 0.207 0.197 0.520 0.274 0.123 0.083 0.256 0.400 0.219 0.125 0.300 0.398 0.191 0.111 0.671 0.220 0.076 0.032 0.412 0.392 0.137 0.059 0.742 0.191 0.051 0.016 0.651 0.256 0.071 0.022 0.262 0.412 0.213 0.113 0.298 0.392 0.194 0.116 0.674 0.220 0.075 0.031 0.261 0.415 0.215 0.108 0.417 0.394 0.134 0.055 0.750 0.190 0.047 0.014 0.657 0.257 0.066 0.020 Table B.6.11R: The Role of Observed and Unobserved Factors in Racial/Ethnic Differences in Component Scores, Baseline sample Academic Extracurricular Teacher 1 Preferred Model Teacher 2 Counselor Alumni Personal Alumni Overall Personal -1.129 -0.712 0.120 -1.237 -0.791 0.259 -0.663 -0.427 0.109 -0.759 -0.451 0.142 -0.722 -0.415 0.116 -0.849 -0.514 0.049 -0.253 -0.191 0.027 -0.637 -0.421 0.073 -0.374 -0.268 0.020 1.503 0.922 -0.136 -0.006 -0.112 0.136 -0.217 -0.146 0.171 0.012 -0.023 -0.159 0.104 0.024 -0.203 0.164 0.017 -0.095 0.236 0.062 -0.181 0.126 0.001 0.160 0.682 0.279 -0.398 Overall Academic Extracurricular Counselor Alumni Personal Alumni Overall -1.099 -0.696 0.102 -1.237 -0.791 0.260 -0.605 -0.393 0.086 -0.677 -0.402 0.093 -0.635 -0.361 0.069 -0.755 -0.460 0.006 -0.245 -0.187 0.020 -0.627 -0.423 0.074 1.451 0.906 -0.081 -0.001 -0.109 0.132 -0.291 -0.177 0.198 -0.105 -0.074 -0.104 -0.009 -0.027 -0.150 0.002 -0.050 -0.022 0.207 0.051 -0.165 0.126 0.000 0.159 Overall Linear Index (z-score) Differences (relative to whites) African American Hispanic Asian American Coefficients African American Hispanic Asian American Include Personal Rating Teacher 1 Teacher 2 Linear Index (z-score) Differences (relative to whites) African American Hispanic Asian American Coefficients African American Hispanic Asian American *Constructed using results from ologitComponentsIndices.do Table B.6.12R: The Role of Observed and Unobserved Factors in Racial/Ethnic Differences in Component Scores, Expanded sample Academic Extracurricular Teacher 1 Preferred Model Teacher 2 Counselor Alumni Personal Alumni Overall Personal -1.145 -0.724 0.098 -1.249 -0.795 0.247 -0.678 -0.443 0.082 -0.771 -0.460 0.121 -0.736 -0.426 0.098 -0.870 -0.527 0.026 -0.267 -0.200 0.014 -0.651 -0.427 0.058 -0.399 -0.290 -0.010 1.500 0.928 -0.135 -0.006 -0.109 0.131 -0.208 -0.136 0.175 0.015 -0.018 -0.152 0.100 0.026 -0.197 0.176 0.021 -0.094 0.234 0.066 -0.179 0.129 -0.007 0.155 0.691 0.286 -0.391 Overall Academic Extracurricular Counselor Alumni Personal Alumni Overall -1.115 -0.709 0.078 -1.249 -0.795 0.249 -0.622 -0.412 0.058 -0.689 -0.414 0.072 -0.650 -0.375 0.049 -0.776 -0.475 -0.018 -0.259 -0.196 0.006 -0.641 -0.429 0.059 1.444 0.911 -0.081 0.000 -0.105 0.128 -0.283 -0.167 0.200 -0.103 -0.069 -0.098 -0.014 -0.026 -0.144 0.012 -0.048 -0.022 0.204 0.055 -0.164 0.129 -0.008 0.155 Overall Linear Index Differences (relative to whites) African American Hispanic Asian American Coefficients African American Hispanic Asian American Include Personal Rating Teacher 1 Teacher 2 Linear Index Differences (relative to whites) African American Hispanic Asian American Coefficients African American Hispanic Asian American *Constructed using results from ologitComponentsIndices.do Table B.6.13R: The Role of Non-Acacadmic Characteristics in Racial/Ethnic Differences in Personal Rating Component Scores Non-Academic Index (z-score) Differences (relative to whites) African American Hispanic Asian American Coefficients African American Hispanic Asian American All non-academic characteristics included -0.390 -0.295 0.073 Include only non-academic ratings -0.502 -0.388 -0.017 0.682 0.279 -0.398 0.682 0.279 -0.398 Table B.7.1R: Logit estimates of Harvard's admission decision, baseline sample African American Hispanic Asian American Missing Year=2015 Year=2016 Year=2017 Year=2018 Year=2019 Female Disadvantaged First Generation Early Decision Waiver Applied for Financial Aid Humanities Biology Physical Sciences Engineering Mathematics Computer Science Unspecified Model 1 0.531 (0.040) 0.425 (0.039) 0.057 (0.032) 0.012 (0.054) -0.227 (0.169) -0.526 (0.181) -0.490 (0.177) -0.676 (0.180) -0.760 (0.176) -0.044 (0.025) 1.183 (0.042) -0.004 (0.052) 1.616 (0.032) -0.153 (0.041) 0.054 (0.032) 0.133 (0.039) -0.269 (0.035) 0.305 (0.047) -0.043 (0.040) 0.161 (0.053) -0.023 (0.068) -0.721 (0.089) Model 2 2.417 (0.050) 1.273 (0.044) -0.434 (0.035) -0.283 (0.057) -0.098 (0.179) -0.432 (0.191) -0.498 (0.189) -0.894 (0.191) -0.743 (0.186) 0.254 (0.027) 1.257 (0.048) 0.174 (0.059) 1.456 (0.035) 0.484 (0.047) 0.073 (0.033) 0.145 (0.043) -0.403 (0.038) -0.189 (0.051) -0.348 (0.044) -0.191 (0.057) -0.390 (0.073) -0.418 (0.097) Admit Model 3 Model 4 2.671 2.851 (0.074) (0.078) 1.286 1.339 (0.063) (0.067) -0.565 -0.378 (0.052) (0.055) -0.348 -0.330 (0.093) (0.099) -0.094 -0.159 (0.179) (0.223) -0.429 -0.405 (0.191) (0.234) -0.501 -0.393 (0.189) (0.233) -0.908 -0.818 (0.191) (0.237) -0.745 -0.733 (0.187) (0.231) 0.228 0.271 (0.064) (0.088) 1.497 1.606 (0.071) (0.108) 0.161 -0.018 (0.059) (0.127) 1.371 1.348 (0.055) (0.084) 0.499 0.387 (0.046) (0.049) 0.057 0.114 (0.034) (0.081) 0.208 0.251 (0.064) (0.109) -0.347 -0.262 (0.055) (0.102) -0.187 -0.081 (0.066) (0.133) -0.397 -0.451 (0.058) (0.119) -0.135 -0.236 (0.073) (0.150) -0.469 -0.262 (0.090) (0.214) -0.474 -2.103 (0.136) (1.053) Model 5 3.772 (0.105) 1.959 (0.085) -0.466 (0.070) -0.379 (0.122) -0.875 (0.277) -0.701 (0.287) -0.528 (0.290) -1.102 (0.298) -1.142 (0.289) 0.163 (0.110) 1.660 (0.138) -0.014 (0.167) 1.410 (0.104) 0.697 (0.063) 0.343 (0.101) 0.206 (0.137) -0.031 (0.128) 0.027 (0.173) -0.082 (0.149) -0.276 (0.192) 0.191 (0.274) -1.297 (1.146) Model 6 3.876 (0.112) 2.027 (0.091) -0.330 (0.074) -0.211 (0.128) -0.654 (0.296) -0.727 (0.307) -0.729 (0.307) -1.392 (0.315) -1.350 (0.307) 0.141 (0.116) 1.535 (0.147) 0.058 (0.178) 1.440 (0.110) 0.717 (0.067) 0.405 (0.107) 0.256 (0.145) 0.037 (0.136) 0.159 (0.183) 0.041 (0.159) -0.106 (0.203) 0.462 (0.294) -1.227 (1.149) Table B.7.1R Continued: Logit estimates of Harvard's admission decision, baseline sample Model 1 Academic index AI Sq. X (AI>0) AI Sq. X (AI<0) Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspecified Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Model 2 2.011 (0.137) 0.268 (0.078) -0.950 (0.167) Admit Model 3 Model 4 1.937 1.930 (0.136) (0.142) 0.339 0.390 (0.079) (0.082) -0.926 -0.933 (0.165) (0.171) -0.116 -0.095 (0.085) (0.089) -0.103 -0.107 (0.075) (0.077) -0.002 -0.018 (0.104) (0.108) 0.136 0.131 (0.087) (0.090) -0.130 -0.153 (0.117) (0.121) 0.285 0.216 (0.155) (0.159) 0.117 0.060 (0.191) (0.203) -0.035 -0.067 (0.086) (0.089) 0.063 0.068 (0.079) (0.082) 0.107 0.095 (0.065) (0.067) 0.106 0.130 (0.112) (0.118) -0.984 -1.094 (0.107) (0.111) -0.270 -0.350 (0.098) (0.104) 0.015 0.006 (0.092) (0.095) -0.204 -0.061 (0.167) (0.173) -0.017 -0.009 (0.113) (0.118) -0.070 -0.026 (0.103) (0.108) 0.256 0.225 (0.077) (0.079) 0.129 0.165 (0.122) (0.127) Model 5 0.673 (0.189) 0.108 (0.106) -1.236 (0.219) 0.070 (0.110) -0.132 (0.096) -0.101 (0.136) 0.062 (0.111) 0.028 (0.154) 0.284 (0.202) 0.215 (0.251) -0.099 (0.114) 0.117 (0.104) 0.229 (0.082) 0.074 (0.146) -1.577 (0.143) -0.582 (0.133) 0.144 (0.119) -0.035 (0.222) 0.039 (0.152) 0.031 (0.139) -0.048 (0.100) 0.192 (0.162) Model 6 0.646 (0.200) 0.195 (0.113) -1.355 (0.230) 0.042 (0.116) -0.211 (0.102) -0.125 (0.144) 0.003 (0.118) -0.052 (0.163) 0.125 (0.213) 0.182 (0.266) -0.088 (0.121) 0.098 (0.110) 0.200 (0.087) 0.005 (0.153) -1.540 (0.151) -0.583 (0.140) 0.147 (0.126) -0.008 (0.230) 0.037 (0.162) 0.042 (0.150) -0.017 (0.106) 0.172 (0.169) Table B.7.1R Continued: Logit estimates of Harvard's admission decision, baseline sample Model 1 Model 2 Disadvantaged X Year=2015 Disadvantaged X Year=2016 Disadvantaged X Year=2017 Disadvantaged X Year=2018 Disadvantaged X Year=2019 Early X Year=2016 Early X Year=2017 Early X Year=2018 Admit Model 3 Model 4 0.044 (0.123) -0.002 (0.129) -0.011 (0.134) -0.262 (0.138) -0.493 (0.131) -0.097 (0.101) 0.002 (0.101) 0.215 (0.101) Academic Rating=4 Academic Rating=2 Academic Rating=1 Extracurricular Rating=5 Extracurricular Rating=4 Extracurricular Rating=2 Extracurricular Rating=1 Athletic Rating=5 Athletic Rating=4 Athletic Rating=2 >=2 Academic and Extracurricular >=2 Academic and Athletic >=2 Extracurricular and Athletic Model 5 -0.041 (0.157) -0.263 (0.165) -0.240 (0.171) -0.496 (0.175) -0.541 (0.165) -0.304 (0.128) -0.171 (0.126) 0.241 (0.127) -3.990 (0.626) 1.425 (0.090) 4.094 (0.156) 1.147 (0.215) -1.301 (0.393) 1.990 (0.082) 4.232 (0.169) 0.761 (0.087) -0.182 (0.038) 1.368 (0.114) -0.143 (0.088) -0.149 (0.116) -0.446 (0.101) Personal Rating=2 Personal Rating=1 >=2 Academic and Personal >=2 Personal and Extracurricular >=2 Personal and Athletic Observations Pseudo R Sq. 142728 0.078 142700 0.26 142700 0.262 136061 0.283 128422 0.556 Model 6 0.012 (0.167) -0.232 (0.174) -0.246 (0.182) -0.409 (0.185) -0.631 (0.174) -0.303 (0.135) -0.188 (0.134) 0.247 (0.134) -3.915 (0.633) 1.941 (0.128) 5.122 (0.185) 1.109 (0.227) -1.122 (0.408) 1.810 (0.108) 4.215 (0.187) 0.734 (0.092) -0.043 (0.041) 1.354 (0.155) -0.041 (0.095) -0.153 (0.125) -0.444 (0.107) 2.415 (0.118) 3.594 (0.538) -0.433 (0.112) 0.026 (0.080) -0.053 (0.117) 128082 0.604 Table B.7.2R: Logit estimates of Harvard's admission decision, expanded sample African American Hispanic Asian American Missing Year=2015 Year=2016 Year=2017 Year=2018 Year=2019 Female Disadvantaged First generation Early Decision Legacy Double Legacy Faculty or Staff Dean's/Director's List Waiver Applied for Financial Aid Humanities Biology Physical Sciences Engineering Mathematics Computer Science Unspecified Academic index AI Sq. X (AI>0) AI Sq. X (AI<0) Model 1 0.486 (0.038) 0.393 (0.037) 0.047 (0.030) 0.010 (0.048) -0.207 (0.164) -0.536 (0.176) -0.542 (0.174) -0.697 (0.174) -0.814 (0.173) -0.025 (0.023) 1.172 (0.041) 0.012 (0.051) 1.632 (0.029) 1.238 (0.046) 0.511 (0.090) 1.260 (0.139) 1.495 (0.053) -0.147 (0.041) -0.066 (0.028) 0.136 (0.035) -0.241 (0.032) 0.308 (0.044) -0.033 (0.038) 0.174 (0.049) -0.006 (0.064) -0.566 (0.076) Model 2 2.290 (0.047) 1.190 (0.042) -0.400 (0.032) -0.233 (0.051) -0.073 (0.172) -0.434 (0.185) -0.544 (0.185) -0.909 (0.183) -0.786 (0.182) 0.245 (0.025) 1.243 (0.047) 0.180 (0.057) 1.448 (0.032) 1.650 (0.051) 0.372 (0.101) 1.410 (0.159) 1.931 (0.059) 0.453 (0.046) -0.043 (0.030) 0.149 (0.038) -0.380 (0.035) -0.186 (0.048) -0.341 (0.041) -0.198 (0.054) -0.379 (0.068) -0.357 (0.084) 1.811 (0.121) 0.323 (0.069) -0.554 (0.115) Admit Model 3 Model 4 2.604 2.815 (0.071) (0.075) 1.271 1.338 (0.061) (0.064) -0.529 -0.321 (0.050) (0.053) -0.290 -0.264 (0.087) (0.092) -0.068 -0.210 (0.173) (0.213) -0.434 -0.430 (0.185) (0.223) -0.544 -0.428 (0.185) (0.223) -0.922 -0.836 (0.184) (0.224) -0.794 -0.746 (0.183) (0.221) 0.247 0.258 (0.058) (0.081) 1.494 1.616 (0.070) (0.106) 0.165 -0.033 (0.058) (0.124) 1.394 1.426 (0.047) (0.075) 1.697 1.720 (0.059) (0.123) 0.377 0.337 (0.101) (0.106) 1.692 1.875 (0.310) (0.319) 2.379 2.449 (0.356) (0.366) 0.464 0.365 (0.045) (0.048) -0.056 0.022 (0.030) (0.073) 0.203 0.260 (0.057) (0.099) -0.335 -0.247 (0.051) (0.095) -0.185 -0.076 (0.062) (0.125) -0.378 -0.368 (0.054) (0.110) -0.175 -0.278 (0.069) (0.142) -0.456 -0.435 (0.083) (0.211) -0.363 -2.048 (0.116) (1.039) 1.740 1.730 (0.120) (0.126) 0.401 0.444 (0.069) (0.072) -0.535 -0.531 (0.114) (0.120) Model 5 3.596 (0.097) 1.908 (0.081) -0.389 (0.066) -0.272 (0.112) -0.825 (0.260) -0.629 (0.268) -0.596 (0.273) -1.014 (0.278) -1.089 (0.271) 0.177 (0.099) 1.640 (0.132) -0.066 (0.159) 1.480 (0.092) 2.141 (0.155) 0.689 (0.130) 2.472 (0.359) 3.301 (0.417) 0.619 (0.060) 0.200 (0.090) 0.213 (0.123) -0.029 (0.117) 0.034 (0.158) 0.046 (0.136) -0.286 (0.177) 0.063 (0.260) -1.302 (1.127) 0.472 (0.163) 0.186 (0.092) -0.669 (0.154) Model 6 3.674 (0.103) 1.959 (0.086) -0.257 (0.070) -0.108 (0.116) -0.615 (0.277) -0.664 (0.285) -0.789 (0.288) -1.243 (0.292) -1.284 (0.288) 0.155 (0.105) 1.527 (0.139) -0.001 (0.168) 1.531 (0.096) 2.329 (0.164) 0.738 (0.135) 2.630 (0.353) 3.246 (0.417) 0.632 (0.064) 0.247 (0.095) 0.242 (0.130) 0.028 (0.124) 0.140 (0.166) 0.133 (0.145) -0.115 (0.185) 0.293 (0.275) -1.281 (1.159) 0.428 (0.171) 0.285 (0.097) -0.730 (0.160) Table B.7.2R Continued: Logit estimates of Harvard's admission decision, expanded sample Model 1 Female X Humanities Female X Biology Female X Phys Sci Female X Engineering Female X Math Female X Comp Sci Female X Unspecified Female X African American Female X Hispanic Female X Asian American Female X Missing Disadv X African American Disadv X Hispanic Disadv X Asian American Disadv X Missing Early X African American Early X Hispanic Early X Asian American Early X Missing Legacy X African American Legacy X Hispanic Legacy X Asian American Legacy X Missing Other Special X African American Other Special X Hispanic Other Special X Asian American Other Special X Missing Model 2 Admit Model 3 Model 4 -0.108 -0.103 (0.077) (0.080) -0.088 -0.092 (0.070) (0.072) -0.020 -0.045 (0.098) (0.101) 0.101 0.109 (0.082) (0.085) -0.061 -0.105 (0.108) (0.113) 0.272 0.200 (0.145) (0.150) 0.027 0.020 (0.166) (0.177) -0.065 -0.099 (0.081) (0.084) 0.018 0.028 (0.074) (0.078) 0.063 0.047 (0.060) (0.062) 0.113 0.127 (0.101) (0.107) -1.023 -1.121 (0.104) (0.108) -0.278 -0.356 (0.096) (0.102) 0.020 0.023 (0.090) (0.093) -0.182 -0.006 (0.162) (0.168) -0.008 -0.054 (0.104) (0.109) -0.081 -0.091 (0.095) (0.100) 0.212 0.149 (0.070) (0.073) 0.016 0.018 (0.109) (0.113) -0.725 -0.716 (0.214) (0.223) -0.536 -0.672 (0.183) (0.192) 0.398 0.331 (0.142) (0.150) -0.080 -0.074 (0.161) (0.171) -0.882 -0.788 (0.349) (0.364) -0.729 -0.692 (0.230) (0.243) 0.377 0.491 (0.160) (0.175) 0.253 0.436 (0.191) (0.203) Model 5 0.044 (0.098) -0.118 (0.088) -0.195 (0.125) 0.003 (0.103) 0.079 (0.140) 0.218 (0.185) 0.107 (0.217) -0.111 (0.105) 0.079 (0.096) 0.172 (0.076) 0.038 (0.130) -1.582 (0.135) -0.618 (0.127) 0.159 (0.115) 0.026 (0.210) -0.092 (0.137) -0.036 (0.125) -0.083 (0.090) 0.037 (0.142) -0.792 (0.281) -0.779 (0.235) 0.626 (0.187) 0.156 (0.215) -1.261 (0.485) -1.343 (0.287) 0.515 (0.208) 0.348 (0.247) Model 6 0.022 (0.103) -0.170 (0.093) -0.233 (0.130) -0.033 (0.109) 0.000 (0.148) 0.052 (0.194) 0.081 (0.227) -0.092 (0.110) 0.076 (0.101) 0.141 (0.080) -0.038 (0.135) -1.565 (0.142) -0.616 (0.133) 0.162 (0.121) 0.024 (0.218) -0.134 (0.145) -0.055 (0.133) -0.069 (0.095) 0.030 (0.147) -0.872 (0.297) -0.736 (0.240) 0.612 (0.195) 0.092 (0.219) -1.267 (0.529) -1.328 (0.295) 0.471 (0.219) 0.291 (0.253) Table B.7.2R Continued: Logit estimates of Harvard's admission decision, expanded sample Model 1 Model 2 Disadvantaged X Year=2015 Disadvantaged X Year=2016 Disadvantaged X Year=2017 Disadvantaged X Year=2018 Disadvantaged X Year=2019 Early X Year=2016 Early X Year=2017 Early X Year=2018 Legacy X Year=2015 Legacy X Year=2016 Legacy X Year=2017 Legacy X Year=2018 Legacy X Year=2019 Admit Model 3 Model 4 0.026 (0.120) -0.049 (0.126) -0.022 (0.131) -0.240 (0.135) -0.475 (0.127) -0.167 (0.093) 0.013 (0.093) 0.196 (0.093) 0.059 (0.158) -0.291 (0.161) -0.308 (0.161) -0.513 (0.165) -0.349 (0.162) Academic Rating=4 Academic Rating=2 Academic Rating=1 Extracurricular Rating=5 Extracurricular Rating=4 Extracurricular Rating=2 Extracurricular Rating=1 Athletic Rating=5 Athletic Rating=4 Athletic Rating=2 >=2 Academic and Extracurricular >=2 Academic and Athletic >=2 Extracurricular and Athletic Model 5 -0.065 (0.150) -0.289 (0.157) -0.223 (0.163) -0.413 (0.167) -0.507 (0.157) -0.382 (0.115) -0.142 (0.114) 0.146 (0.114) 0.264 (0.198) -0.283 (0.200) -0.242 (0.201) -0.424 (0.204) -0.474 (0.201) -2.426 (0.291) 1.206 (0.077) 3.806 (0.145) 0.931 (0.203) -0.952 (0.289) 1.689 (0.070) 3.795 (0.157) 0.713 (0.082) -0.164 (0.035) 1.362 (0.093) 0.076 (0.076) -0.158 (0.098) -0.483 (0.091) Personal Rating=4 Personal Rating=2 Personal Rating=1 >=2 Academic and Personal >=2 Personal and Extracurricular >=2 Personal and Athletic Observations Pseudo R Sq. 148769 0.136 148741 0.294 148741 0.297 141701 0.318 134365 0.555 Model 6 -0.026 (0.158) -0.274 (0.165) -0.241 (0.172) -0.342 (0.176) -0.576 (0.164) -0.421 (0.120) -0.161 (0.120) 0.142 (0.120) 0.209 (0.207) -0.406 (0.208) -0.311 (0.211) -0.571 (0.216) -0.668 (0.210) -2.328 (0.299) 1.512 (0.102) 4.573 (0.163) 0.881 (0.211) -0.739 (0.296) 1.417 (0.092) 3.672 (0.171) 0.694 (0.086) -0.041 (0.037) 1.357 (0.123) 0.216 (0.081) -0.148 (0.104) -0.465 (0.097) -3.346 (1.334) 2.075 (0.095) 3.276 (0.485) -0.224 (0.091) 0.117 (0.071) -0.109 (0.099) 134349 0.599 Table B.7.3R: Share of each race/ethnicity in each admissions index decile, expanded sample Preferred Model Admissions Decile 5 or lower 6 7 8 9 10 White 0.451 0.111 0.111 0.108 0.109 0.109 Admissions Decile 5 or lower 6 7 8 9 10 White 0.452 0.110 0.111 0.108 0.109 0.112 African American 0.781 0.053 0.044 0.041 0.042 0.039 Hispanic 0.690 0.071 0.062 0.061 0.059 0.057 Asian American 0.397 0.112 0.118 0.125 0.124 0.125 Hispanic 0.688 0.072 0.062 0.055 0.058 0.064 Asian American 0.395 0.115 0.121 0.128 0.123 0.117 +Personal Rating * created using admissionsLogitsIndices.do. African American 0.780 0.048 0.040 0.041 0.047 0.045 Table B.8.1R: Logit estimates of Harvard's admission decision with interactions between race and year African American 2015 X African American 2016 X African American 2017 X African American 2018 X African American 2019 X African American Hispanic 2015 X Hispanic 2016 X Hispanic 2017 X Hispanic 2018 X Hispanic 2019 X Hispanic Asian American 2015 X Asian American 2016 X Asian American 2017 X Asian American 2018 X Asian American 2019 X Asian American Observations Pseudo R Sq. Baseline dataset Model 5 Model 6 3.894 3.992 (0.165) (0.175) -0.011 -0.108 (0.195) (0.206) -0.359 -0.240 (0.208) (0.221) -0.191 -0.187 (0.210) (0.222) -0.028 -0.052 (0.204) (0.216) -0.200 -0.111 (0.209) (0.221) 1.850 1.978 (0.148) (0.158) 0.138 -0.003 (0.185) (0.196) -0.190 -0.203 (0.196) (0.209) 0.247 0.268 (0.195) (0.207) 0.348 0.219 (0.193) (0.205) 0.063 -0.012 (0.200) (0.212) -0.524 -0.411 (0.112) (0.119) 0.042 0.068 (0.140) (0.148) 0.075 0.147 (0.149) (0.159) 0.128 0.123 (0.152) (0.162) -0.042 0.005 (0.155) (0.163) 0.200 0.212 (0.156) (0.165) 122,303 119,896 0.531 0.623 Expanded dataset Model 5 Model 6 3.659 3.722 (0.153) (0.160) 0.017 -0.047 (0.181) (0.190) -0.328 -0.225 (0.193) (0.203) -0.049 -0.040 (0.194) (0.204) -0.003 -0.021 (0.189) (0.199) -0.023 0.093 (0.194) (0.204) 1.728 1.802 (0.139) (0.147) 0.239 0.157 (0.173) (0.182) -0.173 -0.170 (0.183) (0.193) 0.353 0.416 (0.182) (0.192) 0.408 0.322 (0.179) (0.189) 0.240 0.217 (0.186) (0.196) -0.490 -0.390 (0.106) (0.112) 0.076 0.114 (0.132) (0.140) 0.146 0.213 (0.139) (0.148) 0.235 0.240 (0.143) (0.152) -0.001 0.056 (0.145) (0.152) 0.248 0.293 (0.145) (0.153) 149,425 144,189 0.569 0.649 *Standard errors in parentheses. Bold and italicized coefficients are statistically different from zero at the 5% level *See Figure 7.1 For the full set of controls Table C.1R: Difference in characteristics for those labeled Standard Strong by race/ethnicity Share Standard Strong Academic Index SAT Math SAT Verbal Share Academic 2 or better Share Extracurricular 2 or better Share Personal 2 or better White 0.120** 227.04* 749.84* 758.06 0.500* 0.159 0.087 African American 0.010* 206.40* 625.00* 615* 0.333 0.000 0.000 Hispanic 0.036* 220.86* 733.64* 685.45* 0.417** 0.083 0.083 Asian American 0.151 230.56 769.50 758.67 0.684 0.175 0.096 Number labeled Standard Strong 127 3 12 114 *indicates statistically different from Asian American rating at the 95% level

Disclaimer: Justia Dockets & Filings provides public litigation records from the federal appellate and district courts. These filings and docket sheets should not be considered findings of fact or liability, nor do they necessarily reflect the view of Justia.


Why Is My Information Online?