The Shape of the River: Long-term Consequences of Considering Race in College and University Admissions
by William G. Bowen & Derek Bok
472 pages (hardcover)
I meant to finish reading this, I really did; but again, library time constraints beckon. For future reference, I got to page 112, a little under 1/4 of the way through. I will try to write up what I wrote with relative objectivity, but I shan't hold back from interjecting (biased) commentary. From skimming the rest of the book, Bowen and Bok seem to have done a great job documenting the success of minority students admitted to highly selective institutions, years after they graduate (or don't graduate).
For those who don't know, Bowen & Bok are two former Ivy League presidents (of Princeton and Harvard, respectively) better-known within higher education for their later, controversial work on SAT scores and admissions expanded to all races; in comparison, this earlier publication consciously simplifies the world to black and white. In the later study, B&B found that when using white as the baseline, blacks and Hispanic [not necessarily Latin@, since federal race classifications include Spain origin as Hispanic] students received a significant boost while Asian students were actually negative, i.e. in order to be considered equivalent to a similar white applicant, s/he had to score higher on the SAT.
All data in this study was based on a detailed database combining information from 28 colleges and universities, who are listed on p. xxviii-xxix. B&B do a reasonably good job of straddling the fence, providing evidence for and against both sides of the affirmative action debate. I'll be discussing mostly the information that I want to remember, i.e. anti-, but first I'll summarize the opposing details and encourage you to read the book for yourself. In essence, the black students who would not have been admitted under a hypothetical race-neutral process go on to exceptional achievements despite their weaker academic skills.
So let's start with the definition of race-neutral: it is NOT admitting by the numbers. B&B grouped black applicants by SAT ranges and posited that they would have the same probability of admission as white applicants in the same ranges. This causes the overall chance of admission for black students to drop from 42% to 13%, versus an overall chance of 25% for white students. The real-life Berkeley example bore out this hypothesis remarkably, although B&B make no mention of the large increase in Asian students as a result of race-neutral at the UCs (31-3). Similar drops in enrollment would occur at the most selective law and business schools (45), where "black applicants [to business schools] were 2.7 times as likely to gain admission as whites with comparable records. Hispanic applicants were 2.8 times as likely to be admitted" (45-6). For med school, the median pre-med science GPA for accepted students was 3.1 for blacks vs. 3.6 for whites; correspondingly, "the median test scores of blacks accepted to medical schools was lower than the median for whites who were rejected" (46). This establishes that regardless of what happens later--and much does happen later--admitted black students (and presumably Hispanic students) are less qualified in comparison to white students (and presumably Asian students). Speaking of which--I find myself constantly wondering how Asians fit into this world, as Hispanic students are much more similar to black students in profile than Asians are to whites. "The real racial divide in America was and remains black and white" (xxvii, Shelby Steele qtd. by Scott Shepard 11)--I absolutely, positively disagree. Such a statement--from someone implied to be one of the "scholars from the black, Hispanic, and Asian American communities" (xxvii)--erases the existence of Hispanic, APA, and Native American peoples from "America" and the importance of a multicolored discussion of race.
Prior to affirmative action in the late 1960s, "the selective colleges would rather be selective than integrated" (qtd. in 5)--nothing wrong with that wish, in my mind, as the selective colleges should not be lowering standards for the sake of diversity. "Harvard [Law School] began admitting black students with test scores far below those of their white classmates....and black enrollment began to rise" (5). That cause and effect seriously concerns me with its lack of acknowledgment of racial discrimination--to discriminate on the basis of race, which is exactly what happened. A graph on p.27--before recentering, but nevertheless interesting: at a combined SAT score (out of 1600) below 1000, about 10% of whites and 20% of blacks gained admission (I am estimating based on a graph, so bear with me). The gap is largest in the "1200-1249 range, [where] the probability of being admitted was 19 percent for white candidates, as compared with 60 percent for black candidates" (26). Finally, at the 1500+ mark--close to perfect scores--about 70% of whites and 100% of blacks gained admission. Even in the '90s, admission at the most selective schools was never guaranteed by "high stats"--except for those few and exceptional black candidates, it seemed. "To sum up, black candidates are consistently admitted at higher rates than legacies, who in turn are admitted at consistently higher rates than non-legacies, but the 'advantage' enjoyed by legacies is concentrated at the upper end of the SAT range" (28-9).
Socioeconomic AA would not "substantially cushion the effect of ending racial preferences" (Thomas Kane, qtd. in 47), because while URMs are disproportionately poor, "they still make up a minority of all college-age Americans with low incomes" (47). B&B also briefly references the prohibitive cost of implementing such a program to the same degree as current racial AA. To which I say: well, yes, and remind me why a poor black person is more deserving of admission than a poor Asian person, other factors held constant? Socioeconomic AA is not meant to substitute for racial AA; it is meant to help those who will always need more help. In capitalist America, there is no doubt whatsoever that the poor student is always disadvantaged over the identical well-to-do student. Even Questbridge, which I wholeheartedly support, helps only the very tippy-top of the poor students (it is more useful in guaranteeing affordability than in affecting admissions, since most Questbridge finalists would have been admitted on stats and extraodinary personal qualities anyway). [/soapbox]
B&B does unequivocally disabuse the notion of academic "mismatch." Average college GPAs may differ by over half a point (72), but black students still go on to do well in life after college (no doubt aided by affirmative action at law/med/grad schools, says my cynical mind). Interestingly, the SAT remains useful in significantly predicting both academic performance in college and postgrad achievements, although the prediction is stronger for whites than for blacks. Graduating from a top college does affect one's postgrad prospects within higher education. James Thomas, former dean of admissions at Yale Law School: "Someone who has played--and succeeded--on a tough field lets us take a little more risk and admit someone who strikes our faculty as having that something that's going to make a difference in the world. Does this mean that we don't take someone from an off-the-beaten-track school? No, but that person has to have every single thing in line--there's no margin for error" (personal correspondence, qtd. in 101). Note: Blacks who are "specially" admitted go on to disproportionately achieve exceptional accomplishments. I have not discussed that data here because it is irrelevant to the basis of my position on affirmative action, although I have read every word of it.
A very good book, highly recommended as long as you aren't afraid of social science stat-talk (and even if you are). Alas, it didn't change my personal opinion. (I'm still waiting for someone to come up with a brilliant argument that will push me across the fence.) The ends never justify the means, and however beneficial AA is/would be, its very existence remains discriminatory and hypocritical to statements of nondiscrimination.