October 11, 2012

A 'Magisterial' Work on Affirmative Action

By John S. Rosenberg


"Mend it, don't end it" was the famous advice on affirmative action from Bill Clinton, who did neither. There are, of course, other useful slogans, such as "Muddle it," which the Supreme Court essentially did in the 2003 Gratz and Grutter cases. The Court held that the University of Michigan could not give a fixed number of points to minority applicants but that its law school could give even more substantial preferences based on race so long as it sufficiently disguised what it was doing under the smokescreen of individualized, "holistic" review.

Now under new leadership and with a few new members, the Court will see if it can do better when it decides, after hearing oral arguments this week, whether the University of Texas is allowed to supplement its successful, facially race-neutral diversity-producing "top 10%" admissions policy by taking race into account in the admission of other students. 

Is Race a Tiny Factor?

If the Court continues to accept racial preference so long as it is sufficiently disguised, Texas should win easily. Consider this labyrinth: For applicants not in the top 10 percent of their high school classes (about 19% of those currently admitted), the measure of their qualifications is a combination of their Academic Index (AI) score, which reflects grades and test scores, and Personal Achievement Index (PAI) score, which is where the process goes subterranean. According to the University's brief,

an applicant's PAI score is based on two essays and a Personal Achievement Score (PAS). Essays are reviewed by specially trained readers, and are scored on a race-blind basis from 1 to 6. The PAS score ranges from 1 to 6 as well, and is based on holistic consideration of six equally-weighted factors: leadership potential, extracurricular activities, honors and awards, work experience, community service, and special circumstances. The "special circumstances" factor is broken down into seven attributes, including socioeconomic considerations, and -- as of 2005 -- an applicant's race. Race is one of seven components of a single factor in the PAS score, which comprises one third of the PAI, which is one of two numerical values (PAI and AI) that places a student on the admissions grid, from which students are admitted race-blind in groups. In other words, race is "a factor of a factor of a factor of a factor" in UT's holistic review.

Of course if race is such a teeny-weeny "factor of a factor of a factor of a factor," it's not at all clear why its use should be regarded as compelling enough to carve out a large and permanent exception to the equal protection principle prohibiting the distribution of benefits and burdens based on racial favoritism. In fact, however, as Richard Sander and Stuart Taylor, Jr., point out in their powerful amicus brief,  "among freshmen entering the University of Texas at Austin in 2009 who were admitted outside the top-ten-percent system, the mean SAT score (on a scale of 2400) of Asians was a staggering 467 points above (and the mean score of whites was 390 points above) the mean black score."

As Roger Clegg and I have summarized the first of ten reasons we give to be against diversity as it is practiced and defended in higher education today, "Holistic, Schmolistic"!

Outshining 'The Shape of the River'

The highly anticipated Sander-Taylor book, Mismatch: How Affirmative Action Hurts Students It's Intended to Help, and Why Universities Won't Admit It, was published Tuesday, on the eve of the oral argument in Fisher v. Texas. It is, in a word, magisterial. No matter what the Supremes decide, this work will be regarded as a major--perhaps the major--discussion of the use and abuse of race in American higher education, easily displacing Bowen and Bok's unduly influential The Shape of the River, which it respectfully but effectively eviscerates.

Most readers of Minding The Campus are no doubt familiar with at least the broad outlines of Sander's mismatch theory -- that minorities who receive preferential treatment in admissions, and thus attend institutions where their qualifications are below average, cluster in the bottom of their classes, have much lower grades and  graduation rates, and pass the bar at lower rates than their non-preferred peers. As a result, there are fewer black lawyers, engineers, and other STEM professionals than there would have been in the absence of racial preferences.

But even if you know all this, and more, read the book! As someone who has attempted to follow racial issues closely, I can assure you that you will learn, as I did, a great deal that you didn't know and be impressed by the wealth of social science evidence ably and judiciously presented to support and extend the mismatch theory.

For instance, blacks at elite colleges have grades at the 15th-20th percentile of white students, 5% blacks in top 20%; 4 to 5 times more likely to wind up in the bottom 10% of class than whites, but if you look at preferences and not race, race disparity disappears.

Prop. 209 Actually Increased Diversity

One of Sander's and Taylor's most impressive contributions is their thorough and detailed analysis of the effects of Prop. 209 on the University of California system, effects that are widely and routinely misunderstood, ignored, or denied. Among them:

  • the number of blacks entering the UC system after the passage of Prop. 209 declined hardly at all;
  • after the passage of Prop. 209 the number of Hispanic freshmen increased by 22 percent in 2000-2003;
  • the number of blacks and Hispanics receiving Bachelor's degrees increased substantially;
  • from 1995-1997 to 2001-2003 the number of black and Hispanic freshmen who graduated in four years rose 55 percent, the number who graduated with STEM degrees in four years rose 51 percent, and the number who graduated in four years with GPAs of 3.5 or higher rose by 63 percent.

Evidence Authorities Worked Hard to Ignore

Opponents of Prop 209, a group that included all University of California leaders, were almost hysterically fearful that its ban on racial preference would discourage minorities from applying and, if they did apply and were accepted, from enrolling. Sander and Taylor have dramatically demonstrated, however, that exactly the opposite happened. "Black applications rose at seven of the eight UC campuses, and Hispanic applications rose at all eight.... Prop 209 did not 'chill' the interest of minorities in attending the University of California; rather, if anything, Prop 209 'warmed' their interest." Not only did the number of applications increase, but the yield rate increased as well. "In 1998," they found, "the black yield rate at Berkeley was nearly 52 percent, by far the highest in many years and probably an all-time record."

Not only were these results astonishing at the time, but they still are because they were so widely ignored or denied, not only in the media coverage but also by UC officials themselves, whom Sander and Taylor describe as almost willfully ignorant of the good news around them.

Indeed, perhaps the most depressing aspect of the story told here is the rigid, close-minded ideological resistance of academic leaders and their liberal allies even to consider the evidence that Sander, Taylor, and other scholars have amassed demonstrating the damage done by racial preferences and the rewards that would flow from eliminating or minimizing them.

Anyone who doubts the pall that defending racial preference at all costs has cast over campuses today will have a hard time explaining the efforts detailed here to block access to data that might prove embarrassing and the examples given of data being manipulated and in one case even being manufactured by ostensibly reputable scholars.


Can They Publish a Dissenting Article?

A particularly telling example is that Sander's Stanford Law Review article that ignited the debate over mismatch was initially accepted for publication by the University of Pennsylvania Law Review, but "a few days later a very embarrassed editor called back" to say that "a battle had erupted over whether the UPLR should be associated with something so controversial" and the members of the review had voted to rescind the offer of publication.

Mismatch, in short, is a major contribution to the debate over affirmative action, a model of vigorous but fair and balanced argument and analysis. Its only flaw, so far as I can see, is its failure to agree with me. Well, not just with me but with all those who believe that preference based on race should be ended, not mended.

Sander and Taylor refrain from calling for the abolition of race preferences in large part because they believe that academia is so committed to preserving them that the subterfuges and substitutes they would (and do) create to avoid implementing race neutrality would be worse than a well-regulated regime of carefully limited preferences.

They believe this could be accomplished through insisting upon transparency, limiting the size or scope of preference based on race to preference based on socio-economic status, and "outlawing race-based aid awards." When I asked Taylor Tuesday at a Cato Institute Forum what legal argument would prohibit race-based aid that would not also prohibit race-based admissions, his only reply was a cost-benefit analysis.

Social science thus has its limits.


John S. Rosenberg blogs at Discriminations.

Post a comment

Leave comments here. Unless they are vicious or obscene, they will be printed.




Do We Over-Invest in Non-Traditional Students?
Richard Vedder

The Liberal Arts Are in Trouble--Should We Celebrate?: Part 2

The Liberal Arts Are in Trouble--Should We Celebrate?

Walter Russell Mead: The Coming Reformation of Higher Ed

A Sorry Attack on the Common Core
Sol Stern

All Essays >>>

Published by the Manhattan Institute
The Manhattan Insitute's Center for the American University.