Toward Sensible Affirmative Action Policies 
                 in School Admissions

                    Norman Matloff
            University of California at Davis

                      Asian Week
                     March 17, 1995

Affirmative action makes an easy target for critics. Conservatives dismiss it as "social engineering," while even liberals wince at horror stories in which middle-class job applicants with 1/16 Portuguese blood qualify as "Hispanic."

In the area of admissions to academic institutions (our focus here, which differs from the workplace), affirmative action has seen some abominable implementations. For instance, the (court-ordered) admissions policy for San Francisco's prestigious Lowell High School is so overtly race-conscious that it even invents its own "races," setting criteria for Chinese-American applicants which are different from those for other Asian-American applicants.

Yet even affirmative action's worst critics agree that all of us, not just the poor or disfranchised, lose when society's underclass grows and becomes increasingly distant from the rest of the population. When implemented well, affirmative action can help to remedy this problem.

Though the concept may sound a bit trite these days, role models are indeed important, and affirmative action helps to provide them. Equally important are "reverse role models": I remember seeing two computer science students on campus, one black and the other Asian. The black was explaining some technical point which had mystified the Asian. If the latter had grown up with little exposure to blacks, this experience would help counter any negative stereotypes he had harbored.

But shouldn't the best qualified applicants be given priority? The answer to this question requires a critical examination of just what is meant by "best qualified."

First, one must keep in mind that neither SAT scores or any other numeric measure will be a very accurate predictor of future grades. It thus makes no sense to admit one applicant over another simply because the first applicant had somewhat higher test scores.

For example, in a controversy over law school admissions at the University of Texas, critics complained that the median percentiles on the admissions test were 94 for whites, but "only" 89 for blacks. But of course this really was a minor gap, not the chasm perceived by the critics. These numbers show that _both_ black and white applicants are talented, with typical members of both groups being in approximately the upper 10 percent of the population.

Sadly, many of the students (and even their parents) do not see it this way. They view it like the Olympics, where being a tenth-second faster can bring one gold instead of silver. But what on earth is the relevance of the Olympics to school admissions?

Many students who complain about affirmative action in university admissions policies feel that higher SAT scores give one some inherent "right" to admission to a famous school. But when I ask whether they take advantage of the special virtues of such a school---do they attend the public lectures by world figures who visit, do they make use of the extensive library facilities, do they make it a point to take courses from Nobel laureates---the answer is almost invariably no.

On the contrary, these students usually admit that they simply want the prestigious name that the school offers. With such a confession, they clearly lose any right to the moral high ground they claim on the basis of something like SAT scores.

The case of Lowell High School mentioned earlier is even murkier. Its principal, Paul Cheng, says that one of the main factors behind the fervent desires of parents to get their kids into Lowell is the school's wide variety of Advanced Placement (AP) courses. These prepare students to take national AP examinations which, if passed, grant college credit. The attraction of this, Cheng points out, is parental savings in college tuition dollars. In that case, maybe admissions priority for Lowell should be given to students intending to study at expensive colleges, rather than applicants with higher test scores.

That last suggestion is of course intended to be facetious. Yet it again illustrates that "merit" is an ill-defined concept with ill-defined social goals.

On the other hand, numeric measures do have _some_ power in predicting success at a school. And make no mistake about it---it is undesirable to admit someone who is "not in the same league" with typical students at that school. Admitting students with little hope of success makes it difficult for faculty to assign challenging work to their classes, for example.

And more importantly, it sets up the unprepared students for failure. This is anathema to a central goal of affirmative action, which is to build self esteem, not destroy it.

Thus, some threshhold values should be established for numerics such as SAT scores, below which there is little expectation of success at this school. But, after having set such threshhold values, there is, except for very special cases, really no defensible reason for further comparison of test scores among applicants.

Once the applicant pool has been narrowed in this manner, a sensible policy would be that school admissions officers use a lottery for selecting applicants. At most schools, this would produce the desired diversity in race and gender which we affirmative action advocates consider so important. But at the same time such a procedure would be simple, fair and race/gender-blind, which would go a long way toward answering affirmative action's critics.


Norman Matloff teaches at the University of California at Davis, where he was formerly chair of the Affirmative Action Committee.