secrets of the sat
photo of william g. bowen
William G. Bowen: He is co-author of The Shape of the River and President of the Andrew W. Mellon Foundation.  He is also  former president of Princeton University.
navigation, see text below
In your book, you say affirmative action was what you call a principle success story.

Well it is. It is an extraordinary success story. I mean by any normal measurement, the African-American students, other minorities admitted through these programs have done extraordinarily well. Both in terms of their accomplishments in school but then, I think, even more impressively in what the coaches love to call the game of life. Their progress through graduate and professional schools into occupations of importance and consequence.

I think most striking is the leadership they are contributing in civic and community life. The ratios of Black matriculants leading civic and community organizations to their white classmates, almost twice in many areas, if you group them by professional degrees obtained, whatever. So I think by any normal reckoning, these students have done exceptionally well. And certainly they think that.

I mean you don't find evidence of dissatisfaction, disillusionment, on the contrary. Many students had to work very hard. They didn't always do as well academically as they would have like or as we would have liked. But did they learn a lot? Absolutely. And are they putting what they learned to good use? Not just for themselves but for the society? No question about that.

Talk about some of the measures of success that you mention in the book.

Well first, of course, the achievement of advanced degrees in many of the most demanding fields, law, medicine, business, Ph.Ds. Because increasingly in American life, these are the pathways to opportunities to lead. And what we find is an even higher proportion of Black graduates than of white graduates went on to earn professional degrees, Ph.Ds and so forth. So that itself, I think is significant. But then beyond that they have gone on to very important positions in, really, every walk of life.

So much depends on personal qualities--the ability to get up off the floor and start over, to accept correction and the benefit from it.  Those are traits of character, those are not things that any test scores are going to measure very well. One crude measure that occurs to an economist is, of course, wages, earnings which measure in a crude but not foolish way, what the market is willing to pay somebody. And the average earnings of both the men and the women, African American men and women who graduated from these institutions, when they were 38 years old, when we surveyed them, extraordinary. Huge jump up from the average earnings of all Black holders of B.A.s.

But beyond that, beyond...let me not stop there. Economists are sometimes tempted to stop there. But, of course, these schools admitted these students because they thought that they would not only do well for themselves but that they would contribute to the larger society which, heaven knows, we need. And when you look at their contributions, their participation involvement, volunteer activities of every kind, including civic and community activities, back home if you will. You just find them present in remarkable numbers. In larger numbers than their white classmates and I think that is a tribute to them. To their willingness to take on these tasks. But it is also an indication of what the need has been for minority participation on civic and community boards, every kind of activity of that kind.

Is there a problem with looking at average SAT scores to determine the reliability of how much preference is given?

There are so many problems with the SAT emphasis that one hardly knows where to begin in answering the question. Let me just say first, to make the most basic point. The schools in our study have never believed that test scores are the end all and the be all for anybody. And there are large numbers of students of every race and every background who are being turned down, who have higher test scores than those who are being admitted. Because the people doing the admitting realize that there is much more to potential and promise and ability to contribute both on campus and after campus than ever can be captured in a test score measure.

So this obsession with test scores is, I think, contrary to the way these institutions function and have ever functioned. I think that is what people sometimes lose sight of.

The second point I would make that I think is also basic, is that admission to college is not a matter of rewarding people for what they have done up to some point in time. That is what high school graduation is for. Admission to college is betting on performance, placing bets when you are investing in human capital. And when you make decisions among very well qualified applicants... Of course, all the people we are talking about are over an extremely high threshold. What you have to ask is that if this student, compared with this student and looked at in the context of the class as a whole, is going to contribute more to the learning environment on campus and then has a better chance at making a big contribution to society than some other students. And the fact that there are significant differences in average test scores, I think in and of itself tells us little.

Can you talk about merit, and the mission of the university?

I think merit has to be thought about not as some abstract thing, handed down from on high that anyone who can count can ascertain. If only that were that simple. I believe profoundly in admitting on the merit. Absolutely. Which means there should not be favoritism, that you should be looking at what individuals can contribute to the school and afterwards.

But does that mean that there is a simple numerical measure of merit? Of course not. Adlai Stevenson, one of the distinguished graduates of one of the institutions in the study used to be fond of saying that he was never threatened by Phi Beta Kappa. No he wasn't? Was he a contributing citizen in this country? Certainly I think so.

One of the important statistical findings in our study is, while SATs along with high school grades are helpful in determining who is over threshold, that is who can do the work and you don't want to admit anyone who can't. But once you get above some level, let's say about 1100 before the SATs were recentered, small differences, incremental differences in test scores make surprisingly little difference in terms of prediction of either success in college or even more fundamentally, success after college. There is just too much else going on.

So much depends on personal qualities. On the ability to get up off the floor and start over. The ability to accept correction and the benefit from it. Those are traits of character, those are not things that any test scores are going to measure very well.

Are we just taking a problem and postponing it for four years?

There are others who will tell you more about life in professional schools especially than I can, what I can tell you on the basis of a lot of evidence in the book is that these students are graduating from the very best business, law, medical schools in extraordinary numbers. And after that they are doing awfully well. They are doing awfully well not just in monetary terms, but in terms of service to their community.

As I recall the statistics in the book, something like twice as many Black doctors as white doctors from this group of students are serving as leaders in civil and community activities. You know, from my perspective that is significant. Because that was one of the purposes of this whole enterprise. We have to keep going back and ask what were they trying to do?

And the answer was that in part we were trying to increase the flow of very talented minority students in every aspect of American life so they can contribute and give back. That is what they are doing.

What about the criticism that by admitting students with lower scores than whites, you are setting them up for failure.

There is this argument, this so-called "fit hypothesis" that we have somehow harmed the intended beneficiaries of this process by encouraging them to go to Tufts or Wellesley or Tulane or Yale--whatever the school you want to choose. An African American student with a test score of, let's say, 1100 where the average for everybody is 1300. And the argument goes--this student would have been much better off if only he or she had gone to a school where the average for everybody was 1100, where the student would have fit in better.

Well the data just completely rebut this hypothesis. There is just nothing in it. If you compare African American students with 1100 test scores who went to schools where the average was 1300, with African American students with 1100 test scores who went to schools where the average was about 1100, the ones who when to the schools where the average was 1300 just did much better along any dimension and were more satisfied with their school. They graduated in higher numbers, more likely to go to graduate and professional schools.

So this notion that we have somehow harmed the intended beneficiaries, that they are victims in fact, is really nonsense. There is just not an iota of evidence, certainly in what we have looked at that would support that proposition.

Is looking at elite schools misleading?

No. Looking at the most selective schools is precisely what one has to do if one wants to sort out this debate over race sensitive admissions because it is in those schools that picking and choosing with race as a factor is a really important issue.

Of course, the large majority of institutions in America are not selecting them. They admit all candidates over some threshold because they have places for them. And that is terrific. I mean that is one reason why the country has done so well, is because there is broad-based access to higher education. And we don't believe for one second that these highly selective schools that we are studying are, in some sense, nobler, more virtuous. I don't believe that for one second.

Every sector of American higher education contributes and contributes in very, very important ways. The reason we focus on the academically selective sector is because these schools are in the center of the debate, as are the leading graduate and professional schools of law and medicine and business. Why? Because they have many, many more highly qualified candidates than they have places and so they have to choose on the basis of some criteria.

And the issue for these schools and the country is race relevant as one criterion among others within these schools? And if you want to examine that question, you have to look at those schools.

How much does affirmative action hurt? Talk about the benefits and the perceived benefits that would have resulted from race blind admissions?

Right. One of the complications in this entire discussion is, of course, that so many wonderfully qualified students of actually all races including Black students are turned away by schools that would to have them if they had places. When I was in the President's office, one of the worst aspects of the job was explaining to extraordinary people, parents, children, why we just couldn't take them. Not because we wouldn't love to have them but because on an all things considered basis, we thought that some other set of people would comprise a class that would advance the interest of the university more completely.

Now, does that satisfy the parent of a disappointed child? Of course not of course not. And I had to make that speech at the trustees, whose children were not admitted. So not being admitted is an endemic fact of life in these schools. Now, of course, because there are so many white students and so many white applicants relative to Black applicants, there are large numbers of white applicants who think that they would have had the ten spaces freed up if there had been no race sensitive admission.

Now, the probability of any one of them gaining that space is, of course, low. And in fact we estimate that if there were no race sensitive admissions at all, the probability of being admitted to one of these schools for a white applicant would go up from like 25% to 26.5%. So quantitatively there is not much of an impact. But that does not cause me to dismiss the concerns of the white applicant or of the parents. I mean I think they are genuine. I have children, I understand this entirely.

But we come back to the point of what is the purpose of the exercise and the purpose of the exercise is not to reward people for past achievement. Not to hand out brass rings. But rather to make the best corrective set of bets you can on the future. That is what admissions is all about. And inevitably you will disappoint some high achieving people if you have lots of good candidates.

Can you talk about the analogy you use in the book regarding the perception of being disadvantaged by the affirmative action policy.

Well, Tom King used the analogy of the handicap parking space to suggest that there is a tendency for the disappointed people, the disappointed parkers, the disappointed applicants to exaggerate their chances of gaining admission, either to the school or to the parking space. You know, everyone drives around the parking lot and thinks, ah, only if there were not that space, I would have been able to park.

And the analogy is helpful in explaining that perception but it is, of course, also misleading in many ways in that, whereas the handicapped parking space reserved a space for a person, in these schools race sensitive admissions does not reserve space. And it is really important to know that we are not talking about quotas or set asides or anything like that. We are talking about a probabilistic admission process in which race is considered along with other things.

Would eliminating the SAT be harmful?

Yes. I think the data in our study and other data indicates that it is useful as one factor, again, among a host of factors in helping schools decide especially who is over threshold. It is a helpful measure. It is not perfect and there are other things to look at, but would I want to through away a useful piece of information? Certainly not.

What do you think the downside would be?

If we were to throw out the SAT we would, I think, lose one useful instrument in identifying students of high potential who might otherwise be missed. Who went to some small school someplace. Who took an odd set of courses but who, nevertheless, have a lot of talent. And let's remember why the SAT was invented in the first place.

It was invented in the first place precisely so that the most selective schools would not rely just on the secondary schools that they knew and on and on. That it was intended as a way to democratize institutions of higher education, by alerting the admissions officers of the most competitive schools, Berkeley and others to the highly qualified who might otherwise be missed. And so, would I regret losing that tool? Certainly I would regret losing that tool.

Let's talk about the test score gap.

The black/white test score gap, in terms of the test scores that applicants present is, of course, very important, because in any system in which test scores play a role, it affects the numbers of students who will be admitted from difference groups if one looks only at test scores. And so from that perspective it is inescapable and consequential.

I think that the factors that create the test score gap are still less well understood than they should be, though much more is being learned about all of that. And, of course, I agree with those who say, shouldn't we be working to improve elementary and secondary education so that the preparation of minority students as well as other students will be stronger and better, that there will be a bigger pool, if you will.

And the answer to that is, of course we should be doing that. But we should be doing that, in my view, with our left arm but with our right arm we are doing the best we can with the pools that we have at present. And what I think the evidence in our study shows is that knowledgeable admissions offices can choose among students with somewhat lower test scores, very effectively. They can choose from within the pools of students with still very good scores by a national standard, but lower scores for these schools. Those individuals are likely to do well.

I think one of the more troubling findings in our study is that this gap in performance is, if anything, greater at the high SAT levels than it is at the low SAT levels. Contrary to what a lot of the critics seem to assume.

Put another way, removing from the applicant pool or from the accepted group, those African American students with the lower SAT scores, which I presume is what some people want us to do, would not address the problem of the performance gap. The performance gap problems, which is a serious problem, which we need to work on is more pronounced, if anything at the higher SAT levels and when we...well, I don't know why that is.

I do agree that there are risks in consideration of race and that consideration of race can be abused as can consideration of anything else. And so that is why, at least for me, it is important to combine the flexibility and the freedom to exercise judgment and to consider race along with other factors with a clear obligation to be accountable. To look at the results of what you have done and see if there are results that you can defend and are satisfied with. And if there are not, then you ought to change your policy.

And if the results of our study had shown, for example, the African American student admitted to the most selective schools with SATs in, let's say, the 1100 range were doing poorly, after school especially, in life and in contributing to society, then I would have said wait a minute. Maybe we are not following good practices. Let's look again. But that is not what the data show.

A professor at UCB said Shape of the River is a wonderful book but it is a eulogy for a fallen hero. What is your response?

Well I have not heard the book described before as a eulogy for a fallen hero and I don't believe it is. I really don't. As I talk to the Presidents of Colleges and Universities, of all kinds across the country and to Boards of Trustees, I sense, if anything, a renewed conviction that these policies make a lot of sense and that they are the policies we ought to pursue. Nor am I willing to concede, I just don't think it is going to happen, that the courts in their ultimate wisdom will conclude that the Bakke Rule was a mistake.

Fundamentally, I have too much confidence in the good sense of the American public to believe that when the full stakes are understood that the American public is going to want to see leading selective colleges and universities resegregated, in large measure. I just don't believe that that is what most people are going to want. And so that is why I remain confident that in one way or another the wisdom of enrolling diverse student bodies, Hispanics and African Americans and white candidates together, to live in a world that is ever more pluralistic, will prevail and that ways will be found to permit that to happen.

So I hope that won't happen. I hope people will understand that that kind of dramatic shift in direction, really it would be in some ways diminishing hope. It would be a very serious thing for the country at large. For all of us. For our children.

Is that why you wrote the book? Is Shape of the River an amicus brief?

No. River is not an amicus brief. River is what its forward introduction says it is. It is an effort to look as clearly as one can at how these policies have worked at a set of schools that is really quite representative of the academically selective sector in American higher education. The database on which it tests, the college and beyond database was not created with this purpose in mind. It was created to serve a range of scholarly objectives since those of us who created it thought that we needed to know more about outcomes in higher education generally.

...admissions policies overwhelming majority of academically selective schools in the late '60s and early '70s have achieved the two fundamental objectives that they were intended to achieve.

What were those objectives? First, to enroll a more diverse student body so that the kind of learning through diversity, through living with, going to class with individuals who were different from you could operate more effectively and more fully. This was thought to be important both for the quality of the educational experience itself and as preparation for people who were going to go out and live and work in an increasingly pluralistic society.

So improving the educational experience on campus was one major objective. And this was, of course, the objective that Justice Powell said was entirely appropriate in his decision in the Bakke case.

One of the most telling findings in the book, I think, is the amount of support for race sensitive admissions policies that is reported by the white graduates of these schools who were in the cohorts that we studied. Now 80% of them or 78%, something like that believe that these policies should either be maintained or should be strengthened. Only about one in five favor weakening these policies. Now I think that is eloquent testimony from the white students who were there and the African American students believe in these policies and think that they learn from them, as a result of them, even more convincingly.

What is your response to people who say you are putting far too much weight on these elite institutions?

I would certainly not suggest for a second, the book doesn't suggest that these students would have somehow been driven to oblivion in the absence of these policies. That is crazy. That is nonsense. But it also is true that going to the most selective schools, not only at the undergraduate level, but certainly at the professional level increases ones chances of going on and doing a great many things in life. Many of which are important to the public at large as well as to the individuals and so I don't see a case for excluding the most talented minority students, or large numbers of them, from access to what, at least many people think, are exceptionally good educational opportunities.

But there is also a second point. The second point is, that if you were to exclude, in effect, significant numbers of minority students from leading colleges and universities, you would harm not only those students and what they can contribute, but their white classmates. I mean this line of argument that it doesn't really matter if they don't go to Yale because they will go somewhere else, somehow forgets of ignores the effect of diversity at Yale on the educational experience at Yale, for everyone. And that is a big thing to ignore or forget. It is a critical thing to ignore or forget.

And the book contains so many vignettes that tell us in moving language how much learning occurred for both Blacks and whites from the kinds of exchanges that diversity permitted. I lived through, on the Princeton campus, searing debates over South Africa and whether divestment was an appropriate response to apartheid. I was a staunch opponent of divestment, to the aggravation of the more liberal elements in the students and the faculty. But that isn't the point.

The point is that, that kind of searing debate had enormous educational value. Enormous educational value. The students who were Black and who saw South Africa in a certain way, the student who were white may have seen it quite differently. Did that produce learning that would have been diminished had there not been substantial African American presence? For sure.

What would you say to someone like Sheryl Hopwood when she turns and says, "Look, I was poor and I worked hard all my life and here I get a certain score and someone who gets a lower score was admitted."

What I say to Sheryl Hopwood or to any disappointed white applicant is that I understand your disappointment. I really do, and if there were more places for a well-qualified candidate, we would have been happy to have you as many of the minority candidates who were also disappointed and turned away.

But the fact is, very hard choices have to be made in the admission process. And they have to be made not on the basis of who has achieved a certain test score result at this point in their lives, but on the basis of which set of applicants will really contribute most to the quality of education at this institution and to the larger purposes for American society, to the need of the society for diverse leadership that have got to be taken into account.

The purpose of admissions is not to confer rewards, not to distribute goodies. It is, rather, to advance broad social objectives. The very objectives that have been used since the beginning to justify public support for these schools. To justify tax exemption because they are thought to serve purposes that are important in a democracy. I believe that they do.

home | discussion | who got in? | interviews | the race issue | sat & test prep | history of the sat
the screening process | test score gap | getting in to berkeley | bibliography | links | tapes & transcripts | press | synopsis

FRONTLINE | pbs online | wgbh

web site copyright 1995-2014 WGBH educational foundation



../test/ ../race/ ../interviews/ ../who/ ../talk/ ../