William G. Bowen’s Responses to Discussion Session Comments by Howard Gardner and John Hennessy

THERE IS SO much here that I cannot possibly respond to everything that has been said. Let me first offer a few comments that may be controversial and challenge some of our accepted thinking—including some of my own. The first observation I would make cuts across the two sets of remarks by our commentators. It concerns the stratification issue in higher education.

The difference in circumstances between Harvard, Yale, Stanford, Princeton—the top-tier places in terms of wealth—and everybody else is dramatic, and so it is not surprising to hear what John reported about Stanford’s ability to provide more money to students who really need it, and more services for the students who really need them. The same thing is true at Princeton. It cannot be true at more than five or six places, maybe ten; and this tiny number of institutions together educates a very, very small fraction of the student population. Much more research needs to be done on both the extent of stratification (measured in many ways, not just in terms of money), and what all of this is going to mean over time. This is, I think, a big question.

Much as I would like to see more of the wonderful features of residential life that Howard described so well, and so very eloquently, it is by no means obvious that the resources are going to be there to make this possible—except in the wealthiest sector of higher education. One of the things that I am going to say this afternoon, in my second Tanner lecture, is that it is going to be crucially important to develop skillfully defined mixes of teaching methods that use online technologies to complement the best features of face-to-face interaction. But that is going to require considerable changes in decision-making models and in how we think about who in the academy should make what kinds of decisions. The models we have across the country for governance are models that are fifty, seventy-five years old, and whether they make sense in the digital age is an important question. I have my doubts.

I have my own recollections of the glories of wonderful student-teacher interactions, and of how much I benefited from them. Yet my good friend, William Baumol, who is smarter than essentially everybody, keeps reminding me that teachers are no different than doctors in that they tend to exaggerate their importance. They believe that, whether it is the bedside manner of the doctor or whether it is the twinkle in the eye of the engaging faculty member, they change lives profoundly; however, in most instances, this is very much an open question. As we think about the transition to sophisticated kinds of online teaching, we do well not to be too quick to accept what we want to believe about our own talents and their consequences.

The next point I would make has to do, Howard, with something else you said with which I agree—namely, that we should stop doing so much pandering to student taste of one kind or another. I am glad, John, that you are not proposing that at Stanford. That is good. But there is an important consequence which I am not sure very many universities and colleges either understand or are willing to accept—namely, that if you restrain yourself in this way, you are going to lose some students. My response: so what? I cannot tell you how many times at Princeton I met with admissions people who wanted to talk about yield. I said “Please, please, I do not want to hear one word about yield. I do not care what the yield is. What I care about is the quality of the students that we end up with in the fall. The fact that some number of other students to whom we offered admission chose, rightly or wrongly, wisely or unwisely, to go somewhere else is just fine—no issue for me. It does not make any difference.” The pool of excellent candidates at places like Princeton is so deep that it is unimportant if the university loses a few of them on the margin. Of course, I recognize what a privileged place Princeton is, that the great majority of colleges and universities cannot be so cavalier about yield. But those that can be cavalier should be.

There is also a related point, which plagues the liberal arts colleges especially—namely, fear of rankings. It is even true that some presidents are, apparently, paid on the basis of the rankings. Lunatic! And the rankings are based far too heavily on SAT scores, which predict actually one thing very well: family wealth. They do not predict much else. Please go back and reread the relevant chapter in Crossing the Finish Line.

[Audience comment] Nothing predicts very well.

Well, that is not true. Your statement is an exaggeration. If you look at the Crossing the Finish Line book that my colleagues and I wrote not that long ago, you find a dramatic difference in the predictive power of achievement-based measures versus so-called aptitude tests. What matters is how students did in secondary school, and how they did on achievement tests, not how they did on the ACT or SAT. The differences in terms of predictive effects are dramatic. And when you stop and think about it, this is not really surprising. What does achievement-based performance measure? What does it tell you? It tells you who can get up off the floor after having been knocked down and do fine the next day. It tells you about coping skills. It tells you about motivation. It tells you about a whole set of things that really matter in terms of success both in college and in life. And yet, if places downplay the traditional aptitude-based, test-score measures of “quality,” they pay a price in terms of rankings. My answer to the presidents with whom I discuss this is, “Fine. Let us hope that your board has brain enough to understand that rankings are not the currency in which things should be measured.” So I think, Howard, that to move in the direction you suggest, places have to be willing to let the yield go where it will, to let the rankings go where they will, and to live with the consequences of focusing on the kind of education that we ought to provide—to put our resources where they are most needed.

Back to stratification. We need to distinguish very carefully between what is feasible at places with substantial resources and at places that do not have such wealth. When my colleagues and I were testing out a very sophisticated interactive statistics course developed by Carnegie Mellon, I asked someone who serves with me on the ITHAKA board, Steve Stigler, a world-famous statistician at the University of Chicago, to look at the statistics course and tell me, as an expert, what he thought of it. He said, “Well, you know, it is pretty good. But of course I teach differently, and better than that.”

I said, “Steve, no question. If I have a choice between having a statistics course from you or having this machine-guided form of instruction, which am I going to choose? I am going to choose Steve Stigler in a heartbeat.” But how many students have that choice? The choice for most people is not Steve Stigler versus machine-guided instruction; it is machine-guided instruction versus some unknown quality of instruction. Instructional quality is extremely varied. And so I think we need to be realistic in thinking about what the real choices are, and not glorify a Steve-Stigler option that is not real.

Finally, a word about collaboration. I believe in collaboration. It is, by and large, a good thing if it is handled wisely and in a tough-minded way. But there are dangers in collaboration. Absent efficient modes of decision-making, it is very hard to get good judgments made when nimbleness and truth telling are required. There can be too much politeness, too much inclination to say, “Oh, let me not force that answer on you, even though it is the right answer.” And so too often we end up with lowest-common-denominator outcomes—with one participant in the collaboration insisting on this kind of a test which others know makes no sense. We go along. But it is very important in addressing some problems—not all, and there certainly are areas where collaboration works beautifully—not to be too collaborative. Let me cite athletics as an example. The membership decision-making models in athletics can be terrible. I participated in some of those for too long. It is the same cursed problem of wanting to be nice, of not wanting to offend.