I have often been asked what (if anything!) I have learned from my many years of experience as a university provost and president. In my attempt to answer that question, Lessons Learned: Reflections of a University President, a recurring theme is that on more than one occasion I failed to look closely enough at real evidence. I sometimes relied too much on what I assumed to be reality—or feared might be reality. I learned that it was easy to succumb to the temptation to believe what I wanted to believe or to accept at face value assertions made by others. There is simply no substitute for framing questions carefully and then looking with a cold eye at properly organized sets of facts. Let me give three examples, each relating to that bedrock function of any educational institution, the admissions process.
Affirmative action. I have been a vigorous advocate of affirmative action in admissions. In company with many others, I have believed that carefully constructed “holistic” admissions policies that take account of race—but without using quotas or mechanical formulas—could improve educational outcomes for all students by allowing individuals from different backgrounds to learn with and from one another. It has also been clear to me that the country badly needs the talents of its minority population. Moreover, the social fabric requires, in Justice Sandra Day O’Connor’s words, that “the path to leadership be visibly open to talented individuals of every race and ethnicity.”
But has affirmative action really worked? In the early years of minority recruitment at Princeton, I was reluctant to analyze data on educational outcomes for minority students because of a fear that the results would “look bad” or be misinterpreted. That was a mistake on my part. As the later findings in the book Derek Bok and I published in 1998 demonstrated, minority students at selective colleges and universities did just fine.
The extensive empirical research that Bok and I presented in The Shape of the River, alongside even more substantial evidence presented in other books, put a stake through the heart of what is sometimes called “the mismatch hypothesis,” which opponents of affirmative action have sometimes used to suggest that minority students are cajoled into attending schools for which they are unqualified and are, as a consequence, stigmatized, demoralized, and doomed to failure. The facts, when properly assembled, show that there is no truth to this allegation. On the contrary, the effects go entirely the other way around: Minority students with given qualifications have done better when they attended selective institutions, where many of their classmates had even stronger credentials, than they did when they attended less selective institutions. The facts are compelling, and it was gratifying to see this research cited in the Supreme Court decisions on the University of Michigan affirmative-action cases.
But no matter how minority students had performed, it would have been better to look directly at outcomes and acknowledge any shortcomings that were found, rather than to suggest by silence that maybe there really was something wrong. Daytime realities are almost never as bad as nightmares.
Early decision. During my years in the president’s office, I thought Princeton’s policy of need-blind admissions and its practice of offering enough financial aid to meet students’ full needs were effectively helping the school enroll students from all backgrounds and providing special consideration to students from modest circumstances. This turned out to be an optimistic assumption, in part because I, like many of my counterparts, misunderstood the degree of countervailing advantage that our schools were conferring on candidates who applied for “early admission.”
As the Harvard economics professor Chris Avery and his colleagues have demonstrated, that misunderstanding represented our failure to ask the right questions of the data. Evidence showing that average test scores and grades were roughly the same for students admitted “early” and students admitted in regular admissions rounds does not prove (as some of us believed) that students applying early received no advantage. Instead, the right question to ask is whether students of given qualifications who apply early have the same odds of being accepted as students with the same qualifications who apply in the regular round.
The answer is: They do not. Students applying early have far higher odds of being admitted than do similarly qualified students who apply in the regular round—the reason being that admission officers are admitting the “early applicants” from far smaller pools of similarly qualified candidates. The value of early-decision programs is even more questionable given that they inevitably favor applicants who are more knowledgeable about the process and who are well connected.
Athletic recruitment. As president of Princeton, I was at least dimly aware that there were problems with the ever-increasing intensity of athletic recruitment, even for athletes in the so-called minor sports, and even in colleges without big-time basketball and football programs. Many such colleges and universities admit large numbers of recruited athletes whose academic credentials do not measure up to those of their fellow students.
One of the mistakes I made in attempting to address this problem in the Ivy League was to put too much emphasis on the presumed curative powers of an “academic index” that focused on the academic qualifications of athletes when they are admitted. But, in fact, what is really important is the academic performance of athletes once they are in college.
Evidence indicates that, in general, recruited athletes do even less well academically than one would expect them to do on the basis of their (below average) entering credentials. A far from inconsequential number of these students fail to take full advantage of the educational opportunities that they are offered. Research that Sarah Levin and I later carried out demonstrates that the root of the problem is not so much the time athletes spend practicing and playing as it is the recruitment process itself, and the interests and motivations of the different sets of students captured by it.
Other examples of assumptions that fail in the face of facts abound, as Lessons Learned illustrates. To be sure, some problems and some opportunities are unique to particular institutions, but many of the most important questions that present themselves are more common than not. And, as I now know, framing the right questions and seeking out the facts, while often difficult and sometimes awkward, ultimately pay big dividends.
