Web Exclusives:Features
a PAW web exclusive column

May 16, 2001:
The culture and value of sports: a roundtable discussion

Introduction

With James Shulman, his colleague at the Andrew W. Mellon Foundation, William G. Bowen *58, the president of Mellon and a former president of Princeton, has written a provocative book on intercollegiate athletics. Published earlier this year by Princeton University Press, The Game of Life: College Sports and Educational Values, has been praised as possibly "one of the most important books on higher education" of the last 20 years (Louis Menand, The New Yorker). The book is novel for its statistical approach and focus on academically competitive schools whose athletic programs seldom come in for criticism.

The Shulman-Bowen study draws from the Mellon Foundation's College and Beyond database, which has tracked the college and postgraduate careers of 90,000 students who entered 30 selective schools in 1951, 1976, and 1989. In terms of the visibility of their athletic programs, the institutions represented in the database run the gamut from Division IA public universities (the University of Michigan and Penn State, for example) and Division IA privates (Duke and Northwestern) to IAA Ivy League universities and three categories of Division III schools: universities (Tufts, Emory), coed liberal-arts colleges (Denison, Williams), and women's colleges (Barnard, Wellesley).

The authors plumb their statistical mother lode to examine prevailing assumptions about the benefits of intercollegiate sports to participants and the schools they represent. Many of the book's more surprising conclusions relate to Ivy League schools and selective liberal-arts colleges, which have long prided themselves on nurturing the student-athlete - with the emphasis on "student." Even while acknowledging athletes' high graduation rates and their general successes in other ways, to Shulman and Bowen, the data suggest that at these schools a growing emphasis on athletics increasingly threatens to undermine institutional goals. It is a trend, they note, that results from an understandable and commendable desire to foster "successful" (i.e. winning) sports programs - a kind of athletics "arms race," as they put it, that is incremental but cumulative in effect. (For an excerpt from The Game of Life about Princeton's decision to cut, and later to restore, wrestling from its athletics program, see "Paying the Price for Sports," paw, February 21, pp. 18-19.)

In the authors' view, the issue for academically selective schools comes down to resource allocation, particularly as it relates to admissions and the recruiting of athletes, who as a group disproportionately rank in the bottom third of their class. As Shulman and Bowen asked in an op-ed article they wrote for The New York Times, "With intellectual capital ever more important, how great a role should hand-eye coordination play in deciding who is given educational opportunity?"

Shulman and Bowen stress that as participants and fans they love sports as much as the next person - indeed, as an undergraduate at Denison University, Bowen was a singles and doubles champion for two years in the Ohio Conference, and he can usually be found in Jadwin Gym cheering on the men's basketball team whenever the Tigers play at home. As they also note, the issues raised in The Game of Life are imbedded in larger societal trends, including the growing commercialization of sports and the increasing competence and specialization of precollege athletes, who often begin concentrating on a single sport in grade school. They note, too, that the academic and social experiences of student-athletes at smaller selective schools are much more akin to those of their nonathlete peers than is true for athletes at "big-time" Division IA schools.

Not surprisingly, many present and former athletes and athletic administrators take issue with the book's findings and question some of its implicit assumptions - for example, that the higher proportion of athletes at Ivy League schools today has the negative effects the authors claim. One such skeptic is Gary Walters '67, Princeton's director of athletics. The following roundtable discussion was initiated in response to Walters's request for an airing of issues raised in The Game of Life. Its participants were Walters, Bowen, Shulman, and Jeff Orleans, the executive director of the Council of Ivy League Presidents (the Ivy League). Representing paw was one of its former editors, Jim Merritt '66. Emily D. Johnson '01 transcribed the session, which took place in the Mellon Foundation's Princeton office.

Discussion

Paw: My first question is for Bill and James. Why did you write this book, and whom did you write it for?

Bowen: For a long time, certainly dating back to my days at Princeton, I was concerned about the slide in the SAT scores of admitted athletes at Ivy League schools and what I saw as the drift of athletic recruitment. That concern led me and several others in the mid-1980s to create the Academic Index. As you know, this is a formula that considers an athlete's admissions credentials in relation to the overall academic profile of students at his or her school. The effort to create the Academic Index made me extremely conscious of how little I or anyone else really knew about the role of intercollegiate athletics in higher education. There were plenty of anecdotes about the subject but very little in the way of systematic data. So we set out to build the College and Beyond database to study athletics at a range of academically selective institutions. This is the same database that Derek Bok and I used for The Shape of the River, our 1998 study of affirmative action at selective schools.

Paw: And what did you find that most surprised you?

Bowen: The extent and spread of what James and I call the athletic divide: the degree to which athletes, especially recruited athletes, increasingly differ from other students at their respective institutions. As a sometime economist and social scientist I've spent most of my adult life looking at data, and I don't think I've ever seen a set so relentlessly consistent. It shows how a phenomenon that was once confined, more or less, to high-profile sports at big-time schools has spread - to lower-profile sports at those schools, to the Ivies, and to coed liberal-arts colleges. The spread is mostly a product of many decisions made over many years by well-meaning people who care deeply about athletics. There's nothing evil or conspiratorial going on here.

Shulman: I was surprised to see the same pattern repeating itself at the Ivies and Division III schools, not only in men's sports but in women's. It isn't a problem confined to football and basketball at Division IA schools.

Walters. I've read your book through twice - including all the end notes and appendices - and some sections three or four times. Along with Jeff and some other Ivy League athletic directors, I take issue with some of the ways you've interpreted the data and how you've arranged and presented it. There are times when you seem to selectively aggregate or disaggregate statistics to reinforce your arguments. I'm concerned in large part about impressions. The way some of the data are presented has led some media commentators to oversimplify your conclusions. It can also leave alumni and the public generally with some wrong impressions about athletics at Princeton. Take, for example, the percentage of varsity athletes at Princeton. On page 33 of the book you have a graph showing male athletes as a percent of all male students, by cohort and division. For the 1989 Ivy League cohort it shows that 27 percent of male students took part in intercollegiate athletics. That's certainly not the case at Princeton, and I doubt it's representative of the Ivy League as a whole.

Bowen: That figure represents, of course, not the entire Ivy League but the four institutions - Columbia, Penn, Princeton, and Yale - in the database. More to your point, you also need to keep in mind that these are students who participated in varsity or J.V. athletics at some point during their four undergraduate years, so that in any given year the actual percentage would be lower.

Walters: Okay. Then, if you turn to supplementary data in the appendix and look at the table on page 366, you find a totally different set of numbers for athletes as a percentage of undergraduates - 9 percent at Columbia and Penn, 18 percent at Yale, and 20 percent at Princeton. The numbers here reflect statistics filed in 1997-98 under the Equity in Athletics Disclosure Act, or EADA. The EADA data include both men and women, and there's duplication and even triplication because two- and three-sport athletes are counted twice and three times, respectively. So in a sense we're dealing with apples and oranges. Even allowing for that, the graph on page 33 appears to exaggerate the percentage of male varsity athletes in the Ivies. Do you agree?

Bowen: I don't at all. The chart is correct for the four schools in the database. It's based on a name-by-name, person-to-person tagging of people in the cohort. We learned early on that the only way to get reliable data was to build up the numbers by individual matriculants. The database was laboriously constructed one person at a time, based on names supplied to us by each school. The data here are absolutely reliable, and the figure of 27 percent is correct - again with the caveat that it applies to students who played on intercollegiate teams at some point over a four-year period.

Paw: That figure includes both recruited athletes and walk-ons?

Bowen: That's right - but not intramural or club-sport people.

Walters: The point I'm making is that the EADA data are absolutely reliable as well, and as I mentioned, for Princeton they show a figure for male and female athletes combined of 20 percent. If you include J.V. athletes it's probably 23 or 24 percent, but then if you eliminate the duplication and triplication in the EADA numbers the most accurate figure is probably 18 or 19 percent. Again, I'm talking about impressions. Alumni who look just at this graph in the front of the book and ignore the supplementary tables in the back would wrongly conclude that 27 percent of Princeton undergraduates are varsity athletes. They wouldn't be aware that the correct figure is 18-19 percent, which includes not just recruited athletes but a substantial number of walk-ons.

Paw: Many alumni probably assume that the day of the walk-on - when a guy like Dick Kazmaier '52 could be recruited to play basketball, then decide to go out for football and wind up winning the Heisman Trophy - is gone forever. Do varsity teams still have walk-ons?

Walters: I have some figures from our exit surveys for the last four years. These are questions we ask senior student-athletes before they graduate. About 30 percent of team members reported they were walk-ons. We also asked how many years they were members of either J.V. or varsity teams. For recruited athletes the average was 3.39 years and for walk-ons it was 2.25.

As an aside, I should also mention that the exit surveys ask questions that help us evaluate coaches on issues of character and values. On a scale of 1 to 5, with 5 being the best, recruited athletes gave coaches a 4.15 for sportsmanship, while walk-ons gave them a 4.13 - statistically, there's no difference. For "ethical behavior" the coaches rated 4.17 and 4.21, respectively. When we asked students if their coaches were good role models, 75 percent answered yes. As director of athletics I've tried to stress what I call "value-based coaching." These results reinforce the notion that at Princeton, coaches are also teachers in the best and broadest sense.

Paw: Do high-profile sports like football and basketball have walk-ons too?

Walters: They do, although you find more of them in sports like track, water polo, volleyball, and crew.

Bowen: One of the admission people we interviewed commented that rowing is the last of the really amateur sports because so few secondary schools offer rowing. Most people who end up rowing in college are spotted by team members in freshman registration lines.

Orleans: My impression is that there are two kinds of walk-ons in the Ivies. One group are kids who try a new sport in college - sports like fencing, rowing, and sometimes squash. Then there are the kids who played a sport in high school but weren't recruited to play in college, but show up at practice anyway. Chances are they aren't going to start on a team, but they still have a reasonable chance of making it and having a good experience. Gary's numbers bear this out.

Walters: I'd like to get back to the matter of student numbers. Through most of the 1980s, anywhere from 21 to 23 percent of the students admitted to Princeton were recruited athletes. We've cut that down significantly in the last 10 years, so that recruited athlete now constitute a little more than 17 percent of an entering class - roughly 200 positions out of 1,150. We've done this by getting coaches to focus on fewer, but better, prospects.

Bowen: That's certainly good news, but it comes with a dilemma which we discuss in the book. If you reduce the number of spots, there is a natural tendency on a coach's part to look for recruits who are very single-mindedly focused on their particular sport, because, if you have fewer good players, your team can't afford much attrition. So if a prospect tells a coach, "I love tennis, but I'm also interested in philosophy and working for the student newspaper," the coach may have reservations about adding that person's name to the list he sends to the dean of admission. Understandably, a coach is reluctant to commit a spot to people who may play for only a year or two because of competing interests.

Walters: Actually, I'm pleased to say that the attrition levels in Princeton athletics have remained roughly the same over the years. The attrition we do have at Princeton and other Ivy schools is one of the prices we pay for not offering athletic scholarships, but that's not something I would want to change.

Paw: To jump around a bit, I have a question about one of the food-for-thought proposals the authors make in the last chapter of the book. You suggest that the Ivy and Division III schools ought to think about the possibility of returning to one-platoon football. In the same breath you say that such an idea is probably doomed from the start. Would it get anywhere if an Ivy League president suggested it to his peers?

Bowen: My answer is no, it would not.

Shulman: This was suggested to me by a Division III admissions director who believed it might accomplish several purposes. Some alumni might welcome a return to the days when players were less specialized and played both ways. The football played would be less "professional" in a positive sense - who says that Bowdoin should be playing the same kind of game as the Baltimore Ravens? And of course, reducing the number of players a football team has to field would help on the gender-equity front.

Paw: What would be the objections?

Bowen: My own objection would be that the fewer football players you would need to recruit doesn't really address the problems raised in the book. And the amount of capital, in terms of alumni relations, that one would have to expend to make that change would be better spent on the broader issues.

Shulman: Can I go back to the Gary's comments about the virtues of athletics? I think it's important to say something here. All of us in this room love sports. I had coaches in little league and high school. Did I find them to be ethical? Yes. Were they role models? Yes. Did I enjoy playing sports and gain from it? Absolutely. None of us would call into question that kids who play sports enjoy the experience and gain from it. As the authors of this study, Bill and I are questioning institutional policy. How big a role should athletic talent and accomplishment play in admissions? How much recruiting should there be? These are policy issues completely separate from whether or not kids have a good experience playing sports. Attending a place like Princeton is a huge opportunity, and the tough questions relate to how spaces in a class are allocated.

At the same time, it's not for us to say what an institution's mission should be. Obviously, that's up to each institution, and we hope the data we've presented and the questions we've raised help schools do a better job of defining their missions. For example, it's fine to recruit people who bring diversity to your campus if diversity is one of your educational goals. And it's fine to recruit lacrosse or softball players if part of your mission is having a winning lacrosse or softball team.

Walters: I'd like to get back for a moment to problems I see with how information in the book is presented and the potential for misreading it. Data drive this book, and it has so much information that a reader has to make a real effort to digest it all.

If you turn, for example, to the spread on pages 48 and 49, you see four graphs, one for each of the four divisions represented in the study. The graphs show the divergence in SAT scores between students at large and athletes in various sports. They present that divergence in a striking way. Now turn to the table on page 313 - labeled Scorecard 2.3 - showing average combined SAT scores for male athletes by athlete status, cohort, and division. Comparing Ivy statistics for the 1951 and 1989 cohorts, you see that for male students generally, SATs went from 1185 to 1337, an increase of 12.8 percent. For athletes in general, SATs went from 1145 to 1270, an increase of 11 percent, and for high-profile athletes they went from 1114 to 1212, an increase of 8.8 percent.

SATs of the high-profile athletes increased less than those of the other groups. Still, high-profile athletes in 1989 had higher average SATs than students at large in 1951 (1212 versus 1176). Then look at the SATs for high-profile athletes at the Division IA privates - the Stanfords and Dukes and Northwesterns of the world. SATs for 1951 aren't available for this division, but if you compare SATs for the 1976 and 1989 cohorts, you see they actually declined, from 1035 to 1003.

I think this affirms that the Ivies must be doing something right. I can be sensitive about your concern over the growing academic divide, but the fact is that many of the policies put in place in the Ivy League have managed it effectively.

One reason for our relative success is the Academic Index, which of course is unique to the Ivy League. But there's at least one negative to the AI. SATs have been rising at Ivy schools. This in turn has raised the SAT threshold for recruited athletes, which makes it harder to maintain the kind of minority participation we like to have in high-profile sports, where minorities tend to be disproportionately represented. Look again at Scorecard 2.3 and you see that the 1989 SATs for high-profile sports were higher in the Ivies than they were in the Division III coed liberal-arts schools - schools that aren't constrained by the AI. I should add that Division III schools in general have a 17 percent minority representation on their high-profile teams, which is higher than the Ivies.

Orleans: You have to be very careful with this comparison. The liberal-arts cohorts in the book include a number of schools whose student-body profiles, whether measured by Academic Index or SATs, are lower than Ivy profiles. So these schools are naturally going to have access to students, including students of color (whether athletes or not), to whom the Ivies won't. Any schools you compare with the Ivies have to have admission standards similar to the Ivies'.

Bowen: Let me say a bit more about the Academic Index, which as mentioned was meant to insure that the SATs of recruited athletes stayed within some reasonable range relative to the SATs of other students. When we did this study I was curious about the actual academic performance of athletes once they'd matriculated. There's a table on page 319 - Scorecard 3.3 - showing the percent of male students with GPAs in the bottom third of their classes by athlete status, cohort, and division. If you look at the 1989 Ivy cohort, you find that in the high-profile sports, nearly 69 percent of men ended up in the bottom third of the class. The comparable figures for the 1951 and 1976 Ivy cohorts are 43 percent and 65 percent, respectively. The 1989 cohort came along after the Academic Index was put in place with the goal of raising athletes' SAT scores. What's surprising - and disappointing - is that the effort to improve things by instituting the AI didn't carry over to how athletes performed in the classroom. It is troubling that 69 percent of these folks wound up in the bottom third of their class. I bet that we'd all be happier if that number were different.

The SATs are a predictor, in part, of how students should perform once they get to college, and as a former teacher I'm concerned whenever students underperform. We need to work together to find ways to change this.

Orleans: I have a specific question here. For Ivy students in the 1989 cohort, 10 percent of the athletes and 9 percent of the students at large attained M.D. degrees. The findings show that athletes cluster "low" in some class-rank or GPA respects, but they also show that a higher percentage of athletes are getting medical degrees. As educators, how should we correlate these performance data with what happens in terms of degree attainment?

Walters: The book has some interesting comparative statistics on the careers athletes choose. Look at the tables on pages 324-327 (Scorecards 4.1 to 4.4). For the 1989 Ivy cohort, 11 percent of athletes pursued an M.B.A., versus 6 percent of students at large. Comparable figures were 10 percent versus 9 percent for a medical degree and 11 percent versus 12 percent for a law degree. At the other end of the spectrum, you have students who pursued Ph.D.s - for athletes the figure is 6 percent, while for students at large it's 15 percent. To my mind, this shows that athletes are pursuing the basic professions and contributing to society in a really constructive and practical way. We all know that the academic job market can't support nearly the number of Ph.D.s turned out by universities, but athletes seem to know this better than most.

To pursue this point further, turn to the table on page 331 - Scorecard 4.8 - and you'll see that for every cohort, male Ivy League athletes have higher earned incomes than nonathletes. If you look at the 1976 and 1989 cohorts, you find that the athletes do even better vis-a-vis students at large than athletes in the 1951 cohort, a group that in my view the book tends to romanticize as being somehow "purer" because athletes back then were supposedly less single-minded about sports and were recruited less intensely.

Bowen: That table, of course, lumps together all occupations, whether they're in the for-profit or nonprofit sectors. As we discuss in the book, the earnings advantage of athletes is concentrated entirely in the for-profit sector and more specifically in financial services. So the shift over time that you point out is primarily a function of the movement of athletes toward financial services. If you look at people in other occupations - law, medicine, and so forth - athletes do not enjoy any earnings advantage.

Our data show that Ivy male athletes go disproportionately into financial services and the corporate world generally. This difference in the career choices of males athletes is striking. That's not a value judgment, just a fact. It's obviously extremely important that colleges and universities turn out people who are going to become leaders in the for-profit sector, where I spend a fair amount of my own time these days.

Walters: My take on this is that as our economic base has changed from manufacturing to service, athletes by and large have made wise choices by shifting to the service sector.

Bowen: Well and good. As the data make clear, we certainly are not suggesting that athletes attending places like Princeton are doing badly in life. By any objective national standard they're doing very well indeed and no doubt have benefited from the education they received. The question that I think universities generally have to ask -I'm not speaking just about Princeton - concerns the emphasis given to athletics in the context of academics: Is it the right emphasis, too much, too little? Most faculty members, I'm sure, would like to see athletes (and everyone, for that matter) do better academically relative to their potential.

We ought to be thinking how to deal with that. I know, Gary, that this is something that you and Jeff are concerned about on a league-wide basis. In support of something you said earlier, it's true that academic standards at all Ivy and Division III schools in our study have risen significantly. That's a welcome fact, but without any question, part of the academic-athletic divide is a consequence of it.

We should also bear in mind - and this is something, Gary, that you pointed out when we approached you at the beginning of the study - that along with the rise in academic standards have come changes in faculties. In terms of quality, background, and expectations, professors today are significantly different from their counterparts in the 1950s. In general, they are better qualified, conduct more research, and have more specialized interests. Many more are women who came of age when fewer women participated in sports at the college level, or they come from abroad and are products of campuses where sports aren't critical to students' lives.

All this is good - we want universities to be stronger academically and their faculties to be more diverse. But to fall back on an athletic metaphor, higher academic standards raise the hurdle for students generally, including athletes. And that is unquestionably an aspect of this dynamic that we have to think about.

Orleans. It strikes me that athletics may be only the most visible of a number ways that our selective institutions have changed over the last 20 years. During this same period, in general, faculty members have become more committed to their academic disciplines and less committed to their institutions. I doubt these institutions have thought as carefully as they might about how these changes in faculties have affected the undergraduate experience.

Bowen: I would agree. Your point also relates to admissions. So much has changed in the way admission offices go about "crafting" a class. The changes are due in part to the shifting expectations of faculty concerning the kinds of students they want in their classrooms. It's not just athletics that has been affected. We talk about this at the end of The Game of Life, and it's pursued in more detail in Crafting a Class: College Admissions and Financial Aid, a book written under the auspices of the Mellon Foundation by Elizabeth Duffy '88 and Idana Goldberg.

Shulman. To get back to Gary's observations about the accomplishments of athletes in the work force: As Bill said, these athletes are very successful, and they wouldn't have gotten into Ivy schools if they weren't smart. We know from the study that a higher percentage of athletes go into business, and particularly into financial services, than students in general, and that in the financial-services sector they tend to earn more money than those who were not athletes.

That's great, but it's not the issue. The issue - the critical question for institutional policy makers and admission officers - is the allocation of academic resources. If you want students inclined toward the world of finance, fine - you can diversify a campus of eggheads by recruiting people with real-world interests, whether or not they are athletes. The fact that athletes tend to be more politically conservative is sometimes invoked as a reason for admitting them for the sake of increasing political diversity on campuses that are overwhelmingly liberal. But we need to be candid about why we're recruiting lacrosse players or swimmers - it's not for the sake of diversity, even though that might be a positive byproduct, but because we want to have a winning lacrosse or swim team. The same holds for minorities. An athletic director at a Division IA school said to me, "If it weren't for sports, you wouldn't see a black face on this campus." I told him, "I would if you offered African Americans full-ride scholarships and made up for lower levels of preparation with academic tutoring. There are plenty in this country who would love to come to your school."

Orleans. I think James is absolutely right. And whether one agrees or disagrees that Ivy institutions "over-recruit" for athletes, it's clear that we have avoided using athletics as a "front" for diversity. A lot of schools have not been as candid about this as we have been, or as successful, given our standards, in recruiting a diverse student body.

Walters: I know that at Princeton, at least, we have avoided using athletics as a front for diversity. At the same time, I have been extraordinarily proud of the role our coaches have played in recruiting really talented student-athletes of color who otherwise wouldn't have come to Princeton.

Bowen: I wouldn't question that for a second. But in this regard, let me add yet another finding from the study that surprised me. When we calculated the effect of athletic recruitment on African-American representation in the student body for the 1989 cohort - this was for all divisions - we found a positive difference of just one percent. More specifically, about six percent of students in the 1989 cohort were African American; without athletic recruitment that figure would have been five percent. Why is that? Because, while you may find more African Americans in football, basketball, and track, they are underrepresented in most other sports. So the net effect is much less than we might have thought.

Orleans: And that goes right to James's point, that you shouldn't confuse the two issues of recruiting minorities and recruiting athletes. Yet the fact remains that overall, recruiting athletes makes at least some small contribution to diversity. A few years ago, the Ivy League in effect "tested" the Academic Index for its possible exclusionary effect on African Americans. We found that, consistently in football and basketball and to some extent in track, these teams have a higher black representation than the student body as a whole. This isn't surprising, given the difficulty of attracting black males to selective institutions. So in that sense, you might say that the Academic Index has given admission officers some structure that allows athletic recruitment also to contribute to racial diversity.

Shulman: That may be true, but I would argue that the recruitment of African American athletes sends a questionable message. I believe that selective schools like Princeton should be sending the right signals to the African-American community - they should be saying, in effect, that we're more interested in you as a writer of poetry or computer code or as a student of organic chemistry or philosophy than as a basketball or football player.

Bowen: One of the people I've stayed in touch with over the years is Jeanne Ashe, Arthur Ashe's widow. She's a friend, and anytime we talk about these subjects, she reminds me that Arthur said over and over, that for African Americans what is important is education, not athletics. He felt strongly that the recruitment of athletes has a very bad effect on the African-American community. I think there is a "signaling" problem, without any question, and it's a national issue.

Orleans: I don't think the issue is that Ivy schools send the wrong signal about athletics. I think it's the very difficult perpetual issue of balancing our overall admission standards with the academic profiles of African-Americans (especially males) produced by America's secondary schools.

Bowen: I'd like to bring up one point we haven't touched on. The book shows over and over again the importance of selection versus treatment, a big theme in contemporary social science. So many of the results we found are clearly a function of selection. By that I mean that the kind of student-athlete you admit to your school ("selection") matters more than that person's athletic experience ("treatment") once he or she is on campus. The discipline and ability to work with others as part of a team - qualities we associate with athletes - may indeed, as we surmise, carry over into vocational life and contribute to professional success, but chances are that the recruited athlete already possessed those qualities in high school. Our data show that, in general, a recruited athlete who may have played two sports for four years doesn't do any better in later life than an athlete who played on one team for a single season. More training - greater doses of "treatment," if you will - does not correlate with outcomes. Which brings us full circle to the question we raised early on: What sorts of students should a school be looking for in the first place?

Walters: I'd like to know how you reconcile that comment with the book's finding that faculty mentoring is a powerful component of academic performance. I agree with that finding. I'm interested because at Princeton we've tried to encourage faculty mentoring of athletes through our faculty fellows program.

Bowen: That's a very helpful question. My comment on selection versus treatment is strictly in the context of athletics. That isn't to say that other kinds of treatment aren't consequential. Faculty mentoring certainly is, and for everybody, athletes and nonathletes alike. But causation works both ways. When a professor takes an interest in a student, the student benefits. But some students are more predisposed than others to seeking out professors.

Walters: I'm concerned that some faculty members, because they've stereotyped athletes, don't reach out to them. Our faculty fellows program is helping to break down those stereotypes, and I'm hopeful that the result will be improved academic performance by athletes.

Certainly no one would argue against the importance of academic performance, but I would make a distinction between academics and education. In that broader context, I believe that quality, value-based coaching provides at least as much long-term benefit as faculty mentoring. Ron Kinchla, a psychology professor, has told me that cognitive, emotional, and social development happens as much on the playing field as it does in the classroom. I'm sorry your book doesn't put more emphasis on the positive educational values of athletics.

Orleans: Being a successful athlete at any age implies that you've assimilated a set of skills and attitudes and disciplines and passions, and perhaps that subject is a different book. But I agree with Gary that The Game of Life doesn't give enough credit to what it means to be an athlete. Someone with natural athletic ability who hasn't developed his talent through hard work and dedication is not going to be either a good athlete or a good student, much less both.

Bowen: I agree with everything you say about the value of athletics, and the book more or less takes all that as a given. And there are places where we do extol the virtues of athletics. I'm thinking particularly of the example set by the 1976 Ivy League cohort of women athletes, who arrived at their schools only a few years after they'd become coeducational. For the most part they weren't recruited, and they played sports with little or no institutional support - no uniforms, no travel budgets. They weren't encouraged to play in any way, shape, or form. They were also less likely to be in the bottom half of their class, and they were more likely to go on to graduate and professional schools. They arrived on campus with a lot of different arrows in their quivers and made the most of their experience, on and off the field. In many ways, I think their experience represented the ideal of college athletics, and that they were the ideal college athletes.

Walters: In critiquing the book, I don't mean to suggest that it doesn't raise provocative issues and serve a very useful purpose. I would concede that at the Division I level there appears to be a growing academic-athletic divide, reflected most egregiously in the disappointing graduation rates among high-profile athletes. Furthermore, the intensification and specialization of high-school and intercollegiate athletics at all levels, including the Ivies and Division III, has certainly increased the amount of time devoted to athletic training in and out of season, perhaps fundamentally changing - in subtle and not-so-subtle ways - the student-athlete experience. At Princeton and the other Ivies, concerns have been raised about the costs, both educational and personal, associated with a commitment to sports. What is happening to the balance in our student-athletes' lives? Is there ample time for engaging in cultural activities on campus? My biggest concern centers on the student-athlete on financial aid who also has to work 10 hours a week, the ultimate challenge in multi-tasking. Addressing these and other issues raised by Bill and James in their book is important, and the 1994 trustee report on athletics provided the impetus for us at Princeton to confront these challenges and attempt to better integrate the student-athlete experience with the educational mission of the university.

Bowen: And I would emphasize the importance of thinking freshly about these matters and working collaboratively. These are tough issues, and they don't lend themselves to instant, magic solutions. At the same time, I think the Ivies and the Division III schools have a greater opportunity going forward to effect constructive change. And Gary gets our vote as the athletic director who's probably most engaged in effecting change.

Walters: I appreciate that. You'd be interested to know that I had a call from a parent the other day, an Ivy alum who's son is an athlete here. He said, "My son was recruited by any number of Ivy League schools, including my alma mater, but he chose Princeton instead. He did so after visiting the campus and seeing how the guys on the team really had interests beyond athletics." It was gratifying to learn that the kid made his decision based on the quality of people here.

Any reader wishing to discuss the Game of Life with Jeff Orleans can contact him at jorleans@princeton.edu