Top Colleges Should Select Randomly From a Pool of ‘Good Enough

Is the admissions process a good one?

by Barry Schwartz (Swarthmore)

(Originally published in the Chronicle of Higher Education, February 25, 2005.)

 

Jane is preparing for a dinner party. For dessert she intends to make a Grand Marnier soufflé. She’s made it before, and it’s come out fine, but she’s been wondering whether all the elaborate steps in the recipe are really necessary. She’d like to experiment, to see if the preparation can be simplified. But she won’t experiment today. Today she’ll follow the recipe as she has before, because she wants to be sure the soufflé works.

The point of that example is twofold. First, the way to get to the bottom of things is to experiment. By manipulating the world, we can find out how it operates. Experimentation has no equal for unlocking the world’s secrets. But second, experiments carry risks. If we actually care about results, we can’t always afford to experiment. Both the power and the risks of experimentation are embodied in science. To cope with the risks, science invented a domain for experimentation that is essentially risk-free — the laboratory. There, even failed experiments are informative, and nobody has to eat a leaden soufflé. Engineers can’t do experiments with the bridges they build. But they can do experiments in laboratories that lead to new techniques of bridge design.

Schools should be laboratories, too. Since the intended “product” of education is learning, in an important sense there can be no failed experiments. Real mastery in the classroom demands risk taking; it demands experimentation.

Yet recent college-admissions seasons have shown us that experimentation today is discouraged in many secondary-school classrooms — because so much is riding on results. Among today’s top-achieving high-school students, the future seems to depend on their getting into selective institutions like Harvard, Stanford, Yale, or my own college, Swarthmore. Despite the fact that those institutions now cost about $40,000 a year, they’ve been receiving record numbers of applicants, sometimes approximately 10 for every available place. Those who apply are the best students at their respective high schools. Almost every one of them is good enough to succeed at their college of choice. But only one in 10 is given the chance.

Why such intense competition? It is probably a reflection of a widespread belief that the United States has become, in Robert H. Frank and Philip J. Cook’s words, a “winner-take-all society.” Although a good deal has been written about the unfortunate consequences of living in such a society, the focus has characteristically been on the losers. I want to focus here on the winners.

I believe that intense competition creates a classroom where only results matter. It makes the stakes so high that students can’t afford to take risks. Everything they do is calculated to produce better credentials — high grades, great SAT scores, impressive extracurricular activities. They choose classes that play to their strengths, rather than those that might correct their weaknesses or nurture new interests.

As to their parents, they spend as much as it costs for a year at an elite college on test-prep courses, personal tutoring, and college counselors who help “psych out” the people in admissions. Parents write their children’s college-application essays. Now on the scene: “getting into college” summer camps, costing as much as $3,000 for two weeks, to buttress the strategic work done during the academic year. Following the carefully crafted recipe begins so early that decisions about preschool are now made with an eye on the college-admission prize.

Such intense competition sacrifices risk taking, intellectual curiosity, and the desire for mastery on the altar of demonstrable success. As a result, even though applicants look better than ever, they may actually be entering college with less learning. Moreover, since they are doing the work they do in and out of school for the wrong reasons — not because they are interested in learning, but because they are interested in succeeding — the intense competition undermines their motivation to continue to learn for the sake of gaining understanding once they enter college. A great deal of laboratory research by psychologists like Teresa M. Amabile, Mark R. Lepper, Edward L. Deci, Richard M. Ryan, and myself has shown that when extrinsic rewards (money, prizes, awards) are provided to people for participating in activities that are interesting enough to sustain engagement for the intrinsic satisfaction they bring, intrinsic motivation is undermined; people no longer participate in the activities in the absence of the rewards.

In addition, the quality of performance deteriorates as people start doing the minimum necessary to produce rewards. To cite one example, when nursery-school children were given “good-player awards” for drawing with special pens, they were less likely than those who had drawn with the pens but had not been given awards to use the pens at all, and they made less elaborate, less creative drawings if they used the pens when the awards were not available. The prizes had turned play into work, as the authors of that study put it. The lesson from such research is that even those who excel enough to get into Harvard are likely to burn out — to be less inspired students — once they have achieved their goal.

By making students so competitive, our selective institutions are subverting their own educational aims. They are admitting students who have done things for the wrong reasons in high school, and who are likely to be disappointing in college.

Another side to the competition is as wasteful, if not as destructive. Just as students compete for admission to selective colleges, colleges compete for outstanding students. Thanks in part to the perverse effects of college rankings, the profiles of admitted students and the percentage “yield” from those who have been admitted are public information that plays a significant role in the status of a college.

All selective institutions fear the “death spiral” in which lowered rankings and reputation lead to worse students, which lead to worse rankings, and so on, until an institution falls out of the selective group. Thus we have merit-based financial aid crowding out aid based on need, ever more elaborate recreational facilities on campuses, glitzy viewbooks and videos, fancy Web sites, fun-filled weekends for admitted students — untold millions of dollars spent each year trying to convince the students you’ve admitted that yours is the school for them. That’s even though, if they choose not to come, there would be a half-dozen perfectly good but rejected students who could fill their spaces.

In the college-admissions game, both the applicants and the institutions are behaving like what I have elsewhere called “maximizers.” Both believe that only the best will do. Research by my colleagues and I has shown that such maximizing is a self-defeating strategy, at least subjectively, in that it leads to stress, anxiety, frustration, regret, and, ultimately, disappointment with outcomes that are excellent but fail to meet either expectations or aspirations. Maximizers may do better than nonmaximizers (we call them “satisficers”) objectively, but they tend to feel worse about how they do.

Maximizing is almost certainly a self-defeating strategy objectively as well. Many years ago, the social scientists Detlof von Winterfeldt and Ward Edwards articulated what they called the “principle of the flat maximum.” The principle asserts that in many situations involving uncertainty — and college choice is certainly such a situation, from the perspectives of both applicants and institutions — the likely outcomes of many choices are effectively equivalent. Or, to put it perhaps more accurately, the degree of uncertainty makes it impossible to know which excellent school (or student) will be better than which other excellent school (or student). Said another way, there are many “right” choices.

Uncertainty of outcomes makes the hair-splitting to distinguish among excellent colleges or students a waste of time and effort. There is more uncertainty about the quality of the student-college match than there is variation among colleges — at least within the set of excellent, selective colleges. So once a set of “good enough students” or “good enough colleges” has been identified, it probably doesn’t matter very much which one you choose; and if it does matter, there is no way to know in advance (because of the inherent uncertainty) what the right choice is.

Given that both institutions and students are distorting their goals in what amounts to a fool’s errand, is there anything, other than hand-wringing, to be done? After all, selective colleges can admit only so many students; if 10 times that many seek admission, competition seems inevitable.

Not so. There is a simple step that selective institutions can take that will sharply reduce competition and thus change the distorted adolescence that many of our most talented students now experience. All that is required is this: When Harvard, Stanford, Yale, and Swarthmore get their applications, they can scrutinize them — using the same high standards they currently use — and identify students good enough to be admitted. Let’s assume that would cut the pool by half or two-thirds. Then the names of all the “good enough” applicants could be placed in a metaphorical hat, and the “winners” drawn at random. While high-school students might have to distort their lives to be the “best” to gain admission to Harvard, they won’t have to distort their lives nearly so much to be “good enough.” The only reason that would remain for participating in all those enrichment programs and attending high-powered pre-schools would be interest, not competitive advantage.

This modest proposal may seem preposterous at first blush, but it isn’t. There is little doubt that any random fifth of the applicants who might survive an initial screening would make a fine first-year class at Harvard. Stanford could fill its entering class with applicants who had near-perfect scores on their SAT’s and still have plenty with such scores left over.

Further, while admissions people like to believe that they have the discernment to look at 8,000 wonderful applicants and pick, with high accuracy, the 1,600 “superwonderful” ones, there is a huge literature on decision making, much of it reviewed in a classic article in Science 15 years ago by Robyn M. Dawes, David Faust, and Paul E. Meehl, which makes clear that people in such positions are much more confident of their abilities than the data warrant. In other words, picking a fifth of the 8,000 at random might be just as good a way of producing a great class as the tortured scrutiny of folders that is the present practice.

It is easy to understand why admissions officers might deceive themselves into inflating their powers of discernment. After all, every year they admit a class, and it does just fine — even better than fine. So they must be doing something right, they tell themselves. That is an example of what psychologists call “confirmation bias.” The real test of discernment would be to admit the near-miss rejects and see how well they do. To my knowledge, no one has done that particular experiment.

In discussing this proposal with friends and colleagues, I have heard several objections. First, people argue that in a meritocracy like ours, anything as important as college admissions should not be determined by a roll of the dice. To that I have two responses: If you accept the principle of the flat maximum, which particular selective school a student attends is not that important. And if you accept how imperfectly admissions are currently done, it is largely a crapshoot already. We just pretend that it isn’t.

Second, people argue that all a proposal like mine does is focus competition on getting to the right side of the cutoff line between good enough and not. As long as the competition continues, and applications keep looking better and better, the cutoff line will be nudged upward and nothing will have been accomplished. To that, I also have two responses: Nothing short of total randomness in admissions decisions can eliminate competition completely. And admissions professionals must resist the temptation to keep adjusting their criteria upward. If they have a sound idea of what a “good enough” applicant looks like, there is no reason why they should change it because some applications appear to exceed it by a wide margin. For my proposal to work, colleges must honestly adhere to the “good enough” criterion.

Third, people point out that an admissions procedure like mine would have no effect unless many, if not all, selective colleges adopt it. That is true, and since colleges and universities are not permitted to collude in setting admissions procedures, one can only hope that many institutions would see the good sense in the proposal and embrace it.

Fourth, what does one do about early decision? Exactly the same procedures could be followed for early decision as for regular decision. Still, there is something to be said for instituting a lower cutoff line for early decision, because one of the best indicators that an applicant will do well at a college is the student’s conviction that that college is the right choice. Early-decision applications are an index of that conviction, although, because of the intense competition for admission and the belief that early-decision applications improve the chances of admission, it is increasingly becoming an application strategy rather than a sign of real commitment.

Fifth, what about students with special talents or personal attributes — violinists, dancers, committed social activists, members of ethnic and racial minority groups, and, of course, athletes? That is a real problem. If people with special attributes are excluded from the lottery and evaluated in the old-fashioned way, then word will quickly filter down to high schools about the subset of activities likely to alter chances for admission. The set of activities that high-school students will engage in for the wrong reasons will shrink, but they will not be eliminated. Therefore, for the lottery idea to work, it must be adopted with virtually no exceptions. Criteria for “good enough” can be sufficiently flexible that applicants who are athletes, violinists, members of minority groups, or from Alaska get credit for those characteristics, but there can be no guarantees of admission. I know that many institutions say they don’t guarantee admission. If that is true, it will make my proposal easier to put into effect. My suspicion, however, is that some special talents are currently weighted so heavily that it is hard to imagine what could possibly tip the balance against admission.

Finally, what about the possibility that some unlucky but deserving student will get rejected everywhere? If a procedure like mine were widely instituted, there would no longer be any “safety” schools — colleges that were selective, but not as selective as Harvard, Stanford, Yale, and Swarthmore. I have nothing wise to say about that problem. If students apply to enough colleges, the probability of meeting admissions criteria at many of them but being admitted to none would be low. But it would never be zero.

I have no doubt that there are other potential problems with my proposal, that some of them will be really difficult to solve, and that some will become apparent only after such a procedure is in place (we must always be on the lookout for unintended consequences). If it were simply a question of improving a system that isn’t broken, only a fool would undertake the task of anticipating all the difficulties and finding ways around them. But I believe strongly that it is not a question of working to improve a system that is serving us well. The system we currently use is badly broken, and no amount of minor tinkering will set it right.

With an admissions procedure like the one I’m proposing, for the most part the desperate efforts of high-school students to climb to the top on the backs of their classmates could stop. Secondary schools could once again be places for experimentation. Learning could once again be guided by curiosity rather than competition. Adolescents could once again devote at least some of their time to figuring out what kind of people they are and want to be. And the result of decreased selectivity, I’m convinced, would not be worse students at our most selective institutions, but better ones. We would still have to work on reducing the competition among colleges for students, but at least the students they were competing for would be more worth having.

—

Barry Schwartz is a professor of psychology at Swarthmore College and his most recent book is The Paradox of Choice: Why More is Less.