ale University president Richard Levin launched an important debate this year by charging that early-decision plans at American colleges and universities promote inequity, unhealthy homogeneity, and mission creep. The only one who benefits, he told a reporter, is the admissions office.

A stern editorial in The New York Times on December 16, 2001, agreed with him that the 200 or so institutions that use early decision stand to benefit more than their students do. Although a Harvard study has shown that students improve their odds of acceptance by applying early, still "the early-decision process is harmful to most students," says The Times.

Critics claim that early decision is primarily an institutional gimmick for boosting the admissions yield, since candidates agree in advance to matriculate if accepted. It saves a lot of trouble, counter the advocates, by relieving high-school students who are completely certain of their college choice of the burden of multiple applications, and the stress of awaiting a decision late in their senior year.

Duke, like most of our peers, uses early-decision plans to admit a portion of the class. We believe that such plans can serve both students and institutions well under certain circumstances. However, like a number of our peers and some parents and guidance counselors, we are concerned that the increasing use of early-decision plans is skewing college choices prematurely for many students, threatening to compromise the diversity and heterogeneity of admitted classes, and, as The Times editorial says, making the senior year of high school "almost irrelevant."

The main problem with early decision is that it has become so popular--or at least so prevalent--that it sweeps up within its purview many students whom it was never designed to serve, and for whom it may actually be counterproductive. If students believe, with some justice, that their chances of admission are greater if they apply early, and if most of their peers are doing so, the pressure to choose a college by the summer after their junior year becomes very hard to resist.

Early decision was originally designed to cater to the student who has wanted to attend a specific college all her life, never wavering in her loyalty. This is the student who has had a Blue Devil on her pillow and a Duke poster on her wall since birth, and for whom Duke is seen to be a good fit by those who know her and the institution well. For such a student, there is no point in applying to several schools unless the first choice doesn't work out, and early decision gives her that information in plenty of time to apply elsewhere. If the student is admitted, the college gains a committed matriculant without having had to screen large numbers of applications. Everybody wins--but only so long as other equally strong students, unsure for very good reasons of their college choice until they go through the application process, are not excluded.

The key to the integrity of the program, then, is to keep the number of early-decision admits to a reasonable proportion of the whole admitted class. Duke has normally admitted about a third of the class under early-decision plans. As the number of early-decision applicants goes up, as it did last fall, we simply admit fewer of them and protect the same number of slots for regular-decision applicants.

However, the clear short-term advantages of early-decision plans for the university and for the student mean that some institutions, including some that are very highly regarded, now fill a higher percentage of their classes with early-decision candidates. It is not uncommon for more than half an entering class to be filled through the early-decision process.

What's wrong with that? The problem is that early-decision applicant pools are heavily weighted toward certain kinds of students--particularly white, upper-middle-class Northeasterners from private schools. These are the students who have the information and advice that leads them to understand that early application can increase their chances of admission. At Duke, for instance, while we admitted 32 percent of the early applicants this year, we average around 23 percent from our regular pool. But some of these early applicants may not really have their hearts set on coming to Duke. They may simply feel the pressure to choose a college--any college--by October of their senior year in order to be part of the success system. With more information, and more opportunity to reflect on that information, they might in the end choose another institution that fits their talents and ambitions more closely.

Equally important, other students from other parts of the country or the world, from smaller schools or families with less sophisticated counseling--students for whom Duke might be the best possible fit--have a slightly lower probability of being admitted if they apply regular decision. We want plenty of these students, too. Among them are many of the kinds of students not well-served by early decision. They have eclectic and wide-ranging tastes, may not have a well-formed preference by their junior year of high school, and will gain useful information on the strengths of different colleges from the admissions process itself, frenzied as it may seem. They may need more time to build their résumés; their self-image and goals may still be rapidly developing at age sixteen or seventeen or they may need more information about financial aid available at different institutions before making their decision.

From the institution's point of view, if early admission allows too many students to come in early, it denies the chance to admit large numbers of students who would enrich the class with their diverse perspectives, greater maturity, independence, and adventurousness. The more than 15,700 applications to Duke this year for the 1,100 berths available through regular decision has set a record for the university, breaking our previous record set in 1987. Regular-pool applications are up and at all-time highs for all minority groups--including African Americans, Native Americans, Latinos, and Asian Americans--and for international students, to whom financial aid has become available for the first time. In short, we are looking at a bumper crop of future leaders from all parts of our country and, increasingly, the world. Not only are many of these students personally well served by Duke, they enrich and diversify our classes, and everybody benefits.

Thus, in my view, we should all be concerned about the surge in students applying early and troubled by the downside consequences of this trend, both for the students themselves and for high schools and colleges. Early-decision systems need careful monitoring and a large dose of institutional skepticism, restraint, and self-discipline, as well as better counseling in high school to help students avoid a herd mentality. Institutions should not succumb to the temptation to take advantage of this windfall surge in early applications by increasing the number of such students whom they accept, thus disadvantaging other, equally deserving students, and contributing to a situation where the pressure to apply early could spiral out of control.

At Duke, we want to enroll each year a goodly number of those students who have dreamed of being Dukies since they knew there were colleges in the world. We also want plenty of students who come to the decision that they want to be Dukies relatively late in their young lives, on very good grounds and after careful consideration of some very fine alternatives. Both types of students bring a lot to Duke, and to each other. So long as early decision, properly controlled, can contribute to that mix, we will happily maintain the system. If the general consequences of an uncontrolled system threaten the more basic health of our national admissions processes and our colleges and universities, we, like Yale and other institutions, will have to consider disbanding it.