Lacking available slots, the 2004 financial aid initiative boosted enrollment from families earning under $60,000 by 51 students out of a freshman class of 1,620—a significant but not overwhelming increase. It's a good bet that most of them nudged out outstanding unconnected candidates, rather than legacies, athletes, or development cases.

  One unhooked middle-class applicant whom Harvard passed over was Samantha Baras, who was in the top 10 percent of her class at Brook-line High School in Brookline, Massachusetts. (Among her high school classmates who enrolled at Harvard: President Summers's stepdaughter.)

  Samantha scored 1480 on the SATs—without test preparation or retaking the test—as well as an 800 on the SAT II writing test and the maximum 5 on three advanced placement tests. A vegetarian who plans to be a doctor, she wrote her college essay on her mixed emotions about dissecting animals in biology class: “Nothing feels as bad as that first cut. Nothing feels as good as finding what I'm looking for—whether it's the sixth stomach chamber of an obese earthworm or a specific vein of a frog…. How is it possible for me to live as an enthusiastic vegetarian and dissector?”

  Rejected by four Ivy League universities—Harvard, Brown, Penn, and her first choice, Columbia—she enrolled at New York University, which gave her a merit scholarship. As a freshman there, she compiled an impressive 3.8 average. “NYU is a great school and it has worked out very well for me,” she told me. “There is no question that there are many exceptionally high-performing students at NYU, sharing a similar story to mine, who deserve the same opportunity as Ivy League graduates to rise to the top of our society.”

  Samantha's credentials were “way up there,” said Brookline High principal Robert Weintraub, who added that he called Columbia and other Ivies on her behalf, to no avail. “Samantha does not have legacy. That's sort of a strike against her in terms of getting into highly selective schools. We have tons of parents who've gone to Harvard who live in Brookline.”

  To expand access for low-income students without abandoning the middle class, colleges have little choice but to dismantle the preferences of privilege. If they don't do so voluntarily, public pressure may eventually prod Congress to penalize schools engaging in these preferences by withholding federal aid or revoking their nonprofit status.

  As we have seen, Senator Kennedy proposed requiring colleges to report data on admissions of alumni children. That would be a valiant first step, but more sweeping change is needed—and would likely draw wide public support. I recommend the following reforms:

  1. End legacy preference. Most Americans want to abolish legacy preference, as Oxford and Cambridge in England have already done. Requiring alumni children to compete on even terms would help restore fairness to admissions and improve access to elite universities for unconnected middle-class and working-class applicants.

  Colleges” fears that their traditions and endowments would decline without legacy preference are exaggerated. Since alumni often can afford private schools, test-prep services, tutors, and independent counselors, their children would still enjoy many advantages over other candidates. Moreover, without an admissions entitlement, legacies would likely strive harder in high school. As a result, some alumni children would be accepted on merit, enabling them to share college customs and family lore with classmates. Those who get in would have the self-confidence of knowing they did it on their own, rather than by relying on their parental pedigree.

  Nor would alumni contributions to colleges necessarily plummet, as the examples of Caltech, Berea, and Cooper Union show. For most alumni, giving to their alma mater doesn't hinge on whether it accepts their children; it stems from other motives, such as gratitude for their own education or desire to promote research or teaching in an overlooked field. The few alumni who do respond to an admissions rebuff by reducing or withholding donations are dismayed largely because they were counting on legacy preference to save the day. If they didn't expect preference, they wouldn't be so disappointed. They might also offset their reduced charity to their alma mater by giving to the colleges where their children do enroll.

  Elite colleges could compensate for losing legacy preference by creating other rewards and naming opportunities for donors—much as Caltech named an asteroid after Gordon Moore. And with their endowments in the billions of dollars, they could easily withstand a small decline in giving.

  Even a prominent higher education lobbyist says colleges could prosper without legacy preference. The credo that legacy is essential to fund-raising is “the kind of traditional thinking it's very hard to break out of,” said Gerald Cassidy. “If you keep track of your alumni, and you initiate them early into giving small amounts so they get used to giving, you can raise a lot of money without legacy preference. Particularly if you've conditioned them to the fact that there aren't preferences that could work for their children.”

  Ending legacy preference would require a seismic shift in attitudes among college administrators—a cultural upheaval that apparently hasn't taken hold at Texas A&M. In January 2004, the state engineering school announced that it was dropping legacy preference, which had amounted to a maximum of 4 points on a 100-point admissions scale. It acted under pressure from legislators and civil rights groups; upset that the university had decided not to practice affirmative action for minority applicants, they demanded that A&M also toss out legacy preference for alumni children, an overwhelmingly white group.

  According to A&M administrators, the change has not decreased legacy enrollment at all, apparently because alumni children still have an unofficial edge. Robert Walker, vice president of development, explained that he and his staff advise children of alumni and donors to answer essay questions on the application by writing “everything they can about grandparents, parents, aunts, and uncles that were Aggies. We don't give them any credit for that, but the people who read it can't help” but notice the family ties. “It helps them psychologically.”

  2. Establish a firewall between fund-raising and admissions. As we've seen, most elite universities profess to separate these two functions but ac tually pursue children of prospective donors, celebrities who create buzz, and politicians who bring government funds to the campus.

  A true firewall would mean that college fund-raisers and the president's staff would respect the integrity of the selection process and stop imposing “development cases” on admissions. No longer would movers and shakers such as Dave Zucconi and Joel Fleishman have a say in both fund-raising and admissions. Colleges would rein in the parent fund-raising committees that have morphed into admissions networks for affluent suburbs such as Lake Forest, Illinois. To avoid the appearance of a quid pro quo, every college should adopt a policy that it will neither accept nor negotiate donations from any parent whose child has applied to the institution or is on the waiting list. If the child has been admitted and decided to enroll, there should be a minimum waiting period of, say, two years until the parents can donate.

  Once colleges could no longer dangle admissions slots, they would be forced to compete on quality for nonalumni donations and government funding. Like Caltech, Berea, and Cooper Union, they would attract gifts based on the excellence of their academic (and athletic) programs and the appeal of their missions. To meet its fund-raising target, each college would then have a financial incentive to improve its educational quality and define what makes it valuable and unique—ultimately benefiting students, employers, and American society as a whole.

  3. Develop conflict-of-interest policies for college admissions staff. Any college admissions staff member who is related to or has socialized with an applicant or an applicant's family should not evaluate, advocate for, or vote on that candidate. Obviously, admissions staffers are likely to be biased in favor of applicants they know. But even if personal relationships furnish valuable insights into an applicant, as Harvard and other elite colleges contend, the trouble is that admissions deans and staffers are more likely to be acquainted with children of alumni and donors than with first-generation students from immigrant or low-income families. Conflict-of-interest rules would give underprivileged applicants a fairer shake.

  4. Abolish athletic preference and scholarships for upper-crust sports. I would not entirely eliminate preference for athletes in admissions. Unlike the legacy and development boosts, which credit an applicant for parental achievement, athletic preference rewards a candidate's own hard work and excellence. But the growing number of slots set aside for athletes at elite institutions supposedly devoted to academics has gotten out of hand.

  Recognizing the problem, several colleges—including Amherst and other members of the New England Small College Athletic Conference— have considered capping athletic admissions in some sports. But, as in the fallout from Title IX, such caps are likely to hurt working-class sports such as wrestling, while wealthy alumni and donors shield upper-class sports from cuts.

  In my view, athletic preference and scholarships should be reserved for recruits in major sports that most American children have an opportunity to try. Currently, for boys, the most popular sports are football, basketball, track and field, baseball, soccer, wrestling, cross country, golf, tennis, and swimming and diving; for girls, basketball, track and field, volleyball, Softball, soccer, tennis, cross country, swimming and diving, competitive spirit squads, and golf. Colleges discriminate against minorities and low-and middle-income students by reserving slots—and, outside the Ivy League, scholarships—for athletes in sports only rich white people tend to play. Also, strictly by the numbers, it takes more skill to be recruited in a mass sport such as basketball or baseball than in sailing or squash.

  Colleges could still field teams in less popular sports, but recruits would be evaluated on their academic credentials and would not be eligible for athletic scholarships. Without admissions preference for prep-school sports such as women's crew and horseback riding, colleges would comply with gender-diversity requirements by taking more female athletes from low-and middle-income families and public school backgrounds.

  5. Eliminate admissions preference for faculty children and tuition assis tance plans that put pressure on admissions offices to do so. It has be come standard practice for universities to compensate faculty members and other employees by reimbursing part or all of their children's college tuition. It would be fairer to professors without college-age children for universities to drop this benefit and pay higher salaries across the board instead. If a college does offer such a reimbursement program, the tuition break should be fully “portable”—that is, the children should receive the same subsidy no matter where they enroll. As we've seen, if the college lim its the benefit to faculty children in its own student body, then the admis sions staff feels obliged to accept them—and avoid the wrath of professors forced to pony up full tuition elsewhere. Coming from highly educated families that value learning, faculty children start with an academic edge and don't deserve an admissions advantage.

  In addition, to avoid an appearance of conflict of interest, university presidents, provosts, and other highly paid administrators should be discouraged from sending their children and relatives to their own institutions—particularly if they stand to benefit from a tuition break.

  6. Provide equal access for Asian American students and for international students who need financial aid. If elite colleges were truly com mitted to socioeconomic diversity, they would regard the proliferation of outstanding Asian American applicants as an opportunity, not a problem. They would rush to propel into the higher ranks of American society a group of students who not only boast outstanding test scores and grades but also are immigrants or immigrants” children from low-or middle-income families that sacrificed in hopes of a better life for the next generation. Asian American students also bring a variety of cultures, languages, and religions to stir the campus melting pot. Colleges should counter anti-Asian bias through sensitivity training sessions and hiring more Asian American admissions deans, directors, and staff.

  In 1990, the federal Office for Civil Rights found that Asian American students needed higher test scores than whites to be admitted to Harvard. But it bought Harvard's argument that preferences for legacies and athletes, who were mostly white, accounted for the gap, rather than race discrimination. Scaling back those preferences would deprive elite universities of excuses for setting a higher bar for Asian Americans and compel them to confront the question of racism.

  By the same token, elite universities that proclaim themselves to be “need-blind” in domestic admissions should dig into their billion-dollar endowments to extend that same policy to foreign students. Limiting international recruits to the jet-setting children of royalty and business tycoons discriminates against talented lower-income applicants, induces cynicism abroad about American higher education, and gives domestic students a narrower understanding of the rest of the world.

  BY ELIMINATING preferences for legacies, development cases, faculty children, and athletes in patrician sports, elite colleges would open up a substantial number of seats—an estimated one-quarter of the freshman class. These slots should be filled on merit in the broadest meaning of the term. Candidates should be evaluated not on their parents” wealth but on their own achievement and potential. That doesn't mean that colleges should automatically accept applicants with the best test scores or grades, although those are important credentials. Admissions decisions inevitably are, and should be, subjective. Relying on the best judgment of their staff and faculty, colleges should consider anything germane to a candidate's cause, including essays, recommendations, and economic or social disadvantage. Thus Anthony Marx could fulfill his goal of socioeconomic diversity and revamp Amherst's economic profile by allocating more spaces to outstanding students from low-income families. Other colleges might supplant the beneficiaries of privilege with students who are especially gifted in music and art, or tenacious in overcoming illness, abuse, family disruptions, racial discrimination, or inferior high schools. Any of these approaches would be more attuned to American values of hard work, equal opportunity, and upward mobility and more likely to yield a capable leadership class than kowtowing to wealth.

  There remains the thorny question of affirmative action. Eliminating admissions breaks for legacies and other predominantly white groups would weaken the case for minority preference. Ideally, a fair, accessible college admissions system based on individual merit would make no exception for any entitled group, but would inspire students of all races— and their teachers—to redouble their educational efforts. Perhaps in twenty-five years, as Justice O'Connor foresaw in her Michigan opinion, affirmative action will no longer be needed.

  For now, though, only an admissions preference can compensate for the persistent achievement gap between white and Asian students, on one hand, and black and Hispanic students, on the other. By ensuring diversity, affirmative action both enhances intellectual and social interaction on campus and fosters minority representation in the upper realms of government, business, and the military, to which selective colleges provide a gateway.

  Affirmative action aside, substituting outstanding students for those who coasted in on their parents” wallets would improve the colleges” academic rigor and social cohesion. The Ivy Leagues could toss out “gut” courses patronized by legacies and development cases and make their humanities and social sciences curricula as demanding as Caltech's engineering program, yielding more-knowledgeable and better-prepared graduates. The meritocratic ethic would spill over into daily campus life; with everyone arriving through the same admissions process and fewer students from fancy neighborhoods and prep schools, social segregation would abate. To survive, eating clubs and other insular groups would become more open to students of all backgrounds.

  All American schoolchildren are taught that if they work hard enough, they can fulfill their ambitions, even become president. The elite colleges that unlock the door to success in our society contradict that promise, and sow anger and disillusion, by compromising standards to admit children of the rich, famous, and powerful. Ending the preferences of privilege would revitalize the social compact, replenish the ranks of leadership with deserving newcomers, and bear out de Tocqueville's description of America as a democracy with “an almost universal equality of social conditions.”

  EPILOGUE

  A week after this book was published in September 2006, Harvard made what it portrayed as a dramatic bid to attract low-income students. The nation's oldest university eliminated its early action program, which had allowed high school seniors who applied by November 1 to find out on December 15 whether they were being admitted, rejected, or deferred to the regular pool. Instead, all applicants for the 2008 freshman class would have to abide by the same January 1 deadline.

  “We think this will produce a fairer process,” interim president Derek Bok told the New York Times, “because the existing process has been shown to advantage those who are already advantaged.” Admissions dean William Fitzsimmons added, “There are lots of very talented students out there from poor and moderate-income backgrounds who have been discouraged by this whole hocus-pocus of early admissions.”

  In fact, Harvard's explanation contained a bit of hocus-pocus. The main advantage for wealthy students does not rest on when they apply, but on the preferences they receive whenever they apply. While early candidates often do get into Ivy League schools with borderline credentials, that's largely because a disproportionate number of them qualify for preference as alumni and faculty children, athletes, and development cases. Except at a handful of schools—the University of Pennsylvania, for example, offers legacy preference only to alumni children who apply early—these candidates have an edge whether they're admitted at Thanksgiving or Easter.