Legacy Preferences Make a Lot of Sense, Up to a Point

A great many colleges give preference in admission to children of alumni. (This includes public colleges because few colleges any more rely entirely on public funding.) It is widely believed, with some evidence, that alumni give more generously to their alma mater if their children are admitted to, and especially if, having been admitted, they attend, the college that they (the alumni) attended. This is the passionate belief of admissions officer, and if it were false it would have been abandoned, because it attracts a good deal of criticism.

The "legacy" preference is not a mere tie breaker; as in the case of preferences for blacks, Hispanics, and athletes, it amounts in effect to pretending that the applicant had a substantially higher score on the Scholastic Aptitude Test than he or she actually had.

The preferences for blacks and Hispanics are defended on the ground (possibly spurious in most cases, reflecting pressure from government rather than educational theory) that they are better than their high-school records or SAT scores suggest, because of a disadvantaged upbringing; in fact the colleges, at least the elite ones, try to admit the non disadvantaged members of these minorities because they are more likely to graduate. But the athletic and legacy preferences have strictly financial motivations. Alumni donations amount to some 28 percent of total college revenues, and both athletic prowess and admission of alumni offspring are considered important to alumni donations.

The practice of favoring alumni children seems to be particularly strong in elite colleges, though statistics are hard to come by because most colleges will not reveal the admission rate of alumni children relative to the admission rate for applicants who do not belong to preferred minorities, are not good athletes, and are not alumni children. Not that such statistics would be meaningful by themselves, at least in the case of alumni children admitted to elite schools. By definition one of the parents attended such a school, namely the one their child has applied to; and given assortative matching in marriage, there is a good chance that the other parent went to the same, or another, elite school, as well. Hence their children are probably of above-average quality (unless their parents were second- or third-generation legacy admits), in which event they would be admitted at a higher rate than non-legacy applicants even without legacy preference. What one needs to know is how great the preference for alumni children is and therefore how many are admitted who wouldn't be without the preference. We don't know how great it is, but it must be significant; otherwise colleges wouldn't bother to give the preference—most alumni children would be admitted without it.

Admitting an applicant to an elite college because one or both of his parents went there is offensive in a nation whose constitution forbids the granting of titles of nobility. It confers an arbitrary advantage, based on ancestry rather than on merit or promise. The advantage is not primarily that the preferentially admitted student receives a better education; he may not, because the fact that he would not have been admitted without a preference suggests that he may not be a motivated or talented student; and you cannot make silk purses out of sow's ears. But his degree from an elite college will open doors; prospective employers, even if they discover that an applicant is the child of an alumnus of the college he attended, will not know whether he was one of the alumni children who wouldn't have been admitted had it not been for the connection. The student will also derive social benefits (prestige, maybe some extra refinement) and valuable contacts from hobnobbing with the other students; elite colleges have a disproportionate number of elite students.

The same argument is made against allowing gifts and bequests by wealthy parents (or, more realistically, against light or no taxation of such transfers): that they confer an arbitrary advantage. It is further argued that they blunt the children's ambition. But there are counterarguments, which cumulatively are compelling: heavy taxation of gifts and bequests increases consumption at the expense of investment; reduces parents' utility and their incentive to work hard; weakens family bonds; and reduces constructive risk taking by young people by reducing the number of young people who have financial security. Abandoning legacy admissions would have much smaller effects. The major one would be a reduction in alumni giving, and no one knows how much of a reduction there would be, or the extent to which it could be offset by more vigorous solicitation of alumni donations. But the benefits from abolishing legacy admissions would also be slight, if I am right that one of the principal benefits to the legatees is social—the benefits from rubbing shoulders with elite students. For the students whom the alumni children displace at the elite schools still go to college. They just will drop down a notch in the college pecking order; and their presence in colleges at that level will be a boon to the other students in those colleges.

Furthermore, if one asks who exactly will replace the alumni children if legacy admissions disappear, the answer—better qualified students—must itself be qualified by recognition that parents invest money to improve their children's college admission prospects, and their investment may enable a somewhat less able applicant to be admitted in preference to an abler one. The parents may have financed a highly expensive elementary and high school education for their children at an elite private school, hired tutors for them, and financed extracurricular activites of the kind that impress admissions officers. The greater the inequality of income in the society—and it is very great in our society at present—the more that wealthy people will be able to gain access to elite colleges for their children, and by doing so underwrite the children's future economic success and social prestige.

This point suggests that, as long as colleges are private (including de facto private, which is increasingly the case with public colleges, many of which like the University of Michigan, receive only a small fraction of their money from the state), they will follow a business model, and a business model of education is inherently tilted in favor of applicants who may not be the most promising students, whether they are members of preferred minorities, athletes, children of very wealthy people even if neither parent is an alumnus (such children are favored by many colleges), children with an impressive resume attributable to a significant effect in their parents' investment in them, or alumni children. Although most colleges—and all elite colleges—are nonprofit institutions, the only difference between most nonprofit institutions and for-profit institutions is that the former cannot distribute surplus revenue (profit) to the people who supply them with capital: there are no shareholders or dividends. They are no less competitive than private institutions. One can imagine a system in which all education, from kindergarten to graduate and professional school, would be financed by the government, but few would argue that such a system would be an improvement over what we have, even though what we have has many unattractive features.

The ultimate question from an economic standpoint is whether legacy admissions create a harmful externality, which is to say impose a cost on society that the parties to the practice that causes the externality do not take into account. The college and its legacy admits like the practice of legacy admissions and do not want to change it; should they be forced to because of the harm the practice causes to other members of society?

I think not. The harms are of two kinds. First, legacy admissions undermine the matching of the best students with the best colleges. That kind of "assortative mating" is probably most efficient from the standpoint of maximizing the gains from education; and education creates benefits for society as a whole. With legacy admissions, there are fewer places for the best students in the best colleges; some of those best are relegated to the next tier of colleges. On the other hand, students learn from other students as well as from teachers, and those "best" who are knocked down a tier thus benefit their fellow students in the second-tier colleges to which they are admitted. The net loss to educational efficiency caused by legacy may well be slight—it may even be negative if abolition of legacies would result in a significant reduction in alumni donations and therefore in college resources.

Second, legacy admissions increase inequality, or at least may increase it (this qualification is explained below). Inequality, beyond a point we may have reached, is harmful to society as a whole; the enormous federal deficit is in part a result of efforts by the Bush Administration as well as the Obama Administration to offset wage stagnation and rsulting increased inequality by heavy borrowing to fund entitlement programs. But the increase in inequality brought about by legacy admissions is probably infinitesimal. Moreover, legacy admissions may actually reduce inequality, because the abler person pushed down to a lower-tier school may have lower lifetime earnings as a result, and if as is likely he is upper middle class, the reduction in his lifetime earnings will reduce inequality.

I continue to find the practice distasteful. But I don't think public intervention (which might take the form of federal and state regulations conditioning financial assistance to colleges on the abandonment of the practice) is warranted.