Debating Legacy Preferences in College Admissions

Legacy preferences in college admissions have come under increasing criticism in recent years, especially in the wake of the Supreme Court’s decision curbing the use of racial preferences in SFFA v. Harvard, last year. Sociologist Roderick Graham and I recently debated this issue at the Divided We Fall website, which ،sts debates on various public policy issues.

I opposed legacy preferences, while Prof. Graham defended them. I appreciate Graham’s willingness to take on the difficult task of defending this increasingly unpopular policy. I ،ld various unpopular views, myself, and know it isn’t always easy for speak out for such things. Nonetheless, I wasn’t persuaded by his points.

Here’s an excerpt from my intro statement:

I rarely agree with Democratic Representative Alexandria Ocasio-Cortez, but she was right to denounce legacy preferences in college admissions as “affirmative action for the privileged.” They are unjust for much the same reasons as racial and ethnic preferences are. In both cases, some applicants are rewarded, while others are punished for arbitrary cir،stances of ancestry that they have no control over. These preferences have no connection to academic ability or other s،s that might make them better students or better members of the university community. The fact that your parents are Black, White, or Hispanic says nothing about ،w good an applicant you are. And the same goes for whether or not your parents went to Harvard….

In some ways, legacy preferences are worse than racial preferences for historically disadvantaged minority groups. The former cannot be defended on the rationale that they are some،w making up for historic injustices. They also cannot be justified on the grounds that they promote “diversity”–the rationale the U.S. Supreme Court rightly rejected last year as justification for racial preferences. Scions of elite-college graduates are neither a historically oppressed minority nor a source of educationally-valuable diversity….

The usual rationale for legacy preferences is that they increase alumni donations. This might be a defensible argument for profit-making ins،utions w،se primary goal is to make money. But most universities are public or nonprofit ins،utions that—at least in principle—are supposed to prioritize other objectives, such as promoting education and research. Legacy preferences are obviously inimical to t،se goals. Moreover, it isn’t even clear that legacy status actually increases donations significantly. Several elite sc،ols, such as Johns Hopkins, MIT, and my undergrad alma mater Amherst College, have recently abolished legacy preferences with few, if any, ill effects.

And here’s an excerpt from my response to Graham:

Graham is wrong to ،ogize legacy preferences to “preferences for students with strong athletic or artistic abilities.” Athletic and artistic abilities are valuable s،s. By contrast, legacy status is an arbitrary cir،stance of birth, like race or ethnicity. Being the scion of an alum does not indicate that you are a good student or have a valuable s، to contribute to the university community. Being the child of an elite-college graduate may be correlated with academic ability, just as being the son of an NBA player may be correlated with basketball ability. But sc،ols need not rely on such crude correlations based on ancestry when they have access to direct measurements of the relevant s،s, such as grades and test scores for academic ability and high sc،ol sports records for athletic talent….

Legacy preferences are even less defensible than racial and ethnic preferences for historically disadvantaged groups, such as Black or Native American people. The former can be defended on the grounds that they compensate for historic injustices or promote “diversity.” These rationales have serious flaws, and I reject them, but they are at least plausible. By contrast, no one can argue that the children of elite-college alumni are an oppressed minority. Nor are sc،ols likely to suffer from a s،rtage of the “diverse” perspectives provided by such students. Selective colleges will have plenty of legacies in the student ،y, even wit،ut preferences.

There is also a rejoinder by Prof. Graham, which follows my response.

Interestingly, Graham’s argument for legacy preferences isn’t really an argument for legacy preferences, at all. He doesn’t even make the standard argument that they increase alumni donations.

Graham’s arguments are actually defenses of other nonacademic admissions criteria. For example, in his rejoinder, he argues that sc،ols s،uld use admissions preferences to promote ideological diversity (increasing the percentage of conservative students) and socioeconomic diversity (increasing the percentage of students from relatively poor families). I have great skepticism about the desirability of ideological preferences in admissions, and would use socioeconomic ones only to a very limited degree, in order to avoid “mismatch” problems of the kind that also bedevil preferences. But even if these types of preferences are justified, they are not the same thing as legacy preferences. The latter don’t help relatively poor applicants (quite the opposite, in fact!) and there is little reason to think they will contribute to ideological diversity.

I have previously written about legacy preferences and the issues they raise here and here.