The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent

Affirmative Action

Rights and Wrongs of Replacing Traditional Affirmative Action with Preferences for Descendants of Slaves

This approach would avoid many of the flaws of traditional racial preferences. But it has some downsides of its own.


The Supreme Court recently decided to hear two cases challenging racial preferences in college admissions. While we cannot know for sure, it seems highly likely that the conservative majority on the Supreme Court will take this opportunity to either strike down or severely curtail the dubious "diversity" rationale that it previously used to justify such preferences.

This possibility has stimulated renewed interest in possible alternatives to the use of racial preferences. One possible option is to replace them with preferences for descendants of American slaves. After I published an article in the Boston Globe criticizing the diversity rationale, a reporter for a Boston NPR station contacted me to ask about this alternative. Co-blogger David Bernstein also highlighted this option in a recent blog post, and in his excellent new book Classified: The Untold Story of Racial Classification in America. Georgetown University has already adopted a small-scale version of this policy, by granting admissions preferences to the descendants of slaves owned and sold by the school in the 19th century. This post expands on the answer I gave the reporter.

In my view, replacing race-based affirmative action with preferences for American descendants of slaves ("ADOS," as David Bernstein refers to them) would be a significant improvement over status quo policies. But this approach would also have crucial downsides.

Depending on how it is structured, this policy might well avoid many of the flaws of traditional race-based affirmative action. Most obviously, it would not require the use of racial classifications. This avoids the well-known moral pitfalls of race discrimination.

It would also  likely avoid the legal problems, as well. Unlike race discrimination, preferences based on ADOS status are not presumptively unconstitutional. They probably would only be subject to minimal "rational basis" scrutiny under current Supreme Court precedent, which they would likely easily pass. Preferences for descendants of slaves also would not violate Title VI of the Civil Rights Act of 1964, which bars discrimination "on the ground of race, color, or national origin" in any program receiving federal funds.

In addition, limiting admissions preferences to descendants of slaves would refocus affirmative action on its original purpose of compensating groups that have been victims of massive historic injustices, a much more compelling justification than the badly flawed "diversity" theory. While slavery and the segregation that followed it are far from the only racial injustices in American history, they are by far the biggest. And similar preferences could potentially be extended to descendants of victims of other massive historical wrongs, such as the forcible displacement of many Native Americans from their lands.

But, despite its attractions, giving preferences to descendants of slaves raises several difficult problems of its own.

First, how would we verify whether a given applicant qualifies as a member of the relevant group? If we rely on self-identification, there will be obvious incentives for fraud or deception. But having university officials investigate applicants' ancestry or demand verification thereof also creates problems. In many cases, there might not be reliable records available, going all the way back to the days of slavery. Conducting investigations into the ancestry of applicants is also likely to be costly and intrusive.

Second, there is the closely related issue of how to classify the large number of people of mixed ancestry. Millions of Americans who look "white" and are perceived as such by society have slaves or former slaves in their family trees. If everyone with such ancestry is allowed to qualify for ADOS preferences, it would mean extending it to a large number of applicants whose claim to be victims of racial injustice is, at best, highly tenuous.

On the other hand, if only sufficiently "black" descendants of slaves qualify as true descendants, then we are right back to using racial classifications. Doing so would bring back the very problem the ADOS strategy is intended to solve. In addition, any attempt to determine who is "black enough" to qualify as a "real" descendant of slaves is all too likely to degenerate into subjectivity and bias. The history of such attempts at racial classification is, to put it mildly, not an encouraging one.

Finally, there are also difficult moral and philosophical issues with assuming that anyone who is a descendant of slaves (even if they are "authentically" black) is automatically a victim of injustice himself or herself. Consider an example from my own family history. I am a Russian Jew and descendant of people who suffered from pogroms, the Holocaust, and a variety of other czarist, Nazi, and Soviet injustices. Does that make me a victim of anti-Semitism myself, thereby worthy of compensation of some kind (perhaps from the Russian or German governments)? Maybe. But the issue is highly contestable.

If not for the many injustices perpetrated by various Russian and German regimes, I almost certainly would not have even been born. My grandmother lost nearly all of her family, including her first husband, in World War II; many of them perished in the brutal 900 day siege of Leningrad. If not for these horrific events—the responsibility of Hitler's regime, with an assist from Joseph Stalin and the Nazi-Soviet Pact—she likely would never have married my grandfather (whom she met after the war), nor given birth to my father. Indeed, even a slightly different chain of events from that which actually happened would have prevented me from ever existing. In a perverse, but real, sense I am at least as much a beneficiary of Hitler and Stalin's injustices as I am a victim.

What is true for me is also true for nearly everyone alive today. If not for our horrible history of injustice, hardly any of us would be here—whether we are descendants of slaves, descendants of slaveowners, or descendants of neither.

This consideration is irrelevant to policies that seek to compensate people who are direct victims of unjust policies, as opposed to descendants of such victims. For example, it does not weaken the rationale for compensation payments to actual Holocaust survivors, or the belated and inadequate compensation paid to Japanese-Americans forcibly detained in camps during World War II. But it is a serious issue when we are considering compensation for the descendants of victims of historic wrongs, on the ground that they have been negatively impacted by the long-term effects of those injustices.

Even aside from the possibility that, but for the history of injustice, they wouldn't exist in the first place, it is often difficult to determine where a given individual would be if not for the wrongs suffered by their ancestors. Some descendants of slaves—like some descendants of Holocaust survivors, Japanese-American internees, and other victims of great historic wrongs—are nonetheless affluent and otherwise well-off today. I myself am a relatively wealthy law professor.

Perhaps all these people would be richer and happier still, if not for the awful history endured by their ancestors. But it's very difficult to tell one way or the other.

This last problem could perhaps be addressed by limiting compensatory preferences to relatively poor and disadvantaged members of the relevant group. But it will not be easy to figure out where to draw the line, and not clear that any institution can be trusted to do so objectively.

In sum, replacing traditional affirmative action with preferences limited to descendants of slaves has some important advantages. But any such program would also have to grapple with grave difficulties of its own.

Despite these reservations, I would not legally forbid private universities from trying this. Public ones are, I think, a closer call, though unlike racial preferences, these programs would not be presumptively unconstitutional. Government institutions, I believe, have stronger nondiscrimination obligations than private ones, and thus should have stronger presumptions against any kind of discrimination based on ancestry, which too often cause grave harm, even if not based on race. Indeed, in most situations, discrimination based on parentage and place of birth is unjust for many of the same reasons as racial discrimination is. And even if such policies are legal and not inherently unjust, they should only be enacted if we can come up with at least reasonably good solutions for the problems described above.

It is perhaps worth noting that my concerns about preferences based on ancestry also apply to the "legacy preferences" still used by many selective universities. Unlike ADOS preferences, they cannot even be defended on the grounds that they help remedy historic wrongs. Schools should, therefore, abolish them, as my own undergraduate alma mater, Amherst College recently did.