A Tepid Defense of the "Diversity" Rationale for Affirmative Action
by Michael Dorf
My most recent Verdict column (now also available via Newsweek) discusses the recent announcement that the DOJ is committing resources to investigating and potentially suing American colleges and universities over their affirmative action programs. I describe the potential peril that the announcement--if the DOJ follows through--holds for colleges and universities that push the edge of the envelope of what the SCOTUS precedent currently permits and even for those that hew strictly to the line between permissible "plus-factor" affirmative action and impermissible de facto quotas. For the latter, the greater danger would be another SCOTUS appointment for President Trump (or, in the event of his removal, any other Republican president).
Here I want to discuss a related issue. Defenders of affirmative action--and even some of its critics--frequently criticize Justice Lewis Powell's controlling opinion in the Bakke case on the ground that it rejected the most compelling justification for race-based affirmative action while validating a relatively weak justification. I'll offer a hesitant defense of Powell's position (which subsequently was adopted by a SCOTUS majority).
The UC-Davis Medical School affirmative action program in Bakke used a separate admissions pool for "disadvantaged" students that was, in essence, race-based. Powell's opinion in Bakke ruled that the program was invalid but rejected the position of four of his colleagues that race could never be a factor in higher education admissions. He gave Harvard's undergraduate admissions program--which used race as a "plus factor" in a "holistic" evaluation of each applicant--as an example of a permissible program.
En route to distinguishing quotas from plus factors, Powell also considered the various purported compelling interests that the medical school offered in defense of its program. First, while acknowledging that a state university has an interest in remedying its own past acts of racial discrimination, he rejected the argument that past (and ongoing) racial discrimination by other social actors justified the use of race in admissions. Second, the medical school argued that educating minority doctors would increase medical services available to under-served minority populations. Powell thought this argument too attenuated.
In doctrinal terms, Powell's Bakke opinion thus stands either for the proposition that interests in remedying societal discrimination and educating people to serve under-served populations are not compelling interests or that, if they are, race-based affirmative action is not narrowly tailored to advance these interests. However understood, these justifications are off the table. Instead, since Bakke, background diversity (as a means of achieving viewpoint diversity) has been pretty much the only game in town.
As noted above, Powell's opinion has come in for its share of criticism for its rejection of the remedial rationale. The main criticism (mostly from liberals but also occasionally from conservatives) goes like this: Given our history of slavery, segregation, and discrimination against African Americans--and the way in which that history pervades every institution of American life--simple justice demands (or at least permits) that race be considered in order that African American applicants whose educational preparation has left them at a disadvantage relative to most other applicants be given opportunities for higher education. A remedial program of affirmative action solely for African American applicants would be easier to justify than the diversity programs that carve up educational opportunities among multiple groups.
I think there is undoubtedly something to the foregoing complaint. But I think that it tends to overlook the difficulties that are also inherent in a remedial approach and to exaggerate its differences from a diversity approach.
(1) Race-based slavery is one of two original sins of American history. The other is genocide against Native Americans. At the very least, then, a remedial program of affirmative action would need to include Native Peoples. But there are other instances of severe--if not enslaving or genocidal--discrimination that have had substantial impacts. Why should the threshold be enslavement or genocide?
(2) Powell thought that race was too imprecise an instrument for remedying "societal" discrimination. One can disagree (as I do) with the suggestion that colleges and universities bear no responsibility for what happens in the broader society, but still be troubled by the same basic analytic problem that troubled Powell: figuring out who is entitled to the remedy. Consider one possible rule: anyone whose ancestors were enslaved qualifies for affirmative action. This leads to at least some curious results. E.g., Barack Obama would not have been eligible for affirmative action but his daughters (whose ancestors on their mother's side were enslaved) would be--even though they are more privileged than he was.
To be sure, there's a ready response to that objection: Race in America is and has always been a cultural category. Regardless of whether your ancestors were brought here in slave ships three hundred years ago or immigrated here twenty years ago, you will be seen the same way.
But that answer itself has problems. On average, African and Caribbean immigrants to the U.S. have substantially higher socioeconomic performance than African Americans whose ancestors were enslaved here. Thus, programs of strictly race-based affirmative action will not benefit those persons of African descent who are--from a U.S. societal perspective--most deserving.
We can put that consideration aside and say that current discrimination--which does not sharply distinguish between recent immigrants and their children versus African Americans whose families have lived here for centuries--provides a remedial justification for affirmative action, but then we need to ask who faces the most, or suffers the most from, current discrimination. A program of affirmative action responsive to that question will be substantially more complex than the simple one envisioned by the standard criticism of Powell's rejection of the remedial rationale. It will apply to a fair number of groups because, unfortunately, a fair number of groups face current discrimination. This returns us to point (1).
(3) Meanwhile, critics of affirmative action sometimes argue that by the time students apply to colleges and universities it is too late to provide an effective remedy for prior discrimination. A serious effort to increase the representation of under-represented minorities in higher education and American public life more broadly would target pre-school and elementary school. When voiced by conservatives, this complaint is frequently disingenuous, because they also oppose anything like the sort of early investment in human capital that would be required to carry out this sort of program. And, defenders of affirmative action say with some justification, this is not an either/or proposition: We can make the large investment in earlier interventions that will some day render affirmative action in college and university admissions unnecessary, while practicing the latter in the meantime.
That all makes sense to me, but I would note that the background-and-viewpoint-diversity rationale for affirmative action makes more sense of the overall system we actually have. Given that we are not making the kinds of investments in the wellbeing of minority citizens that would be needed to render affirmative action in higher education admissions unnecessary, one should ask what exactly such admissions programs are doing.
One answer might be: They're doing what they can. Many college and university administrators (and faculty) would, if it were up to them (i.e., us), make those larger investments at an earlier stage, but of course, it's not up to them (us). All that they (we) can do is to make some corrections at the end of the pipe. I think this hypothesis explains a great deal, as evidenced by the fact that many of the people who criticize Powell's Bakke opinion's rejection of the remedial justification are academics. To the extent that this is what is going on, colleges and universities are shoehorning remedial programs into a diversity rationale.
But the foregoing explanation--while causally accurate--does not provide a very good justification for a system in which society makes insufficient investments in early education (and wellbeing more generally) of minority children but then tries to make up for a little bit of that injustice by providing a small number of minority college and university applicants with a benefit in the admissions game. A much better justification would focus on a rationale that is in some way distinctive to colleges and universities. And that is exactly what Powell's endorsement of diversity actually does.
My most recent Verdict column (now also available via Newsweek) discusses the recent announcement that the DOJ is committing resources to investigating and potentially suing American colleges and universities over their affirmative action programs. I describe the potential peril that the announcement--if the DOJ follows through--holds for colleges and universities that push the edge of the envelope of what the SCOTUS precedent currently permits and even for those that hew strictly to the line between permissible "plus-factor" affirmative action and impermissible de facto quotas. For the latter, the greater danger would be another SCOTUS appointment for President Trump (or, in the event of his removal, any other Republican president).
Here I want to discuss a related issue. Defenders of affirmative action--and even some of its critics--frequently criticize Justice Lewis Powell's controlling opinion in the Bakke case on the ground that it rejected the most compelling justification for race-based affirmative action while validating a relatively weak justification. I'll offer a hesitant defense of Powell's position (which subsequently was adopted by a SCOTUS majority).
The UC-Davis Medical School affirmative action program in Bakke used a separate admissions pool for "disadvantaged" students that was, in essence, race-based. Powell's opinion in Bakke ruled that the program was invalid but rejected the position of four of his colleagues that race could never be a factor in higher education admissions. He gave Harvard's undergraduate admissions program--which used race as a "plus factor" in a "holistic" evaluation of each applicant--as an example of a permissible program.
En route to distinguishing quotas from plus factors, Powell also considered the various purported compelling interests that the medical school offered in defense of its program. First, while acknowledging that a state university has an interest in remedying its own past acts of racial discrimination, he rejected the argument that past (and ongoing) racial discrimination by other social actors justified the use of race in admissions. Second, the medical school argued that educating minority doctors would increase medical services available to under-served minority populations. Powell thought this argument too attenuated.
In doctrinal terms, Powell's Bakke opinion thus stands either for the proposition that interests in remedying societal discrimination and educating people to serve under-served populations are not compelling interests or that, if they are, race-based affirmative action is not narrowly tailored to advance these interests. However understood, these justifications are off the table. Instead, since Bakke, background diversity (as a means of achieving viewpoint diversity) has been pretty much the only game in town.
As noted above, Powell's opinion has come in for its share of criticism for its rejection of the remedial rationale. The main criticism (mostly from liberals but also occasionally from conservatives) goes like this: Given our history of slavery, segregation, and discrimination against African Americans--and the way in which that history pervades every institution of American life--simple justice demands (or at least permits) that race be considered in order that African American applicants whose educational preparation has left them at a disadvantage relative to most other applicants be given opportunities for higher education. A remedial program of affirmative action solely for African American applicants would be easier to justify than the diversity programs that carve up educational opportunities among multiple groups.
I think there is undoubtedly something to the foregoing complaint. But I think that it tends to overlook the difficulties that are also inherent in a remedial approach and to exaggerate its differences from a diversity approach.
(1) Race-based slavery is one of two original sins of American history. The other is genocide against Native Americans. At the very least, then, a remedial program of affirmative action would need to include Native Peoples. But there are other instances of severe--if not enslaving or genocidal--discrimination that have had substantial impacts. Why should the threshold be enslavement or genocide?
(2) Powell thought that race was too imprecise an instrument for remedying "societal" discrimination. One can disagree (as I do) with the suggestion that colleges and universities bear no responsibility for what happens in the broader society, but still be troubled by the same basic analytic problem that troubled Powell: figuring out who is entitled to the remedy. Consider one possible rule: anyone whose ancestors were enslaved qualifies for affirmative action. This leads to at least some curious results. E.g., Barack Obama would not have been eligible for affirmative action but his daughters (whose ancestors on their mother's side were enslaved) would be--even though they are more privileged than he was.
To be sure, there's a ready response to that objection: Race in America is and has always been a cultural category. Regardless of whether your ancestors were brought here in slave ships three hundred years ago or immigrated here twenty years ago, you will be seen the same way.
But that answer itself has problems. On average, African and Caribbean immigrants to the U.S. have substantially higher socioeconomic performance than African Americans whose ancestors were enslaved here. Thus, programs of strictly race-based affirmative action will not benefit those persons of African descent who are--from a U.S. societal perspective--most deserving.
We can put that consideration aside and say that current discrimination--which does not sharply distinguish between recent immigrants and their children versus African Americans whose families have lived here for centuries--provides a remedial justification for affirmative action, but then we need to ask who faces the most, or suffers the most from, current discrimination. A program of affirmative action responsive to that question will be substantially more complex than the simple one envisioned by the standard criticism of Powell's rejection of the remedial rationale. It will apply to a fair number of groups because, unfortunately, a fair number of groups face current discrimination. This returns us to point (1).
(3) Meanwhile, critics of affirmative action sometimes argue that by the time students apply to colleges and universities it is too late to provide an effective remedy for prior discrimination. A serious effort to increase the representation of under-represented minorities in higher education and American public life more broadly would target pre-school and elementary school. When voiced by conservatives, this complaint is frequently disingenuous, because they also oppose anything like the sort of early investment in human capital that would be required to carry out this sort of program. And, defenders of affirmative action say with some justification, this is not an either/or proposition: We can make the large investment in earlier interventions that will some day render affirmative action in college and university admissions unnecessary, while practicing the latter in the meantime.
That all makes sense to me, but I would note that the background-and-viewpoint-diversity rationale for affirmative action makes more sense of the overall system we actually have. Given that we are not making the kinds of investments in the wellbeing of minority citizens that would be needed to render affirmative action in higher education admissions unnecessary, one should ask what exactly such admissions programs are doing.
One answer might be: They're doing what they can. Many college and university administrators (and faculty) would, if it were up to them (i.e., us), make those larger investments at an earlier stage, but of course, it's not up to them (us). All that they (we) can do is to make some corrections at the end of the pipe. I think this hypothesis explains a great deal, as evidenced by the fact that many of the people who criticize Powell's Bakke opinion's rejection of the remedial justification are academics. To the extent that this is what is going on, colleges and universities are shoehorning remedial programs into a diversity rationale.
But the foregoing explanation--while causally accurate--does not provide a very good justification for a system in which society makes insufficient investments in early education (and wellbeing more generally) of minority children but then tries to make up for a little bit of that injustice by providing a small number of minority college and university applicants with a benefit in the admissions game. A much better justification would focus on a rationale that is in some way distinctive to colleges and universities. And that is exactly what Powell's endorsement of diversity actually does.
* * *
My defense of the diversity rationale is tepid because I have not denied the force of the criticisms of the rationale. I have simply argued that the chief alternative is also problematic. Some of my more conservative readers may therefore conclude that I ought to style this blog post a criticism of affirmative action rather than a tepid defense. But that's not my point at all.
I agree with the complaint that we do not do nearly enough at the stage where it matters most to ensure that all children have opportunities to maximize their potential. I therefore favor programs that would eventually make affirmative action unnecessary. In the meantime, however, I think that a combination of rationales--remedial, diversity, role-model, community-serving, etc.--provides adequate justification for a variety of affirmative action programs in higher education. None of these rationales standing alone or even in combination is perfect, but then, one can say the same about the relation between rationales and implementation of nearly every public policy. And with the liberals in Bakke and subsequent cases, I have always regarded it as perverse to say that because of a historical injustice perpetrated against African Americans and members of other racial minority groups, the law should be less forgiving of efforts to provide greater opportunities for members of those groups than it is of nominally race-neutral programs that perpetuate group-based inequality--such as alumni preferences.
It seems to me passing strange that of all the policies pursued by our important institutions, one would single out the relatively modest efforts to increase representation of under-represented minorities as especially worthy of attack. But then, it also seems to me passing strange that Donald J. Trump and Jefferson Beauregard Sessions III should think they have the moral authority to pursue policies that purport to aim at racial justice. And of course it seems stranger still that they are respectively the president and the Attorney General.
________
Postscript: In response to the column, I received a number of emails critical of my characterization of Trump as a "real-estate-developer-turned-reality-TV-star who became president thanks to support from racists." The authors of these emails said they voted for Trump and were not racists. Yet I didn't say that Trump won the election thanks to support that came exclusively from racists. The Nation article to which I linked showed that Trump voters were, on average, more racist than other voters. Perhaps I ought to have said in the column that Trump won the election thanks to the support of racists, self-described non-racists who do not understand statistics, and others.
________
Postscript: In response to the column, I received a number of emails critical of my characterization of Trump as a "real-estate-developer-turned-reality-TV-star who became president thanks to support from racists." The authors of these emails said they voted for Trump and were not racists. Yet I didn't say that Trump won the election thanks to support that came exclusively from racists. The Nation article to which I linked showed that Trump voters were, on average, more racist than other voters. Perhaps I ought to have said in the column that Trump won the election thanks to the support of racists, self-described non-racists who do not understand statistics, and others.