The Libertarian Case for Affirmative Action

The Supreme Court yesterday decided to take up a new affirmative action case from the University of Michigan, along with the case from the University of Texas they are already considering. I don’t know what the Supreme Court will decide, or what it ought to decide on the basis of constitutional text and judicial precedent (although the appeals court’s logic in upholding Michigan’s program seems wildly tortured to me), but libertarians do not necessarily oppose state-sponsored affirmative action. Yes, that’s right: even state-sponsored racial discrimination can be justified for libertarians.

Now, affirmative action is generally not a big issue for libertarians. I’ve never heard a libertarian spontaneously bring up affirmative action as an important area of government overreach. It certainly doesn’t exercise libertarians the way the drug war, perverse economic regulations, protectionism, failed foreign wars, and even Bloomberg’s soda ban do. Nevertheless, my impression is that most libertarians oppose all state-sponsored discrimination on the basis of ascriptive criteria (race, gender, religion, etc.), except where necessarily related to the legitimate task or purpose the government is undertaking.

But deontological libertarians also believe in restorative or restitutive justice. And it cannot be denied that governments in the United States have historically harshly discriminated against African-Americans. For libertarians, those governments are morally required to compensate the victims of their injustice. For that reason, libertarians support financial “reparations” for victims of segregation.

But what about broader programs like those in university admissions that are intended to benefit all blacks? The last Jim Crow laws and practices were forcibly stamped out in the 1960s, almost 50 years ago. Does that mean that blacks born after the days of legally mandated segregation deserve no compensation? Not at all. Descendants of those discriminated against still carry some of the burdens of that discrimination. As generations pass, identifying the present-day victims of past injustice becomes more difficult and, eventually, impossible. Reparations for slavery would be difficult to implement justly; reparations for dispossession of Native Americans even more so. But within one or two generations, we can still identify victims and come up with plausible measures of the compensation they deserve. By analogy, I can’t imagine that any libertarian would say that the family of someone murdered by police don’t deserve any compensation, even though they were not the “first-order” victims of injustice. They still suffered injustice.

But don’t racial preference programs benefit all members of a racial minority group, whether or not they are descended from victims of gross, systematic, state-sponsored injustice? For instance, an immigrant from Africa could benefit from Michigan’s affirmative-action admissions policy. Here there are two points to make. First, courts have ruled out compensatory justice as a rationale for affirmative action in university admissions. So to satisfy the moral principle of restorative justice, universities have to come up with a different stated rationale: promoting “diversity” on campus. Racial preferences, rather than preferences for victims of gross injustice, are a “second-best” solution to this legal problem. Second, procedural justice might also justify affirmative action.

Suppose you believe that private, social discrimination against people who are visibly black remains a significant problem in American society, as seems likely. In their education and home environments, blacks might on average experience more significant challenges than whites and others. Adopting a formally race-blind admissions policy would fail to take these factors into account and would not actually satisfy the procedural justice principle of racial neutrality. Implementing racial neutrality in university admissions might require taking race into account, in the absence of better metrics of the amount of discrimination a student has faced.

Isn’t this procedural justice principle of racial neutrality an “end-state” or “patterned” principle of justice, which libertarians oppose? Not at all. The principle does not aim at any particular pattern of racial diversity in admissions. Rather, it aims at maintaining a fair procedure for deciding admissions. Achieving fairness toward people of all racial backgrounds might require violating a purely formal principle of racial blindness. To put the point another way, racial blindness can mean systematic racial bias.

But doesn’t affirmative action in university admissions actually hurt the intended beneficiaries, by admitting students who are not capable of performing the work required of them? That may be; it is an empirical question. Still, it is an odd argument for anti-paternalist libertarians to make: “We should provide people with fewer opportunities, because by giving them opportunities we are duping them into failure.” Libertarians don’t generally accept that kind of logic. If black students saw affirmative action as hurting them, they wouldn’t want it. Ultimately, it shouldn’t be up to mostly white admissions officers to decide for black applicants what is good for them, should it?

Again, all of this is aimed at moral theory. Can state-sponsored affirmative action be justified in principle? Yes. Is it constitutional? Is it effective at achieving its purposes? Those are separate questions I am not trying to answer here.

22 thoughts on “The Libertarian Case for Affirmative Action

  1. “blacks might on average experience more significant challenges than whites and others. Adopting a formally race-blind admissions policy would fail to take these factors into account and would not actually satisfy the procedural justice principle of racial neutrality. Implementing racial neutrality in university admissions might require taking race into account, in the absence of better metrics of the amount of discrimination a student has faced….it aims at maintaining a fair procedure for deciding admissions. Achieving fairness toward people of all racial backgrounds might require violating a purely formal principle of racial blindness. To put the point another way, racial blindness can mean systematic racial bias.”

    The problem with this sort of approach to the argument is that it ends up smuggling compensatory justice back in, and at the same time causes the very injustice it’s meant to mitigat. On this model, race is elevated above socio-economic class, so it perversely privileges people who have had fewer significant challenges over those with more, as when, e.g., the kids of a black doctor or lawyer are regarded as more challenged than the kids of poor white kids from marginalized ethinc groups. At my institution, e.g., we have many kids whose families are of Portuguese, Italian, or Irish descent, often second-generation Americans and the first in their families to finish HS let alone go to college. They grew up in poverty or to low-income working parents and attended K-12 schools which may have been sub-optimal. But on your model, they’re “white” and therefore privileged, whereas Sasha and Malia, who are black, have “faced signiifcant challenges.” You’ll reply I’m cherry-picking an easy example, but this gets to why “race” is too simplistic a measure for the goal you’re shooting for.

    1. That’s a fair point, but it would imply that other criteria, such as family income, should be used in admissions decisions in addition to race, not necessarily that race should not be considered. There are also all sorts of other considerations against affirmative action that I did not consider here: stoking up white resentment, establishing an unpleasant precedent for gov’t racial discrimination, etc. I just want to claim that there is a libertarian rationale for A.A., not necessarily that it is, in the final analysis, compelling.

      1. If an institution wants to diversify racially, that’s its prerogative, but I’m not sure it’s connected to libertarianiam per se.

      2. I agree, but some people would claim that libertarianism implies that state-sponsored AA is always morally impermissible.

  2. You ask: Can state-sponsored affirmative action be justified in principle? The answer is yes. You also state that deontological liberatarians believe in restorative justice, maybe or maybe not. But you have made your case wholly on the basis of positive rights theory and group rights.

    I don’t doubt your bona fides as a libertarian, or that you have made a case for state sponsored affirmative action but you have not made a libertarian case for affirmative action, because positive rights and group rights are decidedly un-libertarian approaches.

    Libertarians might agree that restitution can and should be provided for individuals after a hearing of the facts in an impartial tribunal. But few would agree that restitution should be provided, via legislation and state action, for whole groups, at the expense of other whole groups on the basis that, in the past, a similar group suffered a loss from another similar group. It is not just a matter of scale; those are categorically different cases.

    1. Why do you need a tribunal to decide individual cases? Sounds expensive. Lots of taxpayer money. Much simpler to make reparations via statute. Of course, there’s a point at which justice has been done, and no more reparations are due. Affirmative action isn’t justified forever.

      1. I am not certain that there is such a thing as group justice. No two members of any group will have suffered the same or deserve the same recompense. Justice if there is any, can be served fairly, only on a case by case basis. I think justice requires that the sufferer be able to identify the wrongdoer. One cannot justly do this at a group level. You are right, there is a point at which justice is done, or I would add; it will never be done. That’s why there are statutes of limitations in contract, tort and criminal law.

      2. Right, but consider the case of the Tuskegee syphilis experiments in the ’40s. The unwilling victims of those experiments were recently compensated — rightly so, I think, even though taxpayers had to foot the bill. There’s no other choice. And if taxpayers have to pay for government injustice, perhaps they will make more of an effort to prevent those injustices. Segregation seems analogous.

      3. The Tuskegee case was a smaller, readily identifiable group of victims with a readily identifiable perpetrator. That is categorically different.

        I don’t argue the morality of Affirmative Action. I only suggest, that in the spectrum which is libertarianism, and even within the subset of deontological libertarianism; when you speak in terms of positive rights and group rights, it is possible, you may have wandered, philosophically, off the spectrum which is libertarianism.

      4. The argument I laid out in this post has nothing to do with positive rights or group rights.

  3. Yes. But doesn’t this necessarily violate the rights of the white students?

    By choosing a black student over an identically qualified white student for purposes of restorative justice you are implicitly blaming punishing the white student for actions he had nothing to do with. This would then create a need for more restorative justice action on the part of the school. And wouldnt this newly affected student be essentially suffering an identical form of racially biased discrimination?

    Ultimately, this policy would create more moral justice issues than it solves.

    1. If black applicants in general have likely suffered from relative disadvantages due to ongoing social discrimination, then necessarily nonblack applicants in general have enjoyed relative advantages. Discriminating against nonblack applicants is not invidious or unfair if it is solely intended to “adjust” the qualifications of applicants for those unearned advantages and disadvantages. The point here is not to blame or to punish particular people, but to come up with a fair process for admissions.

      1. “black applicants in general” But that’s my objection: what is true of black applicants in general is not true of Sasha and Malia, ands what’s true of nonblack applicants in general is not true of my 1st-gen Portuguese-American students. If the institiution wants to take account of relative disadvantages, fine, but if you do it based on _just_ race (or just gender), then it’s failing to treat people as individuals.

      2. I agree. But the other implication of that line of argument is that selectively banning racial preferences while allowing legacy preferences, which account for a much larger share of admissions, effectively privileges advantaged whites over blacks, recent immigrants, and first-generation college students. And it’s undeniable that in this country, race and class are correlated. It’s an imperfect correlation, but admissions officers have to use imperfect statistical proxies because they don’t know every applicant’s life story.

  4. “Ultimately, it shouldn’t be up to mostly white admissions officers to decide for black applicants what is good for them, should it?”

    I’ve used similar arguments myself lots of times about lots of paternalistic policies. But the rejoinder is that they are not just black applicants, they are black applicants who are 17 & 18-year-old kids. Their decision making abilities (including both knowledge of the world and knowledge of themselves) is limited–which is one reason they are going to college in the first place!

    The admission officials are giving these kids not only an opportunity but also a signal: “we think you can make it here.” Suppose the letter of admission said this: “you have the opportunity to attend our college, but you should know that your academic qualifications alone would not qualify you for admission; therefore, you should think carefully and talk with adult advisers about whether it is wise to come compete against students who are, according to objective indicators, more intelligent and more academically prepared than you are.” If that were the signal colleges were giving, I’d be more supportive of the anti-paternalistic theme here. A useful reform might be that colleges had to give students a clear sense of where their grades and test scores put them in terms of the overall distribution of students who go to the college. That would help them make more informed choices.

    PS: This is a nicely done essay, overall. I assigned it to my students in Theories of Public Policy. More hits for you!

    1. Yes, that is a good rejoinder, and it is the rejoinder I hoped to provoke, actually. 🙂 Now, whether the paternalistic claim is actually right or not is another question. I haven’t studied the AA issue deeply enough to know — but it might be true. Glad you enjoyed the piece!

  5. In principle, I think you are right that particular people can be the victims of injustice committed against other (related) persons. And I understand why it might make sense to use race as a rough rubric for distinguishing between those who are victims of injustice and those who are not, since the original injustice was perpetrated against a group of persons on the basis of their skin color.

    I don’t understand, though, why this entails affirmative action rather than a massive class action. Showing damages and then compensating the victims of an injustice is the usual way we would rectify this kind of wrong isn’t it? It would be odd if the family of the person murdered by police (in your example) received compensation in the form of governmental assurance that it would be slightly easier for them gain entry into college or find a job. It seems particularly strained to me to claim that it is fair that particular employers or universities (especially if private) should have to bear the brunt of compensating the victims, if we are taking for granted that the injustice is one that all white people as a group have either committed or benefited from. Not to mention the particular white applicants that would otherwise have been successful in their applications to universities or businesses. Shouldn’t all white people have to bear the compensatory burden equally for the injustice? Or at very least, should have to pay in proportion to the relative benefit they have received from the inequality?

    I don’t think that money would really “solve” the inequities that are so strongly associated with race now. I don’t think money is normally enough to ever really compensate the victims of severe injustices. But just settling in a lawsuit and giving the victims of injustice money seems a lot less arbitrary to me than making some employers and businesses and applicants pay a disproportionate percentage of the compensation, as would undoubtedly occur with AA.

    1. I don’t think there is any good libertarian case for forcing private employers or universities to practice AA; I’m just thinking that there are reasons why state-owned universities and employers could practice AA in a world of the second-best, i.e., one in which a class action suit of the sort contemplated here will not be permitted.

Leave a Reply to Aeon J. Skoble Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s