An article in the current issue of The Economist (“A Lawsuit Reveals How Peculiar Harvard’s Definition of Merit Is”) raises again the problem of discrimination. A private organization has sued Harvard University for discrimination against Asian-American applicants. The discovery process forced Harvard to reveal much information about its admission process, which confirmed what everybody suspected–that Harvard does discriminate, directly and indirectly, on the basis or race.
It is difficult not to accept that a private entity, including a private university like Harvard, should be able to discriminate as it wishes. One might reject the morality of some forms of discrimination. For example, discrimination against Jews in the Progressive Era, including at Harvard, was certainly objectionable. But some of us will have different views on different forms of discrimination. That private discrimination should not be banned (as many forms are now banned by federal law, including on the basis of race) is supported by good moral and economic reasons.
The moral basis of value judgments in public policy evaluation can perhaps be summarized in the principle proposed by Robert Nozick: the law should not forbid “capitalist acts between consenting adults” (Anarchy, State, and Utopia).
There also exist good economic reasons to favor the private liberty to discriminate. Economic reasons have to do with consequences in terms of individual income, opportunities, and social coordination. Only a private entity can know all its circumstances. For example, only Harvard can determine whether its practice of discriminating in favor of alumni’s children is desirable.
Moreover, competition forces discriminators to pay the cost of their discrimination, as famously demonstrated by Gary Becker. Consequently, one has an incentive not to indulge in discrimination too much, if at all. If Harvard discriminates on the basis of race, it will (with some probability) lose out to MIT or Caltech, two other private universities. (Incidentally, Lauren Landsburg has a fascinating EconLog article on Becker as a teacher.)
A general habit of judging individuals on the basis of the groups they belong to could have dire consequences in the long term. Could this consideration provide a libertarian basis for some antidiscrimination laws? Probably not. At any rate, government cannot help by mandating discrimination to fight discrimination, that is, with affirmative action. The case under consideration supports these doubts.
What is sure is that discrimination by public institutions and bodies should not be allowed. It is contrary to the rule of law, which treats all individuals equally. Public discrimination is unjust to the individuals who are forced to pay taxes like others but do not benefit from an equal treatment by the state. From a narrower economic viewpoint, public discrimination or a legal obligation to discriminate privately (as was long the case against Blacks in America or South Africa) limits exchange with the individuals discriminated against and thus reduces the general benefits of exchange.
I would argue that the subsidies that Harvard gets, which are, I suppose, open to all other universities and represent only a small portion of Harvard’s budget, are not sufficient to justify regulating it like a public university. Such subsidies show the danger of government subsidies to private organizations.
The Economist article reminds us how affirmative action is just discrimination in reverse. The federal government has imposed affirmative action for several decades, even if numerical quotas are now deemed illegal. Forced antidiscrimination preferences are disguised quotas. By mandating discrimination in favor of certain races or ethnic backgrounds, the federal government fuels discrimination, for it implies discriminating against individuals from the non-protected races, that is, Asians and Whites. One more non-Asian or non-White admitted at university because of his race or ethnic background means closing the door to one Asian or White.
As Friedrich Hayek wrote in The Constitution of Liberty,
It is ironic that Harvard University is being sued for discriminating the way the government—and probably most of its politically correct faculty—wanted, that is, in favor of Hispanics and Blacks. A compounding irony is that Harvard is also being investigated by the Department of Justice for the same affirmative action. (It would not be surprising if the DOJ had some political motivation.)
One cannot obey legislation that mandates to both discriminate and not discriminate. This is what affirmative action is. Shouldn’t Harvard now claim a private right to discriminate? Will this bring the Harvard crowd to discover the benefits of liberty?
READER COMMENTS
Jackson
Jun 25 2018 at 2:16am
I agree that the government shouldn’t discriminate based on race, but I find it odd that you said “Public discrimination is unjust to the individuals who are forced to pay taxes like others but do not benefit from an equal treatment by the state.” Aren’t all people who don’t get into a public college being discriminated against by the state? Even if it’s not based on race, but rather IQ (SAT or equivalent scores) or whatever other factors, aren’t the rejected parties still taxpayers who aren’t being afforded the right to access public resources?
Thomas Sewell
Jun 25 2018 at 11:43am
Yes, everyone discriminates all the time, whenever they choose one person or thing over another.
The question here is specifically racial discrimination, based on whatever imagined category of “race” the discriminator came up with to classify individual human beings. Pierre is using a shorthand for racial discrimination, which I agree can be confusing when not clarified, but is a habit many people have developed.
In this case, it’s generally accepted that merit-based discrimination (i.e. SAT scores) which indicate someone’s ability to benefit themselves and/or others based on their abilities are ok to use as a factor in decision making. Instead, if you are using some general characteristic such as “race”, when you can instead make decisions based on the individual’s actual characteristics, then that’s very poor decision making which also isn’t fair to the individuals involved.
It’s the difference between deciding to purchase a particular appliance because it’s less expensive, or does the job better, vs. deciding which appliance to purchase because of the skin color of the people who are selling appliances. One of those decision-making heuristics is typically going to lead to better outcomes over the long term than the other one.
Pierre Lemieux
Jun 25 2018 at 5:38pm
@Thomas Sewell: I agree, Thomas, I should have been more careful.
Jackson
Jun 26 2018 at 5:37pm
Correct me if I am wrong (it wouldn’t be the first time), but this seems like a materially different argument than the one being made in the post in an important way. Prof. Lemieux’s argument (simplified) seems to be that government (racial) discrimination is wrong because it leads to unequal treatment of citizens, which undermines the rule of law.
Your argument seems to be that racial discrimination in colleges is a bad idea because race is irrelevant in determining the value of one student over another. Your argument creates an exception to the “treat everyone equally” argument. Specifically, if I understand you correctly, you are saying it is not ok for the government to discriminate based on any factor that is irrelevant to the quality or efficiency of the government services, and that race is one of those irrelevant factors.
This is a much weaker argument than Prof. Lemiuex’s, as one would only need to show that race is indeed relevant to educational quality and efficiency in order to show that it is an acceptable form of discrimination. Indeed, I believe that is what colleges try to do when they make their cases for doing it. Their arguments often include something like “having a racially diverse student body increases the quality of our institution because people from different racial backgrounds have different perspectives and combining different perspectives is what our goal is.”
One may certainly disagree with that, but at that point it’s just weighing whether racial diversity improves educational quality or not (and the fact that many of the most successful private institutions such as Harvard do it is at least some evidence that it might). From the way I read it, Prof. Lemieux’s argument implies that racial discrimination by public institutions is wrong even if quality and efficiency are improved. But that argument seems too strong to me, as I explained in my previous comment.
Thomas Sewell
Jun 26 2018 at 10:00pm
@Jackson,
I agree that there may be some argument for a University using some definition of “race” in admissions if that’s all the information they have about applicants. Back in the real world, they have a whole host of individualized criteria and information sets about applicants available to the admissions staff. If there is anything else they wanted to know, they could add that to their application process.
In that situation, information about “race” doesn’t add anything to the process. First, you’d have to come up with a rational definition, rather than the sort of fuzzy bureaucratic mix based mostly on historical misunderstandings and non-scientific groupings most places currently use. Universities are supposed to base things on science, right? Not accidents of geography or similar skin color. From aAs early as 1950 in “The Race Question”, it’s been clear there isn’t much scientific basis for the races as defined (from 3 to 11 different or more?) over various decades.
BTW, I didn’t say “quality or efficiency of the government services”, I said “merit-based discrimination which indicate someone’s ability to benefit themselves and/or others based on their abilities”. The difference is that one can be argued for on the basis of justice/merit, while the other may or may not be.
Using some psuedo-science racial classification based on skin color/appearance is impossible in my opinion to meet that standard better than whatever more specific criteria it’s being used as a proxy for (poverty, life experience, background, etc…). For example, even if what you really care about is if someone has sickle cell, you can just make that specifically your criteria instead of “race”.
Otherwise you end up with the absurd result of the lesser-prepared child of rich and famous black parents getting a massive admissions advantage over a much harder working and better prepared student whose family struggled with poverty as refugees from Cambodia, for example. In what sane system does the first person get bonuses on their admissions compared to the second?
TL;DR version: If your goal is different perspectives and backgrounds, then select for different perspectives and backgrounds, not race. Two children of State Department Yale graduates have more in common background and perspective-wise even if they are classified as black and white than either of them compared to the child of a agriculture worker, even if they are classified as the same race by admissions staff.
Hazel Meade
Jun 26 2018 at 1:28pm
Interesting comment. there are actually some states, such as Arizona, where colleges are mandated to admit every student who meets a threshold criteria, so there aren’t a fixed number of spots. A valid argument could be made that admitting every student who qualifies and then setting the threshold low enough to get a reasonable distribution of society would be less discriminatory. If you want to be more inclusive, you just admit more people of all socioeconomic backgrounds instead of selecting individuals. This might be a more viable solution for public institutions.
Hazel Meade
Jun 25 2018 at 11:16am
A good an nuanced article. I’d like to point out a recent post by David Henderson on movie theater segregation:
http://stageeconlib.wpengine.com/archives/2018/05/was_segregation.html
In general it is not always true that discriminators pay the cost of their discrimination, as argued by Becker. Sometimes, discrimination is driven not by the preferences of owners, but by their customers. Indeed, in the case of Harvard, it seems that some of their concern is driven by a belief that white students will avoid the university if it becomes “too Asian”.
As I’ve argued elsewhere, while I don’t support government imposed anti-discrimination laws, I do support the development of anti-discriminatory <i> social norms </i>. In other words, the problem here is the attitude of white students towards Asians along with the administrators response to those attitude, and the solution is changing those attitudes and educating white and Asian students to create a truly integrated campus where whites and Asians (and everyone else) can feel comfortable working together, as well as challanging the administrators to overcome their fear of having a campus with too many Asians.
Of course Harvard is a private institution which is entitled to tune admissions to accomplish various ends it desires. But those should be positive ends, not ends that are based on fear or prejudice or entrenched racial divisions. Harvard should be working to overcome racial prejudice in all it’s forms, whether it involve African Americans or Asians. There is no contradiction between practicing Affirmative Action when it comes to African Americans and not discriminating against Asians. As always, admissions should be made on a case by case individual basis rather than based on racial quotas. That is to say, there is a role for affirmative action in admissions, but it should not lie in attempts to achieve any particular racial makeup of the student body. It should be focused on individuals in a case by case way.
Pierre Lemieux
Jun 25 2018 at 5:49pm
@Hazel Meade: You are right that this is an important argument that I overlooked. It does reinforces the fear I expressed in the next paragraph. My only quibble with your comment is that your last paragraph does not take “affirmative action” in it usual sense of positive discrimination; if you use it in that sense, Hayek’s observation is relevant.
Hazel Meade
Jun 26 2018 at 1:34pm
I actually am trying to use it in the positive sense. I do think that it should be acceptable for admissions officers to discriminate slightly in favor of black people as a means of adjusting for historical and ongoing factors that make it harder for black people to succeed. I just think it should occur more on the margins than with quotas. If you have a fixed number of spots and you have to decide between two nearly equally qualified candidates who are both on borderline of the selection process, it should be okay, as a private entitiy, to decide that you’re going to admit the black person over the marginally more qualified white or Asian, because it serves a beneficial social purpose and you want to really give people from the most disadvantaged groups a little extra help. As a private institution, they are entitled to do that. And that person who doesn’t get in will just end up in a school one tier lower, not completely kicked out of college.
In the case of Harvard, it’s apparent that what is going on is not marginal. It’s over 10 percentage points of difference.
Hazel Meade
Jun 26 2018 at 1:43pm
And to add, it should be based on some individual judgement about how much the individuals background may have really hindered them. Like if there’s a black candidate from a really poor district who is somehow amazingly outperforming his peers, you might think that person would do even better if he was put in an environment where he could really excel. And if there’s someone whose getting straight A but comes from a rich family and went to a prep school where he got a lot of hand-holding and SAT prep classes and so forth, you might think that his scores might not be due to superior skills, but due to having a lot of hand holding on the way up. it should be acceptable for admissions officers to handicap people a little bit based on their backgrounds, to try to seek the diamonds in the rough.
Weir
Jun 25 2018 at 7:18pm
Bill de Blasio should try that line: The concern that white students will avoid the Bronx High School of Science if it doesn’t keep enough Asians out. He should try telling people it’s all the fault of those racist white geeks and nerds losing interest in Bronx Science, and see if anyone thinks that’s remotely plausible as an excuse for his racist policies.
R Richard Schweitzer
Jun 25 2018 at 12:30pm
“What is sure is that discrimination by public institutions and bodies should not be allowed. It is contrary to the rule of law, which treats all individuals equally.”
From the article
“The expression “the rule of law,” taken precisely stands for a mode of moral association exclusively in terms of the recognition of the authority of known, non-instrumental rules (that is, laws) which impose obligations to subscribe to the adverbial conditions in the performance of the self-chosen actions of all who fall within their jurisdiction. This mode of association may be opprobriously branded as “legalistic” and other modes may be considered more interesting or more profitable, but this I think is what the rule of law must mean. Like all other modes of association it is an abstract relationship of personae — persons solely in respect of being alike and without exception the subject of these obligations to one another.”
Michael Oakeshott: The Rule of Law; part 6.
R Richard Schweitzer
Jun 25 2018 at 1:01pm
Law describes, defines, but does not necessarily delineate observed social order and the relationships within it; generally, in terms of the obligations in the relationships in particular circumstances..
Legislation differs from law, since legislation is only Rules of Policy.
Rules of Policy (legislation, regulations, ordinances and their excrescences) are attempts to describe, define and delineate desired social order and the relationships necessary for it.
Confusing Rules of Policy (legislation, etc.) with Law, creates great difficulties when the systems and institutional frameworks that determine Law are employed to give effect to Rules of Policy. This is undoubtedly due to the problems associated with the determinations necessary to delineate a commonly accepted desired social order and the relationships within it.
Pierre Lemieux
Jun 25 2018 at 5:59pm
@R Richard Schweitzer: Thanks for your comment. I agree with you if one takes “law” in Hayek’s sense, which he opposed to legislation or, more precisely perhaps, to rules of organization. (You are obviously familiar with his Law, Legislation and Liberty, especially Vol. 1.) In the sentence you quote from me, I was using “law” in its degenerated sense (which tends to be the commonly accepted sense today), as you rightly pointed out.
Gary Anderson
Jun 25 2018 at 9:15pm
We have ghettos because black people were discriminated against for a century by Jim Crow Laws. How about giving affirmative action a century?
Pierre Lemieux
Jun 25 2018 at 10:30pm
@Gary Anderson: We see at Harvard what it leads to. And after the century is over (half is already over, in fact), will there be a century of affirmative action for Asian Americans, and then another one for Whites, and back to the Blacks, and so forth?
SaveyourSelf
Jun 26 2018 at 8:07am
Public discrimination is unjust to the individuals who are forced to pay taxes like others but do not benefit from an equal treatment by the state.
Yes, I completely agree, but this is a problematic argument because the taxes themselves discriminate. They are not the same for everyone. It could be argued that, using a flat tax design, everyone is taxed the same percentage if not the same amount. But neither a fixed quantity tax system nor a simple flat percentage tax system exists anywhere that I know of. So, in light of an unequal tax structure, how could public institutions then not discriminate? The Supreme Court decision to justify the “individual mandate” in Obamacare makes this problem more than hypothetical academic curiosity.
John Alcorn
Jun 26 2018 at 11:06am
(My comments sidestep two issues raised by Pierre Lemieux’s stimulating blogpost: Is affirmative action justifiable? Is the distinction between public and private universities crucial?)
Customers (students) are also ‘inputs’ in the production of education, perhaps especially at highly selective, residential universities. They learn more from one another than they do from Faculty. They form social networks that are potentially useful in a variety of careers, and perhaps key in elite finance and law. A subset of alumni eventually become major donors to the alma mater. Insofar as the composition of the student body shapes interactions and then philanthropy, students, alumni, and Trustees have a fundamental interest in the composition of the student body.
Politics constrains and distorts Admissions criteria. There is evidence that recent applications of affirmative action change the demographic proportions of the student body, but largely fail to achieve interaction and integration:
“African American students had a significantly higher fraction of same-race friends (and thus a lower fraction of other-race friends) in college than in high school. That colleges seem to be less conducive to interracial friendship formation is surprising, as students in most colleges live in closer proximity to students of other races and ethnicities than typically was the case in the neighborhoods from which they came. Because of housing segregation, college campuses tend to be more diverse than the areas in which students were raised, and because most students live on campus for several years, one might predict the closer contact would lead to more interracial friendships. That colleges do not tend to foster much interracial friendships thus suggests substantial room for improvement in promoting cross-racial interaction.” — Peter Arcidiacono, Michael Lovenheim, & Maria Zhu, “Affirmative Action in Undergraduate Education,” Annual Review of Economics 7 (2015) 487-518, p. 510. Gated publication: https://goo.gl/JNPN7A Ungated MS: https://goo.gl/WSVyVC
There is also evidence that affirmative action may be either a curse or a blessing for the target student:
“affirmative action policies can influence graduation rates through two conflicting mechanisms: (a) the mismatch effect, in which graduation rates decrease due to students being paired with colleges at which they are less likely to perform well, and (b) the college quality effect, in which graduation rates increase due to access to better colleges and more resources.” (ibid., p. 500)
If universities wish to pursue affirmative action in Admissions in good faith, then they should design policies to achieve a) productive interaction and integration among diverse students and b) the best interest of the target student. Given political pressures, universities must take special care not “to use” affirmative-action students in virtue-signaling games.
A saving grace in U.S. higher education is that any talented, motivated student who suffers adverse discrimination in Admissions at a particular university (e.g., the Asian-American plaintiffs at Harvard) may readily land a place at another highly-selective university, perhaps a notch or two down the pecking order. Given the range of opportunities, higher education is not a locus of highly-consequential injustice. Barriers to immigration and highly punitive prohibitions of consensual behaviors produce much greater injustice.
Hazel Meade
Jun 26 2018 at 1:55pm
So, actually I had an idea a while pack about this. It occurred more in the context of people with Asperger’s Syndrome, but it could be applied equally well to interracial friendships.
What if colleges had some sort of course where people were randomly assigned friends, during their first semester? Like, instead of random roommate assignments, you would be required to actually hang out with someone and get to know them, and then write a term paper on them at the end of the semester. And then you would be graded on how well you actually got to know that person, based on what you reported in your paper. Papers could be checked against online profiles and admissions applications to verify honesty, but would have to involve a level of intimacy more than a facebook page to get an A.
At the very least, it would be an interesting sociological experiment. What lengths would people go to to avoid having to hang out with someone from another race? Which people would take it seriously? Would those friendships last?
John Alcorn
Jun 26 2018 at 11:47am
I hasten to add a clarification to my previous comment. I have in mind Bryan Caplan’s distinction between individual and social perspectives, in his tour de force, The Case Against Education (Princeton U. Press, 2018).
An individual applicant who suffers adverse discrimination in Admissions at, say, Harvard, will land on her feet at another highly selective university. The impact of injustice on her is minor by comparison to impact of injustice on the poor migrant who is denied entry to any prosperous country.
However, higher education produces great injustice at the social level, if higher education is mostly a zero-sum signaling game (a wasteful rat race), which relegates to an underclass youths who fare poorly in school.
Pierre Lemieux
Jun 27 2018 at 11:35pm
Many interesting threads here. Let me come back to a point made by Hazel (with the help of David Henderson). On second thought, is it sure that the discriminators don’t always pay? It seems that customer-discriminators do pay a cost, even if perhaps lower or not as easily measurable as profits. Suppose you have two groups in society: the Blues and the Yellows. (Suppose also that they are in equal number.) Two thirds of the Blues never want to eat in a restaurant where some customers are Yellow. Two thirds of the Yellows never want to be in a restaurant where some customers are Blue. One third of the Blues and one third of the Yellows don’t care: rename this intersection set the Greens. Entrepreneurs will open restaurants “For Blues Only” and “For Yellows Only,” and presumably also some “No Discrimination” restaurants. The Greens will have a choice of two sorts of restaurants, while both the Blues and the Yellows will patronize only one sort. The Blues and the Yellows have less choice, which is the cost they pay for their discrimination.
An illustration with a little tweak: In a free society, where both smoking restaurants (permitted, not compulsory smoking) and non-smoking restaurants can legally exist, the die-hard anti-smokers pay a cost. To eliminate or reduce this cost, the government bans smoking restaurants.
John Alcorn
Jun 28 2018 at 4:36pm
Dr. Lemieux,
Your point that customer-discriminators pay a cost seems overdrawn. Yuppies prefer clubs where yuppies gather, bikers prefer biker hangouts, and maybe no one prefers to socialize with a representative cross-section of the population. People who care a lot about sorting and mixing have more options in cities, where entertainment markets are more complex. And they vote with their feet. Surely, they see high benefits where you see costs.
Comments are closed.