10 Notes on the End of Affirmative Action

In a landmark decision handed down today, the Supreme Court ruled that the use of race in college admissions violates the 14th amendment––effectively ending “Affirmative Action” overnight. For some, this is a shocking step backwards. For others, it’s a long overdue endorsement of color-blindness and non-discrimination. 

I am in the latter camp. I think “Affirmative Action” is a misguided, discriminatory policy whose end is long overdue. But I know many intelligent and well-meaning people who disagree. In the coming weeks, I expect many media outlets to push an alarmist viewpoint that we have dialed the clock back to the days of Jim Crow. So I’d like to lay out, as clearly as possible, my reasons for believing that this decision is a net good for American society. 

1. “Affirmative Action” is a Euphemism for Racial Discrimination

We use euphemisms to hide ugly realities. During the War On Terror, for example, the phrase “enhanced interrogation” was used instead of “torture.” Whereas “torture” calls to mind hellish dungeons and blood-curdling screams––realities the public might not support––“enhanced interrogation” calls to mind…well, nothing in particular. That’s the point. 

“Affirmative Action” is a euphemism too. If you didn’t know what it meant, you would never guess that it referred to a policy of racial discrimination against certain people (whites and Asians), and racial preferences for others (blacks and Hispanics). 

Do “discrimination” and “preferences” sound too harsh? Isn’t race just used as a “tie-breaker” between otherwise identical applicants? 

Princeton sociologist Thomas Espenshade studied a representative sample of elite colleges and found that when other factors were held equal, Asians and whites had to score 450 and 310 SAT points higher than black applicants, respectively, to have the same odds of being admitted. (Espenshade, No Longer Separate, Not Yet Equal, chapter 3).

Harvard’s 2013 internal memo found that if they ended “Affirmative Action”, the number of Asian-American students would more than double, comprising 43 percent of the student body. That doesn’t sound like a tie-breaker to me. If the shoe were on the other foot––if black and Hispanic kids were at a disadvantage equal to hundreds of SAT points––no one would be tempted to dismiss it as “small.”

What goes on in admissions offices is shrouded in secrecy. If it were shown openly, it would seem ugly and racist. Here, for instance, is an actual conversation (revealed in the lawsuit) between admissions officers at UNC: 

“I just opened a brown girl who’s an 810 [SAT].”

“If its brown and above a 1300 [SAT] put them in for [the] merit/Excel [scholarship].”

“Still yes, give these brown babies a shot at these merit $$.”

“I am reading an Am. Ind.”

“[W]ith these [URM] kids, I’m trying to at least give them the chance to compete even if the [extracurriculars] and essays are just average.”

“I don’t think I can admit or defer this brown girl.”

“perfect 2400 SAT All 5 on AP one B in 11th” 

“Brown?!”

“Heck no. Asian.”

“Of course. Still impressive.”

“I just read a blk girl who is an MC and Park nominee.”....

“Stellar academics for a Native Amer/African Amer kid.”....

“I’m going through this trouble because this is a bi-racial (black/white) male.”

“Holistic admissions” is a propaganda phrase. Here’s the reality: racial identity––a fact which no one chooses about themselves––is the but-for cause of admission and rejection for a significant portion of American students applying to elite colleges. 

Imagine if every college rejection letter contained an honest account of why every kid was rejected. Imagine, for example, if the Asian-American kid who would have gotten into Harvard were she not Asian received an honest statement attesting to that fact in her rejection letter: “We regret to inform you that you’ve been rejected in part because you are Asian-American. Had you been black or Hispanic with otherwise identical qualifications, we would have accepted you.” 

How long would “Affirmative Action” survive in this scenario? If your intuition is “not very long” (that is mine), then what does it tell you that secrecy is the main reason “Affirmative Action” survived as long as it did?  

Sadly, euphemisms work. They shape public opinion. Consider the results of two surveys about college admissions. Pew asked: “Do you think race/ethnicity should be a major factor, minor factor, or not a factor in college admissions?” 73 percent of Americans said race/ethnicity should not be a factor in college admissions at all, including a majority of black respondents.

But the same year, Gallup asked: “Do you generally favor or oppose affirmative action programs for racial minorities?” And 61 percent of respondents said “yes”.

How can that be? How can majorities both support and oppose racial preferences? 

The key is how you ask the question. When you ask people if they support “Affirmative Action,” they’ll say “yes.” But if you instead ask the same question using a straightforward definition of “Affirmative Action”–– like “using race as a factor in college admissions”––those same people will say “no”. That’s the power of euphemism. (I’m sure “enhanced interrogation” polled better than “torture” too.) 

“Affirmative Action” may be popular. But the underlying reality it represents––racial discrimination and preferences––is unpopular with Americans of all races. 

2. “Affirmative Action” Affects the Elites, Not the Masses

Who is really affected by “Affirmative Action”? From how it’s discussed, you would think “Affirmative Action” affects a wide swathe of the black and Hispanic public. But you’d be wrong. 

By Princeton sociologist Thomas Espenshade’s estimate, in any given year, only 1 percent of black and Hispanic 18-year olds get into a college as a result of racial preferences. The other 99 percent either don’t go to college at all or don’t go to colleges selective enough to “need” racial preferences. Schools with acceptance rates over 50% generally don’t use affirmative action. 

“Affirmative Action” is an elite policy. Problems that affect elites get more attention than problems that affect the masses because elites have more media influence. This dynamic is as true with elites of color as it is with white elites. In the coming weeks, we’re going to hear that this decision restricts “access” to higher education. But “access” to higher education will not be affected by this ruling; what will be affected is the ease of entrance to a select few very prestigious schools. 

Take a look at the colleges attended by Fortune 500 CEOs. By and large, they are schools with acceptance rates between 60-95%––not the elite schools affected by this verdict. Did these CEOs lack “access” to higher education because they didn’t get into an Ivy? I don’t think so. Without attributing malice or dishonesty to anyone, I believe that framing “Affirmative Action” as an “higher education access” issue is a way for us “top 1% elites of color” (I include myself in this category) to give our own narrow self-interest a more noble gloss. 

3. The Benefits of “Affirmative Action” are Dubious

Many individual black and Hispanic students within the “1%” feel that they’ve benefitted from attending a more prestigious university than they would otherwise have gotten into. Perhaps they get access to better professors, networks, opportunities, and a stronger brand signal. 

That’s a separate question from whether Affirmative Action is responsible for the economic and social gains that minorities as a whole have made in the past half century. At first blush, it’s implausible that a policy which only affects 1 percent of black and Hispanic kids each year could be responsible for any major trends (positive or negative) among blacks and Hispanics as a whole. 

Secondly, there is a correlation-causation issue here. Minorities were making steady progress before Affirmative Action started in the 1970s. The economist Thomas Sowell often points out that the black poverty rate nearly cut in half––from 87 percent to 47 percent––between 1940 and 1960. It fell another 17 percent during the decade of the 60s, before Affirmative Action began.

When “Affirmative Action” truly launched in the 1970s, the downward motion of the poverty rate stopped, and did not resume going down again until the 1990s. To be clear, I’m not suggesting that Affirmative Action was the cause of that; I’m suggesting that it had little effect either way. The idea that “Affirmative Action” in college admissions has been responsible for the economic progress of black Americans as a whole has no empirical foundation.

Source: https://www.census.gov/library/stories/2020/09/poverty-rates-for-blacks-and-hispanics-reached-historic-lows-in-2019.html

It is even worth reexamining the belief that black individuals who get admitted under affirmative action policies benefit from a stronger brand signal to employers. For instance, people are dunking on Clarence Thomas as a hypocrite because he “benefited” from Affirmative Action himself  but is now destroying that same ladder to success for others. 

They are conveniently failing to mention that Thomas actually blamed Affirmative Action for his inability to get a job after law school. He put a “10-cent” label on his Yale Law degree, feeling that employers devalued it because they assumed he only got in because he was black. Don’t we all have those sorts of intrusive thoughts, rightly or wrongly, about legacy admits and student athletes? How many of us openly said that of President Bush? It would be strange if race-based affirmative action were the only anti-meritocratic policy about which people did not have such intrusive thoughts. And what good is the stronger brand signal if prospective employers are silently (and rationally) canceling it out in their heads? 

(Ultimately, the notion that Thomas is a hypocrite wouldn’t make sense even if he did believe he had benefited from Affirmative Action. You can derive benefit from a system while also believing that that system is unjust. President Bush, for instance, believed that the legacy system was unjust even though he likely benefited from it personally. Was he a hypocrite? What should Thomas have done to avoid hypocrisy, opt out of law school entirely? You don’t get to choose whether you benefit from a given system of admissions. But you do get to choose your opinion of that system.)   

4. Mismatch is Real

Affirmative Action has negative side effects even for those students it is intended to help. 

What happens when you put someone who scored 1900 on the SAT (a very good score) into a school where the average student got a 2300 (a great score). Perhaps they rise to the occasion because their potential was not accurately measured by the standardized test. Perhaps.

Or perhaps they feel insecure and unprepared compared to their peers. Perhaps they will feel like they’re barely treading water among skilled swimmers. They may even get demoralized, switch to an easier major, or quit school altogether. 

What I’ve described is called “mismatch” theory––the idea that students admitted under lower standards end up struggling more than their peers. Defenders of racial preferences have tended to dismiss evidence of mismatch. I disagree. I think there’s no doubt that mismatch has happened and continues to happen––though reasonable people could disagree on its scope and degree. 

In 2012, a group of social scientists at Duke University studied this topic. They found that black students entering Duke initially got worse grades than their white peers in the first year––exactly what you’d expect if you believe that admission standards predict performance. But by the end of four years, they found that black students had substantially closed the GPA gap with their white peers. An affirmative action success story right? 

Wrong. It turns out that the reason black students as a whole performed better as college progressed was because so many of them switched from harder majors to easier majors. Incoming freshmen of all races showed similar levels of interest in hard sciences. But the attrition rate out of hard majors for black students was far larger than for white students. Over 54% of black men who initially planned to major in hard sciences eventually switched to the humanities or social sciences. The figure for white men was less than 8%. (Women showed the same disparity but to a lesser degree).

How do we know the attrition from harder majors was a product of lower standards of entry, rather than some other race-specific cause (like racism)? Well, if the problem is lower standards of entry, then it should apply to any identifiable group of students who are admitted under lower standards––including legacy admits. Sure enough, the Duke researchers found the same effect for legacy admits: an illusory convergence of GPA with their non-legacy peers resulting from a higher attrition rate out of hard majors into easier majors. 

Whatever benefits racial preferences in college admissions have passed on to black and Hispanic students, it has certainly hurt many of them as well. No doubt, many students who would have followed through on their hard science majors at a less prestigious school ended up majoring in humanities simply in order to survive––hardly the intended result. 

5. “Affirmative Action” is Not the Product of The Civil Rights Movement

The phrase “Affirmative Action” was first used in a 1961 executive order by JFK which mandated that government contractors take “affirmative action” to ensure that employees are hired “without regard to their race creed, color, or national origin.” Four years later, Lyndon B. Johnson issued a similar executive order, using nearly identical language.

At that time, “affirmative action” simply meant serious efforts at non-discrimination––milquetoast to modern ears, but radical at a time when discrimination against blacks and other minorities was considered normal. It was not until Nixon’s 1969 “Philadelphia Plan” that the meaning of “Affirmative Action” shifted from non-discrimination to the realm of preferences and quotas. 

Martin Luther King Jr. was killed in 1968, before “Affirmative Action” came to have its modern meaning. Thus, we cannot know for certain what he would have thought of it. 

However, there are clues. There’s a commonly-circulated Dr. King quote, from his 1964 book Why We Can’t Wait, that is often alleged to indicate his support for Affirmative Action: 

Whenever this issue of compensatory or preferential treatment for the Negro is raised, some of our friends recoil in horror. The Negro should be granted equality, they agree; but he should ask nothing more. On the surface, this appears reasonable, but it is not realistic. For it is obvious that if a man is entered at the starting line in a race three hundred years after another man, the first would have to perform some impossible feat in order to catch up with his fellow runner. (Why We Can’t Wait, pgs. 158-159)

Shortly after this passage, King even mentions that he is aware of Affirmative Action programs in India. All of this, however, is only a prelude to his actual recommendation, which occurs 4 pages later: 

I am proposing, therefore, that, just as we granted a GI Bill of Rights to war veterans, America launch a broad-based and gigantic Bill of Rights for the Disadvantaged, our veterans of the long siege of denial…While Negroes form the vast majority of America’s disadvantaged, there are millions of white poor who would also benefit from such a bill. The moral justification for special measures for Negroes is rooted in the robberies inherent in the institution of slavery. Many poor whites, however, were the derivative victims of slavery…It is a simple matter of justice that America, in dealing creatively with the task of raising the Negro from backwardness, should also be rescuing a large stratum of the forgotten white poor. A Bill of Rights for the Disadvantaged could mark the rise of a new era, in which the full resources of the society would be used to attack the tenacious poverty which so paradoxically exists in the midst of plenty. (Why We Can’t Wait, pgs. 163-165)

Thus, Dr. King’s proposal for “compensatory or preferential treatment for the Negro” was in fact a class-based anti-poverty program he called the “Bill of Rights for the Disadvantaged” that would target the poor without regard to race––but would, of course, disproportionately benefit blacks because blacks are more likely to be poor. 

Years later, in his final book, Where Do We Go From Here?, Dr. King devotes an entire chapter to critiquing the Black Power movement. One of his core critiques is that they should focus on class instead of race. He even suggests that they change their name from “Black Power” to “Power for Poor People”. (King, Where Do We Go From Here?, pg. 51).

While Dr. King did not live long enough to comment directly on racial preferences in college admissions, we know that he had a stable preference, up until his death, for class-based policies––policies that would target the black and white poor alike––over race-based policies. And we know that he was aware of the existence of Affirmative Action in India and could easily have said “let’s copy India,” but instead recommended a class-based anti-poverty program. 

Bayard Rustin was a close friend and associate of Dr. King throughout his life and in fact ghost wrote parts of Dr. King’s Why We Can’t Wait. Rustin also led the 1963 March On Washington and persuaded Dr. King to found his flagship organization, the SCLC. 

Unlike Dr. King, Rustin did live long enough to see the rise of racial preferences and quotas, and he was clear in saying that they had nothing to do with the Civil Rights movement.

In a 1974 letter to the editor of the Wall Street Journal, Rustin wrote: 

The controversy over quotas and preferential treatment did not originate in the agenda of the civil rights movement, except insofar as that movement provided the impetus for all subsequent efforts to enhance the status of minorities. The leaders of the civil rights movement—King, Randolph, Wilkins, and others—were explicit in opposing reverse discrimination. They were opposed on philosophical grounds, but were also motivated by pragmatic political considerations. For they recognized that to transform the demand for Negro rights into a call for the displacement of whites would inevitably elicit instantaneous and widespread resistance from a society otherwise disposed to view the civil rights agenda favorably. Which brings us to a basic point, albeit one which is often overlooked or dismissed as of secondary importance. And that is that quotas are the progeny, not of the program of the civil rights movement, but rather of the economic policies of the Nixon administration and of the shortcomings of the administration’s bureaucracy…Weakening the merit principle and legitimate standards does no benefit to society, least of all to minorities. (Rustin, I Must Resist, pgs. 394-395)

To be clear, he is speaking of preferences and quotas in employment here. But his argument––that the whole idea of racial preferences did not stem from the Civil Rights Movement––is nonetheless a historical fact. 

Put simply, race-based “Affirmative Action” was not part of the agenda of the Civil Rights Movement, though many of its defenders would have you think it was. 

6. Quotas are a Red herring

Ever since the Bakke decision in 1978, racial quotas have been illegal. Schools have been able to “take race into account”––i.e. discriminate––as long as they don’t have specific numerical targets for each race in mind. 

The moral logic of this decision has always puzzled me. If you’re racially discriminating either way, why should it matter whether you have a predetermined quota in mind? 

If the shoe were on the other foot, would the quota vs. no-quota distinction matter? That is, if it were black people getting discriminated against, would anyone care whether that discrimination was quota-based or not? Would anyone mount a defense of a white supremacist institution by saying “Well, they may be discriminating against blacks, but at least they aren’t using quotas”? Of course not. Why does this distinction suddenly become important in the case of discrimination against Asians and whites?

The quota vs. no-quota distinction is the greatest red herring in the “Affirmative Action” debate. Perhaps it made sense as a strategic compromise, but let’s not pretend it ever made sense logically or ethically. 

7. We’re Confused About Diversity 

The original rationale for racial preferences was to repay black Americans specifically for the harm of racial discrimination. In fact, in the 1970s, the policy sometimes went by a different name: “compensatory justice”. (Urofsky, The Affirmative Action Puzzle, pg. 95)

Bakke marked a shift in the rationale from compensation for past racism to promoting racial diversity in general—–a change that was reified in Grutter v. Bollinger (2003). Since then, promoting racial diversity has been the official rationale for racial preferences in college admissions. 

Yet our attitude towards diversity is schizophrenic. Apparently, diversity is so important that colleges can discriminate in order to achieve it. But the moment you get on campus, you can live in all-black or all-POC student housing, take black studies courses, and go to an all-black graduation. In other words, you can segregate your life so that you experience no racial diversity at all. 

Not to mention, aren’t we proud of our HBCUs, despite the fact that they are some of the least racially-diverse colleges in the nation? If diversity is a crucial prerequisite for learning, as suggested in Bakke and Grutter, wouldn’t that imply that students at HBCUs aren’t learning?

So which is it? Is racial diversity so important that we ought to discriminate to achieve it, or do we love our non-diverse, single-race “safe spaces”? 

Or is it that diversity is somehow crucial for white kids, but not important for kids of color? And if so, why? 

As much lip service as people pay to diversity, the truth is that our attitude towards it is incoherent. And the invocation of diversity as a rationale for racial preferences has always seemed like a strategic, post-hoc rationale––especially when compared to the original impetus for the policy: compensating black Americans specifically for past racism. None other than Ta-Nehisi Coates, no opponent of Affirmative Action, wrote: “Isn't the ‘diversity’ argument actually kind of weak? Isn’t the recompensation argument actually much more compelling?”.

My personal view is that diversity is like love. When it happens naturally, it’s the most beautiful thing in the world. But the moment it’s arranged, legislated, or mandated, you’ve somewhat missed the point.  

8. Affirmative Action as Reparations?

What about the reparations rationale for “Affirmative Action”? What’s wrong with the idea that people who have been tangibly harmed by racist government policies should be made whole? 

This argument would have been very compelling to me in, say, 1970. In 2023, it appeals to me less.

Any claim of reparations loses strength over time. There’s a straightforward reason for this: as time passes, the descendants of the person harmed become more and more distant from the original harm.

Here’s my limiting principle on reparations claims––which I have applied consistently to all groups of people: reparations are due to the individuals harmed by government policy in their lifetimes. If that opportunity is missed––as it has been for the great majority of historical crimes––reparations should be paid to their immediate family (spouse or children). But if that opportunity is also missed, reparations are not therefore owed to their more distant ancestors (grandchildren and beyond). 

That has been my consistent principle––a kind of statute of limitations––which I apply even-handedly to all historical crimes. You may disagree. But my question is then: what’s your limiting principle on reparations, and do you apply it even-handedly? 

The implied theory behind reparations for slavery is that reparations claims never expire. If that’s what people believe, then they should make it explicit and apply it even-handedly. 

Finally, it is worth noting that “Affirmative Action” as reparations would have to selectively exclude recent black immigrants from places like Jamaica, Haiti, Ghana, and Nigeria––something which few outside of radical groups like ADOS seem eager to do. This is no small point: one of the few studies which has looked at the question found that 41 percent of black students at elite schools are not descendants of slavery and Jim Crow, but the children of post-1965 black immigrants.

9. The Equilibrium Will Change

The result of banning “Affirmative Action” is not only up to the Supreme Court. It is also up to the many actors––with their own incentives and constraints––who will change their behavior in response to the ban. 

First, admissions offices will find clever ways to create diverse student bodies without directly using race. They could institute a “geographic diversity” system; they could give priority to the top X percent of students at every high school; they could give every student an “adversity score” that combines variables (like neighborhood crime rate, parental income, and more). They may implement systems so clever that we could not even predict them now. 

Admissions offices have strong incentives––due to internal culture and workplace values––to find clever, legal ways to create diversity. They will not simply continue with business as usual minus the use of race. A new equilibrium will be reached. 

Student applicants may change their behavior as well. If colleges are not allowed to ask the race of applicants, for instance, we may see black and Hispanic applicants volunteer their race with greater frequency in personal essays. So I expect there will be much less change than the alarmist media want you to think.

It’s worth reflecting on the fact that many states have already banned Affirmative Action––for decades in some cases. I don’t remember anyone mentioning this to me when I was applying to colleges as a black/Hispanic student in 2013.

In fact, I’m guessing a lot of people just learned about state-level affirmative action bans in the past month. If our nation’s colleges were about to become segregated hellscapes, don’t you think you would have heard––at some point in the last 20 years––that California, Washington, Florida, Michigan, Nebraska, Arizona, New Hampshire, Oklahoma, and Idaho were already segregated hellscapes?  

I expect less to change than do the media outlets that make money by creating fear. Nevertheless, I support this decision by the court, because, regardless of its downstream effects––which will be in part determined by how colleges and applicants change their behavior in response––it sends a strong signal that racial discrimination is immoral, illegal, and unconstitutional. I believe that is a good signal to send, in line with the ethos of the Civil Rights movement. (See Point #5). 

Finally, many people are rightly pointing out that legacy admissions are also anti-meritocratic and should be ended. I agree. Yet some have leveled a hypocrisy accusation––arguing that the Court should have ruled against legacy admissions in this case, based on the logic that legacy admits are disproportionately white. This is absurd. If they had ruled against all admissions policies which had a racially disparate impact, they would be presumptively banning all the other strategies schools might adopt to increase diversity––including geographic preferences and all the rest. 

10. If Not Affirmative Action, then What? 

“Affirmative Action” in college admissions is often viewed as a way of reducing inequality. As I argued in point #3, the policy’s track record on this point is dubious. But if “Affirmative Action” isn’t the solution, then what is? 

To speak narrowly about black and white Americans––who are at the core of the issue––I believe one of the most important drivers of outcome gaps is the “human capital” gap. By “human capital,” I mean skills, knowledge, abilities, habits that lead to higher productivity. This is by no means the only driver of outcome gaps. Racial discrimination is a factor. Differences in social capital are also important. But the human capital gap is, in my view, the most important. 

Much of the human capital gap is determined by peer culture and parental investment––which are beyond the direct reach of the state. But education is a major lever that the state can pull to influence human capital. From that observation, it might seem to follow that racial preferences in college admissions play a role in addressing the human capital gap. 

But recall from point #2 that only 1 percent of black and Hispanic 18-year olds get into college as a result of Affirmative Action. Not to mention, Affirmative Action kicks in at 18-years old, after the most crucial years of child development are already over. 

It is between the ages of 0 and 18 that we can have the greatest impact on the human capital of black and Hispanic kids. A program which begins after kids have turned 18 is not a serious intervention. A serious intervention would focus on the younger, more malleable years.  

What kind of intervention do I have in mind? Harvard economist Roland Fryer has done some of the most groundbreaking but neglected work on this question. From studying 39 charter schools in New York City, Fryer identified five variables that correlated with success: (1) frequent teacher feedback, (2) the use of data to guide instruction, (3) high-dosage tutoring, (4) increased school time, and (5) a culture of high expectations.

Observational data such as this is useful, but suffers a fatal flaw: you can never control for all the pre-existing differences between the kids who end up at better charter schools and the kids who end up at worse ones––even with charter “lotteries” (which are often not as random as they seem). 

So Fryer went further. He gained unprecedented access to the Houston public school system and designed a randomized experiment. He took a set of 16 low-performing schools––containing just over 5,000 students––and created “matched pairs” of schools based on the relevant student demographics. He then randomly assigned one school in each pair to the control group and one to the treatment group. 

Schools in the treatment group received complete overhauls based on the principles learned from high-performing NYC charter schools: The principals were fired and replaced––along with about half of the teachers. School days were lengthened. Extra tutoring was provided. Extra tests were given and the results were used to tailor each student’s tutoring. And a culture of excellence and high expectation was consciously fostered from the top-down. The overall result was a 0.103 standard deviation increase in math scores for black kids, with even larger gains for elementary-aged kids, and kids from lower socioeconomic backgrounds.  

If one-tenth of the energy directed at preserving “Affirmative Action” were directed at implementing the Fryer protocol at schools across the nation, that would constitute, in my view, a serious intervention aimed at attacking the root causes of racial inequality.