Why affirmative action had to go
As America diversified, a policy designed to help Black people was turning into a complex, confusing racial spoils system.
As of this writing, the decision hasn’t been handed down yet, but everyone thinks that the Supreme Court is likely to strike down affirmative action — i.e., race-based admissions policies at selective colleges. The lawsuit in question alleges that these policies, which generally are set up to give an admissions boost to Black and Hispanic students, result in de facto discrimination against Asian Americans.
To be honest, the lawsuit is probably right. The way our university system is set up, admission to top schools is basically a zero-sum game — the number of spots is harshly limited in order to pump up the prestige of these schools and their expenditure per student. Putting racial preferences into that game makes it, to some degree, a zero-sum game between racial groups. Removing those racial preferences — along with legacy preferences (which pretty much everyone hates) and athlete preferences — would probably leave White admissions mostly unchanged, and boost Asian admissions at the expense of Hispanic and (especially) Black admissions:
A lot of people are very upset about this outcome, alleging that Black students in particular are still uniquely and severely disadvantaged. And to be honest, they’re probably right. But the affirmative action system probably couldn’t last in its current form — in a rapidly diversifying country, affirmative action as we know it was evolving away from a measure intended to address the Black-White divide, and toward a complex and opaque racial spoils system.
As for what will replace affirmative action, the answer is probably class-based preferences — focusing on admitting kids from poor families instead of kids from specific races. That system isn’t perfect either, though, and there are additional things we should do to improve college admissions — most importantly, expanding the student bodies at elite schools.
Why affirmative action can’t handle a highly diverse country
It’s odd to say that diversity was affirmative action’s downfall, since diversity is the exactly the principle that affirmative action has long rested on. A Supreme Court case in the 70s decided that although explicit racial quotas were forbidden, race could be one factor among many, for the purpose of ensuring diversity in the classroom.
In 1978, White people were still around 80% of the populace, and most nonwhite Americans were Black:
Therefore, “diversity” really just meant that Black people got included. SCOTUS’ diversity rationale depicted having Black kids in the classroom as something that would benefit White kids as well. But really everyone knew that Black kids would be the main beneficiaries — that the most important purpose of affirmative action was not diversity but equity. The idea came directly from Lyndon B. Johnson, when he said:
You do not take a person who, for years, has been hobbled by chains and liberate him, bring him up to the starting line of a race and then say, “you are free to compete with all the others,” and still justly believe that you have been completely fair…We seek not just legal equity but human ability, not just equality as a right and a theory but equality as a fact and equality as a result.
Though I don’t have concrete evidence, I suspect that there was another purpose as well — to make Black Americans feel more of a sense of inclusion in, and ownership of, the United States of America. In other words, I suspect that affirmative action was intended to be what Daron Acemoglu and James Robinson call an “inclusive institution”. Acemoglu and Robinson believe that inclusive institutions are fundamental to nations’ ability to generate popular buy-in for the creation of public goods, and, ultimately, for economic development. If Black Americans — one out of eight Americans — believe that the country just isn’t set up for them, and that they’re not really full citizens, that will tend to make the country function less effectively.
In any case, affirmative action obviously didn’t bring about perfect racial equity in American society, but it probably did make some progress in that direction. For example, a 2020 paper by Akhtari, Bau & Laliberté found:
[W]e study a U.S. Supreme Court ruling that reinstated affirmative action in three states…When affirmative action is re-instated, racial gaps in SAT scores, grades, attendance, and college applications fall. Average SAT scores for both whites and minorities increase, suggesting that reductions in racial gaps are driven by improvements in minorities' outcomes.
Notice that this paper actually increased SAT scores, suggesting that affirmative action works via increased motivation, not just by putting its thumb on the scale for some races.
And a 2012 paper by Amalia Miller and Carmit Segal found:
We find that AA increased black employment at all ranks by 4.5 to 6.2 percentage points relative to national trends. We also find no erosion of these employment gains in the fifteen years following AA termination, although black employment growth was significantly lower in departments after AA ended than in departments whose plans continued.
And a 2017 paper by Conrad Miller found:
Affirmative action increases the black share of employees over time: in 5 years after an establishment is first regulated, the black share of employees increases by an average of 0.8 percentage points. Strikingly, the black share continues to grow at a similar pace even after an establishment is deregulated.
In other words, it’s probably true that affirmative action was a significant factor in the creation of the Black middle class in America. Income and wealth gaps between Black and White Americans didn’t really shrink much (mainly due to a few White people getting incredibly rich), but many Black Americans won a place in the nation’s elite and in the ranks of the economically secure.
But as the nation grew more diverse, affirmative action stopped being mostly about Black people. The Hispanic share of the U.S. population tripled, and now stands at around 19%, while the Black share basically didn’t change. The Asian percent grew to around 6%, meaning that about a quarter of the nation is now neither White nor Black. As you can see from the chart above, that percentage is expected to continue to grow.
But even that chart dramatically understates the racial complexity of the future America. The percent of mixed-race Americans is growing very rapidly, due to very high outmarriage rates by native-born Hispanics and Asians. In a few decades, single-race charts like the Pew chart above just won’t make sense anymore. As the sociologist Richard Alba writes in his book The Great Demographic Illusion, racial self-identification among mixed-race Americans doesn’t necessarily describe as strong or consistent of a racial identity as among single-race Americans, especially back in the days when the Black-White divide defined America. Many Hispanic Americans are already mixed-race, and we’ve already seen big swings back and forth in the percent of Hispanics that identify as White.
On top of that, we have shifting patterns of disadvantage. The persistent Black-White economic gaps suggest that Black disadvantage will not be dislodged easily. But at the same time, Hispanic Americans are steadily catching up with their White counterparts. This is from Chetty, Hendren, Jones & Porter (2018):
Hispanic Americans are moving up significantly in the income distribution across generations. For example, a model of intergenerational mobility analogous to Becker and Tomes (1979) predicts that the gap will shrink from the 22 percentile difference between Hispanic and white parents observed in our sample…to 6 percentiles in steady state…
Hispanics are on an upward trajectory across generations and may close most of the gap between their incomes and those of whites…Their low levels of income at present thus appear to to be primarily due to transitory factors.
Meanwhile, disadvantage is very difficult to quantify for many groups. For example, if you just look at averages, Asian Americans seem advantaged relative to White Americans, since they have higher income and lower poverty rates. But “Asian” is a bit of an artificial category, including people with ancestors from China, India, and many other disparate countries. When you break the Asian category down by national origin, you find that some subgroups have very high poverty rates, while others have extremely low rates. Asian poverty is also much higher in certain regions of the country.
In other words, all the assumptions that went into the creation of affirmative action — a society divided along a Black-White line, where White people had a clear economic advantage — increasingly don’t hold true in America, and will hold even less true over time. In 1978, affirmative action usually meant awarding some preference to a working-class Black applicant over a middle-class White one. In 2050, today’s affirmative action policies would mean awarding some preference to a rich White applicant with one White Cuban grandparent over a working-class Bhutanese-Hmong American, at a time when Hispanic average income is 87% as high as White average income, simply based on the theory that Hispanics are structurally disadvantaged and Asians are not.
Does that make any sense whatsoever? No, it does not. And yet this is where we’re headed. In their book The Diversity Paradox, sociologists Jennifer Lee and Frank Bean interviewed some Americans with partial Hispanic ancestry who identify as White in their daily life but who “check the box” as Hispanic for the sake of college admissions.
An alternative, of course, is to make affirmative action more complex. We could call in armies of economists to reevaluate privilege and disadvantage for various groups every 10 years or so — an evaluation that would certainly be bitterly contested and would depend a lot on opaque statistical assumptions, as the academic debates in the current affirmative action lawsuit show. And we could then award racial preferences based on some weighted average of Americans’ ancestries.
If this sounds dystopian to you, it’s because it would be dystopian. Doing college admissions by a combination of blood quantum and econometric assumptions would turn opportunity in America into a bizarre, opaque racial spoils system of Byzantine complexity. If you’re looking to create inclusive institutions in America, this is not how you do it.
In fact, people don’t even think the current system is fair. Though support is fairly high when you just ask people what they think of “affirmative action”, when you actually spell out what that means, they are strongly against. A recent poll by the Washington Post asked specifically about the possibility of SCOTUS banning all racial preferences in admissions, and found strong support among White, Asian, and even Hispanic Americans, with Black Americans split:
In other words, despite the benefits of the current system, it’s pretty clearly not a good long-term fit. SCOTUS is simply doing what Americans wanted anyway.
What comes after affirmative action?
The question, then, is what comes after affirmative action is gone. In a 2018 Bloomberg post, I suggested replacing the current system with a combination of class-based preferences and racial preferences restricted only to Black and Native Americans, reasoning that those two groups have been subject to uniquely severe exclusion in America. Whatever you think of that reasoning, it appears that SCOTUS is not going to allow that kind of needle-threading. Thus, it looks as if the likeliest replacement for affirmative action is even simpler: class-based admissions preferences.
The idea of affirmative action based on familial income and wealth rather than race has been around a long time, and it has many attractive qualities. It elegantly solves the problem of changing group disadvantage over time, as well as the disaggregation problem for broad groups like “Hispanic” and “Asian” — not to mention mixed-race folks. It will probably strike a lot of Americans as fair as well — the WaPo poll found that Americans think poor kids as a whole are more disadvantaged than any racial group in particular. Other polls find strong support for class-based rather than race-based affirmative action. Class-based affirmative action is strongly in keeping with Americans’ traditional support for the ideal of equality of opportunity.
And of course, class-based admissions preference involve the elimination of legacy preferences, which are despised by strong majorities of every racial group. I would also curb athletic preferences as well, given how easy it is to game these with rich-kid sports like sailing.
If paired with policies to admit the top 10% of high school graduating classes (such as the one used in Texas), class-based affirmative action might even increase the number of Black students at top colleges. At least, that’s the prediction of the Century Foundation:
That’s not a certainty, of course. When the University of California was forced to drop racial preferences, class-based alternatives didn’t restore the status quo:
Despite adopting aggressively class-based, race-neutral admissions policies to stay in line with Prop 209, the state's flagship universities have not been able to return black student enrollment to their pre-1996 levels. Today, California's university system is educating an increasingly economically diverse but racially isolated crop of students. Roughly one-third of University of California undergrads qualify for federal Pell Grants, but in 2012-2013, black students made up just 2.4% of UC-Berkeley's enrolled undergrads. African-Americans, meanwhile, make up about 6.6% of the state's population. The economic workaround is not working alone.
Some argue that adding wealth to measures of SES, in addition to income, will make things better for Black applicants, but that remains to be seen. If you believe that Black Americans are uniquely disadvantaged in America, there is some reason to worry that class-based preferences won’t be enough. We should do more to make help Black kids at the k-12 level.
The real problem, though, is that tweaking the formula for who gets a handicap in the college admissions game doesn’t actually change the game itself. Admission to top colleges in America will still be zero-sum. And as the U.S. population grows, and top university student bodies stay roughly constant, the game will get steadily more desperate, high-stakes, and vicious. When I got into Stanford, the admissions rate was around 16%; now, it’s more like 4%.
If these schools are providing quality educations instead of just prestige and signaling value, then why don’t they scale up? Why force the nation’s most talented kids to fight tooth and nail over a tiny chance of admissions when we could just…educate more kids? Obviously you can’t scale up infinitely — not everyone is prepared to handle education at a selective college, and eventually each campus will get overcrowded. But if our higher ed system worked two decades ago when one out of six kids could get into Stanford, then it would work just fine if one out of six kids could get in today. When Harvard’s Twitter account bragged about how many talented kids the school rejects, it should have been read as an admission of inadequacy rather than a sign of prestige:
Signaling and prestige are not valuable to our nation as a whole; human capital is. And the more we expand opportunity at top universities, the less focused our society will be on fighting each other in a bitter racial and class struggle over who gets in. Fairness is an importance goal, but abundance is even more important.
Update: As expected, the Supreme Court struck down affirmative action. Jay Caspian Kang has an excellent post about the decision, excoriating the Court’s liberal justices for ignoring Asian Americans in their dissent, and lamenting how affirmative action had become divorced from its original intended purpose:
Asian Americans, the group whom the suit was supposedly about, have been oddly absent from the conversations that have followed the ruling. The repetitiveness of the affirmative-action debate has come about, in large part, because both the courts and the media have mostly ignored the Asian American plaintiffs and chosen, instead, to relitigate the same arguments about merit, white supremacy, and privilege…
The end of affirmative action really started in 1978…[when] a divided Court [argued] that the race of a candidate could be considered, but not as part of a reparative, quota-based program that tried to reduce the harms of slavery and injustice...It is easy and perhaps virtuous to defend the reparative version of affirmative action; it is harder to defend the system as it has actually been used…
[The liberal justices’] opinions betray the corruption of affirmative action’s original righteous, reparative promise, and the way in which a program that was designed for a racially binary America never got meaningfully updated for today’s multiracial democracy.
I agree 100%.