Americans’ views about affirmative action are notoriously slippery. To give you some idea of how slippery they can be, consider two polls:

  • A February 2019 Gallup Poll determined that over 60% of all Americans favor affirmative action policies in education for both women and minorities, a pronounced increase in favorable responses from the previous 2016 Gallup poll.
  • A 2019 Pew Poll determined that about three-quarters of all Americans oppose affirmative action in education based on race or ethnicity and only 7% believe it should be a major consideration in college admissions.

Why These Polls are Irrelevant

What these poll results reveal isn’t so much a sampling of informed opinion as a demonstration of how the various arguments for and against affirmative action in education, some well-intended, some subtly partisan and others wildly slanted for or against, have left most Americans confused and, too often, resentful.

The resentment isn’t confined, as you might think, to minorities. A wide swath of white Americans, 55%, feel discriminated against. According to a 2019 Robert Wood Johnson poll, 11% of these white Americans believe that when they applied for college, they were discriminated against on the basis of race.

The Problem With Fact-Free Opinionating

One source of the confusion surrounding affirmative action policies in education is the way polling has evolved since it was first introduced in the 1936 U.S. Presidential election.

In an effort to avoid intervening in ways that may alter respondent opinion, almost all polls avoid carefully or extensively defining the terms used in their questions. In this way, the reasoning goes, respondents’ current states of mind, which also reflect their understanding of the terms used, are captured without distortion.

This, it would seem, is an unexceptionable process. But here’s the problem, which I’ll introduce by way of a question: “How many other lives are you simultaneously living currently, including those in other galaxies?” Posed this way, without giving you the string theory perspective (which you may or may not accept anyway) that poses the existence of infinite universes like ours, with your identical selves alive in all of them, the question is nearly meaningless and nonsensical.

Even after hearing the theory explained by a competent physicist, you may well disbelieve the idea of multiple versions of yourself because it is so far from what you think of as “common sense.” Your own experience tells you this can’t possibly be true.

This is the problem pollsters run into when they ask questions about “affirmative action,” but especially when they avoid any nominally neutral definition of what the “affirmative action” is they’re asking questions about. We are still a highly segregated society and our experiences differ accordingly.

Americans don’t have a single agreed-upon definition of affirmative action. What the term means to an African-American, a Latinx, an Asian, a white male and a white woman will differ. And each of us, depending on our ethnicity, gender, education, economic class and a few more variables, will have personal experiences that are completely unknown to other Americans, who may determine that the experience described by another person contradicts their own experience so fundamentally as to be untrue.

In this circumstance, asking poll respondents to answer questions about affirmative action without describing or defining what they are being asked to give opinions on will almost inevitably lead to the directly contradictory results described above.

What We Know and What We Don’t

It would be presumptuous to attempt to straighten out the various misunderstandings (and consequent mistrusts) related to affirmative action in this short article. Here I’d only like to suggest how the differences of opinion arise in just one subject area: equality of opportunity. The misunderstandings surrounding this one term are several, but they’re finite.

The conservative argument against affirmative action often centers on the belief that while all Americans do not have precisely the same educational opportunities, we do at least all have a fair shot.

A clear instance of this from the conservative view, is the practice of using SAT scores as an important determinant of a prospective student’s fitness for college. These scores aren’t the only criterion for admission, but they are often determinative. And black and brown students don’t do nearly as well as white and Asian students.

A normal distribution of scores would find black and brown students with the same percentage of high and low scores. In reality, they’re clustered near the bottom. 60% of the highest scores are by Asian students, 33% by white students, with only 5% by Latinx students and 2% by African-American students.

So, with this evidence, conservatives ask, why should colleges admit numbers of African-American and Latinx students in proportion to their population percentages? Wouldn’t it be a lot fairer to admit students without respect to race, but rather on the basis of their achievement, with the SAT test being one reliable way of determining that achievement?

Another less rigorous approach that nevertheless rejects admissions based on race, is to allow the relative wealth of a student’s family to be taken into account. As Jonathan Chait acknowledges in an article on this subject in The Economist, “Wealth is just distilled opportunity…kids who grow up poor in single-parent homes don’t do well.”

Other conservatives reject this approach as well, because (according to one frequently expressed conservative opinion), poverty “is almost always the consequence of poor decision-making.” Why this justifies refusing to help the children of the poor is not explained.

The argument against both positions — rejecting race as a valid determining criterion for college admission, but allowing poverty or rejecting both criteria — is that this is not the way things work in the real world. A student — never mind race and the marital status of the parents for a moment— who is educated in a school district where schools are underfunded and overcrowded, as in Perlita, Texas, where annual household income averages a little over $16,000 — does not have an educational opportunity that remotely equals that in Scarsdale, New York, with average annual income of about $240,000.

Property taxes, as I pointed out in the first of these four articles on the educational disadvantages suffered by black and brown students, are largely what funds our schools, district by district. The national annual per-pupil average spent on education is under $11,000; in the wealthiest 10 school districts the average is around $20,000. This is not equal opportunity.

Another factor that conservative commentators either overlook or discount without comment when opposing affirmative action is the debilitating effect of white prejudice against black and brown students in education.

The third of these four articles describes the published views of University of Pennsylvania law professor Amy Wax, who is an instance of how this works. Professor Wax has publicly and repeatedly prejudged the capabilities of her minority students, finding them almost all inferior. Not only will Professor Wax’s views almost inevitably incline her to grade black students lower; the transmission of these harmful attitudes to those students negatively affects their self-worth, which, in turn, further contributes to underperformance — the Pygmalion effect.

Is There a Solution?

It would be reassuring to believe that there is some way for Americans to come together on this contentious issue. In the near term, this seems doubtful. Until social integration increases, we will continue to have experiences very different from one another, with those differences profoundly determined by race and class. According to The Washington Post, as recently as 2014 three-quarters of white Americans didn’t have a single non-white friend. Another Washington Post article notes that a similar social separation exists between Americans who are rich and those who are poor. The unpleasant truth is that we are a highly divided society and are unlikely to resolve these divisions anytime soon.

Note: This is the final article in a series of four on equal opportunity in education. Previous articles are Why Black and Brown Students Get Inferior Educations: Segregation; School Funds and Teacher Expectations; and Why Minority Students Get Bad Grades: The Pygmalion Effect.