Harvard Has a Choice on Diversity — and It’s Not About Race

Stuart Bradford for The Chronicle

How should educators think about affirmative action in the age of Donald Trump? On the one hand, President Trump’s polarizing rhetoric underlines why it is so important to provide educational settings that bring students of different backgrounds together to learn to understand and respect one another. On the other hand, Trump’s rise suggests that the old ways of creating diversity — using large racial preferences that tend to benefit wealthier students of color — come at a considerable price for the nation. Research has long found that the mere mention of “affirmative action” triggers negative racial attitudes. Such policies can feed the idea that white working-class people have been “forgotten.”

At the recent trial examining Harvard’s use of race in admissions, one of the important questions raised was whether there is a better, less divisive way of creating diversity. What would happen if colleges adopted a more robust vision of diversity that resulted in inclusion by both race and socioeconomic status? While the dueling placards of demonstrators suggested the battle was between those defending diversity and those advocating race-neutral admissions, could race-neutral strategies be structured in a way that recognized the impact of discrimination? In my personal capacity, I served as an expert witness for Students for Fair Admissions, the plaintiffs in the case, I concluded from the evidence that there are new, better paths to diversity.

Advocates of racial preferences suggested that diversity would plunge if Harvard stopped using race — and if nothing more was done, that would be true. Mark G. Yudof and Rachel F. Moran, writing in The Chronicle, cited a 2013 Harvard Office of Institutional Research study that found that if grades and test scores were the sole basis of admissions, African-American admissions would plummet from 10 percent to 1 percent and Hispanic shares would drop from 9 percent to 2 percent.

Harvard’s expert witness, David Card of the University of California at Berkeley, likewise projected that if Harvard dropped explicit racial preferences and kept its holistic admission process that rates students on academics, extracurricular activities, athletics, and personal characteristics, black admissions in the class of 2019 would drop from 14 percent to 6 percent, and Hispanic admissions would decline from 14 percent to 9 percent.

But what would happen if Harvard took additional race-neutral steps to make the admissions process fairer? Using data from 160,000 Harvard applicants over a six-year period, Students for Fair Admissions modeled the effects of various changes in the admissions process. We began by eliminating the various preferences Harvard currently employs that disproportionately benefit white and wealthy students, such as being a legacy candidate, a child of faculty or staff, or being on a special “dean’s interest” list. We then simulated what would happen if Harvard gave a preference to socioeconomically disadvantaged students of a magnitude that is about half the size of the preference Harvard currently gives to recruited athletes.

The result? Student diversity levels — racial/ethnic and socioeconomic — would rise. For the class of 2019, the share of first-generation college students would increase from 7 percent to 25 percent. This presumably would make classroom conversations more interesting in an institution that Raj Chetty, then at Stanford University, found in a 2017 study has had 23 times as many rich students as poor students. Under the simulation, the shares of underrepresented minorities would also increase from 28 percent currently, to 30 percent. Hispanic student shares would double from 10 percent to 20 percent. African-American shares would drop from 14 percent to 10 percent — a level that was common at Harvard for many years, including as recently as the admitted class of 2016.

And the truth is that Harvard could do even better, if it would consider the wealth or net worth of applicants’ families. Harvard’s failure to do so now effectively penalizes African-American applicants in the aggregate. Professor Dalton Conley, of Princeton, has found that a family’s wealth better reflects the nation’s legacy of slavery and segregation than factors like income because wealth is handed down from generation to generation. African-Americans typically have incomes that are 60 percent of white incomes, but African-American median wealth is just 10 percent of white wealth.

Wealth is important to a student’s life chances, Conley explains, because “educational advantages are acquired through major capital investments and decisions,” such as purchasing a home in a neighborhood with good public schools. Adding wealth data into the simulation would be the fair thing to do and would boost African-American shares.

Notably, Harvard could produce strong levels of racial, ethnic, and socioeconomic diversity without compromising academic excellence. The simulated class would have SAT scores at the 98th percentile and high-school grades as high as the current class — even though many more of the students would have had to overcome economic obstacles. And with its $39-billion endowment, Harvard could easily increase the amount of financial-aid dollars it would need to give to the growing number of socioeconomically disadvantaged applicants admitted under such a system.

Some people argue that even if socioeconomic preferences can achieve racial diversity, race should nevertheless be considered as a factor to signal that society believes that race continues to matter. But consideration of socioeconomic indicators does not suggest race is irrelevant; instead, socioeconomic preferences produce racial diversity precisely because race matters.

Martin Luther King Jr. clearly recognized this when he suggested that America needed affirmative action as a remedy for its history of discrimination in the form of a Bill of Rights for the Disadvantaged. The program would disproportionately benefit black victims of discrimination, King said, but would also benefit poor whites as “a simple matter of justice.” That inclusive idea of affirmative action is a far cry from Harvard’s current system.

If Harvard ultimately loses its case, its admirable commitment to racial diversity will almost surely cause it to transition to a more robust vision of affirmative action that results in not only substantial racial diversity but socioeconomic diversity as well. Doing so would move America’s oldest college away from its current lopsided economic profile, in which there are more students from the top 10 percent of the nation’s income distribution than from the bottom 90 percent. Along the way, Harvard would recognize more fully the rich talent that researchers have found exists throughout the socioeconomic and racial distribution of the country.

Richard D. Kahlenberg is a senior fellow at the Century Foundation and author of The Remedy: Class, Race, and Affirmative Action (Basic Books, 1996). The opinions expressed in this essay are his own and not necessarily those of the foundation.