On August 10, only a few days before the white supremacist murder in Charlottesville upended the national discussion of anything race-related, The New Yorker published an intriguing essay on affirmative action by Harvard law professor Jeannie Suk Gersen. The piece, entitled “The Uncomfortable Truth About Affirmative Action and Asian Americans,” was a response to reports from the previous week that the Department of Justice’s Civil Rights Division was looking into “investigations and possible litigation related to intentional race-based discrimination in college and university admissions.” Contrary to initial speculations, the DOJ would not be targeting affirmative action policies that allegedly harmed whites, but reviving a complaint against Harvard for discriminating against Asian applicants.

Suk’s “uncomfortable truth” is that the complaint is probably right as far as it goes. Asians and Asian Americans are almost certainly disadvantaged in elite college admissions and possibly subject to illegal quotas. They must have substantially higher standardized test scores than other groups to gain admission, and their share of the Ivy League student population has remained stagnant since the early 1990s—a period over which their population has increased at four times the rate of the U.S. population as a whole. If elite schools’ holistic admissions practices do not involve Asian quotas, then their enrollment figures are explicable, as Suk notes, only if “Asian applicants are severely less likely than white ones to have the special personal qualities that colleges seek.”

This point by itself should not be particularly controversial, but the dynamics of the affirmative action debate has led defenders of racial preferences to try to obfuscate it. A representative Vox essay on the subject studiously avoids the question of whether Asians are disadvantaged in college admissions while offering a variety of non-sequiturs meant to remind them that they are on Team POC. It notes, for instance, that Asian groups like Laotians and the Hmong are underprivileged; that Asians are underpaid and under-promoted relative to their education; that conservatives have used the “model minority myth” to praise Asians as a “foil” to blacks and Latinos; and that some white students have come up with racist names for heavily Asian colleges. But Laotians and the Hmong account for less than five percent of the Asian American population, Asians may not be underpaid if one looks only at those born in the United States (and if they were they would need a fair shake in admissions anyway), and the other two claims are simply irrelevant to the matter at hand.

Although substantively weak, such arguments recur time and again thanks to the partisan contours of the debate (and, perhaps, thanks to a folk belief that any program benefitting one minority must as a matter of logical necessity benefit every other). Suk rejects such evasions. Whatever elite colleges say they are doing with their admissions policies, she is clear that they are in fact engaged in racial balancing—that is, managing their demographic makeup such that they are not “swamped by members of any particular race.” And in practice they are most worried about being swamped by Asian students. Although taboo and possibly illegal, Suk argues that such balancing is realistically “unavoidable” and may in fact be desirable, as “we should not want the composition of our élite universities to be wildly out of proportion to the racial composition of our country.”

Suk is probably right that balancing of some form is warranted. It is hazardous to build an elite that looks nothing like the country it presides over. But openly admitting to penalizing Asians in the college admissions process threatens not only to doom affirmative action in the view of the courts; it also seems to subvert its moral logic, which rests on the idea that America’s white majority owes a leg up to groups it oppressed in the past. Affirmative action becomes more difficult to justify if those most subject to reverse discrimination are members of a minority group that was not responsible for those historical injustices—and indeed was itself subject to many of them.

This dilemma leads Suk to a proposal that is dangerous precisely because it is so compelling. She agrees it is unjust to discriminate against Asians, yet wishes to maintain affirmative action programs benefitting blacks and Latinos; to square this circle, she separates the two policies, suggesting we jettison the former while maintaining the latter. After all, “The problem is not race-conscious holistic review; rather, it is the added, sub-rosa deployment of racial balancing in a manner that keeps the number of Asians so artificially low relative to whites who are less strong on academic measures.” What is needed is thus “race-conscious affirmative action” to “address the historic discrimination and underrepresentation of blacks and Latinos, in combination with far less severity in the favoring of whites relative to Asians.” That is to say, keep the current preferences for blacks and Latinos while eliminating the penalty for Asian Americans.

This is a deeply elegant solution for liberals. It defangs one of the most commonsensical objections to race-conscious admissions—that they unfairly hurt a minority group—while neutralizing any conservative attempt to offer Asians a quid pro quo on affirmative action. It would also remove a source of tension between Asians and other minority groups that tend to vote Democratic. Although some polls have suggested a majority of Asian Americans support affirmative action, these rely on framing that dodges the issue of anti-Asian penalties. When affirmative action has come up in the California legislature, however, Asian-American Democrats have split with blacks and Hispanics and voted with Republicans to defend race-blind admissions.

But what would Suk’s proposal look like in practice? Harvard’s incoming freshman class is 49.1 percent white and 22.2 percent Asian; for the United States as a whole, those numbers are 63.7 percent and 4.7 percent, respectively. What does Harvard look like without white-Asian “balancing,” assuming that both groups continue to compete for the same roughly 70 percent of slots? 40 percent white and 30 percent Asian? 35 and 35? 30 and 40? At CalTech, which admits mostly based on test scores, the undergraduate population is 29 percent white and 41 percent Asian; at UC Berkeley, which is prohibited by state law from using race-conscious admissions practices, the numbers are almost identical. Granted, both of these schools are in California, where Asians make up 15 percent of the population. But if one assumes an even split at 35-35, then whites’ share at the country’s top school would fall to just over half of their share in the total population; Asians would be overrepresented by around a factor of seven. Crucially, from whites’ perspective, this would be the result not of an academic horse race but of an inconsistent selection process where balancing is applied only when doing so benefited non-whites.

One could certainly argue for such a scenario on moral or ideological grounds. Whites have dominated the American elite since its founding. A handicap in elite college admissions is by most standards a small price to pay for helping to level the playing field. Yet I believe Suk was right in her original suggestion that for reasons of social stability, colleges should be wary of allowing their demographics to diverge too far from those of the country as a whole. The problem is that keeping traditional race-based affirmative action while scrapping Asian balancing might do exactly that.

Morals aside, the case for ethnic balancing is essentially a pragmatic one: in a diverse country, it is important for different groups to buy in to the system. This is probably the best justification for the current arrangement, in which access to the upper echelons of American society—as mediated through elite higher education—is granted on the basis of an obscure calculus combining individual merit, family circumstances, and de facto recognition that the country’s racial and cultural groups need to see themselves reflected in the corridors of power. This may be unfair to individuals, but it is better than a nominally meritocratic alternative in which some groups monopolize access to the elite while others are shut out of it altogether.

Such disproportion is politically dangerous enough when a group shut out of power is a minority—in order to ensure stability, for instance, Singapore works hard to ensure that ethnic Malays, at around 13 percent of the population, feel represented in the political system. The problem is far worse when the out-group is a majority, and is especially delicate when that majority perceives it is being shut out of important corridors of power by a culturally or ethnically alien elite. As Amy Chua argued in World on Fire, such arrangements—in which a country’s elite is ethnically distinct from the majority of the population as a whole—can be highly unstable. The elite may feel little obligation toward the general population and is often forced to rule through undemocratic means. Members of the ethnic majority, in turn, becomes susceptible to blood-and-soil demagoguery that promises to give them back the fruits of “their” land. The turn toward majoritarian chauvinism is, as Chua argues, not a unique product of any one country’s history, but a pattern that recurs around the world—in Latin America, Malaysia, Myanmar, the Philippines, and Russia.

American politics is already beginning to see the effects of whites becoming more conscious of themselves as a declining majority; and though many whites accept this as a positive or merely neutral development, many view it with fear or anxiety. One of the great political challenges of the coming decades will be to soothe the latter group’s anxieties such that the country doesn’t wind up with a significant plurality of whites whose basic desire is to smash a system they mistrust and resent. Although elite college admissions are only a tiny part of this larger puzzle, they go a long way in determining the makeup of the American elite, and their policies are important as signals of how that elite thinks and what it values. If purely meritocratic admissions would create an ethnic makeup we are unwilling to accept, then the current policy of balancing may be the best of bad options.