What About Legacy Mismatches?

2012 stats for undergrads at UC Berkeley, which under CA law doesn’t allow affirmative action

A reader of Asian descent makes a key distinction:

I strongly oppose the notion that Asian Americans should view affirmative action as being a competitive process only between minorities. Specifically, I am writing in to partially disagree with the Asian American reader who wrote:


When we talk at the group level, AA [affirmative action] is about “blacks getting the same advantage whites always had,” but at an individual level, it means smart Asian kids getting shut out in favor of black or other underrepresented minority kids.


I don’t think the statistics warrant his/her notion that Asians are being rejected for only underrepresented minorities. I think the unspoken quotas currently in place at Ivies are for protecting the Caucasian ratio. If you look at Cal-tech and the UC systems that have done away with affirmative action, it would seem to validate this view.

That seems to be true. Take UC-Berkeley: Their diversity statistics show 3 percent Black, 49 percent Asian, and 29 percent White. Harvard, where affirmative action is allowed, had a record high last year of 12 percent Black—a figure school’s website promotes alongside its 21 percent Asian figure … but it doesn’t provide a White percentage (though that figure appears to be roughly 50 percent when Harvard’s four non-White categories are subtracted from the whole). So if more Asians were allowed into Harvard, they would likely cut into that 50 percent, not the 12 percent Black or 13 percent Hispanic—figures that affirmative action was created to maintain.

This reader proposes a solution to prevent direct competition between underrepresented minorities:

The quotas and “negative affirmative action” practiced upon Asians (i.e. they must perform at a higher level than whites to have equal chances of admissions) can be ended without ending the “positive affirmative action” that exists for blacks (and Latinos, Filipinos, Cambodians, etc.). Admissions officers could simply place applicants into either a race-blind or race-conscious pool depending on the representation of that applicant’s race at the university relative to the overall population of the country (easily available from census data).

Overrepresented whites, Chinese, Indians, etc. would be in the race blind pool, while underrepresented blacks, Latinos, Vietnamese, etc. would be in the race-conscious pool (until they too are “overrepresented” in the university, after which point they would be placed in the race-blind pool).  In this way, positive affirmative action can continue, while quotas for Asians will be forced to end.  Furthermore, Asians will no longer be usable as rhetorical weaponry by the political right, as it could no longer be claimed that Asians are disadvantaged relative to whites by affirmative action.


The drop in black admissions and rise in Asian admissions at schools that ban affirmative action are due to two distinct factors: the ending of negative action for Asians and the ending of positive action for blacks. You can end the former without ending the latter, despite what the political left might believe.

From a “white/Asian” opponent of Asian quotas who poses some “practical questions”:

How should school admissions decide who is Asian? Should it be based purely on self-identification, or should we institute compulsory genetic testing or perhaps an “eyeball” test, just in case some try to pass as white? If an Asian person identifies as white in order to avoid discrimination, would that be morally wrong? In the case of a biracial (Black-Asian) candidate, should they be discriminated against? Or should we put them at the very top of our list?

That email reminds me of the following passage from Malcolm Gladwell’s “Getting In,” a superb essay on the history of college admissions:

In the nineteen-twenties, when Harvard tried to figure out how many Jews they had on campus, the admissions office scoured student records and assigned each suspected Jew the designation j1 (for someone who was “conclusively Jewish”), j2 (where the “preponderance of evidence” pointed to Jewishness), or j3 (where Jewishness was a “possibility”).

An Asian American critic of AA looks ahead to a lawsuit:

Now that the model minority myth is of no political use to the liberal elites who actually want to promote affirmative action, Asian American “over representation” is now considered a problem rather than a promise. We are now an invasive species that only affirmative action can help manage.


(Ironically, it is a discriminatory immigration policy that fuels the overrepresentation: Asians with highly developed skills or a promise to invest capital to hire U.S. citizens are pushed to the front of the line. We do well because it takes doing well just for us to get here.)


So it’s no wonder that Edward Blum, of the right-wing American Enterprise Institute (AEI), has seemingly taken Asian American issues to heart and has filed lawsuits on our behalf to halt or defeat the race-tinkering of the liberal elite. [CB note: Blum also encouraged Abigail Fisher to sue the University of Texas, resulting in the current Fisher case before the Court, and he’s also behind the current Evenwel v. Abbott, which, according to Garrett Epps in a recent piece for us, “would limit representation to eligible voters—favoring wealthier, whiter, and more conservative citizens.”] Asian American students are understandably apprehensive about his intentions, even if he is advocating for their immediate interests.

From a Reuters profile of Blum:

“It took a Jewish white guy to start this case,” said James Chen, who runs a San Francisco-based company called Asian Advantage College Consulting. The firm advises students to underplay their ancestry as one strategy for maximizing their chances of getting into a top school. Among some recent immigrants he has worked with, Chen said, “there is a tendency to keep their head down.” […]


Because Harvard reveals few details about its admissions process, Blum’s lawsuit relies mostly on data provided by outsiders. One example is a 2009 report co-written by Princeton University sociology professor Thomas Espenshade. The study contends that Asian Americans needed SAT scores 140 points higher than whites, all other variables being equal, to get into elite schools. Blum also cited research by professors he used in the Fisher case. In its court filing, Harvard denied that it discriminates and said it lacked information about the studies to specifically respond to them.

As always, hit up hello@theatlantic.com to join the debate. Update from a reader:

Some readers have previously suggested that “Asian Americans should [not] view affirmative action as being a competitive process only between minorities” and that “Admissions officers could simply place applicants into either a race-blind or race-conscious pool depending on the representation of that applicant’s race at the university relative to the overall population of the country (easily available from census data).”


This is a wonderful idea, but it is nearly impossible given the legal realities of affirmative action in America. A university’s ability to employ affirmative action at all is heavily predicated on its ability to justify how such policies are consistent with furthering its educational mission. Beyond general hand-waving about the value of diversity, this requires universities to espouse some metric for determining when they have “enough” diversity. After all, without such a metric, who’s to say they aren’t already diverse enough? Not the university, certainly; this was the central holding of Fisher I. And not only must they have such a metric, it must also be consistent and non-arbitrary, lest they open themselves up to attack from the likes of Fisher.


Most universities solve this problem by loosely employing a “proportion of the population” approach, wherein representation at the university should model representation in the population at large, which works very well for underrepresented minorities. But, under such a framework, overrepresented minorities necessarily get the short end of the proverbial stick, because their over-representation means that some other group must be underrepresented and is therefore a problem.


This is not the end of the story, though. Universities could adopt a different metric of representation. If they did so, overrepresented minorities could be fine. But to adopt a different metric would mean adopting that metric across the board: applying it only to some groups but not others would be deeply constitutionally suspicious. And this, in turn, could potentially harm underrepresented minorities, who just lost a very defensible argument for increased representation.


And operating in orbit around all of this is the ever-present threat of litigious anti-affirmative action groups who relish any opportunity to haul universities into court. This puts universities between a rock and a hard place, because they have to find some consistent, legally defensible way to argue that under-representation is both harmful (in the case of blacks, etc…) AND beneficial (in the case of whites) to their educational mission. I won’t go so far as to say it’s an intractable problem, but it’s certainly hard to find a happy ending for both over- and under-represented minorities.


The path of least resistance appears to be the proportion of the population approach, with the ensuing harm to over-represented minorities, at least until such time as it is ruled unconstitutional.