Wednesday, May 09, 2012

The Uncomfortable Question of Who Counts as a Member of a Minority Group

By Mike Dorf


My latest Verdict column wades into the controversy over Elizabeth Warren's past claim to be part Native American. (For a useful summary of the blawg coverage of this issue, click here.)  My conclusion?  That the issue is good for the Brown campaign because affirmative action is a useful wedge to divide minority voters from working class and middle class white voters.  In the course of  explaining how the controversy is really about affirmative action, I mention some common arguments both for and against affirmative action.  One of the arguments against is that affirmative action utilizes a two-wrongs-make-a-right logic.  Here I want to expand on that theme and also subject it to a little critical analysis.

The Warren controversy highlights the difficulty of deciding whether someone "counts" for purposes of entitling that person to whatever bonus he or she gets for being a member of the relevant group.  And that determination can be uncomfortable.  Even Supreme Court Justices now regarded as liberal or moderate have at one point taken note of the difficulty.  In his dissent in the Fullilove case, Justice Stevens compared the method for determining eligibility for affirmative action benefits to Nazi race laws.  Justice Kennedy quoted that language in his dissent in the 1990 Metro Broadcasting case, adding in a comparison to South African apartheid laws.

Yet it's relatively easy to distinguish these examples based on the uses to which the classifications were put.  Nazi race laws were not appalling because they identified who was a Jew.  They were appalling because, for the Nazis, identifying Jews was the first step to murdering Jews.  Likewise, apartheid South Africa obsessed about degrees of whiteness, blackness, and colored-ness because of the legal privileges and disabilities that went with those statuses.  Once Justice Stevens had migrated from the center-right of the Court to its left, he saw this important distinction, stating in his 1995 dissent in the Adarand case that such comparisons "disregard the difference between a 'No Trespassing' sign and a welcome mat."  No rational person thinks that the classifications appearing in programs of affirmative action in higher education are a prelude to genocide or apartheid.

But even if the Nazi and South African comparisons are unfair, there remains something inherently troublesome about the government (or powerful private actors) trying to figure out whether someone is Black enough, Latino enough, or Native enough to qualify for an affirmative action program.  Why?  I think it's because doing so reminds us of why we object to racial and ethnic discrimination in the first place: It renders critical what should be an irrelevant criterion.  If we're not even sure whether someone is African American, Latino, or Native American, then why, we find ourselves asking, should we be distributing various benefits based on that status?

There are answers to that question, of course, but I think that even most on-net supporters of affirmative action (like myself) are at least somewhat uncomfortable with the "who counts" inquiry.  For one thing, it tends to pull apart different justifications for affirmative action.  If we are interested in intellectual diversity, then someone who strongly identifies with a particular group should "count" more than someone who doesn't, even if the strong identifier is less connected to the group as a matter of genetics.  By contrast, if we are concerned about role models or in combating stereotypes, then it matters whether someone will be clearly identifiable as a member of the relevant group.

Perhaps the "who counts" question is more or less problematic depending on the use to which it is put.  Even opponents of affirmative action will have to engage in some determinations of "who counts," at least for the purpose of detecting old-fashioned discrimination.  Suppose that a white plaintiff sues his employer claiming that he was passed up for a promotion in favor of a less qualified African American candidate.  In order to make out a prima facie case, the plaintiff will need to establish his own race and the race of the candidate who received the promotion.  That will not always be obvious.

But unless one wants to do away with antidiscrimination law itself, it is hard to see how one can avoid such "who counts" inquiries.  (One might say that it doesn't matter what the promotion applicants' races were but whether the employer perceived them as belonging to different races, and acted on that basis, but as a practical matter, what the employer perceived will usually have to be proved by reference to the applicants' respective actual races, and so we'll be back to the "who counts" inquiry.)

Certainly it would go much too far to say that we ought to give up antidiscrimination law entirely because it requires certain classifications that are, in certain respects, reminiscent of classifications used by Nazi Germany and apartheid South Africa for wholly nefarious purposes.  In the end, the discomfort that most of us feel when confronted with borderline cases--whether for affirmative action purposes or simply for antidiscrimination purposes--may just be part of the price we need to pay to build a more equal society.

12 comments:

AF said...

I agree that the "who counts" issue is one of the most difficult issues for affirmative action supporters. However, I disagree that the same issue arises for purposes of detecting "old-fashioned discrimination," if by that you mean disparate treatment.

As you acknowledge, for purposes of discrimination the issue is the alleged discriminator's intent, which means it their perception of someone's race, not the person's actual race, that is relevant. You say that as a practical matter the question of perceived race will boil down to the "who counts" inquiry, but I don't think that's right. In the rare case where there is a dispute about someone's race in a discrimination case, there will need to be a fact-intensive inquiry as to the defendant's sources of knowledge and perception: did the person self-identify as a particular race? to whom? did the company collect data on race? if there was no self-identification, did the relevant decision-makers know what the relevant individual looked like? what did the relevant individual look like? This only partially overlaps, at most, with the "who counts" inquiry. Genetics, for example, will be almost entirely irrelevant.

More importantly, even to the extent that the factual issues in a discrimination case overlap with those in the "who counts" inquiry (eg, evidence as to the individual's appearance), they are being put to a different purpose: determining the defendant's intention, rather than the individual's "actual" race. Thus, the truly uncomfortable aspect of the "who counts" inquiry -- as you put it, the "government (or powerful private actors) trying to figure out whether someone is Black enough, Latino enough, or Native enough" -- is not in play. The decision-maker does not need to figure out whether someone "is" a particular race, only whether the defendant thought so.

Michael C. Dorf said...

AF: Thanks for the comment. It's a fair point but:

1) Title VII literally forbids discrimination "because of . . . race . . . ." It's true, as I acknowledge, that judicial decisions expand that to cover perceptions of race, but the core prohibition is on race discrimination, not discrimination on the basis of perceived race;

2) As you acknowledge implicitly, a disparate impact case will look at actual race--and there "who counts" inquiries will occur too. Again, it would be theoretically possible to measure disparate impact based on perception of race, but that seems practically impossible for a case against a large enterprise.

Now maybe this is a reason to be skeptical of disparate impact liability in the first place, but if one thinks--as I do--that disparate impact is often the only realistic way to get at intentional discrimination in a large organization, that's a very high price.

AF said...

I agree that disparate impact can theoretically raise the same boundary problems as affirmative action.

I still don't think they have any practical relevance to disparate treatment, because in practice the question of intent to discriminate based on race depends on the defendant's perception of race. Consider two hypothetical cases. First, the plaintiff can prove that she "is" black, but the defendant can prove that she didn't know that. There is no liability because there is no intent to discriminate. Second, the plaintiff "is not" black but the defendant discriminated against her because of a perception that she was. I suppose it is theoretically arguable that this would not be discrimination, but I am not aware of any court that has come to that conclusion, or even of any defendant that has made that argument. If there are such cases, I stand corrected.

Stephen R. Diamond said...
This comment has been removed by the author.
Stephen R. Diamond said...

The "who counts" issue is necessarily both raised and resolved by the requirement that the protected group be a _discrete_ suspect class. (Why this is a good requirement I discuss at http://tinyurl.com/c6lw4tg)

In this light, the definition of the protected class _must_ correspond to the definition based on which it is subject to persecution--and such a definition must exist. Racism, being a doctrine of purity, has minimal requirements for animus.

Unknown said...

More importantly, even to the extent that the factual issues in a discrimination case overlap with those in the "who counts" inquiry (eg, evidence as to the individual's appearance), they are being put to a different purpose: determining the defendant's intention, rather than the individual's "actual" race. Thus, the truly uncomfortable aspect of the "who counts" inquiry -- as you put it, the "government (or powerful private actors) trying to figure out whether someone is Black enough, Latino enough, or Native enough" WOW Gold-- is not in play. The decision-maker does not need to figure out whether someone "is" a particular race, only whether the defendant thought so.

Unknown said...

Now maybe this is a reason to become skeptical of disparate impact liability initially, when one thinks--as I do--that disparate impact is usually the only realistic way of WOW Gold Buy WOW Gold Cheap WOW Gold getting at intentional discrimination in a very large organization, that's a very high price.

Sam Rickless said...

Perhaps as a side note...

You mention three possible justifications for affirmative action: promotion of intellectual diversity, creation of role models, and combating stereotypes. All of these are forward-looking, consequentialist justifications. I think the strongest argument for affirmative action is that it is required by the principle of equal opportunity. The claim is that individuals who are perceived as belonging to such-and-such a group suffered discrimination in the past and suffer discrimination now. By the principle of equality of opportunity, members of such a group should be compensated for the disadvantages resulting from unjust discrimination. Here there IS a potential lack of connection between self-identification as a member of a group and identification by others as a member of that group. With respect to the administration of AA programs, this will be a problem with respect to some groups, but not with respect to others.

Tenantlaw said...

Isn't the basic justification for affirmative action that it prevents ongoing discrimination? Most employment and admissions decisions are subjective, even though affirmative action opponents want to dress them up as objective and neutral. As long as there's an imbalance in power, affirmative action prevents those making subjective hiring and admissions decisions from engaging in discrimination. Simple as that.

toryburchoutlet said...

Buy Windows 7 Key
associates of such a team should be paid for for the negatives as a result of unfair elegance. Here there IS Buy Cheap Windows 7 ultimate Key a prospective deficit of relationship between self-identification as a participant of a team and recognition by others as a participant of that team. With regards to the management of AA applications, this will be a issue with regard to some categories, but not with regard to others. Buy Windows 7 ultimate activation Key

Jet Li said...

The perfect!These articles written too great,they rich contents and data accurately.they are help to me.I expect to see your new share.
-----------------
http://www.kavkey4u.com/

jet said...

More significantly, even to the level that the informative problems in a elegance situation overlap with those in the "who counts" questions (eg, proof as to the person's appearance), they are being put to a different purpose from http://www.mmomesos.com/: identifying the defendant's objective, rather than the person's "actual" competition. Thus, the truly unpleasant element of the "who counts" questions -- as you put it, the "government (or highly effective personal actors) trying to determine whether someone is Dark enough, Latino enough, or Local enough" -- is not in perform. The decision-maker does not need to determine whether someone "is" a particular competition, only whether the accused believed so.