Disparate Impact Disputation, Continued... - Garrett Moritz is back with more disparate impact defense. Or is it forth? As he points out, we have gone back and forth so many times now that it is hard to keep straight who's on forth and who's on back.
Rather than beginning my response at the beginning let me first turn to the end and both acknowledge and second Moritz's generous comments in his closing paragraph about the substance and the tone of our long-running exchange. (Go read it now.) Such civil and pleasant disagreements are unfortunately rare. Indeed, he is too kind; even for a geezer like me, keeping up with "a whippersnapper" like him is a smaller feat than he imagines.... Now, if I can just force these decrepit old fingers (and the mind to which they are tenuously connected) to slow-dance across the keyboard (boogieing is out of the question), let us go back to the beginning and return to our debate over disparate impact.
Now that you've read Moritz's last paragraph, you need to return and read the preceding paragraphs if you haven't already, for I'm not going to summarize his points here. These posts are already too long, and in any event my summary would no doubt not do justice to his own thoughtful presentation.
One more preliminary observation: my basic argument, put forth to some degree in most of my exchanges with Moritz and elsewhere on DISCRIMINATIONS, is that disparate impact -- theory and practice -- represents a degradation of the civil rights ideal. By draining the concept of discrimination of any requirement for intent, it is of necessity based on the assumption that statistical disparities alone (absent compelling justification) are a violation in need of a remedy and that fairness itself requires something approaching proportional representation.
Seem too strong? Consider the following, from Rick Hertzberg's review of a new book by political theorist Robert Dahl in the New Yorker (link via Stuart Buck, whose comments on Hertzberg/Dahl are devastating):
Imagine, if you can, that African-Americans were represented "fairly" in the Senate. They would then have twelve senators instead of, at present, zero, since black folk make up twelve per cent of the population.
This assumption, and the new conception of fairness that it has spawned, in turn requires that racial preferences be implemented, that group rights take precedence over individual rights, and that the government be engaged in a never-ending process of racial and ethnic regulation to ensure that the numbers in all areas of American life don't harbor "underrepresentation" or "overrepresentation." It requires, in short, abandoning the core American value -- arguably THE core American value -- that every individual be judged "without regard" to race, religion, or national origin.
Now, notice that virtually nothing in Moritz's current long post, or his earlier ones, actually defends disparate impact. Instead, he is at pains to show that such policies are legal, that Congress has a right to implement them, that courts shouldn't interfere, etc., etc. I am definitely not one to disparage lawyers or law -- quite the opposite -- but most of this discussion misses, if not the point, at least my point. For the sake of argument, assume for a moment (but only for a moment) that everything Moritz says is persuasive. That still would tell us nothing about whether disparate impact is a Good Thing. It isn't. Even if Congress or private organizations can impose it, they shouldn't.
Turning, finally, to a few specifics:
1. In my last post I took issue with Moritz for leaning on the weak reed of "legislative acquiescence" to bolster his view that the Supremes should avoid ruling now on the propriety of disparate impact in the Fair Housing Act, but in doing so I noted that "pursuing this controversial theory of 'legislative acquiescence' here would take us too far afield." Like a determined hound chasing stray rabbits into every nook and cranny when he should remain on the trail of bigger game, however, Moritz does pursue it, although he admits in the end that it was "a detour." Nevertheless, as he says, it was "fun" even if diverting. It was a virtuoso performance, and may even be right, whatever that would mean here.
2. Moritz obviously relishes word games, but it's hard to believe he seriously believes that an insistence on some consistency in our understanding of what equality entails -- equality, after all, is a value at the very heart of our Constitutional enterprise -- is comparable to such silly examples of words having different meanings in different contexts as "He's on fire! Get him the ball!" v. "He's on fire! Get a bucket!" Even Sasha Volokh's entertaining example of the changing meanings of "gift" hardly rises to the level of Constitutional gravity.
It is certainly true that words can have different meanings in different contexts, a point that was addressed head on by Justice Powell in his Bakke opinion:
The concept of "discrimination," like the phrase "equal protection of the laws," is susceptible of varying interpretations, for as Mr. Justice Holmes declared, "[a] word is not a crystal, transparent and unchanged, it is the skin of a living thought and may vary greatly in color and content according to the circumstances and the time in which it is used." Towne v. Eisner, 245 U.S. 418, 425 (1918)
The question in Bakke was whether the vision of equality Congress wrote into Title VI of the Civil Rights Act imposed a different standard from what the 14th Amendment required. All agreed that it did not, that Title VI was intended to mean what the 14th meant (though there was disagreement over precisely what that was, with the minority arguing that both required a colorblind standard). Our society can survive different meanings of "fire" in different contexts, and even different meanings of "gift," far more easily than it can conflicting meanings of the requirements of equality -- individual rights here, group rights there; official neutrality here; racial/ethnic/religous preferences there; etc.
Before blowing the whistle on our word games, however, I think it is necessary to observe that Moritz's own apparent understanding of exactly what "disparate impact" means misses the mark. Once some confusing conceptual underbrush is cleared we may be able to narrow somewhat the range of our disagreement. Consider:
[T]here is a fundamental difficulty when it comes to stopping discrimination: some discrimination, the worst kind of discrimination, discriminates without speaking its name. Sometimes an employer fires every black employee and no whites, and yet does not say "I hate blacks, so I fire them"; instead, they say, "I am committed to racial equality; however, I am also a businessman and these particular workers were simply not cost-justified." If that's true, it's a legitimate business purpose and will rebut the prima facie case proved by a showing of disparate impact; for instance, if the employer shows data on employees which proves the point. However, if the employer has no data, or data that shows that the workers were just as productive as anyone else, they'll probably lose.
The trouble here is not with who would win or lose; it is the larger point that this is not a disparate impact case at all. Back to basics: disparate treatment is depriving someone of some right or benefit because of/based on/due to race, and it requires intent. Disparate impact is following a policy or practice that adversely affects minorities disproportionately, without any intent to do so, unless the policy or practice can be justified as reasonable or necessary (the threshold a challenged policy must surpass tends to range from amorphous to elastic).
Moritz's hypothetical employer fired all and only black employees. If he can justify each firing with relevant business or performance data he should win; if he can't he should lose. This is a classic disparate treatment case. Not even the harshest critics of disparate impact argue that "effects" should be ignored. They are nearly always relevant (it's hard to imagine victims of discrimination in the absence of any discriminatory effects); they can raise a prima facie case of discrimination; and in certain circumstances, such as the hypothetical Moritz has in mind, they can even provide sufficient evidence of ... disparate treatment. If the adverse affect on minorities is sufficiently severe and is unpersuasively explained, discrimination can be inferred even in the absence of clear evidence of a discriminatory intent. Moritz seems to believe that if disparate impact were done away with no discrimination complaints would succeed unless evidence of clearly stated intent were found. Not so, but in order to avoid that imaginary fate, disparate impact theory severs the concept of discrimination from intent altogether, and that is going too far, for reasons I will address below.
3. One of the biggest problems with disparate impact is determining what standard a challenged policy or practice must meet. Is it absolute business necessity, such as the business failing without it? Pretty high business necessity, such as less profitability without the challenged policy? Or mere reasonableness, such as it seems like a good idea? Moritz argues "it is not true now, nor was it ever, that disparate impact under civil rights statutes necessarily equals a finding of discrimination. Rather, it serves a burden-shifting function." This is pure fancy, or wishful thinking. On the contrary, burden-shifting is not mere burden-shifting. It can be onerous if the justification standard the challenged policy must meet is too high, as it quite often has been in real cases.
Take the original case of Griggs v. Duke Power Company, where Duke Power's policy of requiring all new hires to have high school diplomas was found to be discriminatory. What would a justification of such a policy look like?
Duke Power's argument was that a diploma was a valid and reasonable proxy for such essential qualities as ability to complete a task, punctuality, responsibility, willingness to work, delayed gratification in the form of foregoing grasshopper pleasures for ant drudgery (I'm putting words in Duke's mouth here, but you get the drift). There was no evidence that this argument was disingenuous or dishonest. Perhaps Justice Moritz disagrees. He and at least four of his brethren or sistren think Duke Power (and its shareholders) should suffer whatever loss would flow from abandoning this policy so that it could hire more minorities. (In doing so the majority dismisses Justice Jessie's cost-benefit analysis demonstrating that the devaluing of education in the majority's decision will produce far greater long-term social harm than the benefits that result from hiring a rather limited additional number of minority employees.) Is society better off with Justice Moritz (or Justice Jessie, for that matter) making these decisions instead of management, admissions officers, etc.?
Speaking of admissions officers, what exactly is the defense of the SAT against disparate impact attacks? Assume that it does predict with some accuracy grades in the first year of college. So what? Does it predict grades in later years? Do grades predict success in life? What is success in life? Should Congress and even judges of the quality of Justice Moritz be making these decisions for all of us?
Now for one of my favorites, the under-noticed but real phenomenon of what could be called reverse disparate impact (though I would never use such a term, for the same reason I don't believe there is such a thing as reverse discrimination). Take, for instance, the newly popular policy of giving admission preferences based on overcoming adversity, etc. (Look here for a discussion of the spreading "sob story sweepstakes.") What if such a policy were challenged by white plaintiffs claiming that this policy benefited minorities so disproportionately that it amounted to disparate impact discrimination, citing Justice Moritz's penetrating observation in an earlier case that "some discrimination, the worst kind of discrimination, discriminates without speaking its name."
4. Should it matter whether the college in the above example is private? I ask, because to Moritz it seems to make all the difference in the world. Once again he spends a large portion of this even larger post insisting on the centrality of the distinction between "the Equal Protection context," which he describes as "a limitation on what legislatures can do," on the one hand, and "the axis of Congressional power," usually under the Commerce Clause, on the other. He is untroubled by courts striking down as discriminatory activity in one area what Congress allows or even requires in another.
And once again, I think Moritz makes far too much of this distinction, real though it is. State action, for example, extends far beyond "legislatures," and similarly Congress regulates far more than private businesses. Title IX, as readers of these exchanges know too well, has been interpreted (by colleges themselves, and others) to require colleges to eliminate men's teams in order to equalize participation rates between men and women.
Whatever weight one chooses to place on this distinction, however, if a future Supreme Court were to decide that the 14th Amendment requires pure colorblindness (hope springs eternal on my side of the aisle as well) that would put the kibosh on Congressional power to decide otherwise. And in the alternative, even though the 14th is not now interpreted to require pure colorblindness, on Moritz's own theory Congress is free to do so. Indeed, in the foundational Bakke case a four-justice minority held that Title VI of the Civil Rights Act of 1964 went beyond the 14th Amendment and required a colorblind standard. (I think they were right.) If a future Congress, over the strenuous objections of Rep. Moritz (D, Harvard), were to adopt (or re-adopt, if you agree with the Bakke minority) this strict colorblind standard, would Justice Moritz still argue so eloquently for judicial restraint, or would he side with today's liberals who argue (like yesterday's conservatives) that the original intent of the framers of the 14th Amendment clearly was to allow racial discrimination when it was reasonable?
If this uncertainty over the depth of Justice Moritz's commitment to judicial restraint seems uncalled for, recall that in the legal attack on Proposition 209 in California, which committed the state to colorblind racial neutrality in contracting, admissions, etc., the ACLU and its friends actually argued that the 14th Amendment barred the state from banning preferences! Not the legislature, but the state itself, in the form of the people amending their Constitution. Really. I'm not making this up. I don't have ready access to a law library, but I've searched all Justice Moritz's prior opinions that I can find and I can't find anything he's written rejecting this bizarre view held by a number of his ideological fellow travelers.
5. Moritz has gotten himself so entangled in the separation of powers and judicial restraint that he has not focused on what discrimination is or what I regard as the central problem of disparate impact -- that it is based on, and indeed requires, a conception of equality that distorts what has been (and should remain) a fundamental core value, the belief that rights inhere in individuals and not groups and that foremost among those rights is that the state must treat its citizens "without regard" to race, creed, or color.
Wherever policies that have a disparate impact are regarded as discriminatory (again, in the absence of a persuasive justification), the "remedy" -- whether imposed by a court or undertaken "voluntarily" by an organization, quite possibly to head off litigation -- will inevitably require "racially conscious" hiring, admissions, etc. Further, it will require constant monitoring and regulation to keep the numbers in line with various guidelines, expectations, and requirements. In short, it will tend to insure that race remains contentious, which I suspect is one of the goals of its adherents. This will inevitably result -- indeed, for substantial numbers of people it has already resulted -- in a conception of equality that elevates proportional representation (since that is regarded as the "natural" result of any non-discriminatory system) over the "without regard" principle. Indeed, failure to "regard" race will come to be seen, as it is already seen by many, as the essence of discrimination. "Diversity" will become, as it has already become for many, the most fundamental value, and its implementation and maintenance will have turned yesterday's vice -- assigning benefits and burdens on the basis of race -- into today's mandatory virtue. Disparate impact theory, in short, turns civil rights inside out and upside down. Finally, as I have argued (here and here), if the assumptions on which disparate impact is based catch on it will not only do away with neutrality regarding race and ethnicity but regarding religion as well. This is not good.
6. Boy, I sure hope the Whippersnapper delays his reply long enough for me to get my warm milk and nap....