"The Spoils of Victimhood,
or the case against the case against affirmative action"

by Michael Kinsley (from The New Yorker, 3/27/95)


YES! MICHAEL KINSLEY, in standard type: The moral principle that is wielded against affirmative action is "color-blind equal opportunity." This principle, as an aspiration or as a description of the United States right now, is full of logical problems -- problems so severe that they make any self-righteous anathema on affirmative action nonsensical.

NO! Interspersed REBUTTAL by Curtis Crawford, in italics: Amid bursts of (sometimes enlightening) intellectual fireworks, Mr. Kinsley constructs and demolishes a straw man. Onto the clear and simple principle of racial nondiscrimination, he tacks an ambiguous and thoroughly muddled notion of equal opportunity. The incoherence of the latter he then attributes to the former.

"The Spoils of Victimhood"

Affirmative action is one of those controversies, like abortion, in which opponents have the advantage of moral clarity. They are defending a seemingly absolute principle, while supporters are defending something much more ambiguous.

If you believe that human life begins at the moment of conception, abortion is murder, plain and simple. The opposing principle—a woman has the right to ''control her own body''—can't possibly be so absolute. (It doesn't give the right to kill your Siamese twin, for example.) Few abortion opponents actually carry their morally clear principle to its logical conclusion: that women who procure abortions should be prosecuted for murder—and, in death-penalty states, executed. But the seeming moral clarity is a great rhetorical plus.

In the case of affirmative action,* opponents have the further advantage that the morally clear principle* they are defending is one we all profess to share. We don't all agree that human life begins at conception, but we do all agree—or, at least, fear to disagree—that the United States *ought to be a society of ''color-blind equal opportunity'' (henceforth ''CBEO'')*. We all ''oppose on principle any counting, discriminating or allocating by race'' (as one conservative columnist recently stated the anti-affirmative-action credo)—or, at least, we would like to do so. When opponents of affirmative action quote Martin Luther King, Jr., about judging people by the content of their character, not the color of their skin, who can object? You can argue about the reality, but you won't get far by challenging the ideal.


* Mr. Kinsley fails to define "affirmtive action" and mislabels the opposing moral principle. By affirmative action he evidently means preferential treatment based on race or ethnicity. The moral principle that condemns such treatment is not "color-blind equal opportunity" (CBEO), but Racial Nondiscrimination: i.e., no difference in treatment based on race or ethnicity.

This principle is prescriptive, not descriptive. It is an aspiration, but also a command, both moral and legal. ''Color-blind'' is close, but not quite the same: obeying the rule against discrimination sometimes requires awareness of race, as when checking to see whether discrimination has occurred. ''Equal opportunity'' is irredeemably ambiguous. It can mean something specific and feasible, like the right of a worker not to be treated differently because of race; or something indeterminate and unrealizable, like the right of everyone to an equal chance for success.

Title VII of the 1964 Civil Rights Act, which forbids employment discrimination based on race, ethnicity, sex, or religion, is named "Equal Employment Opportunity." The federal agency established to enforce Title VII is the "Equal Employment Opportunity Commission." Employers who promise to obey Title VII are called "Equal Opportunity Employers." In this statute, "equal opportunity" is firmly tied to nondiscrimination on a few forbidden bases. Is this what Kinsley means by "equal opportunity"?

*Opposition to affirmative action has a second great advantage in today's political culture: It feeds that powerful hunger for the moral prestige and political spoils of victimhood. It gives white males whining rights in the victimization bazaar, just like minorities and women.* This also is an advantage not shared by opposition to abortion, which—agree with it or not—is selfless. Right-to-lifers are out to protect the interests of what they see as a powerless other person. Arguments against affirmative action, by contrast, offer a moral gloss for self-interest, which is always appealing. And the appeal cuts across economic and cultural lines. *It tempts lower-middle-class workers concerned about their stagnant incomes and job security. It tempts yuppie professionals concerned about their promotions at work and their kids' college admissions. It tempts intellectuals concerned about their tenure.* Republicans have decided, with almost sadistic delight, to make affirmative action their ''wedge'' issue for 1996—the issue that will pry various voting blocs away from the Democratic party for good. Democrats are struck dumb with terror, or reduced to pathetically disingenuous denials that affirmative action violates the sacred CBEO principle.


* These are ad hominem arguments, designed to impugn the morality of people who uphold the moral principle that Kinsley attacks. One could respond in kind concerning the supporters of affirmative action, by citing the readiness—of beneficiaries to enlarge and expand racial preferences, while denying their existence; of minority leaders to condemn discrimination when adverse to minorities, while ignoring it when adverse to whites; of politicians to condone and increase minority preferences in the pursuit of minority votes. But such arguments would not decide the validity of the policy.

There's plenty to be cynical about in the current campaign against affirmative action. Conservatives, who ordinarily sneer at victim chic, embrace it in this case. *Today's sanctimonious calls for color-blind equal opportunity for the benefit of whites come primarily from political elements that opposed the principle when it was enshrined, in the 1950s and 1960s, to help blacks.* **(Republicans as moderate as George Bush were against the 1964 Civil Rights Act.)** And even today much of the political power of anti-affirmative-action rhetoric is based on its appeal to emotions that are anything but color-blind. But all that is mere irony. Most affirmative-action opponents are not racists. The issue inflames emotions that are understandable, if not logical.


* Implementing the Racial Nondiscrimination principle helps not only blacks but whoever would otherwise be discriminated against. Most Americans supported it "in the 1950s and 1960s," and still do: a splendid moral achievement. One should rejoice that conservatives have embraced it, and lament that liberals have abandoned it.

** Bush was a congressman from Texas. Almost all Southern politicians, Democratic or Republican, were "against the 1964 Civil Rights Act." The leadership for passage was mostly Democratic, but the votes were largely Republican.

Especially in our sound-bite-and-spin political culture, it is hard for logic to prevail over emotion. You cannot trump a simple and ringing principle like CBEO with 'Wait a minute—it's more complicated than that.'' But in the case of affirmative action it is more complicated than that. Even as an aspiration, let alone as a description of the United States right now, ''color-blind equal opportunity'' is *full of logical problems—problems* so severe that they make any self-righteous anathema on affirmative action nonsensical.

  * I deny, and Mr. Kinsley cannot show, that the racial nondiscrimination rule is "full of logical problems." When he tacks on the phrase, "equal opportunity," does he mean something other than racial nondiscrimination?

Opponents of affirmative action have a potted history and a potted analysis that go together. The potted history is that the civil rights movement was good through about 1964, when it pushed to eliminate discrimination against blacks, and went bad after about 1969, when it started pushing for discrimination against whites. The period 1964-69 was an alleged golden age of near-unanimous agreement that there should be no discrimination either way. *The potted analysis is that there is a simple spectrum of discrimination, with a zero point in the middle—the CBEO point—where discrimination against blacks stops and discrimination against whites begins.* But it is not possible either historically or analytically to draw a sharp distinction between ''good'' civil rights and ''bad'' affirmative action. The two are entwined.


* What Mr. Kinsley calls the "zero point" bars racial discrimination against everyone, not just blacks and whites. It is the core of the state and federal civil rights laws of the 1950s and '60s. As examples, here are two provisions of the US Civil Rights Act of 1964:

"SEC. 601. No person in the United States shall, on the ground of race, color or national origin, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance."

"SEC. 703. (a) It shall be an unlawful employment practice for an employer (1) to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual's race, color, religion, sex, or national origin.."

Does Kinsley contend that the rules stated here are "full of logical problems"?

We needn't dwell on the obvious and boring fact that there is still plenty of racism and discrimination of the traditional white-on-black sort, which is a significant handicap to blacks trying to make it in American society. In 1990, the Urban Institute sent out young blacks and young whites with equivalent résumés and references, and who had been trained to behave identically, to apply for entry-level jobs. The whites did substantially better. You're not surprised, are you? Although affirmative action of course gives blacks an advantage in specific situations, *a black face is still a net minus in the climb up the American greasy pole.*

  Perhaps so. Testing the presence of racial discrimination by using people with "equivalent resumes and references" is a sound approach. In its Fall 1990 issue, the Journal of Economic Perspectives published a comparison of black and white hourly wages for men, ages 22-29. The research statistics, based on the Nationl Longitudinal Survey of Youth, were able to control for each man's region, education, work experience, aptitude test scores, and other variables. The results: the wages of blacks and whites with the same characteristics were within four percentage points of each other. [The article, by June O'Neill, is entitled "The Role of Human Capital in Earnings Differences betwen Black and White Men."

*Then, there's the equally obvious fact that the typical black American does not run the race of life with ''equal opportunity'' at the starting gate.* **Most blacks are at a disadvantage that begins with prenatal nutrition and runs through education to almost every other factor in career success.** It hardly matters whether this disadvantage is due to the legacy of past racism or the failures of the Great Society, or whatever. The fact remains.


*In this sentence, "equal opportunity" means something quite different from Racial Nondiscrimintion. A Mr. Kinsley uses the word here, one lacks "equal opportunity" if one starts "the race of life" at a disadvantage, whether one suffers racial discrimination or not. This is a common and legitimate definition of the phrase, but it is not what Title VII and other antidiscrimination statutes require or can produce.

**Disadvantages that many blacks suffer, such as family poverty, inadequate nutrition, and incompetent education, are also suffered by many who are not black. Insofar as it is the state's responsibility to address these disadvantages, it can and should do so without regard to the race of those who suffer them.

Indeed, leave aside the racial issue and consider the alleged principle of ''equal opportunity'' for a moment. It is an odd one. We all pay lip service to it, but in many respects it is not even accepted, let alone realized. It is violated in ways we can all agree to deplore—e.g., nepotism—but it is also violated in ways we tolerate, or even celebrate. Elite colleges, for example, routinely seek geographical diversity in their student bodies, turning down better-qualified applicants from Brooklyn or Brookline to make room for a North Dakotan. Yet people are not tearing their hair out over that. Preferential treatment for alumni children is also fairly uncontroversial.

  In this paragraph, by "equal opportunity" what Mr. Kinsley appears to mean is neither equal advantage nor racial nondiscrimination, but no difference in treatment based on such factors as residence or alumni parentage. His definitions of "equal opportunity" are all over the lot, confirming my assertion that the term is irredeemably ambiguous.

These are narrowly mechanical examples. More generally, consider child rearing*. You cannot watch what frenzied yuppie parents do for their kids—the intelligence-nurturing* toys, the infinite variety of special lessons, and so on—without concluding that either these parents are wasting a lot of money and energy or their kids are gaining a major leg up in life. I suppose the correct answer is both. Yet who would criticize these parents for wanting the best for their kids? Then, there are connections.* A friend recently asked me to help her son get a summer job in journalism. And, of course, I did what I could. Should I have refused to help my friend's kid, in the name of ''equal opportunity''? It wouldn't be human. The child of a Supreme Court justice, to choose a ripe example, will enjoy vast quantities of this kind of unequal opportunity—whatever the child's race, and whatever the parent's stated views on affirmative action. Racism need not be involved. But this sort of thing does, as a practical matter, put most blacks at a real disadvantage. And it puts the whole notion of ''equal opportunity'' in a strange light.

  * There is no moral consensus or legal mandate in the United States requiring "equal opportunity" defined as an equal chance of selection despite differences in upbringing, education, connections, state of residence, etc. The only kind of "equal opportunity" that is mandated in the Civil Rights Act is that a person's chance of selection not be affected by his race, ethnicity, sex, or religion. Thus Racial Nondiscrimination is a much clearer description of the moral principle that condemns race-based affirmative action than the term "color-blind equal opportunity."

It is important to distinguish two different arguments for ''equal opportunity.'' There is the economic-efficiency argument: that the economy works best, for the good of the entire society, when an irrelevant factor like race (or connections or geography) is not allowed to interfere with selecting the best-qualified person for the job. Then there is the fairness argument: that it is unfair to a particular individual to deny him or her a job or other career opportunity because of an irrelevant factor like race (or connections or geography). The efficiency argument is valid, at least in theory. The fairness argument is much murkier. Yet it is the fairness argument that sustains opposition to affirmative action.

  By representing the principle as "equal opportunity" rather than Racial Nondiscrimination, Mr. Kinsley misses the moral difference between different treatment based on factors like connections or geography and different treatment based on race. All impede the selection of the best-qualified person, and all may seem unfair, but only the latter has become a crucial issue in state, federal and constitutional law. This status is warranted by the fact that racial or ethnic favoritism, throughout history and around the world, has resulted in racial or ethnic partisanship, prejudice, animosity, oppression and internecine war.

In The Bell Curve, their best-selling book on race, genes, and intelligence, Charles Murray and the late Richard Herrnstein willfully or sloppily confuse the two arguments. The authors make clear their hostility to affirmative action. Indeed, it doesn't require too much psychological over- reaching to speculate that hostility to affirmative action was a chief motivation for their project. They ask, in reference to affirmative action: ''To what extent is a society fair when people of similar ability and background are treated as differently as they are now?'' And they answer: ''Such a society is manifestly unfair." *Although they acknowledge that blacks are still behind whites in job status and pay, they argue that this is not true ''after controlling for I.Q.'' * In other words, blacks are already doing as well as they deserve to, given their lower intelligence. Therefore, Murray and Herrnstein conclude, affirmative action imposes ''costs in economic efficiency and fairness.''

  *Many people mistakenly assume that lower average racial achievement or earnings are proof of discrimination, whereas they may actually be based on less ability. It would be important evidence that discrimination against blacks is no greater than discrimination in their behalf if, after controlling for individual I.Q. scores, blacks wee found not to be "behind whites in job status and pay." The Herrnstein/Murray conclusion that blacks are not behind in these respects "after controlling for I.Q." is based on data from the National Longitudinal Survey of Youth, the source of the findings in the Journal of Economic Perspectives article quoted above.

One reviewer of The Bell Curve—Alan Ryan, writing in The New York Review of Books—labels this the ''dollars per IQ point'' argument, and dryly observes: ''If your notion of justice is that people should be paid according to their IQs, then [affirmative action] is unjust.'' But, further- more, Murray and Herrnstein have the logical connection between their factual assertions about IQ and their moral outrage about affirmative action exactly backward. The more ability is determined to be innate and inexorable—something you're born with and bear no moral responsibility for—the less sense it makes to talk about equal opportunity in moral terms as a matter of ''fairness.'' What kind of ''equal opportunity'' is it to be born with a low IQ?

  *Here Mr. Kinsley returns to his notion of "equal opportunity" as the absence of any disadvantge -- whether innate or developed, parental, intellectual, physical, educational, social, economic -- in one person as compared to another. The only society in which this kind of "equal opportunity" could exist is one in which all persons were identical in all respects. If there are readers inclined to this definition of "equal opportunity," the mixture of force and confusion in Mr. Kinsley's arguments may be curative. But they have nothing to do with the logical coherence of Racial Nondiscrimination, the moral principle that is violated by race-based affirmtive action.

Whatever combination of genes and environment may be responsible, each of us plays the game of life with cards dealt by forces largely beyond our control. The content of our characters is the least of it. This makes ''equal opportunity'' a dubious, if not meaningless, standard of fairness. *Dispensing with the fairness argument* and sticking with economic efficiency would leave affirmative action opponents on a firmer logical footing (though published estimates of the cost of affirmative action—4 percent of gross national product, in one case—are highly tendentious). But abandoning the banner of ''fairness'' would drain the cause of most of its demagogic appeal. And, to the reasonable mind, it would raise the question of whether the social benefit was worth a certain economic cost.

  *Where was Mr. Kinsley when the segregationists really needed him? From the abolitionist petitions against racial discrimination in the 1830s through the antidiscrimination statutes and constitutional amendments following the Civil War down to the state and federal antidiscrimination laws of the 1950s and '60s, the rallying cry was that racial discrimination was "unfair." The supporters of racial discrimination, under heavy moral fire, would have welcomed Kinsley's proofs that the claim of unfairness is absurd.

After all, traditional civil rights enforcement of the kind almost everyone claims to believe in can also impose a real economic cost. Much discrimination in the economy is motivated not by irrational racism but by rational economic calculation. It is easier and cheaper to hire on the basis of group generalizations than to judge each person on individual merit. If those generalizations made no economic sense, a law would not be necessary to prevent most employers from making them. But we do have such a law, having correctly decided that social justice is worth some economic cost. In this way, among others, affirmative action is on a continuum with traditional civil rights principles, not a departure from them.

Of course, it is widely believed that the proper goal of civil rights is a society where people are judged as individuals and not as members of a group. This ideal, many people say, is what divides ''good'' civil rights from ''bad'' affirmative action. But, like ''equal opportunity'' and other shibboleths of the affirmative action debate, the appealing notion of ''individuals, not groups'' is more a slogan than a coherent principle.

What does it mean to judge people ''as individuals'' and not as members of a group? Does it mean that each person should be evaluated in a way that does not involve assumptions based on that person's membership in some group? In that sense, no one is ever judged as an individual. Every criterion used in the meritocratic selection process, whether for a job or a place at school, is a generalization from some group trait. Those generalizations are always approximate, and sometimes wrong. Even the most seemingly scientific and individually tailored measurement—say, a score on a medical-school entrance exam—cannot predict with perfect accuracy who will make the best doctors. People who score high on the exam are being judged, as a group, more likely than others to become good doctors, and as a generalization that may be perfectly valid. (More refined selection machinery may be impossible or too costly.) But it is a group generalization.

  *It is as true or truer that "every criterion used in the meritocratic selection process" is based on a generalization concerning an individual trait. It is individuals who have the traits that employers and admissions officers measure when deciding whom to select. It is also true that the helpfulness of the measurements in choosing the best candidate is only probable. Being black or a veteran is as much an idividual fact as being smart or industrious. So Mr. Kinsley is right that the principle of judging people "as individuals" does not provide a sufficient reason for insisting on racial nondiscrimination.

True, there are groups and groups. ''People who score high on their medical-school entrance exams'' are not self-identified as a group, and feel no group affinity. Other groups that get systematic group preference in our society—veterans, for example—have more of the symptoms of ''groupiness.'' Giving veterans preference (e.g., for civil service jobs) violates the ''equal opportunity'' principle that a job should go to the best-qualified candidate. It does so in favor of the apparently superior principle of rewarding those who have sacrificed for their country. But that, too, is only a generalization. Some veterans may have sacrificed nothing in particular, while nonveterans who have sacrificed greatly in other ways get no preference, or may even lose out to a less deserving veteran. The point is not that giving veterans preference is a bad idea—only that, like any group generalization, it is approximate. Yet we live with it.

What is different about group generalizations based on race? *Generalizations that discriminate against blacks have been outlawed for obvious historical reasons.* But what of group generalizations that discriminate in favor of blacks? How are they different from the group generalizations that are the warp and woof of daily life? **The answer cannot be that they are different simply because race is innate and immutable. So, according to its enthusiasts, is IQ—yet opponents of affirmative action generally wish to see the role of such innate characteristics (if not IQ, then ''merit'' or ''talent'') enhanced, not reduced.** ***The answer must be that race is such a toxic subject in American culture that it should not enter into calculations about people's places in society—even in order to benefit racism's historic victims.*** That is a respectable answer. But it understandably rings hollow to many blacks, who see this ****sudden and ostentatious anathema on racial consciousness as a bit too convenient. Where was color blindness when they needed it?****


* It would be clearer to say that what has been outlawed is discriminatory actions, not "generalizations." "Generalizations" as Mr. Kinsley uses the term are the reasons for a rule, not the acts it forbids. Moreover, when discrimination against blacks has been outlawed, it has almost always been by forbidding racial discrimination against any person, regardless of race. There were indeed "obvious historical reasons" for that general prohibition. If Kinsley disagrees, does he want us to withdraw the legal right not to be racially discriminated against from certain racial and ethnic groups? If so, would he specify which?

** Good point.

*** I believe that some such claim is crucial for the moral principle that racial discrimination is wrong. However, I would amend his sentence to read: that race or ethnicity is so easily a source of prejudice, partisanship, hatred, oppression and civil war in any multi-racial or multi-ethnic society that it should not be a basis of unequal treatment, even to benefit persons whose racial or ethnic group has suffered from such discrimination in the past.

**** Sudden? Racial Nondiscrimination as a moral principle has been the watchword of the American struggle for racial justice since the 1830s.

In practice, moreover, even the most devout preachers of CBEO tolerate violations of the alleged color-blind principle, either because of a more important principle or because of some consideration more pressing than principle. The Republican party, for example, has never acted as if it regarded color blindness as a supreme moral axiom. *Republicans, who were slow to approach the CBEO zero point in the 1960s, have often charged past it in the years since.* If color blindness is a virtue, hypocrisy in this case is the tribute virtue pays to vice.

  * True, alas!

Like many of the supposed excesses of the Great Society, affirmative action actually got its big push under President Richard Nixon. The first official government affirmative action program was the Philadelphia Plan, designed to cure racial segregation in the construction industry by making federal contracts contingent on the racial composition of the workforce. The plan was created and then killed by the Johnson administration; Nixon not only revived it but lobbied hard to prevent Congress from killing it again.

During the Reagan years, administration officials like Clarence Thomas, the chairman of the Equal Employment Opportunity Commission, enforced racial hiring ''goals.'' The 1988 Republican platform promised to ''increase, strengthen, and reinvigorate'' government programs that earmarked contracts and subsidized loans for minorities. Although the 1992 platform declared, ''Promoting opportunity, we reject efforts to replace equal rights with quotas or other preferential treatment.'' David Frum points out in his book, Dead Right: ''Every committee convened at the [1992] convention was required to be half composed of women and the speaking roster was filled with the names of nobodies selected only because of their sex or race.'' Even as the GOP plots to make racial favoritism its hot-button issue for 1996, Newsweek reports on a ''GOP Minority Outreach Strategy'' memo that calls for preferential committee assignments for minority House members and special efforts to find minorities to testify at congressional hearings.

But this kind of garden-variety political hypocrisy is less interesting than the larger, and possibly sincere, moral confusion of those who maintain that there's a good kind of affirmative action, which upholds the CBEO principle, and a bad kind, which betrays it. This is a widespread view, even among the harshest critics of affirmative action as it is practiced today. Something like it may be what House Speaker Newt Gingrich had in mind when he remarked the other day, semicoherently, ''It is my belief that affirmative action programs, if done for individuals, are good, but if done by some group distinction, bad, because it is antithetical to the American dream to measure people by the genetic patterns of their great-grandmothers.'' In The Be!l Curve, Murray and Herrnstein are a bit clearer. The context is a discussion of college-admissions policy, but the logic applies to affirmative action in any field:

"Universities should cast a wide net in seeking applicants, making special efforts to seek talent wherever it lives—in the black South Bronx, Latino Los Angeles, and white Appalachia alike. In the case of two candidates who are fairly closely matched otherwise, universities should give the nod to the applicant from the disadvantaged background. This original sense of affirmative action seems to us to have been not only reasonable and fair but wise."

Like others who take a similar line, Murray and Herrnstein claim to be asserting a prelapsarian civil rights ideal. This ''good'' kind of affirmative action has three elements. First, it involves only the preliminary stages of the selection process (often summarized as ''recruitment and training''), not the offer of a job or other selective goody itself. Second, it is only a tiebreaker: No one of superior merit will ever be displaced by someone of inferior merit. Third, ''disadvantage'' is defined in nonracial terms. (This is sometimes called affirmative action by social or economic class.) Do these elements describe an effective form of affirmative action that nevertheless passes CBEO muster? Not really.

First, discrimination in recruitment and training is still discrimination. Someone in the disfavored class will not get the job who otherwise would. If the special recruitment efforts are aimed specifically at racial minorities, as affirmative action opponents less zealous than Murray and Herrnstein would allow, that means that whites who might also benefit from recruitment and training are being victimized.

Second, a tiebreaker matters only for the last available slot. If two people are equally qualified for admission to a college class of a 1,000, the solution is to admit or to reject them both, unless—by an extraordinary coincidence—they are worse qualified than exactly 999 other applicants. It is affirmative action critics—especially Murray and Herrnstein, with their IQ obsession—who seem to imagine that everyone in America can be ranked with scientific precision, from No. 1 to No. 260,000,000, in terms of his or her qualification for any desirable career opportunity. Under that assumption, or even under a less extreme version of it, tiebreaker affirmative action would be meaningless. But if you start to take a more relaxed and nuanced view of ''qualifications''—that people are often equally qualified, or, at least, that the machinery for ranking people by qualifications is often crude—you undermine both the fairness case and the efficiency case against traditional affirmative action.

Third, what about affirmative action based on ''disadvantage'' rather than on race? In recent weeks, this has become the hottest idea in the affirmative action debate. Even President Clinton, answering a press conference question about affirmative action on March 3, said that he favors "need-based programs."

Reverse discrimination based on something other than race would indeed make affirmative action color-blind, but it would not cure affirmative action's other supposed defects. It would still be social engineering. There would still be ''victims.'' It would still mean passing over a better-''qualified'' candidate for a less ''qualified'' one. If that is unfair and economical1y costly, the injustice and the cost remain even though the color consciousness would be eliminated. The alleged stigma—to oneself and others—of having got the desirable slot through affirmative action rather than by one's own just deserts would remain as well.

Then, there is the problem of defining and comparing ''disadvantage.'' Affirmative action opponents have justifiable fun with the puzzle of deciding, in a society of racial mongrels, who, exactly , is black—and, beyond that, what other ethnic groups should qualify for official ''victim'' status. Changing the standard to disadvantage in the abstract would make the puzzle a nightmare. Is it worse to be a cleaning lady's son or a coal miner's daughter? Two points if your father didn't go to college, minus one if he finished high school, plus three if you have no father? (Or will that reward illegitimacy, which we're all trying hard these days not to do?) Communist societies tried this kind of institutionalized reverse discrimination-penalizing children of the middle class without any enviable success. Officially sanctioned affirmative action by ''disadvantage'' would turn today's festival of competitive victimization into an orgy.

  *Mr. Kinsley's arguments against preferential college admissions based on various measures of social or economic "disadvantage" (not defined as race) are cogent. However, such disadvantage is rightly the basis for many valuable government programs. Why not provide federally-supported supplemental education designed to help children who attend incompetent elementary and secondary schools to acquire the ability to meet regular college standards? The best answer to the demand for lower standards is training that helps people meet high standards.

Once you accept ''disadvantage'' as a legitimate ground for favoritism, moreover, the question arises of whether it is so terrible to use ''black'' as at least one rough-and-ready shorthand for ''disadvantaged.'' After all, advocates of this kind of affirmative action acknowledge that many, if not most, of its beneficiaries would, in fact, be black. To be sure, not all blacks are disadvantaged, and not all disadvantaged people (however you choose to define the word) are black. But generalizations, as has been noted, are an inevitable part of any selection process. And the generalization ''Black equals disadvantaged'' is probably as accurate as many generalizations that go unchallenged, such as ''High test score equals good doctor'' or "Veteran equals sacrifice for the nation.'' *So opponents of today's affirmative action cannot take refuge in some ideal version of it that would avoid giving offense to their stated principles.* **And even today there are many situations where affirmative action opponents don't uphold the alleged principle of color blindness—or where they preach it while practicing the opposite.** How much is left of the CBEO principle when all these situations are added up?


* Not so. I easily think of three kinds of affirmative action that are perfectly compatible with Racial Nondiscrimination: 1. As defined in Title VII of the Civil Rights Act of 1964: action by an employer to hire, reinstate or provide back pay to persons against whom he has discriminated. 2. As defined in President Johnson's Executive Order 11246 (1965): measures by an employer to ensure that racial discrimination does not occur. 3. As often proposed in the early days of affirmative action: to make sure that notice of available positions is widely disseminated.

** True, but inconsistent supporters do not an illogical principle make.

Almost everybody accepts, for example, the idea that sometimes diversity of races, or even race itself, is a valid job consideration. Hiring policemen is a common illustration. Being black is a positive job qualification for a cop patrolling a black neighborhood (or, at least, racial diver- sity is a valid hiring goal for a city police force). The selection of Supreme Court justices is widely accepted to be another situation where it seems to be okay to take race or sex into account. Why? Because the notion that measuring ''qualifications'' alone can scientifical1y pinpoint the best Supreme Court justice is especial1y absurd, and because the notion that a diverse Supreme Court has value for its own sake is especial1y appealing. Ronald Reagan promised outright to appoint a woman justice, and did so. George Bush swore that race had nothing to do with his selection of Clarence Thomas, but nobody believed him.


Racial diversity is popular now; racial homogeneity was popular at a former time. Neither goal ordinarily justifies racial discrimination. If an exception is sought, the burden of proof should be on those who seek it. Racial discrimination in the employment or assignment of police officers seems to me a plausible exception; in judicial appointments, dead wrong.

A more complicated example of acceptable reverse discrimination is the whole area of ''remedies.'' *Although there are arguments about the details, even hard-core affirmative action opponents accept the principle that racially conscious ''remedies''—e.g., reverse discrimination, even quotas—may be imposed by the courts (or sometimes by lesser authorities) when an employer or an organization is found to have practiced traditional racial discrimination in the past.* This is supposedly part of the ''true,'' pre-1969 meaning of ''civil rights.'' It's an odd exception to the alleged color-blind principle, when you think about it. Can the same civil rights precepts really forbid you to practice reverse discrimination except when they require you to do the same thing? (That is indeed how the 1964 Civil Rights Act is generally interpreted.) The idea of a judicial ''remedy'' is that the victim of injustice has a right to be made whole and the wrongdoer has no right to complain. But the reverse discrimination remedy usually benefits people who were not victims of the specific act of past discrimination, and the burden falls on people who were not responsible for it or beneficiaries of it.


* I agree that this remedy is inconsistent with the Racial Nondiscrimination principle. I also agree that this is how the EEOC and the courts have interpreted Title VII, although the statutory language seems clearly against them. If this remedy were necessary for adequate enforcement of Racial Nondiscrimination, we would indeed have a logical conflict in the rule itself. Such a necessity has not been shown. It is better that some violators of the principle go free than that racial discrimination be used to enforce it.

In a way, the ''remedy'' exception, if you accept its logic, swallows the ostensible color-blindness rule. After all, a major justification for affirmative action in general is making up for past discrimination. The whole thing is a remedy. Indeed, the ''remedy'' logic works better in general than it does in the particular. Unlike past specific episodes of racial injustice, past racial injustice in general real1y does have some bearing on the relative places of random blacks and whites today . If people real1y could be ranked from No. 1 to No. 260,000,000 in terms of ''qualifications,'' past discrimination would clearly have moved blacks down the queue and moved whites up—and affirmative action really does begin to look like rough justice for that. Of course, in real life there is no such simple queue. So the black person who moves up the line thanks to affirmative action may not logically ''deserve'' the place he gets. But, for the same reason, the white person who loses that place doesn't ''deserve'' it, either.

A third widely acknowledged exception to CBEO involves private, voluntary efforts that favor blacks—as opposed to rules imposed by government or the courts. Of all aspects of the affirmative action debate, this one leads to the most comic somersaults of logic by professed defenders of the sacred CBEO principle. In 1990, the Bush administration declared that university scholarships earmarked for minorities violated CBEO and were against the law. A fuss ensued, and the administration retreated to the position that the anathema on minority scholarships applied only to scholarships financed by universities themselves. Minority scholarships financed by private individuals, corporations, and foundations were not merely permissible but praiseworthy. ''I've long been committed to them,'' George Bush said. The alleged difference was that universities are federal contractors, and therefore semigovernmental actors. But it was never explained how racial1y earmarked scholarships could be immoral in the one case and praiseworthy in the other.

Is private affirmative action good or bad? Even the right-wing Heritage Foundation has had scholarships earmarked for minorities. Private efforts directed specifically at minorities, amounting to overt racial favoritism, are widespread, and it is unthinkable that conservatives, of all people, would want the power of the federal government used to smash them. What's more, it is a perfectly coherent position to hold that private individuals should have more leeway to discriminate than the government has. But that doesn't solve the mystery of why the identical violation of the sacred CBEO principle should be positively evil when practiced by the government and positively good when practiced privately. After all, private, voluntary affirmative action is not ''voluntary'' for the victim—the white person who can't get a job or scholarship because a minority was favored. The alleged harm of affirmative action is exactly the same whether it is voluntary and private or is government-imposed.

And there's another logical problem: We do not permit—let alone glorify—private, voluntary racial discrimination against blacks. You cannot offer jobs or scholarships exclusively to whites. We have decided that the government can and should intrude in those circumstances. That decision, and the decision to regard private reverse discrimination differently, may be right or wrong. But, taken together, these two decisions grossly violate the principle of color blindness. Ban private racial discrimination in one case; tolerate it, and even celebrate it, in the other. Yet that represents the consensus view of most affirmative action opponents.

There are only two ways to escape from this logical box while preserving a commitment to CBEO. One would be to support a government crackdown on private efforts to help minorities—which is hardly what those plotting to make affirmative action a big political issue have in mind. (Some might favor a crackdown on brazen reverse discrimination—such as overt racial quotas—even when it is private. But few would use the civil rights laws against reverse discrimination as rigorously as we use them against discrimination of the classic sort.) The other option would be to support repeal of the 1964 Civil Rights Act, thereby allowing private, voluntary racial discrimination of all kinds. Whatever its other merits, this second option would certainly bury the notion that affirmative action opponents wish to fulfill the true meaning of the Civil Rights Act. A third option, of course, would be to abandon the pretense that CBEO is a coherent principle.


A fourth option is to distinguish between violations of the principle that should be outlawed and punished by government, and violations that should be left to individual conscience. A fifth option is to admit exceptions to the principle if their justification meets a high standard of scrutiny. As a strong supporter of Racial Nondiscrimination, I would keep "the 1964 Civil Rights Act," except for enlarging somewhat the size of firms not covered by the employment provisions law. I would urge private charities to adopt nonracial criteria for their beneficiaries, but would not seek to control this behavior by law.

*The point is that a pure, discrimination-free society is not merely a hopeless ideal; it is a logical mirage.* **One final example: actuarial justice. Whites live longer than blacks; women live longer than men. Actuarially, whites should have to pay more each year for the equivalent Social Security pension, and pay less for the equivalent life-insurance policy. Ditto women and men. Does race-neutral-and sex-neutral-justice require that whites and blacks, women and men, pay the same rates, or that they get the same insurance value per dollar paid? There is no logical solution to this puzzle.** (That's why our real-life solutions are all over the lot.)


* By a "pure, discrimination-free society," I assume Mr. Kinsley means a society completely free of racial or ethnic discrimination. ''Completely free of'' would mean that any conflicting good must give way to this one. Such an ideal would be wrong, not because of incoherence, but because of arrogance: claiming always to be greater than any conflicting good. By the way, would Kinsley contend that a society free of discrimination against blacks is a ''logical mirage''?

** I realize that many people interpret nondiscrimination in this case to require "the same rates." That has always seemed to me a preference for the appearance over the reality of equal treatment.

To be sure, affirmative action supporters have their own muddles and false distinctions, and pay their own hypocritical obeisance to CBEO. They cling to a mythical difference between affirmative action and reverse discrimination, or between goals and quotas. In the 1978 Bakke case, the Supreme Court's first major ruling on affirmative action, Justice Lewis Powell made much of the distinction between using race as a ''factor'' in the selection process and using it as the ''decisive'' factor. But, for any given applicant, the selection process is an all-or-nothing proposition. And a factor is not a factor unless it sometimes makes the difference. When race is the deciding factor, it might just as well have been the sole factor: Someone has gained an opportunity, and someone else has lost one, simply because of race. Likewise, there is no sensible moral distinction between a flexible goal and a rigid quota. Both are reverse discrimination. If the goal never leads to a specifìc advantage for a specifìc individual, it is meaningless. And if that advantage causes some other individual to lose out, it might as well have been a quota of one as far as he is concerned.

Good people find this hard to accept. They cling to Justice Powell's comforting belief in the ambiguity of ''factors.'' But consider a law-school class of a hundred. Case 1: Ten places in the class are reserved for blacks. It is the ''decisive'' factor. This is roughly the situation that Powell found unconstitutional in Bakke. *Case 2: In the admissions formula, whatever it may be—board scores plus grades minus parking tickets, and so on—blacks are awarded extra points. It's just a ''factor''—okay by the Powell standard—but some number of extra points will produce exactly ten blacks in the class.* Indeed, under the assumption that everyone, black or white, can be ranked by formula, they will be the same ten blacks who would be admitted under a rigid quota. And the same whites—some number up to ten of them—will lose their places. What, morally, is the difference?

  *Although Powell's standard was (deliberately?) unclear, this probably misrepresents it. His idea was that extra points based on race (e.g., black) or ethnicity (e.g., Italian-American) could be given to individuals who promised an interesting mix, but not to every member of such a group.

The federal courts are currently revisiting another false distinction enshrined in their own doctrine: that between reverse discrimination in hiring and reverse discrimination in layoffs. In 1986, the courts ruled against reverse discrimination in layoffs, reasoning: ''While hiring goals impose a diffuse burden . . . layoffs impose the entire burden of achieving racial equality on particular individuals.'' Of course, hiring goals also impose the entire burden on a particular individual—the one who would have got the job but didn't. The burden is ''diffuse'' only in the sense that in hirings, as opposed to layoffs, we don't know who that person is. This may make affirmative action more politically tolerable, in the way that we often impose burdens in the abstract (such as, ultimately, the burden of dying in defense of one's country) that would be unbearable to impose in the particular. But is there a real moral distinction between hirings and layoffs? *I can't see it.* **Affirmative action defenders, like affirmative action opponents, must have the courage of their convictions—for opponents, that CBEO is an absolute principle; for defenders, that reverse racial discrimination is morally justified—or must revise them.**


* Neither can I.

**Opponents of affirmative action need not and should not hold Racial Nondiscrimination as an absolute principle. Principles need not be absolute in order to be legal and moral. Killing is not wrong in every circumstance; nevertheless we properly hold that killing is wrong. The antidiscrimination statutes do not outlaw racial discrimination in every circumstance; nevertheless they incorporate a crucial moral and legal principle. Under the Civil Rights Act of 1964 there are two important kinds of exceptions: areas of social life to which the prohibition is not extended, such as marriage, private charity and Supreme Court appointments; and special situations in prohibited areas, such as a bona fide occupational reason for selecting by ethnicity. There is another kind of exception not contemplated by the Act: reserving for members of some races only the right not to be racially discriminated against.

If a rule has exceptions, one needs criteria for determining them. The needs to safeguard personal liberty and to restrain the power of government are good reasons for limiting the scope of the legal prohibition. Some occupational needs may require ethnic or even racial discrimination in selection. But unless such exceptions are very narrowly defined, they can destroy public understanding of and adherence to Racial Nondiscrimination. A good criterion for whether to create an exception would be that the consensus in favor is overwhelming, including majorities of the racial or ethnic parties adversely affected. An important safeguard would be that such decisions be made by the legislative, not the executive or judicial branch. One limit I would place on exceptions: that the law should never, explicitly or implicitly, establish a racial or ethnic double standard, by which the right not to be racially discriminated against depends upon a person's race. Those not inclined to observe such a limit should be especially careful that the parties whose right is withdrawn consider the exception reasonable and fair.

The current layoffs case that is drawing a lot of attention, Taxman v. Piscataway, involves two public-school teachers, one black and one White, who were equal by every available standard of CBEO, such as academic credentials and written evaluations. They were equal, as well, by seniority—a typical sort of non-CBEO standard that doesn't seem to bother anybody. If it weren't for affirmative action, the choice between them would have been made by the flip of a coin. So this is what the affirmative action debate comes down to. Is the color-blind principle so sacred, so inviolable, that the school board must flip a coin rather than decide that America might be better off with more black teachers?

Of course, what l have been arguing here has not really been the case for affirmative action—more like the case against the case against affirmative action. And l have not meant to suggest that a color-blind society—or, at least, a color-blind economy—is a meaningless or worthless goal. The point is that it's a complicated and ambiguous goal, and the notion that all other goals—including other civil rights goals—must give way before it belongs in the world of agitprop.

CBEO is not the only argument made against affirmative action. There is the stigma argument: the assumption of inferiority—by the affirmative action beneficiaries themselves as well as by others. The stigma, as critics point out, attaches even to minority-group members who are making it, or could make it, without affirmative action. There is the undeniable social tension and resentment caused by the perception that affirmative action is unfair, whether that perception is rational or not. (This argument, of course, is often advanced by the very people working to aggravate the social tension they claim to deplore.) There is the problem of limiting the principle. What groups—women? Hispanics? Asians?— are entitled to affirmative action, and how do you define the members of the group?

And, as with any form of social engineering, there is the problem of politicizing outcomes that—just or unjust—would otherwise be written off as the workings of fate. The last thing our political culture needs is more opportunities for people to take umbrage. By making the question of who gets what job a matter of public policy, affirmative action opens it up to politics. In politics, almost by definition, minority interests are at a disadvantage. So, to some extent, affirmative action brings its current problems on itself.

All these problems suggest that affirmative action is a social policy that should be used with care, *weighing the costs against the benefits* in each circumstance. What they do not suggest is that a ban on affirmative action is anything like a moral imperative.

  * Let us indeed weigh costs and benefits, but not as if we had learned nothing from the past. The harm from racial and ethnic discrimination throughout history must be placed on the scales. The proven social utility of the racial nondiscrimination rule justifies its status as a moral imperative, not absolute but with a formidable presumption in its favor.

One form of affirmative action that certainly does not meet a reasonable cost-benefit test is affirmative action for businesses owned by minorities and women: preferential access to government contracts, subsidized loans, and communication licenses. The House of Representatives recently voted to kill one of these preferences—a fat tax break for companies that sell broadcast and cable stations to minorities—and rightly so. Anyone in a position to buy and operate a television station either is already affluent or is just fronting for White males. Rich blacks undoubtedly face disadvantages compared with rich whites, but they are not disadvantages worth incurring the real costs of affirmative action to rectify. The Federal Communications Commission's affirmative action policy has been especially farcical. Members of favored groups who get valuable licenses, for nothing or at a discount, are more or less free to resell them, at market rates, to white males or anyone else. The policy amounts to the simple anointment of black millionaires. And, because black millionaire businessmen are such an obvious exception to the generalization that ''black equals disadvantaged,'' these policies help to discredit affirmative action even in situations where the generalization makes more sense.

The FCC insanity illuminates an important fact about the reverse discrimination controversy: Reverse discrimination usually occurs in situations where someone—often the government—is allocating inequality. Somebody gets a valuable government license; somebody else doesn't. Somebody gets into medical school; somebody else doesn't. And, rather than obsessing about whether the inequality ought to be allocated on the basis of race, we might do better to ponder whether reducing the inequality might make more sense than arguing about the allocation.

The FCC has been anointing millionaires for decades. Most of them have been white males. Only recently has it begun auctioning off (rather than giving away) some of its valuable licenses, so that their value is shared by all of us. That change is surely a greater triumph for fairness than either applying or repealing affirmative action. An economist would look at the excess demand for medical school slots and say that this is a market out of whack. The selection process—even assuming that applicants can be ranked accurately according to their doctoring potential—excludes many who would become good doctors. Instead of arguing about the race of those who make the cut, why not try to bring the market into equilibrium? A variety of government policies—such as overrestrictive rules about what services nurses and paramedics may perform—artificially inflate the advantage of being a doctor. Reform those policies (which is starting to happen), and the market will begin to sort itself out. More of those who wish to become doctors will be able to do so. Surely that serves egalitarian principles more than a change—in either direction—in the racial composition of medical school admissions.

Affirmative action has become a scapegoat for the anxieties of the white middle class. Some of those anxieties are justified; some are self-indulgent fantasies. But the actual role of affirmative action in denying opportunities to white people is small compared with its role in the public imagination and the public debate. Black unemployment is still higher than white unemployment, and blacks still trail whites in every major prestige occupation outside sports. That reality raises many troubling questions—about affirmative action, among other things. But it also suggests that *few whites have actually lost a career opportunity to a less qualified black*—certainly far fewer than the number who have been whipped into a fury of resentment over affirmative action.

  * If it were true that, thanks to affirmtive action, "few whites have actually lost a career opportunity to a qualified black," it would also be true that, thanks to affirmative action, few blacks have gained a career opportunity at the expense of a more qualified quite. But this was emphatically not the view of pro-affirmative action organizations in the Bakke case. The Association of American Law Schools, in its amicus brief, calculated that without racial preference, the number of black students entering law school each year would fall by more than 1,000. Conversely, the number of white students would increase by the same amount.

What is most poisonous about the campaign against affirmative action is that it invites whites to blame blacks for their problems. What is almost as poisonous, though, is the way attacks on affirmative action reinforce the 1-to-260-million myth: the myth that, in the absence of racial discrimination, we would all find our places in life according to our just deserts. Oddly, the philosophy of affirmative action itself derives from, and reinforces, the same myth. This myth nurtures frenzies of resentment that do our politics plenty of damage. American society would be healthier—and not only on the race issue—if people, black and white, would keep in mind that neither just deserts nor nefarious discrimination is as important, in determining one's lot in life, as simple luck.

Original article, © Michael Kinsley, reprinted with permission.

[William Morrow & Company (now owned by Harper Collins) reprinted "The Spoils of Victimhood" in a 1995 anthology of Kinsley's essays entitled Big Babies]