Affirmative action, in theory, is a matter of distributive justice, which is why liberals and progressives tend to look benevolently on it while conservatives and libertarians consider it a travesty. But in practice, does it actually advance the cause of justice in America? In some respects it has attempted to remedy the long-term effects of slavery and racial segregation; then again, programs that benefit women and recent immigrants cannot be defended on those grounds. Is it supposed to enhance “diversity” in colleges, professional schools and workplaces, or is it more specifically a set of programs for enhancing the class mobility of people who are structurally disadvantaged from birth? How can we decide whether affirmative action has been successful–and how can we decide whether affirmative action could ever be so successful as to become obsolete?

Advocates of affirmative action have three arguments at their disposal: It is a corrective for past inequality and oppression; it is a corrective for present inequality, much of which derives from past inequality and oppression; and diversity in the classroom or the workplace is not only a positive good in itself but conducive to greater social goods (a more capable global workforce and a more cosmopolitan environment in which people engage with others of different backgrounds and beliefs). Critics of affirmative action, for their part, have insisted that it is unfair to people who have never practiced race or gender discrimination; that it is counterproductive insofar as it damages the self-esteem of its beneficiaries and the public trust of institutions that practice it; and that it has deleterious consequences for society at large, because it promotes the unqualified and the incompetent over the talented and industrious.

Affirmative action in college admissions has been problematic, sometimes rewarding well-to-do immigrants over poor African-American applicants–except that all the other alternatives, like offering admission to the top 10 or 20 percent of high school graduates in a state, seem to be even worse, admitting badly underprepared kids from the top tiers of impoverished urban and rural schools while keeping out talented students who don’t make their school’s talented tenth. In the workplace, affirmative action has been checkered by fraud and confounded by the indeterminacy of racial identities–and yet it’s so popular as to constitute business as usual for American big business, as evidenced by the sixty-eight Fortune 500 corporations, twenty-nine former high-ranking military leaders and twenty-eight broadcast media companies and organizations that filed amicus briefs in support of the University of Michigan’s affirmative action programs in the recent Supreme Court cases of Gratz v. Bollinger and Grutter v. Bollinger (2003).

The thorny legal and political issues affirmative action raises are made still more complex by the fact that affirmative action is inevitably recursive: The extent to which it works in practice depends considerably on our interpretations of it. Administrators and organizations that oppose it tend to be quite bad at implementing it; likewise, female and minority students and employees who are told that they are taking the places of more qualified applicants tend to doubt themselves and perform less capably than women and minorities who are told that they have every right to be right where they are.

The Pursuit of Fairness, Terry Anderson’s comprehensive history of affirmative action, starts off in the 1930s, when the cause of justice is unambiguous: The bad guys wear white hoods and signs around their necks reading No Jobs for Niggers Until Every White Man Has a Job!–a slogan used by unemployed whites in Atlanta in 1930. By 1944 race relations had “progressed” to the point where 14 percent of white workers responded affirmatively to the question, “Should a Negro be your foreman or supervisor?” It was a hypothetical, since there weren’t any black supervisors or any African-Americans being groomed for management positions in any integrated institutions in the public or private sector. But global conflagration has a way of transforming the national landscape: “it was the need for manpower in a total war that wounded Jim Crow,” Anderson writes, and after the war, “many began to consider it fair that all taxpayers should have the opportunity to hold jobs supported by their taxes. That concept was the foundation for a policy that eventually became known as affirmative action.”

Though he was hardly free of racial prejudice, Harry Truman gets his due in this history, both for the content of his character and for his exceptional political courage in desegregating the armed forces and the federal government (and provoking Strom Thurmond’s Dixiecrat revolt from the Democratic Party in 1948, a campaign still recalled fondly by certain prominent Southerners). Still, progress was glacial, even measured by the very slowest glaciers on the globe: Though the federal government had hired African-American employees and soldiers in unprecedented numbers (by 1962 African-Americans were 13 percent of the federal workforce, 20 percent of the post office and a third of the General Services Administration), and though the Supreme Court had delivered its decision in Brown v. Board of Education, most of the country was still partying like it was 1899. In 1963, Ebony magazine noted, “two of 3,500 apprentices in all trades in Newark are Negro and in Chicago, where a quarter of the population is Negro, the apprentice figure is less than 1 percent.” In 1964, in Holly Bluff, Mississippi, school officials spent $190 for every white student and $1.26 (note the decimal point) for every black one. As Anderson sums up, the pace of black hiring in the 1950s meant that “minorities could not expect jobs proportionate to their percentage of the population…among business managers and owners until 2730.” As Satchel Paige might have said, don’t look back–some glacier might be gaining on you.

That was the world in which the 1964 Civil Rights Act intervened; that was the world changed by JFK’s creation of the Committee on Equal Employment Opportunity, Congress’s passage of the Voting Rights Act and LBJ’s executive order 11246, which dictated that “the contractor will take affirmative action to ensure that applicants are employed, and that employees are treated during employment, without regard to their race, creed, color, or national origin.” Up to this point, affirmative action is simply a matter of chipping away at the ossified and obsessive American version of apartheid. “But then,” Anderson writes, “the next question: If the government had outlawed discrimination against African Americans, how could it encourage, even demand, that employers begin hiring a race that traditionally had suffered discrimination?”

From that question, Anderson takes us on what is by now a familiar bumpy ride, past Nixon’s “goals and timetables” and the controversial “Philadelphia Plan” to integrate trade unions and the municipal civil services. Griggs v. Duke Power Co. (1971) establishes the principle of “disparate impact” under Title VII; minority “set-asides” are created by the Nixon-era US Small Business Administration Section 8(a). These are followed by the mid-1970s backlash, the era of Nathan Glazer’s Affirmative Discrimination and the American Federation of Teachers and B’nai B’rith amicus briefs for Allan Bakke. Meanwhile, the Equal Employment Opportunity Commission grows teeth, issues new regulations for the hiring of women and wins back pay for race and gender discrimination at AT&T and Bethlehem Steel. In 1981 Reagan takes office with an all-white Cabinet save for Samuel Pierce, the black Secretary of Housing and Urban Development, whom the President often fails to recognize and who rarely shows up for his job; the EEOC entertains a handful of preposterous cases, filed by genuine troublemakers and incompetents; minority set-asides are abused by front organizations, Cuban émigrés, wealthy immigrants and well-connected white women. Then comes the first wave of affirmative action layoffs, as white civil servants with twenty years’ seniority are canned while recently hired blacks and Latinos are spared; by 1989, Griggs–and the principle of “disparate impact”–is overturned by the Supreme Court, and in 1991 Congress passes a new Civil Rights Act and the Senate confirms Justice Clarence Thomas. Four years later Bill Clinton says, “mend it, don’t end it” after conducting an extensive review of all federal programs, and set-asides are eliminated by Adarand v. Peña. California and Washington pass state referendums abolishing affirmative action, students and conservative organizations file suits against the Universities of Texas and Michigan, and finally we reach the present, still trying to understand what remains of the righteous social crusade launched sixty years ago.

So what do Americans really think of affirmative action? It depends on how you pose the question. In 2003, Anderson reports, “opinion polls found that Americans approved 2 to 1 of ‘programs designed to increase the number of black and minority students’; the same people disapproved 3 to 1 of ‘giving preferential treatment’ to minorities and that included a majority of minority respondents.” At the end of two generations’ debate about racial and social justice, it would seem, Americans emphatically approve of affirmative action programs, even as they passionately oppose any race or gender “preferences” that would make affirmative action programs work.

Thomas Sowell, a prominent black conservative sociologist at the Hoover Institution, has been arguing against affirmative action for more than thirty years, and in Affirmative Action Around the World, a couple of Sowell’s arguments are incontrovertible. He’s right, for example, that “no historic sufferings of blacks in the United States can justify preferential benefits to white women or to recently arrived immigrants from Asia or Latin America who happen to be non-white, but whose ancestors obviously never suffered any discrimination in the United States.” He cites the federal largesse showered upon the Fanjul family of Cuban émigrés, who possess “a fortune exceeding $500 million” (but then so does Anderson, in the course of acknowledging that affirmative action, like any system, can be gamed by unscrupulous players). And Sowell is right that the implementation of affirmative action in the United States since the mid-1960s has gone well beyond the race-neutral language of the Civil Rights Act and LBJ’s executive order 11246.

But Sowell is not content with merely plausible arguments; he wants to insist, by way of surveys of India, Malaysia, Sri Lanka and Nigeria, that “quotas and preferences” can lead to mass murder and civil war, as when Sri Lankan Sinhalese attempted to marginalize the dominant Sri Lankan Tamils by means of punitive quotas, and Tamils fought back with protests and guerrilla war, or when Malays instituted such a draconian set of “racial preferences” as to provoke an exodus of high-skilled Chinese from Malaysia. At times Sowell has the honesty to admit that he has no idea whether these “affirmative action” programs are the cause of all the civil unrest he documents: “If it is difficult to isolate the effects of preferences and quotas, as such, on Nigeria’s troubled history, it is much clearer that the group polarization which preceded and produced the preferences and quotas has been deadly in its effects.” But this is like saying that it’s not clear whether we should attribute “race-related violence” in the United States to affirmative action or to slavery and segregation: While it’s difficult to isolate the impact of affirmative action on America’s troubled history, there’s no denying that the group polarization that preceded and produced affirmative action–segregation, Jim Crow, lynching–has been deadly in its effects.

Yet when Sowell finally gets to discussing the United States, it turns out that affirmative-action-inspired mass murder and civil war are the least of our worries. Never mind that the quota systems of Nigeria and Malaysia have never been enacted here; never mind that while India has extended its terribly amorphous category of “backward classes,” the United States has backed away from controversial practices like set-asides and race-norming. For Sowell, affirmative action in America is literally a question of life and death, as illustrated by the case of Patrick Chavis, a black man admitted to the medical school of the University of California, Davis, the same school that rejected Allan Bakke. For years, the liberal affirmative action establishment touted Chavis, as Sowell notes: “The Lawyers Committee for Civil Rights made the usual comparison between Chavis and Bakke–to Chavis’ advantage–in 1997, just two weeks before the Medical Board of California suspended Chavis’ license to practice medicine in the wake of a suspicious death of one of his patients.”

For Sowell, Chavis clinches the argument: “the role of academic institutions is not to play God in judging individual souls. It is, among other things, to see that people like Patrick Chavis do not end up with scalpels in their hands and ‘M.D.’ after their names to lure unsuspecting patients to their deaths.” And Chavis is “not an isolated example”; on the contrary, Chavises have infected the system at every level, corrupting American meritocracy and placing us all at risk:

In other fields as well, it is the ignored third parties who have the biggest stake in what institutions of higher learning do and how well they do it. Applicants for engineering schools do not have nearly as large a stake as those millions of other people whose lives depend on the quality of the engineering that goes into the bridges they drive across or the planes they fly in or the equipment they work with.

Be afraid. Be very afraid. Affirmative action does not merely lead to mass murder and civil war–that’s just for starters. Before long, bridges will collapse and planes will plummet from the sky, and every time they do you’ll know that some incompetent affirmative action hire is to blame.

Why, then, has the private sector gradually embraced affirmative action, with all these incompetent minorities eating away at our social infrastructure? Here’s Sowell’s stab at an explanation:

Court decisions legitimizing affirmative action under prescribed conditions then provided businesses with a set of guidelines that could minimize their legal jeopardy. Therefore, when efforts were made to end group preferences and quotas during the 1980s by some within the Reagan administration, big business support for the continuation of affirmative action helped doom the efforts to rescind it.

What in the world does “therefore” mean in this passage? Sure, minimal “legal jeopardy” is good, but wouldn’t no legal jeopardy be better? Dodging this question, Sowell then writes, “In addition, large corporations tend to have their own internal affirmative action officials and departments, with their own vested interests in the continuation of such policies.” Aha, those vested interests. The idea that large corporations might not want to be seen as racist or sexist employers, like the idea that large corporations really mean it when they say that affirmative action has given them a deeper pool of efficient employees, seems never to have crossed Sowell’s mind.

What social scientists will find most puzzling–or galling–about Sowell’s book, though, is his dogged insistence that affirmative action has never been tested empirically, that there are no studies of its effects, no cost/benefit analyses of its policies. Faye Crosby’s latest work, Affirmative Action Is Dead; Long Live Affirmative Action, provides a compact, ready-to-hand rebuttal of claims like these, crammed as it is with studies, surveys, assessments and a battery of psychological and attitudinal tests. At the same time, Crosby’s book might also stand as an index of the naïveté of some of the people devising those tests. She repeatedly professes bewilderment at opposition to affirmative action: “Why, I wonder, has affirmative action in employment and education not been universally and vigorously supported in the United States?”

Perhaps Crosby’s profession of puzzlement is merely for rhetorical effect, but it’s hard to imagine how any researcher could wonder aloud why white guys, who’d once competed for college placements, jobs and promotions with about 44 percent of the population, might resist policies that put them in competition with the other 60 percent. And Crosby seems to think that many people oppose affirmative action largely because they’re misinformed about it: “For some time I have had no doubt that much of the resistance to affirmative action is based on misconceptions about the policy…. The policy implications of this conclusion would seem to be clear. Just teach people how affirmative action really operates, and the controversy will die down.” Really? Crosby herself seems unsure of this. Although she says that for many Republicans and libertarians, “opposition to policies that resemble affirmative action derives more from a dislike of government intervention than from racial prejudice,” she also concedes that “for overt and covert racists and sexists, the knowledge that affirmative action benefits those who have been previously excluded or oppressed may be what bothers them about the policy.” But I doubt any of this is news in 2005. At the close of her book, Crosby proposes experiments that test people’s support for affirmative action relative to their ideas about justice, but it’s a mystery to me why she didn’t devise those experiments and conduct them herself.

Then again, thanks to Crosby, I’ve gotten the distinct sense that such tests are part of the problem: She cites–and has conducted–study after study in which people are asked whether they approve of Company X, which gives “slight preferences for qualified women and ethnic minorities,” or Company Y, which gives “strong preferences to qualified women and ethnic minorities, even if we must turn away better qualified or more highly skilled non-minority applicants.” And guess what? In study after study, people approve of Company X’s policies but not Company Y’s–without being told that, as a matter of fact, Company Y’s policies are illegal in the United States. On the one hand, this result suggests that more people would support affirmative action if they knew what it does and does not entail; on the other hand, it suggests that these very tests may be muddying rather than clarifying the issue. Perhaps it is no wonder that so many Americans believe affirmative action involves precisely the kind of racial quotas and set-asides that have been struck down by courts time and again; to gauge by studies like these, professional psychologists and sociologists have managed to spread disinformation about affirmative action as effectively as a whole think tank full of Sowells.

As Anderson points out at the close of The Pursuit of Fairness, the nation’s demographics will render affirmative action increasingly troublesome in the next quarter-century, not only because the sons and daughters of the new African-American professional class are now entering colleges and workplaces but also because integration and immigration have produced something like a melting pot or a glorious mosaic or a dead metaphor of hybrid hyphenated identities. This outcome has further complicated a system that, as Nathan Glazer (who recently recast himself as a supporter of affirmative action) has remarked, already grants affirmative action status to people from Argentina and Spain but not Brazil or Portugal, and considers everyone from the Middle East to be white.

But our confusions are not simply demographic. As Scott Jaschik reported last summer in the Boston Globe, “The University of Michigan released enrollment figures for next fall showing that the number of black students in its freshman class would be declining by as much as 13 percent. That same day, Texas A&M University–a school that refuses to consider race or ethnicity in admissions–announced its own numbers for the fall. Enrollment would be going up–dramatically–for all minority groups, including a whopping 57 percent increase for black students.” Apparently, Michigan’s response to Gratz v. Bollinger produced a more complex application process that contributed to an overall 18 percent drop in applications; meanwhile, Texas A&M, an all-white school until 1963, announced earlier this year that it would abolish “legacy” admissions and began recruiting low-income students from urban areas. Clearly, Texas A&M sent somebody the right signal about race and justice, while Michigan is left to explore the imponderables of unintended consequences. And if the history of affirmative action is any guide, all we will be able to predict about the experiment is that both its intended and unintended results will surprise us–and compel us to think again, and yet again, about how best to foster justice for all.