Affirmative action, the policy of giving preferences for jobs, university admissions, or government contracts to members of designated racial and ethnic groups, has never been popular, and it could soon be abolished. In 2003, the Supreme Court struck down an undergraduate admissions policy at the University of Michigan that provided extra points for minority applicants. At the same time, the Court approved by a single vote the more subjective practice of taking race into account as one factor among several in admissions to the university’s law school. The change of one vote (by the recently confirmed Chief Justice John Roberts?) would have meant the end of overt affirmative action in higher education. The trend against affirmative action in the states is even more pronounced. In California and Washington constitutional referendums have banned the government from using affirmative action in any of its activities. Other states have ended or severely limited affirmative action by executive authority.

More remarkable than the current opposition to affirmative action is the fact that it ever came into existence in the first place. On its face, the policy seems to violate one of the most basic American values—the idea that individual merit as manifested in a fair and open competition should be rewarded. A practice that seems to go against the individualistic and meritocratic American ethos is clearly vulnerable to an attack that is likely to be persuasive to many of those who do not stand to benefit from it. Moreover, affirmative action seems contrary to the emphasis on colorblindness that was characteristic of the civil rights movement of the Fifties and early Sixties, and was expressed in the language of its greatest achievement—the Civil Rights Act of 1964.

Two very different arguments have been advanced for affirmative action. One claims that it is just compensation for historical injustices and disadvantages. In the case of claims by African- Americans the emphasis is usually on the wounds inflicted by centuries of slavery, segregation, and discrimination. President Lyndon Johnson made one of the most elegant and influential statements of this position in his Howard University speech of 1965, which is quoted by Ira Katznelson in When Affirmative Action Was White:

You do not take a person who, for years, has been hobbled by chains and liberate him, bring him up to the starting line of a race and then say, “you are free to compete with all the others,” and still justly believe that you have been completely fair…. It is not enough just to open the gates of opportunity. All our citizens must have the ability to walk through those gates…. We seek not just legal equity but human ability, not just equality as a right and a theory but equality as a fact and equality as a result.

The other argument, which is reflected in recent Supreme Court decisions and is currently much heard, is based on the assumption that racial and ethnic diversity among “elites”—relatively well-off people who have some degree of responsibility for others, whether private or public—is beneficial to society and its institutions. Prominent among those who defend affirmative action, for example, are spokesmen for the American military who lent conspicuous support to the University of Michigan’s side in the 2003 Supreme Court case. Since the enlisted ranks are disproportionately black and Latino, discipline and morale are presumably inspired by having the same groups represented among the officers, including those of the highest rank. Corporations that deal with a multicultural and multiracial clientele, sometimes on an international scale, find obvious advantages in being represented by people who reflect the racial and ethnic diversity of those with whom they are doing business. Many large corporations practice affirmative action voluntarily even when there is no significant pressure from the government.

In higher education the diversity argument takes a slightly different form. Racially and ethnically heterogeneous student bodies are said to create an appropriate educational environment for students who will encounter many different kinds of people when they go out into the world. Faculties, moreover, must be diverse if they are to provide inspiration and suitable “role models” for minority students.

Clearly affirmative action has had its greatest success in producing more diverse elites, particularly in the much-heralded emergence of a substantial African-American middle class, something that never existed before. But as the sociologist William Julius Wilson has argued for many years, this process of embourgeoisement has been accompanied by the equally substantial growth of “the truly disadvantaged,” the economically marginal black inhabitants of the urban ghettos. Since the advent of affirmative action in the 1960s, the overall differences between blacks and whites have changed very little with respect to average incomes, property holdings, and levels of educational attainment. What is new is the gulf that has opened in the black community between the middle class and the lower or “under” class.1


Affirmative action originated as a pragmatic response by those in the federal government responsible for enforcing the fair employment provisions of the Civil Rights Act of 1964.2 The Equal Employment Opportunity Commission (EEOC) set up under the act lacked the staff to investigate most individual claims of discrimination in employment. It also lacked legal authority to act effectively on behalf of the complainants. As a result, the only way that the EEOC could begin to do its job was to request government contractors to provide statistics on the racial composition of their labor force. If blacks (and by the 1970s other minorities as well) were underrepresented among the workers relative to their percentage of the local population, the EEOC set numerical goals for minority recruitment sufficient to correct this disproportion. Employers were then required to make “good faith efforts” to meet “quotas” for black workers. If they didn’t hire more blacks, they risked losing contracts. The professed aim was equal opportunity, not racial favoritism; but the paradox that bedeviled the program from the start was that it appeared to require preferential means to reach egalitarian ends.

After its fitful beginnings during the Johnson administration, affirmative action took a dramatic turn under Richard Nixon, whose administration put into effect a controversial plan to integrate Philadelphia’s construction trades. Historians have concluded that the Philadelphia Plan of 1969–1970, which set firm racial quotas for hiring for one industry in one city, was a political ploy. It was designed by the Nixon Republicans to cause friction between two of the principal constituencies of the Democratic Party—organized labor, which opposed the plan because of the threat it posed to jobs under its control, and African-Americans, who had overwhelmingly supported the Democrats since the New Deal.3 At the same time, Nixon was trying to appeal to Southern whites by doing little to enforce desegregation, especially in the schools.

When rising opposition to the war in Vietnam became the critical issue for his administration in 1970 and 1971, and hard hats like the construction workers of Philadelphia were in the forefront of those opposing the anti-war protesters, the Philadelphia Plan was quietly shelved. From then on, Republicans were, for the most part, strongly opposed to affirmative action and benefited from the backlash against it, attacking the Democrats as the “party of quotas” because of their continued support for the policy.

Affirmative action was declared constitutional in 1971 when the Supreme Court ruled in Griggs v. Duke Power Co. that discrimination in employment could be subject to affirmative action even if it were not intentional or motivated by prejudice. The Court found that the standardized aptitude tests given by the company to employees prevented blacks from moving to higher-paying departments. Such requirements could be “fair in form,” the Court said, but they could still be described as “discriminatory in operation” if they had an “adverse impact” on blacks. Hence the EEOC was legally entitled to set goals for increasing minority employment and to require periodic reports on the progress being made on fulfilling such goals by any of the 300,000 firms doing business with the federal government.

In 1978 in Regents of the University of California v. Bakke, the Court held by five votes to four that strict numerical quotas, such as those that the medical school of the University of California at Davis had set for minority applicants, could not be permitted. But the concurring opinion of Justice Lewis Powell, who cast the deciding vote, held that race could still be used as a positive factor in considering the qualifications of candidates for admission so long as two criteria were satisfied. If a university were to give preference to blacks it had to establish a direct connection between the claim to such special consideration and a specific historical injustice such as the exclusion of blacks from professional schools over the years (generalized claims of past racism would not suffice). Racial preferences must also serve a “compelling public interest” or purpose. The constitutional foundation for affirmative action laid by Justice Powell has endured for the past twenty-six years. The recent Michigan decisions forbid numerical point systems as well as statistical quotas; but it continues to be constitutional to use race among other factors to determine qualifications for university admissions and employment.


Ira Katznelson has made a major contribution to the affirmative action debate in his book When Affirmative Action Was White. He accepts Justice Powell’s criteria and uses them to justify a much more ambitious governmental attack on racial inequality than currently exists. He presents a new version of the argument that affirmative action is justified as compensation for historical wrongs against black people. Instead of going back to slavery, he maintains that people who are still alive (or have living children or grandchildren) and have been the victims of specific historical injustices can provide strong claims for restitution from the United States government, the direct source of these injustices.


Most of Katznelson’s book is devoted to showing how the economic and social legislation of the 1930s and 1940s favored whites over blacks. Katznelson is not the first historian to argue that the New Deal and Fair Deal widened the gulf between whites and blacks in the United States, but he is the first to consider such discrimination as the principal justification for an ambitious affirmative action program that would include reparations for blacks.4

The undeniable fact is that, by comparison with whites, blacks became relatively worse off during this period. But this relative failure has been obscured by the equally undeniable fact that the material circumstances of African-Americans improved and were, on average, significantly better in 1950 than they had been in 1930. What Katznelson shows is that the Democratic social and economic policies of the Thirties and Forties were rigged so that whites got much more than a fair share of the benefits.

The primary cause of this inequity, Katznelson contends, was the influence of Southern segregationists within the Democratic Party. In the 1930s, when the first New Deal policies were being enacted, white Southern congressmen provided necessary votes for liberal measures that strengthened the labor movement, set minimum wages, and gave relief or temporary work to the unemployed. But they did so only on the condition that the Southern racial order remain insulated against federal actions that might threaten it. The cooperation of New Dealers and segregationists broke down in the 1940s, when a strengthened labor movement began to look south and consider organizing blacks as well as whites. At that point, a new coalition of Northern Republicans and Southern Democrats succeeded in stopping the advance of organized labor, especially by passing the Taft-Hartley Act of 1947, which put heavy restrictions on union organizing.

In 1948 the Democratic Party, with labor support, took up the cause of civil rights for the first time, and Harry Truman was elected president despite the defection of much of the South to the States’ Rights or “Dixiecrat” Party. But this change of heart by the Democrats was, Katznelson points out, less than a complete conversion to the cause of racial justice. He reminds us that the Democrats of the 1950s, trying to keep the South’s electoral votes, backtracked on civil rights and made renewed overtures to Southern white supremacists. In support of his argument, Katznelson might have noted that Adlai Stevenson’s first running mate was a solid segregationist and former Dixiecrat—Senator John J. Sparkman of Alabama.

The New Deal policies that worsened the situation of blacks were not overtly discriminatory. The primary device used by Southern white supremacists was to exclude agricultural laborers and domestic servants from coverage under the Social Security Act and National Labor Relations Act of 1935 and the Fair Labor Standards Act of 1938. Since these were the occupations of most Southern blacks and of much smaller proportions of Southern whites, such exclusions meant that most blacks were being left out of the new welfare state and denied the same chance to escape from poverty that was available to many relatively poor whites. In the South, therefore, the New Deal actually had the effect of strengthening the economic basis of white privilege. It is true that at the height of the Depression African-Americans received some help from the WPA and other emergency measures to provide relief and work, but since Southern white supremacists locally administered these programs, racial discrimination continued.

Service in the military during World War II provided blacks with some opportunities for education and for developing valuable skills. But as Katznelson points out, smaller proportions of blacks than whites actually served in the armed forces (more were considered physically or mentally unfit for military service) and the separate but unequal segregation of the armed forces meant that blacks had relatively fewer chances to acquire new skills and advance to higher ranks. Although he mentions it, Katznelson pays little attention to one bright spot in the World War II experience for African-Americans—the increased access to industrial jobs, especially in the North, resulting mainly from the tight wartime labor market.

The federal government made a modest contribution to diversifying jobs through the activities of the Fair Employment Practices Committee (FEPC) established in 1941 as the result of protests led by the African-American labor leader E. Philip Randolph. The FEPC, by hearing complaints from blacks and demanding explanations from businesses, allowed more blacks to benefit from the new welfare state and narrowed the difference between the average white and black incomes. Here for the first time since Reconstruction the federal government was acting against racial discrimination rather than facilitating it. The federal FEPC did not survive the war but it established an important precedent for later civil rights campaigns.

In the immediate postwar period, Katznelson convincingly argues, the GI Bill widened further the economic and social differences between the races. Southern segregation meant that educational opportunities available to whites were withheld from blacks, who were forced to compete for a very limited number of places in all-black institutions. Even in the North many colleges and universities either excluded blacks or admitted only a handful. GI loans for buying houses or financing small businesses were very difficult for blacks to obtain because of the discriminatory policies of banks and other lending agencies. Katznelson concludes that most government social policies during the 1930s and 1940s were, in effect, part of a vast affirmative action program for whites that left blacks further behind than they had been at the beginning of the period. He makes a chilling case.


Katznelson is somewhat more effective in describing the problem than in suggesting how to solve it. The general principle behind the kind of affirmative action that he recommends is clear enough: “Under affirmative action,” he writes,

[blacks should be] compensated not for being black but only because they were subject to unfair treatment at an earlier moment because they were black.

In an effort to fulfill the requirements that Justice Powell prescribed in the Bakke case, Katznelson offers two possible approaches. One would have the government identify all the people, or their immediate descendants, who were injured by exclusions from the various social and economic programs of the 1930s and 1940s. The government would calculate how much they would have gotten had they not been left out, and pay it to them in a lump sum.

Acknowledging that such a program would be “administratively burdensome” (clearly an understatement), he proposes an alternative, an all-out assault on poverty in general, based on the assumption that most of the people who are currently poor have been put at a disadvantage by the unjust policies of the 1930s and 1940s. To some extent, the program he favors would function like the GI Bill with “subsidized mortgages; generous grants for education and training; and active job searching and placement.” Health insurance and childcare could also be provided.

What is striking and somewhat unexpected is that these proposals do not depend on racial categories. All who suffered from policies of exclusion, whether in education or jobs, including some whites, would be entitled to compensation. It occurs to me that Katznelson’s essentially colorblind proposals, especially his second approach, could easily be justified on other grounds than as an antidote to racial inequality. If one simply assumed, as a good social democrat would, that poverty itself is an evil and that wealth should be more equitably distributed, similar policies could be justified. What he may be implying is that white America needs to face up, psychologically as well as materially, to its current as well as its past history of racial oppression, and that basing a colorblind antipoverty program on the need to redress racial discrimination will further this goal. But it remains unclear to me how Katznelson’s proposals would differ substantially in practice from those of William Julius Wilson, who distinguishes between affirmative action, which he defines as a way of producing more diversity among elites, and the kind of class-based assault on economic inequality that he believes is needed to raise up what he calls “the truly disadvantaged.”5

A question that can be raised about the adequacy of Katznelson’s historical analysis may also have implications for our understanding of the current prospects for racial equality. Are the South and Southern politicians as fully to blame for the increase in white advantages as he contends? Arguably the most important source of the current economic gulf between the races is the vast difference in average net worth or property ownership. Although average black incomes may be around two thirds those of whites, their average net worth, as Katznelson shows, is only about one tenth. Much of this difference is explained by the fact that whites own far more homes than blacks and therefore their net worth is higher.

How did this vast inequality come about? It was mainly the result of the greater white access to home mortgages that were insured and subsidized by the federal government. Before the 1930s a home buyer had to put down 50 percent of a house’s price and could get only a relatively short-term mortgage, perhaps only ten years. By the 1950s, as a result of a series of federal housing programs, including the GI Bill, most Americans could get long-term mortgages—up to thirty years—with a down payment as low as 10 percent. By 1984 seven out of ten whites owned their own homes, worth on average $52,000. But only one in four blacks owned a home, worth, on average, less than $30,000.

Katznelson outlines these facts toward the end of his book, and they illustrate dramatically his general point about the widening economic gulf between the races during the middle decades of the twentieth century. But he makes no effort to explain them as manifestations of Southern influence within the Democratic Party. The advantages of whites over blacks that he’s describing were more characteristically Northern than Southern; they manifested themselves in the growth of virtually all-white suburbs outside the major cities and virtually all-black ghettos within them.6

This new form of racial segregation was not simply the product of private choices, among them the refusal of white home-owners to sell to blacks, blockbusting and the racial “steering” of home buyers by real estate agents, and the personal prejudice of bankers asked to approve loans for blacks. The urban segregation that has contributed so much to the persistence of black inequality came about in large part because between the 1930s and the 1970s federal housing agencies refused to approve mortgage loans in neighborhoods that were “redlined,” which meant property values were deemed uncertain because of the presence of blacks.7

It is difficult to see the hand of Southern segregationists in these policies. It seems that Northern politicians were responding more directly to the racist attitudes of Northern whites who refused to live close to blacks. They were in effect underwriting the spatial segregation of the metropolitan North. It is not entirely clear how Katznelson’s proposal would try to rectify this aspect of affirmative action that benefited whites. Perhaps people could be compensated for the mortgages they were denied; but this would be extraordinarily difficult and would omit those who did not apply for mortgages because they expected they would be turned down.

Also in need of clarification is whether Katznelson’s attempt to justify large affirmative action programs for blacks applies to other minorities that are not black. He says virtually nothing about them. Since Mexican-Americans in the Southwest during the New Deal era were, like blacks in the South, disproportionately servants and farm laborers, they were similarly excluded from coverage by social security and labor rights legislation. Many of the same factors that make African-Americans eligible would thus apply to Chicanos, or at least to those who were in the United States between the 1930s and the 1950s, and to their descendants. But what about more recent Latino immigrants? They cannot claim as forcefully as blacks can that they were historically denied opportunities, such as obtaining mortgages, that were open to non-Hispanic whites.

A case can in fact be made that affirmative action was stretched out of shape and rendered incoherent when it was extended beyond African-Americans, Indians, and long-resident Mexican Americans to include recent nonwhite immigrants.8 Katznelson might agree with this view, but he does not address the question specifically, or even mention Latinos or other non-black ethnic groups (an omission that will be particularly striking, for example, to readers who live in multicultural California). But Katznelson’s book makes as strong a case as I have ever seen made for vigorous action to bring about equal opportunities for African-Americans.

This Issue

November 17, 2005