Do Schools Make a Difference?

Since the days of Thomas Jefferson, we have believed that education is a means of achieving equality in our society. But in the last few years, social science has brought that assumption into question. The battle now rages; controversy surrounds the participants—Jencks, Moynihan, Coleman, Pettigrew, Jensen, Herrnstein, Armor; and where it all will end matters not just to the experts, but to all of us.

THE Atlantic

FOUNDED IN 1857

by Godfrey Hodgson

The day Daniel Patrick Moynihan arrived at Harvard in the spring of 1966, he met some of his new colleagues at the Faculty Club in Cambridge. One of those present that evening was Professor Seymour Martin Lipset of the Harvard government department. “Hello, Pat,” said Lipset, “guess what Coleman’s found?”

“Coleman” was James S. Coleman, professor of social relations at Johns Hopkins, who had been charged by the Johnson Administration with conducting an extensive survey of “the lack of availability of equal educational opportunities” by reason of race, religion, or national origin. And what the Coleman survey had found, as Lipset paraphrased its voluminous findings, could hardly have come as more of a surprise. He had found, as Lipset told Moynihan excitedly, that “schools make no difference; families make the difference.”

Some six years later, Moynihan arrived a few minutes late for lunch with a friend at the same club, in a mood of jubilant intellectual pugnacity unusual even for him. Both the delay and the mood, he explained, resulted from a demonstration he had run into on his way across Harvard Yard from a class. Some students were handing out leaflets. It was their content which had produced Moynihan’s mood of sardonic amusement. “Christopher Jencks,” they said, “is a tool of reactionary American imperialist capitalism.”

Christopher Jencks a tool of capitalism? In the dozen years since he graduated from the Harvard Graduate School of Education, where he is now an associate professor, Sandy Jencks (as he is called) had moved perceptibly from the liberal toward the radical position. While an editor of The New Republic he began working with the distinctly New Left Institute for Policy Studies in Washington. He got into the neighborhood community control thing, and he helped to found the Cambridge Institute, which looks for “alternative visions of the American future”: decidedly one of the rising intellectual reputations on the American left. Now he has written a book, Inequality, in association with other researchers, working in large part from the same Coleman report data which, in Lipset’s words, showed that “schools make no difference.”

Pat Moynihan had started on the left, too. Trained as an orthodox social scientist, he grew up among liberals, and then discovered a most unorthodox flair for polemical prose and persuasive speech. After a political apprenticeship working for Governor Averell Harriman of New York, he took office as a liberal intellectual in good standing as an Assistant Secretary of Labor in the Kennedy Administration, and stayed on for a period under Johnson.

For some years now, however, he could hardly have been called a man of the left. His intellectual voyage can be dated from the publication of his report on the Negro family in 1965. Many liberals and blacks reacted with outrage to his dour assessment of the likelihood that orthodox liberal policies could eliminate the problems of the black under-class. He was outraged in turn by what he perceived as the liberals’ dishonest and antiintellectual refusal to follow where social science led. And by the time he returned to Washington in 1969 to serve Richard Nixon as a Cabinet-rank Counselor, he could no longer be called a man of the left at all. In recent years he has in fact established himself as the shrewdest strategist and most flamboyant impresario of an intellectual movement which can perhaps be called neo-conservative— though Moynihan maintains that it is radical. Whatever the label, it is almost contemptuously skeptical of the New Left and of conventional liberal shibboleths alike.

Moynihan’s amusement at seeing his colleague Jencks leafleted in Harvard Yard was not due to malice or Schadenfreude. On the contrary, it seemed to him to confirm that the argument between him and his friends and the left was over: that he had won. “Jencks ends up,” he told me, “where Richard Nixon was in 1969.”

Christopher Jencks does not see it that way. In fact, it is a strange kind of argument: one in which the participants largely agree about the Coleman survey, greatly as it surprised them when they first grasped it, but disagree, sometimes vehemently, on what it implies. The fight calls into question certain propositions which, until the Coleman report, few social scientists and few liberals dreamed of doubting: principally, that one of the main causes of inequality in American life has been inequality in education; and that education could be used as a tool to reduce inequality in society. The crucial role which education has been assigned in the United States is under heavy challenge. Is there now to be a retreat from the traditional faith in education as a tool of social change in America?

Since the days of Horace Mann and John Dewey—indeed since the days of Thomas Jefferson, that child of the Enlightenment—education has occupied a special place in the optimistic vision of American progressives, and of many American conservatives, for that matter. As the historian David Potter pointed out in People of Plenty, the American left, encouraged by the opportunities of an unexhausted continent and by the experience of economic success, has always differed sharply from the European left in that it has generally assumed that social problems could be resolved out of incremental growth: that is, that the life of the have-nots could be made tolerable without taking anything from the haves. Education has always seemed one of the most acceptable ways of using the national wealth to provide opportunity for the poor without offending the comfortable. As a tool of reform, education had the advantage that it appealed to the ideology of conservatives, to that ethic of self-improvement which stretches back down the American tradition through Horatio Alger and McGuffey’s Readers to Benjamin Franklin himself. This was particularly true in the age of the Great Migration. The public school systems of New York and other cities with large immigrant populations really did provide a measure of equality of opportunity to the immigrant poor. By the time the New Deal coalition was formed (and educators of one sort and another were to be a significant part of that coalition), these assumptions about education were deeply rooted. And they were powerfully reinforced, and virtually certified with the authority of social science, by the Supreme Court’s 1954 desegregation decision in Brown v. The School Board of Topeka. Plessy v. Ferguson, the 1896 Supreme Court decision by which statutory and customary segregation in the South were reconciled with the Thirteenth, Fourteenth, and Fifteenth Amendments, was not a school case. (As it happens, it concerned segregation on a Lake Pontchartrain ferry steamer.) But when, in the late 1930s and the 1940s, the NAACP, its lawyers, and its allies began to go to court to lay siege to segregation, they deliberately, and wisely, chose education as the field of attack. This was not accidental; they well knew that education was so firmly associated with equality in the public mind that it would be an easier point of attack than, say, public accommodations or housing. Not coincidentally, they worked their way up to the main citadel of the 1954 Brown decision by way of a series of law school cases: lawyers would find it hard to deny that segregation in law school was irrelevant to success in professional life.

In Brown, the NAACP’s lawyers deployed social science evidence in support of their contention that segregated education was inherently unequal, citing especially work done by psychologists Kenneth and Mamie Clark with black children and black and white dolls. The Clarks’ conclusions were that segregation inflicts psychological harm.

The historical accident of the circumstances in which school segregation came to be overthrown by the Supreme Court contributed to the currency of what turned out to be a shaky assumption. The great majority of American liberals, and this included large numbers of judges. Democratic politicians, and educators, came to suppose that there was incontrovertible evidence in the findings of social science to prove not just that segregated education was unequal but that if you wanted to achieve equality, education could do it for you. Or, to put the same point in a slightly different way, the prominence given to footnote 11 in the Brown judgment, which listed social science research showing that education could not be both separate and equal, had the effect of partially obscuring the real grounds for overthrowing segregation, which were constitutional, political, and moral.

Then a contemporary development put education right at the center of the political stage. President Johnson’s “Great Society” was to be achieved without alienating the power structure and, above all, the Congress. Education was an important part of the Great Society strategy from the start, but as other approaches to reducing poverty and racial inequality, notably “community action,” ran into political opposition, they fell apart, and so the proportional emphasis on educational programs in the Great Society scheme grew. In the end, the Johnson Administration, heavily committed to reducing inequality, was almost equally committed to education as one of the principal ways to do it.

Each of the events and historical developments sketched here increased the shock effect of the Coleman report—once its conclusions were understood. A handful of social scientists had indeed hinted, before Coleman, that the effect of schools on equality of opportunity might have been exaggerated. But such work had simply made no dent in the almost universal assumption to the contrary.

James Coleman himself has confessed he does not know exactly why Congress, in section 402 of the Civil Rights Act of 1964, ordered the Commissioner of Education to conduct a survey “concerning the lack of availability of equal educational opportunities for individuals by reason of race, color, religion or national origin.” The most likely reason is that Congress thought it was setting out to document the obvious in order to arm the Administration with a public relations bludgeon to overcome opposition. Certainly James Coleman took it for granted that his survey would find gross differences in the quality of the schools that black and white children went to. “The study will show.” he predicted in an interview more than halfway through the job, “the difference in the quality of schools that the average Negro child and the average white child are exposed to. You know yourself that the difference is going to be striking.”

He was exactly wrong. Coleman was staggered— in the word of one of his associates—to find the lack of difference. When the results were in, from about 600‚000 children and 60,000 teachers in roughly 4000 schools, when they had been collected and collated and computed, and sifted with regression analysis and all the other refinements of statistical science, they were astonishing. A writer in Science called them “a spear pointed at the heart of the cherished American belief that equality of educational opportunity will increase the equality of educational achievement.”

What did the figures say? Christopher Jencks later picked out four major points:

(1) Most black and white Americans attended different schools.

(2) Despite popular impressions to the contrary, the physical facilities, the formal curricula, and most of the measurable characteristics of teachers in black and white schools were quite similar.

(3) Despite popular impressions to the contrary, measured differences in schools’ physical facilities, formal curricula, and teacher characteristics had very little effect on either black or white students’ performance on standardized tests.

(4) The one school characteristic that showed a consistent relationship to test performance was the one characteristic to which poor black children were denied access: classmates from affluent homes.

Here is how James Coleman himself summed up the 737 pages of his report (not to mention the additional 548 pages of statistical explanation):

Children were tested at the beginning of grades 1, 3, 6, 9 and 12. Achievement of the average American Indian, Mexican American, Puerto Rican, and Negro (in this descending order)1 was much lower than the average white or Oriental American, at all grade levels . . . the differences are large to begin with, and they are even larger at higher grades. Two points, then, are clear: (1) these minority children have a serious educational deficiency at the start of school, which is obviously not a result of school, and (2) they have an even more serious deficiency at the end of school, which is obviously in part a result of school.

Coleman added that the survey showed that most of the variation in student achievement lay within the same school, and very little of it was between schools. Family background—whatever that might mean—must, he concluded, account for far more of the variation in achievement than differences between schools. Moreover, such differences as could be attributed to the schools seemed to result more from the social environment (Jencks’s “affluent classmates,” and also teachers) than from the quality of the school itself.

This was the most crucial point. For if quality were measured, as it had tended to be measured by administrators and educational reformers alike, in material terms, then the quality of the school, on Coleman’s data, counted for virtually nothing.

When other things were equal, the report said, factors such as the amount of money spent per pupil, or the number of books in the library, or physical facilities such as gymnasiums or cafeterias or laboratories, or even differences in the curriculum, seemed to make no appreciable difference to the children’s level of achievement. Nothing could have more flatly contradicted the assumptions on which the Administration in Washington, and urban school boards across the country, were pouring money into compensatory education programs.

As we shall see, the report exploded with immense force underground, sending seismic shocks through the academic and bureaucratic worlds of education. But on the surface the shock was not at first apparent. There were two main reasons for this. The first was that the report was, after all, long, tough, dry, and technical. It had been written in five months in order to comply with a congressional deadline, and it therefore made no attempt to point a moral or adorn a tale: it was essentially a mass of data. All of these characteristics militated against its being reported in detail by, for example, the Associated Press, the source from which most American newspapers get most of their outof-town news.

The Office of Education, which realized all too clearly how explosive the report was, didn’t exactly trumpet the news to the world. The report was released, by a hallowed bureaucratic stratagem, on the eve of July 4, 1966. Few reporters care to spend that holiday gutting 737 pages of regression analysis and standard deviations. And to head off those few who might have been tempted to make the effort if they guessed that there was a good story at the end of it, the Office of Education put out a summary report which can only be described as misleading. “Nationally,” it said, to take one example, “Negroes have fewer of some of the facilities that seem most related to academic achievement.” That was true. But it was not the significant truth.

The point was that the gap was far smaller than anyone expected it to be. To take one of the summary report’s own examples, it was true that Negro children had “less access” to chemistry labs than whites. But the difference was that only 94 percent of them, as compared to 98 percent of whites, went to schools with chemistry labs. That was hardly the kind of difference which could explain any large part of the gap between white and black achievements in school, let alone that larger gap, lurking in the back of every educational policy maker’s mind, between the average status and income of blacks and whites in life after they leave school.

A few attempts were made to discredit the survey. But the Coleman findings were in greater danger of being ignored than of being controverted when, at the beginning of the academic year in the fall of 1966, Pat Moynihan began to apply his talents to make sure that the report should not be ignored. He and Professor Thomas Pettigrew of the Harvard School of Education organized a Seminar on the Equality of Educational Opportunity Report (SEEOR). The seminar met every week at the Harvard Faculty Club, and by the end more than eighty people had taken part.

Moynihan had taken the precaution of getting a grant for expenses from the Carnegie Corporation, some of which was laid out on refreshments, stronger than coffee or cookies. “It was quite something, that seminar,” says Jencks, reminiscing. “Pat always had the very best booze and the best cigars.” But if Moynihan is a connoisseur of the good things in life, he also knows how to generate intellectual excitement, or to spot where it is welling up.

“When I was at the School of Education ten years ago,” Jencks says, “almost nobody who was literate was interested in education. The educational sociologists and psychologists, the educational economists, they were all pretty near the bottom of the heap. Suddenly that’s changed.”

“That seminar taught me something about Harvard,” Moynihan says. “People here are not interested in a problem when they think it’s solved. There are no reputations to be made there. But when something which people think was locked up opens up, suddenly they all want to get involved.” People started coming up to Moynihan in Harvard Yard and asking if they could take part: statisticians, economists, pediatricians‚ Professor Abram Chayes from the Law School (and the Kennedy State Department). Education had become fashionable. Jason Epstein of Random House and Charles E. Silberman from Fortune magazine started coming up from New York.

Harvard had seen nothing quite like it since the arms control seminars of the late 1950s, at which the future strategic policies of the Kennedy Administration were forged and the nucleus of the elite that was to operate them in government was brought together. In the intervening decade, domestic social questions had reasserted their urgency. Education had emerged as the field where all the agonizing problems of race, poverty, and the cities seemed to intersect.

If schools, as Seymour Martin Lipset paraphrased Coleman, “make no difference,” what could explain the inequalities of achievement in school and afterwards? One school of thought was ready and waiting in the wings with an answer. In the winter of 1969, the following words appeared in an article in the Harvard Educational Review:

There is an increasing realization among students of the psychology of the disadvantaged that the discrepancy in their average performance cannot be completely or directly attributed to discrimination or inequalities in education. It seems not unreasonable, in view of the fact that intelligence variation has a large genetic component, to hypothesize that genetic factors may play a part in this picture.

The author was Professor Arthur Jensen, not a Harvard man, but an educational psychologist from Berkeley with a national reputation. He had jabbed his finger at the rawest, most sensitive spot in the entire system of liberal thinking about education and equality in America. For after more than a generation of widespread IQ testing, it is an experimental finding, beloved of racists and profoundly disconcerting to liberals, that while the average white IQ is 100, the average black IQ is 85. Racists have seen in this statistical finding confirmation of a theory of innate biological inferiority. Conservatives have seen in it an argument against heavy expenditures on education, and against efforts to desegregate. And liberals have retorted that the lower average performance of blacks is due either to cultural bias in the tests used or to unfavorable environmental factors which require redoubled efforts on the part of social policy makers.

Jensen marched straight into the fiercest of this cross fire. He argued two propositions in particular in his article: that research findings suggest that heredity explains more of the differences in IQ between individuals than does environment, and that heredity accounts for the differences between the average IQ’s of groups as well as between those of individuals.

The article was scholarly in tone. In form it was largely a recital of research data. And it was tentative in its conclusion that perhaps more of the differential between blacks’ and whites’ average IQ’s was due to heredity than to environment. That did not stop it from causing a most formidable rumpus. It became a ninety days’ wonder in the press and the news magazines. It was discussed at a Cabinet meeting. And Students for a Democratic Society rampaged around the Berkeley campus chanting “Fight racism! Fire Jensen!”

Two years later, a long article in The Atlantic by Professor Richard Herrnstein on the history and implications of IQ provoked a reaction which showed that the sensitivity of the issue had by no means subsided. Herrnstein touched only gingerly on the racial issue. “Although there are scraps of evidence for a genetic component in the blackwhite difference,” he wrote, “the overwhelming case is for believing that American blacks have been at an environmental disadvantage ... a neutral commentator (a rarity these days) would have to say that the case is simply not settled, given our present stage of knowledge.”

Neutral commentators certainly proved rare among those who wrote in to the editor. Arthur Jensen wrote to say that Herrnstein’s essay was “the most accurately informative psychological article I have ever read in the popular press”; while a professor from the University of Connecticut said: “This is not new. Hitler’s propagandists used the same tactics in the thirties while his metal workers put the finishing touches on the gas ovens.”

If Herrnstein—understandably enough—tiptoed cautiously around the outskirts of the black-white IQ argument, he charged boldly enough into another part of the field. The closer society came to its ideal of unimpeded upward social mobility, the closer he predicted it would come to “meritocracy,” a visionary state of society described by the British sociologist Michael Young. A new upper class composed of the descendants of the most successful competitors with the highest IQ’s would defend its own advantage far more skillfully and successfully than did the old aristocracies. Herrnstein did not welcome this trend; he merely argued that it might be inevitable. “Our society may be sorting itself willy-nilly into inherited castes,” he concluded gloomily. Or, as his Harvard colleague David K. Cohen neatly epigrammatized Herrnstein’s long article in a rejoinder in Commentary, “His essay questioned the traditional liberal idea that stupidity results from the inheritance of poverty, contending instead that poverty results from the inheritance of stupidity.”

Cohen went on to disagree with Herrnstein’s prediction. “America is not a meritocracy,” he wrote, “if by that we mean a society in which income, status, or power are heavily determined by IQ. . . . Being stupid is not what is responsible for being poor in America.”

But that still left the original question open.

If differences in the quality of schools, as measured by money, facilities, and curricula, don’t explain inequality, because the differences between the schools attended by children of different racial groups are simply not that great in those respects, then what does? Genetic differentials in IQ, perhaps, says Jensen. Nonsense, says a majority of the educational community; the explanation is more likely to be integration—or rather the lack of it.

“I’m a Southern liberal,” says Tom Pettigrew. “There are only about thirty of us, and my wife says we all know each other.” Pettigrew comes from Richmond, Virginia, but his father immigrated from Scotland, and there is something about him that strikes one as more typically Scots than Southern. He is a shy man with a passion for methodological precision: “I really believe that data can free us,” he says. He also has a deceptive, because quiet, commitment to the liberal faith.

The Coleman report gave only three pages to the effects of desegregation, and Pettigrew didn’t think that was enough. At Jim Coleman’s explicit insistence, the data bank of the survey was to be made generally available for the cost of the computer tapes. Pettigrew persuaded the Civil Rights Commission to take advantage of this and to reanalyze the data to see what light it cast on the effects of desegregation. David Cohen and Pettigrew were the main authors of the resulting survey, which came out in 1967 as Racial Isolation in the Public Schools and gave the impression that the Coleman data supported desegregation. This was true up to a point. Coleman had concluded that desegregation did have an effect. But his report also showed that social class had a greater effect. Pettigrew is not much troubled by this, because of the close connection between race and social class in America. “Two-thirds of the whites are middleclass,” he says, “and two-thirds of the blacks are working-class.”

Pettigrew also draws a sharp distinction between desegregation and integration. By integration he means an atmosphere of genuine acceptance and friendly respect across racial lines, and he believes that mere desegregation won’t help blacks to do better in school until this kind of atmosphere is achieved. He is impressed by the work of Professor Irwin Katz, who has found that black children do best in truly integrated situations, moderately well in all-black situations, and worst of all in “interracial situations characterized by stress and threat.”

Pettigrew believes, in other words, that integration, as opposed to mere desegregation, will be needed to bring black children’s achievement up to equality with whites’. And he argues that no one can say that integration hasn’t worked, for the simple reason that it hasn’t been tried.

“The U.S. is going through a period of self-flagellation,” he said to me. “I dispute the argument that Moynihan is forever putting out. He says liberalism was tried and didn’t work.” This, as we shall see, misstates Moynihan’s views. The difficulty is partly semantic. Moynihan believes that past policies, which can be called “liberal,” have “worked” in the sense that they have produced a surprising degree of equality in terms of all the resources that go into schools, without, however, achieving equality of outcome. “I say liberalism hasn’t been tried,” Pettigrew goes on. “Racial integration has yet to be tried in this country.” Desegregation proceeded so slowly, Pettigrew says, that the courts “got mad and started ruling for busing in 1969 and 1970.” Until desegregation is achieved, he argues, we won’t know whether integration works.

The Civil Rights Commission’s report on racial isolation did recommend that the federal government set a national standard that no black children should go to a school that was more than 50 percent black. In practical terms, that meant busing. And, in fact, Pettigrew argues that some busing will be needed to achieve desegregation—and thus to produce the physical circumstances in which integration as he understands it can take place. He has been actively involved as a witness in several desegregation suits in which he has advocated busing.

It is, therefore, as Pettigrew himself wryly remarks, an irony that he should have suggested to one of his junior colleagues at the Harvard School of Education that he do a study on busing. The colleague’s name was David Armor, and Pettigrew’s idea was that it would be interesting to take a look at Project Metco, a scheme for busing children out of Roxbury, the main Boston ghetto, into nearby white suburban schools.

That was in 1969. Three years later, a paper by David Armor called “The Evidence on Busing” was published in The Public Interest. Armor said he had concentrated on the question of whether “induced integration”—that is, busing—“enhances black achievement, self-esteem, race relations, and opportunities for higher education.” In a word‚ Armor maintained that it did not.

The article used data not only from Project Metco but from reports of four other Northern programs for induced integration: in White Plains, New York; Ann Arbor, Michigan; Riverside, California; and New Haven and Hartford, Connecticut. And on the basis of this data,2 Armor maintained that “the available evidence . . . indicates that busing is not an effective policy instrument for raising the achievement of blacks or for increasing interracial harmony.”

“None of the studies,” said Armor, “were able to demonstrate conclusively that integration has had an effect on academic achievement as measured by standardized tests.” Aspirations, indeed, were high among the black children in Project Metco. But they might be too high, in view of the fact that, while 80 percent of them started college, half of them dropped out. As for race relations, Armor found the bused students not only more militant but actually more hostile to integration than the study’s “control group,” which was not bused. Militancy, as measured, for example, by sympathy with the Black Panthers, seemed to be particularly rife among those children who had high aspirations (such as going to college) but were getting C grades or below in competitive suburban high schools.

But Armor did not limit himself to reporting the results of his own Metco study and the other four studies. His article was a sweeping, slashing attack on the whole tradition of liberal social science. He described what he called the “integration policy model,” based on social science research going back to the time of John Dollard and Gunnar Myrdal. Though the “real goals of social science and public policy are not in opposition,” Armor said, he claimed that almost all of the “major premises of the integration policy model are not supported by the data”—by which he meant the studies he quoted.

It was a frontal assault on the liberal tradition in the social sciences for a generation: on “forty years of studies,” as one of his opponents put it. At one point Armor came close to accusing his opponents of deliberate dishonesty: “There is the danger that important research may be stopped when the desired results are not forthcoming. The current controversy over the busing of schoolchildren affords a prime example.”

It was not likely that such an attack would go unanswered, and, in fact, the response was both swift and severe. Pettigrew and three colleagues fired back a critique which called Armor’s article “a distorted and incomplete review.” To back up their charge, they argued that the studies Armor had cited as “the evidence on busing” were highly selective. Armor had not discussed seven other studies which they said met his own methodological criteria—from New York, Buffalo, Rochester, Newark, Philadelphia, Sacramento, and North Carolina—surveys which had reported positive achievement results for bused blacks. The integrationists also found what they claimed were disastrous weaknesses in Armor’s own Metco study. For one thing, they said, he compared the bused children with a control group which included children who were also attending desegregated schools, though not under Project Metco. “Incredible as it sounds,” Pettigrew and his colleagues commented, “Dr. Armor compared children who were bused to desegregated schools with other children many of whom were also bused to desegregated schools. Not surprisingly he found few differences between them.”

“We respect Dr. Armor’s right to publish his views against mandatory busing,” they said. “But we challenge his claim that those views are based on scientific evidence.” (Armor is replying to the critique in The Public Interest, the journal which published his paper.)

If the tone of the public controversy sounds rough, it was positively courtly compared to the atmosphere inside William James Hall, the new Harvard high-rise where Pettigrew and Armor had their offices, two doors apart.

Armor, too, started out on the left. He was president of the student body at Berkeley in 1959-1960, and head of SLATE, a forerunner of the radical Free Speech Movement there. He was also a protégé of Pettigrew’s at Harvard, and indeed a close friend. But by the spring of 1972, Pettigrew realized that Armor had become vehemently opposed to mandatory busing.

Both men became very bitter. Armor failed to get tenure at Harvard, and has now moved to a visiting professorship at UCLA. Armor accused one of Pettigrew’s assistants of breaking into his office to steal his Metco data. Pettigrew wrote to the New York Times: “There is no evidence beyond the allegation itself for the charge, much less any link between the paper’s critics and the alleged intrusion.” Armor accused Pettigrew of suppressing his paper; Pettigrew does concede that he told Armor that he had done “incomparable harm” by publishing it.

Tempers, in short, were comprehensively lost over the Armor affair. Much of the bitterness, no doubt, must be put down to personal factors. But it would be wrong to dismiss the episode as a mere squabble between professors. For it shows just how traumatically a world where consensus reigned half a dozen years ago has been affected by the pressure to abandon certain cherished premises. And the issue, after all, is the interrelationship of education, race, and equality in America, which is not exactly a recondite academic quibble.

To an unbiased eye (“a rarity these days,” as Richard Herrnstein might say), Armor’s paper has been rather seriously impugned. It does not follow that his central thesis is entirely discredited. Even Pettigrew was quoted, at the height of the row, as saying that “nobody is claiming that integration has been a raving success.” “That’s not what they were saying before,” says Moynihan. And Christopher Jencks, who can hardly be accused of conservative prejudice, has summed up the evidence in the most cautious and equivocal way. Blacks, he says, might do much better in “truly integrated schools, whatever they may be.” Failing that consummation, devoutly to be wished, the benefits of desegregation appear to be spotty, and busing can be expected to yield contradictory results.

Jencks’s position is easily misunderstood. In an interview, he drew some distinctions for me. He reminded me that he had himself written that the Coleman report “put the weight of social science behind integration.” It was not until Armor’s article was published, he said, that social scientists began to argue that desegregation itself might not work. Jencks personally feels that Armor’s data were shaky, and that the effect of Armor’s paper came from its review of other studies—a review which, as Pettigrew pointed out, does not refer to all the available studies. Jencks himself thinks that desegregation is probably necessary, simply in order to meet the constitutional requirements of the Fourteenth Amendment, in virtually every urban school district in the country. He does, however, have personal reservations about mandatory busing, on libertarian grounds. The furthest he would go was to say, “I think that a case can be made out that busing might be a useful part of an overall strategy of desegregation.” That is not to say that he has any tenderness toward segregation. On the contrary, he rejects it as absolutely as any of the “integrationists.” The difference is that Jencks does not think that segregation explains nearly as much of existing inequality as the integrationists think it does.

But with Armor’s paper and its reception, we are getting ahead of the story. The Coleman report came out in 1966. It was not until 1972 that two major books appeared, each an attempt to reassess the whole question of the relationship between education and equality in America in the light of the Coleman data. Each was collaborative.

The first of these two books was the Random House collection of papers arising out of the SEEOR seminar, which was published as On Equality of Educational Opportunity, with Frederick Mosteller (professor of mathematical statistics at Harvard) and Daniel Patrick Moynihan as co-editors. Most of the leading participants in the debate contributed chapters: Pettigrew and Armor, Coleman, David Cohen, and Christopher Jencks among them. The introductory essay was signed jointly by Mosteller and Moynihan. If much of the technical analysis and of the drafting were Mosteller’s, the essay’s style and conclusions are vintage Moynihan.

Later in the year, Christopher Jencks and seven of his colleagues (two of whom, Marshall Smith and David Cohen, had also contributed to the Mosteller-Moynihan volume) published an only slightly less massive book: Inequality: A Reassessment of the Effect of Family and Schooling in America. This work displays considerably more intellectual cohesion than the Mosteller-Moynihan book, presumably because Jencks actually wrote his group’s text himself from start to finish and, according to the preface, it “embodies his prejudices and obsessions, and these are not shared by all the authors.” But again, though the book draws upon data from dozens of other largeand smallscale surveys, the data from the Coleman survey are the bedrock and foundation.

The enormous body of analysis and reinterpretation in these two books represents the completion of the first stage of the reaction to Coleman. I began by quoting Professor Lipset’s hasty shorthand for Coleman’s central discovery: “schools make no difference.” Professor Pettigrew draws an important distinction. “Never once was it said that schools make no difference. The belief that Coleman hit was the belief that you could make a difference with money.” (He added: “Americans are crazy in the head about money; they think you can do everything with money.”) However that may be, the nub of the discovery that has set off the whole prolonged, disturbing, confusing, sometimes bitter debate can be expressed as a simple syllogism:

(1) The “quality” of the schools attended by black and white children in America was more nearly equal than anyone supposed. (2) The gap between the achievement of black and white children got wider, not narrower, over twelve years at school. (3) Therefore there was no reason to suppose that increasing the flow of resources into the schools would affect the outcome in terms of achievement, let alone eliminate inequality.

Among the social scientists, the central ground of debate about the meaning of those findings now lies between Jencks and Moynihan. It is a strange debate, for the two protagonists have much in common, even if one does have New Left loyalties, and the other served in Nixon’s White House and now as Nixon’s Ambassador to India. Both use the same data. Indeed, the spectacle of social scientists reaching into the same data bank for ammunition to fire at each other is sometimes reminiscent of war between two legs of the same octopus. Both agree on many of the implications of the data, and on many of the conclusions to be drawn from them. Yet those who lump the two professors together, as many practical educators and civil rights lawyers do, as “Moynihan and Jencks and those people up at Harvard,” could hardly be more wrong. The two men are divided by temperament and ideology in the preconceptions they bring to the data, and ultimately in the policy prescriptions they draw from that data.

Perhaps the very heart of their disagreement, after all, comes down to a matter of temperament. Is a glass half empty, or is it half full? A pessimist will say it is half empty when an optimist says it is half full. Pat Moynihan (and his co-author Mosteller—but I should be surprised if these particular thoughts were not Moynihan’s contribution, since they coincide with so much that he has said and written elsewhere) looked at the Coleman data and made the very reasonable inference that if the differences in quality between the schools attended by different groups of children in the United States were so much smaller than everyone had expected to find them, then the United States had come much closer to realizing the goal of equality of educational opportunity than most people realized. He then chose to relate this to the general question of social optimism versus social pessimism. At the time of the Coleman report’s publication, “a certain atmosphere of ‘cultural despair’ was gathering in the nation,” they wrote, “and has since been more in evidence. Some would say more in order. We simply disagree with such despair.”

One of the specific recommendations of the Mosteller-Moynihan essay is optimism. The electorate should maintain the pressure on government and school boards, the essay urges, “with an attitude that optimistically expects gains, but‚ knowing their rarity, appreciates them when they occur.” Yet on examination this is a strange use of the word optimism. For optimism normally connotes an attitude toward the future. But the emotion that is being evoked here has more to do with the past: it is not optimism so much as pride. “The nation entered the middle third of the twentieth century bound to the mores of caste and class. The white race was dominant. . . . Education beyond a fairly rudimentary point was largely determined by social status. In a bare third of a century these circumstances have been extensively changed. Changed! Not merely a sequence of events drifting in one direction or another. To the contrary, events have been bent to the national will.” True, the essay concedes, the period ended with racial tensions higher than ever before, and with dissatisfaction with the educational system approaching crisis. Nevertheless, say Moynihan and Mosteller, we should accentuate the positive. “It is simply extraordinary that so much has been done. . . . No small achievement! In truth, a splendid one. . . . It truly is not sinful to take modest satisfaction in our progress.”

Swept along by the dithyrambic rhythm of these tributes to past policies, it would be easy to conclude that Moynihan thinks they should be pressed to the utmost. But he does not. When I asked him why not, he replied promptly, if cryptically: “Production functions.” In an article in the fall, 1972, issue of The Public Interest, he spells out what he means. The argument is characteristically simple, forceful, and provocative.

Proposition 1: “The most striking aspect of educational expenditure is how large it has become.” It has now reached $1000 per pupil per annum, and it has been rising at 9.7 percent annually for the last ten years, while the GNP has risen 6.8 percent.

Proposition 2 (the Coleman point): Maybe not much learning takes place in a school without teachers or a roof. But “after a point school expenditure does not seem to have any notable influence on school achievement.”

There are, Moynihan concedes, considerable legional‚ class, racial, and ethnic variations in achievement, and he would like to see them disappear. “But it is simply not clear that school expenditure is the heart of the matter.”

This is where the production function, or what is more familiar to laymen as the law of diminishing returns, comes in, according to Moynihan. The liberal faith held that expenditure of resources on education would produce not merely a greater equality in scholastic achievement, but greater equality in society. On the contrary, says Moynihan, additional expenditure on education (and indeed on certain other social policies) is likely to produce greater inequality, at least of income.

The day the students leafleted Christopher Jencks in Harvard Yard, Moynihan said to me: “They’re defending a class interest.” What he meant was that as future teachers, or social workers‚ or administrators of education or social policies, left-wing students had a vested economic interest in the high-investment “liberal” policies they defended.

“Any increase in school expenditure,” Moynihan wrote in The Public Interest, “will in the first instance accrue to teachers, who receive about 68% of the operating expenditure of elementary and secondary schools. That these are estimable and deserving persons none should doubt”—Brutus is an honorable man—“but neither should there be any illusion that they are deprived.” With teachers earning some $10,000 a year on the average, he argues, and with many of them married women with well-paid husbands, “increasing educational expenditures will have the short-run effect of income inequality.”

As a matter of statistical fact, that may be literally true. But it is a peculiar argument nonetheless, for several reasons. For, leaving aside the matter of their spouses’ incomes, teachers are not, relatively, a highly paid group. Marginal increases in their salaries have an imperceptible effect on inequality in the national income distribution.

Whatever its merits, however, Moynihan’s position is plain. But it is worth noting that this position fits oddly with an exhortation to optimism. There is indeed nothing sinful about taking satisfaction in past progress; but when this attitude is combined with skepticism about the benefits to be expected from future public expenditure, it is usually called not optimistic but conservative.

Like Moynihan, Christopher Jencks is concerned with equality, not only in the schools but also in the world after school. The essence, and the originality, of his thinking lie in the use he makes of two crucial, though in themselves unoriginal‚ distinctions.

The first distinction is between equality of opportunity and equality of condition. Most Americans say they are in favor of equality. But what most of them mean by this is equality of opportunity. What we have learned from the Coleman report, says Jencks, and from the fate of the reforms of the 1960s, is that contrary to the conventional wisdom, you cannot have equality of opportunity without a good deal of equality of condition—now and not in the hereafter.

This is where the second of Jencks’s distinctions comes in. Where the Coleman survey, and most of the work published in the Mosteller-Moynihan volume, looked at the degree of equality between groups, Jencks is more interested in inequality between individuals. Coleman’s conception of equality looked at the distribution of opportunity between two groups. For Coleman, as Marshall Smith puts it, if you laid the distribution curve of one group over the distribution curve of the other, and they coincided exactly, then you could say that the two groups were equal. And Coleman found that between white and black Americans, this was closer to being true than most people had suspected. “Sandy Jencks is saying that though this may apply as between groups, this approximate equality disappears when you look at individuals.”

It is cause for shock, he says in the preface to his book, “that white workers earn 50 percent more than black workers.” But it is a good deal more shocking “that the best-paid fifth of all white workers earns 600 percent more than the worstpaid fifth. From this point of view, racial inequality looks almost insignificant”—by comparison with economic inequality.

Is the glass half empty, or half full? If Moynihan’s instinct is to emphasize the real progress that has been made toward reducing inequality in America, Jencks stresses how much inequality remains, not only in educational opportunity, in learning skills, and in educational credentials but also in job status, in job satisfaction, and in income.

The trouble is, he points out—and here I am summarizing an argument which is based, step by step, on mountains of statistical data—that whatever measure you take—income, socioeconomic status, or education—there is plenty of inequality among Americans. But the same people by no means always come out at the same point on each measure. In the social scientists’ terms, these different kinds of inequality don’t “correlate” very closely. It follows that school reform is not likely to effect much greater equality outside the school. The “factory model,” which assumes that the school’s outcome is the direct product of its inputs, must be abandoned, says Jencks. For him, a school is in reality more like a family than a factory.

This idea underlies a surprising strand in Jencks’s thought. If there is no direct correlation between expenditure on schools and effects on society—for example, in producing greater equality between racial groups—some would draw the lesson that it is not worth spending more than a (possibly quite high) minimum on schools. (That is something like Moynihan’s theoretical position, as we have seen.) No, says Jencks, spend more money; not because of the benefits it will bring in some sociological hereafter but simply because people spend something close to a fifth of their life in school, and it is better that they spend that time in a pleasant and comfortable environment.

“There is no evidence,” Jencks writes, “that building a school playground will affect the students’ chances of learning to read, getting into college, or earning $50,000 a year when they are fifty. Building a playground may, however, have a considerable effect on the students’ chances of having a good time during recess when they are eight.” And in a recent statement protesting the use of the conclusions which Inequality reaches “to justify limiting educational expenditures and abandoning efforts at desegregation,” Jencks writes that “educators will have to keep struggling,” and that “they need more help than they are currently getting.” But he concludes that the egalitarian trend in American education over the last thirty years has not made the distribution of either income or status outside the schools much more equal. He writes: “As long as egalitarians assume that public policy cannot contribute to economic equality directly, but must proceed by ingenious manipulation of marginal institutions like the schools, progress will remain glacial.”

“Marginal institutions like the schools”! The phrase sets Jencks every bit as far outside the old liberal orthodoxy as Moynihan’s suggestion that spending money on schools may actually increase inequality. Fourteen words from the end of his book Jencks unfurls a word which startles many of his readers. “If we want to move beyond this tradition, we will have to establish political control over the economic institutions that shape our society. That is what other countries usually call socialism. Anything less will end in the same disappointment as the reforms of the 1960s.”3

Norman Drachler was superintendent of the huge, tormented Detroit public school system from 1966 to 1971. When I talked to him recently, he was going through the anguish of liberal educators who had the intellectual honesty to try to reconcile the new teachings of the social scientists with the working assumptions of a lifetime of effort.

He showed me a headline from the New York Times of December 4, 1966, which perfectly summed up the pre-Coleman orthodoxy, WHEN SPENDING FOR EDUCATION IS LOW, it said, ARMY INTELLIGENCE TEST FAILURES ARE HIGH. And he showed me figures to prove that when federal money under Title I of the Elementary and Secondary Education Act of 1965 was concentrated on the schools with the greatest need in Detroit, reading scores improved by two months from 1965 to 1971, while the city-wide average declined by two months. “In the worst schools, Title I helped to arrest a disastrous fall,” says Drachler. “Where we spent more money, we did do better.”

How did he square this with the Coleman report?

“I think Coleman is basically correct. With better schools we can only make a small difference. But it is worth that investment.”

The post-Coleman challenge to the case for spending money on education is beginning to echo through the halls of Congress, ominously for the supporters of federal aid to education, who include both Representative John Brademas, Democrat of Indiana, the chairman of the House Select Subcommittee on Education, and one of his Republican colleagues‚ Representative Albert Quie of Minnesota. In a recent speech Quie has made it plain that he remains to be convinced that compensatory education makes no difference. John Brademas is afraid that the social science findings, misunderstood or deliberately misrepresented, will be used to justify savage cuts in federal aid to elementary and secondary education and to make opposition to such programs respectable. He is deeply skeptical of the case against the efficacy of educational spending, pointing out not only that federal aid still amounts to only 7 percent of the cost of elementary and secondary schooling but also that in many cases funds intended under Title I for compensatory education for underprivileged children have been indiscriminately spent for political reasons on middle-class children, so that few valid conclusions can be drawn from the experience of Title I. He feels adrift without adequate information, while the opponents of educational spending are able to use the social scientists’ evidence, often disingenuously. In his own reelection campaign in Indiana last fall he was amused, but not happy, to find his Republican opponent quoting what he called the “Colombo report” (meaning the Coleman report) at him.

Education lobbyists claim that the “Jencks report” has been freely cited by the Nixon Administration’s Office of Management and Budget on Capitol Hill in justification of the cuts in the fiscal 1974 budget. And even in some of the more conservative governors’ offices, one lobbyist for elementary and secondary education told me there is a widespread feeling that “Coleman and Jencks” have the effect of giving education a low priority.

Money is one issue; integration is another. Although‚ as Christopher Jencks put it to me, “the impact both of Coleman and of the MoynihanMosteller book is to put the support of social science behind integration,” and even though a majority of the social scientists who have spoken up remains integrationist, there is no mistaking the chill which the Armor paper, supported as it has been to some extent by various influential figures in the intellectual community, has sent down the spines of the integrationists. Last November, for example, Harold Howe, U.S. Commissioner for Education in the Johnson Administration (he is now with the Ford Foundation), conceded that “the lively researches of statistically oriented social scientists have cast some shadows on conventional assumptions about the benefits of integration, particularly in the schools.”

The first place where those shadows would fall is in the courts, which are now jammed with cases arising from the tough desegregation orders made by federal judges in all parts of the country since 1969, Integrationists insist that the law requires school desegregation under the Fourteenth Amendment, wholly independent of social science data regarding its effect. As former Chief Justice Earl Warren put it in a recent interview with Dr. Abram Sachar of Brandeis, Brown was a race case, not an education case. And so far the judges have upheld the principle that the requirement of desegregation in the law is independent of evidence about its effect.

But already the courts have begun to hear social science evidence about the equality of achievement in schools. In Keyes (the Denver school desegregation case which the Supreme Court has already heard, but on which it has not yet handed down its opinion), Judge William Doyle, in the district court, asked for evidence about the achievement of seventeen schools which he found to be segregated, though not as a result of public policy. James Coleman himself was one of the witnesses, and he testified that while compensatory education had proved disappointing, desegregation might be helpful.

David Armor was a witness on the other side in one of the Detroit desegregation hearings. But in the Memphis case, where his paper was produced in evidence, the court of appeals gave it short shrift. Judge Anthony Celebrezze (a former Democratic Secretary of HEW) dismissed it as “a single piece of much criticized sociological research,” and said “it would be presumptuous in the extreme for us to refuse to follow a Supreme Court decision on the basis of such meager evidence.”

Judicial reaction generally, says Louis Lucas, a Memphis lawyer who appeared for plaintiffs in both the Detroit and Memphis cases, “has been to say ‘a plague on both your houses’ to the social scientists. They have noticed how much criticism of the new findings there has been, and they say in effect, ‘We are not going to re-try Brown.’”

But that is exactly what less sanguine integrationists are afraid the Supreme Court will do, with respect to the most difficult Northern desegregation cases: not frontally, but by erosion. Norman Drachler‚ for example, told me he thought it very probable that the Burger Court would find some way to re-try Brown without seeming to do so. Nick Flannery, of the Harvard Center for Law and Education, told me that “the Burger Court will almost certainly be looking for distinctions to draw that will narrow the scope of Brown. ”

Flannery suggested some possibilities. The Court could adopt Judge Doyle’s argument (in the Denver case) that not all segregation results from public policy. Or it could adopt the Justice Department’s contention that the wrong to be remedied is not segregation itself, but discrimination, so that the plaintiff can get relief only when he can show not merely generalized segregation but particular instances of discrimination. In the Swann case, in 1971, having to do with Charlotte, North Carolina, Chief Justice Warren Burger laid down the principle that the scope of the remedy need not exceed the scope of the violation. That might seem to lay the groundwork for limiting Brown in this way. Alternatively, the Court might reverse the integrationist doctrine that has been developing in the lower courts, by imposing burdens of proof on the plaintiffs which would make the process of bringing a school desegregation case even lengthier and more expensive than it is already.

Some years ago, the great historian of the South, C. Vann Woodward of Yale, compared the civil rights movement of the 1960s to the Reconstruction period after the Civil War and said that he thought this second Reconstruction was ending. There is a parallel in the intellectual world that Woodward did not draw. The 1870s—the years of “reunion and reaction,” when the nation wearied of the political impasse created by white resistance to the Radicals’ drive for Negro equality—were also the years when American intellectual life was swept by the ideas of Herbert Spencer and his followers, the Social Darwinists. Their enthusiasm for ruthless competition that would drive the weakest to the wall, for “anarchy with a policeman” as the type of society most likely to produce the highest evolution of man, did much to rationalize and to justify public indifference as white supremacy reasserted itself after Reconstruction. The skepticism about the efficacy of social reform which seems to be emerging from the social science of the Nixon era in itself, of course, bears no resemblance to the harsh Social Darwinism of the age of the Robber Barons. The only parallel would lie in the danger that this new skepticism which is eroding the confident liberal assumptions could be distorted and used to rationalize a second period of indifference in a nation once again weary of the stress of reform.

What can be said, at the end of the first stage of the reception of the Coleman doctrine, is that— whether you believe with Daniel Patrick Moynihan that liberal education policies of the last few generations have succeeded so well that they have run into diminishing returns, or with Christopher Jencks that they have proved disappointing—those policies, and the intellectual assumptions on which they were built, are in bad trouble. They have lost support in the ranks of the social scientists who provided America, from Roosevelt to Johnson, with a major part of its operating ideology. □

  1. Coleman oversimplified his own report slightly on this point: in the first grade blacks did better than Puerto Ricans, while in the twelfth grade Mexican Americans did better than American Indians.
  2. Armor mentioned three other studies: one from Berkeley, California, one from Evanston, Illinois, and one from Rochester, New York.
  3. In one sense‚ Moynihan is closer to Jencks than is generally supposed. When he went to work for President Nixon, both he and the President were fully aware of the Coleman conclusions. At that point, in February, 1969, two documents arrived on Moynihan’s desk within seventy-two hours. The first was Arthur Jensen’s article, which started from the proposition that compensatory education wasn’t working. The second document, the Ohio Westinghouse report, was a gloomy appraisal of one major experiment with compensatory education. Project Headstart. Moynihan says that the conception of his Family Assistance Plan was directly influenced by the social science findings about education and equality. “The argument was put to the President,” he says, “that enormous expectations had built up that you could achieve racial equality through compensatory education‚ and it was not working. Point two: a proposition had been put forward by Dr Jensen which the democracy could not live with. Therefore, point three: you had to move directly to income redistribution.” There is an ironic parallel here—if a distant one—to the way in which Christopher Jencks concludes his book Inequality.