Abigail and Stephan Thernstrom
Abigail and Stephan Thernstrom; drawing by David Levine


As a Harvard graduate student and junior professor in the 1950s and 1960s tutored by Oscar Handlin, Stephan Thernstrom resolved to write “from the bottom up,” to reveal the lives of ordinary Americans.1 His theme in Poverty and Progress (1964) and The Other Bostonians (1973) was that mid-nineteenth- to early-twentieth-century America offered less social mobility than was believed. He found that the social system was not completely rigid—some sons achieved a status that was higher than their fathers’—but upward movement was mostly in fairly small steps. If a working-class boy rose to the middle class as an adult, he was likely a skilled craftsman’s son who became a clerk. Movement from bottom to top, from poor to rich, was rare; even movement from poor to middle class was an anomaly. Statistical analysis of trends in occupation, income, and property ownership, Thernstrom wrote, “yielded rather pessimistic conclusions about social mobility in nineteenth-century America.”2

So we might expect a now-older and still-wiser Stephan Thernstrom to be suspicious of claims that differences in class, which is closely related to race in America, could be obliterated if only the public schools did a better job. Yet this is the view that he and his collaborator (and wife), Abigail Thernstrom, promote in their recent book No Excuses: Closing the Racial Gap in Learning. The Thernstroms claim that when black youths have test scores that are equal to whites’, their adult earnings will also be equal. Blacks’ average academic achievement, in the Thernstroms’ view, is now not equal to whites’ mostly because black community culture doesn’t value education the way the culture of middle-class whites does. As a result, black students don’t try hard in school, and schools, in turn, don’t expect much of black students or press them to try harder. If schools had higher expectations, tougher discipline, a concentration on basic skills, and a resistance to the loose standards of progressive education, the test scores, and ultimately the working lives, of low-income blacks would reach white middle-class levels.

If the Thernstroms claimed only that good schools could narrow the gap in achievement between blacks and whites, or if they tried to show only that poor blacks could score as well as poor whites on tests, their analysis would be less vulnerable. But they believe that their prescriptions for school reform will enable poor black children to do as well on tests as middle-class white students. Consider, for example, the claims of the federal No Child Left Behind education law, whose aims the Thernstroms endorse: according to the law, not only are whites and blacks expected to show identical (and high) achievement a decade from now, but so are middle-class and “economically disadvantaged” students.

The premise that racial discrimination has been erased and that the remaining reason for differences in the relative earnings of blacks and whites is a difference of skills has become an article of faith among conservatives. (The Thernstroms are fellows at the Manhattan Institute, which promotes a variety of right-of-center causes.) The evidence for this, however, is selective, based mostly on statistics showing only that the earnings of black and white college graduates are similar. Discrimination in the labor market against “highly educated blacks [was a problem] four decades ago,” the Thernstroms write, but no longer. Therefore, they conclude, if black students worked as hard as whites in elementary and secondary school, and if blacks attended schools that expected them to work harder, there would be little difference in their success as adults.

But only a minority of Americans, black or white, are college graduates (18 percent of blacks and 34 percent of whites in their late twenties have college degrees).3 If the Thernstroms are correct in denying any rational basis for a black culture of underachievement, they need to show that black high school graduates’ later earnings are comparable to those of whites. But they don’t.

This is not to say that there has been no progress. Male black workers with only high school educations earn on average 79 percent of what similarly educated whites do. If only high school graduates with similar test scores are compared, then black wages are 87 percent of white wages. These figures, however, understate the difference in annual earnings between white and black high school graduates because black high school graduates work fewer annual hours and annual weeks than do white high school graduates.4

While some black high school graduates may be less aggressive in their search for jobs, discrimination keeps many, especially males, out of the labor market. As firms become more sophisticated about avoiding explicitly racist practices, discrimination is harder to prove, but there are enough lawsuits to suggest that it exists. Earlier this year, the Equal Employment Opportunity Commission charged Kodak, Rochester’s most prominent employer, with paying black workers less than similarly situated whites. Kodak promoted blacks less often and retaliated against those who complained by firing or harassing them. Blacks were subjected to racial epithets and slurs by senior Kodak employees. While Kodak has promised to correct these practices, it would not be irrational for black high school students in Rochester to conclude that however well they do, they will not be appropriately rewarded for their efforts.


Studies of black and white job-seekers with identical résumés who apply for publicly advertised jobs provide solid evidence of systemic discrimination that cannot be attributed to differences in skills between comparably educated blacks and whites. In the early 1990s, for example, teams sponsored by the Urban Institute in Chicago and in Washington, D.C., trained applicants for jobs with nearly identical résumés to present themselves in the same way in their interviews. Black males were three times as likely to be rejected as white males. Other studies have found that among applicants who were offered jobs, whites were offered higher salaries. A recent study found that whites’ applications were more successful than blacks’ even when the whites had criminal records and the otherwise identical blacks did not.5

Certainly discrimination would help to explain the culture of underachievement among blacks, the main concern of No Excuses. For his study Black American Students in an Affluent Suburb, John U. Ogbu, a Nigerian-born anthropologist at the University of California, Berkeley, who died unexpectedly last year, investigated the high school in Shaker Heights, Ohio, trying to understand why middle-class black students did far less well than middle-class whites. Ogbu, who has reported similar findings for over thirty years, lists in this most recent book the familiar claims about black students, claims that he found largely accurate. They put pressure on one another not to “act white” by doing well in school. They do not work as hard in school as white students from economically similar families. Black students watch TV and socialize with friends more than whites do. More black students than whites come to class unprepared and are more disruptive. Black students spend much more time at after-school jobs. Many think grades are unimportant because they can go to college on athletic scholarships. When given the opportunity to take more academically challenging courses in high school, they frequently decline.

Moreover, black parents supervise homework less frequently than white parents do, while black adolescents are exposed daily to skepticism about white-controlled institutions, such as schools. Black students also accept conventional notions that they have less academic ability and make less effort as a result. Ogbu supports these conclusions with his own observations and interviews with Shaker Heights’ black students, their parents, and their teachers. One student told Ogbu, “[Black] kids seem to…have…this unconscious way of thinking that Blacks are inferior to Whites. And I think that takes a toll….”

Along with other black conservatives—John McWhorter, Shelby Steele, and Thomas Sowell—Ogbu places the blame for ongoing inequality on black communities. He recommends a variety of self-help strategies to raise black students’ achievement, such as publicizing black students’ academic successes, reinforcing parents’ commitment to monitoring their children, and so on. But he departs from conservative thinking in one important respect: he acknowledges that blacks have understandable reasons for cultural hostility to white society. Ogbu supports campaigns in black communities to encourage successful black scholars by holding award ceremonies for them. He also approves of schoolteachers’ attempts to convince blacks that there are no limits to their ambitions. But he acknowledges that such measures are unlikely to raise achievement to white middle-class levels.

Even if there were no social discrimination, and if black students were willing to believe in the value of hard work, lower-class blacks would still face obstacles to social mobility similar to those Stephan Thernstrom long ago documented for nineteenth-century white working-class youths. For one thing, the culture of underachievement is not exclusively black. Lower-class children, both black and white, do less well in school than middle-class children. Even if the race gap could be eliminated, the problem of improving the achievement of lower-class children of both races to middle-class levels would remain.

For decades now, sociologists have documented how differences in social class, regardless of race, affect young children’s readiness for schooling.6 Annette Lareau, who has observed middle- and working-class children, writes in her book Unequal Childhoods that middle-class children today are encouraged from an early age to negotiate with their parents over what to wear or eat, to question adult statements if they seem implausible, and to interact with adults as equals. Children from the white middle class are expected, for example, to describe their symptoms to pediatricians. Money tends to be less frequently discussed in middle-class families, so it less frequently occurs to middle-class children that their ambitions might be blocked by a shortage of resources.


Working-class children have no such sense of entitlement. Most of them, black and white, speak to adults only when spoken to; they are not expected to express opinions that challenge what adults say. Money or lack of it is a frequent topic of conversation at home, and children become aware early of the limits to their futures.

Lareau describes the hectic schedule middle-class parents organize for their children—music lessons and soccer leagues, for example. These are expensive, not only in money but in time, since children usually have to be driven from one activity to another. Children earn trophies and parental praise for their performances, which reinforces their sense of entitlement. These activities also promote teamwork and easier relations with strangers. The working-class children Lareau observed mostly stayed in their neighborhoods, playing games only among themselves.7

Middle-class parents were more likely to encourage children to figure out problems for themselves. Working-class parents were more likely to tell them what to do. Lareau’s middle- and working-class parents both encouraged their children to read, and parents from both classes read aloud to their children when they were young; but middle-class parents were more likely to read themselves, thus showing the importance of reading by their own behavior. Moreover, Lareau’s middle-class parents more frequently intervened in schools when they felt it in their children’s interest to do so. In high school, as John Ogbu observed, middle-class white parents are aggressive in guiding their children’s decisions on curriculum, while Ogbu’s black parents and Lareau’s working-class parents are not. Indeed, in many ways, Ogbu’s middle-class blacks are similar to Lareau’s working-class whites in attitudes toward education.

Definitions of class distinctions are often imprecise. Ogbu observed black and white middle-class families in Shaker Heights who, despite having similar incomes, may nevertheless belong to different social classes. In Whitewashing Race, Michael Brown and his colleagues try to give a more sophisticated description of race-based class differences.8 They are critical of arguments against affirmative action in the Thernstroms’ previous book, America in Black and White (1997), pointing to many differences between blacks and whites that the Thernstroms ignore. While the lives of blacks and whites may be becoming more similar, there remain vast differences that affect blacks’ ability to compete with whites in school.

Black median family income is now 62 percent of white income, up a little from 58 percent thirty years ago. But, as Brown and his colleagues point out, the median net worth of black families is still only 8 percent of whites’. Part of the reason for this difference is that government and private lenders have discouraged or prohibited black families from buying homes in suburban communities where values have appreciated in the decades after World War II. Part of the reason, too, is that black middle-class adults are more likely than whites to be the first generation in their families to have risen to the middle class; many have low-income parents and other close relatives to help support, making it more difficult for them to save money.

According to Brown and his colleagues, “The best predictor of current net worth for young black and white families is their parents’ net worth.” Especially in today’s inflated housing market, young couples who are not the first generation in their families to be middle class often get financial help from their parents to buy a house. Even in comparatively well-off suburbs like Shaker Heights, black families, lacking such help, are likely to be under more economic strain. Blacks and whites in Shaker Heights may all be middle-class, but blacks have a more tenuous hold on their status. Young people who know their parents have saved money for their college tuitions are likely to have different attitudes toward their educations than similarly middle-class young people (more often black) whose parents have no such savings. The Thernstoms acknowledge that current income alone is an insufficient explanation of difference in status:

A more refined measure of social class that took into account the socioeconomic status of grandparents as well as parents and that included information about wealth would doubtless narrow some of the differences that we have attributed to race.

But they nonetheless attribute these differences to a black culture of underachievement, even though culture is at least partly an inextricable expression of social and economic life.

No one can say how much of the gap in academic achievement between blacks and whites is caused by racially neutral class differences, how much is attributable to black culture, and how much of that cultural difference is itself a defensive reaction to continuing discrimination. The Thernstroms write that poverty (defined by current income), parents with little education, and residence in poor neighborhoods explain only about a third of the gap. It doesn’t take much reflection to think of other noncultural factors that may help to explain the differences in academic success. Poor black families, for example, are likely to have low incomes for longer periods than poor whites, whose poverty is more episodic. But when the test scores of “similar” black and white students are compared, current-year family income is usually the only evidence available.9 Taking into account poverty’s duration would account for more of what falsely appears to be a gap in race and culture.

Partly because urban rents have climbed faster than earnings of low-wage workers, low-income families move frequently. And black families move more often than whites, perhaps because poor urban blacks are poorer than poor urban whites (partly, in turn, owing to more single-parent families and to higher black unemployment). A government survey a decade ago found that black children are more than twice as likely as whites to have attended at least three different schools by third grade.10 An analysis of demographic and test score data in Texas recently concluded that if the mobility of black students could be reduced to the frequency of whites’, it would eliminate about 14 percent of the racial gap in test scores. Taking income into account, blacks’ average test scores would rise by 7 percent if poor children moved only as often as the non-poor.11


What is most puzzling about the Thernstroms’ argument is the lack of connection between their analysis and the policies they prescribe. On finding that the gap in achievement between blacks and whites derives from some combination of social class and black culture, they conclude that it can be closed by establishing more charter schools, issuing vouchers to black children to attend private schools, improving the quality of teachers (largely by weakening the power of teachers’ unions so that administrators will have more discretion over hiring and firing teachers), and concentrating more on basic academic skills (“facts are good”).

Spending more money on schools, lowering class sizes, and making efforts to integrate schools, they say, have proven to be failures. Nor do they endorse expensive early childhood programs for black infants and toddlers that might expose them to well-educated adults who could help provide the more literate (middle-class) intellectual environment these children lack at home. This is not a radical idea. Last year, President Bush’s assistant secretary of education responsible for elementary and secondary school policy resigned, and shortly thereafter denounced the administration’s insistence that blacks’ inadequate achievement was caused by “low expectations” held by schools. Rather, she said, by age three, black children are already far behind, making necessary high-quality early childhood programs beginning in infancy.12

The Thernstroms do not recommend interventions that could modify the enormous out-of-school differences that contribute to the lower achievement of black students—neither early childhood programs nor the funding of after-school or summer programs that could offer artistic and athletic activities to lower-class children. Instead, they hold up as models a few charter schools that report high test scores for black children, and a single teacher in Los Angeles who is reputedly successful. The charter schools the Thernstroms describe are doubtless excellent, but they are questionable models for raising the achievement of lower-class blacks. For example, the Thernstroms praise a chain of middle schools called the KIPP Academies, which try to enforce an academic culture that the Thernstroms say most black families lack: strict discipline, constant talk of college plans, and a focus on basic skills.13 But these charter schools do not enroll black children from typical low-income families. For example, the Bronx KIPP Academy, cited by the Thernstroms, entering fifth graders have fourth-grade test scores that are considerably higher than the averages in the regular schools from which they come. Parents who send their children to such schools are already unusually interested in their education; children are accepted only if parents agree to monitor their homework, enforce approved disciplinary measures, and limit television-watching. If children or their parents violate these agreements, the children can be expelled—a rare occurrence, but a threat nevertheless.

KIPP also provides after-school, Saturday, and summer programs devoted to artistic activities, yet they do so without spending more than regular schools whose students are in attendance for 40 percent less time. The Thernstroms see this as proof that more money for schools isn’t needed, and attribute KIPP’s accomplishments to its “creative budget decisions.” But KIPP survives largely by hiring young, energetic teachers who do not intend to make teaching a career. For working many more hours and weeks than most teachers, the KIPP faculty gets extra pay but still earns less than experienced teachers in the regular public school system. KIPP teachers are frequently replaced, partly because the program cannot sufficiently raise their salaries as they grow older. KIPP is an interesting model but one that cannot be widely reproduced.

The founders of KIPP were inspired by the example of Rafe Esquith, a teacher at the Hobart School in Los Angeles who, the Thernstroms believe, is an exception to the ineffectiveness of other public educators. In his memoir There Are No Shortcuts, Esquith describes teaching at a school whose students are almost entirely the children of Korean and Central American immigrants, many of whom were well educated in their home countries. (Some of the Central Americans, for example, were professionals who fled the civil wars of the 1980s.) Esquith has had virtually no experience with black students, which makes him a curious model for the Thernstroms, who make so much of black culture as the cause of blacks’ low achievement in school.

Esquith is an inspiring teacher whose pupils perform Shakespeare, learn to play stringed instruments, read literary classics, and win academic competitions. His students arrive in school early for extra instruction, stay late, come to school on Saturdays, and take cross-country trips in the summer, all under Esquith’s tutelage. He initially paid for these extras himself with what he earned by taking second and third jobs, and later, as his reputation grew, from foundation grants. Esquith only mentions in passing the crucial fact that his students were specially selected because of their unusually high IQ scores (his were “gifted and talented” classes). There is no reason to think that what works for these children will work for poor children of average ability.

Would providing additional money to public schools serving black students help them? Convinced that public schools are incompetent, the Thernstroms say no. For some schools they are probably right, although they underestimate how many public schools in impoverished communities are staffed by dedicated teachers who impose appropriate discipline and high standards. Still, if the causes of the discrepancy in achievement are as rooted in class and culture as they appear to be, more money for the schools won’t itself close it. But more money could somewhat narrow the gap, as is suggested by evidence which the Thernstroms ignore.

A puzzling aspect of the discrepancy in test scores is that it was cut in about half between 1970 and 1990. Since then it has remained steady, narrowing slightly on some tests, widening on others. (The Thernstroms mention only the widening.) David Grissmer, a RAND research scientist, and his colleagues have shown that some social class differences between whites and blacks narrowed during the 1970s and 1980s, explaining about half of the reduction in the gap in test scores.14 A decrease in black family size gave black parents more time to devote to each of their children. The percentage of black mothers who completed high school also rose substantially, giving black families greater comfort with formal education.

The Thernstroms believe that government spending on the education of black children has had no effect. Yet if only half of the closing gap from 1970 to 1990 can be attributed to some improvement in blacks’ social and economic circumstances, it is possible, indeed likely, that public programs (in which there was then greater interest than there is now) contributed to the other half. Between 1970 and 1990 new money was invested in Head Start, Title I compensatory programs, special education for slow learners, and reduced class size. This was also a period of school integration, another reform the Thernstroms dismiss as ineffective.15 Much of the new money in these decades may have been spent inefficiently, but enough may have been well spent to help narrow the discrepancy between blacks’ and whites’ achievement.

In his recent book Final Test, Peter Schrag argues that unequal distribution of money among schools accounts for much of the achievement gap. Many educators and activists have concluded that smaller classes, better-paid teachers, better facilities, and access to pre-kindergarten classes for four-year-olds are needed for poor and minority students to keep up. With low achievement so much a product of class and culture, and so well established by the time children are three, the idea that more money can itself solve the problem is as dubious as the idea that tough discipline, higher standards, and more basic education alone will work. Still, more money, like better instruction, can certainly help, and in Final Test, Schrag documents efforts to find in state constitutions a right to have more money spent on the education of disadvantaged children.

In more than thirty states, plain-tiffs representing low-achieving students have sought court orders to increase school funding on constitutional grounds. Beginning in the early 1970s, these lawsuits mostly challenged school districts’ dependence on property tax revenue. Communities with high tax bases can afford to pay teachers more, and have smaller classes, better facilities, even better football equipment. Plaintiffs reasonably claimed that if education was a constitutional responsibility, states could not lawfully provide different educations to children based on the wealth of their com-munities. The suits have been only moderately successful; spending on education has been rising in almost every state, with or without court orders. Perhaps the fear of litigation has provoked some governors and legislatures to increase school funds more than they would otherwise have done.

But even equal spending may not be enough, and lawsuits now have taken a different tack. Plaintiffs argue that state constitutions implicitly guarantee an education for disadvantaged children that is minimally “adequate,” and the money that a state is obliged to spend on schools cannot simply go to poor and rich districts equally. Instead, adequacy must now be defined as providing sufficient money to generate “adequate” outcomes. Yet asking judges to determine not only what the specific outcomes of schooling should be, but what specific pedagogies, processes, and resources are guaranteed to achieve them, all from vague state constitutional guarantees of, say, “a thorough and efficient” education, requires a degree of judicial activism that even the most dedicated loose constructionists should abhor.

This is why so many of the state legal disputes described by Schrag end inconclusively. In Ohio two years ago, for example, judges found the system for financing education unconstitutional, but acknowledged that the courts are not competent to design a new system. As a result the legislature must do so without further judicial review. Schrag’s response is clever, and somewhat cynical. Instead of trying to find precise pedagogical descriptions of adequate schools in state constitutions, he urges plaintiffs to endorse the Bush administration’s No Child Left Behind law, and then argue that it is a denial of fairness to expect poor children to meet the law’s high standards without guaranteeing better-financed schools. This makes the definition of “adequacy” subject to legislatively adopted academic standards, not to the platitudes of constitutional aspiration.

To merge what Schrag describes as conservative goals (tough accountability for test scores) with liberal means (more money to achieve them) requires that low test scores be used as a lever to get states to put up more money. Schrag acknowledges that this tactic raises cause for concern:

…Instead of teaching kids to think and encouraging them to be creative, the accountability system foster[s] questionable drill-and-kill teaching practices and cram courses designed mainly to bolster scores, cause[s] schools to neglect things like art and music and higher-order skills that were not tested, and [drives] more students to drop out.

But to pursue these objections to a test-based accountability system, he concludes, would only undercut the case for more money—so he advises that they be soft-pedaled.

Schrag is right to insist that if spent properly, new money for things like smaller class sizes, better-qualified teachers, pre-kindergarten, and better facilities (the sorts of improvements that may soon be decreed as a result of New York City’s “adequacy” lawsuit) can do a lot to improve outcomes from poor children. But even if lawsuits demanding an “adequate education” could produce more money, class differences will continue to ensure an achievement gap because schools alone, no matter how well financed, can’t by themselves overcome the cultural, social, and economic causes of the differences in academic achievement. The drive for “adequacy” is bound to be seen, when someday it has played itself out, as only another failed education crusade.

Today’s education debates are poisoned by the insistence of partisans on finding a single cause, whether it be low standards, parental inattention, or lack of money, for continuing racial inequality, and then to denounce remedies that address others. Of recent books bemoaning the black–white achievement gap, Pedro Noguera’s City Schools and the American Dream attempts to avoid this single-solution approach. Noguera acknowledges that conservatives are right in saying that many urban schools are failing. But in view of the social and economic problems of their students, he writes, these schools are a lot better than critics think. In contrast to the desperation of their surrounding communities, urban public schools are the safest and most enriching influences in many children’s lives. Adding health clinics and other social services would make them even better. Some schools do better than others but that doesn’t mean that even the best schools can succeed on their own. Still, Noguera argues, by promoting intensive adult–child relationships, schools can help black students to develop more positive attitudes toward their education.

But Noguera, too, fudges the central issue of national education debates: Does closing the test score gap mean raising the achievement of lower-class black children only to the level of lower-class whites? This may well be possible, but few consider it a goal worth striving for. Or does Noguera hold out hope for schools that produce classless outcomes, in which not only exceptional poor black youths, but typical ones, reach the level of typical middle-class whites? Such a revolutionary outcome is no more likely today than it was in the nineteenth-century New England towns that Stephan Thernstrom once described. Since the US today has less social mobility across generations than most other industrialized nations, modest improvement might be a more appropriate aspiration. If we fail to accept gradual progress, education reform will remain a hopeless enterprise.

This Issue

December 2, 2004