Minding the Gap
Addressing differences in student and school performance
Abigail Thernstrom, a member of the Massachusetts State Board of Education, and Stephan Thernstrom, Winthrop professor of history at Harvard, have written an impassioned, informed, curiously uneven book about one of our country's most complex and important social problems: inequalities in what students learn at school. They frame the problem as a "racial gap" and focus on the disparities between test scores of non-Hispanic whites and Asian Americans, on the one hand, and blacks and Hispanics on the other. I shall ask later whether this is the best formulation, but let us start with their chosen perspective.
The problem the Thernstroms point to is now both well-known and widely agonized over, but it always warrants another examination. They lay it out clearly: African Americans, and to a lesser extent Latinos, do not learn as much in their passage through elementary and secondary school, at least as measured by test scores, as do Anglos and Asians. The average gap in test scores narrowed during the 1970s and 1980s, then widened in the 1990s; more than 50 percent (and in one case more than 70 percent) of black and Hispanic twelfth graders score "below basic" on national standardized achievement tests in most subjects, compared with roughly a quarter of Anglos and Asians. This failure to learn makes the former groups less able to graduate from college, less likely to attain or keep high-paying jobs, and less fully engaged in all aspects of American society. It is a "heartbreaking picture."
Although the Thernstroms appropriately emphasize the breadth and depth of the test-score gap, they also want to insist that it can be bridged. To do so, they believe, one must understand what causes it; here, too, their argument is not new but warrants careful attention. In their view, the educational gap for African Americans is no longer caused by direct racial discrimination but rather by a "cultural inheritance" that is itself "the product of a very long history of racial oppression." Black parents read less to their young children than white parents do, and black children are more disruptive in school than white children from the earliest grades, spend a "dismaying" amount of time watching television, and generally live in a culture that does not easily "connect...black students to the world of academic achievement."
Hispanic students mostly suffer from the recency of their families' immigration, because many grow up in families and communities in which English is not the default language and because many immigrant parents urge their children to get jobs rather than stay in school. In an important analysis, however, the Thernstroms point out that the children of native-born Latino parents do move toward closing the test-score gap with Anglos so the Thernstroms' main solution in their case is, implicitly, time (and English immersion), rather than any active intervention by public or private agencies.
The Thernstroms note that Asian-American students, in contrast, benefit enormously from their groups' cultures. (The authors recognize that there is no single "Asian culture," but perforce fall back on monolithic language and data that aggregate all Asians.) "In large part, Asian students typically do well in school because their parents insist upon it, and they feel obliged to comply with their parents' wishes." This finding, the Thernstroms assert, is "grounds for optimism" because "parental pressure to work extraordinarily hard in school...is a culturally transferable trait." Before turning to this implied solution to the test-score gap, however, let me note that the Thernstroms do not investigate Anglo culture; readers are given no help, therefore, in interpreting the fact that 60 percent of Americans would rather have their child "make average grades and be active in extracurricular activities" than "get A grades" (according to a 1996 survey by Phi Delta Kappa, the professional association for educators). More generally, we are left to assume that white parents endorse high achievement more than do black or Hispanic parents, but less than Asian parents. This is a plausible argument, but it is not self-evidently true and a direct discussion would have been valuable.
In any case, the Thernstroms move on to provide reasons for rejecting most liberal solutions to the test-score gap, such as more funding for urban public-school systems, smaller classes, more black and Hispanic teachers, or maintenance of federal programs targeted at disadvantaged students. Instead they endorse small, semi-independent schools that face the cultural challenge head on especially the several dozen KIPP (Knowledge Is Power Program) charter schools now operating in a few cities. In KIPP schools, which remain within the public-school system but have considerably more autonomy (and a little less funding) than regular public schools, teaching is very carefully organized, controlled, and monitored ("every teacher's...blackboard...looks the same") and the only extracurricular activity is music.
These schools (or, very occasionally, classrooms in a regular public school) hold teachers, parents, and students to an extraordinarily high standard of achievement through written contracts, extra hours and weeks of schooling, strict order and discipline, and relentless attention to achievement. They also promote cultural change through pep talks, posters, and what appears to be continuous student chanting: "The day we visited Amistad, the kids were chanting...North Star has almost the identical chant...." Or "at KIPP students chant rules in unison..." and "'Why are you here?' a teacher asks in a 'call-response' chant that students run through in the morning circle." And so on. The justification for all of this is that "we are fighting a battle involving skills and values," according to one of the Thernstroms' exemplary school leaders. "We are not afraid to set social norms."
The Thernstroms show that the students in their model schools and classrooms score well on state tests. And their descriptions are moving and inspiring, just as their descriptions of horrible teachers in regular public schools are depressing and infuriating. I would rather have my children in a KIPP school, chanting, identical blackboards, and all, than in a lot of schools in America's largest cities (where a majority of black and Latino children live). But can an effort to expand these "little islands of true heroism" really solve the problem that the Thernstroms laid out so elegantly in the first chapter of No Excuses?
In principle, yes. As they point out, it requires no new funds or changes in the governance structure of schooling for all parents, teachers, and principals to simply decide that henceforth they will tolerate no disruption and will demand academic excellence from all students. But that is unlikely to happen, to put it mildly. Thus we need a more careful examination of the nature of the test-score gap, its causes, and the available solutions, before signing on to their program.
For one thing, the racial gap is at least partly a class gap. The Thernstroms concur, but before and after an elegant analysis of the role of social and economic standing in the achievement gap, they drop the point. By their reckoning, a third of the test-score gap is due to differences in family background. But their measures of family background pay no attention to levels of wealth (which is vastly more unequally distributed than income, and which has been shown to have an impact on schooling achievement); to an assessment of the education of both parents (rather than just of the parent with the higher level of schooling); or to other variables that could plausibly raise the explanatory power of class to, say, half of the gap. Class does matter, a lot: data from the Digest of Education Statistics, 2001, show that the test-score gap in math between 17-year-old students whose parents completed college and those whose parents did not complete high school is now greater than the comparable test-score gap between Anglos and African Americans. In addition, the social and economic status of one's schoolmates matters a great deal. Middle-class children achieve more than poor children do when both are in schools with mostly poor classmates. However, middle-class children in poor schools achieve less than do middle-class children in mostly middle-class schools sometimes, in fact, they fare worse than poor children do in mostly middle-class schools.
The Thernstroms' solution to the achievement gap might remain the same even if they focused on class rather than, or in interaction with, race and ethnicity. After all, plenty of scholars argue that poor Americans lack the right culture for "connect[ing]...to the world of academic achievement." But perhaps new solutions would become available with this new lens on the problem. Attention to class disparities, for example, would allow many districts to "desegregate" high- and low-achieving students in a way that the Thernstroms insist is not possible with regard to racial desegregation. (They take school-district boundaries as given, rather than seeing them for what they are, a constitutionally malleable creation by a state.)
Or perhaps the racial gap is at least partly a state-policy gap. The Digest of Education Statistics also shows that if we compare eighth graders in central cities to those in the "urban fringe and large towns," the average national disparity in reading scores is 12 points. But in Connecticut, New York, and Minnesota, the disparity is double that, whereas in California, Florida, and Delaware it is half that or less. If states are acting so differently that children in some big cities are less harmed by their urban residence than are children in other big cities in different states, appropriate policy recommendations for closing the test-score gap would look very different from those offered by the Thernstroms.
Even if we accept that a large part of the reason for the achievement gap really is cultural, the Thernstroms' solution is insufficient. Pace their single statement to the contrary, the evidence has been strong and consistent for four decades that children's classmates have an enormous impact on their commitment to learning, on the environment in which they learn, and on the outcomes of their schooling. Culture matters. If black students need to absorb achievement-oriented values, or Hispanic students need to become acclimated to American values, or poor children need to learn middle-class values, then putting them in a classroom with students who already have those values (or at least, whose parents and teachers do) is the most effective way to reduce the test-score gap across the nation as a whole. That conclusion implies as much racial desegregation (across district lines if necessary), economic mixing, and ethnic or linguistic mixing as possible. (After all, once boys and girls started sitting in the same classrooms, they became more like each other in goals and outlooks for better and for worse.)
Such a policy choice would no doubt be difficult and controversial, but it is no more difficult than struggling to find enough "heroes" to create the schools needed to transform the lives of millions of underachieving students. Such a policy would also call on middle-class, predominantly white, suburbanites to shed their own excuses for their racial and economic isolation.
The Thernstroms avoid this challenge by proposing that all schools containing underachieving African-American and Hispanic children be turned into charter schools. That solution would have the pseudo-advantage of letting affluent white suburbanites off the hook. It would also have the real advantage of enabling energetic, idealistic, imaginative citizens to try their hand at improving the terrible conditions in which too many children waste their days.
But a massive switch to charter schools is unrealistic as the central policy solution. To begin, manage, and invigorate a school for years on end, especially in a community with few resources, ingrained mistrust of neighbors and authorities, and deep needs, is very difficult. The charter schools that already exist show why this reform is not a panacea. Some are superb, some are hopelessly mismanaged or cynically corrupt, some are hoped-for profit centers for private corporations, and most produce about the same test results as the public schools near them. Our nation cannot rely on solving the problems of public education by retreating from the public arena.
Abigail Thernstrom and Stephan Thernstrom deserve praise for their passion for racial equality, their powerful prose, and their fascinating arguments. That their policy proposal seems more a counsel of despair than a solution is as much a commentary on the complexities of race, class, and American public schooling as it is a shortcoming of their book. Let us hope for more heroes in our children's classrooms along with public policies that will truly leave no child behind.
Jennifer L. Hochschild is Jayne professor of government and a member of the African and African American studies department; she is currently on leave as a Radcliffe Institute fellow for the 2003-2004 academic year. She is coauthor, with Nathan Scovronick, of The American Dream and the Public Schools (Oxford, 2003).