FROM ISSUE NUMBER 5 ~ FALL 2010 GO TO TABLE OF CONTENTS

TEXT SIZE A A A

The Widest Achievement Gap

DAVID L. KIRP

Over the past few decades, improved academic-achievement data have given us unprecedented insight into the strengths and weaknesses of America's system of education. And among the most striking and worrisome findings has been the degree to which African-American boys and young men lag behind their fellow students. By almost every measure of academic performance and achievement, black males are on the wrong side of a staggering divide.

This so-called "achievement gap" has of course drawn a great deal of attention in education-policy circles. But efforts to address it have, by and large, been guided by one of two sets of assumptions, neither of which is well rooted in real-world evidence. One theory holds that the potential of black students, like that of white students, is limited by their innate intelligence — or IQ — and that the gap in performance thus reflects the difference between these upper limits. Even the most ambitious educational interventions, then, will be largely futile. The contending theory asserts that black students underperform because they are likely to be poor and underprivileged, and that addressing the academic achievement gap therefore requires first changing the economic and social conditions in which these students grow up.

Since both sides believe that the reasons for the achievement gap lie almost entirely outside the classroom — and, in many cases, beyond the control of students, parents, or teachers — they also contend that education reforms focused on changing schools, helping parents, and aiding students would be of little use. Both camps are, in essence, education-policy fatalists.

But while this ideological debate has dragged on, the facts have begun to tell a different story. Just as we have learned more about the nature of the achievement gap, we have also learned a great deal about how it might be narrowed. A growing body of evidence points to the effectiveness of approaches that incorporate intense individual attention for students, support for parents, and a continuum of age-appropriate strategies to improve reading and math skills. These strategies have generated some extraordinary successes. And they offer reason for cautious optimism that many American children who have been consigned to lives of frustration and poverty might instead become productive, fulfilled citizens.

LEFT BEHIND

It is hard to overstate the plight of African-American boys and young men in our education system today. On every measure of educational attainment, they fare the worst; despite waves of reform, their situation has not changed appreciably in 30 years.

The gap between their performance and that of their peers is perceptible from the first day of kindergarten, and only widens thereafter. In the 2008 National Assessment of Educational Progress — the massive, federally mandated report card on student performance, measured in grades 4, 8, and 12 — the reading scores of African-American boys in eighth grade were barely higher than the scores of white girls in fourth grade. In math, 46% of African-American boys demonstrated "basic" or higher grade-level skills, compared with 82% of white boys. On the National Education Longitudinal Survey, 54% of 16-year-old African-American males scored below the 20th percentile, compared with 24% of white males and 42% of Hispanic males. Having well-educated parents did not close the gap: In 2006, 43% of black high-school seniors with at least one college-educated parent failed to demonstrate even basic reading comprehension, nearly twice the percentage of whites.

According to a College Board report published earlier this year, black male students are 2.4 times as likely to have been suspended and twice as likely to have repeated a grade as white males. High-school graduation rates tell the same story — just 42% of black males graduated on time in 2006, compared with 71% of white males. After leaving school, these dropouts generally seem to encounter only more failure: Among 16- to 24-year-old black men not enrolled in school, fewer than half have jobs; about a third are in prison or jail, or on probation or parole.

Black men also fare badly when compared to black women — who, it is worth remembering, grow up in similar family and cultural circumstances and attend the same schools. Black men are three times more likely than black women to be suspended; their high-school graduation rate is 9% lower; and they are only half as likely to get a college degree.

These comparisons have led some social scientists and geneticists to suggest that the chief explanation for such disparities is innate intellectual capacity. After all, academic achievement is clearly linked to intelligence. And more than a century's worth of quantitative-genetics literature contends that each person's IQ tends to be remarkably stable over time, and that about three-quarters of IQ differences among individuals can be attributed to heredity. This would suggest that there are different upper limits on intellectual ability for different people, so that differences in academic success are only natural.

In their 1994 bestseller, The Bell Curve, Richard Herrnstein and Charles Murray relied on such evidence to argue that the United States is increasingly a genetic meritocracy, and that government interventions cannot be expected to improve the performance of students who are limited by their low IQs. Prominent geneticists, like Arthur Jensen at the University of California, Berkeley, have concluded on similar grounds that "compensatory education" — which aims to correct for social or economic disadvantages — is a losing bet.

But a growing body of evidence suggests that this argument is at best incomplete, especially when it comes to the poorest and most disadvantaged children. This flaw is due in no small part to the blinkered nature of past research on the stability and heritability of IQ, which has relied heavily on studies of twins. Identical twins are essentially clones; fraternal twins are siblings born together, but not genetically identical. If heredity explains most differences in intelligence, the logic goes, the IQ scores of identical twins will be far more similar than the scores of fraternal twins. And this is what the research has typically shown. Only when children have spent their earliest years in the most wretched and abusive of circumstances has environment seemed to make a noticeable difference. Otherwise, genes rule.

In the past few years, however, researchers have cast serious doubt on this conclusion. In 2003, psychologist Eric Turkheimer combed through the twin-study data and found that the numerous evaluations share a fatal flaw: They focused only on middle-class twins. (The poor, of course, do not generally volunteer for academic research projects.) But Turkheimer was able to find a large-sample study from the 1970s that tracked more than 50,000 American children — many from poor families — who had taken IQ tests at age 7. The data showed that, among twins with well-off parents, virtually all the variation in IQ could be attributed to genes; the result was consistent with the conventional wisdom. But Turkheimer's other discovery was startling: Among twins from poor families, heredity explained almost none of the IQ variation. The IQ scores of identical and fraternal twins were both remarkably similar; the shared experience of growing up poor had leveled the fraternal twins' intelligence as much as shared genetics had shaped the intelligence of the identical twins. This, of course, would indicate that people's intelligence is not simply a matter of genetics: In Turkheimer's analysis, being raised with little social or economic capital overwhelmed the children's genetic capacities. So while genetics may well place an upper limit on intelligence, poor children are often denied the opportunity to reach their own maximum potential — meaning that, in the right environment, the achievement gap might be significantly narrowed.

Other research in recent years has supported Turkheimer's findings. When William Kremen, a psychologist at the University of California, San Diego, examined the reading ability of middle-aged twins in 2007, he found that, even half a century after childhood, family background still had a major effect — but only for children who grew up poor. French psychologist Michel Duyme has shown that, when four- and five-year-old French children who were abused and neglected as infants were adopted by caring families, their IQs increased by as much as 25 points. (The children who were raised in well-to-do homes gained the most.)

In short, the case for genetic fatalism in the face of the achievement gap has grown far weaker. Increasingly, it seems there is room to improve the educational performance of black boys and young men. And new data point to a range of strategies that enable underperforming youth generally, and African-American boys in particular, to thrive.

Ferreting out these initiatives has proved to be surprisingly difficult, because most researchers analyzing education data have tended not to isolate the "race-plus-gender" effects of programs that benefit poor and minority children. The assessment offered in the following pages therefore includes several previously unpublished analyses of data, undertaken by the studies' authors at my request, which do take both race and gender into account.

Though the most promising educational approaches differ in many respects, they share one key element: They emphasize a continuous supply of individual attention, regardless of whether the child in question is an infant, toddler, or teenager. Such intense engagement can benefit all children, but it turns out to be especially important for African-American boys. (While social science has not provided a clear explanation for this difference, the fact that many of these young people are raised without fathers is of course one likely factor.) Successful approaches also tend to involve a progression of age-appropriate assistance, and not just a single intervention at some particular age.

None of these strategies is a silver bullet, and none would be easy to implement on a national scale. But the evidence shows that solutions are possible, and worth pursuing.

STARTING EARLY

More than anyone else, mothers and fathers shape their children's futures. And so, as the old joke goes, the smartest decision a child can make is to pick the right parents.

While several studies have shown that family wealth and neighborhood quality account for nearly half the IQ gap between blacks and whites at age five, recent work by developmental psychologists Jeanne Brooks-Gunn and Lisa Markman suggests that more than half of that gap is actually attributable to how mothers relate to their children. In a large-scale observational study of parent behavior published in 2005, Brooks-Gunn and Markman found that, as a group, black mothers ranked lower on seven measures of parenting — ranging from nurturance to the availability of books in the home — than white mothers. The most striking differences concerned children's exposure to language, which is critical to success in school. Brooks-Gunn and Markman found that these multiple differences in parenting styles generated a 6- to 12-point racial gap on a common school-readiness test (structured like an intelligence test with a score of 100 as the median). Other work supports this conclusion: A 1994 study by psychologists Betty Hart and Todd Risley showed that children from poor, mainly black families engaged in literally thousands fewer conversations than children from wealthier, mainly white families; by the age of three, those poorer children had vocabularies just half the size of their well-off counterparts.

Disparities in child-rearing are not easy to mitigate through public policy. But well-designed parenting programs have been shown to have a meaningful impact. Evidence-based home-visiting programs, like the Nurse-Family Partnership (which relies on trained nurses to support parents from pregnancy through the first two years of a baby's life) — as well as center-based programs that also include home visits, like Early Head Start — have been shown to enhance parents' sensitivity to their infants' and toddlers' cues, lessen reliance on spanking, and increase the number of age-appropriate learning materials around the house (as well as the amount of time spent reading to kids).

As economist Steven Barnett, director of the National Institute for Early Educational Research, concluded in a five-state study, these strategies boost children's performance in the early grades. The effect of the Nurse-Family Partnership on African-American boys specifically has not been isolated, but in one study in Memphis — where the participants were overwhelmingly African-American (though not all boys) — the children whose parents participated in the program performed significantly better during their first years at school than did those in the control group.

Unfortunately, such high-quality parenting support is available to only a handful of poor families. The biggest program, Early Head Start, is so meagerly funded by the states and the federal government that it can enroll fewer than 1% of eligible (below poverty line) infants. Its expansion, particularly in African-American communities, is essential to any strategy for bridging the black male achievement gap.

But improving interventions at the earliest stages is not enough. Early-parenting programs can make a lasting difference only if they are followed up with effective early education — meaning that African-American boys must continue to receive quality instruction and focused attention throughout childhood, and especially in their first years of school.

The argument for investing in early education as a way to narrow the achievement gap rests on the results of three groundbreaking long-term studies: the Perry Preschool study, which followed a group of at-risk black children in Ypsilanti, Michigan, starting in the mid-1960s; the Abecedarian study, run by the University of North Carolina, which enrolled children from "low socioeconomic backgrounds" in Chapel Hill starting in the early 1970s; and the Chicago Longitudinal Study, which began observing children from the city's poorest neighborhoods at the Chicago Child-Parent Centers in the 1980s. In each instance, children who enrolled in a high-quality early-education program were tracked into their twenties — in the case of Perry Preschool, into their forties — and their life trajectories were compared with those of a matched control group. While these three experiments differed in some particulars (Abecedarian children remained in the program from the time they were infants until they finished kindergarten, for instance, while Perry enrolled three- and four-year-olds), in each case, highly trained teachers — assigned to a small number of children — used curricula designed to harness children's curiosity while enlisting parents to bolster their efforts.

The results of the studies were eye-opening. Compared with the control groups, significantly fewer participants were left back or assigned to special-education classes, while significantly more children graduated from high school and enrolled in college. More remarkable still were the effects on participants later in life — including lower crime rates, better health self-reports, less reliance on welfare, and greater earnings. Several economists have converted those outcomes into a cost-benefit metric, finding that, for every dollar that was spent on these early-education programs, the investment generated between five and 17 dollars in societal returns (such as taxes collected from the participants' higher earnings and savings achieved through lower rates of incarceration and dependence on public assistance). No other educational intervention — indeed, no other social-policy innovation — has yielded remotely comparable results, and these findings have fueled a nationwide movement to expand pre-kindergarten programs.

Since almost every child who attended Perry, Abecedarian, or the Child-Parent Centers was African-American, these studies established that high-quality early education could narrow the racial achievement gap. But did these programs help African-American boys in particular? The results from the Perry Preschool and Abecedarian studies are inconclusive on this point, largely because the number of students involved in each was relatively small (around 100 children in both cases). There can be no question, however, about the effects of the third pre-school initiative, the Chicago Child-Parent Centers, on the life trajectories of black males. Unlike Perry and Abecedarian, which were structured as short-term experiments, these pre-kindergartens have been run by the Chicago public-school system since 1967, and have enrolled tens of thousands of students. What's more, unlike the Perry and Abecedarian pre-schoolers, these poor, mostly black children live in inner-city neighborhoods, where life stresses are especially acute.

The CPC study followed 1,520 children — half of whom participated in the centers' language-focused early-education program — into their mid-twenties. The research found significant effects on educational attainment, as well as on incarceration rates and income. The largest impact was on high-school completion rates: While the girls in the program had graduation rates identical to girls who did not participate, 74% of the boys who attended a Child-Parent Center had graduated by age 24, compared with 57% of non-participating boys. These young men were also half again as likely to attend college, according to an analysis published last year by University of Minnesota professor Arthur Reynolds and his colleagues (who have been conducting the Child-Parent Center studies for a quarter-century).

Because these centers are staffed by regular school-district personnel, not specially trained professionals (as was the case for Perry Preschool and Abecedarian), and have been operating for more than 40 years, there is cause for confidence that the model can be scaled up. But the fact that only 10.4% of the male CPC students attended college — compared with 6.6% of males in the control group and 16.2% of the girls who enrolled in a Child-Parent Center — offers a pointed reminder that, while this program has been a success, it is no panacea. Society's investment in young black males can't stop with pre-school.

Unlike the Child-Parent Centers and state-funded pre-schools, which place a heavy emphasis on language, Head Start — the $7 billion-a-year federal early-education program — has historically stressed child development rather than academic preparation. Such an approach seems intuitively sensible, because Head Start-eligible children often lack basic health care and have a hard time regulating their emotions — both major obstacles to classroom success. But the program's implementation has been uneven, and multiple evaluations have yielded mostly disappointing results. A national study released earlier this year (the most rigorous to date) finds that, while Head Start modestly boosts reading and math skills during the time children are in the program, those gains disappear by first grade.

Still, we do not know if such dismal results tell the entire story. Remarkably, even though Head Start has been operating since 1965, no long-term experimental studies of the program have been conducted. In seminal articles published in 1995 and 2000, however, economists Janet Currie, Duncan Thomas, and Eliana Garces adopted an ingenious research approach: They compared the life histories of siblings who were of Head Start age in the mid-1970s, only one of whom attended Head Start. The economists found that white children who had been in Head Start were significantly more likely than their siblings to graduate from high school and to attend college; black children, meanwhile, were significantly less likely to have been convicted of a crime, but appeared to receive no education-related benefits from the program.

A 2009 study, by Carnegie Mellon public-policy professor David Deming, followed up on the work of Currie, Thomas, and Garces and yielded more encouraging results. Deming tracked the Head Start and non-Head Start siblings into their early twenties. What he discovered was striking — and did include positive educational effects for black students. Overall, Head Start participants were significantly less likely to have been left back or assigned to special education. They were more likely to have graduated from high school or earned a high-school equivalency diploma, and to have enrolled in college; they were also more likely to have had some work experience and to have stayed healthy.

While the studies do not specify the outcomes for black males, they do show that African-American children and the most disadvantaged children gained the most. Black students who participated in the program were 10.7% less likely to repeat a grade, 7.1% less likely to be diagnosed with a learning disability, 11.1% more likely to graduate from high school, and 13.6% more likely to enter college. Male students were 20.4% less likely to repeat a grade, 11.4% more likely to complete high school, and 10% less likely to be out of school and not working. It is reasonable to assume that Head Start had noteworthy long-term benefits for black males, but this is one area in which the familiar call for additional research is certainly warranted.

Early-education programs clearly seem to make a difference, but not by themselves. In a 2010 paper, Rucker Johnson — a public-policy professor at the University of California, Berkeley — examined the life histories of several thousand children who attended Head Start in the 1970s. By the time they were in their thirties, those who had attended well-funded elementary and secondary schools were more likely to have graduated from high school; they also had higher earnings, and were in better health, than their classmates in those schools who had not attended Head Start. For these children, Head Start mattered a lot. But for those who had attended poorly funded schools, Head Start made absolutely no difference; whatever advantages they had acquired as three- and four-year-olds were subsequently undone. The conclusion is obvious: For African-American boys, as for children generally, there are no quick fixes, no birth-to-five "solutions." Continuity from the earliest years onward is the key to success.

REFORMING SCHOOLS

Following up on successful early-education efforts through the K-12 education system has proved enormously difficult. As economists Roland Fryer and Steven Levitt showed in a 2006 study, the achievement gap for black male students grows steadily from kindergarten into the early grades, and only gets worse thereafter. For some researchers, this has justified abandoning the schoolhouse as a potential means for bridging the divide. "None of the various school-related policies is likely to play a major role in reducing the black-white achievement gap," Duke public-policy professor Helen Ladd asserted in 2008.

But here, too, a careful review of the data offers some cause for hope, as several evidence-based strategies do show promise. Some, like reducing class sizes, are well known. Other approaches — like comprehensive school and district reforms, or increased reliance on character-building charter schools — have received less attention from researchers in the field.

The logic behind reducing class sizes is straightforward: When classes are small, teachers can pay more attention to each pupil. Students experience continuing pressure to stay engaged; their attention to learning improves, as does their performance. But whether class size really improves student outcomes has been much debated.

The most carefully designed test of the effect of reducing class sizes has been the Tennessee STAR Project. During the 1980s, 11,600 students in kindergarten through third grade and 1,300 teachers in 76 schools in 42 districts participated in the experiment. Students were randomly assigned to small (13 to 17 students) and regular (22 to 28 students) classes, and the teachers were also randomly assigned. Being in a small class, economists Alan Krueger and Diane Schanzenbach found, made a noticeable difference in a student's academic performance: Math and reading scores improved, and more students took college-entrance exams, signaling at least an interest in continuing their education beyond high school. The published studies emerging from the experiment did not look at race-plus-gender differences, but Schanzenbach's unpublished re-analysis of the data (undertaken at my request) shows that, more than any other group, African-American boys benefited from small classes. Their test scores improved by a statistically significant 7.2 percentage points, slightly more than the improvement in the scores of African-American girls.

But while the African-American girls who had been in small classes enjoyed persistent academic gains through sixth grade — three years after they had returned to regular classes — and were more likely to take college-entrance exams several years later, the gains for African-American boys tapered off dramatically in fourth grade, as soon as they returned to regular classes. In a nurturing environment, black boys did well — just as they had in the intimate enclaves of the Child-Parent Centers and Abecedarian — but they lost that edge in the more impersonal world of the large classroom.

Experiments are one thing; widespread practice is quite another. In 1996, California adopted a policy of reducing class sizes for grades K-3 in an attempt to help poor and minority youth, but those students actually fared worse. A badly designed plan was the culprit. As economists Christopher Jepsen and Steven Rivkin reported,

The burdens of implementation [in California] fell disproportionately on urban schools suffering from poverty, overcrowding, and language barriers, and the need to provide many special services. The possible positive effects attributable to smaller classes were often mitigated in these schools because teacher quality was lower than in other schools, as more experienced teachers left to fill new openings in less troubled schools. Urban schools were left to fill not only the vacancies created by those who transferred out, but also the newly created slots. They did so by hiring inexperienced and uncertified teachers, with the result that one-quarter of the black students in high-poverty schools had a first- or second-year teacher, and nearly 30% had a teacher who was not fully certified.

The lesson is all too clear: If class-size reduction is to narrow the achievement gap, as it did in Tennessee, schools need to get the plan right; otherwise, their efforts can easily backfire.

Another often-cited means of helping schools better serve black students is desegregation. The underlying assumption, note economists Jacob Vigdor and Jens Ludwig, is that segregation "affects both the motivation of the students and their perception of the larger opportunity structure they face in society and their exposure to high-quality school resources or even the academic climate in the school."

Oceans of ink have been spilled by those attempting to show the relationship between segregation and student achievement. And historical trends do suggest some connection: As public schools became more desegregated, beginning in the 1960s, the achievement gap narrowed; as school segregation increased again, beginning around 1990, progress in closing that gap ground to a halt. What's more, blacks' biggest gains were in the Southern states, where the effects of desegregation were greatest.

Economists have done much of the recent research in this field. One study, by Jonathan Guryan of the University of Chicago, concluded that court-ordered desegregation in the 1970s reduced black high-school students' dropout rates by 2 to 3% during the 1970s — explaining half of the overall decline during that period — while having no comparable effect on white students. That improvement was reversed when desegregation orders were terminated in the 1990s. An even stronger case for linking desegregation and black achievement was made in a 2009 study of Texas public schools carried out by economists Eric Hanushek, John Kain, and Steven Rivkin, whose work compared test-score variations among students who attended the same school at different times, when the school's racial composition varied. Simply equalizing a school's racial composition, the analysis concludes, can reduce the achievement gap by about 25% — an impact as great as any reform strategy has had to date.

For the foreseeable future, however, desegregation on such a sweeping scale is not in the cards. The trend has been for public schools to become more, not less, segregated, a process abetted by the judiciary. Over the past generation, federal courts have stopped monitoring desegregation plans that school districts had implemented because of earlier court orders; in 2007, the Supreme Court went so far as to overturn voluntary desegregation plans in Seattle and Louisville. For those who remain committed to desegregation, the one hopeful sign is the long-term trend toward greater residential integration — though this trend is far more pronounced among middle-class families of different races than among the poor, where the achievement gap is greatest.

REFORMING DISTRICTS

Some of the most ambitious efforts to address the underachievement of African-American boys and young men have involved broad, systemic reforms — carefully conceived and clearly articulated transformations of entire school districts that are designed to run seamlessly from pre-kindergarten through high school.

The best-documented example is that of Montgomery County, Maryland — located just outside Washington, D.C. — whose experience was described by Stacy Childress, Denis Doyle, and David Thomas in their 2009 book, Leading for Equity. The Montgomery County school district encompasses two demographically different communities, one composed almost entirely of white and Asian professional families — residing in neighborhoods that school authorities refer to as the "green zone" — and another composed of mainly poor and minority families, who live in what the school district labels the "red zone." Since the late 1990s, Montgomery County has put enormous effort into "greening" the red zone. The statistics tell a story of considerable accomplishment.

 On each of a dozen metrics, from kindergarten reading evaluations to SAT scores, there has been improvement among all racial and ethnic groups in Montgomery County's schools. And by almost every measure, black as well as Hispanic students have narrowed the achievement gap. For instance, in 2009, 95% of third graders in the county scored at the proficient or advanced level on the state reading test — among them 80% of black children, a 100% improvement since 2003. More than 94% of white and 79% of black third-grade students scored at these levels in math; in six years, African-American children narrowed the racial gap in math from 22 to 16 percentage points.

In high school, 74% of black students took the SAT in 2009, compared with 66% in 2006 — again shrinking the gap with white students, whose participation rate rose from 82% to 84%. The number of Advanced Placement exams taken by African-American students almost tripled between 2003 and 2009 (the number taken by whites increased by 50%). Scores on one-third of the AP exams taken by black youth were three or higher (on a scale of one to five; three is the minimum score accepted for credit by most colleges). That's twice as many such scores in six years, compared with an increase of 40% for whites during the same period. (Though one-third is still well below the 61% national average for students of all races.)

System-wide reform drove these results. The pre-school-to-12th-grade "pipeline" in Montgomery County starts with a pre-kindergarten curriculum that emphasizes literacy and numeracy; a uniform, evidence-based curriculum is in place through high school. Student evaluation is ongoing, with regular feedback and coaching for teachers and tutoring for students who are falling behind. High-school students are pushed to take the SAT exam and Advanced Placement courses in order to increase their odds of being admitted to good colleges. Across the district, many schools have been transformed from 9 a.m.-to-3 p.m., 180-day-a-year operations into community hubs that are open during evenings, weekends, and summers.

Montgomery County has some unique advantages, especially its wealth. Although it has both high-income and low-income populations and is racially mixed, its wealthy residents are very wealthy: It is the tenth-richest county in the nation, and in 2008 had a median household income of $94,000 — almost twice the national average. Thus the county's substantial property-tax base has given administrators serious money to work with. Yet it is worth noting that the public-school systems in Montgomery County and the neighboring District of Columbia both spent about $15,000 per student in the 2007-2008 school year — and while Montgomery County has obviously gotten a respectable return on its investment, D.C. has performed dismally, ranking dead last in the nation on 2007 and 2009 NAEP math and reading scores for fourth graders. Montgomery County's success therefore suggests that a thoughtfully conceived, capably managed, and well-funded set of reforms can make a major difference.

Such efforts, however, have often proved difficult to scale up beyond the school-system level. Several comprehensive school-reform strategies — which entail changing curricula as well as restructuring evaluation and teaching practices — have shown decidedly mixed results. As Janet Quint and her colleagues at the Manpower Development Research Corporation reported in 2005, interventions with good results in one locale often flop when tried elsewhere.

Even so, there have been a few promising exceptions. For example, "Success for All" — a school-wide, first-through-fourth-grade strategy that emphasizes improvement in reading skills — has largely overcome these implementation problems. Since its launch in Baltimore in 1987, the program has been used with more than a million students in some 1,600 schools across the country; almost all of the schools have been located in high-poverty communities, and about half of the students have been black. Evidence of the program's overall effectiveness in boosting reading scores — not just while children are participating in the program, but for several years afterward — has been confirmed in more than 40 evaluations, including a national randomized experiment. The biggest gainers are low-achieving and minority students: The minority-white achievement gap is actually halved in the typical Success for All school district. And an unpublished analysis carried out at my request by Johns Hopkins University psychologist Robert Slaving — a co-author of a number of these studies, and the founder of Success for All — found that, on average, black males participating in Success for All gained between one month and four months of reading ability (a level of improvement comparable to that of black females).

Why might Success for All work so well for black male students? Close attention to the particular needs of each student is the centerpiece of the program's educational model. Unlike in most schools, where what happens in the classroom stays in the classroom, teachers in Success for All programs regularly discuss with one another how each student is performing. The reading classes are grouped by achievement, not grade level, and are taught by someone other than the student's regular classroom teacher, so that more than one teacher is intimately familiar with each student. Students who are struggling academically receive extra tutoring. And each school has a parent coordinator whose job is to engage families, enlist their support, and help them with pressing concerns like access to social services. The key elements of Success for All are precisely the interventions that, from pre-school to high school, turn out to be most helpful in meeting the unique needs of African-American boys.

REFORMING STUDENTS

Changing schools can make some difference — but what about changing the mindsets of students themselves? Some remarkable research in recent years has shown that students' theories and beliefs about intelligence can play a powerful role in how they assess their own learning capacity. If they believe that intelligence is fixed and outside their control, they pay "an emotional tax that is a form of intellectual emasculation" — as psychologists Catherine Good, Joshua Aronson, and Michael Inzlicht put it in a recent study — thus becoming more likely to give up and disengage from learning.

The good news is that this destructive psychological dynamic can be reversed. When students appreciate the fact that intelligence is malleable, and therefore within their control, they grow inclined to work harder — and naturally perform better in school. College students who are exposed to information about brain development that shows the plasticity of intelligence, Aronson and several colleagues have demonstrated, "reported greater enjoyment of the academic process, greater academic engagement, and obtained higher grade point averages than students in the control group." Moreover, Aronson and his fellow researchers report, "While white students benefited to some extent from this effort to change students' mindsets, the benefits to black males were far more substantial." What is remarkable about Aronson's findings is how little exposure to such material is required to improve students' perceptions: Just three sessions in which intelligence was presented as malleable "created an enduring and beneficial change. [Students] reported enjoying and valuing academics more and they received higher grades."

What works for college students can also work for middle-school students; their beliefs about the nature of intelligence can be changed, and that change can improve their performance in school. Psychologists Lisa Blackwell and Kali Trzesniewski have demonstrated that when children who believe intelligence to be immutable are taught (in just four class sessions) about how learning changes the brain, they set higher learning goals, are more likely to think that making an effort can pay off, are more motivated to succeed — and do significantly better in math. "Theories of intelligence can be manipulated in real world contexts and have a positive impact on achievement outcomes," Blackwell and Trzesniewski conclude.

Initiatives designed to change students' attitudes more generally — not just attitudes about their own potential, but attitudes about the value of work, discipline, and education — can also make a real difference. Here, charter schools have been at the forefront. Those that succeed in narrowing the achievement gap for African-American boys often do so by stressing old-fashioned character education, which they regard as a prerequisite for academic success. Preliminary research supports this contention: A 2006 study of 164 mostly African-American eighth graders, carried out by psychologists Angela Duckworth and Martin Seligman, examined the correlation between students' grade-point averages and their IQ scores, and then looked at the correlation between their GPAs and scores on a test that measures self-discipline. The self-discipline test proved twice as good a predictor of GPA as IQ.

The best examples of charter schools that make the development of character and self-discipline an important part of students' education are systems such as the Knowledge Is Power Program (or KIPP), a national network of 99 schools, and Green Dot Public Schools, an organization that runs 19 inner-city high schools in Los Angeles and New York. Both KIPP and Green Dot enroll poor and mainly minority children, and both have an enviable track record with black adolescent males. Their success can be attributed partly to the strategies used in Montgomery County and many of the Success for All schools: a tight link between evaluation and instruction, and a culture of accountability and ongoing self-examination. But the explicit emphasis on character is what sets the charter-school networks apart.

Alt hough these schools are open to all, the fact that students must apply for admission — rather than being admitted automatically, as is the case in most public schools — makes comparisons tricky. Yet even taking into account the possible differences in students' and parents' levels of motivation, the academic performance gap between these charter schools and public schools that serve similar students is striking. At a Green Dot school in one of the poorest neighborhoods in Los Angeles, for instance, 68% of the African-American male students graduated in four years, matching the national average for all high-school students. By comparison, at one nearby high school, only 9% of black males graduated on time, and at another, only 3% did so. At a KIPP Academy in the South Bronx — one of the worst-off sections of New York City — 86% of eighth-grade students scored at grade level in math in 2006, compared with 16% of all eighth graders in the community. "I think we have to teach work ethic in the same way we have to teach adding fractions with unlike denominators," says Dacia Toll, co-CEO of Achievement First, which operates successful inner-city charter schools in Connecticut and Brooklyn, New York. "But once children have got the work ethic and the commitment to others and to education down, it's actually pretty easy to teach them."

It is unlikely that charters like Green Dot and KIPP can be replicated on a scale big enough to "tip" public education nationwide. Still, says Toll, these schools "change the public conversation from ‘you can't educate these kids' to ‘you can only educate these kids if....' " When it comes to closing the achievement gap, this is surely a change for the better.

BEYOND THE SCHOOLHOUSE

American students generally attend school for just six hours a day, 180 days a year. Since children spend most of their time — and do much of their learning and developing — outside of the classroom, what happens beyond the schoolhouse obviously shapes how they fare in class. Unfortunately, for too many students, what happens "on the outside" undermines the best efforts and intentions of teachers and administrators at their schools. Many schools lock up tightly at 3 p.m., sending children into empty houses, barren neighborhoods, or crime-ridden streets — often with dire consequences.

As education researcher Patricia Lauer has found, the amount of time that children spend hanging out on street corners with their friends after school is actually a better predictor of poor academic performance than is family income. This is particularly problematic for poor black students: They are disproportionately likely to live in neighborhoods of "concentrated disadvantage," where once-vibrant institutions have been shuttered; where streets and parks and other public spaces are danger magnets; and where crime, poverty, unemployment, and teen-pregnancy rates are sky-high. A 2008 study by Harvard sociologist Robert Sampson concluded that growing up in a neighborhood of concentrated disadvantage has the same effect on a five-year-old's verbal ability as missing an entire year of school. What makes Sampson's finding particularly striking is that his research compared the consequences of being raised in Chicago's best-off and worst-off African-American neighborhoods; no comparison was made with white neighborhoods, since no white neighborhood was remotely comparable to the worst-off parts of the city's South Side.

Transforming the traditional school into a "community school" — one that is open from dawn until dusk, including on weekends and during the summer, and that offers medical, social, and psychological supports as well as academic help, sports, and activities — can help to offset these disadvantages. Though not parsed by race and gender, a number of studies show that such schools have positive effects on an array of educational outcomes: Students who spend time on after-school and summer projects have higher math and reading scores than their gone-at-3 p.m. classmates; a 2006 study by Lauer and her colleagues also finds that these students have better attendance records and exhibit better classroom behavior.

Mentoring is another strategy that has proved effective in combating the social isolation pervasive among black youth. "African-American males are less likely than females to feel capable academically," observes University of Pittsburgh social-work professor Larry Davis. "There is little social pressure to graduate from high school and they often don't understand the economic returns from education." Introducing a stable and caring adult — a mentor — into these students' lives can combat those attitudes, since mentors can help make connections between school and life while fostering students' self-confidence.

Experience Corps, a national non-profit that enrolls adults age 55 and older to serve as reading tutors for children in kindergarten through third grade, offers one example of how this approach can work. A 2009 evaluation that randomly paired 1,000 very poor readers with mentors found that, over the course of a single school year, the students who had been tutored by Experience Corps volunteers made 60% more progress in reading comprehension than those who did not have tutors. An unpublished re-analysis of that data, performed at my request, shows that the effects did not differ across gender or race.

Friends of the Children, which was founded in 1993 in Portland, Oregon, and now has affiliates in five additional communities, is much more intensely involved with the students it supports. Staffers visit kindergarten classrooms to identify children who, even at their young age, are seen by their teachers as likely future candidates for prison or early pregnancy (because the students carry a great deal of family and social baggage and because, even as five-year-olds, they are extremely hard to manage in the classroom). Once selected by the non-profit, these children are mentored until they graduate from high school. The program's track record is astonishing: Of the more than 200 graduates of Friends of the Children's Portland affiliate, about half of whom are African-American, 82% have earned high-school diplomas. This rate is 13% higher than the national average — a feat made even more remarkable when one considers that 68% of these children were the first members of their families to complete high school. Moreover, 40% of the high-school graduates have gone on to college, which is also higher than the national average. Even though 60% of the students who have participated in Friends of the Children have had at least one parent who has been incarcerated, 92% have stayed out of the juvenile-justice system. The single best predictor of teen pregnancy is being the child of a teen mother — but while 61% of the girls in the program were born to unwed teenage mothers, only 2% became teen mothers themselves. The one downside of the Friends of the Children program is its expense: about $100,000 for each child over the course of 12 years. But a 2010 study by the Oregon Harvard Business School Association found that the program's societal benefits — derived from greater educational achievement and lower rates of incarceration and teen motherhood — outweighed costs by more than six to one.

In New York City, almost all of the roughly 100 students in Friends of the Children are African-American. While that program has not yet graduated its first cohort, its record through the early years of high school is impressive: Not a single student has dropped out, and the promotion rate in school is 98% (the only exceptions are a student who transferred to a parochial school where he was asked to repeat a grade, and a boy who lost a month of schooling because of family turmoil). The students have done well in some of the city's top charter and private schools, as well as in selective public schools: In 2009, for the fifth straight year, their reading and math test scores were better than the averages for their schools. All of the students but one have stayed out of the juvenile-justice system (one boy brought a BB gun to school). The only girl who had a child put her baby up for adoption and stayed in school. And according to Robert Houck, executive director of Friends of the Children, the boys in the program have done as well in school as the girls.

Friends of the Children is too labor intensive and expensive to serve as a national model. But like the Green Dot and KIPP charter-school systems, it shows that the doomsayers are wrong: The intergenerational cycle of disadvantage can be broken for black boys and young men, and there is more than one way to break it.

AN EQUAL CHANCE

This same conclusion applies across the entire range of techniques described above. No single strategy or program offers an easy cure-all for the achievement gap in education — but each helps to demonstrate that fatalism is not the right response. Solutions are possible, and indeed a surprising range and number of strategies result in genuine improvement.

Implementation is, of course, the make-or-break factor in any reform program, and effective implementation requires addressing communities' unique circumstances and meeting students' particular needs. After all, as the research demonstrates, the same idea can succeed in one time and place while failing in another. And just as no approach would work in every setting, no single approach can best serve the same child at every point in his development. Closing the achievement gap requires a well-planned sequence of interventions, beginning with help for parents of infants and continuing all the way through college. What is needed — to borrow the words of Geoffrey Canada, founder of the much-touted Harlem Children's Zone — is the "conveyor belt" approach.

This should be self-evident. After all, would any of us think that our obligation to a child we love ends after two years at a high-quality pre-school or a top-flight primary school, and that he should then be left to sink or swim? Why should things be different for other people's children — especially for African-American boys, many of whom wear a doomed-to-fail label from the day they arrive at kindergarten?

It is economically smart to invest in these children, since those investments will repay society many times over. It is also the right thing to do. And it can work. Careful, thorough, empirical research points not to fatalism and surrender, but to genuine hope that our best intentions might actually be compatible with responsible public policy. When those two forces align, it is hard to justify remaining idle — especially when so much is at stake.

David L. Kirp is a professor in the Goldman School of Public Policy at the University of California, Berkeley, and the author of, most recently, Healthy, Wealthy, and Wise: Five Big Ideas for Transforming Children's Lives (forthcoming). A version of this essay will appear in Building Healthy Communities: A Focus on Boys and Young Men of Color (Christopher Edley and Jorge Velasco-Ruiz, eds.), forthcoming this fall from the University of California Press.