Closing the Achievement Gap

But that is changing. Good. Because if we don't get the numbers out on the table and talk about them, we're never going to close the gap once and for all. I worry, though, about how many people head into discussions without accurate data. And I worry even more about how many education leaders have antiquated—and downright wrong—notions about the whys beneath the achievement gap.

I want to respond to both these worries by putting some crucial data on the table and by sharing what both research and experience teach us about how schools can close the gaps between groups of students. Most of the data are from standard national sources, including the National Center for Education Statistics (NCES) and the National Assessment of Education Progress (NAEP), as well as from states and local school districts that have been unusually successful at educating poor and minority students.¹ 

Understanding Achievement Patterns

The performance of African American and Latino youngsters improved dramatically during the 1970s and 1980s. The 1990s, however, were another matter. In some subjects and at some grade levels, the gaps started growing; in others, they were stagnant (National Center for Education Statistics, 2001).

• Reading achievement among 17-year-old African Americans and Latinos climbed substantially through the 1970s and 1980s, but gaps separating them from other students widened somewhat during the 1990s.
• The patterns in mathematics achievement look similar for 13-year-olds, with the African American and white gap reaching its narrowest in 1990 and the Latino and white gap narrowing until 1992, and the gaps widening thereafter.

In 1999, by the end of high school

• Only 1 in 50 Latinos and 1 in 100 African American 17-year-olds can read and gain information from specialized text—such as the science section in the newspaper (compared to about 1 in 12 whites), and
• Fewer than one-quarter of Latinos and one-fifth of African Americans can read the complicated but less specialized text that more than half of white students can read.

The same patterns hold in math.

• About 1 in 30 Latinos and 1 in 100 African Americans can comfortably do multistep problem solving and elementary algebra, compared to about 1 in 10 white students.
• Only 3 in 10 African American and 4 in 10 Latino 17-year-olds have mastered the usage and computation of fractions, commonly used percents, and averages, compared to 7 in 10 white students.

By the end of high school, in fact, African American and Latino students have skills in both reading and mathematics that are the same as those of white students in 8th grade. Significant differences also persist in the rates at which different groups of students complete high school and in their postsecondary education experiences.

• In the 18- to 24-year-old group, about 90 percent of whites and 94 percent of Asians have either completed high school or earned a GED. Among African Americans, the rate drops to 81 percent; among Latinos, 63 percent.
• Approximately 76 percent of white graduates and 86 percent of Asian graduates go directly to college, compared to 71 percent of African American and 71 percent of Latino graduates.
• Young African Americans are only about half as likely as white students to earn a bachelor's degree by age 29; young Latinos are only one-third as likely as whites to earn a college degree (see fig. 1).

Figure 1. Highest Educational Attainment for Every 100 Kindergartners

What's Going On?

Over the past five years, staff members at the Education Trust have shared these and related data on the achievement gap with hundreds of audiences all over the United States. During that time, we've learned a lot about what people think is going on.

When we speak with adults, no matter where we are in the country, they make the same comments. "They're too poor." "Their parents don't care." "They come to school without an adequate breakfast." "They don't have enough books in the home." "Indeed, there aren't enough parents in the home." Their reasons, in other words, are always about the children and their families.

Young people, however, have different answers. They talk about teachers who often do not know the subjects that they are teaching. They talk about counselors who consistently underestimate their potential and place them in lower-level courses. They talk about principals who dismiss their concerns. And they talk about a curriculum and a set of expectations that feel so miserably low-level that they literally bore the students right out the school door.

When we ask, "What about the things that the adults are always talking about—neighborhood violence, single-parent homes, and so on?"—the young people's responses are fascinating. "Sure, those things matter," they say. "But what hurts us more is that you teach us less."

The truth is that the data bear out what the young people are saying. It's not that issues like poverty and parental education don't matter. Clearly they do. But we take the students who have less to begin with and then systematically give them less in school. In fact, we give these students less of everything that we believe makes a difference. We do this in hundreds of different ways.

Let me be clear. It would help if changes were made outside of schools, too: if parents spent more time with their children, if poverty didn't crush so many spirits, and if the broader culture didn't bombard young people with so many destructive messages. But because both research and experience show that what schools do matters greatly, I'll concentrate on what works in education.

Lesson 1: Standards Are Key

Historically, we have not agreed on what U.S. students should learn at each grade level—or on what kind of work is good enough. These decisions have been left to individual schools and teachers. The result is a system that, by and large, doesn't ask much of most of its students. And we don't have to go far to find that out: Ask the nearest teenager. In survey after survey, young people tell us that they are not challenged in school.

The situation is worse in high-poverty and high-minority schools. For the past six years, our staff at the Education Trust has worked with teachers who are trying to improve the achievement levels of their students. But while we've been observing these high-poverty classrooms, we've also looked carefully at what happens there—what kinds of assignments teachers give, for example—compared to what happens in other classrooms.

We have come away stunned. Stunned, first, by how little is expected of students in high-poverty schools—how few assignments they get in a given school week or month. Stunned, second, by the low level of the few assignments that they do get. In high-poverty urban middle schools, for example, we see a lot of coloring assignments, rather than writing or mathematics assignments. Even at the high school level, we found coloring assignments. "Read To Kill a Mockingbird," says the 11th grade English teacher, "and when you're finished, color a poster about it." Indeed, national data make it clear that we expect so little of students in high-poverty schools that we give them As for work that would earn a C or D anywhere else.

Clear and public standards for what students should learn at benchmark grade levels are a crucial part of solving the problem. They are a guide—for teachers, administrators, parents, and students themselves—to what knowledge and skills students must master.

Kentucky was the first state to embrace standards-based reform. Ten years ago, the Kentucky legislature put out an ambitious set of learning goals and had the audacity to declare that all of its children—even the poorest—would meet those goals. Leaders in Kentucky are the first to acknowledge that they are not there yet. But their progress is clear and compelling. And poor children are, in fact, learning in all subjects. For example, in reading, 7 of the 20 top-performing elementary schools are high-poverty; in math, 8 of the top 20 are high-poverty; in writing, 13 of the top 20 are high-poverty.

Lesson 2: All Students Must Have a Challenging Curriculum

Standards won't make much of a difference, though, if they are not accompanied by a rigorous curriculum that is aligned with those standards. Yet in too many schools, some students are taught a high-level curriculum, whereas other students continue to be taught a low-level curriculum that is aligned with jobs that no longer exist.

Current patterns are clearest in high schools, where students who take more-rigorous coursework learn more and perform better on tests. Indeed, the more-rigorous courses they take, the better they do.

• In mathematics, students who complete the full college preparatory sequence perform much higher on the National Assessment of Educational Progress (NAEP) than those who complete only one or two courses.
• The reverse is true of watered-down, traditional "vocational" courses. The more vocational education courses students take, the lower their performance on the NAEP.
• Although some of these differences are clearly attributable to the fact that higher-scoring students are often assigned to tougher classes, careful research shows the positive impact of more-rigorous coursework even on formerly low-achieving students.

Since 1983, we've made progress in increasing the number of students who take a rigorous, college-preparatory curriculum. But the pace is not fast enough.

Almost three-quarters of high school graduates go on to higher education, but only about half of them complete even a mid-level college-preparatory curriculum (four years of English and three years each of math, science, and social studies). If we also include two years of a foreign language and a semester of computer science, the numbers drop to about 12 percent. The numbers are worse for African Americans, Latinos, and low-income students.

These patterns are disturbing because the quality and intensity of high school coursework are the most important determinants of success in college—more important than class rank or scores on college admissions tests (Adelman, 1998). Curriculum rigor is also important for work-bound students (Bottoms, 1998).

A few years ago, the chancellor of the New York City schools required all 9th graders to take the Regents math and science exams. Though many people were worried that failure rates would be astronomical, in one year the number of Latinos in New York City who passed the Regents science exam tripled, and the number of African Americans who passed doubled. Other groups also had gains in science and mathematics. Did they all pass? No, they didn't. But as a principal friend of mine used to say, "At least they failed something worthwhile." And remember, these youngsters previously would never even have been given a chance to learn higher-order content.

Lesson 3: Students Need Extra Help

Ample evidence shows that almost all students can achieve at high levels if they are taught at high levels. But equally clear is that some students require more time and more instruction. It won't do, in other words, just to throw students into a high-level course if they can't even read the textbook.

One of the most frequent questions we are asked by stressed-out middle and high school teachers is "How am I supposed to get my students ready to pass the (fill-in-the-blank) grade test when they enter with 3rd grade reading skills and I have only my 35-minute period each day?"

The answer, of course, is "You can't." Especially when students are behind in foundational skills like reading and mathematics, we need to double or even triple the amount and quality of instruction that they get.

Around the United States, states and communities are wrestling with how best to provide those extras. Kentucky gives high-poverty schools extra funds every year to extend instruction in whatever way works best for their community: before school, after school, weekends, or summers. Maryland provides a wide range of assistance to students who are not on track to pass its new high school graduation test. And San Diego created more time, mostly within the regular school day, by doubling—even tripling—the amount of instructional time devoted to literacy and mathematics for low-performing students and by training all of its teachers.

Lesson 4: Teachers Matter a Lot

If students are going to be held to high standards, they need teachers who know the subjects and know how to teach the subjects. Yet large numbers of students, especially those who are poor or are members of minority groups, are taught by teachers who do not have strong backgrounds in the subjects they teach.

• In every subject area, students in high-poverty schools are more likely than other students to be taught by teachers without even a minor in the subjects they teach (see fig. 2).
• The differences are often greater in predominantly minority high schools. In math and science, for example, only about half the teachers in schools with 90 percent or greater minority enrollments meet even their states' minimum requirements to teach those subjects—far fewer than in predominantly white schools.
• The patterns are similar regardless of the measure of teacher qualifications—experience, certification, academic preparation, or performance on licensure tests. We take the students who most depend on their teachers for subject-matter learning and assign them teachers with the weakest academic foundations.

Figure 2. Percentage of Underqualified* Teachers in High School Classrooms

A decade ago, we might have said that we didn't know how much this mattered. We believed that what students learned was largely a factor of their family income or parental education, not of what schools did. But recent research has turned these assumptions upside down. What schools do matters enormously. And what matters most is good teaching.

• Results from a recent Boston study of the effects teachers have on learning are fairly typical (Boston Public Schools, 1998). In just one academic year, the top third of teachers produced as much as six times the learning growth as the bottom third of teachers. In fact, 10th graders taught by the least effective teachers made nearly no gains in reading and even lost ground in math.
• Groundbreaking research in Tennessee and Texas shows that these effects are cumulative and hold up regardless of race, class, or prior achievement levels. Some of the classrooms showing the greatest gains are filled with low-income students, some with well-to-do students. And the same is true with the small-gain classrooms. It's not the kids after all: Something very different is going on with the teaching (Sanders & Rivers, 1996).

Findings like these make us wonder what would happen if, instead of getting far fewer than their fair share of good teachers, underachieving students actually got more. In a study of Texas school districts, Harvard economist Ronald Ferguson (1998) found a handful of districts that reversed the normal pattern: Districts with initially high-performing (presumably relatively affluent) 1st graders hired from the bottom of the teacher pool, and districts with initially low-performing (presumably low-income) 1st graders hired from the upper tiers of the teacher pool. By the time their students reached high school, these districts swapped places in student achievement.

El Paso, Texas, is a community that has taken such research seriously. Eight years ago, despite the extraordinarily high poverty of their city, local education leaders set some very high standards for what their students should know and be able to do. Unlike other communities, though, they didn't stop there. At the University of Texas, El Paso, the faculty revamped how it prepared teachers. New elementary teachers, for example, take more than twice as much math and science as their predecessors. More to the point, though, the teachers of these courses are math and science professors who themselves participated in the standard-setting process and who know, at a much deeper level, what kinds of mathematical understanding the teachers need.

The community also organized a structure—the El Paso Collaborative—to provide support to existing teachers and to help them teach to the new standards. The collaborative sponsored intensive summer workshops, monthly meetings for teachers within content areas, and work sessions in schools to analyze student assignments against the standards. The three school districts also released 60 teachers to coach their peers.

The results are clear: no more low performing schools and increased achievement for all groups of students, with bigger increases among the groups that have historically been behind.

An Academic Core

El Paso and the other successful communities and states have a lot to teach us about how to raise overall achievement and close gaps. Each community, of course, does things a little bit differently. What we learn is the value of a relentless focus on the academic core. Clear and high standards. Assessments aligned with those standards. Accountability systems that demand results for all kinds of students. Intensive efforts to assist teachers in improving their practice. And extra instruction for students who need it.

References

Adelman, C. (1998). Answers in the toolbox. Washington, DC: U.S. Department of Education.

Boston Public Schools. (1998, March 9). High school restructuring. Boston: Author.

Bottoms, G. (1998). High schools that work. Atlanta, GA: Southern Regional Education Board.

Ferguson, R. (1998). Can schools narrow the black-white test score gap? In C. Jencks & M. Phillips (Eds.), The black-white test score gap (pp. 318–374). Washington, DC: The Brookings Institute.

National Center for Education Statistics. (2001). NAEP summary data tables [Online]. Washington, DC: U.S. Department of Education. Available: http://nces.ed.gov/nationsreportcard

Sanders, W., & Rivers, J. (1996). Cumulative and residual effects of teachers on future student academic achievement. Knoxville, TN: University of Tennessee Value-Added Research and Assessment Center.