Skip over navigation

Journal Issue: The Next Generation of Antipoverty Policies Volume 17 Number 2 Fall 2007

Improving the Education of Children Living in Poverty
Richard J. Murnane

Create Incentives for States to Act

Congress could create incentives for states to strengthen high school graduation requirements to reflect the skills that students need voluntary interdistrict school choice programs.

Strengthen High School Graduation Requirements
Today twenty-two states require high school students to pass exit exams in mathematics and English language arts to earn a high school diploma.18 But passing these exams does not mean that students are ready either for college or for the demands of jobs with promising futures. Although more than 70 percent of high school graduates enter twoand four-year colleges, more than a quarter must take remedial courses in English and mathematics before registering for courses that provide college credit, and the share is much higher for disadvantaged students. More than 60 percent of employers rate high school graduates’ skills in writing and basic math as only “fair” or “poor.”19

To give educators and students clear signals about the adequacy of the work they do together in high schools, states should align high school standards, assessments, and graduation requirements with the knowledge and skills needed for postsecondary education and work. Public colleges and universities could create incentives for high school students to master the more demanding skills required for high school graduation by committing to base college course placement on students’ scores on recalibrated state exams. Knowing that scoring well on high school exit exams would guarantee acceptance into college courses that count toward a degree (as opposed to being funneled into “developmental courses,” which do not) would encourage students to do the hard work needed to master important skills.

States that strengthen high school graduation requirements would be likely to strengthen content standards in the earlier grades to prepare students to do more demanding high school work. The variation across states in standards and assessments would likely diminish. Moving toward a common set of national standards and assessments makes sense in a country with a mobile population and an increasingly integrated economy.

Care must be taken in determining precisely which skills are important for success after high school graduation. The tendency is to ratchet up standards in areas such as mathematics, where skills are relatively easy to measure, and to neglect skills such as oral communication, teamwork, and job search and interviewing that are critical to success in postsecondary education and work but are hard to measure.20

A ten-year study of career academies illustrates the importance of skills other than reading and math to success after high school. Career academies are small learning communities embedded within a larger high school, whose students take classes together for at least three years from a team of teachers drawn from different disciplines. The academies offer a college preparatory curriculum with a career theme, which enables students to identify relationships among academic subjects and understand how they are applied in a broad field of work. The academies generally include partnerships with local employers, who provide work-based learning opportunities for their students.

In 1993 MDRC, one of the nation’s leading contract research firms, undertook an experimental study of the effect of nine career academies serving large shares of students living in poverty. Because there was excess student demand for all nine academies, lotteries determined which interested students were offered places. Both the students who were offered places (the treatment group) and those who lost out in the lottery and enrolled in other school programs (the control group) were followed through high school and for four years after graduation. A variety of indicators of success (reading and math scores, course grades, on-time graduation, college enrollment and completion, labor market earnings) were measured for all participants.

The findings of the evaluation are striking. In both treatment and control groups, academic skills, high school graduation rates, and college enrollment rates were higher, on average, than the national average for students with similar demographic characteristics. (These credentials reflect the greater than average motivation of students who wanted to enroll in career academies.) However, at the end of high school the math and reading skills of students in the treatment group were no higher, on average, than those in the control group. Nonetheless, young men who had been offered places in a career academy earned $10,000 (18 percent) more than men in the control group during the four-year follow- up period after high school. The labor market benefits were especially large for male students who were at risk of dropping out of high school as the experiment began. The explanation for this striking pattern is that enrollment in career academies and the associated opportunities for workplace internships and jobs enabled students to acquire skills that were important to labor market success even though they were not captured by scores on standardized reading and math tests.21

Congress could provide funding to help states strengthen high school graduation requirements when it reauthorizes the Higher Education Act or NCLB. Some states have already begun work. Through the auspices of the American Diploma Project (ADP), a project of the organization Achieve, five states worked together in 2003–04 to develop benchmarks describing the specific English and mathematics skills needed for success in postsecondary education or in jobs with growth potential. Thirty states are now working to align high school standards with the demands of postsecondary education and work, and a dozen are also upgrading their high school graduation requirements.22

In most states realigning high school graduation requirements will entail redesigning exit examinations. As is almost always the case when new exams are introduced, scores will initially be poor but will improve as educators learn to prepare students for them. The question will inevitably arise whether improved exit exam scores reflect better preparation of students for postsecondary education and work or simply score inflation resulting from narrowly focused test preparation. To answer this question, it is necessary also to align the twelfth-grade NAEP English language arts and mathematics examinations with the skills needed for postsecondary education and work and require all states to administer these examinations.

Today the federal government requires states to participate in the NAEP assessment of the English language arts and mathematical skills of students in grades four and eight. As noted, comparisons of the performance of students on the NAEP tests and on mandatory state tests have revealed how undemanding many state tests are and how low many states have set their thresholds for proficiency. It is important to have similar nationally comparable benchmarks against which to judge states’ high school graduation requirements, including their exit examinations. Requiring all states to administer twelfth-grade NAEP tests could provide these benchmarks if two challenges can be overcome. The first is to redesign the NAEP tests to be sure they reflect the skills needed for postsecondary education and work. Progress on this front is under way. In August 2006 the National Assessment Governing Board, the group that sets policy for the NAEP, voted to redesign the NAEP twelfth-grade mathematics examination in accord with skills necessary for postsecondary education and jobs with growth potential.23 In taking this action the board accepted advice requested from Achieve, sponsor of the American Diploma Project (ADP). The revised NAEP twelfthgrade math examination will thus likely be informed by the work of the ADP.

The second challenge is to convince twelfthgrade students to make their best efforts in answering questions on the NAEP examinations when their scores not only have no consequences for them but are never even known to them. Only if the students give their best effort will the scores serve as a useful audit of the consequences of revising high school graduation requirements. Whether it is possible to elicit the full attention and effort of twelfth-grade students under those circumstances remains to be seen, though several recent experiments show some promise.24

Some readers may wonder why states do not simply require that students score above predetermined cut-offs on redesigned twelfthgrade NAEP tests in order to receive a high school diploma. In other words, why not get high school students to take the NAEP tests seriously by making the scores count? There are two complementary answers. First, the NAEP uses a matrix sampling design under which different students are asked to answer different questions. The design permits reliable estimation of the extent to which groups of students have mastered a much broader range of skills than would be the case if all students answered the same questions. But as a consequence, scores are not computed for individual students. Second, critics of test-based accountability often complain that test score gains on high-stakes tests stem from extensive drilling and do not reflect increases in students’ mastery of the relevant subject domains.25 The only way to assess the extent to which this is true is to compare score trends on the high-stakes test with those on a different, broad-based examination. The NAEP tests are designed to serve this audit function.

Promote Interdistrict School Choice
No Child Left Behind requires school districts to give students the option of transferring to a more successful public school if their own school fails to make adequate yearly progress for two years in a row. And the law gives low-achieving children from low-income families priority in requesting transfers. To date, however, this school choice option has been little used, for several reasons. Successful public schools, especially in urban areas, rarely have empty seats and often have long waiting lists.26 And many school superintendents give parents little or no information about the school choice option. Finally, scores on state tests taken in the spring of one school year are often not available until after the next school year has begun. (A solution to this problem is to base the NCLB choice option on the most recent accountability data available when families are choosing schools for the next school year.)

Because NCLB as written gives neighboring school districts no real reason to accept students from failing urban schools, the next round of legislation should create strong incentives for states to develop voluntary interdistrict choice programs. Several promising precedents exist. METCO, a grant program funded by the Commonwealth of Massachusetts with the voluntary participation of thirty-four suburban districts near Boston and four near Springfield, has been in operation since 1966 (when it was funded in part by a grant from the U.S. Department of Education). Today it enables 3,300 low-income Boston and Springfield students to attend public schools in other communities. St. Louis also has a voluntary program, under which 12,000 African American children, 75 percent of whom are from low-income families, attend public schools in sixteen suburban districts.27 Significant state funding provides an incentive for suburban districts to participate in these interdistrict choice programs.

As the METCO and St. Louis programs show, with appropriate incentives, suburban school districts serving primarily middle-class children are willing to educate a nontrivial number of low-income urban students. A recent evaluation found METCO a promising approach to improving the education of some children living in poverty. It increased the reading achievement of participating urban children attending suburban elementary schools and had no lasting negative effects on the achievement of their suburban classmates. 28 Evidence from the St. Louis choice program is also encouraging.29 Competitive grants to states for the design and implementation of interdistrict choice programs could make school choice under NCLB a real option for many children from low-income families.

Congress should also amend adequate yearly progress regulations to ensure that suburban districts are not penalized for accepting urban students from low-performing schools. The option of satisfying adequate yearly progress requirements by demonstrating gains in the achievement of initially lowachieving children could be important in this regard.

Although creating interdistrict choice options for low-income children who attend poorly performing schools is important, such programs are likely to serve only a minority of urban children. The reason is that the willingness of suburban communities to voluntarily accept low-income students from urban school districts would diminish as the share of these students in their schools grew. Improving teaching and learning in schools serving high concentrations of poor children must thus be a central part of federal education policy.