Most conventional education reforms concentrate on achieving equality between separate schools for rich and poor. From private-school vouchers and charter schools to class-size reduction and education testing, mainstream efforts ignore a central finding of education research: that schools that are majority poor tend to fail to produce high levels of academic achievement, no matter the level of funding or model of school governance employed. Forty years ago, legendary sociologist James Coleman found that, after the socioeconomic status of a child’s family, the biggest predictor of academic achievement is the socioeconomic status of the school he or she attends. While there are anecdotal stories of high poverty schools that work, University of Wisconsin researcher Douglas Harris found that middle-class schools are 22 times as likely to be consistently high performing as high-poverty schools. So in a majority–middle class country, why not allow more children to attend majority–middle class schools?

Middle-class schools perform better in part because middle-class students on average receive more support at home and come to school better prepared. But the vastly different educational environments typically found in middle-class and high-poverty schools also have a profound effect on achievement. On the 2005 National Assessment of Educational Progress (NAEP) given to fourth graders in math, for example, low-income students attending more affluent schools were almost two years ahead of low-income students in high-poverty schools. Indeed, low-income students given a chance to attend more-affluent schools performed more than half a year better, on average, than middle-income students who attend high-poverty schools. And it’s not just money that matters; it’s the people in and around a school. Every school community has three sets of actors: children; teachers and administrators; and parents. Middle-class schools tend to provide positive, well-disciplined peer role models, good teachers, and an active parent volunteer community, which benefit all the students in a school.

Recognizing this research, about 40 school districts have begun to take steps to provide a better socioeconomic balance in schools. In Wake County (Raleigh), North Carolina, for example, the school board adopted a policy in 2000 that no school should have more than 40 percent of its students eligible for free or reduced-price lunch. Unlike the old compulsory-busing programs of the 1970s, most districts pursuing socioeconomic integration today use public-school choice and incentives (like magnet schools) to create integrated schooling. Districts should poll parents to find out what sorts of schools they would like for their children. Cambridge, Massachusetts, for example, uses a system of universal choice: Every school is a magnet school, offering a distinct educational approach, and parents rank their preferences among the schools. School officials then honor preferences with an eye toward ensuring schools are economically integrated.

So far, these efforts are producing substantial results. In Wake County, low-income and minority students are outperforming similar students in other large North Carolina districts that have not addressed concentrations of poverty. And the middle-class pupils in Wake County are in no way harmed academically by the presence of some low-income students in school; indeed, administrators have reported that students benefit from exposure to racial and economic diversity. On the 2006 High School End of Course exams, 60.5 percent of low-income students in Wake County passed, compared with 43 percent in comparably sized Durham County; meanwhile, 85.4 percent of Wake’s middle-class students passed, compared with 67.9 percent in Durham.

Districts like Wake County emphasize economic integration rather than racial integration in part because using economic status avoids the legal problems associated with using race, and in part because of the overwhelming research showing that black kids don’t learn more sitting next to whites, but rather poor kids do better in middle-class environments. To be sure, pursuing economic integration is politically tricky for school districts, but the federal government could play an important supporting role. Under the No Child Left Behind Act, students trapped in failing schools are given the right to attend better-performing public schools, and that right could be extended to include transferring across school district lines. Former senator and presidential candidate John Edwards has proposed a creative way to encourage this movement: provide financial incentives for suburban middle-class schools to accept low-income students, and at the same time increase magnet-school funding to draw middle-class students into excellent urban magnet schools.

If the ultimate goal of a school system is to produce high-performing students, then economic integration is a clearly superior option. Rather than trying to make high-poverty schools work against all odds, we should eliminate high-poverty schools altogether and strike at the fountainhead of educational inequality.