While most education reform is focused on testing, charter schools, and merit pay for teachers, the new issue of the New York Times Magazine highlights the importance of a different strategy: giving low-income students a chance to attend middle-class schools. Adam Davidson’s terrific piece—“Who Knew Greenwich, Conn., Was a Model of Equality?”—reports that the low-income children (of gardeners, handymen and housekeepers) who attend school in wealthy Greenwich perform substantially better at the high school level than the low-income students in a nearby Connecticut town with higher rates of poverty.

Davidson’s reporting on Connecticut is consistent with national data. On the fourth grade National Assessment of Educational Progress in Math, low-income students in low-poverty schools are about two years ahead of low-income students stuck in high-poverty schools (see figure 2).

Skeptics point out (correctly) that the low-income children whose parents make special efforts to gain access to strong schools like those in Greenwich may be particularly motivated—a criticism that also applies to parents who apply to charter schools like KIPP. Maybe it’s something about what these parents are providing at home, not what the school is doing, that helps explain the positive results, critics suggest.

As Davidson notes, however, research that seeks to control for “self-selection” bias also finds positive results for students who attend middle-class schools. In 2010, Heather Schwartz, published a report with The Century Foundation (Housing Policy Is School Policy), finding that among low-income families randomly assigned to public housing units in Montgomery County, Maryland, those students in public housing in low-poverty neighborhoods and schools performed far better than those in higher-poverty neighborhoods and schools.

Some might think it’s unsurprising that students perform better in wealthier districts like Greenwich because such districts tend to spend more money per pupil than higher-poverty school districts do. More money, not integration, could be the answer, some might suggest. But in fact, in the Montgomery County study, students in higher poverty schools received about $2,000 more per pupil than those in the wealthier schools, and yet the advantages of middle-class schools proved more important than per-pupil expenditure. Nationally, those advantages include being around classmates who are highly engaged academically, parents who volunteer in class at high rates and excellent teachers, with high expectations. In tight fiscal times, the finding that integration matters even more than money should be something that liberals and conservatives could rally around.

Davidson cites Century Foundation research that 80 school districts pursue socioeconomic integration plans in order to give more low-income students a chance to go to middle-class schools. Moreover, as I argue in my book, All Together Now, while there is strong evidence that low-income students benefit, there is no good evidence that the achievement of middle-class students declines in majority-middle class schools that are socioeconomically integrated. And middle-class peers benefit from economic diversity in important ways that don’t always show up in test scores.

If it is possible to raise the achievement of low-income students, while enhancing the overall educational experience of middle-class pupils, shouldn’t socioeconomic integration be part of our national conversation of school reform?