THE MOST HEATED debate in American education just took a turn for the bitter. Yesterday, in a vitriolic split decision, a federal appeals court upheld a ruling that a Cleveland, Ohio, voucher program violated the separation of church and state by allowing state money to go toward the tuition at sectarian schools, schools that tend to be the most affordable of the nonpublic set. This was an issue of government and religion. But eventually, America's courts will have to deal with a state or municipality that pursues a school choice strategy that excludes parochial schools. And at that time (which may come soon in Florida where the legal debate is over the state's mandate to provide public education), the larger question of what exactly constitutes a public education will have to be addressed. However, even this broader debate is the wrong one to be having, for the dirty little secret of the educational world is that -- in the end -- schools matter little.

Almost thirty-five years ago, sociologist James Coleman and his coauthors issued what remains today the most controversial report on inequality in schooling. The document, which later came to be known as the Coleman Report, reached the troubling conclusion that the strongest predictor of academic performance was not school-based dynamics, but rather the student's family background, that is, her household income, parents' occupations, and so on. The most shocking and important lesson of this report is that the indicators upon which we focus our educational policy debate -- per pupil expenditures and student-to-teacher ratios -- did not appear to matter much.

While the report was initially assailed on methodological grounds, many researchers have gone back to reanalyze the original data (which comprised information on over six hundred thousand students in four thousand schools) and the overall pattern of findings has held steady. Scholars may not agree as to what the net effects of schools are on educational achievement. But across the ideological spectrum there is a tacit consensus that the effects of other dimensions of background -- be they community, family, genetics, whatever -- are stronger. If anything, school vouchers would make these factors more important and the dynamics inside schools less critical. This is because under a voucher plan even more of the "sorting" of students and families would take place before a kid even walks into his chosen school building. Savvier, better educated parents will get their kids well placed, and families with fewer financial and non-financial resources will be left behind. With whom one goes to school -- not where -- is the most important factor, since school is all about internalizing one's place in society's pecking order.

Why, then, after thirty-five years of evidence that schools are marginal in the grand scheme of academic achievement do educational policy debates continue to focus almost exclusively on schools and issues such as vouchers? The short answer is politics. Schools are a government responsibility, so they represent an area about which campaign promises can be made, a realm where we can have a valid public debate. However, if we acknowledge that schools matter little, what is society to do? If parenting styles -- or worse still, genetics -- are really much more important for kids than schools are, one can only imagine what kind of public policies that would lead to, beginning with invasive home visits and extending to the domain of eugenics.

Still, there are other non-school-based policy choices. The United States could pursue -- as Gore had promised during the campaign -- a goal of universal day care for the critical early years before kids arrive at kindergarten. Much research shows that rates of illness strongly affect academic performance, so we might improve the health care of school-aged children. We could also expand the housing stock, for recent research also shows that when kids live in crowded conditions at home -- and are unable to sleep well, have no privacy to do homework, and so on -- they tend to concentrate poorly in school and perform worse overall. These issues are not on the lips of educational policymakers and pundits. Many players -- not just elected officials, but also educational researchers, teachers unions, school voucher advocates, even custodians -- have an institutional stake in keeping the debate focused on the time period between 9 A.M. and 3 P.M. After the bell, however, the policy possibilities are endless.