Louis Menand sketches out the cases for and against. He notes two elements in "the standard argument against homework don’t appear to stand up":
The first is that homework is busywork, with no effect on academic achievement. According to the leading authority in the field, Harris Cooper, of Duke University, homework correlates positively—although the effect is not large—with success in school. Professor Cooper says that this is more true in middle school and high school than in primary school, since younger children get distracted more easily. He also thinks that there is such a thing as homework overload—he recommends no more than ten minutes per grade a night. But his conclusion that homework matters is based on a synthesis of forty years’ worth of research.
The other unsubstantiated complaint about homework is that it is increasing. In 2003, Brian Gill (then at rand) and Steven Schlossman (Carnegie Mellon) showed that, except for a post-Sputnik spike in the early nineteen-sixties and a small increase for the youngest kids in the mid-nineteen-eighties, after the publication of "A Nation at Risk," by the Department of Education, which prescribed more homework, the amount of time American students spend on homework has not changed since the nineteen-forties.