Small classes in the early grades: One policy, multiple outcomes

The idea that small classes are desirable for instruction is not new; indeed, it dates back at least to biblical times (Angrist & Lavy, 1996). In the 20th and 21st centuries, small classes, a distinguishing feature of American private schools and a fundamental component of all remedial programs, have been posited as a way to bolster student achievement in public schools generally. Tennessee's Project STAR (Student Teacher Achievement Ratio) is not a program in the traditional sense. It was a large-scale experiment to test the effectiveness of this basic principle of school organization-that is, arranging classes in the early grades so they have a relatively small number of students in a classroom with a full-time teacher. Because of the reported effectiveness of small classes in STAR, the experiment has been followed by class size reduction (CSR) initiatives across the country. According to one recent survey, 33 states had statewide CSR programs for the early grades, of which 13 were mandatory (McCabe, 2006). The number of districts with their own CSR programs is undocumented but probably quite large. President Clinton introduced a nationwide CSR program in 1998, which was discontinued by the Bush administration. STAR has also been followed by additional research on small classes, by several analyses of CSR costs and benefits, and by some arguments that the research on CSR or the conclusions drawn from the research is faulty. This chapter summarizes these developments.