Current educational policy in the UK and elsewhere has emphasised initiatives to reduce class sizes

Current educational policy in the UK and elsewhere has emphasised initiatives to reduce class sizes. To a large extent these policies derive from recent research and research reviews that have demonstrated moderate learning gains from class size reductions. Any new research, therefore, that appears to contradict this deserves careful attention and evaluation. One such recent study is that of Hoxby (2000) who presents an analysis of data on class size which concludes that 'reductions in class size have no effect on student achievement'. Such a categorical statement clearly has important implications, if true, and apart from anything else appears to contradict most existing research findings. In this review I evaluate the research that underpins these claims and similar ones by Rivkin, Hanushek and Kain. (2000), and conclude that various methodological weaknesses render such conclusions unsound.