Abstract
Knowledge of the effect of school resources on student outcomes is important for policy decisions concerning expenditure on schools. However, empirical research has so far produced equivocal findings. This paper examines the methodological and data requirements for good quality estimation of the education production function and reviews four UK studies that use pupil-level longitudinal data with a range of resource and control variables. These have produced some evidence of small resource effects on student outcomes and indicate the importance of model specification in affecting reported findings. If research in this area is to progress, high quality datasets are essential.