School outcomes in New South Wales and Queensland: a regression discontinuity approach

This paper examines the differences in school (NAPLAN) outcomes between New South Wales and Queensland. It shows that there are pronounced differences in Year 3 NAPLAN results between these states, though these dissipate when later class years are considered. The reasons for these state effects in school outcomes are explored using an empirical framework grounded in the regression discontinuity literature. There is no systematic evidence of school-specific unobservables that are linearly related to the distance of schools from the New South Wales-Queensland border that account for this state effect. In other words, the lower outcomes of schools in Queensland relative to schools in New South Wales appear to be a pure institutional phenomenon. This conclusion carries over to the various sensitivity tests undertaken based on the partition of the sample by school sector and by distance from the border. Accordingly, the border effects are expected to be eroded by the recent reforms to the school starting age in Queensland and by the national curriculum currently proposed for implementation in Australia.