Heterogeneous Treatment Effects? An examination of Australian Non-Government Primary Schools

On average, students in Australian non-government schools consistently outperform their counterparts in government schools on standardized tests of literacy and numeracy. However, when differences across school sectors in student characteristics are taken into account there is no evidence that this performance differential is attributable to the nature of the schools. Nevertheless, non-government schools may have heterogeneous effects, that is, they may benefit particular groups of students. This study investigates the extent of non-government school advantage for specific primary school student groups. Test scores from the National Assessment Program Literacy and Numeracy for a nationally representative sample of students from the Longitudinal Study of Australian Children are analysed using inverse-probability-weighted regression adjustment. We find no evidence that attendance at non-government primary schools has a positive effect on academic outcomes of children in general, nor for children categorised by gender or socio-economic status. Although children with an Australian-born primary caregiver perform no better in non-government schools, there is some evidence that children with a non-Australian-born primary caregiver benefit academically from attending non-government schools. Our findings challenge common perceptions of non-government school efficiency and raise some important questions about current policies for funding Australian primary schools.