Around 800,000 children experience exclusion from school every year in England, which can have significant and long-lasting effects, especially on educational achievement. Existing papers on the impact of school exclusion on children's outcomes mainly investigate the correlational relationship using US data. There is a lack of empirical evidence from other countries examining whether this relationship is causal. Longitudinal administrative data provide the ideal opportunity to test the hypothesis that children who are excluded go on to have lower educational achievement as a result.
This paper aims to measure the effect of exclusion in school-aged children. We use individual-level administrative data: the National Pupil Database (NPD) from school year 2004/05 to 2016/17. The NPD is maintained by the UK Department of Education and contains information on all pupils in state-funded schools in England. Detailed characteristics about the pupils and the schools are recorded. Student outcomes include the pupils'test scores, prior attainment, and progression at each Key Stage. In terms of school exclusion, we have detailed information on the time, school term when the exclusion occurred, and the type, reason and number of sessions for each exclusion.
To estimate the impact of school exclusion on the educational achievement of pupils, fixed effect models at individual level will be applied using the longitudinal dataset. We will control for the socio-economic characteristics of pupils, households and schools, for example age, gender, ethnic group, cohort, special education needs, location, siblings, school type, and other aggregate level characteristics. Having already conducted a comprehensive literature review, and with data expected to be available in September, we anticipate preliminary results this autumn.