Independent-school competition and Sweden's performance in TIMSS

This paper analyses the effects of independent-school competition on Sweden's performance in TIMSS, an international low-stakes test in mathematics and science among students in year 8. Exploiting variation in independent-school enrolment shares across counties over time, it finds that increasing competition has improved TIMSS scores, an impact that appears only after 2003 and is driven by for-profit schools. The results suggest that competition both slowed down Sweden's performance decline between 1995 and 2011 as well as contributed to its improving scores between 2011 and 2019. A simulation based on the estimates indicates that Sweden's average score in TIMSS 2019 would have been 20 points, or 0.24 standard deviations, lower without the expansion of the independent-school sector.