Free to improve? The impact of free school attendance in England

We evaluate the impact of attending two secondary free schools in England - new autonomous state-funded start-ups - using admission lotteries and a distance-based regression discontinuity design. We characterise each school's ethos through text analysis of vision statements: one follows a no excuses paradigm common among US charter schools; the other adopts a classical liberal, knowledge-rich approach. These features distinguish them from each other and from counterfactual schools attended by rejected applicants. Despite pedagogical differences, both schools significantly improve test scores, reduce absences, and lower student mobility. Our findings support policies promoting horizontal differentiation in publicly funded education.