Politics, Markets and Schools: Quasi-Experimental Evidence on the Impact of Autonomy and Competition from a Truly Revolutionary UK Reform

Abstract

Supporters of market-based education reforms argue that school autonomy and between-school competition can raise student achievement. Yet U.S. reforms based in part on these ideas - charter schools, school-based management, vouchers and school choice - are limited in scope, complicating evaluations of their impact. In contrast, a series of remarkable reforms enacted by the Thatcher Government in Britain in the 1980s provide an ideal testing ground for examining the effects of school autonomy and between-school competition. In this paper I study one reform - described by Chubb and Moe (1992) as ‘truly revolutionary’ - that allowed public high schools to ‘opt out’ of the local school authority and become quasi-independent, funded directly by central Government. In order to opt out schools had to first win a majority vote of current parents, and I assess the impact of school autonomy via a regression discontinuity design, comparing student achievement levels at schools where the vote barely won to those where it barely lost. To assess the effects of competition I use this same idea to compare student achievement levels at neighbouring schools of barely winners to neighbouring schools of barely losers. My results suggest two conclusions. First, there were large gains to schools that won the vote and opted out, on the order of a onequarter standard deviation improvement on standardised national examinations. Since results improved for those students already enrolled in the school at the time of the vote, this outcome is not likely to be driven by changes in student-body composition (cream-skimming). Second, the gains enjoyed by the opted-out schools appear not to have spilled over to their neighbours - I can never reject the hypothesis of no spillovers and can always reject effects bigger than one half of the ‘own-school’ impact. I interpret my results as supportive of education reforms that seek to hand power to schools, with the caveat that I do not know precisely what opted-out schools did to improve. With regards to competition, although I cannot rule out small but economically important competition effects, my results suggest caution as to the likely benefits.