Why Education Markets Fail: Assessing the Impact of Market Reforms in Chile

• In Event: Market Mechanisms and Education Inequities: Experiences From Chile, India, South Africa, Canada, and the United States

Abstract

Pupose. This paper examines the effect of education markets reforms in Chile, using the accumulated existing evidence for the last three decades. It analyses the main outcomes of the system in relation to education quality, equality and social integration. It also specifically examines the role of competition among schools as a potential driver of education quality improvement. It shows that the voucher system in Chile has not delivered the promised learning outcomes. Moreover, it has promoted social segregation and prolonged educational inequality among different socio-economic groups.

Theoretical Perspectives. The main hypothesis behind the introduction of vouchers is that it would allow choice for parents and promote competition between schools (e.g. Manski, 1992; Hoxby, 1994; Filmer and Pritchett, 1999). In contrast, some authors highlight the negative effect of competition in education (e.g. Smith and Meier, 1995; McMillan, 1999; Maranto, Milliman and Stevens, 2000; Bellei, 2015). In order to conceptualize the effect of competition between schools on outcomes, this study uses an approach based on contestability of markets (Baumol, 1982) and school turnover indicators.

Method and Data Sources. Using administrative data comprising the universe of students and schools in Chile for the years, it empirically estimates the effect of competition over academic outcomes, using a school turnover indicator as a proxy for competition within districts. The paper then offers an explanation of these findings, analyzing each one of the main underlying assumptions of the voucher model in education, demonstrating through empirical evidence why they do not work in reality. The models are estimated using administrative data provided by the national education quality assessment (SIMCE) taken by all schools in Chile. Both Language and Mathematics 4th grade tests scores are used at the student level for the period 2002-2008. In addition, the Directory of Schools 2002-2008 and enrolment datasets for years 2004-2008 are used.

Results. This study econometrically estimates the effect of competition over academic outcomes, showing that this effect is significantly negative, although virtually negligible in terms of its size (under 0.07 SD). In light of these results, several plausible explanations are explored based on the empirical analysis of each one of the main underlying assumptions of the voucher model in education.

Scholarly Significance. The Chilean case study is important since it constitutes one of the oldest and most extensive voucher systems in the world and thus represents a valuable experience from which numerous general lessons may be drawn. From an empirical perspective, this study provides new evidence, based on extensive administrative data, which builds into previous research. It also addresses several methodological issues not properly taken into account by previous studies in the area, including correcting for endogeneity, using instrumental variables, and selection bias. The paper is relevant for both academics and policy makers as it provides robust evidence on the detrimental effects of competition and vouchers on the quality, inclusion and equity of education systems.

Author

• Javier Gonzalez, University of Cambridge