Educational Reforms’ Impact on Students’ Test Scores in Eleven Countries and their Ensuing Distributive Effects

Based on a multi-level (student, school, and country) analysis of the OECD Programme for International Student Assessment (PISA) 2012 on mathematics, science and reading for the 11 countries that also surveyed school principals and parents, we have tested the tenability of three presuppositions underlying the mix of educational reforms implemented in these countries and containing elements of the four models of public service delivery – choice, voice, target and performance measurement, professional trust – distinguished by Julien Le Grand (2007a/b). Corroborating the first presupposition we have established that parental voice and target performance measurement positively influence the test results of students, but contrary to this presupposition we find that parental choice has no impact on the test results while professional trust has a negative impact. Contrary to the second presupposition, which states that parental choice increases the positive influence of parental voice on test results, we demonstrate that, instead, parental choice undermines the positive influence of parental voice on test results. Contrary to the third presupposition, which states that parental choice produces less inequality than parental voice does, we show that parental voice reduces inequality, while the impact of parental choice does not vary in terms of parents’SES. Our findings suggest the presuppositions underlying much of the debate on educational reforms are basically untenable.