Institutional arrangements in educational systems and student achievement: a cross-national analysis

Abstract

Student achievement, School quality, Institutions, Quantile treatment effects, C14, C33, I21, I28,

The Impact of Institutional Arrangements on Educational Efficiency

“…This paper does confirm the results of some of the previous research on institutional arrangements in education. Consistent with Collier and Millimet [9] and Woessman [15], we find that school autonomy in purchasing supplies has a positive relationship with student achievement and efficiency, and that teachers' unions having a large influence on curriculum is associated with lower levels of student achievement and school efficiency [9,15].…”

“…Studies analyzing inefficiency in public schools have been able to locate the missing relationship between school resources and student performance [7,8]. Additionally, Collier and Millimet [9] find some evidence to support the notion that increased competition leads to increased efficiency in public schools, but only in districts that operate in financially flexible environments. Duncombe et al [10] use Data Envelopment Analysis (DEA) and report lower levels of cost efficiency in *Address correspondence to this author at the Department of Economics and Finance, University of Dayton, 300 College Park, Dayton, OH, 45469, USA; Tel: 937-229-5302; Fax: 937-229-2477; E-mail: collier@udayton.edu districts facing greater competition as measured by private school enrollment.…”

“…There are plenty of exceptions to this rule [11,12]; however, even less research has included data collected from multiple countries. A number of recent papers have used the Trends in International Mathematics and Science Study (TIMSS), an international project that collects data every four years on: student achievement scores, family background, class, school, and country characteristics [3,9,[13][14][15].…”

“…Similarly, Collier and Millimet [9] take a distributional approach, testing the institutional arrangements using quantile treatment effects, and find that many of the conditional mean results in Woessman [15] do not hold for the entire distribution of student test scores.…”

“…The drawback to Collier and Millimet [9] and Woessman [15] is that they allow only for a direct relationship between the institutional characteristics and student test scores. From a theoretical standpoint, it is more likely that the institutional arrangements have an indirect relationship with student achievement.…”