School Autonomy, Education Quality and Development: an Instrumental Variable Approach

In this paper, I manly seek to test the robustness of Hanushek et al.'s (2013) hypothesis, according to which the impact of autonomy in terms of learning outcomes differs across levels of development, being positive for developed countries but not for developing countries. I do so by constructing a school-level measure of autonomy, which I instrument using the distinction between de jure and de facto autonomy, as laid out by Gunnarsson et al. (2009). I also follow them in differentiating between parents participation and school autonomy, thus providing an explanation to their results, in line with the conceptual framework of Hanushek et al. (2013).