Better settings for better education: does decentralization work?

This dissertation examines the effect of decentralization on coverage and efficiency of basic education in Mexico. It evaluates whether the transfer of educational responsibilities, resources, and decision-making power to subnational governments is associated with better educational outcomes. The effects of the reform are estimated at the national and the state level using multivariate statistical techniques, and help to unveil the reasons explaining the observed results and outcomes. Three dimensions of explanation are discussed: the institutional setting present in each state before the reform; the role of national compensatory programs after decentralization; and the level of social capital existing in the different regions of the country. The study is based on the educational production function framework. For the analysis, the educational system is understood as a black box where inputs are transformed into outputs by unknown processes, given certain technological setups, organizational structures and institutional settings. At a particular point in time a treatment variable called decentralization is included in the “black box” modifying the processes that generate the educational outcomes such that it is possible to tease out if these outcomes are affected by the treatment. The dissertation also presents a case of study developed using qualitative methods of research such as semi-structured interviews with parents, teachers and public officials, that explores if decentralization effects at the state level are influenced by the state governments capacity to innovate and improve organizational processes after the educational reform; by the level of social participation in education; and by the role of the union of teachers in obstructing or enhancing subsequent educational changes. Results suggest the effect of decentralization is not homogeneous among states and that it varies according to the educational level that is being looked at. The differences are explained by the presence or absence of a state-based educational system before the reform and by the intervention of the central government through compensatory actions. In addition, findings suggest that a favorable institutional and political setting at the state level, coupled with the strengthening of local capacities and a process of educational innovation, will enhance the success of decentralization.