Efficiency of public spending among Mexican universities and its determinants: Data Envelopment Analysis (DEA) with the application of semi-parametric and two stage models.

Abstract

Based on the methodological approach of Simar and Wilson (2007), Data Envelopment Analysis and truncated regression model with bootstrapping are used to evaluate the performance of 34 state universities of Mexico in their public spending and their determinants during 2010-2015. The results show evidence of disparities between total expenditure and per researcher, and a more homogeneous behavior in expenditure per student. Depending on the criteria of efficiency used and year analyzed, institutional and socioeconomic aspects presented differentiated effects. In this sense, it is highlighted that in those entities where there is a greater population in poverty and greater immigration, efficiency was increased, while the geographical location of the university did not show significant differences in the efficient exercise of spending.