Policy and Contextual Determinants of Mathematics Achievement Across Mexican States: Evidence from PISA, 2003-2012

I studied the contribution of individual, school, and state-level factors to mathematics achievement in Mexico using subnational PISA data from 2003–2012 and five sequential OLS models, drawing on human capital and collective social capital theories. Results indicate that school-average socioeconomic status is the strongest predictor of achievement, substantially exceeding the effect of individual socioeconomic status. Mathematics instructional time and school resource quality are positively associated with performance, while teacher-union conflict is negatively associated with achievement. Public spending per student does not predict higher performance. From a policy perspective, these findings suggest that educational authorities should give greater attention to the socioeconomic composition of schools and systematically examine the institutional practices of consistently high-performing states to inform educational improvement elsewhere.