Corruption, 'ghosts', and education: missing education funds in Mexico

Abstract

This work examines whether corruption in the form of missing funds for education grants in Mexico affects student outcomes. I use an objective measure of corruption from audits by the Superior Audit Office of Mexico governments in relation to educational grants that were transferred from the central government to state governments. I exploit the variation in state-level corruption between 2007 and 2014 using time, municipality and state trends fixed effects. The results show a negative relationship between corruption and various student outcomes at the municipal level in primary school and that corrupt states might also inflate the grant-assignment variable to receive more resources. Municipalities that belong to a state affected by corruption in its education funds score 0.41 standard deviations less in some grades and have lower approval rates. The grant is based on reported figures on student enrollment by each state, the evidence suggests that corrupt states inflate these figures. These municipalities report 2% more students in primary and 4% more in secondary. I find little evidence of heterogeneous effects of corruption depending on political alignment between mayors and governors and none on characteristics related to different costs for improving educational outcomes.