The Impact of IMF-Supported Programs on Government Health Expenditures: An Analysis of 29 Latin American Countries

Abstract

How do IMF-supported programs impact government health expenditures? Several studies have examined the effects of IMF-supported program participation on social expenditures, but there has been no consensus on the answer. This study aims to shed further light on the impact of the IMF on government health expenditures as a share of total government expenditures for 29 Latin American countries for the years 2000 to 2015 by utilizing rigorous econometric practices and sensitivity tests, whilst addressing the prominent sample selection bias problem. Methods applied include incorporation of instrumental variables, Inverse-Probability Weighting, the Nearest Neighbor Matching Method, and the Propensity Score Matching Method. Three findings are established: (1) it is important to use treatment effects to account for the sample selection bias problem – using one instrumental variable alone to address the problem is not sufficient; (2) there is a statistically significant and positive relationship between IMF participation and government health expenditures as a percent of total government expenditures when the average treatment effect in the population is examined; and (3) there is no statistically significant relationship between IMF participation and government health expenditures as a percent of total government expenditures when the average treatment effect on the treated is examined.