Objective: To assess the relationship between government expenditure on maternal health (GE) and maternal mortality (MM) in Mexican poor population between 2000 and 2015 in the 2457 Mexican municipalities.
Methods: Using administrative data, we performed the analysis in three stages: First, we tested the presence of selection bias in MM. Next, we assessed the presence of spatial dependence in the incidence and severity of MM. Finally, we estimated a spatial error model considering the correction of estimates for the spatial dependence and selection bias assessed before.
Results: MM and GE were not randomly distributed throughout the Mexican territory; the most socially vulnerable municipalities exhibited the highest levels of MM severity but the lowest levels of GE and available human and physical resources for maternal health; the incidence of MM was independent of GE; elasticity of GE-severity in MM was -4% (P < 0.01).
Conclusions: Resource allocation for maternal health must move towards a more comprehensive vision, and efforts to achieve an effective delivery of universal health services must improve, particularly regarding the most vulnerable municipalities.
Keywords: government expenditure; maternal health; primary health care; spatial econometrics.
© 2019 John Wiley & Sons, Ltd.