Determinants of Child Mortality in Africa: A Methodological Discussion

Current literature is ambiguous regarding the significance of public health expenditure in reducing mortality rate among children in cross country studies. In fact, several previous studies found the relationship between these two variables to be insignificant. Such findings indicate the existence of a huge inefficiency in public sector of struggling countries and discourage supports provided by donating entities. This study addresses the disagreement in the literature by pointing out how results are sensitive to the use of non-stationary variables that are used often in the literature. Using a Panel Vector Autoregressive model, no empirical evidence for the role of public health expenditure in reducing child mortality is found when non-stationary variables are used. However, results are significantly different as soon as stationary variables are substituted in the same model. In fact, the elasticity of under five mortality rate with respect to per capita public health expenditure is about -0.22 for African countries in the sample.