Health Expenditures and Child Mortality: Evidence From Kenya

Abstract

In previous studies, health impacts of government and household expenditures have been estimated independently. As a result, the complementarities of these expenditures in improving health have remained unexplored. This thesis fills this gap by estimating own and joint effects of public and private health expenditures on child mortality using Kenyan household data supplemented with county level data. In particular, structural linear probability models of neonatal, infant, and under-five mortality are estimated accounting for potential biases due to endogeneity of expenditures and heterogeneity of child health. A notable finding from the empirical analysis is that the effects of public and private health expenditures on child deaths depend critically on age of the child. In particular, public and private health expenditures have no effect on deaths of neonates but significantly influence the mortality of infants and children below the age of five. In structural models of under-five mortality, effects of the interaction between the private and public health expenditures are statisticaIly significant, suggesting that the expenditures complement each other in reducing child mortality. However, after accounting for the interaction effect, the separate impacts of the expenditures on mortality are statistically insignificant. Thus, in controlling childhood diseases, there is need for recognition that whereas the government should invest adequately to provide public health services, households should similarly provide for treatment of nonimmunizabJe diseases. More generally, there is need to design and implement policies that promote synergy between public and private-health expenditures in the control of all diseases.