This paper uses a common household survey instrument and a common set of imputation assumptions to estimate the pro-poorness of government health expenditure across 69 countries at all levels of income. On average, government health expenditure emerges as significantly pro-rich, but there is heterogeneity across countries: in the majority, government health expenditure is neither pro-rich nor pro-poor, while in a small minority it is pro-rich, and in an even smaller minority it is pro-poor. Government health expenditure on contracted private facilities emerges as significantly pro-rich for all types of care, and in almost all Asian countries government health expenditure overall is significantly pro-rich. The pro-poorness of government health expenditure at the country level is significantly and positively correlated with gross domestic product per capita and government health expenditure per capita, significantly and negatively correlated with the share of government facility revenues coming from user fees, and significantly and positively correlated with six measures of the quality of a country's governance
it is not, however, correlated with the size of the private sector nor with the degree to which the private sector delivers care disproportionately to the better-off. Because poorly-governed countries are underrepresented in the sample, government health expenditure is likely to be even more pro-rich in the world as a whole than it is in the countries in this study.