Assessing the impact of health care expenditures on mortality using cross-country data

A significant body of literature has examined the impact of public health expenditure on mortality, using a global cross-section or panel of country-level data. However, while a number of studies do confirm such a relationship, the magnitude of the impact varies considerably between studies, and several studies show statistically insignificant effects. In this paper we re-examine the literature that identifies this effect using cross-country data. Our analysis builds on the two instrumental variables (IV) approaches embodied by key publications in the field – Bokhari et al. (2007) and Moreno-Serra and Smith (2015). Using exactly the same data and econometric specifications as the published studies, we start by successfully replicating their findings. However, further analyses using updated data and ‘streamlined’ econometric specifications, plus statistical data imputation and extensive robustness checks, reveal highly sensitive results. In particular, the relevance of the IVs is seriously compromised in the updated data, leading to imprecise estimations of the relationship. While our results should not be taken to imply that there is no true mortality-reducing impact of public health care expenditures on mortality, the findings do call for further methodological work, for instance in terms of identifying more suitable IVs or by applying other estimation strategies, in an effort to derive more robust estimates of the marginal productivity of public health care funding.