Joint Discussion Paper Series in Economics 48-2012
Various panel models were presented to resolve the ranking of global health care systems according to efficiency. However, in terms of the spatial distribution of statistical units, spatial dependence as a result of various forms of spatial interactions caused biased estimators in classical regression. To our knowledge, this is the first paper which analyzes the healthcare systems of WHO members with regard to spatial dependencies while distinguishing between heterogeneity and inefficiency. It was possible to determine a significant spatial autocorrelation. Therefore one have to consider these spatial spillovers when assessing the performance of healthcare systems. The most meaningful way of implementing these effects appears to be by regressing the health output on various explanatory variables through a combination of the fixed effects spatial lag and the fixed effects cross regressive model. This allows spatial spillovers due to level of education, healthcare expenditure, and the quality of the healthcare system itself, to be diagnosed. Modeling these spatial effects allows previous results to be given more precision with regard to the quality of the healthcare systems of WHO members.
panel data fixed effects production of health efficiency measurement heterogeneity spatial effects spatial autocorrelation