Michael Grossman's human capital model of the demand for health has been argued to be one of the major achievements in theoretical health economics. Attempts to test this model empirically have been sparse, however, and with mixed results. These attempts so far relied on using-mostly cross-sectional-micro data from household surveys. For the first time in the literature, we bring in macroeconomic panel data for 29 OECD countries over the period 1970-2010 to test the model. To check the robustness of the results for the determinants of medical spending identified by the model, we include additional covariates in an extreme bounds analysis (EBA) framework. The preferred model specifications (including the robust covariates) do not lend much empirical support to the Grossman model. This is in line with the mixed results of earlier studies.
Keywords: Extreme bounds analysis; Grossman model; Medical spending; OECD panel.