Classical regression estimates of the determinants of the OECD health expenditures are useful for policy formulation and evaluation. However, if the underlying timeseries data are not collectively stationary in levels, the estimated parameters are faulty and can misguide health policy. Until...