What drives Health Care Expenditure in France since 1950? A time-series study with structural breaks and nonlinearity approaches

Using the French annual database (1950-2009), we conducted a time-series analysis to explain the role of GDP per capita on HCE (Health Care Expenditure) per capita taking into account structural breaks and non-linearity in the long-term economic relationship between HCE and GDP, controlling for price effect, population ageing, innovation proxy and medical density. We show that the non-linearity of the long-run relationship between HCE and GDP comes from both the presence of a structural break and non-linearity explained by a transition variable (by constructing a smooth transition cointegrating regression). More precisely, lower GDP elasticity is explained by an exogenous shock linked to health system policies in the mid 1980’s (break analysis) and endogenously driven changes in the health care system via medical density in France.