Gastos com saúde: uma análise por domicílios para a cidade de São Paulo

Abstract

This paper addresses the question: what happens to health expenditres when a developing country population gets older? One implication of the analysis is that the older the population the higher the demand for the public health system. The analysis is based on a household consumption survey for the city of S„o Paulo, the largest Brazilian city. The estimation procedures consisted of parametric methods, such as Ordinary Least Square and Tobit Censored Regression, and the Least Absolute Deviation, a semiparametric method. Contrary to Ordinary Least Square and Tobit Censored Regression the Least Absolute Deviation does not require the hypothesis of a positive impact of age on health expenditure is sustained by the data with an estimator presenting large sample properties.