Inequality in health coverage, empirical analysis with microdata for Argentina 2006

Abstract

The literature on Health Economics does not usually study health coverage1 in itself but only in an indirect way. Specifically, the approach is often focused in health coverage as an explanatory variable of other variables which are related to the current stock of health. The intertemporal influence of health coverage over the past and future health stock is clear and evidently important beyond its influence over the present health status and its effects over the probability of sickness and recovery. Therefore, it should be clear that the interest of economic agents is to optimally preserve their health status and improve it throughout their entire life. This is an important microeconomic foundation that sustains the interest to study the health coverage as one of the important explanatory factors of the health stock. This work intends to perform a descriptive study on health coverage and explore its relationship with other variables that condition and modify the probability that economic agents receive coverage. Given the dichotomic nature of health coverage, the realization of a descriptive and conditional analysis presents a slightly different agenda than the usual one. This work is organized in the following way: Section I, utilizing microdata from the encuesta permanente de hogares (EPH) for the first semester of 2006 in Argentina, summary statistics are presented and a general description of the health coverage in Argentina for the people who belong to families without salary workers in it. In section II a binary regression model is estimated, and the probability of having coverage is studied. Subsequently, a Concentration Index is calculated on an individual basis. Next, following an adapted methodology than Wagstaff-Doorslaer-Watanabe (2002), a decomposition of the explained part of index which generates the probability of coverage is performed. In Section III, a decomposition of the change in the Concentration Indexes between years 2004 and 2006 takes place, utilizing microdata from the EPH corresponding to the first semester of 2004.