A contribution to the study of health inequalities in France through self-assessed health indicators.

Summary This thesis is in the field of measuring and explaining health in the context of analyzing health inequalities. A first chapter considers the health indicators commonly used in empirical work and returns to the debate on the use of self-assessed health. It highlights the relevance of methodological refinements in health measurement proposed in the international literature that have not yet been applied to France. A second chapter proposes an original methodology for measuring health. The construction is based on an individual health status data considered less subjective, namely the number of diseases and their degree of severity, and considers variables classically collected in health surveys. A third chapter describes the tools of stochastic dominance and the indices commonly used in the analysis of inequalities in a health framework. The fourth chapter analyses social inequalities in health in France in 2004, then over the period 1998-2004. It highlights social inequalities in health in favour of the highest social groups. However, these inequalities decreased between 1998 and 2004, due to a lower elasticity of health with income and a decrease in the unequal distribution of income within social groups. Moreover, the analysis conducted on different measures of health reveals an inffluence on the magnitude of inequalities of the number of categories of the discrete health variable and of the health distribution chosen to cardinalize it. The fifth chapter focuses on the influence of the social background of origin and the relative longevity of the parents with respect to their birth cohort on the health status in adulthood, using three approaches. The first approach highlights the fact that the distributions of health status of persons born to a father or mother belonging to the upper social categories significantly dominate those of persons with parents from lower social categories. The parametric approach confirms an effect of the occupation of each parent on health status in adulthood. It also shows that health status depends significantly on the longevity of each parent. Finally, the concentration index approach highlights an inequality of health opportunities in favour of individuals whose parents have experienced a high longevity and then an inequality of health in favour of individuals from more privileged backgrounds. The chapter concludes that there are inequalities of opportunity in health in France.