Protection sociale et redistribution.

Summary In this thesis, we analyze, from a theoretical point of view, different interventions of the government in the choice of social insurance mechanisms, in the framework of a global redistribution policy. In the first chapter, the government intervenes on the demand side of the agents. In response to the presence of moral hazard, it introduces incentive insurance coverage. Characterized by two parameters, risk and income, individuals can in fact reduce the frequency of risks. Whatever the correlation between the inequality factors, including when the poorest are the most at risk, the rules determining optimal social insurance are modified and the coverage may be partial but retains, despite the presence of an optimal tax, a redistributive role. The second chapter examines, in the field of health, government intervention on the supply side. When there are demand induction phenomena, the physician remuneration mechanism that guarantees the social optimum is a capitation payment system. When quality is endogenous, this same government cannot induce optimality and the remuneration system set up favours one or other of the aspects, quality or cost reduction. In the third chapter, we ask to what extent a government has an interest in favoring opting out when the public provision of insurance is characterized by exogenous inefficiency. When the government adopts the rawls criterion, we show, public insurance never goes away. However, it is optimal to favor the exit of some individuals. The choice of regime that determines the types of individuals who remain beneficiaries of public insurance depends strongly on the structure of inequalities in the economy and the individual taste for quality.