Moral hazard and health insurance: an evaluation based on Rubin’s causal framework[Aléa moral en santé : une évaluation dans le cadre du modèle causal de Rubin]

The purpose of this article is to estimate the sensitivity of healthcare consumption to the presence of health insurance. In France, where social security guarantees a homogeneous basic coverage, the answer to this question traditionally goes through the comparison of medical recourse for people who have complementary health insurance with medical recourse for people who do not. Due to possible selection effects, related to the fact that having a complementary health insurance is most often a personal choice, this comparison is difficult. This article moves the usual line of comparison and answers the question of the impact of insurance on healthcare use by comparing the medical consumptions of employees with supplementary health insurance from companies with mandatory membership, with those of employees who do not have one. Indeed, the organization of health protection in France has this dual feature: many rely on company supplementation and notably encourage collective contracts covering all employees of a company. This type of contract covers one in four policyholders. The fact that the subscription to supplementary health insurance is then attached to the employment contract leads us to believe that it is independent of disease risk if we restrict the analysis to employees only. Our estimate is based on this assumption. In a setting where the impact of insurance is heterogeneous within the population, we find that this hypothesis allows us to estimate what would be the average medical consumption of people without supplementary insurance, if they were provided with one. We find that the probability of going to the doctor at least once a year would increase by almost 20 points. This probability of around 75% when people do not have complementary, would thus rise to 95%.