In the area of health insurance, social welfare is commonly evaluated from an ex ante approach, which computes the expected level of well-being of each individual before a risk occurs. This approach offers the advantage of considering risk preferences, which are a fundamental component of the intrinsic value of insurance. However, some welfare economists advocate for an ex post approach that focuses on the distribution of welfare once all uncertainty is resolved. Rather than contributing to the broad philosophic and ethical debates on the relevance of the ex ante and ex post approaches in the health insurance context, we compare and discuss the empirical differences of using either the ex ante or the ex post approach and of introducing aversion to inequalities to evaluate the effects of health insurance policies on social welfare. Our work is based on the simulated effects of four health insurance reforms that lead to more and more variations in financial risk and that increasingly affect worst-off individuals. We consider a unidimensional function of social welfare that relies on disposable income and use an Atkinson index for measuring social welfare. While accounting for risk aversion in the ex ante approach amounts to considering an aversion to rare but catastrophic situations that is comparable to the introduction of inequality aversion in the ex post approach, the results show some differences in the case of high heterogeneity of risk preferences in the population and when the loading cost of insurance reforms is high. Moreover, the ex ante approach may take more account of catastrophic situations than the ex post approach when individuals' risk aversion is high while the regulator's inequality aversion is low and when the most risk averse are also the worst off.
- Poster