When Can Partial Public Insurance Produce Pareto Improvements?

Wilson (1977) provided the striking result that the government can always Pareto dominate a pooling equilibrium in a private insurance market with adverse selection by providing the pooling policy as a compulsory public policy and allowing individuals to buy supplementary private insurance. I show that this Pareto improving role for the government does not derive from its unique capacity to compel participation in a public insurance program. Rather, it stems from the fact that, with the introduction of the public policy, individuals may now hold multiple insurance policies: one public and one private. If, instead, we relax the assumption of the Wilson model that individuals may only hold one private insurance policy, the private market equilibrium is always second best Pareto efficient and there is no possibility of Pareto improvement through government intervention. Whether in fact individuals are restricted to purchasing only one private insurance policy - and hence whether there is scope for Pareto improvement through government policy in this model - varies in a predictable manner across different insurance markets.