Static and Dynamic Incentives in Selection Markets

We study the optimal combination of static and dynamic incentives in community-rated markets. We use a simple two-period model to derive general lessons. Overall, under worsening health, prices are generally increasing over time, pointing to the desirability of dynamic (late-enrollment) penalties as used in markets like Medicare Part D. However, the desirability of dynamic incentives is reduced when consumers fail to internalize them due to a lack of understanding or forward-looking behavior. To provide policy lessons, we calibrate a multiple-period version of our model to the context of Medicare Part D. Under our preferred assumptions, the optimal late enrollment penalty in Medicare Part D is positive but lower than the level currently set by the regulation.