Optimal Managed Competition Subsidies

When markets fail to provide socially optimal outcomes, governments often intervene through ‘managed competition' where firms compete for per-consumer subsidies. Subsidies are generally set across geographies according to estimates of the cost of government provision, a method which may not be welfare-maximizing. We introduce a framework for determining the optimal subsidy schedule that features heterogeneity in consumer preferences and inertia, and firms with heterogeneous costs that can set prices and product characteristics in response to changes in the subsidy. We apply it to the Medicare Advantage program, which offers Medicare recipients private insurance that replaces Traditional Medicare. We calculate counterfactual equilibria as a function of the subsidies by estimating policy functions for product characteristics from the data and solving for Nash equilibria in prices. The optimal schedule increases consumer surplus by 30% over the current policy and is well-approximated with a linear rule using market-level observables