A welfare criterion with endogenous welfare weights for belief disagreement models

Belief disagreement generates a fundamental tension between two desirable features of a resource allocation: Pareto optimality and risk sharing. While Pareto optimality generally opposes restrictions to trade, a growing literature rejects it in the presence of heterogeneous beliefs and proposes welfare criteria that instead assume risk sharing as fundamentally desirable. We propose a welfare criterion that balances out these two desirable features by endogenously determining admissible welfare weights based on competitive equilibrium allocation as a benchmark. Applying our method to several belief disagreement models, we show how the welfare-optimal degree of risk sharing is between those suggested by Pareto optimality (which implies less) and by other existing approaches (which imply more).