Optimal long-run fiscal policy: Constraints, preferences and the resolution of uncertainty

We construct a computational dynamic stochastic overlapping generations general equilibrium model with uncertain lifetimes and explore the impact of policy stickiness (specifically, a major reform will preclude future reforms for a generation) on optimal long-run fiscal policy. Under such circumstances, entitlement reforms exhaust a valuable option to move in the future. We explore the conditions under which the gain to waiting is large enough to induce optimizing policymakers to delay reforming a suboptimal system. We also allow for the uncertainty to have ARCH characteristics and explore the impact of time-varying uncertainty on the optimality of delayed policy action.