Econometrica: Mar, 1988, Volume 56, Issue 2
Optimal Time-Consistent Fiscal Policy with Finite Lifetimes
Guillermo A. Calvo, Maurice Obstfeld
This paper analyzes aspects of optimal fiscal policy for economies with capital accumulation and finitely-lived, heterogeneous agents. For a particular utilitarian social welfare function, the problem faced by a central planner can be broken down into two subproblems, a standard problem of optimally allocating aggregate consumption over time and a problem of distributing aggregate consumption optimally at each moment among those alive. If it can use a sufficiently rich set of lump-sum taxes and transfers, the government can replicate the command optimum as a market equilibrium outcome. No issue of government debt is needed to achieve this decentralization. The discussion emphasizes time-inconsistency problems arising from two sources. First, some social welfare functions give rise to time-inconsistent command optimums, for reasons first discussed by Strotz (1956). Second, in a market setting, optimal fiscal policy can be time-inconsistent when the government's fiscal tools are too limited to allow it to decentralize a time-consistent command optimum. This second possibility is illustrated by an sample in which the government is constrained to levy the same lump-sum tax or transfer on everyone alive on a given date. Unless fiscal policy has no influence over factor prices, optimal fiscal policy can be time-inconsistent if the economy's horizon is finite.