Macroeconomic Policy Analysis When Planning Horizons are Finite
AbstractThe paper will argue for the desirability of a form of bounded rationality in which economic decision makers are not assumed to form complete state-contingent plans extending indefinitely into the future, but instead to only think forward a finite distance into the future. It will illustrate how this modification of the usual assumptions about the optimality of household and firm decisions changes the form of the structural equations of an otherwise standard New Keynesian model of the effects of monetary and fiscal policy, and will show that allowance for finite planning horizons provides a potential explanation for aspects of the data that have been embarrassing for NK models based on full rationality.
It will further show that while this assumption will still allow for many conclusions about the consequences of forward-looking behavior that are qualitatively (even if not quantitatively) the same as in standard analyses (in particular, precommitments of future policy can be valuable), it allows one to avoid paradoxical conclusions that are sometimes drawn from rational-expectations analysis in infinite-horizon models. It will discuss in particular the consequences of a more realistic assumption of limited forward planning for (i) the arguments of Cochrane and others for "neo-Fisherian" effects of commitments about future interest-rate policy, and (ii) debates about "the fiscal theory of the price level."