in practice governments almost always resort to active policies and rarely rely
on the approach of benign neglect. This paper, while not being equipped to
resolve fully the ongoing dispute, delivers several predictions and delivers clear
policy recommendations.
First of all note that there are imperfections introduced into the paper by
assumption. The assumption of imperfect competition in itself calls for an ac-
tive intervention. Similarly, the fact that an OLG model is used as a benchmark
allows for the possibility of dynamic inefficiency and makes potentially govern-
ment policy viable. The paper, however, takes these two imperfections as given
and does not attempt to provide prescriptions that could counteract them. The
paper shows that fluctuations in macroeconomic variables are costly and fo-
cuses only on measures that could limit the impact of equilibrium fluctuations
on welfare.
Observe that whenever the error of perception is positive economic agents are
overoptimistic. This overoptimism affects the amount saved in equilibrium and
leads to intertemporal welfare redistribution. First of all overoptimism decreases
saving and increases current output and the welfare of current old. Moreover, it
is ex post costly to the currently young since they eventually find out that they
were overly optimistic and that they under saved. Furthermore, it is costly to
the members of the future generations since they have too little capital to work
with. Similarly, overpessimism leads to intertemporal welfare shifts harming
currently alive and unnecessarily rewarding the future generations. Therefore,
inability to foresee the future perfectly leads to suboptimal outcomes and makes
government intervention viable.
Fluctuations in this paper are an outcome of aggregate demand disturbances.
The disturbances result from the fact economic agents revise, in an individu-
ally rational manner, their expectations. Specifically, fluctuations are due to
the fact that economic agents willing to economize on information gathering
and processing costs choose not to foresee the future correctly. Naturally, a
very simple policy can restore the first best outcome. The government should
gather and process information and should endow the public with the best pos-
sible estimates of the future economic variables. This simple strategy can fully
eliminate oscillations in the framework of this paper. In practice, however, this
strategy need not be a realistic option as the viability of the option stems from
the fact that the paper uses a representative agent framework. In general, in
an environment with heterogenous agents the government need not have the
capacity to estimate future income streams of all agents and the basic strategy
must be replaced with second best instruments.
Tax cuts and government spending remain the two major policy instruments.
Moreover, during recessions attempts to restore consumer confidence are a key
component of stabilization programs. Implicit for these policies is the assump-
tion that for some reason consumers choose not to spend the amount they ought
to. However, the assumption is inconsistent with the rational expectations par-
adigm. The framework of this paper allows to justify that these policies are
indeed welfare improving and consistent with consumer behavior. Fluctuations
arise because expectations held by economic agents are based on information
27