occasional dramatic buildups of consumer credit are thought to signal some
form of irrational behavior on the part of consumers. All these possibilities
are considered to lead to output growth that is not justified by corresponding
growth of the productive capacity of the economy. The discrepancy, if it exists,
is not sustainable and normally leads to active policy measures. Clearly, pol-
icy design invokes the concepts of overheating and output gap. Yet, economic
theory has not given any formal meaning to these terms. The framework of
this paper allows to define these concepts in a natural manner. Note that the
optimal amount invested at time t differs from the amount actually invested.
The difference, given by,
k∣'l'' - kt = -(1 - 02) (Etxt+-∖ - Xt+1 ) (41)
accounts for consumer overconfidence. When the above quantity is positive
consumers save too little and the economy departs from its trend upwards and
operates above its potential. In such a situation the economy can be thought as
being overheated as the temporary spending boost cannot be sustained. More-
over, when the actual level of investment matches the level of investment under
perfect foresight then the level of output can be thought as being equal to its
potential. Revision of expectations either upwards or downwards lead to ei-
ther an upward or a downward deviation of output. These deviations allow
actual output to differ from potential output. The difference can be naturally
interpreted as a measure of output gap.
In general the model allows to define several measures of output. First of all,
the model defines the equilibrium level of output. The level is determined by the
supply side characteristics and by expectations, which influence the composition
of aggregate demand. In addition, there are other concepts. Specifically, the
model defines the level of output that would materialize if agents always formed
expectations using the best available expectation formation technology. The
level can be interpreted as the natural level of output or the potential level of
output. Moreover, the level of output that would exist under the assumption
of perfect foresight need not be, and typically is not, identical with the average
level of output. Finally, there is a level of output that would materialize if
economic agents chose to form expectations in a perfect foresight manner at
a given point in time rather than the way they actually chose to. Note that,
this measure is not the same as the level of output generated by an economy
in which all agents form expectations using perfect foresight technology at all
points in time. This is due to the fact that any change in expectations not only
changes the level of consumption, but also the level of investment. Figures (7)
and (8) present samples of measures.
In summary, expansions and recessions exist because consumers’ expecta-
tions are either overly pessimistic or overly optimistic. Overoptimism leads to
low savings and to a higher level of output than the potential level of output,
while overpessimism leads to high savings and to a level of output below the
potential level of output. The deviations of output from the potential level of
output lead to intertemporal welfare redistribution. Therefore, countercyclical
29