Disinflation and the Recession-Now-Versus-Recession-Later Hypothesis: Evidence from Uruguay


WP/95/99-EA
Disinflation and the Recession-Now-Versus-Recession-Later Hypothesis:
Evidence from Uruguay
by Alexander W. Hoffmaister and Carlos A. VĂ©gh


Casual empiricism suggests that the timing of the recessionary costs
associated with an inflation stabilization program depends on the nominal
anchor used. While under money-based stabilization the recession occurs in
the early stages of the program, under exchange rate-based stabilization the
recession appears to take place in the late stages of the program. The
choice of a nominal anchor would thus entail a choice between recession now
(money anchor) or recession later (exchange rate anchor).

This paper offers empirical evidence on the recession-now-versus-
recession-later hypothesis for the case of Uruguay, a chronic inflation
country. Formally, the paper estimates a vector-autoregression model (VAR),
which includes the rate of depreciation, the rate of monetary growth,
inflation and output, and controls for Argentina's influence on the
Uruguayan economy. This VAR model is used to simulate the output response
to both a money-based and an exchange rate-based stabilization.
Technically--and departing from standard practice--the model is subjected
to a series of innovations of the policy variable to ensure that it follows
a predetermined path.

The impulse responses for output indicate that a money-based
stabilization results in an initial contraction, while a (temporary)
exchange rate-based stabilization leads to an initial expansion followed
by a later contraction. The econometric evidence is thus broadly consistent
with the recession-now-versus-recession-later hypothesis. Furthermore, the
evidence suggests that the high degree of dollarization of the Uruguayan
economy may hinder severely the effectiveness of a monetary anchor.