There is large body of empirical literature devoted to study the
relationship between inflation and long-run growth. Recently, levine and Renelt (1992)
encouraged by new developments in growth theory investigated, within a unified framework,
the effet of a number of variables on per capita growth. The authors found that there was
no robust srelationship between the two variables. On the contrary, fisher (1991-1993)
using the Levine and Renelt growth equation approach supports the convetional view that
inflation is an important determinant of the rate of economic growth and that the effects
of important determinant of the rate of economic growth and that the effets of inflation
are stronger at low and moderate inflation levels. Levine and Zervos (1992) include in the
same framework an index of economic policy and concluded that growth and low inflation low
budget deficit are positively correlated. Additional evidence supporting a negative
relationship between inflation and growth can also be found in De Long and Summers (1992)
and De Gregorio (1993), among others.
The predominantly negative correlation between inflation and growth observe in the data
has not been properly rationalized in models where identical agents behave rationally and
where money has a significant impact of the evolution of real variables. In monetary
versions of the neoclassical growth model the quantitative importance of money is quite
modest inducing only small growth and welfare effects and playing almost no role in
explaing the fluctuations of real variables. Because of the same reason, these models have
not been successful at identifying a channel through which inflation plays a more
meaningful role in the economy.
There are numerous plausible channels through which may affect growth and welafre.
However, the implications of many of them have not been fully explored or the simply have
not been successful. Feasible channels are nominally denominated depreciation allowances,
partially indexed tax bracketing, eserve requiriment on bank deposists, invetment
purchases subjet to cash-in advance (CIA) constrain (Stockman, 1981), investment purchases
and labor service payments subject to CIA constraint (Chistiano, 1991) etc.
Nevertheless, as a result of this research program, the distorting effects of inflation on
the labor leisure choice has risen as the basic mechanism at work in monetary models . In
models with no growth (Cooley and Hansen, 1989), inflation reduces labor effort through
its effect on the return to working because part of the labor income has to be carried
over,as cash balances, into the next period’s cashgood trade. In models with
endogenous growth (Gomme, 1993; Jones and Manuelli, the rate of growth of the economy.
Within the first type of models, the welfare cost of a 10% inflation rate was calculated
in 0.4% of income; within the second, Gomme (1993) computes a welfare cost of less than
0.03% of income for a 8.5% inflation rate. This kind of evidence endorses the generally
accepted conclusion that welafre costs of inflation are very samall and they are even
smaller in models with endogenous growth(1).
In this paper I explore one alternative avenue through which inflation can have real
effects and estimate its quantitative importance. The assumption that taxes are directly
collected in money is imposed to capture the real world feature that money is the
requiered means of taxation payment. Most, if no all, of the literature has studied
econmies in which money exclusively has a private use (to buy goods or assetes or factor
payments) ignoring its public use in taxation and the fact that they are closely related
in modern economic arrangements where the value of money is not tied down to gold or any
other kind of backing. It has long been recognized that if the government “(..)
declines to accept some kind of money in payment of obligations to itself, it is difficult
to believe that it would retain much of its general acceptability. (...) Its general
acceptability, which is its all-important attribute, stands or falls by its acceptability
by the state” (Lener, 1947). In consequence, it is natural to consider an economy in
which money fulfills two functions: the government accepts money from households in the
settiement of tax liabilites and money is used as a medium of exchange.
The paper is organized as follows. In sections 2 and 3 I study three model economies
sharing the common features of steady state growth and tax payments explicity modeled as a
monetary obligation. I assume that taxes have to be paid with fiat money accumulated in
advance. Welfare and growth effects of inflation are studied in an exogenous growth model,
and endogenous growth model with human capital accumulation. The principal finding is that
the size of growth and welfare effects are igher than those found in comparable monetary
models. In contrast to the exisiting literature, welfare costs are driven by the effect of
inflation on the rate of growth instead of the effect on the labor-leisure choice. In an
economy with monetary taxation, inflation strikes the growth rate directy through the
after-tax real rate of returm on investment. This is the same channel through which
distortionary taxation has important real effects (Rebelo, 1991).
In section 4 a real business cycle model (RBC) with monetary taxation is parameterized,
calibrated and simulated. I address the cuestion of how the ability of the RBC model is
affected when the tax payment technology is imposed. Section 5 extends the business cycle
model to incorporate liquidity effects. The paper provides a “monetary” economy
in which the observed labor market anomalies related to the correlation and relative
volatility of hours worked and average productivity are not present. Section 6 presents a
summary and conclusions. |