International Evidence on the Long-Run Impact of Inflation David E. Rapach Abstract In this paper, I use a structural vector autoregression framework to analyze the effects of a permanent change in inflation on the long-run real interest rate and real output level in 14 industrialized countries. Long-run monetary superneutrality is rejected for all 14 countries using annual data: the results indicate that a permanent increase in inflation lowers the long-run real interest rate in each country; a permanent increase in inflation also increases the long-run real output level in a number of countries. Long-run monetary superneutrality is also rejected for four out of the five countries examined using quarterly data. Under long-run monetary superneutrality (LRSN), a permanent increase in the growth rate of the money stock has no real effects—apart from real balances—in the long run. In particular, a permanent increase in the growth rate of the money stock leads to an equal increase in the long-run inflation rate, but leaves the long-run real interest rate, capital stock, and real output level unchanged, as in, for example, the simple classical and well-known Sidrauski (1967) models. While there is general agreement that permanent changes in inflation arise solely from equal permanent changes in money growth,1 a substantial body of theory suggests that permanent changes in money growth and thus inflation can alter the steady-state values of the real interest rate, capital stock, and real output level. In such models, money is not superneutral in the long run.2 In the full-employment model of Mundell (1963), a permanent increase in inflation lowers the real interest rate if saving depends on real money balances. Higher inflation raises the nominal interest rate and reduces wealth (via a lower demand for real money balances); saving increases and the real interest rate must fall to [End Page 23] restore goods market equilibrium.3 In the augmented Solow growth model of Tobin (1965), a permanent increase in inflation leads agents to substitute capital holdings for real money balances in their portfolios. This reallocation increases the steady-state capital stock and real output level and lowers the long-run real return to capital, which equals the real interest rate in equilibrium. The negative long-run real interest rate response and positive long-run real output response to a permanent increase in inflation is known as the Mundell-Tobin effect. More recently, overlapping generations models have been used to demonstrate that the Mundell-Tobin effect emerges when agents are explicitly optimizing; see, for example, Weiss (1980) and Espinosa-Vega and Russell (1998).4 Stockman (1981) demonstrated the possibility of a reverse-Mundell-Tobin effect in an inflation-tax model. In his explicitly optimizing model with a representative infinitely lived agent, a permanent increase in inflation leads to a lower steady-state capital stock and real output level and a higher long-run real interest rate due to a cash-in-advance constraint on investment spending.5 Given the ample cause to question LRSN on theoretical grounds, it is important to evaluate its empirical relevance. To this end, I estimate the long-run real interest rate and real output responses to a permanent increase in inflation for 14 industrialized countries using annual data from the postwar era. Estimates of the long-run responses are obtained through a trivariate structural vector autoregression (VAR) model in the inflation rate, nominal interest rate, and real output level. Inflation, preference, and technology structural shocks are identified in the VAR using long-run restrictions motivated by standard neoclassical theory. This paper extends the recent studies of King and Watson (1997), Weber (1994), Bullard and Keating (1995), and Koustas and Serletis (1999). These studies employ bivariate VARs to estimate the long-run real interest rate response or the long-run real output response to a permanent increase in inflation. Canova (1994) observes that it is more satisfactory to analyze these responses in a single model, as is allowed under the present paper's trivariate framework.6 Importantly, the present framework identifies and thus controls for two nonmonetary structural shocks with potentially significant real effects—preference and technology shocks—when analyzing the...
Read full abstract