Academic journal article Economic Inquiry

Determining Output and Inflation Variability: Are the Phillips Curve and the Monetary Policy Reaction Function Responsible?

Academic journal article Economic Inquiry

Determining Output and Inflation Variability: Are the Phillips Curve and the Monetary Policy Reaction Function Responsible?

Article excerpt

I. INTRODUCTION

The natural-rate hypothesis that no long-run trade-off exists between inflation and unemployment has won over most macroeconomists. A new trade-off between the variability of inflation and the variability of output has lately attracted considerable attention. Instead of minimizing a weighted sum of the levels of inflation and unemployment, the central bank's objective function minimizes a weighted sum of the variances of inflation and output. Given this loss function for the central bank, a great deal of theoretical research has been undertaken as to which type of monetary policy rule is both efficient and robust. Clarida et al. (1999) and an edited volume by Taylor (1999b) investigate various monetary policy rules.

In this article, I explore the determinants of inflation and output variability. First, in a theoretical study, I present the theoretical studies of Taylor (1994), Svensson (1997), and Ball (1999a) and investigate how the policy parameters in a Taylor monetary policy reaction function are related to the variability of output and inflation. Computational experiments show that first, as monetary policy makers increase their degree of responsiveness to inflation, the variability of inflation falls but the variability of output increases. Second, if policy makers increase their policy response to output gaps, the variances of both inflation and output decrease. In addition, the experiments illustrate that the changes in the variability of output and inflation can be the result of changes in the responsiveness of prices to output gaps.

The second part of the article provides cross-country evidence regarding the relationship of inflation and output variability to both the policy parameters of the monetary policy reaction function and the slope parameter of the Phillips curve. By analyzing data from 20 Organisation for Economic Cooperation and Development (OECD) countries, this study estimates the parameters of the monetary policy reaction function and the slope parameter of the Phillips curve, and examines the relationships between the estimated parameters and the variability of output and inflation.

Although there have been numerous empirical investigations of monetary policy reaction functions (Taylor 1999a, Clarida et al. 2000, among others), there is very little empirical evidence concerning a relationship between the parameters in the policy reaction function and the variability of output and inflation for several countries. Furthermore, there is even less empirical literature that considers the relationship between the policy parameters and output and inflation variability in a cross-country framework. This study attempts to extend the empirical analysis to a group of countries by using international data. (1)

There are two main findings. First, inflation will be low and stable in countries where the monetary policy rule has large reaction coefficients, and second, output will have large fluctuations in countries where the Phillips curve is relatively flat. (2)

The rest of this article is divided up into four sections. Section II provides an overview of the Taylor-Svensson-Ball (TSB) model. The relationship between the model's parameter values and inflation and output variability is explored through simulations. Section III describes the empirical data for the 20 OECD countries and reports the estimated parameters of the TSB model. Section IV presents the results of the cross-country regressions used to evaluate the determinants of inflation and output variability. Finally, section V offers conclusions.

II. THE TSB MODEL

Assumptions

The TSB model describes the economy in three equations:

(1) [[??].sub.t] = -[beta]([i.sub.t-1] - [[pi].sub.t-1] - [r.sup.*]) + [lambda][[??].sub.t-1] + [u.sub.t],

(2) [[pi].sub.t] = [[pi].sub.t-1] + [alpha][[??].sub.t-1] + [e.sub.t],

(3) [i.sub.t] = [[pi]. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.