MONETARY POLICY REGIME SWITCHES AND MACROECONOMIC DYNAMICS*

• DOI: http://doi.org/10.1111/iere.12153
• Date: 01 February 2016
• Published date: 01 February 2016
• Author: Andrew T. Foerster

INTERNATIONAL ECONOMIC REVIEW

Vol. 57, No. 1, February 2016

MONETARY POLICY REGIME SWITCHES AND MACROECONOMIC DYNAMICS∗

BYANDREW T. FOERSTER 1

Federal Reserve Bank of Kansas City, U.S.A.

This article considers the determinacy and distributional consequences of regime switching in monetary policy.

Although switching in the inflation target does not affect determinacy, switches in the inflation response can cause

indeterminacy. Satisfying the Taylor principle period by period is neither necessary nor sufficient for determinacy when

inflation responses switch; indeterminacy can arise if monetary policy responds too aggressively to inflation in the active

regime. Inflation target switches primarily impact the level of inflation, whereas inflation response switches primarily

impact the volatility. Expecting an inflation target switch has minor effects on volatility, whereas expecting an inflation

response switch raises volatility more substantially.

1. INTRODUCTION

During the 1970s, the U.S. economy experienced significant macroeconomic volatility and a

relatively high average inflation rate. In the early 1980s, the Federal Reserve under Volcker

raised interest rates in an attempt to reduce the average inflation rate and lower volatility.

During the recent financial crisis and ensuing slow recovery, economists renewed debate about

monetary policy objectives and the desirability of the Federal Reserve either relaxing its inflation

response or changing its inflation target. Federal Reserve officials responded by suggesting some

tolerance for inflation above its 2% target, but without changing that target. Other economists

suggested a temporary increase in the inflation target to a value in the 4%–6% range (Rogoff,

2008; Blanchard et al., 2010; Ball, 2013). In Japan, after more than a decade of deflation and

low growth, the Bank of Japan responded in 2013 by raising its inflation target to 2% from its

previous 1% inflation “goal.”

In the Volcker disinflation example, a monetary policy switch possibly occurred, either to a

lower inflation target, an increased willingness to fight inflation deviations from target, or both.

In the recent U.S. example, a policy switch could be to a higher inflation target, a decreased

willingness to fight inflation deviations from target, or both. In the Japan example, the change

in a stated inflation goal serves as an explicit switch (Romer, 2013). To the extent that policy

switches occurred in the past and may occur again in the future, economic agents expect that

changes can occur, and these expectations may affect macroeconomic outcomes.

This article examines what two different monetary policy switching assumptions—changing

inflation targets and inflation responses—imply for macroeconomic dynamics. It allows for

discrete changes in the monetary policy rule, both in the inflation target and how strongly the

monetary authority responds to inflation deviations from target, and examines the economy’s

behavior when neither, one, or both policy parameters switch. It studies how policy switches

affect existence and uniqueness of the economy’s equilibrium and how the distributions of

macroeconomic variables change depending upon which parameters switch.

∗Manuscript received June 2013; revised November 2014.

1The views expressed herein are solely those of the author and do not necessarily reflect the views of the Federal

Reserve Bank of Kansas City or the Federal Reserve System. I thank the editor and two anonymous referees as well

as Francesco Bianchi, Troy Davig, Jim Nason, Chris Otrok, and seminar participants at the Federal Reserve Bank of

Philadelphia, the Midwest Macroeconomics, SED, CEF, and Missouri Economics Conferences for helpful comments.

Please address correspondence to: Andrew T. Foerster, Research Department, Federal Reserve Bank of Kansas City,

1 Memorial Drive, Kansas City, MO 64198. E-mail: andrew.foerster@kc.frb.org.

211

C

(2016) by the Economics Department of the University of Pennsylvania and the Osaka University Institute of Social

and Economic Research Association

212 FOERSTER

Much recent research considers how monetary policy impacts macroeconomic stability, in-

cluding Woodford (2003), Christiano et al. (2005), and Smets and Wouters (2007). Changes in

the conduct of monetary policy and changes in macroeconomic performance led to debate over

whether monetary policy remained fixed or changed over time. Using a Markov-switching vec-

tor autoregression (MS-VAR), Sims and Zha (2006) find support for fixed monetary policy with

stochastic volatility instead of switching monetary policy. In a rational expectations framework,

some research supports switches in inflation targets (Schorfheide, 2005), whereas some supports

no switching inflation target (Liu et al., 2011). In addition, several authors (Davig and Doh,

2008; Chib et al., 2011; Bianchi, 2013) provide evidence of switches in the inflation response.

Although these papers only allow for one monetary policy switching type, this article describes

the differences in macroeconomic behavior generated by these different assumptions.

The different monetary policy switching types have different determinacy implications.

Determinacy—the existence and uniqueness of a stable equilibrium—represents an important

consideration for the conduct of monetary policy. Failure to achieve determinacy, Clarida et al.

(2000) and Lubik and Schorfheide (2004) argue, explains the higher macroeconomic volatility

experienced during the 1970s. However, these two papers ignore the potential for repeated

policy changes and the effects of expectations. When the monetary policy rule switches over

time, Davig and Leeper (2007), Farmer et al. (2009), and Cho (2014) show that determinacy

properties change relative to the case when the policy rule remains fixed.

This article makes two contributions regarding determinacy properties of models with switch-

ing parameters: First, it considers inflation target switching, and, second, it considers a model

with predetermined variables instead of a purely forward-looking model. Investigating deter-

minacy previously required a forward-looking model (Davig and Leeper, 2007; Farmer et al.,

2009); models with predetermined variables either could not address determinacy (Svensson

and Williams, 2007; Farmer et al., 2011) or had used a refinement to disregard certain solution

types (Cho, 2014). This article shows that inflation target switching does not impact determi-

nacy. It also shows that satisfying the Taylor principle period by period is neither necessary nor

sufficient for determinacy when inflation responses switch and that indeterminacy can arise if

the central bank responds too aggressively to inflation in the active regime.

In addition to determinacy, this article shows that different monetary policy switching as-

sumptions imply different macroeconomic variable distributions. Switching in the inflation

target primarily affects the average inflation rate in the economy, whereas inflation response

switching primarily affects inflation’s volatility. Both realizations of switches in the policy rule

and expectations about future switches affect the distributions. Different outcomes can result

even in periods when the monetary policy rule is fixed; these outcomes depend on the type of

policy switch that agents expect in the future. If agents expect a switch in the inflation target,

the level of inflation changes, whereas agents expecting a switch in the inflation response cause

the volatility of inflation to change.

The remainder of the article proceeds as follows: Section 2 describes a New Keynesian (NK)

dynamic stochastic general equilibrium (DSGE) model with regime switching and the different

switching types considered. Section 3 discusses determinacy, examining the effects of different

switching assumptions on the existence and uniqueness of the equilibrium. Section 4 discusses

the effects of monetary policy switches on macroeconomic outcomes, considering both the

long-run distributions and the role of expectations. Finally, Section 5 concludes.

2. MODEL WITH MONETARY POLICY REGIMES

This section presents an NK DSGE model where parameters governing the inflation target

and the inflation response change over time. The following subsections describe the model’s

several parts: households, producers, fiscal policy, the monetary authority, and the stationary

equilibrium, followed by the calibration and discussion of the solution method.