RDP 1999-03: Householders' Inflation Expectations Introduction

Inflation expectations have wide-reaching effects on the macroeconomy. Wage bargaining, price setting, asset allocation and investment, for example, all depend on inflation expectations in one way or another. Accordingly, and particularly in the context of an inflation-targeting regime, understanding how inflation expectations are formed and behave is central to setting monetary policy.

Inflation expectations can be measured in a variety of ways. They can be inferred from financial prices, for example, by taking the difference between nominal and indexed-bond yields. Or different groups of people can be asked directly what they expect to happen to inflation. This paper focuses on the Melbourne Institute survey of householders. The paper has two aims. The first is to examine how householders' expected average inflation varies with particular personal characteristics, such as age, occupation, education and place of residence. The second is to identify the macroeconomic fundamentals that affect householders' expectations of inflation.

Section 2 sets the scene by describing the survey and providing some basic statistics on the overall distribution of inflation expectations. The sample period starts in January 1995 since this is when detailed questions about householders' personal characteristics were first asked. Section 3 examines how average rates of expected inflation vary according to type of householder. People with better access to information or with more developed information-processing skills – for example, people in professional jobs, those who have more education, or those who are older – have inflation expectations which are systematically lower and closer to actual inflation. Inflation expectations of householders with these sorts of characteristics are also more responsive to what is happening with actual inflation.

Section 4 shifts focus to the interaction of inflation expectations with developments in the economy in order to understand the structural foundation of householders' expectations. We start with a cross-section of individual responses from January 1995 to April 1998. When asked about prospects for the economy and jobs, people associate ‘good times’ with low inflation and low unemployment. On the face of it, this suggests that people prefer low inflation; high inflation makes people feel worse off and less sure about the future.

In both cross-section analysis of individuals' responses, and time-series analysis of householders' median expected inflation rates, the key direct macroeconomic influences behind inflation expectations are actual inflation and monetary policy. Inflation expectations tend to move with actual inflation and they move inversely with the cash rate, defined in either real or nominal terms. Tighter monetary policy directly reduces inflation expectations, in addition to its indirect effects on inflation expectations through its influence on inflation via the exchange rate and the output gap. Other macroeconomic variables – the output gap, import price inflation, changes in bilateral or multilateral exchange rates, wages growth or changes in unit labour costs – were found to have no systematic effect on households' inflation expectations. These results are robust across specifications and sample periods. Measured inflation expectations are also at variance with economists' notions of ‘rationality’.