RDP 8808: Consumption and Permanent Income: The Australian Case 1. Introduction

The share market crash of October 1987 focussed attention on the effect of changes in wealth or permanent income on aggregate consumption expenditure. Unfortunately, there is very little clear empirical evidence on the role of wealth in determining consumption in the Australian case.^[1] There are numerous studies using U.S. data based on the method of Hall (1978) which find varying degrees of relevance for the permanent income hypothesis. However, very few Australian studies have pursued the Hall approach and the subsequent extensions. Exceptions include Johnson (1983) using data on expenditures on non-durables, and Jubb (1986) who examines both durable and non-durable expenditures.

The purpose of this paper is to examine the role of permanent income in explaining aggregate consumption expenditure in Australia using a variety of tests. One test uses a modification of the approach taken by Hall (1978). An advantage of the Hall approach is that with sufficient assumptions, aggregate wealth data is not required to test the role of permanent income. An alternative approach proposed by Hayashi (1981), uses data on non-human wealth. The relationship between the two approaches is developed in this paper, and applied to test for the roles of permanent and current disposable income in explaining aggregate consumption.

We find that tests extending the Hall and Hayashi approaches both give strong and consistent evidence in favour of the permanent income hypothesis, especially when the aggregate consumption data is recalculated with expenditure on durable goods converted into a flow of consumption services from these goods. However, we also find that there remains a residual, though significant, proportion of consumption expenditure that is explained by current and not permanent income.

Section 2 of the paper derives the testable implications of the life cycle hypothesis following the work of Hall (1978), Flavin (1981), Hayashi (1982) and Mankiw and Campbell (1987). By allowing for some proportion of consumers to base their decisions on permanent income and some part on current income, a more general model can be derived. In Section 3, the data which is used to test the theories are discussed. A modification of Hall's approach is tested using quarterly and annual data for the period 1962(1) to 1987(2) and the Hayashi approach is tested using annual data for the period 1961/62 to 1986/87. A conclusion is summarised in Section 4.

Footnote

Summaries of early Australian studies of the consumption-savings decision are provided by Williams (1979) and Freebairn (1976). [1]