Modelling Real Private Consumption Expenditure in South Africa to Test the Absolute Income Hypothesis

Abstract

This paper explores, the hotly debated topic among economists and policymakers, whether fiscal and monetary policies impact on households by examining the relevance of the absolute income hypothesis in explaining private consumption expenditure and its relationship with household disposable income in South Africa. Worldwide, private consumption expenditure remains a big puzzle for leading consumption function theories. Friedman’s permanent income hypothesis posits that private consumption expenditure is not affected by how much consumers earn on a daily basis, but by what they expect to earn during their lifetime. Friedman’s permanent income hypothesis is at odds with Keynes’s absolute income hypothesis, that private consumption expenditure is affected by fiscal stimulus policies, which are effective for increasing economic activity and employment. Subscribing to the former underrates the potential power of fiscal stimulus policies and other monetary or trade policies that boost short-term income. The overarching objective of this paper is to ascertain whether patterns of private consumption expenditure in South Africa are determined by Friedman or Keynes’s theory. The paper specified econometric equations with quarterly seasonally adjusted data from the South African Reserve Bank for the sample period 1984 to 2015 and estimated them with cointegration techniques consisting of the Engle-Granger two-step approach. The importance of the paper and its scientific novelty are that it is more realistic since it specified models that take into account the reaction time of the dependent variable when the independent variable changes by imposing lags on the variables. The empirical results indicate that in South Africa, when household disposable income changes over time, private consumption expenditure depends more on a household’s previous disposable income than its current disposable income. The main empirical finding is that the absolute income hypothesis is not appropriate in explaining private consumption expenditure in this country. Even when the interest rate was included in a modified absolute income hypothesis, the overall estimates were not robust. Hence, estimates of the short- and long-run regression models were not consistent with the absolute income hypothesis. This is in line with arguments put forward in some extant studies using this model, that the fiscal stimulus policies might not generate the desired increased economic activity and employment. If households use money from the fiscal stimulus policies to bail themselves out of existing debts rather than consume additional goods and services which, would be the catalyzer to increase Gross Domestic Product (GDP).