Did US Consumers 'Save for a Rainy Day' Before the Great Recession?

Abstract

The weak implication of the permanent income hypothesis, entailing that private savings decisions reect rational expectations about future income, is still widely regarded as a relevant for quantitative macro modeling. We investigate the ‘saving for a rainy day’ hypothesis by testing weak exogeneity of income and consumption with respect to cointegration parameters, and by exploring the direction of Granger causality between consumption and income. We use both aggregate macro time series for the US, as well as disaggregate quarterly data for the 100 largest US Metropolitan Statistical Areas (MSAs). The common sample is 1980q1-2007q4, i.e., a period that starts with the stagation era and ends with the Great Moderation. Since the exogeneity of consumption and the endogeneity of income with respect to the savings rate, is a premise for the belief in a low valued scal policy multiplier, our nding that exogeneity of consumption can be rejected more often than it can be accepted, has relevance, both for analysis of the causes of the Great Depression and for discussing the eectiveness of scal policy. We also give proof that house price changes is likely to have aected consumption and income, before, during and after the Great recession.