Abstract

Standard economic theory holds that households rely on precautionary savings and
other insurance mechanisms to smooth consumption in response to income shocks, and that poorer households are more exposed to the business cycle. These propositions imply less dispersion and lower volatility in consumption than in income, and countercyclical movements in both income and consumption inequality. In this paper we document the opposite patterns, drawing on electronic consumption transaction data linked to administrative income records for the full population of Norwegian households. In the short to medium run, consumption is (i) more unequally distributed than income across households, (ii) more volatile than income within households, and (iii) both consumption and income inequality move procyclically. We show that the elevated levels of consumption inequality and volatility primarily reflect idiosyncratic expenditure shocks, rather than household-level life-cycle trends or income risk. Procyclical inequality, by contrast, is driven by heightened business cycle sensitivity among households in the upper tail of the distribution. Our findings call for a reassessment of precautionary saving motives, highlight the role of redistributive policies, and point to an important disconnect between the sources of household risks and the aggregate business cycle.