Partial Insurance and Consumption Inequality

Stanford University, Stanford CA

Investigators

Abstract

The main objective of this research proposal is to study the changes in the distribution of household consumption in the US over the 1980-1995 period. This is an interesting period to study due to the dramatic increase in income inequality widely analyzed in the labor economics literature. There is an open debate on the welfare implications of increasing income inequality. One view is that the increase in income inequality is worrying because it implies declining living standards. Another view is that one should rather focus on consumption inequality because consumption is a better measure of welfare than income. The theory suggests that household consumption depends on permanent, not current income, which differs from the former due to temporary fluctuations. The theory's main insight is that absent formal markets for insuring against income losses, individuals use personal savings to "self-insure" against temporary shocks. Thus, if the increase in income inequality is mostly due to temporary factors, the welfare effects of increasing income inequality are less serious than originally thought. The extensive research on the evolution of income inequality in recent years has not been paralleled by comparable research on consumption inequality. The existing research has reached contrasting conclusions (some papers find an increase in consumption inequality, while others find little or no change), or made little attempt to interpret the empirical findings using theory. The project consists of three parts. First, it uses a variety of microeconomic data sets to analyze trends in consumption inequality and contrast them with trends in income inequality in order to explain why different papers reach different conclusions about the main facts of the consumption distribution, focusing on definitional issues and sample selection. Second, it provides evidence (already available in preliminary work) that consumption inequality increases at a slower pace than income inequality over the period of study. One possible interpretation is that most of the increase in income inequality is indeed of temporary nature (income instability). However, labor economics studies have reached the conclusion that income instability and permanent income inequality have increased at similar rates over the period of interest. If anything, permanent inequality outpaces transitory uncertainty. Third, it investigates the empirical validity of an alternative interpretation, that the increase in income inequality is (at least partly) due to permanent factors but there exists "partial insurance" against income shocks over and above personal savings. This form of insurance is provided informally through a variety of channels: social and financial institutions, family network, and rational use of other family choice variables. The investigator provides estimates of the extent of partial insurance against income shocks and assesses how the availability of insurance opportunities changes over time and across skill groups. The empirical analysis will be conducted on two data sets, the Consumer Expenditure Survey and the Panel Study of Income Dynamics.