Metropolitan Study of Household Consumption and Income Inequality

RSF’s program on Social Inequality is broadly concerned with the impact of rising inequality on the increasing disparity in the quality of life between the rich and poor. But rising inequality may also have a negative effect on the quality of life across all socioeconomic classes. Maria Charles (University of California, Santa Barbara) suggests that one implication might be changes in consumption patterns and savings – as inequality rises, people may consume more and consequently save less and incur more debt. In order to examine this relationship, Charles will look cross-sectionally at the relationship between inequality and consumption in large metropolitan areas in the United States.

Central to the previous work on consumption is the concept that exposure to higher- income individuals may cause households to increase consumption, either because people seek to gain acceptance by emulating the high levels of consumption of those at the top or because the presence of high-income individuals changes standards about what goods are considered acceptable. Charles will examine these possibilities by estimating the effects of aggregate-level income distributions in twenty-one large Metropolitan Statistical Areas (MSAs) on the consumption behavior of households in these areas. Contextual variables will be measured at the MSA level, while indicators of consumer spending will be analyzed at an individual household level. Charles will differentiate between consumption of positional and nonpositional goods to identify differences in patterns of basic and luxury spending. She will also examine in depth two MSAs in the same state, one with high inequality and one with low inequality, to shed light on the direct impact of inequality on consumption holding constant the potential impact of state laws or regional differences. 

Measuring across MSAs is especially relevant because there should be variation in both income levels and income inequality, as well as how much exposure people have to high-income individuals. MSAs are defined by commuting patterns and act as proxies for regional labor markets, in which socioeconomic classes come in regular contact because workplaces tend to be economically diverse. To obtain indicators of consumer spending, Charles will employ public-use microdata from the Consumer Expenditure Survey from 2006 to 2008.  The sample size is approximately 5,000 households, with 250 per MSA. Aggregate level information for each MSA is available from the American Community Survey and the Consumer Price Index, and includes variables such as mean- and median-household income, housing costs, industrial structure, rates of poverty and unemployment, and cost of living. Charles plans to write several articles as a result of this project, and will work with the University of California Office of Public Affairs on other methods of outreach to make the results publicly available.