Inequality Rising in Spending, Not Just Income
Professor Erik Hurst finds that faulty government statistics led to the perception that either the rich were scaling back, or the poor were living beyond their means.
Economics largely agree that the income gap between the rich and the poor has dramatically widened over the past 30 years. But there is another measure of inequality that has confounded the academic community. The consumption gap - the difference between how much the rich and poor spend - has opened up at a much slower rate, about three times slower than the income gap.
It has led some economists to presume that while the rich were getting richer, they weren't necessarily living larger. Or, the poor were living beyond their means. Perhaps both were true. However, "such a conclusion is unwarranted," writes Erik Hurst, V. Duane Rath Professor of Economics and John E. Jeuck Faculty Fellow. According to Hurst, the understanding of consumption has been clouded by faulty data.
Most studies of consumer spending rely on the same error-prone dataset. The data came from the government's Consumer Expenditure Survey (CE), an in-depth accounting of how much households spend on various goods and services conducted by the Census Bureau on behalf of the Bureau of Labor Statistics. Portions of the CE produce reliable estimates, but the survey as a whole generates strange and increasingly unreliable estimates of total household spending. It consistently falls short of more reliable estimates from national accounts data - the broad measures gathered by the Commerce Department's Bureau of Economic Analysis. Furthermore, the survey may not fully capture the expenditures of the richest households, which are crucial to any understanding of inequality.
In their paper, "The Evolution of Income, Consumption, and Leisure Inequality in The US, 1980-2010," Hurst and his coauthors from Stanford University and University College London urge researchers to dig deeper into the Consumer Expenditure Survey data. The summary data may be fraught with errors, but certain sections of the survey can provide more accurate snapshots of household spending.
For instance, one section asks households to keep a diary of expenditures over the course of two weeks. The diaries offer surprisingly detailed estimates of spending on food, entertainment, and cars. Hurst and his coauthors develop new measures of consumption based on these fine-grained datasets. They also draw on data from outside the Consumer Expenditure Survey to see if it tells a consistent story.
A striking feature emerges from these myriad datasets. Whether the measures are based on food, entertainment, or an entirely separate survey of household consumption, they all tell the same story: "Consumption inequality increased substantially," the coauthors wrote. Moreover, all of the measures, "mirror the overall change in income inequality." A line graph shows the different measures of inequality rising and falling roughly in tandem over the past 30 years.
Ezra Klein summarized the study's findings in the Washington Post: "We've got it all wrong," he wrote. "The data we were using is bad, and consumption inequality is going up alongside income inequality." In other words, the debate over inequality is settled, and it's just as bad as the income figures suggested.
However, the authors find one dataset that tempers their findings on inequality. To truly understand if the rich are leading a higher quality of life than the poor, researchers also should consider how much time each group spends in leisure activities. Here, the data show that poor Americans have more time to read books, socialize with friends or family, and generally enjoy life outside of work. Men with less education have gained 2 1/2 hours of leisure time a week, while men with more education have lost 1 1/2 hours a week. As Klein of the Washington Post noted, "the nonrich are 'consuming' more leisure time than the rich, and that should also factor into our thinking on consumption inequality." ■
Photo by Chris Lake