The many ways to measure economic inequality
From President Obama down to local minimum-wage ordinances, issues of economic inequality have been pushing their way back into the national conversation — particularly among Democrats. In a Bloomberg survey earlier this month, 71% of Democrats said the government should work to narrow the gap between rich and poor, while 68% of Republicans said the government should stand aside and let market forces operate freely.
But economists disagree on just how much inequality there is and how best to measure it. As Federal Reserve economist Arthur Kennickell wrote in a 2009 paper, ” ‘[i]nequality’ may seem a simple term, but operationally it may mean many different things, depending on the point of view.” Most researchers agree that wealth is much more unevenly distributed than income, while consumption is less concentrated at the upper end than either wealth or income.
Probably the most-familiar inequality measures involve income. The Census Bureau annually publishes two measures of income inequality each year; according to the most recent report, the top 5% of households received 22.1% of “equivalence-adjusted” aggregate income last year — nearly as much as the bottom 60% of households (27.2%). (The “equivalence-adjusted” estimates adjust for different household sizes and compositions.)
The Census Bureau also reports the Gini index, a summary statistic that measures the dispersion of incomes on a scale of zero (everyone has exactly the same income) to 1 (one person has all the income). The income Gini for the U.S. has been rising for decades: On an equivalence-adjusted basis, the Gini was 0.362 in 1967 and 0.463 last year. (Other economists, such as Berkeley’s Emmanuel Saez, have tracked income inequality over long periods using somewhat different measures and reached similar conclusions.)
But some economists say that income data have too many flaws to be the primary measure of inequality. For one thing, most income-inequality measures use income before taxes and transfer payments (such as Social Security, food stamps and unemployment benefits), which act to reduce inequality. According to the Organization for Economic Cooperation and Development, taxes and transfers cut the United States’ income Gini from .499 to .380 (although that latter figure still is among the highest in the developed world and also has risen over time).
Other researchers note that an individual’s (or household’s) income can vary considerably over time, and may not reflect all the economic resources available to them — such as credit availability, government assistance or accumulated family wealth. They argue that consumption is a better measure of economic well-being.
Such studies typically find that consumption inequality is less than income inequality, though still significant. A 2012 study from the American Enterprise Institute, using data from the Consumer Expenditure Survey, found that the top 20% of households by income accounted for nearly 40% of total expenditures, while the bottom 20% accounted for less than 10% of expenditures. As the chart shows, the gap between the top and bottom has remained relatively constant — a finding that echoes those of other consumption-oriented researchers.
But other economists have looked at consumption data and reached different conclusions. One recent study, prepared by two U.S. Census researchers and University of Wisconsin economist Timothy Smeeding and presented at the 2013 annual meeting of the American Economic Association, found that consumption inequality grew about two-thirds as much as income inequality between 1985 and 2010.
Another study, by researchers from Princeton and the University of Rochester, adjusted the CES data for what they called “systematic measurement error” and concluded that consumption inequality does in fact track income inequality, as high-income households have shifted their spending away from necessities and toward luxuries. (The CES data, they argue, underestimate consumption of certain types of goods and are vulnerable to richer households under-reporting their consumption generally.)
A third way to look at economic inequality involves household wealth. People of great accumulated wealth may not receive much in the way of income (trust income and capital gains on stocks and other investments, for example, often are excluded from income analyses), while people who earn a lot but also have high expenses (such as child care or tuition) may not consider themselves especially wealthy.
Wealth inequality tends to be much higher than either income or consumption inequality, but also to be more stable over time. NYU economist Edward Wolff, for instance, has used data from the Survey of Consumer Finances (and similar prior surveys) to track household net worth over time.
Wolff found that in 1962, the top 1% of households held 33.4% of all wealth; in 2010 their share was 35.4%. The biggest increase came in the tiers of wealthy just below the 1% — their wealth share rose from 33.6% in 1962 to 41.3% in 2010. As for the bottom 40%? Their share fluctuated between 1.5% and -1% (i.e., negative net worth) for the entire four-decade period Wolff studied.
Drew DeSilver is a senior writer at Pew Research Center.