Measuring Poverty: Income vs Spending
Here is the essence of his argument:
Thanks to a loyal reader for the pointer.
Among low-income households in the United States, the gap between reported income and reported spending has widened gradually since the 1960s and now has taken on chasm-like dimensions. In the early 1960s, the poorest quarter of U.S. households spent 12 percent more than their annual incomes. In 1973, spending by America's poorest fifth surpassed their income by almost 40 percent. And in 2004, spending by the poorest fifth of American families exceeded income by a whopping 95 percent; in effect, spending was nearly twice as much as income.
These patterns might be due to easy access to credit, with many consumers maxing out their credit cards or engaging in other unsustainable borrowing. (Curiously, however, recent credit surveys suggest tha the net worth of poorer Americans has been rising, not falling.)
Another important factor could be the increasing instability of American incomes. Scholars such as Jacob Hacker at Yale University and Robert Moffitt at Johns Hopkins University have noted that the income of American families is likely to bounce around much more today than it did three decades ago -- whether due to greater global competition, increasing rewards for education or other factors. Intensified swings, in turn, mean that more households may, in any given year, earn low incomes and be temporarily classified as living in poverty. But they continue to spend as they did before, anticipating that their incomes will bounce back. Such oscillations also mean that the incomes reported by families in annual surveys -- the backbone for the official poverty estimate -- are a steadily less accurate indicator of true living standards.
Addendum: An NBER study by Dirk Krueger and Fabrizio Perri a few years back also suggested that income inequality and consumption inequality tell different stories.
Update: A comment usefully points to a longer version of Eberstadt's views on the topic.