Four Ways to Prune a Rose: Why the NYT Missed the Mark on the Inequality Debate

February 19, 2015

On Tuesday, the New York Times‘ data-minded blog The Upshot posted an article reporting on an inequality study performed by Georgetown University’s Stephen J. Rose. The article‘s title boldly stated that “Inequality Has Actually Not Risen Since the Financial Crisis.”

Although the study’s analysis is mathematically correct, the study is framed and conducted in a way that makes its findings irrelevant to the larger discussion of inequality in America. Here’s why:

1. Rose’s study examines income and ignores wealth, which is actually the most stark indicator of inequality in America today; the top 10 percent owns 93 percent of all stocks and 61.9 percent of all wealth. This is equivalent to judging someone’s wealth by looking at their paychecks without considering the value and appreciation of the businesses, houses, cars, and bank accounts that are also in their name.

2. Rose’s analysis ends in 2011, so he does not consider the ongoing recovery, the majority benefits of which are once again going to the very wealthy.

3. The current inequality dialogue, championed by Capital author Thomas Piketty and Roosevelt Institute Senior Fellow and Chief Economist Joseph Stiglitz, specifies that wealth is the primary driver of inequality, so singling out income makes the Rose study irrelevant to the important public discourse going on today.

4.  Even if we were to accept the income-only study as valid, there are many facts that call Rose’s findings into question:

  • The income group he describes fell from a record historical high to a slightly lower historical high.
  • Higher incomes have continued to recover since the study’s statistical conclusion.
  • The primary reason lower incomes didn’t fall further is that they were bolstered by government transfers (see chart and caption below).

Rose examines post-tax-and-transfer figures rather than pre-tax-and-transfer figures and argues that this supports his thesis of declining inequality. In fact, the difference is largely due to the rising number of individuals who paid less taxes and qualified for more government benefits due to their loss of income. A greater number of people dependent on government assistance would not fit most definitions of a reduction in inequality.