Reports of the 1 percent's decline are greatly exaggerated

Last month The New York Times’ David Leonhardt and his data analysis team at The Upshot dropped one hell of an “actually” in the headline of his piece “Inequality Has Actually Not Risen Since the Financial Crisis.” It’s a very authoritative statement for a news organization, and it goes against the conventional wisdom about how the 1 percent has fared since the 2007 financial crisis.

“Facts matter,” Leonhardt writes about the stakes of his post, and such a definitive claim presents the perfect opportunity to examine exactly how these facts are determined.

Allowing for the possibility that we’ve been wrong about how the rich are doing over the past few years, there’s still something fishy about Leonhardt’s post. The rise and fall of national inequality over time is a complicated methodological question, but he relies overwhelmingly on a new paper by labor scholar Stephen J. Rose. In his paper for the nonpartisan think tank the Information Technology and Innovation Foundation, Rose’s claims are simple enough: First, 2012 numbers overstated the income of the wealthy because coming tax changes encouraged them to backload future pay. Second, measures that exclude government transfer programs ignore the safety net’s positive effects. He argues that the Obama administration’s plans are working and we should keep pursuing a “robust economic growth strategy.”

But does all that really mean inequality has not risen since the financial crisis? First of all, it depends what your definition of “since” is. Rose was responding to a paper by University of California at Berkeley economist Emmanuel Saez, who examined the share of income accruing to the top percentiles of the income distribution. On most measures, the rich haven’t regained their 2007 position, but the 2012 glut put the top 10 percent of earners over the mark, leading less cautious outlets — such as The New York Times — to proclaim the rich were getting newly rich in the recovery. But all the metrics, including Rose’s, show inequality increasing since 2009. So the answer to the question depends whether “since the crisis” means since precrisis 2007 or since postcrisis 2009. Rose argues the former gives a better picture, but when we’re talking about recovery income, the latter seems more honest. Rose does not comment on Saez’s examination of 1993 through 2013.

Even if we agree on a time frame, however, not everything is settled. Income inequality is one thing, but Leonhardt and Rose are making claims about inequality in general. The rich don’t live paycheck to paycheck, and as Rose writes, income from capital gains is “lumpy”; their liquid income doesn’t accurately describe the lifestyles of the rich and famous. And though wealth is supposed to be relatively constant, the housing market crash revealed asset values are flexible too, and many Americans lost a lot.

To get at both income and wealth, Richard Fry and Rakesh Kochhar at the Pew Institute used Federal Reserve data from the Survey of Consumer Finances to look at the relationship between upper- and middle-income families, as defined by their distance from the median. They then compared their wealth numbers, and the gap they found is the largest in the survey’s history. Their data shows people making the least money taking the hardest hit from the financial crisis: While upper-income families lost 11 percent of their wealth from 2007 to 2013, middle-income families lost 39 percent, and low-income families lost 48 percent. These aren’t the ultrarich that Saez and Rose are talking about, but there’s a notable division here, and government transfers clearly haven’t bailed out low-income Americans the way Rose and Leonhardt would have us believe.

No matter how you look at it, the question of whether America’s rich are getting richer is about how you ask. It’s a question of ideology. For Rose and his patrons at the Information Technology and Innovation Foundation, a rising economic tide managed by Democratic policy will raise all boats. The percentage change in incomes since 2007, as Leonhardt graphs it, sets the ground at the top of the precrisis mountain, framing a popped bubble like a downturn to overcome. And overcome they probably will; Leonhardt admits, “It’s even possible that inequality will soon surpass its 2007 peak.” That would be, to say the least, incommensurate with his leading narrative.

Facts are important, but data doesn’t speak for itself. I find the way Saez pursues the numbers more informative because he focuses on share of income captured; it’s a perspective that foregrounds the economy’s class-competitive nature, and past tax data give him a century’s worth of trends about the top 10 percent of earners from which to draw. The rich are either getting more or less of the total product over time, and keeping track of that number reveals a lot about the distribution of value within society. In his illustrations, Saez graphs the climbing income share of the rich, broken down from the top 10 percent. The charts reveal decades’ worth of growth captured disproportionately by the wealthiest of the wealthy.

But I still have a problem with this presentation of the data. When we see a graph like this, it’s natural to associate relative line positions with the welfare of the people the lines represent. It looks like, for example, in 1973 the top 1 percent weren’t doing so hot, at least comparatively. While the Y axis measures a pile of money, the lines represent groups of different sizes. It’s convenient for fitting them all in the same visual space, but comparing the 1 percent’s share to the share of the top 10 through 5 percent is like comparing apples and apple trees. The 1 percent splits its share among fewer people. To get a graph from which we can easily draw the correct inferences about how much the people represented are earning over time, we need to divide each group by its size.

The result looks like this:

From this perspective, the fluctuations of the top 1 percent’s income share is just some turbulence on a decades-long skyward trajectory, during which they have doubled their portion. Looking at this graph, where the ultrawealthy are so high they can barely differentiate the merely loaded’s share from the ground, it is clear that America’s rich are indeed getting richer. It’s not the only way to look at the question, but it’s no less valid than anything Leonhardt offers to the contrary, and I can’t help thinking that the second graph is a truer representation of the same numbers. The Upshot post was a dangerous combination of click-bait and press-release journalism, credulously repeating a big conclusion in bold without the diligent good-faith investigation required.

The public — or at least the bottom 99 percent of it — deserves better.