Piketty’s Erroneous Data

By Chris Edwards of Cato.

"Many people are interested in the distribution of income and wealth and how it may have changed over time. But there is no single and undisputed source for such data. Rather, economists construct historical time series using partial information and many assumptions.

French economist Thomas Piketty and colleagues have for years been publishing data showing extreme changes in top 1 percent income and wealth shares in the United States over the decades. Many news outlets report the information unquestioned, despite evidence that Piketty is sloppy with data and makes bad assumptions that throw his calculations off.

The other day in the Wall Street Journal, Phillip Magness and Vincent Geloso described errors in the top 1 percent income data of Piketty and his colleagues. The Cato Institute published a collection of essays critiquing Piketty’s theories and data in 2017. Alan Reynolds has been finding flaws in Piketty’s data since 2007.

I noticed this article in the New York Times magazine last week by editor Willy Staley that parrots Piketty’s incorrect claim about the persistence of top wealth holdings. It is classic Piketty. He asserts a supposed general rule of wealth, but appears to have done insufficient research to see if it is actually true. Here’s Staley:

In his book “Capital in the Twenty‐​First Century,” the French economist Thomas Piketty notes that the new economic order has made it difficult for the superrich not to get richer: “Past a certain threshold,” he writes, “all large fortunes, whether inherited or entrepreneurial in origin, grow at extremely high rates, regardless of whether the owner of the fortune works or not.” … “Once a fortune is established, the capital grows according to a dynamic of its own,” Piketty notes, adding that bigger fortunes tend to grow faster — no matter how extravagant, their owners’ living expenses are still such a small proportion of the returns that even more is left over for reinvestment.

Piketty’s claims here are incorrect, as Ryan Bourne and I discussed in “Exploring Wealth Inequality.” To back his claim, Piketty points to data from Forbes: “One of the most striking lessons of the Forbes rankings is that, past a certain threshold, all large fortunes, whether inherited or entrepreneurial in origin, grow at extremely high rates.” And he adds that, “the largest fortunes grew much more rapidly than average wealth. This is the new fact that the Forbes rankings help us bring to light.”

Forbes has published an annual list of the 400 Americans with the highest net worth since 1982. Ryan and I found that only 21 from the 1982 list were still on the list in 2019. Where have the others gone? Numerous people have died and their wealth divided among heirs. The wealth of many others has stagnated or declined because of income taxes, consumption, charitable giving, poor investment choices, and business failures.

Piketty appears to have only looked at the winners on the Forbes list and did not account for people who lost wealth and dropped off the list. He doesn’t seem to have recognized that the rich can make bad investment choices, can manage their companies poorly, and are often outcompeted in the marketplace. The world’s richest man on Forbes global list in 1987 was Yoshiaki Tsutsumi, who was worth $20 billion. His fortune plunged to just $1.2 billion in 2006, and then he dropped off the list.

In the Cato Journal, Robert Arnott and coauthors examined the Forbes lists and found that of the 400 individuals on the 1982 list, just 69 individuals or their descendants remained on the 2014 list. They found that the wealth of those 69 people had grown far more slowly than if they had simply invested passively in stocks and bonds in 1982 and let their holdings grow. They conclude that “dynastic wealth accumulation is simply a myth.”

Similarly, the Tax Foundation’s William McBride looked at changes in wealth for the 400 individuals on the 1987 Forbes U.S. list through to the 2014 list. He calculated the growth in wealth for the 73 people who stayed on the list, and he estimated the growth for those who dropped off by assuming that the drop‐​offs had barely missed the wealth threshold for the 2014 list. With that assumption, he found that the average annual real wealth growth rate over 26 years for the people on the 1987 list was at most a meager 2.4 percent. By contrast, the average annual real return on U.S. stocks over the decades has been about 7 percent.

McBride found that people on the Forbes lists who had inherited their wealth grew their fortunes more slowly than those with self‐​made wealth. Active entrepreneurs often generate new wealth, but individuals on the lists who had inherited did not earn outsized returns—instead, their wealth was eaten away over time, as noted, by taxes, consumption, philanthropy, and bad business and investment decisions.

Reporters doing stories on income and wealth should read more than just Piketty. Economists across the political spectrum have found flaws in his data and theories. Since many assumptions are used to construct historical data on income and wealth distributions, reporters should at least convey to readers the uncertainty in such information and the different ways it can be assembled and interpreted."