The Piketty Book Problem: Only as Good as Its Data
Business + Economy

The Piketty Book Problem: Only as Good as Its Data


Here we go again. Another seminal economic work is coming under fire, with questions raised about the accuracy of the data and spreadsheets underpinning its politically charged conclusions.

The subject of scrutiny this time is French economist Thomas Piketty, whose 577-page blockbuster, Capital in the 21st Century, published in the U.S. earlier this year, has shot to the top of bestseller lists and added fresh fuel to the debate over income and wealth inequality.

Related: Why Economists Are Finally Taking Inequality Seriously

Piketty’s influential work — embraced by liberals and challenged by conservatives — argues that, with the notable exception of the post-war boom years, the return on capital has historically been greater than economic growth and gains in income. Western economies, Piketty says, are reverting to that norm. In short, the rich get richer in capitalist societies and will keep doing so unless we take measures to combat the widening gap.

The Financial Times on Friday called those conclusions into question, or tried to. Chris Giles, the newspaper’s economics editor, reported that a review of the data Piketty used (and made available online) turned up spreadsheet errors and other serious data problems. The “rock-star French economist,” the FT says, “appears to have got his sums wrong.”

Giles found that Piketty made transcription errors in copying data into his spreadsheets, and that he averaged data for Britain, France and Sweden, which would give the numbers for Sweden, with its smaller population, undue weight in the results.

Giles’s allegations, though, involve much more than adding two plus two and coming up with five. The FT report takes issue with the way Piketty manipulated raw data, making seemingly arbitrary adjustments. It also charges that the economist “cherry-picked” some figures. “Some numbers appear simply to be constructed out of thin air,” Giles wrote.

Adjusting for those data issues, Giles said, results in a picture very different than Piketty’s. "Two of Capital in the 21st Century’s central findings — that wealth inequality has begun to rise over the past 30 years and that the U.S. obviously has a more unequal distribution of wealth than Europe — no longer seem to hold," Giles wrote.

Related: What Piketty Left Out of His Argument

Some of the FT’s claims are reminiscent of findings last year that substantially undercut research on the relationship between debt levels and GDP growth. “The FT found mistakes and unexplained entries in his spreadsheets, similar to those which last year undermined the work on public debt and growth of Carmen Reinhart and Kenneth Rogoff,” Giles wrote. That work had also been embraced by politicians, though in that case, those on the right, not the left, relied on the economists’ work to back up their policy agenda.

The revelation last year of errors in Reinhart and Rogoff’s work essentially obliterated the idea that a debt-to-GDP level of 90 percent represented a dangerous threshold above which economies struggle to grow. The FT’s findings with regard to Piketty, while they will inevitably be bandied about by politicians, may not have the same impact.

Piketty may have made mistakes, even some embarrassing ones. He will have to clarify his data and offer a more detailed response than he has so far. But it’s not clear yet that those mistakes fundamentally undercut his ultimate findings.

In his response to the newspaper, Piketty generally, if somewhat vaguely, stood by his data choices and broader conclusions. He wrote that he had sought to make “the most justified choices” when he made adjustments to data necessary to smooth comparisons across different countries and different time periods.

"I have no doubt that my historical data series can be improved and will be improved in the future,” the economist wrote, “but I would be very surprised if any of the substantive conclusion about the long-run evolution of wealth distributions was much affected by these improvements.”

Related: Inequality in Capitalist Systems Is Not Inevitable

Others digging into the data have suggested that the FT’s claims are not quite as strong as the newspaper suggests. “FT is massively overstating its case re: Piketty,” conservative columnist Reihan Salam tweeted. “Conservatives shouldn't get roped into this one.”

The biggest concerns appear to involve Piketty’s data on the U.K., but the FT’s charts using revised data for France and Sweden, for example, still look nearly identical to Piketty’s own. For a more detailed comparison and critique of the data issues, see here, here, here, here and here.

Giles himself admits that the FT’s revised data doesn’t necessarily answer all questions about the trends in wealth inequality. “There is one important caveat," he writes. None of the source data at the basis of Piketty’s work is completely reliable. While this post is clear about what is wrong with Piketty’s charts, it is much less certain about the truth.”

What is clear, as of now, is that Piketty will have to explain in further detail his methods and his math — and that, as always, even the most popular and influential of economic theories is only as good as the data supporting it. When it comes to studying wealth in particular, choosing precisely which data to measure, and how to slice it and dice it, is bound to be messy.

Top Reads from The Fiscal Times: