The main defect in Thomas Piketty's "Capital in the Twenty-First Century" has little to do with the questions about data that the Financial Times raised this weekend. Nonetheless, they're good questions. The FT should be thanked for raising them, and Piketty owes his readers better answers than he's offered so far.
I've argued that the main thing wrong with the book is that it makes a large claim that it then fails to support. Piketty says capitalism embodies a fundamental contradiction: In the nature of the system, inequalities of income and wealth tend to rise inexorably, with plutocracy the logical consequence. He provided a theory of sorts to back this up, but it's far from compelling -- not really a theory at all, more a blend of accounting identity and guesswork. He also offered a lot of supposedly corroborating historical data. The numbers in the book didn't back the claim either.
And that was assuming they were correct. In one way, the FT's finding that some of Piketty's numbers are suspect doesn't advance the discussion very far -- because the figures weren't much help to Piketty's Big Idea in the first place. Even so, in digging so diligently through the data, Chris Giles and his colleagues have performed a valuable service. The data in question can't settle the argument about the fundamental contradiction of capitalism, but they influence the political context and matter in their own right.
Did wealth inequality rise in the U.K. between 1980 and 2010? Piketty said it did. This conformed to (even if it didn't prove) his account of capitalism's underlying logic, and it affirmed many people in the belief that inequality is the defining issue of our age.
The FT says that Piketty's own data sources aren't so sure about this trend: It seems to be flat, not rising. And Piketty's finding for the U.K. appears to drive a similarly dubious result about rising wealth inequality in "Europe" (which he takes France, the U.K. and Sweden to represent).
It would be good to get this and other matters straight. Piketty appears to have made some outright mistakes. That's inevitable in this kind of work. He's made a lot of tweaks and adjustments to source data; he's had to construct data where the sources have gaps; he's had to choose among different sources where they conflict. Some of these choices look questionable.
Piketty's first response to the FT's analysis is not substantive, and he'll need to say more. He may have good answers.
Everybody, including critics such as myself, praised Piketty's industry and ingenuity in gathering data on inequality. The FT's findings don't make me think that this praise is undeserved. Unless there's worse to come, Piketty's reputation as a top-rank empirical investigator will survive. It's much to his credit, and attests to his good faith, that he put all his data online. We wouldn't be having this conversation otherwise. Rather, the FT's criticisms are a needed reminder of just how difficult and error-prone this kind of research is.
A lot of judgement goes into compiling data, especially when complex definitions are involved and when comparisons across countries and over long spans of time are being attempted. The FT's discussion makes the hazards plain.
These problems dictate a certain modesty when it comes to drawing conclusions. In much of the book, as I pointed out, Piketty is appropriately modest about the limits of the data. But his mood fluctuates page by page. Before you know it, he's vaulting from his slippery data to propound a new law of capitalism.
What about the idea that the FT has made too much of all this (a charge, by the way, that isn't confined to the book's admirers)? It's nonsense.
Granted, a quarrel about statistics wouldn't normally justify a front-page article with detailed supporting material and an editorial. But "Capital in the Twenty-First Century" isn't just any book, or so its admirers have been telling us. Maintaining a calm sense of proportion, experts greeted this truly superb and awesome tome as a watershed work of extraordinary importance, one that would change the way economic policy is discussed from now on. As for the data -- that was worth a Nobel prize by itself. Piketty's opus is a runaway bestseller, a leading topic of conversation among intelligent people everywhere. Economists and economic commentators have been discussing little else for weeks.
The experts were so busy being enraptured that they didn't bother to do the checking that the FT decided to undertake. Now they're discussing the FT's criticisms -- and so they should be.
This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.
To contact the author on this story:
Clive Crook at firstname.lastname@example.org