What the Piketty Errors Mean

Remember the Reinhart/Rogoff spreadsheet error? In the event that you do not, here is a summary. Those who follow debates between economists will recall that the spreadsheet error led to all kinds of excorations of Reinhart and Rogoff on the part of liberal economists, who claimed that Reinhart and Rogoff were responsible for austerity policies that killed off economic growth. Even Stephen Colbert got in on the act. Their spreadsheet error was considered to be the worst tragedy that befell the planet since that one time when Oedipus and Jocasta had a super-awesome first date.

Of course, the excoriations were vastly overstated, but that didn’t stop intellectual opponents of Reinhart and Rogoff from engaging in hyperbole on a grand scale. Now that Thomas Piketty has been caught making his own significant errors, comparisons have naturally been made between Piketty on the one hand, and Reinhart and Rogoff on the other.

The problem with these comparisons is that they fail. Reinhart and Rogoff may have made a spreadsheet error, but there is a very plausible argument that the error did not affect their conclusions, and there was no serious accusation on anyone’s part–not even Reinhart’s and Rogoff’s more severe critics–that Reinhart and Rogoff engaged in intellectual or scholarly fraud.

Piketty’s case is quite different. To be sure, there are a number of errors that Giles and Giugliano caught that have been deemed “fat finger errors,” meaning that they were honest mistakes. But there are other errors that cannot be explained away as honest mistakes–at least not at this point. Maybe Piketty will be able to show how these errors were honest in nature. But maybe not, and at this point, accusations of fraud leveled against Piketty are very real and a source of serious concern. Reinhart and Rogoff, for all of the grief that they got over a spreadsheet error, never had to deal with accusations of fraud.

Meanwhile, Paul Krugman tries to engage in some pushback against Piketty’s critics. Krugman acknowledges that “Piketty will have to answer [questions regarding his data] in detail, and we’ll see how well he does it.” He stresses again at the end of his post that Piketty “need[s] to respond to each of the individual questions.” But Krugman insists that “Giles is proving too much; if his attempted reworking of Piketty leads to the conclusion that nothing has happened to wealth inequality, what that really shows is that he’s doing something wrong.”

There are several points that need to be made in response to all of this.

The first is that admitting that “Piketty will have to answer [questions regarding his data] in detail, and we’ll see how well he does it,” and Piketty “need[s] to respond to each of the individual questions” is certainly quite a journey from claiming–as Krugman initially did–that “if you think you’ve found an obvious hole, empirical or logical, in Piketty, you’re very probably wrong. He’s done his homework!” Of course, Krugman doesn’t make mention of that latter statement, because he doesn’t want his readers to remember that he all but declared Piketty to be infallible before this entire brouhaha started.

The second point that needs to be made is that Giles and Giugliano were not arguing that “nothing has happened to wealth inequality.” Although the two state that “once the FT cleaned up and simplified the data, the European numbers do not show any tendency towards rising wealth inequality after 1970,” they also state that “[t]here is some evidence of a rise in the top 10 per cent wealth share since 1970.” So Krugman is misrepresenting Giles’s and Giugliano’s conclusions. Quelle surprise.

The third point worth making is that as Tyler Cowen argues, Krugman’s claim that “[w]e all know that wealth inequality has gone up” may not itself be correct. As Cowen writes, “Keep in mind that the “new and improved numbers,” as produced by Chris Giles, are showing doubts about the course of measured wealth inequality in the UK. Maybe wealth inequality hasn’t gone up.” Cowen continues:

Now maybe that does “have to be wrong.” But if the “new and improved” numbers are wrong, it is hard to then argue Piketty’s wealth inequality numbers can be trusted. In which case we are back to knowing that income inequality has gone up, but not knowing so much concrete about wealth inequality. (That is one reason why my own Average is Over focuses on income, and on labor income in particular, because that is where the main action has been.) The data section of Piketty’s book, which has gathered so much praise, then is not so useful, though by no fault of Piketty’s. We might think it likely that wealth inequality has gone up, but if we are going to do these selective overrides of the best available data, we cannot trust the data so much period or otherwise cite it with authority. We also could not map wealth inequality into particular measures of the r vs. g gap at various periods of time.

If there is one big lesson of the FT/Piketty dust-up, it is that we don’t have reliable numbers on wealth inequality.

Do read the whole thing. Note as well that there are other forms of inequality that we might want to focus on instead, but we are fixated in Ahabesque fashion in wealth inequality. And again, it is worth noting that there were plenty of problems found with Piketty’s claims even before Giles and Giugliano issued their blockbuster findings.

So let’s please stop with the comparisons between Piketty on the one hand, and Reinhart and Rogoff on the other. The Piketty imbroglio is far more serious. And neither Piketty, nor Krugman, nor any other defender of Piketty have provided satisfactory response to the charges raised by Giles and Giugliano.

2 Replies to “What the Piketty Errors Mean”

Comments are closed.

%d bloggers like this: