You may remember a week ago, I wrote, Capital in the 21st Century Not Refuted. It was in regard to an article by Chris Giles at the Financial Times, Data problems with Capital in the 21st Century. My article was mostly about what the political consequences of this would be. Regardless of what Giles found or how Piketty would respond, conservatives would use the initial Giles article to dismiss income inequality. As I concluded: “I imagine watching This Week in two years. Paul Krugman will be on it and say, ‘Thomas Piketty showed that income inequality is getting worse.’ And Peggy Noonan will scoff, ‘Oh, the Financial Times proved that was rubbish years ago!’ And George Stephanopoulos will just shrug. ‘Who can say?'”
We will see if this comes to pass, because Piketty was pretty quick with his response—a 10 page paper, Technical appendix of the book Capital in the twenty-first century Appendix to chapter 10. Inequality of Capital Ownership; Addendum: Response to FT (pdf). If you are technically minded, it is quite a fun read. It starts off with Piketty gushing about how grateful he is that the Financial Times is checking his work and that this is exactly why he made all of the data in the book available on spreadsheet. It comes off very much like a lion talking to a lamb about how happy he is that he decided to join him for dinner.
Piketty then goes on to talk about the data on wealth inequality: it is far worse than the data on income inequality. Implied: you did read the book, didn’t you? But they can agree on that, and Piketty talks about how these data will be getting better in the future in part because of his own and his colleagues’ work. At this point, the lion has not touched the lamb; it is smiling with just a hint of disapproval.
But Piketty continues to take it slow. He then goes in and discusses each of the four countries that he studied: Sweden, France, the United States, and the United Kingdom. He deals with each thusly:
- I see we agree on the overall numbers. Excellent! But you note some minor errors in the years? Yes, I explained this all in the book and even more in the appendix, but I should have been more clear. You see, you can’t treat wealth estimates on a yearly basis because there aren’t enough data and the data are too variable to deal with year by year. But thank you so much for pointing out this issue that I clearly didn’t highlight enough in the book. (The lion smiles broadly.)
- I see we agree on the overall numbers. Excellent! But you make much about a single data point regarding how I correct for mortality. This is all clearly explained in the spreadsheets and the papers I put online for you to use. “This looks a little bit like criticism for the sake of criticism.” (The lion squints at the lamb and looks disapprovingly.)
- United States
- I see we agree on the overall numbers except in the most recent period. Yes, this is sad. The data from the United States is not as good as it is for France. This was well documented in the book. I made the best estimates that I could based upon the data. You have made different estimates. But it is interesting that the most recent and reliable estimates by Emmanuel Saez and Gabriel Zucman agree more with my initial uncertain estimates. Their estimates actually find a larger rise in wealth than I did. This is now the state of the art and we should use this new data, but it does make your estimates look even more pathetic than they did based simply on my own careful work. (The lion licks its lips.)
- United Kingdom
- Ah, here we really do see a very big difference! You see wealth inequality actually decreasing in the recent decades. I cannot say that my methodology is correct. But you have broken up the data and used what I consider the more accurate (and higher) estate tax data in the early years and then shift to survey data that are known to under report total wealth. And those data are from a single year: 2010. Why would you do that other than to simply misrepresent the data? (The lion opens wide.)
That’s basically it. In one sentence: Chris Giles manipulated the US and UK data in such a way to prove what he wanted to prove and his methodology is worse than unwarranted—it is disingenuous. Piketty scored a knockout. Or if you prefer: the lion swallowed the lamb whole.
In fact, the case against Giles and the Financial Times and their over-hyping of the errors they found make me think that this may not give conservatives the cover they so desperately desire. One can hope.
Update (30 May 2014 8:23 pm)
I can’t seem to get enough of this. I’m reading what everyone else is writing about it. I was particularly struck with what RA at The Economist had to say, “One suspects that Mr Giles will not be satisfied by the response, and the discussion is likely to continue.” If this is true, it means that Giles is simply an ideologue who is totally resistant to argument and data. The best he can reasonably do is move the argument to minor points. Because his major point was totally destroyed. As Piketty noted:
If Giles continues this debate, he will lose all credibility. I fully expect Giles to lose all credibility.