A New Critique of Piketty Has Its Own Shortcomings
My colleague, Neil Irwin, has written an outstanding account of the charges leveled by Chris Giles of the Financial Times that the conclusions of 'Capital in the 21st Century,' Thomas Piketty's blockbuster book, 'do not appear to be backed by the book's own sources.'
I'm not (yet) convinced by these charges. Before digging in, let's give credit where it is due: The Financial Times has done a superb job in digging deeply into Mr. Piketty's data. And Mr. Piketty deserves enormous credit for having attempted the very hard task of extracting meaning from a disjointed and piecemeal historical record of the evolution through time of the distribution of wealth.
I drew the following five conclusions from The Financial Times's re-analysis:
Not all differences are errors. Some of the issues that The F.T. highlights are quite straightforward, such as where Mr. Piketty appears to have typed in data from the wrong row of a table. But many other issues are less questions about the validity of the data than they are about incomplete referencing. And yet others appear to be judgment calls, albeit made with insufficient documentation.
They essentially agree. While it's quite natural for a journalist to emphasize the differences between his findings and those of a famous author, the most striking fact is how closely The F.T.'s analysis agrees with Mr. Piketty's. Their preferred time series for the evolution of wealth inequality in the United States, Britain, France and Sweden are remarkably similar.
Not all differences are equally important. To the extent that The F.T. and Mr. Piketty disagree - and as I say, I think these instances are relatively minor - it's not yet clear whether the cause is obvious errors as pointed out by the newspaper, or judgment calls where perhaps the professional economist deserves the benefit of the doubt.
The F.T.'s bottom line is muddier than it looks. Mr. Piketty's broader project is about bringing together disparate sources of data on wealth inequality to try to figure out what has happened through time. This involves seeking ways of comparing otherwise not-very-comparable sources of information, which is a tough job, requiring substantial domain-specific expertise and judgment. The F.T. does a nice job in raising specific concerns with specific data points. But in trying to put together their own series, the paper is at least as guilty as Mr. Piketty of making some pretty big assumptions about the comparability of quite different data sets.
This is a debate about wealth inequality, not income inequality. Mr. Piketty is famous for having produced estimates of not just the share of wealth held by the top 1 percent (or top 10 percent), but also for his estimates showing a greater concentration of income among high earners. The F.T. re-analysis is entirely about wealth inequality, and no serious questions have been raised about the veracity of the data on the rising income shares of the top 1 percent.
Of course, one can't help but be reminded of the kerfuffle about an earlier research paper written by Carmen Reinhart and Ken Rogoff. I fear the similarities are deeper than most realize, with partisans already engaged in gaining political mileage out of sloganeering about fairly inconsequential spreadsheet errors, rather than digging more deeply into what the data actually say. The difference is that this time the political football is an empirical result that is an article of faith among liberals, rather than conservatives. Beyond that, though, it's different mudslingers, but similar mud.
The Financial Times analysis is definitely provocative. While it raises important questions, I'm not convinced it does more than that. Mr. Piketty has already written a fairly general response to The F.T.'s analysis, but he has yet to respond to the specific charges made. One hopes that in time Mr. Piketty will write a longer and more detailed point-by-point reply, admitting errors where they exist, and defending his data choices where they are defensible.
And perhaps some good will come from all this. The clearest effect of Mr. Piketty's efforts is that he has brought new attention to the distribution of wealth. I believe that interest will continue to improve and refine our understanding of the evolution of wealth inequality.
Post a Comment for "A New Critique of Piketty Has Its Own Shortcomings"