EKIP– Expert Club for Economics and Politics A Different Opinion
June 16, 2014 by Phil Magnes
In the past few days, Thomas Piketty and a few of his more unscrupulous defenders have begun to declare the data controversy surrounding “Capital in the 21st Century” over, following the publication of Piketty’s response to Chris Giles of the Financial Times.
I argue that the controversy is not over in the least, and I would draw your attention to five specific parts of his work that remain unresolved.
1. Piketty has yet to provide a clear, transparent, and source-specific description of how he calculates his measurements of wealth inequality for the United States.
As I have pointed out here and here, Piketty draws quite liberally on at least four different data sources when assembling his widely acclaimed calculations for the United States. Moreover, he does so in ways that are visible in his raw Excel spreadsheets but that are hidden or completely absent from his annotation—notably the juggling act between the estate tax and his inequality estimates, which are based on surveys dating back to around 1970. Neither Piketty’s technical appendix nor his appendix in response to Giles address this discrepancy.
2. Piketty's methods for estimating the top 10% of the wealth distribution are simplistic and unclear.
Although not perfectly constructed, most of Piketty’s figures for the wealth of the top 1% are drawn from a variety of sources, such as surveys or tax records. However, he often refers to similar calculations for the top 10%, and in some cases suggests clear trends. A look at Piketty’s Excel files suggests that most of his figures for the top 10% are directly transposed from his calculations for the top 1%, although he usually does so in a vague and amateurish way that simply adds a lot of rough calculations or tries to fit the above figure to additional sources that are not clearly cited and explained.
3. Piketty falls into a methodological contradiction due to his tolerance of different techniques for assessing wealth inequality.
Wealth inequality is notoriously difficult to measure, although the two most common approaches attempt to estimate it from (a) tax data or (b) survey samples. Piketty severely criticizes the Financial Times for using data from a survey of wealth distribution in the UK that suggests less pronounced inequality trends than he claims in his book. He also strongly argues for the methodological superiority of the tax-based estimate… for the UK.
However, in constructing his own time series for the United States, Piketty moves quite freely between calculations based on tax data and those taken from studies by other scholars. He also clearly prefers the US survey data to the calculations based on tax data when the latter show a flat trend over the period for which his arguments imply an expected rise. He has yet to clarify this seemingly glaring methodological inconsistency.
4. Piketty's historical data is full of elementary factual errors, inexplicable omissions, and questionable methods.
Errors of this kind are numerous and pervasive in Piketty's book. They range from minor errors and omissions in the handling of historical data to major errors of interpretive significance to his thesis. A partial list of these problems includes:
- Misrepresentation of the amount of income tax in the United States
- Missing historical data on the amount of property tax in the United States
- Misrepresentation of the size of historical minimum wage levels in the US
- Fictional numbers for 19th century US tax revenue.
- Wrong averages in UK inheritance transfer data, which confuse (or perhaps manipulate) data for the wrong decades
- Using dubious and perhaps fabricated data on the Soviet Union and other communist countries, artificially distorting its global capital/income ratio index.
To date, Piketty has not addressed or corrected any of the errors.
5. Piketty greatly exaggerates the confirmation of his conclusions in the works of other scholars.
Piketty largely attempts to circumvent the aforementioned problems in the U.S. wealth data by relying on the results of a new, ongoing study by his longtime collaborators Gabriel Zucman and Immanuel Saez. Although Piketty and many of his supporters have already accepted the validity of their study, it has yet to be published or subjected to academic scrutiny. It currently exists only in the form of a PowerPoint slideshow. A review of the results they claim to provide suggests that their model may be skewed by one or more external factors related to the interest rate data. And while it has gone largely unnoticed by Piketty’s defenders, Wojtek Kopczuk—another leading inequality measurer (and sometime collaborator of Saez) who is more familiar with the unpublished Saez-Zucman methodology than the PowerPoint slideshow audience—recently suggested that it has a fundamental methodological flaw.
Translator: Evelina Sharapanova
Editor: Daniel Vassilev
The original text is published here.
