Detecting Historical Inequality Patterns: A Replication of Thomas Piketty's Wealth Concentration Estimates for the United Kingdom

an article by Phillip W. Magness (American Institute for Economic Research) published in Social Science Quarterly Volume 100 Issue 5 (August 2019)

Abstract

Objective
This article utilizes a replication exercise to evaluate the reliability of the historical time series for top wealth share concentrations in the United Kingdom, as presented in Thomas Piketty's Capital in the Twenty‐First Century (2014a).

Method
Using Piketty's identified source records, an attempt is made to replicate the construction of his time series for top wealth concentrations in the U.K. These results are then compared against the series presented in Capital, and subsequent improvements by other scholars.

Results
Piketty's time series is shown to diverge substantially from its source data from the U.K., and does not appear to be replicable. In particular, Piketty's series introduces a sizable post‐1980 adjustment that suggests a substantially more rapid acceleration of wealth concentration than its source statistics reveal. Issues of reliability in the U.K. time series mirror similar problems with Piketty's wealth estimates for the United States, although their implications for historical interpretation differ in light of subsequent data.

Conclusion
These findings indicate that Piketty's account of changing wealth concentrations in the United Kingdom in the 20th century is unreliable for interpreting recent patterns in the evolution of top wealth shares. An alternative interpretation of the source data is therefore offered, pointing to a century‐long L‐shaped pattern in place of Piketty's depicted U‐curve.

JEL Classification: N14, D63

---