Tags Matching:

John Sayles

  • Originally published 01/06/2015

    Piketty goes cherry-picking for US Wealth Inequality

    Liberty and Power

    One of the main critiques I’ve made of Thomas Piketty’s data, starting on this blog last May and appearing in my newly released paper with Robert Murphy on the subject, concerns his Figure 10.5, purporting to show the trend of wealth inequality in the United States. As I pointed out at the time, this chart suffers from severe distortions starting in the 1970s that come about from Piketty’s unconventional decennial averaging techniques. The result is a downward distortion that gives the appearance of a clear bottom point in 1970, followed by a steady increase in inequality through the present day. Piketty attains this composite trend by essentially cherrypicking from different data sources until he produces the story he wants to depict, even as the trend he claims is far more ambiguous in the raw numbers.Though his citation practices are often opaque, I reconstructed and labeled the source match-up for Figure 10.5 in the following graph so you can see the cherrypicking at play: