The Piketty fallacy
“Piketty? Oh yes; he’s the man who got his sums wrong.”
That’s why I’m so angry with him: not because the sums are wrong, but because his contribution to the debate has now been fatally undermined by Chris Giles’s attack. I predicted that the right would challenge every aspect of his work, and they have; but the one that counts, in terms of deflecting the debate away from what matters, is the revelation of the transposition and spreadsheet copying errors. The rest of the attacks are now unnecessary: the Piketty bomb is now defused, and his banana-skin consigns him to the ranks of the Left’s harmless: remember Neil Kinnock stumbling at the sea-shore or Michael Foot’s duffle-coat.
It’s almost irrelevant that publications as unlikely as the Economist and the New York Times leapt to Piketty’s defence, or at least to attack Giles in turn: whether or not Giles overstated the case no longer matters. To understand why, just turn to another recent non-fiction publishing phenomenon, Daniel Kahneman’s Thinking, Fast and Slow: THE NUMBERS DON’T MATTER.
The contents of Piketty’s spreadsheets will never affect the interpretation any of us brings to this debate. The right will always believe that inequality is a good thing (a necessary incentive to enterprise, resulting in a bigger cake for all, with trickle-down etc.), while the left will always believe that inequality is a fundamental assault on fairness. And some of us believe that incentives are useful up to a point, but when they are unnecessarily excessive they need to be moderated – but we too will probably continue to believe that irrespective of the evidence.
Because fundamentally evidence and observation don’t play the same role in economics as they do in science, but we think they do. And neither does mathematics (professional economists like Chris Giles probably hate Piketty as much for his cynicism about arithmetic as for his leftist tendencies): in economics, complex modelling has exactly the same function as it had in mediaeval astrology, while economists want you to think that their observations and calculations are as precise as modern astronomy.
But even an astrologer should take the trouble to avoid simple, discoverable errors. It may not be a science, but it is a discipline that ought to be conducted to professional standards, however boring that may be.