Justin Wolfers' Zombie Piketty Confusions

Justin Wolfers posted a slideshow regarding Thomas Piketty's Capital yesterday. Wolfers has been an interesting character to watch on Piketty, especially after his excessive praise of Larry Summers' abortive and error-riddled effort at critiquing it. Perhaps not surprisingly, Wolfers' slideshow recreates the exact same reading errors as Summers. Wolfers should correct these errors.

Summers Quote

Let's start with the Larry Summers quote Wolfers includes approvingly:

[Piketty's] argument is that capital or wealth grows at the rate of return to capital, a rate that normally exceeds the economic growth rate. Thus, economies will tend to have ever-increasing ratios of wealth to income, barring huge disturbances like wars and depressions. Since wealth is highly concentrated, it follows that inequality will tend to increase without bound until a policy change is introduced or some kind of catastrophe interferes with wealth accumulation.

This quote from Summers should never be shared with someone else unless it is meant to ridicule Summers' failure to comprehend even the most basic parts of of Piketty's book. There are two undeniable errors in this quote:

  1. Piketty does not claim that wealth grows at the rate of return to capital. He says that wealth grows by the amount of the national income that is saved each year, which is brutally obvious.
  2. Piketty does not claim that the wealth-to-income ratio will be "ever-increasing." He says the wealth-to-income ratio will move towards the the savings rate divided by the growth rate (s/g) and then stay there.

Two out of the three claims in this quote have absolutely nothing to do with what Piketty argued. If Wolfers thinks they are in the book, I'd challenge him to find them in there somewhere. He won't be able to.

Wealth-to-Income Ratio

The Summers' point above might seem like a nitpick, but the rest of the slideshow makes it clear that Summers' mistakes have infected Wolfers' entire understanding of Piketty. Summers' mistakes are now Wolfers' mistakes. The biggest of those mistakes is a confusion surrounding Piketty's theory of why the wealth-to-income ratio will increase.

As I explained in depth before, Piketty has two separate theories for why income inequality will increase. The main one is this: The growth rate will decline but the savings rate will not. This will cause the wealth-to-income ratio to rise. The rate of return on capital will not decrease enough. Thus, capital's share of the national income will increase. Since capital is distributed very unevenly, an increase in capital's share will cause income inequality to rise.

That's Piketty's first inequality theory. Note that it has nothing whatsoever to do with r > g. The thing that is driving the increase in the wealth-to-income ratio is the rise in s/g (savings rate divided by growth rate). It has nothing to do with r > g.

Wolfer's wrong account of Piketty, which he has clearly cribbed off of wrong Larry Summers, goes like this though: The rate of return on capital is higher than the growth rate (r>g). Capital income is saved at 100% and labor income is saved at 0%. Therefore, wealth will increase at a faster rate than national income. Therefore the wealth-to-income ratio will rise. And so on.

This is wrong. Wolfers has mixed together both of Piketty's income inequality theories (the Capital Share Effect and the Capital Concentration Effect) and managed to produced some Frankenstein argument that makes no sense.

Wolfers writes at one point in his slide: "r > g  ... implies that wealth grows faster than income, only if ... all capital income is reinvested (and no labor income is)." Let us count the errors:

  1. r > g has nothing to do with whether wealth will grow faster than income. Whether wealth will grow faster than income depends on s/g.
  2. All capital income does not need to be reinvested for any of Piketty's arguments.
  3. Imagine two worlds. In one, 100% of capital income is reinvested and 0% of labor income is. In another, 100% of capital income is reinvested and some of the labor income is as well. In which world is it more likely that wealth will grow faster than income? The latter obviously because it has a higher savings rate. Wolfers has said the opposite here (spoiler: this is because he has mixed up the Capital Share Effect and the Capital Concentration Effect).


Wolfers slideshow is wrong in basic ways. It is clear that his wrongness is derivative of Summers' wrongness. He should stop misrepresenting Piketty's argument. He should read this post that actually explains Piketty's basic arguments correctly.