Does the Marginal Elasticity of Substitution Have an Ideology?

In my column at The Week, I noted that Capital in the 21st Century argues capital accumulation threatens higher inequality by increasing capital's share of income. And even if capital's share of income doesn't rise due to capital accumulation, as Matt Rognlie and other critics have highlighted, that would be because additional capital is useless. Hardly an argument for making accumulation a policy priority.

In response, Rognlie repeated something he's said before: that a left-wing ideological commitment to a high elasticity reverses the historical valence. In the past, lefties generally highlighted low estimates, precisely because they invalidate arguments in favor of a tax preference for capital.

This morning it occurred to me that this change is more significant than simply an inconsistency in the ideological time series. Specifically: the main contribution of Capital in the 21st Century to economics as a field, beyond the superlative empirical (and some theoretical) work Piketty's done in his papers, is the unification of inequality with macro. This is something that Branko Milanovic has pointed out, in his initial review of C21 and many times since on his blog. If all we had was a representative agent and inequality is completely unmodeled, then estimates of the Marginal Elasticity of Substitution are all that matter for even hinting at the impact of macroeconomics on inequality and assessing policy proposals like a tax preference for capital.

But if we get to bring inequality out into the open, then we can set up what I've come to call the Piketty Puzzle: the best estimates of the Marginal Elasticity of Substitution, interpreted as an actual parameter in production, are that it is less than one. But historically, the capital-output ratio and capital's share of income move together, with the latter driving inequality dynamics.* The adjustment of the price of capital to its quantity dynamics is less than one-for-one. And I personally think it's highly unlikely that the rentier will be euthanized any time soon. As my friend, the brilliant Mike Konczal, said, is it really likely that in 30 years, the wealthy will all be lamenting how poor they are now that their wealth has become worthless?

*As Piketty and almost everyone in the IGM poll pointed out, inequality in the US to date has been driven anomalously by inequality in labor incomes. More on that in future posts. (And in past ones.)

So what resolves the Piketty Puzzle? My guess: simply, that macroeconomics based on pricing factors of production according to the representative firm's first order conditions isn't a very good model of reality. Whether we call it K or W, capital or wealth, whatever, as the pile grows, its value grows too. And since wealthy people own the pile, as it grows, inequality increases.

So, among other things, we need a new macroeconomics that doesn't start with first order conditions (even if it does continue on from there). We need to abandon them entirely and get comfortable talking about and modeling power.