Methodological Contributions of Pikettyâs Capital
When discussing Capital in the Twenty-First Century, we need to distinguish between its analytics and methodology, its recommendations, and its forecasts. One can agree with the analytics without agreeing with the recommendations, or the reverse. The methodology introduced by Capitalâbecause it seems to fit quite well the likely evolution of the rich world in the decades to come, and more importantly because it provides a novel way to look at economic phenomenaâis probably the most significant contribution of the book. It will affect not only how we think of income distribution and capitalism in the future but also how we think about economic history, from ancient Rome to prerevolutionary France.
The most important methodological contribution of Pikettyâs book is his attempt to unify the fields of economic growth, functional income distribution, and personal income distribution.1 In the standard Walrasian system, the three are formally relatedâbut in actual work in economics they were generally treated separately, or some were even simply left out. Functional income distribution was studied much more by Marxist economists. Neoclassical economists tended to assume that capital and labor shares were broadly fixed. This view changed only fairly recently, and we are now witnessing an upsurge of interest in the topic.2 Pikettyâs emphasis on the rising share of capital income contributed to this efflorescence.
Personal income distribution tended to be studied almost as divorced from the rest of economics, because in a Walrasian world, agents come to the market with the already-given endowments of capital and labor. Because the original distribution of these endowments is not the subject of economics (narrowly defined), personal income distribution was assumed to be whatever the market generates. But in Capital, the movements in the capital / income ratioâdriven by âthe fundamental inequalityâ or âcentral contradiction of capitalism,â namely r > g (return on capital greater than the growth rate of overall income)âlead to the rising share of capital income in net product. This, in turn, leads to a greater interpersonal inequality.
This chapter concentrates on the last point, which is usually implicitly taken for granted: A greater share of capital is associated, it is thought almost implicitly, with a rising interpersonal inequality. This view is understandable because during most of economic history, people with high capital income were also people with high overall income. Therefore, a greater share of net product going to capitalists came to be associated with greater interpersonal inequality.
In a recent paper investigating the association between higher capital shares and income inequality over the long run (going back in some cases to the mid-nineteenth century), Erik Bengtsson and Daniel Waldenström find, in a country fixed-effect setting, that the correlation has typically been positive and fairly strong. For the entire sample of fifteen advanced economies, they find that, on average, each percentage-point increase in the capital share was associated with a 0.89-point increase in the (log of) top 1 percent income share. When other controls are introduced, the size of the coefficient is reduced, but it remains positive and statistically significant.3 Margaret Jacobson and Filippo Occhino similarly find that for the United States, a 1 percent increase in the capital share tended to increase Gini by between 0.15 percent and 0.33 percent.4
Maura Francese and Carlos Mulas-Granados use more recent 1970sâ2010 Luxembourg Income Study microdata from forty-three countries, and decompose the overall change in disposable income Gini into its accounting components: concentration coefficients of labor and capital, labor and capital shares, and changes in taxation and social transfers.5 Unlike Bengtsson and Waldenström, they find a negligible impact of higher capital share, and conclude that most of the increase in disposable income Gini was driven by the rising concentration of wages. They complement the decomposition analysis by a regression on a sample of ninety-three countries, for the 1970sâ2013 period, of capital (labor) share on Gini. Once controls are introduced, labor (capital) share is insignificant.6
So the link between greater capital share and increased interpersonal inequality is not as simple and unambiguous as it seems. Even when there exists a positive relationship between the two, the strength of that relationship varies.
The chapter is organized as follows. In the next section I discuss in general the link between the rising share of capital in net income (Pikettyâs α) and the Gini coefficient of interpersonal income inequality. Next I look at this relationship in three ideal-typical societies: socialist, classical capitalist, and ânewâ capitalist....