Capital is back.” That’s the title of the long paper that first presented the core data and many of the concepts for Piketty’s book — the Grundrisse for his Capital, coauthored with Gabriel Zucman. The title captures perfectly what is so striking about Piketty’s approach to inequality. Conventional treatments have for years focused mainly on inequalities of income from work. The book returns the discussion to capital and labor.

But Piketty doesn’t only focus on the labor-capital distribution of income, since the distribution of capital itself obviously matters. If ownership of capital were spread evenly across the population, these functional shares would hardly matter. In fact, however, capital ownership is highly concentrated, and likely to become more so, since capital breeds capital.

The conventional economic wisdom has long been that the labor-capital split doesn’t matter much because labor and profit shares are constant. According to the Cobb-Douglas aggregate production function, which has been the textbook standard since the 1950s, the parameters are such that an increase in the quantity of one factor relative to the other affects their marginal productivities in just the right way to keep their shares stable. This is extremely convenient, because it keeps the analysis simple, and by coincidence it also seemed to fit the data.