The fall of the labor share of GDP in the US and many other countries has been well documented, but its causes remain controversial in part because empirical assessments have relied on industry or macro data. In this paper we use firm and establishment panel data, focusing on the US Economic Census since 1982 (but also drawing on international micro data) to describe some stylized facts that seem consistent with a new explanation of the fall in the labor share based on “superstar firms." If globalization or technology mean that product markets are increasingly “winner takes most," industries will become increasingly dominated by superstar firms with high profit margins. As the importance of such firms increases, the aggregate labor share will fall. The model has a number of predictions. First, industry sales will be increasingly concentrated in a small number of firms. Second, those industries where concentration rises the most should have the largest falls in the labor share. Third, the fall in the labor share should have a large between firm component (rather than being primarily a within-firm phenomenon, where the labor share falls similarly across most firms). Fourth, this between firm fall in the labor share should again be greatest in the sectors that are concentrating the most. Fifth, these patterns should be observed not only in US firms, but also internationally. We find support for all these predictions of the superstar firm model, particularly in sectors where technical change is more rapid, as measured by the growth of TFP or patent intensity.
Concentrating on the Fall of the Labor Share
Submitted by Staff on January 05, 2017
|Date: December 1, 2016|
|Author(s): David Autor, David Dorn, Lawrence Katz, Christina Patterson, John Van Reenen,|
|Affiliation: MIT - University of Zurich - Harvard University|