A Bayesian Latent Variable Mixture Model for Filtering Firm Profit Rate

February 9, 2014

This paper examines the distribution of profit rates in the United States using firm-level data to examine the theory of the equalization of the rate of profit.

Since Adam Smith, classical political economy has theorized that through competition, capital would move from one sector to another in an endless search for higher rates of profit. From this movement, profit rates across all industries would tend to one general rate, the equalization of profit rates. To investigate this claim, the authors developed a Bayesian statistical model that captures this competitive process of the equalization of profit rates as depicted by the distribution of profit rates of firms and applied it to a dataset of U.S. stock-market listed firms from 1950-2012. The major findings are: profit rates are sharply peaked around a general rate of profit at both an economy-wide and industry level; this general rate of profit is roughly 10% throughout their sample, although the average rate of profit appears to be falling over time; and larger firms exhibit a "statistical equilibrium," in that the distribution of profit rates appears to be stationary, having the same characteristics across time.

Authors: Gregor Semieniuk and Ellis Scharfenaker
Download PDF