Comment

Authors Rossi-Hansberg, Sarte, and Trachter have delivered a focused, definitive paper, part of a broader research program on structural change in US product and labor markets. The subject of the paper is concentration. Measurement of concentration starts from the market definition. Concentration depends on the pattern of market shares within a market, defined by a product, a geographic region, and a time period. Amarket is a specification of the product, region, and time such that a hypothetical monopolist would have market power—it could elevate its price substantially abovemarginal cost. The connection to concentration is the general belief that the actual market power of the firms selling in a market depends on the number and importance of the competitors selling in that market. Themost widely usedmeasure of concentration—and the one used in this paper—is the Herfindahl-Hirschman Index (HHI), the sum of the squared market shares of the sellers in a market. This measure is sensitive to both the number of sellers and their market shares, and it is a sufficient statistic for the elevation of price over marginal cost in certain oligopoly models. However, other models, especially those involving collusion, predict substantial exercise ofmarket powerwithpricesmarked up well above marginal cost, in the face of HHI measures suggesting rather less price elevation. Some markets are national or international. The products in these markets have low transport cost across the country relative to price. A substantial body of research has found high and rising concentration in many national markets, primarily for manufactured goods. The paper confirms that finding. The paper does not pursue the key question of