Comment

The paper byCovarrubias, Gutiérrez, and Philippon providesmany useful insights into the rapidly emerging literature on rising concentration in US industries. Importantly, it catalogs some important empirical shortcomings in the literature. It also provides clarity on conceptual issues that have created confusion. The paper goes on to make two types of original empirical contributions. In the first, it focuses on two categories of explanations for rising concentration: “good” and “bad.” The former is associated with technological change that increases the elasticity of substitution among goods (j, and hence greater competition) or increases firms’ accumulation of intangible capital (g, perhaps associated with network externalities and returns to scale). Bad concentration is instead associated with rising barriers to entry, k. The authors argue that the data tend to favor good concentration earlier in their data sample—through 2000. Thereafter, there is increasing evidence of barriers to competition. The collection of evidence, while not dispositive, moves the weight of the evidence toward market power explanations, especially later in the sample. The last part of the paper takes a different tack. Instead of looking for indicators ofmarket power to explain a broad range of facts, the paper looks at combinations of explanations by industry and argues that there is merit in several of them and that the results vary by industry. I will argue that these last insights are especially helpful, as the macroeconomic data are unlikely to be captured by a single simple narrative.