The research funded by this award seeks to narrow the gap between the broad and deep theoretical literature on vertical integration and the more sparse empirical work on the subject. The gap arose in part because detailed microdata on ownership patterns, especially with regard to changes in ownership over time, has been difficult to obtain. The research in this proposal avoids this limitation because the PI has obtained access to Economic Census microdata, which contains the identity of the owning firm for most business establishments in the U.S. This information will be combined with detailed production information from the Census of Manufactures portion of the data to investigate the causes and consequences of integration.
Intellectual Merit of Proposed Activity The proposed research is on three fronts that each intermesh with and expand upon the current literature. The first set of research, on which work has already begun, investigates the predictions of the recent surge in theoretical models demonstrating the harm that can arise through vertical integration's foreclosure, or market power, effect. These predictions are tested using plant-level data from the cement and ready-mixed concrete industries to see if changes in vertical integration are associated with gains in market power. The findings to this point suggest that foreclosure effects in the industries, if they exist at all, are swamped by countervailing influences: rather than increased market power, vertical integration is associated with lower prices, higher quantities, and has no noticeable impact on entry rates. The second set of analyses, also built upon the cement-concrete case study, explores an alternative mechanism-specifically, a link between vertical integration and higher productivity-as an explanation for the counter-foreclosure results. Examples of the questions this research will address include: Is the integration-productivity link observed in the data the result of a particular economy that is harnessed through integration, or is the connection merely a correlation? Do productive integrated firms seek out merger targets that are more or less productive then average? And, Is physical proximity between upstream and downstream plants in integrated firms necessary to obtain productivity gains? The PI will use rich production data to consider these and other questions at an unusually detailed level.
The final set of analyses will look at changes in the broader (industry-level) patterns of vertical integration over a nearly 35-year time span. It trades the case study's ability to hold constant possible confounding influences on vertical structure for the capability of offering more generalized inferences. An important contribution of this broader-scoped research will be construction of a panel data set on the extent of integration within detailed industries (e.g., SIC four-digit) and industry pairs. This panel will be used in tern to see how integration differences are linked to changes in observable industry-level technological (productivity growth, capital intensity, establishment size, etc.) and demand-market (e.g., cyclicality, variability, output substitutability) features.
The research will have a substantial broader impact. The academic literature on vertical integration's welfare impacts has often lead antitrust policy stances toward vertical mergers, and has been explicitly cited by antitrust authorities when announcing decisions on a particular case or broader policy aims. The research set forth in this proposal has the potential to inform the ongoing discussion between researchers and policymakers by providing further insights into vertical integration's driving forces, outcomes, and the mechanisms through which these outcomes obtain. Understanding these underlying processes is important to setting antitrust policies that advance economic efficiency. One of the biggest advantages of the outlined research with regard to improving this understanding is the extraordinarily detailed analysis made possible through Economic Census microdata. This pertains particularly to vertical integration's long-run effects (such as those on productivity growth, entry, and exit) that, while having important welfare consequences and thus being a major focus of policymakers' attentions, have previously been difficult to measure.
