Book Description

This dissertation contains four essays at the intersection of econometrics and industrial organization. In all my chapters, I rely on a detailed set of supermarket scanner data on ready-to-eat cereals. In Chapter 1, I examine identification of price effects for differentiated product markets by relying on a conditional form of exogeneity that is an alternative framework to standard instrumental variables. I simulate price changes in the cereal industry arising from potential mergers between firms, one of which took place in 2008. In Chapter 2, I continue to employ conditional exogeneity to identify the effect of market price on demand for differentiated products. The analysis here departs from past studies of demand in several ways, including relaxing the prevalent assumption that observed product characteristics are exogenous. Estimates of implied price-cost margins based on the conditional exogeneity framework are far more reasonable and stable compared to estimates based on standard instrumental variables procedures. In Chapter 3, we (coauthored with Xun Lu) relax the omnipresent assumption that indirect utility takes a linear-separable parametric form in standard logit models of demand. We rely on conditional independence to structurally identify and nonparametrically estimate the average marginal effect of market price on consumer demand. We find that the effect of price on demand is monotonically increasing in price, resulting in high-priced goods having less elastic own price elasticities, and thus higher implied price-cost margins, which addresses a well-known concern in empirical industrial organization. In Chapter 4, I examine a firm's decision to raise price overtly (by increasing the dollar amount of a good) versus a hidden price change (by decreasing the contents in a good's package). I conduct a comprehensive set of empirical analyses in order to assess the impact of hidden price increases on expenditure share and profitability. During July 2007, General Mills decreased the cereal content for 20 out of 23 of their products in my sample of scanner data. A key finding is that consumers tend to notice hidden price changes on smaller-sized boxes of cereal, leading them to substitute to larger-sized boxes of cereal.