Multi-Category Competition and Market Power: A Model of Supermarket Pricing

Abstract

In many competitive settings consumers buy multiple product categories, and some prefer to use a single firm, generating complementary cross-category price effects. To study pricing in supermarkets, an organizational form where these effects are internalized, we develop a multi-category multi-seller demand model and estimate it using UK consumer data. This class of model is used widely in theoretical analysis of retail pricing. We quantify crosscategory pricing effects and find that internalizing them substantially reduces market power.We find that consumers inclined to one-stop (rather than multi-stop) shopping have a greater pro-competitive impact because they generate relatively large cross-category effects.

Estimating the Demand for Service Bundles under Three-Part Tariffs

“…In particular, we transform the consumer's original non-differentiable maximization problem into a list of problems that are differentiable and have explicit solutions. This is similar to the method in Thomassen et al [2017]. After doing so, we then simulate each consumer's optimal plan choice and usage decision, given a set of model parameters.…”

“…15 Our approach to solving the usage decision problem is similar to the method in Thomassen et al [2017] except that we have four scenarios for each service leading to 4 K combinations, compared with 2 K in their paper, where K is the number of products.…”

“…This paper provides a framework for estimating consumer preferences over service bundles, where temporal separation between service plan choice and the usage decision is present. This paper is also related to the literature on identifying and estimating the demand for multiple products (e.g., Li et al [2005]; Manchanda et al [1999]; Gentzkow [2007]; Lee et al [2013]; Liu et al [2010]; Thomassen et al [2017]). Gentzkow [2007], for example, studies the joint consumption of print and online newspapers and the relationship between them.…”

Measuring competition in spatial retail

“…10 The median number of stores within 10 miles of given tract is 41. In their analysis of the UK grocery market, Thomassen et al (2017) assume that the choice set is the nearest 30 stores, noting that the fraction of expenditures outside that set is only 1.2%. 11 In our setting, the presence of the logit shock could add an element of misspecification if D is set too large, because all outlets in the choice set will attract at least some consumers.…”

“…1 A primary motivation for pursuing this alternative source of variation is the challenge of obtaining price data for all outlets. However, the widespread practice of uniform grocery pricing (e.g., Thomassen et al, 2017) suggests that price effects may not be well identified from cross-sectional variation alone, even if prices were observed. In this sense, our approach complements Conlon and Mortimer (2013), who use variation in the availability of products in lieu of price changes to measure diversion 2 Holmes' analysis of Wal-Mart abstracted from competition to focus on how the dynamics of Wal-Mart's expansion decisions were influenced by the scale economies associated with operating a dense network of stores.…”