Prices are Endogenous. Evidence from a Large-Scale Pricing Experiment.
Abstract
We present experimental evidence that suggests that all meaningful variation in grocery prices that could be used to estimate demand elasticities is tainted by endogeneity. Our data stem from a Midwestern grocery retail chain that conducted a pricing experiment, administering 175,707 different random prices over the course of 11 weeks, across 736 products at 82 “treated” supermarkets. We compare the demand elasticity estimates derived with these experimental prices against the elasticity estimates derived with the conventional prices set at the 34 “control” supermarkets. We find customers more sensitive to the control stores’ price changes. Specifically, our demand elasticity estimates for the butter spreads, candy, shredded cheese, sliced cheese, snacking cheese, other cheese, facial tissues, pasta, and pasta sauce categories are -1.71, -0.71, -1.62, -1.75, -1.63, -1.61, -1.83, -1.10, and -1.64 at the control stores and are -0.49, -0.36, -0.79, -0.17, -0.44, -0.44, -0.42, -0.01, and -0.26 at the treated stores. The t-statistics of the difference are 10.0, 9.7, 13.3, 17.1, 12.5, 18.1, 12.5, 15.8, and 18.0, respectively, and the effect is even stronger in the difference-in-differences. We cannot reconcile the gap between the experimental and non-experimental elasticity estimates under any specification (although factoring temporary price reductions closes much of it).
Bio
Robert Bray is an associate professor of operations management at the Kellogg School of Management, at Northwestern University. He received his B.S. in industrial engineering, operations research from UC Berkeley in 2006, and his Ph.D. in business administration from the Stanford Graduate School of Business in 2012. Most of his work pertains to empirical supply chain management, but he's also written a few theoretical dynamic programming articles. He grew up in Palos Verdes, CA.
