A Large Discrepancy Between Price Elasticities Estimated with Experimental Price Data and Observational Scanner Data

We present experimental evidence that the observational price variation in typical supermarket scanner data is insufficient to recover true price elasticities. We analyze a large, brick-and-mortar field experiment that generated 389,890 random, in-store prices over 35 weeks, across 409 products, at 82 “test” stores of a Midwestern grocery retailer. We compare the demand elasticity estimates derived from these experimental price changes against those derived from observational price changes at 34 “control” stores. During the experiment, the average experimental elasticity is -0.34, whereas the average observational elasticity is about -2.0. The gap is even wider in the difference in differences, and replicates at the category and the product levels. We cannot reconcile this gap by controlling for promotions, conducting an event study around each price change, focusing on base-price changes, accounting for longer-term price effects, or instrumenting with the chain.