Inventory Record Inaccuracy Explains Price Rigidity in Perishable Groceries

Abstract

We propose a simple explanation for price rigidity in perishable groceries: inventory record inaccuracy (IRI). We build our argument in two steps. First, we tailor Gallego and Van Ryzin’s (1994) revenue management model to perishable groceries by adding an inventory waste process, deteriorating product quality, and menu costs. Comparing the model's optimal time-dependent pricing policy (i.e., when prices do not condition on real-time inventories—which corresponds to IRI) with its optimal state-dependent policy (i.e., when prices do condition on real-time inventories—which corresponds to no IRI), we demonstrate that IRI can explain about 90% of the menu costs required to rationalize any given level of price rigidity. Moreover, based on the model, we predict that reducing menu costs from 10% of marginal cost to 1% would increase a representative product's price-updating frequency without IRI by at least three times more than it would with IRI. Second, we test our prediction with data from two case studies: a U.K. grocery store that installed electronic shelf labels (ESLs)—a technology that reduces physical menu costs—and an E.U. grocery store that installed both ESLs and GS1 barcodes—the latter being a technology that reduces IRI. Consistent with our prediction, the technology upgrade increased price-updating frequency by 55% in the first case study and by 853% in the second study.