Asymmetric Price Adjustment in the Small: An Implication of Rational Inattention

TLDR: In this article, a large supermarket chain in the US showed that small price increases occur more frequently than small price decreases, and that consumers may choose to ignore or not to respond to small price changes.

Abstract: Analyzing scanner price data that cover 27 product categories over an eight-year period from a large Mid-western supermarket chain, we uncover a surprising regularity in the data—small price increases occur more frequently than small price decreases. We find that this asymmetry holds for price changes of up to about 10 cents, on average. The asymmetry disappears for larger price changes. We document this finding for the entire data set, as well as for individual product categories. Further, we find that the asymmetry holds even after excluding from the data the observations pertaining to inflationary periods, and after allowing for various lengths of lagged price adjustment. The findings are insensitive also to the measure of price level used to measure inflation (the PPI or the CPI). To explain these findings, we extend the implications of the literature on rational inattention to individual price dynamics. Specifically, we argue that processing and reacting to price change information is a costly activity. An important implication of rational inattention is that consumers may rationally choose to ignore—and thus not to respond to—small price changes, creating a “range of inattention” along the demand curve. This range of consumer inattention, we argue, gives the retailers incentive for asymmetric price adjustment “in the small.” These incentives, however, disappear for large price changes, because large price changes are processed by consumers and therefore trigger their response. Thus, no asymmetry is observed “in the large.” An additional implication of rational inattention is that the extent of the asymmetry found “in the small” might vary over the business cycle: it might diminish during recessions and strengthen during expansions. We find that the data are indeed consistent with these predictions. An added contribution of the paper is that our theory may offer a possible explanation for the presence of small price changes, which has been a long-standing puzzle in the literature.