This paper studies pricing in the fashion retail industry. Online data was collected for approximately 350,000 distinct products from over 65 retailers in the U.S. and the U.K. We present evidence that a fair fraction of retailers implement an extreme form of price stickiness that we describe as quantum prices: a large number of different products are priced using just a small number of sparse prices, with price changes occurring rarely and in large increments. Normalized price clustering measures are used to show that retailers use quantum prices within- and across- categories, and this clustering is not explained by popular prices, ranges of prices, assortment size, or digit endings. This pricing strategy is consistent with a behavioral model where fewer prices makes price advertising more effective. An implication of this model is that advertising is increasingly effective when the same prices are used across product lines, i.e. for new products. Finally, quantum prices affect product introductions and price adjustment strategies at the firm level, while it creates larger deviations of the law of one price and hinders the computation of inflation at the macro level.
We are grateful to Glenn Ellison, Duncan Simester, Alberto Cavallo, and Ignacio Puente for helpful discussions. We also thank Mariana Albano, Manuel Bertolotto, Vivek Bhattacharya, Sara Ellison, Miguel Espinosa, Ricardo Fraiman, Renee Gosline, Paul Joscow, Chris Knittel, Carolina Levinton, Juan Cristobal Lopez, Victor Martinez de Albeniz, John McCoy, Mateo Montenegro, Chela Pedreschi, Pilar Posse, Drazen Prelec, Nancy Rose, Nader Tavassoli, Julian Villanueva, Birger Wernerfelt, Mike Whinston, seminar participants at MIT and at the Golub Center for Finance and Policy, and pricing managers in apparel, for helpful comments. Nicolas Gomez del Campo and Gina Rey provided outstanding research assistance. All errors are our own. Data and codes will be made publicly available. We welcome comments via email. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.