Why Economists are Wrong to Neglect Retailing and How Steiner’s Theory Provides an Explanation of Important Regularities

Despite the fact that almost a third of every dollar spend by consumers goes to pay retail and wholesale distributor margins, and that one of every six non-farm workers is employed in the distributive trades, the economics profession by-and-large ignores them. Economic models usually assume that producers sell directly to consumers; Steiner labels these “single stage models.” In the rare instances where the omission of retailing is acknowledged, the justification offered is that retailing is “competitive” and so the derived demand theorem assures that retail prices will faithfully reflect changes in producer prices. Steiner provides evidence, based on his own business experience and on extensive empirical evidence, that retail price changes often do not simply reflect producer price changes. Steiner has a striking example from his own experience; early TV advertising of a Kenner toy led to an unexpected reduction of retailer margins from 50% to 33%. At an unchanged factory price, the retail price fell by 25%. Steiner, among others, has shown that this is not an isolated example. The prevalence of “Hi-Lo” pricing by many national retailers shows that a disconnection between producer and retail prices is not a rare event. Steiner has pointed to several strong empirical regularities that, though inconsistent with a perfectly competitive model of retailing, offer substantial hope for understanding the complexities of retailing. The first “inverse association” is that the more prominent the brand, the lower its retail margin; the second is that the more prominent the brand, the higher is its manufacturing margin. Evidence for and against these associations is summarized in the paper. I introduce a formal model of retailing that is complex enough to capture some of Steiner's main ideas, yet simple enough to yield some quantitative predictions. The model predicts that the profit maximizing price for a store brand competing with a national brand will equal the retail price of leading brand minus one half of the difference between producer prices that the retailer pays for the two products. A test of this prediction using soft drink data from a Chicago area supermarket chain shows that actual store brand prices differ by less than 5% on average