A Non-Price-Discrimination Theory of Rebates∗

According to the conventional theory, rebates are but one of many ways to price discriminate among consumers. The firm sells at the shelf price to those willing to pay more for the product and at a lower, rebated price, to those willing to pay less, achieving the necessary separation by relying on the higher costs of redeeming rebates of high willingness-to-pay consumers. This theory, however, does not explain several phenomena associated with rebates. Many consumers respond to rebates at the time of purchase, but far fewer redeem them after purchase. Without resorting to irrationality on the part of some consumers, this redemption gap cannot be explained by price discrimination theory. We also see free-after-rebate deals. This practice is also puzzling from the perspective of price discrimination, because a firm will surely lose money from rebate users. Finally, the usage of rebates tends to vary across industries and product categories even when the same consumers participate in each of them. We propose a non-price-discrimination theory of rebates based on the idea of “utility arbitrage.” Key to our theory is a distinctive property of rebates: unlike other promotion vehicles, they are redeemed after purchase and during consumption. When a product is not consumed instantaneously, the utility that a consumer derives from it will depend on future states of nature. A firm can increase a consumer’s willingness to pay for the product by simultaneously increasing the price she pays in a state of low marginal utility of income and decreasing the price in a state of high marginal utility of income. Rebates help implement utility arbitrage by delivering state-dependent discounts. Thus, the paper identifies a unique role for rebates in the marketing tool kit. Notwithstanding this role, rebates can still price discriminate among consumers with different willingness to pay. Indeed, their ability to utility arbitrage may enhance their price discrimination ability. This suggests that, abstracting from redemption cost differences, rebates are likely to be superior to coupons.