"One Discriminatory Rent" or "Double Jeopardy": Multicomponent Negotiation for New Car Purchases by Meghan R. Busse and Jorge M. Silva-Risso. Published in volume 100, issue 2, pages 470-74 of American Economic Review, May 2010

Automobile manufacturers frequently use promotions involving cash incentives. While payments are nominally directed to either customers or dealers, the ultimate beneficiary of the promotion depends on the outcome of price negotiation. We use program evaluation methods to compare the incidence of these two types of promotions. Customers obtain 70 to 90 percent of a customer rebate, but only 30 to 40 percent of a dealer discount promotion, a $500 difference for a typical promotion. Our leading hypothesis is that pass-through rates differ because of information asymmetries: customer rebates are well-publicized to customers, while dealer discount promotions are not.

We investigate whether the first digit of an odometer reading is more salient to consumers than subsequent digits. We find that retail transaction prices and volumes of used vehicles drop discontinuously at 10,000-mile odometer thresholds, echoing effects found in the wholesale market by Lacetera, Pope and Sydnor (2012). Our results reveal that retail consumers devote limited attention to evaluating vehicle mileage, and that this drives effects in the wholesale market. We estimate the inattention parameter implied by the price discontinuities. In addition, our results suggest that estimating consumer-level structural parameters using data from an intermediate market can give misleading results.