Research Papers
RAND Journal of Economics, forthcoming
Abstract: This article considers an auction model in which a seller’s choice of reserve price signals her private information about the object’s quality. We show that the signaling incentive would lower the seller's payoff and the probability of sale. We estimate the model using a novel dataset from a large online auto auction platform. Counterfactual simulations suggest that a secret reserve price could shut down the signaling incentive and improve both the seller’s payoff and the probability of sale, which supports the prevalent use of secret reserve prices in practice.
Under Review
Abstract: Online auctions have declined in popularity, often due to their high transaction costs. This paper focuses on one such cost: the hassle cost of bidding in often prolonged online auctions. We develop a structural model of ascending auctions with costly bidding that resembles a war of attrition. We show that even in private value settings, bidders strategically lower their bids to avoid hassle. Using data from a large online auto auction platform, we estimate that the average winner incurs hassle costs equivalent to 1.7% of the vehicle's sale price, while sellers lose 5.0% in revenue due to bid-shading. Counterfactual simulations suggest that higher reserve prices can limit unproductive bidding and partially offset the loss.
Abstract: In many settings, behavioral economists have documented a price reference effect: the fact that a consumer's willingness to pay for a good is affected by difference between the observed price and the reference price they rationally expect. In this paper, we first show theoretically that when this price reference effect is sufficiently large, it can overturn the standard textbook result that vertical integration improves joint profits. The key intuition is that the increase in quantity is dampened when consumers update their expectations. To test whether this force is large in a real-world setting, we develop a model of a downstream retailer who faces behavioral consumers and bargains with an upstream producer. We estimate this model using a novel dataset from a large online book retailer, where we observed retail prices, quantities sold and wholesale prices. Counterfactual simulations show that vertical integration would reduce joint profits by 11%. These findings highlight the importance of incorporating consumer expectations in the analysis of optimal pricing and vertical integration.
Abstract: Subsidy programs are widely used by governments to stimulate activities in the durable goods sectors. This paper investigates the impact of government subsidies on consumers' automobile replacement decisions in the context of the U.S. Car Allowance Rebate System program in 2009. I develop and estimate a dynamic discrete-choice model of automobile replacement, and conduct counterfactual analysis to evaluate the effects of government subsidies. Results show that 65% of the households would replace their vehicles even without the subsidy, and the reductions in gasoline consumption and carbon emission are limited. To highlight the impact of subsidy design, I show that limiting the subsidies to low-income consumers would generate 85% of the sales with half amount of total government spending. These findings emphasize the importance of balancing policy stimulus with government spending and targeting consumers more efficiently in the design of subsidy programs.