Bounds Approaches to Empirical Market Design

Project Outcomes Statement

Through this project, the researchers developed several new methods for studying large datasets from real-world auction and bargaining settings, resulting in several new findings for empirical economics. First, the researchers derived a method for inferring the willingness to pay of auction participants from online auction settings, such as eBay. The researchers applied this method to data from eBay auctions for used smartphones to study the effects of a 2013 change in digital copyright law. The researchers found that banning users from unlocking their own smartphones from the cellular carrier led to a decrease in consumers’ willingness to pay, suggesting that consumers react to changes in digital copyright law. This research is under revision for Quantitative Economics, a peer-reviewed journal.

Second, the researchers developed a methodology for inferring buyers’ and sellers’ willingness to pay from alternating-offer bargaining data. The researchers applied this methodology to a large dataset of business-to-business used-car transactions containing all back-and-forth actions taken by the negotiating parties. The researchers found that real-world bargaining was inefficient: over half of the cases where a buyer and seller failed to reach agreement, the buyer actually valued the car more than the seller (and hence trade should have occurred, but did not), suggesting that existing models of bargaining—used in courts and by government agencies—that ignore inefficiencies may be inaccurate. This bargaining research is forthcoming in the Review of Economic Studies, a top tier, peer-reviewed journal. The researchers also extended this methodology to apply it to bargaining data from eBay.com.

The researchers have presented the results of this work at numerous seminars and conferences throughout the United States and abroad. The project helped in the creation of large datasets on real-world bargaining, which have been publicly released on the NBER website for other researchers to study. Several PhD student research assistants received mentoring and training while working on the projects related to this grant, some of whom have now moved on to assistant professor positions at high-ranking business schools.