Auctions offer a unique and visceral way to explore how a market functions, but the rapid-fire pace can make them difficult to follow, much less appreciate the nuances and complexities of the process. Take, for example, the following video, which Chicago Booth Associate Professor Devin Pope used to open up his Becker Brown Bag lunchtime lecture on October 14:
Most research has focused on auction design, considering how structural characteristics, like the number of bidders or whether bids are sealed or open, affect outcomes, while assuming no role for the auctioneer. But after speaking with auctioneers and witnessing the control that they exercise over an auction’s pace, Pope sought to examine what effect, if any, the men on the podium have on sales.
With his coauthors Nicola Lacetera (University of Toronto), Bradley Larsen (Stanford University), and Justin Sydnor (University of Wisconsin-Madison), Pope sifted through  a trove of data gathered from a large wholesale auto auction house in Manheim, PA. Each day at the auction house, thousands of used cars are bought and sold by dealers. The average car sells in one minute and 20 seconds, but not all cars sell and those that due sell for widely different prices. For the auction house, the primary concern is the number of cars that sell, the conversion rate. Auctioneers, who see their role as creating a sense of excitement, urgency, and competition among buyers, are thus judged on their ability to make sales.
Pope’s team began by calculating the conversion rates of Manheim’s 60 full-time auctioneers. Their initial analysis found a difference of roughly 30 percent between the worst- and best-performing auctioneers. That difference is stark, but couldn’t be taken at face value because vehicle assignment among auctioneers is not entirely random. Some dealers request specific auctioneers, raising the possibility that favored auctioneers’ performances may be aided by being reliably supplied more desirable vehicles. So Pope’s team began controlling for as many factors as possible: time of day, car type, dealership, and even lane within the auction house. They also compared auctioneers’ performances over the three years of data they were able to collect.
Controlling for these factors reduced but did not eliminate the performance differentials. Between the 90th and 10th percentile of auctioneers a 5.6 percent difference in conversion rate remained, which translates to approximately 123 cars sold each year and $24,600 in fees for the auction house. The robust results suggest that auctioneers do more than simply call prices and take bids, but they don’t pinpoint the mechanism by which auctioneers affect outcomes.
In interviews, auctioneers state that their success comes from an ability to play off bidders’ competitive instincts and to sustain a momentum that keeps bidders excited and engaged. Indeed, when the team asked the auction house to subjectively rank their auctioneers they found that the subjective rankings strongly correlated with the data driven rankings they had generated, lending credence to the auctioneers’ claims.
For Pope and his team, their now-refined notion of the role that auctioneers play has opened  up avenues of inquiry for future work, including examining why auctions have not moved entirely online, how online auctions could incorporate the excitement prevalent in physical auctions, and how field experiments could more precisely identify the mechanisms of the auctioneers’ impact.

–Jeffrey Lamoureux