Some auctioneers consistently achieve better results than others—suggesting that the process by which an auction is conducted can have a significant effect on outcomes, according to research by Chicago Booth Professor Devin G. Pope, together with Nicola Lacetera of the University of Toronto and the National Bureau of Economic Research, Bradley J. Larsen of Stanford University and eBay Research Labs, and Justin R. Sydnor of the University of Wisconsin.
The authors analyzed data from about 850,000 cars auctioned by 60 auctioneers at a leading used-car-auction operator between 2007 and 2013, assessing performance using conversion rates, the fraction of cars auctioned that were successfully sold. While the auction company reported the conversion rate as its key metric, the authors also analyzed secondary performance metrics including speed of sale and residual price, which is the gap between sales price and average market value.
The average sales probability in the data is 53%, the average sales price, $15,141, and the average length of a sale, 103 seconds. But after controlling for factors such as auction day, sellers’ reservation values, and car type, the researchers find that a one-standard-deviation increase in auctioneer performance corresponded to a 2.3-percentage-point increase in likelihood of a sale, a $41.80 increase in residual price, and a 6.1-second increase in speed of sale. This impact is comparable to that found in studies estimating the effects of information dissemination, changes in auction structure, and broader macroeconomic factors, suggesting auctioneers’ performance impacts outcomes as much as any of those theoretical benchmarks.
To test the robustness of their findings, Pope and his colleagues addressed whether the gap in performance between auctioneers was consistent over time and across metrics, as well as whether this discrepancy was corroborated by feedback from the auction house. Comparing 2007–09 data with 2010–13, the authors find that the top performers were the same. Secondary performance metrics show that auctioneers with the highest conversion rates also tended to achieve higher prices and move cars faster, dispelling the possibility that these auctioneers were excelling by one metric to the detriment of others. Finally, the authors find that the auctioneers laid off during a period of downsizing were the underperformers, indicating that the research findings were consistent with the auction house’s assessment of its employees.
What separates the best auctioneers from the rest? The authors conclude that auctioneers do not convey or even have information about car quality, and find little evidence that successful auctioneers persuade sellers to accept auction outcomes. Rather, the best auctioneers are better at “creating excitement and competitive arousal among buyers” by using the pace of their speech and inflection of their voices.
The role of auctioneers has been largely overlooked by economic literature, which has focused instead on auction structure, leaving out the role of the auctioneer in any theoretical framework. However, Pope and his colleagues find that auctioneers are “a fundamental feature of well-established and well-functioning auction environments.” This suggests that the traditional framework for analyzing auctions fails to capture important real-world dynamics, and it lends support to ongoing initiatives to expand auction modeling. That could enable economics to catch up to what used-car salesmen already know.