UC Berkeley economics professor Dr. Steve Tadelis, gave a keynote presentation at an event hosted by the Oeconomia club at the Upper School during long lunch on Friday in the Nichols Auditorium. While Harker events have featured speakers from the local Stanford University and Santa Clara University, Tadelis is the first Berkeley economics professor to come to the upper campus.

Tadelis is currently on leave from his teaching post at UC Berkeley’s Haas School of Business, working in the corporate world with Amazon, where he serves as Vice President of Economics and Market Design at Amazon.

Opening his presentation titled “Economics in the Real World”, Tadelis made an important distinction between micro and macro economics, defining the difference between an economist specializing in each. He explained that a micro-economist analyzes individual decisions and factors in an economy, on the other hand, for macroeconomists, he jokingly remarked that he has “no idea” what they do.

For the majority of his presentation, Tadelis talked about his microeconomic work on Ebay’s web advertising strategies. He outlined an experiment in which he and his team were to determine the effectiveness of the company’s web advertisements–figuring out which channels brought customers to the site and the size of the purchases coming from each channel.

“[Marketers] try to connect data that tells them, what is the channel that someone came to eBay to buy something versus other channels, how much did they spend coming out of that channel,” he said. “I have to ask myself as an economist, what is the behavior that creates this data. People who click on an Ebay ad buy more than people who arrive from other channels.”

The results of the research were contradictory to the hypothesis made by Ebay’s marketing professionals.Tadelis showed a graph that represented total sales with advertising and without. Surprisingly, it turned out that sales remained the same despite the fact that advertising was cut, a fact made true by Ebay inadvertently targeting people who already had the intention of shopping on their site. Tadelis joked that he wished he could take a fixed rate from the money Ebay saved as a result of his team’s research, a number he estimated to be around $20 million annually.

The second section of the keynote was a discussion of price salience, a concept which concerns the perceived or immediately apparent price of a good. Tadelis brought up the example of StubHub, where at checkout, both the price of the ticket and associated fees are added together. He explained that StubHub lost sales to competitors, who used a strategy of showing only ticket price without the addition of the processing and service fees.

He then analyzed the different approaches of analysis for behavioral economics, advocating  . prescriptive analytics–a “question driven” approach with a “focus on causation” and “broad management and strategy.”

Ending his experimental analysis, Tadelis spoke about reasons why students should consider studying economics at a high level. He began by noting the growing role of PHD economists at large companies, citing Valley behemoths like Facebook, Uber, Ebay, and Amazon as examples.

He advised students to refrain from focusing solely on economics in college, telling the audience to pair the study with another skill such as programming, statistics, or advanced math to maximize and diversify intellectual application.

Sahil Kapur (11), an AP Economics student who attended the event, commented on his motivation for attending and what he learned from the presentation.

“I originally went for the Tadelis presentation because I really wanted to hear more about game theory, especially since he was a game theory professor,” he said. “I learned more about the real world views cases about economics beyond the classroom and beyond the examples we do in class. Especially how it can be applied to real world stuff today like such as online advertising.”

Tadelis’ presentation concluded with a short Q&A session. Students asked a range of questions from the problem of ad-blockers for advertising and the potential application of artificial intelligence in understanding behavioral economics.

To learn more about Dr. Tadelis and his research, visit http://faculty.haas.berkeley.edu/stadelis/.