Johannes Leutgeb

Wissenschaftszentrum Berlin für Sozialforschung

Reichpietschufer 50

10785 Berlin, Germany

I am a research fellow at WZB in Berlin and I am interested in behavioral and experimental economics. Currently, I am working on evolution of overconfidence using long-run interactions in high frequency environments, models of (political) influence and AI-human interactions. I am teaching introductory classes to programming experiments in oTree. Also, I dabble in 3D printing of representations of my data.

recent publications

2023

Learning through period and physical time

Teresa Backhaus, Steffen Huck, Johannes Leutgeb, and Ryan Oprea

Games and Economic Behavior, Sep 2023

Abstract Link

We demonstrate in a laboratory experiment in which subjects play a two-player Cournot-Tullock game over hundreds of periods of varying length that full accounts of subjects’ learning requires the consideration of, both, ‘period time’ and ‘physical time.’
How Trump triumphed: Multi-candidate primaries with buffoons

Micael Castanheira, Steffen Huck, Johannes Leutgeb, and Andrew Schotter

European Economic Review, Aug 2023

Abstract Link

While people on all sides of the political spectrum were amazed that Donald Trump won the Republican nomination this paper theorizes that Trump’s victory was not a crazy event but rather the equilibrium outcome of a multi-candidate race where one candidate, the buffoon, is viewed as likely to self-destruct and hence unworthy of attack. Our model conceptualizes primaries as a truel (a three-way duel); we solve for its equilibrium, and test its implications in a laboratory experiment. We find that people recognize a buffoon when they see one and aim their attacks elsewhere with the unfortunate consequence that the buffoon has an enhanced probability of winning. This result is strongest amongst those subjects who demonstrate an ability to best respond, suggesting that our results would only be stronger when the game is played by experts and for higher stakes.