Algorithmic and Human Collusion
As self-learning pricing algorithms become popular, there are growing concerns among academics and regulators that algorithms could learn to collude tacitly on non-competitive prices and thereby harm competition. I study popular reinforcement learning algorithms and show that they develop collusive behavior in a simulated market environment. To derive a counterfactual that resembles traditional tacit collusion, I conduct market experiments with human participants in the same environment. Across different treatments, I vary the market size and the number of firms that use a self-learned pricing algorithm. I provide evidence that oligopoly markets can become more collusive if algorithms make pricing decisions instead of humans. In two-firm markets, market prices are weakly increasing in the number of algorithms in the market. In three-firm markets, algorithms weaken competition if most firms use an algorithm and human sellers are inexperienced.
Volunteering at the Workplace under Incomplete Information: Team Size Does Not Matter (with Adrian Hillenbrand and Fabian Winter)
Volunteering is a widespread allocation mechanism at the workplace and emerges naturally in open-source software development, the generation of online knowledge platforms, and to some extent in “agile” work environments. Using a field experiment with 8 treatments and close to 2,800 workers on an online labor market, we study the effect of team size on volunteering at the workplace under incomplete information. In stark contrast to the theoretical predictions, we find no effect of team size on volunteering behavior. With the use of our control treatments, we can show that workers react to free-riding incentives provided by the volunteering setting in general, but do not react strategically to the team size. We show that the result is robust to several further factors.
[PDF] [Working on new treatements]
Algorithmic Price Recommendations and Collusion (with Matthias Hunold)
Draft in preparation.
Gaming or Gambling? On Selection Neglect and Loot Boxes (with Simon Cordes, Markus Dertwinkel-Kalt and Mats Köster)
Draft in preparation.
Memory Length in Algorithmic Collusion (with Bernhard Kasberger, Simon Martin and Hans-Theo Normann)
Work in progress