Algorithmic Collusion: Working Mechanisms

This hybrid conference will take place in the Hybrid Learning Theatre and explore recent academic research in Economics and Operations Research on algorithmic collusion. What does the literature say? What are the mechanisms by which algorithmic collusion works? What empirical evidence is there of algorithmic collusion? And what can policy-makers or market designers do about it?

There has been a lot of academic debate in the area of pricing algorithms and competition and the extent to which algorithmic collusion presents an actual threat has not yet been demonstrated. In particular, a key question is to what degree and in which settings algorithms can sustain coordinated prices over competitive levels. Theoretical, computational and empirical papers in are showing that various forms of supra-competitive pricing may indeed emerge through the use of algorithmic pricing, sometimes through the use or adaptation of collusive strategies. This virtual conference will host several speakers working on novel projects related to the concerns around pricing algorithms and collusion.

Organisers: Research fellow Maarten Pieter Schinkel (University of Amsterdam) and alumnus Timo Klein (Utrecht University and Oxera Consulting)
Location: Hybrid Learning Theatre, University of Amsterdam (physically accessible, via Zoom and YouTube Live with chat).

Registration for attendance via zoom is almost fully booked, but you can join via YouTube Live and allows chat as well.

Please find Live to YouTube live stream on the conference webpage: ase.uva.nl/content/events/conferences/2020/12/algorithmic-collusion-working-mechanisms.html

Programme

• 14:00 - Timo Klein (Utrecht University and Oxera Consulting) – Moderator: ‘Introductory remarks: Taking stock of the academic literature so far’
• 14:20 - Arnoud den Boer (University of Amsterdam): ‘(Brief) introduction to recent research at UvA on hub-and-spoke colluding pricing algorithms’
• 14:30 - Janusz Maylahn (University of Amsterdam): ‘Learning to collude in a pricing duopoly’
• 15:00 - Thomas Loots (University of Amsterdam): ‘Tacit collusion by self-learning pricing algorithms under a multinomial logit choice model’
15:30 - Screen break
• 15:40 - Justin Johnson (Cornell University): ‘Platform design when sellers use pricing algorithms’ (joint work with Andrew Rhodes and Matthijs Wildenbeest)
• 16:10 - Daniel Ershov (Toulouse School of Economics): ‘Algorithmic pricing and competition: Empirical evidence from the German retail gasoline market’ (joint work with Stephanie Assad, Robert Clark and Lei Xu)
• 16:40 - Screen break
• 16:50 - Joseph Harrington (Wharton School, University of Pennsylvania): ‘Third party pricing algorithms and the intensity of competition’
• 17:20 - Vincenzo Denicolo (University of Bologna): 'Algorithmic collusion under imperfect monitoring' and 'Protecting consumers from collusive prices due to AI' (forthcoming in Science, joint work with Emilio Calvano, Giacomo Calzolari, Joseph Harrington and Sergio Pastorello)