Platform Collusion in Two-Sided Markets

Résumé

One focus of the two-sided markets literature has been on testing the robustness of standard competition policy results. Price collusion between platforms, however, has not yet been studied. Starting from cartel cases involving two-sided markets, this paper takes a first step in analyzing collusion in an Armstrong (2006) type of setting. In particular, it addresses Evans' hypothesis of collusion being harder to sustain because of feedback effects and stronger requirements concerning agreements and monitoring. First results show that collusion is not harder to sustain than in a "normal" one-sided market with direct network effects whenever the two-sided network externalities are symmetric. If valuations for members on the opposing market side are asymmetric, however, price collusion becomes harder to sustain compared to the one-sided benchmark, although the total external benefit to consumers is identical in both scenarios. This implies that asymmetry rather than size of indirect network effects in a two-sided market matters for collusive stability.