Does 'coordination Magic' Exist in Two-Sided Markets? an Experimental Investigation

We propose a two-sided market entry game and present experiments studying coordination behavior in the game. The two-sided market in the game is operated by an intermediary monopoly platform, serving two sides (i.e., customers and service providers) and featuring asymmetric agents, cross-side network effects, and endogenous market capacity. We find theoretically that the game has multiple pure-strategy Nash equilibria if at least one side has a high willingness to enter the market and the other side's willingness is not very low. The multiple pure-strategy Nash equilibria may be Pareto ranked on both sides or Pareto ranked on one side but Pareto equivalent on the other side. We further conduct a laboratory experiment involving three treatments corresponding to different combinations of willingness to enter the market among customers and service providers. The experimental results indicate that willingness to enter the market and cross-side network effects significantly influence coordination behavior in two-sided markets. When the multiple pure-strategy Nash equilibria are Pareto ranked on both sides, customers and service providers can coordinate their behavior to the payoff-dominant equilibrium via tacit coordination under strategic uncertainty. However, when the multiple pure-strategy Nash equilibria are Pareto ranked on one side but Pareto equivalent on the other side, coordination failure and disequilibrium occurred, and the equilibria cannot predict the aggregate behavior well. Also, we propose a variant logit choice model to reproduce the aggregate behavior in the experiment. Our experimental and simulated results indicate that a thriving two-sided market should get both sides on board.