Complexity and Competition, Part I: Sequential Matching

Working Paper
Change log
Gale, Douglas 
Sabourian, Hamid 

This paper uses the complexity of non-competitive behaviour to provide a new justification for competitive equilibrium in the context of extensive-form market games with a finite number of agents. This paper demonstrates that if rational agents have (at least at the margin) an aversion for complex behaviours then their maximizing behaviour will result in simple behavioural rules and thereby in a perfectly competitive outcome. In particular, we consider sequential market games with heterogeneous sets of buyers and sellers and show that if the complexity costs of implementing strategies enter players� preferences, together with the standard payoff in the game, then every equilibrium strategy profile induces a competitive outcome. This is done for sequential deterministic matching/bargaining models in which at any date either the identities of the matched players are determined exogenously or one player is exogenously selected to choose his partner and make a price proposal.

matching, complexity, automation, bounded rationality, Markov equilibrium, competitive equilibrium
Is Part Of