Coordination in the Aggregate Without Common Knowledge or Outcome Information

Coordination of entry decisions in the iterated MEFC (market-entry fixed cost) game with asymmetric players and varying market capacity values can be achieved either without common knowledge of the distribution of types or without trial-to-trial information about number of entrants and individual payoff. Comparison of the results of each of the two experiments with the earlier findings reported by Rapoport, Seale and Winter (1997) shows that removing either of these two sources of information does not affect coordination on the aggregate level and possibly improves it on the type level. The Nash equilibrium solution accounts for the results of both experiments. Coordination without common knowledge is achieved by some sort of learning with players steadily increasing their use of cutoff decision rules that more or less sort themselves into types. The findings suggest that (1) providing the subjects with common knowledge of the game may not be necessary if the game is iterated for sufficiently many periods; and (2)in the absence of an opportunity to learn, players will use whatever cues they have to coordinate their actions