An experimental investigation of market entry problems.

Abstract

This dissertation considers organizational problems of market entry. The research follows the experimental path. Game theoretic models are combined with laboratory experiments to produce a set of empirical findings. Two market entry problems are studied. The first considers the chain store paradox developed by Selten (1978). This game considers an established chain store with locations in numerous towns. In each of these towns a different competitor decides whether to enter and compete with the chain store. When entry occurs, the chain store can respond cooperatively or aggressively. The game proceeds sequentially, the players are not symmetric, and the critical solution concept is the subgame perfect equilibrium. Three experiments are conducted for a total of 550 trials of the game. Experiments differ in the size of payoffs, the number of entrants, the anonymity of the chain store, and whether subjects play in both the role of the chain store and an entrant or in just one role. There is qualified support for the game theoretic prediction that a chain store cannot deter the sequential entry of competitors. Entry occurred on 459 of 550 trials; while some chain stores pursue deterrence, it largely is not effective in these specific experimental environments. It is suggested that deterrence might be effective if the number of entrants or payoffs are increased. The results have implications for discussions on predatory pricing, reputation, and the value of backwards induction as a solution concept. The second market entry problem is based on a simultaneous market entry game developed by Rapoport (1994). In this game symmetric players decide simultaneously whether to enter a market with a specified capacity. The game theoretic prediction for the number of entrants is based on a Nash equilibrium (in pure or mixed strategies). Again, experimental results support game theoretic predictions. Across three experiments the correlation between the number of entrants and the size of the market capacity is consistently above 0.90. Taken together, these experiments on market entry problems provide strong support for the conceptual use of game theory and the methodological use of controlled laboratory experiments in the field of strategic management.