Betting on the Trend: An Experimental Study on Network Externalities With Continuous Demand
40 Pages Posted: 28 Nov 2006
Date Written: October 18, 2006
We study consumer decision making in a market with network externalities, in which the utility of a purchase increases with the number of purchasers. While research in this area typically models the consumer as making a binary choice, we propose a new model with continuous individual demand, thus allowing for sensitive gauging of agents' belief towards the product's success. We derive a payoff function that can be implemented experimentally as a voluntary contribution game with non-standard characteristics including: (1) Free-riding is impossible; (2) Contribution is profitable iff total contribution exceeds a critical mass; (3) There are two pure-strategy Nash equilibria, one incurring zero contribution, while the other has full-endowment contribution from all players to achieve Pareto optimality. We collect experimental data for the repeated game, which show that the Pareto-optimal equilibrium is not reached under some conditions. We propose a behavioral theory to interpret the data and conclude that: (1) Behavior is influenced by social and game-theoretic factors; (2) Convergence to the Pareto-optimal equilibrium is more likely than not when the critical mass is less than approximately half of the population's total endowment; (3) The difference between first-round total contribution and the critical mass predicts convergence type and efficiency.
Keywords: Network externalities, critical mass, consumer behavior, decision-making, coordination, learning in games, voluntary contribution game, game theory, experimental economics
JEL Classification: C72, C92, D62
Suggested Citation: Suggested Citation