Do student responses parallel theory? Evidence from an oligopoly experiment

The authors test whether or not an existing classroom experiment on oligopolies (Lacombe and Ryan, 2003) is successful in getting students to model the behavior of firms that collude or compete. The conditions used under the Basic Structure of this experiment are expected to provide incentives that push students toward collusion. However, we expect an increased likelihood of competition when the experimental conditions are modified to approximate obstacles to collusion. This analysis indicates that the modifications we used to represent an Unstable Demand Curve condition and an Increased Number of Firms condition did not create a significant change in student responses in these classes. However, the modification that represented the Inability to Detect Price Changes led to a significant change in strategy by these same classes. These empirical data suggest that a straightforward classroom experiment can be used to aid students in recognizing the implications of an obstacle to collusion if it is substantial.