Is there no “I” in “Team”? Interindividual-intergroup discontinuity effect in a Cournot competition experiment

Abstract

• We study individual versus team decisions in a Cournot duopoly experiment. • In a “standard” experiment individuals’ and teams’ decisions are indistinguishable. • Outputs decisions of individuals and teams are both around the Cournot-Nash outcome. • We conducted a second experiment with arguably a recommendation to collude. • Now we find that individuals choose lower outputs and collude more than teams. Even though firms consist of teams of individuals, industrial organization experiments are typically conducted under the unitary player assumption with individuals representing firms. Against this practice, there is evidence that individuals’ decisions differ from those made by teams, and also from individuals’ decisions which affect both oneself and others. This study examines the empirical scope of the unitary player assumption (alternatively the “interindividual-intergroup discontinuity effect”) in the context of a Cournot competition experiment. In the first, standard, experiment (216 participants), we find no evidence for the discontinuity effect. Output choices of both individuals and teams are around the Cournot outcome. In a second experiment (198 participants), we use a recommendation to shift participants’ initial disposition towards the (off-equilibrium) collusive output. This triggers the discontinuity effect; individuals choose lower outputs and collude more than teams. This effect is persistent in the second half of the experiment, and is not caused by team-member discussion or by seating team members together. We discuss the interpretations and implications of our results.