The Econometrics and Behavioral Economics of Escalation of Commitment

We examine the phenomenon of escalation from an economist's perspective, emphasizing explanations which do not rule out rational behavior on the part of firms or agents. We argue that escalation cannot be established as a separate phenomenon unless these possible alternative explanations are properly accounted for. We present Staw and Hoang's (1995) study of NBA data as an instance of where evidence of escalation might be overturned upon more careful analysis. After performing several tests of our alternative explanations, we find that evidence of escalation persists, although it is weaker both in duration and magnitude.