Two Essays on Escalation of Commitment

This dissertation focuses on managerial decision making, and specifically explores conditions wherein managers may increase their propensity to escalate commitment towards a failing project. Escalation researchers (e.g. Schmidt and Calantone, 2002) have listed four classes of factors that may impact a manager's propensity to escalate commitment towards a failing project, and have called for research into how exactly these factors impact escalation. In this dissertation, we explore two such factors. The first factor relates to the characteristics of the decision process used by firms to evaluate the project. Here, for example, researchers have looked at whether the manager was also involved in making decisions about the project in a prior period, and Boulding, Morgan and Staelin (1997) have shown that such manager's positive beliefs about the project (formed in a prior period) make a manager more likely to escalate commitment. The second factor relates to project characteristics. Here, for example, researchers have looked at whether or not the project relates to a product that is perceived as new, and Schmidt and Calantone (2002) have shown that managers are more likely to escalate commitment towards a failing project relating to a new product. The first dissertation essay uses three experiments to examine how a hitherto unexplored characteristic of the decision process might lead to increasing escalation of commitment. Specifically, building off research into the illusion of control, we examine whether the opportunity to use managerial skill during the decision process makes a manager more willing to escalate commitment towards a failing project. We find that whenever managers act on cues that cause them to think they can use their managerial skill to control some outside factor (even though in reality they cannot), managers overestimate their ability to "control the odds" related to this outside factor. Such beliefs feed forward and lead managers to make suboptimal decisions about the overall project. The second dissertation essay looks at how project characteristics might make a manager more (or less) likely to escalate commitment towards a failing project. We explore this issue in the hitherto unexplored real options setting. Real options have emerged as an important part of marketing strategy, and have been used to structure new product alliances, value customers etc. We run a controlled experiment and we examine whether differences in option-structure (which is a project characteristic) impact the propensity to make suboptimal option-exercise decisions. We find that managers are more likely to make suboptimal option-exercise decisions in the case of put options (vis. call options), and - as predicted by the endowment effect literature - this increased propensity to make a suboptimal decision is mediated by/ explained by the psychological ownership construct.