Choice, Accountability, and Effortful Processing in Escalation Situations Hans-Georg Wolff and Klaus Moser University of Erlangen-Nuremberg, School of Business and Economics DOI 10.1027/0044-3409.216.4.235 Abstract. Many studies on escalation of commitment observe only a single decision following negative feedback, although theoretical approaches to escalation behavior depict escalation as a progression over time. The present paper builds on Brockner and Rubin’s (1985) “tunnel vision” account that suggests a distinction between early and late stages of the escalation process. We used a dynamic paradigm, observing repeated decisions following negative feedback and manipulated choice and accountability in order to examine effects of justification on the progression of escalation behavior. Furthermore, reading times are used as a measure of effortful processing to investigate the mediating cognitive processes that lead to escalation behavior. Results show that the combination of choice with account- ability leads to escalation behavior at later stages of the escalation process and that effortful processing mediates this interaction of choice, accountability, and escalation behavior. Keywords: decision making, escalation of commitment, self justification, accountability, information processing Escalating one’s commitment to a chosen course of action can be a costly issue and evidence from the laboratory, as well as field research, has shown that decision makers are prone to invest in failing endeavors (e.g., Bobocel & Meyer, 1994; Schoorman 1988; for an overview see e.g., Staw, 1997). In general, escalation situations are character- ized by prior investments in a course of action and the sub- sequent occurrence of negative feedback with regard to goal attainment. The decision maker (DM) must decide be- tween withdrawal and persistence. In the case of withdraw- al, prior investments will be lost and the decision maker accepts the outcome of failure to attain the goal. In the case of persistence, goal attainment is possible, but uncertain. Additional investments will be wasted if the goal is not attained. Escalation behavior becomes manifest in the de- cision to persist or invest further resources in the course of action (Brockner & Rubin, 1985; Staw, 1997). The present paper adopts a dynamic perspective on es- calation (Goltz, 2000) and examines the joint impact of internal and external justification over time. Brockner and Rubin (1985) have coined the term “tunnel vision” for their finding that the effect of variables on escalation interacts with time and that specific cognitive changes occur in the process of escalation (e.g., Brockner et al., 1982). They suggest that economic considerations are important at early stages of the escalation process, but at later stages DMs become less vigilant to additional infor- mation, similar to the limited vision in a tunnel. In fact, there is evidence that effects of internal justification, that is DMs’ need to justify their decisions to themselves, oc- cur at early stages of the escalation process (Goltz, 1993; McCain, 1986; Staw & Fox, 1977), whereas external jus- tification, the need to justify decisions to others, is more important at later stages (Brockner et al., 1982). Howev- er, prior studies have only partially considered the inter- active effects of these factors. The effects of internal and external justification have been examined in separate studies or have been confounded in some studies (cf. Bobocel & Meyer, 1994). We contribute to the literature by examining the interactive effects of internal and ex- ternal justification and their effects at different stages of the escalation process. Another contribution of the present article is the ex- amination of the psychological processes that underlie escalation behavior. We intend to provide further evi- dence for the tunnel vision account, which predicts that DMs expend cognitive effort on rationalizing their deci- sions at later stages. Whereas prior research has repeat- edly administered reactive measures (i.e., questionnaires) whose delivery can itself impact DMs’ behavior, the pre- sent study uses a nonreactive measure (i.e., reading times) to investigate the amount of cognitive effort that DMs expend as escalation unfolds over time. We also ex- amined whether effortful processing acts as a mediator in the escalation process. Theoretical Background In the present paper we use the term “escalation of com- mitment,” which is more common in the organizational lit- erature (Staw, 2005) whereas other research, especially in social psychology, uses the term “entrapment,” which is synonymous with escalation (e.g., Brockner & Rubin, 1985; Staw, 2005). The escalation phenomenon is also con- © 2008 Hogrefe & Huber Publishers Zeitschrift für Psychologie / Journal of Psychology 2008; Vol. 216(4):235–243 Zeitschrift für Psychologie / Journal of Psychology 2008.216:235-243.