When adverse selection problems exist, the decisions made by project managers may be highly rational viewed from their perspective, if not from the firm's perspective, particularly if the potential effects on the agent's future reputation are taken into consideration (Kanodia, Bushman, & Dickhaut, 1989). Kanodia et al. (1989) suggest that project managers' escalation behavior is one example of a larger phenomenon of hiding private information on human capital (e.g. managerial talent).6 Since managers' talent is inferred by others through their decisions, any future events resulting from these decisions will have the potential to reveal to the labor market which managers are talented or not talented. Kanodia et al. theorize that managers who decide to escalate in the second period of a project can expect to receive higher wages in the future than managers who decide to switch. Switching projects in the second period results in a negative effect on a manager's future expected wages, since this decision reveals that the managers have made an incorrect decision in the first period.
(Harrison and Harrell, 1993) and (Harrell and Harrison, 1994) provide empirical evidence to support the theory of Kanodia et al. (1989). They conclude that when project managers realize that the project that they are managing is failing, they tend to terminate the project if the information about the failing project is publicly available. The rationale for terminating the project: (1) others would know that the project was failing; (2) managers had already suffered damage to their reputation, so no further benefit could be obtained from escalation; and (3) the decision to continue under such conditions could even result in further damage to their reputation as talented project managers. However, when the information about the failing project is known only to the project managers (i.e. the information is privately available), they tend to continue the project even though they realize that their decision is contrary to the firm's interest. By doing so, they can further their own economic interest (e.g. future career opportunities) without fear of repercussions.
The above discussion suggests that the type of availability of information (i.e. private or public information) is an important determinant of project managers' escalation tendency. Bearing in mind the relatively greater emphasis on the preservation of reputation among project managers, it is predicted that the availability of information may contribute to the likelihood of obedience pressure affecting project managers' decision making. When information about the future unprofitability of an investment project is publicly available, the effect of obedience pressure may become diluted by the possibility of public censure. Both project managers and others in the industry realize that such pressure (policy) is inappropriate. Both of them also notice that the current project is failing and therefore should be terminated. As a consequence, being shielded from the possibility of public censure by deciding not to obey may protect a project manager's current unfavorable performance from further damage. On the other hand, when information is privately available, the decision to embrace the instruction to continue the project will be more beneficial from the standpoint of the project managers. That is, it will, at least in the short term, delay their unfavorable performance from being detected by others, and the presence of obedience pressure may give project managers additional feelings of security about their decision (e.g. to avoid immediate punishment). This prediction is consistent with Lord and DeZoort (2001) notion that the effect of obedience pressure will depend on subordinates' perceived costs and benefits related to their responses. Lord and DeZoort (p. 217) mention that auditors may prioritize the importance of securing a favorable evaluation from superiors who provide inappropriate direction over the risk of "getting caught."
In summary, project managers who are subject to the presence of obedience pressure under conditions of private information will have a higher tendency towards escalating their commitment to continue a failing project than will their counterparts who are subject to only one or neither of these conditions. Stated formally:
H2
Project managers who experience obedience pressure and private information will exhibit a greater tendency to continue a failing project than project managers who experience only one or neither of these conditions.
