One of the most challenging decisions that a manager must confront is whether to continue or abandon a troubled project. Published studies suggest that failing software projects are often allowed to continue for too long before appropriate management action is taken to discontinue or redirect the efforts. The level of sunk cost associated with such projects has been offered as one explanation for this escalation of commitment behavior. What prior studies fail to consider is how concepts from risk-taking theory (such as risk propensity and risk perception) affect decision makers' willingness to continue a project under conditions of sunk cost. To better understand factors that may cause decision makers to continue such projects, this study examines the level of sunk cost together with the risk propensity and risk perception of decision makers. These factors are assessed for cross-cultural robustness using matching laboratory experiments carried out in three cultures (Finland, the Netherlands, and Singapore). With a wider set of explanatory factors than prior studies, we could account for a higher amount of variance in decision makers' willingness to continue a project. The level of sunk cost and the risk perception of decision makers contributed significantly to their willingness to continue a project. Moreover, the risk propensity of decision makers was inversely related to risk perception. This inverse relationship was significantly stronger in Singapore (a low uncertainty avoidance culture) than in Finland and the Netherlands (high uncertainty avoidance cultures). These results reveal that some factors behind decision makers' willingness to continue a project are consistent across cultures while others may be culture-sensitive. Implications of these results for further research and practice are discussed.
A general framework of different procurement strategies is introduced to help managers review their project portfolio to find more effective ways of using both internal and market resources in information systems development. Major decision criteria-the specificity of system design and the uncertainty involved in requirements specification-are adopted from transaction cost economics to determine what procurement strategies should be used in different situations. According to our Procurement Principle, systems that are company-specific and involve high uncertainty have to be internally developed because they require both the specific knowledge arid intensive interaction between developers and users. More standard requirements indicate the use of outside consultants or software contractors who have experience and knowledge about a similar type of systems. For routine systems common in many organizations, acquisition and tailoring of a software package provides the most efficient procurement strategy. The Procurement Principle is also empirically tested with data from recent system development projects in major Finnish companies. Partial support was gained for the framework, but some interesting deviations were also detected, such as a tendency to rely on in-house development of even routine systems.
Information technology (IT) projects can fail for any number of reasons, and can result in considerable financial losses for the organizations that undertake them. One pattern of failure that has been observed but seldom studied is the runaway project that takes on a life of its own. Such projects exhibit characteristics that are consistent with the broader phenomenon known as escalating commitment to a failing course of action. Several theories have been offered to explain this phenomenon, including self-justification theory and the so-called sunk cost effect which can be explained by prospect theory. This paper discusses the results of a series of experiments designed to test whether the phenomenon of escalating commitment could be observed in an IT context. Multiple experiments conducted within and across cultures suggest that a high level of sunk cost may influence decision makers to escalate their commitment to an IT project. In addition to discussing this and other findings from an ongoing stream of research, the paper focuses on the challenges faced in carrying out the experiments.