Abstract
A vast amount of resources is wasted on software projects delivering less than the planned benefits. The objective of this paper is to investigate the tendency to continue a project even when it is evident that it will not provide the expected benefits, often referred to as “project escalation” or “escalation of commitment.” We aim to identify factors that empirically have been found to promote software project escalation. We examined 1376 papers related to the phenomenon of escalating commitment to software projects and found that 44 of them included relevant empirical research. After reviewing these papers, we synthesized the results. We provide an overview of 46 factors that have been found to promote software project escalation. Thirteen of the factors were project-related, 20 were psychological factors, nine were social factors, and four were structural factors related to the project's contextual dimensions. We contribute to practice by systemizing empirical evidence of software project escalation that can be of help in avoiding it in the first place and uncovering already escalated situations. As most studies are investigations of only a few factors, we propose further research to study how factors potentially interplay as they contribute to escalation of software projects.