When commitment to failing projects is perfectly rational

In 1989, the city of Denver devised a plan to replace its old airport with a state-of-the-art facility. The plan famously went awry, costing three times as much as its original budget and running a year and a half behind schedule. The troubled, delayed 1995 opening is an example of the phenomenon known to those in the information technology sector as “project escalation” — the continued, persistent commitment to a failing project.

Nicholas Berente

Numerous published papers have examined the elements that doomed the project. “It ended up being very politicized,” said Nicholas Berente, professor of IT, Analytics, and Operations at the University of Notre Dame’s Mendoza College of Business. “It wasn’t just, ‘The technology didn’t work.’”

Despite multiple points of failure, city officials stuck with their flawed plan. “The city of Denver wanted a next-generation airport; they were going to pump any money they could into getting it,” he said.

Charging ahead with a plan despite negative feedback has been at fault for failing projects around the world. Berente studied approaches to the problem in the Journal of Management Information Systems paper, “Rethinking Project Escalation: An Institutional Perspective on the Persistence of Failing Large-Scale Information System Projects” with co-authors Carolina Alves de Lima Salge from the University of Georgia Terry College of Business, Venkata K.P. Mallampalli of the University of Arkansas Walton College of Business, and Ken Park of KP consulting.

Before his career in academia, Berente owned a company that specialized in helping companies implement technologies. At Mendoza, he researches and teaches digital transformation — or, as he describes it, organizational change with technology.

In studying project escalation, Berente believes that the Denver airport failure was more than “just rational individuals not making the rational decision.” From the review of hundreds of papers on the topic, the researchers identified four different standards by which people justify long-term projects that they refer to as different “logics”: economic (decisions based around cost); innovation (focusing on ideals like “a strong desire to be first” and “modernizing government”); engineering (relying on the capabilities of technological professionals to complete a project); and managerial (emphasizing bureaucracy, rules and decision rights).

Examined through this lens, Berente said, “Escalation is not some weird exception to rationality. It’s perfectly rational by some logics.”

The paper also noted that the logics applied can change depending on the phase of the project. For instance, the economic logics of return on investment and innovation are most salient before project approval, but issues like economic cost, technological impositions, and managerial concerns become more relevant after the project launches.

The authors’ analysis complements existing literature on project escalation and provides additional context about how rationale factors into such situations. It might seem like cost overruns and technical problems should logically kill a project, they wrote, but “decision makers draw on different legitimizing logics in different phases of a project.” They also noted that reducing commitment to previously legitimate projects typically involves different standards than the ones that escalated the project in the first place, and are often triggered by external criticism or new management.

One takeaway from the analysis is that managers need to readjust expectations of a united perspective on a large-scale project. “They have this idea that anybody who’s not on board with us is resisting. We have to overcome that resistance,” said Berente. Rather than relying on simplistic notions of managing large-scale change projects, managers must realize that “sometimes, those things that don’t align are perfectly legitimate,” he added. “We have to think of how to reconcile large-scale change with those legitimate, rational aspects of the organization.”

Nicholas Berente is an IT, Analytics, and Operations professor at the University of Notre Dame Mendoza College of Business. He currently teaches the executive MBA Data Technology for Senior Leaders and a doctoral seminar on the philosophy of science. His research examines how digital innovations drive change in organizations and institutions.