Purdue News

JOURNALIST SUMMARY: Earnings pressures in a challenging economy translate into a necessity for change. But organizational change, even in the best of circumstances, is fraught with risk. A Purdue University business professor's research shows that change often introduces what he calls "latent errors" that can potentially lead to human and financial catastrophe. Now he's working on organizational factors that give rise to latent errors long before they wreak their organizational havoc, which can result in disasters ranging from the Barings Bank bankruptcy to the Columbia space shuttle disaster.

November 20, 2003

Managers: Organizational change can plant seeds of disaster

WEST LAFAYETTE, Ind. – While change is the order of the day for many companies in the hypercompetitive global marketplace, a Purdue University researcher says managers need to take steps to shortstop errors that those changes can cause.

Rangaraj Ramanujam, a Krannert School assistant professor of management, studies organizational change. In common terms, he describes the broad problem as "how very bad things happen to companies." His research seeks to identify ahead of time, and not in hindsight, organizational causes of errors that lead to disaster.

Ramanujam looks at organizational catastrophes, such as the 1995 Barings Bank bankruptcy following $1.3 billion in losses from unauthorized trading.

"Here's a case of Britain's oldest investment bank being brought down by 28,000 separate trades with a total value of $29 billion executed by one trader," he says. "The question is how could an organization of smart people let this happen?"

Barings, he says, is not a freak case. Similar sequences of events were replayed in other financial institutions such as Daiwa, Sumitomo and AIB.

Although such outcomes cannot be explained in terms of a few causes, they had a common set of precursors – deviations from policies and procedures instituted precisely to avoid such contingencies.

Ramanujam, in his research with Professor Paul Goodman at Carnegie Mellon University, examines the organizational causes of such "latent errors," which he defines as "major deviations from standard operating procedures, professional standards and regulations." His recent work provides evidence that when companies initiate large changes in top management, technology and structure, a significant, unintended consequence is an increase in latent errors.

"This does not mean that organizational change is either necessary or sufficient to cause adverse outcomes," Ramanujam cautions. "I'd describe the findings as making a strong case for being alert to this previously unrecognized aspect of implementing change."

The latent errors Ramanujam describes are not restricted to financial institutions. An all-too-familiar storyline in accident investigations is the discovery of "trigger events," after-the-fact findings of organizations not doing what they should have or doing what they should not have, Ramanujam says. Latent errors leading to trigger events were present in accidents in a range of settings, including the 1987 chemical plant explosion in Bhopal, India, and the Columbia space shuttle disaster this year.

Trigger events are followed by retrospective questions and reports that uncover the latent errors. But Ramanujam says these after-the-fact analyses "are susceptible to hindsight bias." In addition, he says, from a management point of view, once the media have covered a scandal or disaster, management is in a damage-control mode and can only promise to do things better in the future.

Another important aspect of latent errors is the growing consensus that their causes are primarily organizational. Ramanujam says a significant portion of the Columbia investigation report was devoted to organizational causes of what formerly was viewed purely as a technical problem. His current research, which was published in the October issue of the Academy of Management Journal, begins to look empirically at organizational causes.

Ramanujam's approach is to focus on what he calls a "snapshot" of the organization before the latent errors trigger consequences. He examined internal audit data of the operations of 80 units in a large financial institution offering private, retail and corporate banking services. The central question was whether units that experienced changes had more latent errors than units that did not.

"The study design focused on determining whether change had a significant effect and found that it did," Ramanujam says.

"Interestingly, a few units were very effective in reducing latent errors despite undergoing large-scale changes. In a follow-up study, we examine what might be different in such units. But as the overall findings suggest, such units are the exception."

He and Goodman interviewed 150 managers and auditors to find and analyze the latent errors that occurred during periods of organizational change.

"Current explanations of the organizational origins of accidents understate, or even ignore, the role of organizational change," Ramanujam writes. Even changes that organizations institute in order to reduce errors cause an increase in latent errors.

"Overall, this suggests that organizations implementing major change expose themselves to more latent errors," he says.

Ramanujam says there are a number of implications of his study for managers.

"The most important one is that organizations need to think about change management beyond terms of traditional metrics, such as time, cost and employee resistance, and take deliberate steps to minimize the creation of latent errors."

In other words, he says, successful change implementation should not be assessed only in terms of "getting it done" in the fastest, most effective and frictionless way.

"You need to think about whether you are sowing the seeds for a future disaster," he says, "Managers need to protect and augment organizational attention and memory during major change initiatives.

"For example, managers could appoint specific individuals or teams to monitor routine operations during times of change. They could also arrange an audit of operations concurrent with implementing change."

Ramanujam says the study makes a strong case for taking a close look at a previously unrecognized and potentially serious consequence of organizational change.

"Clearly, more research is needed to help us understand the implications fully, and this was the goal of a follow-up study that we're now analyzing the data from," he says.

Ramanujam also is doing research in a three-year study of medication errors and hospital-generated infections at 30 hospitals in western Pennsylvania.

Writer: Mike Lillich, (765) 494-2077, mlillich@purdue.edu

Source: Rangaraj Ramanujam, (765) 496-7521, (765) 496 -6912, ramanujam@krannert.purdue.edu

Purdue News Service: (765) 494-2096; purduenews@purdue.edu

Note to Journalists: A publication-quality photograph at Rangaraj Ramanujam is available at https://ftp.purdue.edu/pub/uns/++Mugshots/+krannert.mugs/ramanujam.r.jpeg

Internal audit data of operations in a financial institution were used to examine the effects of discontinuous change and work-inherent risk on latent errors, defined here as potentially consequential deviations from procedures and policies. Results show that discontinuous change increases latent errors and that this increase varies with the level of risk. Implications for organizational research on errors and reliability are discussed.

To the Purdue News and Photos Page