What do Silicon Valley Bank, Southwest Airlines, and Norfolk Southern all have in common? These organizations, and many like them, suffered business catastrophes in the past year. A run on Silicon Valley Bank led to its seizure by regulators; Southwest Airlines’ crew scheduling software collapsed, canceling more than 16,000 flights over the winter holiday season; and Norfolk Southern’s derailment in East Palestine, Ohio, caused the community’s evacuation and the burning of toxic chemicals. All three catastrophes likely occurred for the same fundamental reason—one that might have been preventable.
Such business catastrophes and industrial disasters are not as uncommon as one might think. Bank failures in the U.S. average about 25 per year, significant disruptions in airline service are well-known by every frequent flyer, and train derailments average about three per day. The increasing volatility, uncertainty, and complexity of external factors, combined with instantaneous global distribution, make catastrophes more strategically important—if not more likely—over a wider array of organizations than ever.
What is fragility drift?
Typically, the subsystems of an organization or industry need to operate in harmony for the overall system to work effectively. However, these systems can become fragile, inviting catastrophe with the next unexpected shock. How can these systems be fragile when they are initially designed for resilience against industrial disasters?
Akhil Bhardwaj from Tilburg University, Joseph Mahoney from the University of Illinois at Urbana-Champaign, and I have developed new thinking to explain the common final limit in these systems. We call the explanation “fragility drift.” In essence, making changes (or “adaptations”) in one subsystem without checking to see if they might affect other subsystems can endanger the overall system. Like the tower-building game Jenga, adjusting one block on top can increase instability at the bottom and precipitate a catastrophe.
With fragility drift, the system does not appear to be “broken,” so typically, it doesn’t “get fixed.” Yet, fragility drift increases the likelihood that a shock, once easily handled by the system, can now spark a catastrophe and necessitate a recovery plan. For example, not replacing a risk officer who dynamically balances risk might make a bank fragile to sudden withdrawals if interest rates change rapidly. An airline’s continuous growth and geographic expansion might leave once-reliable software unable to handle severe weather. And, without sequencing railcars by weight, braking in unexpected locations might cause heavily loaded railcars to bunch up and derail lightly loaded or empty railcars.
Why might fragility drift happen?
Operators, executives and regulators might not know better. Without system knowledge and detailed system analysis, a local adaptation in one subsystem can (wrongly) appear to be a good idea, especially if failure does not instantly follow. Every adjustment has a consequence, and not all consequences are immediately apparent.
Additionally, changes to business continuity and disaster recovery plans don’t always occur knowingly. Operators, leaders or regulators might pursue “subgoals.” They may trade off difficult-to-measure, long-term system resilience to achieve easy-to-measure, short-term goals, such as improved product features and cost-cutting, to earn a bonus. Or they might succumb to internal or external political pressure to make adaptations that others desire.
How can you prevent fragility drift?
1. Establish awareness. Understanding why fragility drift occurs is the first step in preventing it. The goal is to make operators, leaders and regulators aware of the potential for fragility drift so that they will avoid undertaking adaptations that increase fragility and find ways to prevent the pursuit of “wrong” subgoals.
To raise awareness, organizations should conduct regular risk assessments, integrate systemic thinking into decision-making, and establish clear communication among cross-functional teams. These steps ensure that the interconnectedness of subsystems is considered, fostering a culture that preemptively identifies and addresses potential areas of vulnerability.
2. Visualize through heat maps. Create a simple graphical way for laypeople to understand which subsystem pairings will likely experience fragility drift. For example, we created a graphic for freight train transportation that resembles a heat map. Experts in the industry helped us identify 11 vital freight train transportation subsystems, leading to the creation of a matrix of all subsystem pairs. Widely distributing and explaining such a heat map could help everyone become aware of fragility drift and recognize when a stability analysis needs to be conducted before implementing an adaptation.
3. Ensure system analysis. Based on the heat map, incorporate policies requiring system analysis before making adaptations and create an audit function to veto adaptations that have not undergone the requisite analysis. Political influence could undermine the audit function unless it reports directly to the board of directors. Without a direct reporting relationship to the board, audits might not be sufficiently credible to keep political influence at bay.
When considering adopting any new perspective like fragility drift, we need to ask ourselves three questions: Is the idea novel and valuable? What are the net benefits of adopting it? What are the potential costs of ignoring it? The benefit of adopting this perspective is that it could prevent a disaster. The cost of neglecting fragility drift might be best ascertained by imagining walking in the shoes of the chief executives of Silicon Valley Bank, Southwest Airlines and Norfolk Southern after their catastrophes and thinking about all the anguish from the loss of people, profits and property.
