Introduction: The Inevitability of Migration
In the study of High-Reliability Organizations (HROs), “drift” is the term used to describe the slow, incremental departure from established safety procedures. Unlike a sudden mechanical failure or a singular “human error,” drift is a phenomenon of success—it occurs because the system is functioning, adapting, and striving for efficiency. While Jens Rasmussen, Scott Snook, and Sidney Dekker all address this migration toward the boundary of failure, they differ significantly in their diagnosis of why the migration occurs and where the responsibility lies.
1. Rasmussen’s Dynamic Safety Model: The Pressure of Gradients
Jens Rasmussen (1997) provides the structural foundation for drift theory. He posits that organizations operate within a “space of safe operation” bounded by three constraints:
- The Boundary of Economic Failure: The need to stay profitable or funded.
- The Boundary of Unacceptable Workload: The human need to minimize effort.
- The Boundary of Acceptable Performance (Safety): The limit beyond which a hazard becomes an accident.
Rasmussen argues that “gradients”—the constant pressure to be faster and cheaper—push the operating point away from the economic and workload boundaries and directly toward the safety boundary. In this model, drift is an optimization process. Workers are not being reckless; they are successfully navigating a “finish-to-finish” race where the safety boundary is often invisible until it is crossed.
2. Snook’s Practical Drift: The Decoupling of Theory and Practice
Scott Snook (2000) shifted the focus from external pressures to internal organizational sociology. Through his analysis of the 1994 Black Hawk shootdown, he identified “Practical Drift” as the slow decoupling of Work-as-Imagined (the official rules and regulations) from Work-as-Done (the daily reality of the task).
For Snook, rules are “global” and static, while tasks are “local” and dynamic. Over time, operators realize that following every rule makes the job impossible. They begin to “locally optimize,” creating “workarounds” that become the new standard. Drift here is a pragmatic adaptation. The danger arises when these local adaptations are never communicated back to the center, leading to a system where the “paper” reality and the “operational” reality no longer share the same map.
3. Dekker’s Drift into Failure: The Dark Side of Complexity
Sidney Dekker (2011) brings the conversation into the realm of Complex Adaptive Systems (CAS). Dekker argues that in modern, high-tech environments, failure is an “emergent property.” Unlike Snook or Rasmussen, Dekker emphasizes that drift can occur without anyone breaking a rule or feeling excessive pressure.
In Dekker’s view, drift is a series of small, “locally rational” choices. Each step is so minute that it doesn’t trigger an alarm. No one has a “god’s eye view” of the entire system, so no one sees the accumulation of risk. This is the Normalization of Deviance taken to its logical conclusion: the system becomes increasingly brittle while appearing, on the surface, to be performing better than ever.
Comparative Analysis: Where the Models Diverge
The distinction between these three lies in their view of intentionality and visibility:
- Visibility of the Boundary: Rasmussen believes the boundary exists but is poorly marked. Snook believes the boundary is ignored in favor of “getting the job done.” Dekker argues that in complex systems, the boundary is constantly moving and essentially unknowable until the crash occurs.
- The Nature of the Actor: In Rasmussen’s world, the actor is an Optimizer. In Snook’s, they are a Pragmatist. In Dekker’s, they are a Local Expert making sense of a chaotic environment.
- The Solution: Rasmussen suggests marking the boundaries more clearly. Snook suggests “recoupling” the organization through better communication. Dekker suggests building “capacity” and “resilience”—accepting that drift will happen and creating systems that can fail gracefully.
Conclusion
The synthesis of these models suggests that “drift” is not a pathology to be cured, but a fundamental characteristic of any living system. To manage drift, an organization must move away from the “Old View” of human error and instead study the pressures, adaptations, and local rationalities that make drift a “successful” strategy for employees right up until the moment it fails the organization.
Academic Resources & Comparative Literature
For additional research, the following papers are seminal to each theory and provide a comparative analysis:
1. The Foundational Papers
- Rasmussen, J. (1997).Risk management in a dynamic society: A modelling problem. Safety Science.
- Significance: This is the “Big Bang” of drift theory, introducing the idea of “migration of work practices.”
- Snook, S. A. (2000).Friendly Fire: The Accidental Shootdown of U.S. Black Hawks over Northern Iraq. Princeton University Press.
- Significance: Though a book, the introductory and concluding chapters define “Practical Drift” as the decoupling of global rules from local practice.
- Dekker, S. W. A. (2011).Drift into Failure: From Hunting Broken Components to Understanding Complex Systems. Ashgate Publishing.
- Significance: Transitions the conversation from “human error” to “complexity theory.”
2. Comparative & Evaluative Articles
- “Rasmussen and practical drift: Drift towards danger and the normalization of deviance” (Risk-Engineering.org).
- Link to Article — This is an excellent primer that explicitly links Rasmussen’s boundaries to Snook’s normalization of deviance.
- “Drifting into failure: Theorising the dynamics of disaster incubation” (Dekker, 2013).
- Link to ResearchGate PDF — Dekker himself compares his work with Rasmussen’s “migration” and Turner’s “incubation.”
- “Reflecting on Jens Rasmussen’s legacy: A strong program for a hard problem” (Waterson et al., 2015).
- Link to Paper — This provides a retrospective on how Rasmussen’s work evolved into the “New View” championed by Dekker.
| Feature | Jens Rasmussen (1997) | Scott Snook (2000) | Sidney Dekker (2011) |
| Model Name | Dynamic Safety Model / Accimap | Practical Drift | Drift into Failure (DIF) |
| Why Drift Occurs | Gradients. Economic and workload pressures push the “operating point” away from effort and cost toward the safety boundary. | Uncoupling. Local units adapt rules to be “practical” for their specific environment, breaking away from centralized design. | Complexity. Small, iterative, and “locally rational” choices accumulate in a non-linear system, creating emergent risk. |
| Nature of Rules | Rules are “boundaries” that provide a space for safe navigation. | Rules are “Theory” (WAI) vs. Practice is “Action” (WAD). | Rules are often “brittle” or oversimplified versions of a complex reality. |
| Key Metaphor | The Boundary. Flirting with the margin of error to stay competitive. | The Gap. The slow decoupling of “how we say we do it” from “how we actually do it.” | The Non-Linear Descent. A steady, almost invisible path toward a disaster that no one saw coming. |