InterJournal Complex Systems, 140
Status: Submitted
Manuscript Number: [140]
Submission Date: 971026
Hazards, Self-Organization, and Risk Compensation: A View of Life At the Edge.
Author(s): Keith Donnelly

Subject(s): CX.44, CX.41, CX.13, CX.43, CX.46, CX.4

Category: Article

Abstract:

The response of a human system or organization to hazards may often be viewed in terms of self-organized criticality. That is, the operating point of the system, in terms of resources, scheduling, work methods, safety barriers, etc, evolves until a balance is reached between competing pressures such as the simultaneous desires for high work productivity and a low probability of accidents. From a safety viewpoint, the system may be viewed as having migrated away from high safety to a point where it is "on the edge of chaos". Chaos in this case may be defined as a situation where a small perturbation may or may not be absorbed by the system; the outcome is often tragic. At this operating point "on the edge", the frequency of small incidents already gives an indication that more severe incidents are distinctly possible, even though much less likely to occur. Empirically, a power law relates the frequencies of different severity events. Examples of balanced systems or systems with power laws are air traffic control, powerline maintenance work, the Great Fire of London, and the King's Cross Underground Fire. In the safety literature, Risk Compensation is a theory claiming that most safety improvements will be nullified because humans will increase their exposure to or level of hazard and claim the benefit in some other currency such as time or performance, while keeping the net risk unchanged. Driving faster and following more closely after the introduction of seat belts, anti- lock brakes and airbags are often cited as examples. This is often described as being governed by a "risk thermostat" whereby a target level of acceptable risk is maintained. The insight provided by self-organization is that the feedback provided by small incidents, and the protection against severe events afforded by a power law, allows the operating point to be set "on the edge." Ultimately, it is hoped that theories of self-organized systems will lead to better management indicators of safety risk and lead to insights when organizational or economic changes may be inadvertently moving a system into an unacceptable regime.

Retrieve Manuscript
Submit referee report/comment


Public Comments: