, , ,

What do banks and nuclear power plants have in common? They are both the targets of so-called stress tests, that is, checks for testing system robustness under conditions of excessive strain. Stress testing on banks was motivated by the 2008 financial crisis; in the case of nuclear power plants it was triggered by the recent Fukushima disaster. One could probably enlarge the concept and apply it to oil rigs (remember BP?) or aircrafts (… and volcanic ash).

It is always instructive to scrutinize how and why specific terms gain prominence under what circumstances. ‘Stress testing’ has to do with risk analysis that incorporates scenarios, an approach that should have been the norm. Done well, scenario-driven policy assessment provides information as to the drivers of change, visions for the future, external factors impacting on developments, internal systemic variables, the role of regulation, and the costs and benefits of different impact pathways.

The decision to refer to all of the above as ‘stress testing’ has perhaps to do with the fact that ‘stress’ is a notion that everyone understands, therefore policy-makers might be relying on this to ‘nudge’ us into understanding what it is about. However, given that the term ‘stress’ has various connotations there is a certain danger in using it as a metaphor. Background assumptions may spill over and taint (rather than elucidate) our understanding of what ‘stress testing’ is or should be about. This is more likely to happen when detailed information is missing—as is the case regarding banks or nuclear power plants.

Stress is diffuse and nebulous in character as it can encompass pretty much everything—so much so that one is tempted to use it to dismiss uncomfortable feelings or situations of whatever sort, thus avoiding inquiring further into their meaning or genesis. This is the first catch with stress. The second is that we tend to consider forgetfulness the best remedy against it.

Another handicap is the focus on the limits (institutional, technical or otherwise) of a particular approach. Presently, and applied to banks, stress tests are about liquidity. In the case of nuclear plants they are about safety standards. Yet, eventually, it will be equally important to ask questions about so-called systemic failures and consider alternative or supplementary models that might entail less high-risk futures.