When a probabilistic risk event manifests itself with sufficient frequency, we can use the well developed theory of statistics to both identify its probability and loss characteristics, and make decisions that will minimize the overall effects of the event in the long run. When dealing with catastrophic loss potentials, the application of statistics based on the observation of averages over the long run doesn't make sense. Catastrophes are by definition terminal, final and irreversible. The problem with catastrophes is that, in the long run, there may be no long run.
If we want to avoid catastrophe, we need to establish some minimum acceptable risk level, based on the "practical impossibility" of calamity. At the very least, this involves maintaining a level of catastrophic risk in the world that is at or below some "natural" level. These two levels of risk, the minimal natural level and that at which risk become statistically apparent are often referred to as the de minimis and the de manifestis level, respectively. In the chart below (Figure 1), we identify the region of insignificant (de minimis) risk as beginning at around 1 in 100,000 (10-5) to 1 in 1,000,000 (10-6) annual individual risk level (roughly, the annual likelihood of a fatal lightning strike). At the other end of the risk spectrum are those events that occur with sufficient frequency as to be statistically definable.
Figure 1. De Minimis Risk, De Manifestis Risk and the "Safety Cushion"
The "in between" region becomes especially critical when (1) we are dealing with events whose probability can only be assessed very imperfectly, making precise identification of risk difficult if not impossible, and (2) we don't get a second chance to make the right choice. Viewed in this fashion, the region between de minimis and de manifestis becomes a "safety cushion". This cushion is based on a rough understanding of what we don't know, and how the gaps in this knowledge may adversely affect us.
By taking risks in a safety zone, we increase at least the potential for disaster. As a result, we must recognize that the progress we gain by pushing the precautionary envelope is only achieved at the expense of living under the very real threat of disaster. Risk regulation today reflects our decision to acquiesce to danger in return for the benefits of "progress".
This permissive attitude in risk regulation, as opposed to its precautionary alternative, is often expressed in the form of risk tolerability criteria. In effect, tolerability represents a quasi- cost/benefit structure. We aim for the de minimis level, unless benefits (alternatively, costs) are "too high". The criteria remains genuinely restrictive only with respect to the high-end, or de manifestis level. Of course, given the uncertainties involved, 10-3 may be very difficult to distinguish from 10-2, and that puts us at or over the de manifestis level.
Security from catastrophic risk requires that we preserve the safety cushion. The challenge in carrying out such an imperative is that precautionary avoidance of risk is very much an all-or-nothing affair. Piecemeal reductions in risk are based on a variety of "prioritization" schemes, including cost/ benefit analysis and selective ("weak") precaution. They are not effective against the fundamental problem of catastrophe ("in the long run, there is no long run"). Partial solutions maintain the status quo in terms of risk potential, while the proverbial doomsday clock ticks away. In the mean time, the threat of cataclysmic dénouement remains very real. It brings into being what we might call the "clean slate" condition. Post-cataclysm, we are left to start over, presuming of course there is any one (or any thing) left.
In choosing a regulatory structure, we need to recognize that we can not have it both ways. We chose between the (immediate) benefits a permissive attitiude toward risk convey, and security. These goals are fundamentally antithetical. Nor should we feel that we have no choice. Precautionary regulation represents a genuine option, albeit one that may require significant sacrifice, at least in the short term. The decision between precaution and living on the brink is ultimately about what we value most.
