Gray swans: comparison of natural and financial hazard assessment and mitigation

Even advanced technological societies are vulnerable to natural disasters, such as the 2011 Tohoku earthquake and tsunami, and financial disasters, such as the 2008 collapse of the US housing and financial markets. Both resulted from unrecognized or underappreciated weaknesses in hazard assessment and mitigation policies. These policies relied on models that proved inadequate for reasons including inaccurate conceptualization of the problem, use of a too-short historic record, and neglect of interconnections. Japanese hazard models did not consider the possibility of multiple fault segments failing together, causing a much larger earthquake than anticipated, and neglected historical data for much larger tsunamis than planned for. Mitigation planning underestimated the vulnerability of nuclear power plants, due to a belief in nuclear safety. The US economic models did not consider the hazard that would result if many homeowners could not pay their mortgages, and assumed, based on a short history, that housing prices would keep rising faster than interest rates. They did not anticipate the vulnerability of the financial system to a drop in housing prices, due to belief that markets functioned best without government regulation. Preventing both types of disasters from recurring involves balancing the costs and benefits of mitigation policies. A crucial aspect of this balancing is that the benefits must be estimated using models with significant uncertainties to infer the probabilities of the future events, as we illustrate using a simple model for tsunami mitigation. Improving hazard models is important because overestimating or underestimating the hazard leads to too much or too little mitigation. Thus, although one type of disaster has natural causes and the other has economic causes, comparison provides insights for improving hazard assessment and mitigation policies. Instead of viewing such disasters as unpredictable and unavoidable “black swan” events, they are better viewed as “gray swans” that—although novel and outside recent experience—can be better foreseen and mitigated. Copyright Springer Science+Business Media B.V. 2014