Engineers have made great strides over the years in reducing the risk of structural failures in bridges, buildings, aircraft and networks. They have examined past disasters and with the support of regulators have used black boxes to capture much if not all of the detailed events leading up to the critical failure in question. Modelling alternatives and testing them has created far greater safety records and a huge leap in public confidence (and anger when things go wrong). Product safety laws have played a very positive part in this improvement to what the market might otherwise deliver.
Biological and social systems have not fared as well and, in countries and organisations with lower standards of corporate governance, risky behaviour, complacency and outright fraud can still be found in abundance. Regulatory capture and weak law enforcement have been major causes of disasters such as those caused by earthquakes, tidal waves, and ecosystem destruction.
On the other hand, it is not unusual to find a growing tendency especially in the welfare dominated societies of UK and Europe, to try and legislate for every eventuality with so called Health and Safety regulations (driven by considerations of insurance cover and potential claims), stifling people’s natural desire to engage in pro bono work, to help fellow citizens in distress, or otherwise put themselves at risk for the sake of the general good.
Part of the problem stems from a human tendency to believe that, if a known risk has not yet manifested itself, it is increasingly unlikely to happen – the very opposite of an insurers view of the world! It also the case that energetic decision makers may often be most likely to believe that they are less likely than most to be impacted and will have a better chance of surviving – and this is what is often reinforced by their compensation package focused on quarterly results and the ‘bottom line’ while shareholders and other stakeholders with a greater interest in the balance sheet and the longer-term look on with increasing concern at the risks being taken.
Statistics can have a big role in reducing uncertainty around potential risk; even more value can come from modelling a being able to home in on the ‘one in a million’ situation that destroyed LTCM, Fukushima, or the last Concorde and identifying which were the five factors which had to come together to create the disaster and what could be done to eliminate them. Risk cannot be eliminated in the real world but better information can help people to accept price and provision against it in a sensible and cost effective way, providing the necessary capital to survive the worst and to continue to modelling proper.