Under SARF, as humans communicate to each other, the impact of a risk event can amplified through points of contact and under certain conditions where there is fear that the risk may have wide-ranging (societal) implications and if the risk itself is not understood well.
Certain events may be interpreted as signals or clues regarding the magnitude of the risk and the adequacy of the risk management process. When amplified, they cause ripple effects, impacting people and businesses in heretofor unexpected ways or in unexpected magnitude.
Consider consumer reaction to the Tylenol poisonings in the 1980s. Tylenol tampering resulted in more than 125,000 stories in the print media alone and inflicted losses of more than $1 billion upon the Johnson & Johnson company, including a damaged image of the product. Further, consumer demand and regulation following this risk event led to the ubiquity of tamper-proof packages (and associated costs) at completely unrelated firms.
Or, consider the reaction to this decade's terrorist threats, which has led to an enormous cost on all who travel, businesses wishing to hire foreign talent in the United States or businesses involved in import/export.
In 1998, the Asian currency and Russian debt crises had ripple effects that led to the demise of the hedge-fund Long Term Capital Management (LTCM). This demise, in turn, was perceived as having the potential to lead to a catastrophic disruption of the entire global capital markets system and resulted in substantial financial losses (and gains) for firms that believed they had no exposure to either Asia or Russia and certainly not to hedge funds.
In 1992, the researchers who conceived of SARF evaluated their theory by reviewing a large database of 128 risk events. In their study, they found strong evidence that the social amplification of a risk event is as important in determining the full set of risk consequences as is the direct impact of the risk event. Applying this result to internal risk assessments suggests that it would be easy to greatly underestimate the impact of a risk event if only first order effects are considered and not the secondary and tertiary impacts from social amplification or communication and reaction to the risk event.
Again, considering the Tylenol tampering case, an internal risk assessment of a scenario that included such an event might result in the risk being limited to be legal liability from the poisonings and perhaps some negative customer impact. However, it would be unlikely that any ex-ante analysis would have concluded the long-term impact on product packaging and associated costs that were a result of the amplification of the story. Or, if the scenario had involved such an event at a competing firm, the impact might have even been assumed to be positive for the "unaffected" firm.
So, what are the factors that can increase the likelihood of social amplification? How are hazards or risks perceived?
It turns out, not surprisingly, that what people don't understand and what they perceive as possibly having potentially wide-ranging effects are the things they are most likely respond to with some kind of action, e.g. a change in valuation.
One very important method for understanding how people react to risk is psychometric paradigm which has identified people's emotional reactions to risky situations that affect the judgments of the riskiness of events that go beyond their objective consequences. This paradigm is characterized by risk dimensions called Dread (perceived lack of control, feelings of dread and perceived catastrophic potential) and risk of the Unknown (the extent to which the risk is judged to be unobservable, unknown, new or delayed in producing harmful impacts).
Psychology Professor Paul Slovic developed a dread/knowledge spectrum which represented below. This chart depicts the factors that contribute to feelings of dread and knowledge.
The Dread/Knowledge Spectrum (click to enlarge)
"Dread risk", captures aspects of the described risks that make us anxious: lack of control over exposure to the risk, consequences that are catastrophic, or have global ramifications. "Unknown risk", refers to the degree to which exposure to a risk and its consequences are predictable and observable: how much is known about the risk and is the exposure easily detected. In other words, is there transparency and an ability to manage the risk?
The most important factor of these two is Dread risk. The higher a risk's score on this factor, the higher its perceived risk, the more people want to see its current risks reduced.
From this depiction, we can recognize that both dread and our lack of familiarity with something will likely amplify the human response to a risk event. In other words, risks that are in the upper right hand corner of the dread/knowledge chart are the ones most likely to lead to an amplification effect.
On its current trajectory, the 2007 subprime mortgage crisis seems a classic example of a risk event being amplified. Clearly it first affected those with direct holdings of sub-prime mortgages, but quickly went on to affect general liquidity being provided to financial service companies. Next has come damage to the specialized insurers, industries that supply the home marketplace and more generally to equities worldwide as a falling-home-price induced recession in the US is expected to have global implications.
How many trillions of dollars of equity valuation worldwide have been lost since August on what is expected to be a few hundred billion in subprime losses?
One implication of the signal concept is that effort and expense beyond that indicated by a first-order cost-benefit analysis might be warranted to reduce the possibility of high signal events and that transparency may be undervalued, underappreciated or improperly feared. Perhaps it is too late to close that barn door, as an American expression goes. But, something must be done to reduce the Dread and Unknown factors around the subprime crisis.
You've seen many call for full transparency...even security-level detail. If such had been in existence when the first losses emerged, would the panic have been amplified? It seems clear, even now, that the impact of the subprime losses are still being amplified.
Message to every bank and regulator: someone who knows the true exposures needs to reveal all. The message from SARF is that transparency may actually help your firm. In the meantime, without such, you're killing the rest of us.
For more on SARF, see:
Kasperson, J.X., Kasperson, R.E., Pidgeon, N. & Slovic, P. (2003). The social amplification of risk: assessing fifteen years of research and theory, in The Social Amplification of Risk, edited by Pidgeon, N., Kasperson, R.E., & Slovic, P., Cambridge Press, Cambridge, UK.
Slovic, P. & Weber, E. (2002). Perception of Risk Posed by Extreme Events, prepared for discussion at "Risk Management Strategies in an Uncertain World", April 12-13.
Weber, E. (2001), Risk: Empirical studies on decision and choice, in International Encyclopedia of the Social & Behavioral Sciences, Elsevier Science, Ltd
Weber, E. (2004), Who's afraid of poor old age? Risk perception in risk management decisions, in Pension and Design Structure by Mitchell, Olivia S. & Utkus, Stephen P., Oxford University Press
The citations above were used in developing this posting. A wide range of research in this area has been published, but mostly outside of finance. See work by Elke Weber at Columbia for specific attempts to model the impact and perception of risk in financial realms.
