Methods in PRA III: Black Swans and Unpredictability Flashcards
What leads to possibilistic risk assessment?
Raw uncertainty.
What is political about unpredictability?
Labelling risk as unpredictable uncertainty is a political act.
How can uncertainty be managed?
Uncertainty can be managed through simulation, insurance and acceptance.
Can we predict when something is uncertain?
There are problems with data, method and human bias when we are operating in unpredictable environments. Some scholars argue that we can start predicting the future as it is an issue of low predictability. Others (Furedi, Taleb & Makridakis etc.) refer to it as an epistempic uncertainty and therefore we cannot predict.
What are two types of uncertainty?
- Epistemic Black Swans: uncertainty is due to insufficient (or lack of) knowledge - we cannot know how the risk is distributed in advance.
- Aleatory (‘perfect storm’): we do know probability but not what happens when they interact, overlap or coincide - randomness i.e. no distribution.
What kind of event can be labelled as a black swan?
A low probability/ high impact event. It is an epistemic uncertainty, a surprise that it occurs and a huge impact. Afterwards we can rationalise, but before they occur we do not know anything about the probability/ distribution and therefore the forecasting methods are insufficient.
What is the relationship of ‘big data’ with black swans?
We gather more and more data as we hope it will ‘speak to us’ and allow us to predict - hope that trends will emerge from big data. e.g. think surveillance of mass populations (governmentality lecture) or information security & public private partnership session in SRM.
How do ‘narratives’ fit into uncertainty?
More knowledge increases uncertainty as it creates awareness of risk, making the knowledge impotent, leading to powerlessness - culture of fear. (Scenarios rely on the problem of constructing narratives).
What are the politics of unpredictability?
- Choice is political - allow some actions and defy others - debate: low predictability OR complete uncertainty?
- Labelling is exclusionary (categories - think governmentality).
- Individuals, institutions etc are made responsible - when something is labelled unpredictable.
- Think of speech act theory: something is made more or less important. Securitization act will always pinpoint a responsible agent.
- A possibility speech act refers to the unknown = ucertainty.
How can we risk manage the unpredictable (3 ways)?
- Scenarios/Simulation: imagine all the black swans out there, create risk portfolio and prepare/mitigate for them all. But this is very resource consuming - and is this even possible?
- Diversification and insurance: hedging/ buffers/ insurance/ securitisation -buffer to deal with volatility, but ignores the problem of predictability. Requires that protected value is flexible (liquidity) and diversifiable - what is not must be monetized. It needs to be insurable. Problem is that most people are not ready to accept the buffers (idle money), they would have to be made to believe that there is an actual threat present.
- Accepting volatility: accepting uncertainty instead of trying to control and supress volatility. Trying to control volatility may increase the risk of black swans occurring - probability may increase - fragility of the complex system that is the problem, not the black swan itself. - Resilience?
How do we differentiate in the face of unpredictability?
Decision to insure something or not is also a value choice and therefore political.
Is unpredictability in itself necessarily negative?
No. Throughout history uncertainty has been linked with opportunity - precautionary risk and possibility. Why has uncertainty now merged with precaution?
What does it mean that something is unpredictable?
That we do not have the knowledge or means to accurately predict an outcome of something. When talking about risk this refers to probability as well as possibility when talking about future uncertainties.
What do Goodwin & Wright (2010) argue?
That predictability is the capability of carrying out prediction tasks in a valid way. This assumes that there is a way of tackling unpredictability by using the right combination of prediction methods. They argue for the combination of devils advocacy with dialectical inquiry, paired with methods such as Delphi method and scenario planning.
What to Makridakis and Taleb (2009) argue?
That there is no means of forecasting accurately and that future uncertainty cannot be corrected, assessed and controlled. Epistemic Black Swans. They also argue that what we should do is accept this and deal with what we can predict and control, follow protective/ proactive, and preparation-based strategies (resilience measures) and acknowledge the possibly large impact of back swans.
