Why Risk Managers Need To Understand Human Nature
Fear of Flying 1 – Commercial Airlines
A friend of mine, let’s call him Sam, is terrified of flying. His fear stems from a belief that there is a good chance that the plane will crash and he will most likely die in the accident. Sam is not a risk management professional but he would understand that “risk” is “the possibility that an event will occur and adversely affect the achievement of objectives”. If his objective is to visit London, and you were to ask him to rate the likelihood of “dying in a commercial aircraft accident” - which would obviously prevent him from achieving his objective (ever), and even if he survived the accident it would certainly adversely affect the visit - on a scale of 1 to 5 (where 1 is rare, 2 is unlikely, 3 is possible, 4 is likely, and 5 is almost certain), he would probably pick either 3 (possible) or 4 (likely).
However, the statistics do not back up his rating.
The latest IATA Safety Report 2013 shows that nearly 3.5 billion people flew on 36 million trips and there were 16 fatal accidents with 210 fatalities. 2013 was statistically a good year (although no fatality is good) as the five-year average was 517 fatalities per year. Even using the five year average, the probability of dying in a commercial plane crash is 1 in 6.7 million trips.
To put this in perspective, the death rate for car travel is around 72 times higher. In America, there is more chance of dying while walking (1 in 54,538), from falling down stairs (1 in 157,300) or when cycling (1 in 340,845). Incidentally, if you are worried about the risk of Earth being hit by an asteroid the chances are 1 in 75 million and the most likely cause of death is heart disease (1 in 500).
So the risk of “dying in a commercial aircraft accident” is 1 (rare).
Fear of Flying 2 – Light Aircraft
In spite of his fear, Sam is prepared to travel on a commercial jet, although in a state of high anxiety – especially during take-off and landing. This has some logic as the majority of accidents occur at these stages. The alternative forms of transport for long distances are presumably just too inconvenient for him to contemplate. But he absolutely refuses to fly in a small airplane. His definition of small is anything with less than 100 passengers. A trip to a camp in the Okavango Delta inaccessible by anything other than a light aircraft is a journey he will not take. He has clearly rated the likelihood of “dying in a commercial light aircraft accident” as 5 (almost certain).
Once again, the statistics do not back up his rating.
According to the South African Civil Aviation Authority Accidents & Incidents in 2014 there were 23 fatal accidents with 33 fatalities. Although there were more light aircraft fatal accidents in South Africa than commercial aircraft fatal accidents in the entire world it must be born in mind that these statistics include private flights, training flights, crop spraying and fire-bombing flights. They also include not only light aircraft but also home built aircraft, experimental craft, gliders, helicopters and gyros. So, in 2014, there were no recorded fatalities by commercial light aircraft accidents - with the possible exception of a Beechcraft King Air that crash landed with 3 fatalities in February 2014 whose type of operation is “Unknown”.
Looking at the statistics for Botswana, on the 14th October 2011 a Cessna 208B Grand Caravan operated by Moremi Air carrying tourists in the Okavango Delta crashed shortly after take-off with 8 fatalities. Obviously this was of great concern to the transport minister Frank Ramsden (no doubt prompted by the minister of tourism) who issued the statement “Botswana has one of the best civil aviation safety records regionally. In the last 10 years, we have only had one fatal accident. Statistically, over 90 000 000km have been flown by local operators in the Okavango Delta … including last week's crash Botswana's safety record was the equivalent of 0.022 fatal accidents per million kilometres flown”.
A check of the Aviation Safety Network database for Botswana shows that the last fatal accident of a commercial charter in the Okavango Delta (excluding a cargo flight, the Botswana Defence Force, and Botswana Police Service) was 2001 with 5 fatalities.
So the risk of “dying in a commercial light aircraft accident” is also 1 (rare) and Sam should make that trip.
System 1 and 2
So why is Sam’s intuitive assessment of the dangers of flying so far from reality? In his book Thinking, Fast and Slow, Daniel Kahneman presents a view of how our minds work. He refers to two systems in the mind, System 1 and System 2 – the fast and slow thinkers from the title.
- System 1 handles the things that we do almost automatically like driving a car or interpreting emotions. I wonder whether this is the “lizard brain” we used to jokingly refer to as children.
- System 2 handles tasks that require attention and concentration like filling in a form or complex arithmetic. It is also endorses the impressions, intuitions, intentions, and feelings that System 1 generates but, as it is in charge of self-control, may reject them as well.
Usually System 1 runs automatically and System 2 monitors and adopts its suggestions in a relaxed fashion. When System 1 is presented with an issue it cannot handle, it will defer to System 2.
To quote Kahneman, “The division of labor between System 1 and System 2 is highly efficient: it minimizes effort and optimizes performance.”
System 1 does have some limitations. It can make systematic errors in certain circumstances, it is lousy at logic and statistics and it cannot be switched off.
System 2 also has its short comings. It is lazy, requires effort to engage, gets tired and can give up if overstretched.
Kahneman uses a simple test to demonstrate these limitations.
- A bat and ball together cost R110
- The bat costs R100 more than the ball
- How much does the ball cost?
Intuitively you would probably have answered R10. If you check this answer you will see that if the ball costs R10, the bat would cost R110 and together they would cost R120. System 1 gave what it thinks is the obvious answer and System 2 was too lazy to check thoroughly. The correct answer is R5.
For another example, take a look at the following puzzle.
- Jack is looking at Anne but Anne is looking at George.
- Jack is married, but George is not.
- Is a married person looking at an unmarried person (Yes, No, or Cannot be determined)?
As we don't know Anne’s marital status, 80% of those tested go with the answer Cannot be determined. But if Anne is married, she is looking at the unmarried George, and if she is unmarried, then the married Jack is looking at her. So the answer is Yes.
As aircraft accidents and fatalities almost always make the news and dominate the headlines, Sam has allowed System 1 to assume that they are common occurrences. This is System 1 displaying its bias caused by frequent reminders of how dangerous flying seems to be. System 2 has not bothered to check.
Determining probability is something we are all very bad at. So, when rating your risk likelihood, try and make sure System 1 is not providing the answer (keeping in mind that you cannot switch it off) and that System 2 is fully engaged (bearing in mind that it is lazy).
Graham Blunn, Director, BarnOwl
[i] Committee of Sponsoring Organizations, Enterprise Risk Management—Integrated Framework (2004), p. 16
[viii] Thinking, Fast and Slow by Daniel Kahneman