I recently attended an event at the Royal Society of Arts, where I am a Life Fellow, in which Michael Blastland and David Spiegelhalter presented some of the results of their work in analysing actual statistics of various risks and the perceptions people have of the likelihood of such risks. Some of us approach such issues methodically through the examination of statistics. Others of us prefer to assess such situations through the stories we have heard about such matters. The authors contend that understanding chance and risk needs both. They make a good combination because Michael Blastland started out telling stories as an English Literature graduate working in journalism but he moved on to work in the business of explaining numbers to people, devising the excellent Radio 4 programme about numbers, More or Less
, and writing, with Andrew Dilmot, The Tiger that Isn’t
, a guide to numbers in the news. David Spiegelhalter is a statistician who is the Winton Professor for the Public Understanding of Risk at Cambridge University. He also turns up regularly on Radio 4 to comment on some statistical quirk.
In their recently published book The Norm Chronicles: Stories and Numbers about Danger[i]
they tackle this subject head on with a series of stories with fictitious characters - the strictly average Norm, the careful Prudence and the reckless Kelvin, who face a life full of hazards, both fun and fatal. In the process they challenge you to find out which is the bigger risk – an all-body health scan or standing a mile-and-a-half from the epicentre of the Hiroshima atom bomb.
There is not much in literature or the media about non-events. Storytelling is about exceptions. We are unlikely to read in the newspapers the headline: “No children hurt on way to school!”
Anton Chekhov understood one of the basic tenets of storytelling when he said in 1889 “One must not put a loaded rifle on the stage if noone is thinking of using it.”
Risks are overwhelmingly small but the emphasis is always on something happening. The Daily Express
stated in 2012 ‘A daily fry-up boost
s your cancer risk by 20 per cent’
about the effect of processed meat. This ’20 per cent up’ is a calculation based only on things that happen. It begins with people who have or will have pancreatic cancer and ignores all those who don’t or won’t. Five people in every 400 typically develop this type of cancer. If everyone
of those 400 people eats an extra fry-up every day the 20 per cent increase takes the number getting pancreatic cancer from five to six. The absolute risk has moved from 5 in 400 to 6 in 400, an increase of 0.25 per cent. The number remaining clear has fallen from 395 in 400 to 394.
To help us understand the scale of risk the authors have adopted the concept of the MicroMort[ii]
, which is a 1-in-a-million chance of death. In this way risks are reduced to a micro or daily rate on a consistent scale. According to the Office for National Statistics, 18,000 people died from ‘external causes’ in England and Wales in 2010. That is all those people – out of the total population of 54 million - who died from accidents, murders, suicides etc. This corresponds to an average of 18,000/54 = 333 MicroMorts per year for each person, or roughly 1 a day. This is not a perfect measure but gives us a good way to compare risks provided we are consistent.
So let’s make a few comparisons.
Walking is 1 Micromort per 27 miles. Cycling is similar. But riding a motorbike is I per 7 miles, about 4 times the chance of death over the same distance. Since we are more likely to ride a motor bike further than we will walk the risk increases again. Driving a car however is just 1 MicroMort per 333 miles and going on a train 1 MicroMort per 7,500 miles. Flying on commercial aircraft is the same risk as a train but again our journeys are likely to be longer which increases the risk. At the extremes of travel flying in Bomber Command in World War II was 25,000 MicroMorts per mission!
Looking at some typically hazardous pursuits which, it might be argued, are entirely discretionary, scuba diving carries a risk of 5 MicroMorts per dive, hang gliding 8 per jump, rock climbing 3 per climb, sky diving 10 per jump and running marathons 7 per run. But climbing in the Himalayas has been variously estimated at between 12,000 MicroMorts per climb and 43,000 MicroMorts per climb. By the end of 2011, 219 people were known to have died climbing Mount Everest, 1 for every 25 who reached the summit. Of the 20,000 mountaineers thought to be climbing above 8,000 metres in the Himalayas between 1990 and 2006 it was estimated that 238 died, a rate of 12,000 MicroMorts per climb. But in a study of 533 mountaineers on British expeditions above 7,000 metres between 1968 and 1987, there were 23 deaths (1 in 23) equal to about 43,000 MicroMorts per climb, i.e. riskier than a bombing mission in the Second World War. Climbing Mount Everest has, if anything, got more dangerous as the higher peaks have become overcrowded with inexpert tourist climbers.
Aging also plays its part in assessing risk. One of the most dangerous times of life is the first year, roughly that of riding 30,000 miles on a motor bike, or once round the world. Those who get through it will not face such dangers again, on average, until their mid-fifties. Actually most of this danger is in the first few weeks of life and after that risks plummet from 4,300 MicroMorts in the first year to fewer than 100 per year for a seven year old, less than a quarter of a MM per day. This makes age 7 the safest year in a human’s life. At this point 0.01% will not see their next birthday, by age 32 this has increased to 0.1%, by age 60 to 1.0% and by age 83 to 10%. From age 25 to 80 this trend follows a straight line where the risk of death in the following year increases by 9% a year. This was first discovered in 1825 and has remained constant despite vast improvements in mortality rates.
Benjamin Gompertz, who discovered this ‘law of mortality’, was himself barred from university as a Jew but tried hard to defy his own discovery and lived to 86. We can link an alternative metaphor to this, the speed of aging. The annual risk of dying doubles every 8 years, i.e. the trend is exponential. Therefore one year of life increases the hazard ratio by 1.09. Consider the following examples: