(credit: 272447 /Pixabay)
We sometimes treat similar risks very differently, and that affects how we make decisions
Fear of flying is not likely to be something bothering many people any time soon. The latest IATA figures show that, after a drop in global demand for air travel in February of 10% year-on-year, the decline was way more precipitous in March: it was down by 56%. The figures for April are not yet available, but FlightRadar24 (which tracks all flights globally) reports it shrank by nearly 74%. This includes cargo movements, so whatever remains of passenger flights will be negligible.
Fear of flying is often cited as an irrational fear, on the basis that the objective risk of perishing while in an aircraft disaster is comparable to the risk of dying while travelling to or from the airport, especially if that is done by car (a 1000 mile flight poses about the same risk as driving 250 miles). You could indeed argue that categorically avoiding one risk, but happily accepting a similar or larger risk is irrational. But is that really so?
Risk: more than just figures
The risks may be equivalent, but there is more to it. Should something go wrong, the prospect of spending your last minute or so in an airliner going down, well… I can kind of see why some people might have a strong preference to avoid such a fate. And before – if you are lucky enough not to suffer from aerophobia – you laugh at people refusing to travel by air, consider this. Would you happily jump out of a plane in free fall for thirty seconds, before eventually engaging your parachute? If you drive a pretty average 8,000 miles per year, you will have exactly the same risk of dying. And yet, for most people, it doesn’t feel remotely the same.
If you don’t go skydiving (or travel by aeroplane) then what you are doing is to avoid any possibility of being killed during that activity altogether. Like avoiding hangovers by not drinking alcohol, it is hard to see why this would be irrational. Before we dismiss fear of flying (or of skydiving) as irrational, we should realize that there is more to risk avoidance than just numbers.
If we must see irrationality in refusing to travel by plane, then we have to look elsewhere, for example at the failure to consider the trade-offs in a categorical refusal. By unconditionally rejecting (or indeed accepting) a particular possibility, without even considering the consequences, we are potentially excluding the option that is ultimately best for us. That would be reasonable example of acting irrationally, against our own interest.
There is another facet of risk where a question can be asked about rationality, and it is staring policy makers in the countries that have successfully contained the COVID-19 pandemic right in the face. Imposing a lockdown was easy in several ways. The message was exceedingly simple: stay at home (with very few, pretty clear exceptions). There was also, initially, little worry about the consequences. When the number of new victims of a disease doubles every other day, and the health care system is accelerating towards 100% of its capacity, you pull out all the stops, without asking too many questions.
But once the curve has been flattened, attention turns to the drawbacks of a lockdown, and pressure rises to relax earlier rules. That will certainly make communication more difficult (“if we can go to work, then why can we not visit friends?”). More importantly, it also confronts governments with complex decision-making challenges. Easing the restrictions almost inevitably involves an increase in the risk that some people might contract COVID-19 (and die of it).
Not all the dead are equal
One such decision is the reopening of schools. School age children seem to be a low risk population across most countries. In the UK, the latest figures for March (when the total number of deaths ascribed to COVID-19 was 3,912 – it is around 30,000 now), only one person under 20 was among the dead (in the 15-19 range). This would make allowing young people to go back to school an appealing decision: big benefits (children can get properly educated again, and parents can go back to work), small disadvantages. However, there are nearly 9 million pupils in British schools. Might children actually die if the schools reopen?
Data on influenza (ordinary flu) give us some guidance. A study looking at incidence (the chance of contracting it) and case fatality rate (the likelihood that someone dies, having contracted the disease) shows that, on average, the incidence of influenza for children aged 5-14 is 0.85%, and the CFR is 0.12%. Both the incidence and the CFR for people above 14 are higher, so if we use the same numbers for the entire school age population (up to age 18) we know we are underestimating reality. If 0.85% of 9 million pupils contract the disease, and 0.12% of that number die, about 90 COVID-19 deaths can be expected among the school age population.
How would people react to even a fraction of this figure? Imagine the headlines in the media, even after the first death: “Schools reopen, Abigail (6) dies of COVID-19”. Which policy maker would be prepared to face the cameras and explain that this is not much different from the 60 or so children who die of flu each year?
Now consider another measure: the reopening of businesses, and the corresponding pickup of traffic. Over the last eight years, the number of traffic fatalities in the UK has been around 1800 per annum, or about 34 per week. There are no figures yet for traffic casualties during the lockdown, but with car traffic down by 80%, a reasonable approximation would be to say it has been reduced pro rata, to say about 7 fatalities per week. We can truthfully say that the lockdown has saved 27 lives every week since 24th March – that is 160 people. We can be absolutely certain that, as the economy kicks back into action, more people will die on the roads.
Will we see headlines proclaiming “Lockdown ends, 14 extra traffic fatalities in the first week”? Not very likely.
A decision to ease the lockdown at the end of May, will, by the end of the year, be responsible for well over 500 traffic deaths – people who would not have died on the road had the lockdown been maintained. It may also be responsible for maybe 100 school children dying over the same period, compared to the same baseline.
While the risks of travelling by plane and by car may be roughly similar, we can understand why, to some at least, they appear very different – a matter of preference. But is it just a matter of preference, if we find an increase in the number of deaths among young people of about 100 compared to the present status quo so much worse than an increase of at least 5 times that in traffic fatalities?
Even your correspondent struggles with this question. If you feel the same, you are in good company.