Toasters & Terrorism

With heated debates on the topics of terrorism, the migrant crisis currently engulfing Europe, and the impact of  open borders becoming more common place recently, it’s likely you’ve come across news articles claiming that the chances of being a victim of a terrorist attack are extremely low, even trivial to account for.

You may have heard mainstream news outlets pontificating – sometimes in a derisive and holier-than-thou tone – that you are more likely to be killed falling from a ladder, being struck by lightning, or being electrocuted by a home appliance such as a toaster. Apparently even husbands are more deadly than terrorists, so if you’re raising a fuss about the dangers to society from rampant terrorism, while married to a man, you’re obviously illogical and inconsistent, right?

The reasoning is that if you drive a vehicle, make french toast, climb a ladder to clean your eavestrough, or get married to a man, you’re putting yourself in far more peril. You’re incessant worrying about a possible terrorist attack is, therefore, irrational.

the truth is that people who make arguments equating deadly home appliances with terrorism are employing a faulty understanding of risk.

The central concept that needs to be illustrated is the conflation of normal distributions with fat-tailed distributions.

When examining things such as university grades, national IQ levels, blood pressure, and heights of people, we see that the distribution of the data points tend to cluster around a mean (average) with relatively few points collecting at the extremes on both ends (called tails).

For example, the height of most men will be at or near 5″10, with few men being 4″8 and 7″2. There are very clear observable upper and lower bounds when it comes to the range of height in the male population (no man is two inches tall or twenty feet tall). In addition, the impact of the extreme outliers is negligible in terms of the whole population (the height of the tallest man in the country will not even come close to the combined height of the rest of the men).

The distribution of the data points resembles a bell, from which the term “Bell Curve” is derived from. It should be familiar to anyone who has studied basic statistics.

When examining things such as pandemics, hurricanes, and major terrorist attacks, a lesser known type of distribution is required to properly map out the risk and impact: the fat-tailed version. Here, the two tails at both ends are fat, indicating that relatively few data points at the extremes have a significant impact on the overall distribution.

For example, all of the major losses in the financial markets will come from relatively few events such as Black Monday and the Dot-Com Bubble. Similarly, most of the deaths resulting from an infectious agent will be caused by one deadly pathogen such as the Black Death and the Spanish Flu.

How can we use this to analyze risk when it comes to things such as falls from ladders and terrorist attacks?

The answer is we must examine the probability and magnitude of these events. We must also examine how reliable the statistical models employed are at predicting and measuring the risk of such events occurring.

We can reliably measure what the overall number of deaths will be as a result of people falling from ladders. We can be relatively certain that the number of people killed every year will be at or close to the historical average. We can comfortably assert that the magnitude of each death is negligible and inconsequential as it relates to all other deaths from ladders that year. The overall effect on society will be also small or non-existent.

In the world of normal distributions and thin-tailed events, things don’t change that much from year to year.

When it comes to a major terrorist attack, however, the story is different.

We can’t reliably measure what the overall number of deaths will be as a result of people being killed by terrorists; such events don’t fall neatly into a normal distribution. We can’t be certain that the number of people killed every year will be at or close to the historical average. The reason is because a major terrorist attack such as the one on the World Trade Center on September 11, 2001, where nearly 3000 people died, heavily skews the the data, thereby rendering it’s predictive value virtually useless.

One single attack can be responsible for the vast majority of deaths linked to terrorism, and not only during the course of a single year, but more likely a decade or more. Even though the risk of such a major, unprecedented attack is rare, it only needs to happen once.

The chances of the number of deaths caused by toasters or ladders increasing ten fold or hundred fold during the course of a year is nearly zero (it would take all the producers of toasters and ladders to collude and purposefully manufacture models that are unsafe to a nefariously high degree), but the chances of a terrorist attack being deadlier by this many orders of magnitude is highly plausible.

In the case of 9/11, it is plainly obvious that the magnitude of this one attack was colossal and has had a profound effect on the United States across many dimensions: economic damage, increased expenditures on security, dramatic change in foreign policy (leading to the second Iraq War and the continued occupation of Afghanistan to this day), etc.

If you’re old enough, think back to life before and after the 9/11 terror attack. You’ll very quickly realize how that single, solitary event changed almost everything.

In addition to the immediate death and destruction, such large-scale attacks are also socially enervating, resulting in increased paranoia, psychological distress, elevated hostility towards immigrants, and the possibility of a breakdown in social cohesion.

With terrorist attacks on the rise in Europe, there is also the risk for deadly network effects to manifest, multiplying the number of attacks as well as increasing their magnitude (something toasters and ladders are not capable of). As social cohesion and trust erodes it is likely that militant, far right nationalist groups will emerge, launching retaliatory terrorist attacks Muslims and other minorities, as well as anyone heavily associated with left wing political activism (think Anders Breivik).

Let’s hope we find solutions so we don’t go down that dark path. However, making erroneous statistical analogies such as comparing the deaths of people falling from ladders versus terrorist attacks is not a good start. It is critical that we examine the risks posed by rare, fat-tailed events in their own separate domain. After all, these are the ones that ultimately shape and have the potential to devastate societies; we should not take them lightly.

Be the first to comment

Leave a Reply

Your email address will not be published.