Cold numbers

ISIS, bloodshed in Ukraine, conflicts in Africa. It is hard not to become a pessimist when you turn on the news or open the newspaper. However, according to the prevailing Long Peace Theory we are living in the most peaceful times in history. Never mind the headlines, say a growing number of sociologists. Their theory basically states that since the frequency of wars among great powers dropped since World War II, the chances of getting killed in a war has also plummeted.

To Pasquale Cirillo and Nassim Taleb these statements are a thorn in their side. In their recently published article ‘On the tail risk of violent conflict and its underestimation’ , they say that they find no trend away from large conflicts and no evidence that large wars cluster together like episodes of financial volatility are known to do.

With help from historians, Cirillo and Taleb took up the gruesome task to analyse war deaths in conflicts with at least 2,000 casualties all the way back to the birth of Christ. They accounted for changes in population size throughout history.

Their main conclusion as written in their abstract: “We find the estimated mean likely to be at least three times larger than the sample mean, meaning sever underestimation of the severity of conflicts.” Or put in simple words: Wars follow no trend and simply occur with equal likelihood through time.

The researchers are specialised in heavy tailed phenomena, a challenging field in statistics. These are occurrences that do not follow a normal distribution; they are extreme in size and rare over time. Record-breaking insurance losses or financial returns are examples, but also wars and earthquakes.

Cirillo and Taleb continue on their work. They have two students in history in the United States who are collecting extra data. One of them is also a captain in the American army. They are collecting data on conflicts that amounted to 500 deaths or more. “Of course this is a complicated task. Numbers of causalities are often underestimated. And there are many conflicts that go unnoticed to historians.”

Cirillo studied economics and finance. “I prefer working on finance – which is what I mostly do – because it is less gruesome. But I think it is important to do this statistical work on conflicts. It can be dangerous to underestimate the risks of war.”

Big quarrel

A famous proponent of Long Peace Theory is Harvard psychologist Steven Pinker. With his book The better angels of nature: Why violence has declined (2011) he gained a lot of attention. There is a big quarrel online between Nassim Taleb and Pinker. Taleb refers to Pinker as a statistical illusionist. Pinker says that Taleb’s “allegations of statistical malfeasance [are] products of dyslexia”.