A scholarly debate over the risk of war.
In recent years, Harvard psychologist and bestselling author Steven Pinker has argued that the world is generally becoming a safer place than it used to be. When it comes to war, for instance, he’s shown that interstate conflicts have fallen sharply since 1945—especially among global powers. And that armed conflicts of all kinds have been on the decline since the Cold War. And that annual rates of battle deaths have been far lower in the 2000s than in previous decades.
Pinker and Andrew Mack used these and other figures to show that the “world is not falling apart” in Slate last December:
Wars between states—by far the most destructive of all conflicts—are all but obsolete. The increase in the number and deadliness of civil wars since 2010 is circumscribed, puny in comparison with the decline that preceded it, and unlikely to escalate.
NYU risk analyst and fellow bestselling author Nassim Nicholas Taleb isn’t buying it. In a new working paper, Taleb and probability scholar Pasquale Cirillo of Delft University of Technology refute the claims that the world has entered a “long peace” … with a whole lot of math. By using different statistical methods to evaluate war and violent conflicts, they calculate that we should expect three times more casualty figures than methods like Pinker’s suggest.
Here’s the biggest (and, frankly, most lay person-readable) takeaway from their study of 2,000 years of war:
Contrary to current discussions, all statistical pictures thus obtained show that 1) the risk of violent conflict has not been decreasing, but is rather underestimated by techniques relying on naive year-on-year changes in the mean, or using sample mean as an estimator of the true mean of an extremely fat-tailed phenomenon; 2) armed conflicts have memoryless inter-arrival times, thus incompatible with the idea of a time trend.
Basically the two sides boil down to this: if you take the average annual number of global war deaths since 1945, things seem to be getting better; but if you go back far enough and track the extreme unpredictability of huge wars, as well as changes in population sizes, you find “no particular trend” away from these conflicts—and therefore shouldn’t expect respite from them moving forward. Mark Buchanan has a nice summary in Bloomberg View:
Cirillo and Taleb also found no evidence that wars cluster together, as earthquakes and episodes of financial volatility are known to do. Rather, big wars follow no trend and simply occur with equal likelihood through time. Doing the statistics right, they argue, shows that the recent peaceful past is almost certainly causing us to seriously underestimate how much violent conflict we're likely see in the future.
In the macro sense, there’s a lot on the line with this debate. Either we’re on the right track with democracy and international diplomacy, and should feel upbeat about humanity’s progress. Or we’re right to be wary of future conflicts and prepare for them accordingly.
But there are plenty of considerations that make an either-or outcome less meaningful. The changing nature of war over this long timeline limits the value of sweeping conclusions about armed conflict. More everyday threats to existence—such as homicide, and violence against women and children—do seem to have fallen of late by Pinker’s measures. And in the event you ever find yourself staring down some serious violence, the statistical probability that it should or should not have happened isn’t likely to be much comfort.
H/t: History News Network