The randomness of deadly attacks in the small-scale insurgencies that currently define war may not be so random after all. A new paper suggests an “ecology” — or perhaps it’s a pathology — that unifies attacks and terrorist acts regardless of their ideology or locale, and that understanding this ecology might help shorten conflicts like the growing war in Afghanistan.
The authors say that the size and timing of “violent events” show statistical patterns that are consistent whether it’s Northern Ireland or Iraq, Palestine or Peru. (These patterns do not hold true, they stress, in conventional uniformed-army vs. uniformed-army wars.)
The researchers – engineer Juan Camilo Bohòrquez of Colombia’s Universidad de los Andes, economist Michael Spagat of London’s Royal Holloway College, physicists Sean Gourley and Neil F. Johnson with the University of Miami’s Complex Systems Group, and physicist Alexander Dixon of Cambridge — also believe they’ve created a mathematical model, that, as Gourley explained, allows them to “understand the inner workings of an insurgency — how they form groups, how they make decisions, how many groups there are, etc.”
Understanding these complexities could help craft the best policies to tamp down future insurgencies — or help insurgents who are good at math to “optimize” their organizations.
Their findings mimic another messy human endeavor, finance, and may find a role to help fight complex battles without guns and bombs, such as fighting disease.
“Both finance and conflict involve understanding the collective actions of large groups of individuals,” Gourley explained via e-mail. “These individuals form groups and make decisions to optimize their own utilities. There is also intense competition between the groups with only the successful groups surviving. In many ways, it would be surprising if we didn’t find similarities between these two types of ecosystems. Indeed if we look to at the financial structure of the Somali pirate ecosystem we can see the lines between insurgency and finance start to become very blurred.”
Still, a power law is no panacea for peace. Grasping their finding, the five stress in a cover article appearing this week in the journal Nature, is not the same as understanding the causes or origins of war — “one of the ‘messiest’ of all human activities to analyze.”
One Percent Terror
A participant in the American Civil War is credited with the aphorism, “Soldiering is 99 percent boredom and 1 percent sheer terror” – a timeless observation in the hurry-up-and-wait world of the military but apparently provable in today’s asymmetric wars, too, as Gourley discovered while in Baghdad.
“For the most part, nothing happens and the city feels almost normal. But then an explosion will go off and all hell will break loose with screams, crying, AK-47s being fired and sirens ringing out across the street. Then silence, and nothing happens again.”
Analysis in this case confirmed the anecdote. Starting with work in Colombia and Iraq around 2005, the authors noticed similarities in the size, i.e. body counts, of attacks despite the obvious differences between those insurgencies. Such patterns had been noted by those studying traditional warfare, but it was a bit of surprise to see it in the skulking realm of insurgency.
Later, as the researchers stirred in more conflicts — nine are in the paper and more than 54,000 “violent events” are logged — and saw how “well behaved” the numbers were, they started looking for other patterns. A very strong “bursty” phenomenon appeared (after applying a “power law,” a statistical relationship that refers to the exponent in the equation and not to politics) in the timing of attacks.
“What this means,” Gourley explained, “is that instead of being randomly distributed through time, we get long quiet periods where nothing much happens and then periods of time where there are a large number of attacks. This also means that if you are in a war zone and it is a quiet/low-violence day, it is more likely to be a quiet/low-violence day the next day.”
Helping this distribution is that it’s tougher to kill a lot of people than it is to kill a handful. At the same time, a glutted international media requires an ever higher body count to ward off coverage fatigue, while the insurgents’ social media wants to see evidence that their guys are out there killing.
As counterinsurgency expert David Kilcullen has said about insurgent attacks in Iraq, “They’re not doing that because they want to reduce the number of Humvees we have in Iraq by one. They’re doing it because they want spectacular media footage of a burning Humvee.”
In their Nature paper, the authors acknowledge the media as providing oxygen to some conflicts and to terrorist attacks, but they approach that characterization with caveats.
“So if you were to remove media coverage you would potentially see a different distribution of attacks in time,” Gourley wrote. “We know that media plays a role in the timing of attacks; it is harder to know whether this makes it ‘complicit’ in an insurgency’s violent attacks. Of course, our model is based upon the competition for a limited resource. In insurgent conflicts, this is often media coverage — but it could be something else.”
Competition for resources echoes another field, financial trading, where bursty trading is dubbed “clustered volatility.” The warfare model mimics well-established trading models, in part because it was built on some of the authors’ work in finance.
“Instead of a bursty pattern of attacks,” noted co-author Johnson, “one would get a bursty pattern of large movements. … Deep down, both are competition for a limited resource (good press for conflict, or good price for traders).”
Dateline Kabul
Tracking and examining specific, identifiable violent incidents is a straightforward exercise if you ignore the immense difficulty of getting reliable data on death tolls. (The authors intentionally did not collect data on injuries since those are harder to cross-check.)
But explaining how or why these events emerge — and developing an equation that generates numbers that replicate what’s really happening on the ground — is more difficult. The researchers opted to study the insurgents: how many groups of how many members there were, how effective they were at killing people, how they coalesce and fragment, and how they arrive at decisions.
Tracking and examining intentionally secretive insurgent groups is tricky by definition; by their nature these warring parties are “fragmented, transient and evolving,” the authors admit. However, by having determined that “common ecology,” the researchers say their model not only can reproduce real-life distributions of attacks, but can also account for changing rules of engagement by those fighting insurgents.
And while the model presented in Nature assumes the numbers of insurgents and those fighting them remain the same, the model does allow increases, say a surge, and decreases, say a Predator strike.
This flexibility lends itself to real decisions about policy or strategy.
“Using our model of the insurgent ecosystem, we are able to predict the likely effects of strategic decisions, e.g., we can change the model to add in a third population of peace keepers and see what happens to the volume of violence,” Gourley wrote. “Or we can increase the number of troops in a country like Afghanistan and see what happens to the duration of the war.”
So what of Afghanistan, where U.S. President Barack Obama is increasing the Western presence there by 30,000 troops?
While Gourley did not answer using specific numbers, he demonstrated how the model finds that adding more troops – but not adding enough — might actually lengthen the war. Adding enough can, as hoped, shorten the conflict. “Enough” refers a ratio between counterinsurgents and insurgents of roughly 15 to 1. Interestingly, since the model accounts for the “strength” of various fighting forces, the ratio isn’t automatically a reflection of boots on the ground but how forces stack up against each other. The key, it seems, to asymmetric conflicts is maintaining — or defeating — that asymmetry.
Just adding troops is not the only way to tailor the war, even within the limited parameters of the researchers’ equation. Those fighting the insurgents might focus on medium-size insurgent groups instead of the “tall poppy;” they might interrupt insurgent communication, or they can attempt to change the way media are used to spur attacks – all things that can be tried out on a laptop before putting lives in further danger.
While such uses hone in on the battlefield, the researchers see applications beyond.
“These mechanisms are potentially very general and are likely to apply to other conflictual situations, not just war,” explained Spagat. “Neil Johnson is already working on applying these ideas to battles within human bodies between immune systems and pathogens.
“I suppose a successful application shouldn’t be bad … as long as we don’t become cocky and underestimate the new challenges of a new application.”
Sign up for our free e-newsletter.
Are you on Facebook? Become our fan.
Follow us on Twitter.