War Games

On the streets of Chicago, two rival gangs find themselves on the verge of battle. Gang A has possession over a territory—a lucrative drug market—that Gang B is interested in seizing. Gang A wonders whether its rivals will attempt to capture the area. Gang B wonders how long the defenders could hold out in a turf war. 

The battle lines are set. So will there be violence? And if so, what will it look like?

But the gangs don’t have to be gangs, necessarily. They could be oil cartels, or militant groups—any factions that seek to profit from a given region. And it doesn’t matter whether that region is in Chicago or Cairo, just so long as it’s on the perimeter of a circle, near five other regions the same size. Why? Because this scenario is playing out in the mind of Ethan Bueno de Mesquita, professor and deputy dean for research and strategic initiatives at Harris. It’s an iteration of his newest game theory model, a dynamic mathematical machine that’s shedding new light on the shadowy world of factional violence.

“I wanted to write down a model that included the idea of geography, that included the idea of territories from which factions were extracting economic profits and that included the idea that factions might fight over control of that territory,” Bueno de Mesquita explains. The result was his newest paper, “Factional Conflict and Territorial Rents,” in which he uses game theory to see how changing conditions might cause more or less conflict over a particular territory. Abstract as it sounds, this model adds a very real strategic dimension to our understanding of violence. 

The motivations behind violence can seem simple. When more groups are involved, or when there’s more to be gained from the territory in question (from oil production, drug sales, etc.), then we typically expect to see violence escalate. 

At least, those were the ideas that Bueno de Mesquita saw repeated time after time in news reports about gangs in Chicago. “The story in the press was that this run-up in violence was the result of increased factionalization – that you had a change in the structure of Chicago street gangs, from a few very large gangs to a lot of little gangs, each of which controlled a very small amount of territory – and that this was inevitably going to lead to an increase in homicides that we then saw,” he recalls. “I was interested in probing that intuition and thinking about, ‘Well, why should that be true?’” 

He began working on a sophisticated model of conflict that would let him tweak a wide range of parameters. When markets grew, more factions were involved or nearby territories traded hands, he would adjust the model accordingly, and it would spit out a formula estimating the potential payoff for each faction. Based on those payoffs, Bueno de Mesquita could predict how much each faction would invest in conquering or defending the vulnerable territory. “In the case of a Chicago street gang, I literally am making a choice: do I want to buy more guns? Do I want to send my people away from selling drugs and off to take another territory, knowing they might be killed, knowing they might be arrested?” It can seem bizarre to think of gang fights as investments, Bueno de Mesquita concedes, “but what it looks like in the world is buying guns and using them.” A conflict breaks out when two or more factions decide to make such an investment. 

According to the model, conventional expectations about violence often miss the mark. For example, take the basic turf war, one of a dozen scenarios that Bueno de Mesquita fed into his mathematical contraption. He found, to his surprise, that Gang A will often cede the territory without a fight, particularly if the stakes are high. It’s a matter of expectations, he explains. If Gang A and Gang B both control large areas, then they face little competition, and can extract large profits. That means the coveted territory is extremely valuable to whoever has possession. It also means that each side has immense resources at its disposal, so any conflict would be quite costly for both gangs, and both gangs know it.

As Gang B begins piling up guns for a fight, Gang A, wary of an arms race, will likely back down—what Bueno de Mesquita calls the “scare-off” effect. “When I really, really want something, and the other guy knows I really, really want it, sometimes he runs away,” he says. 

Bueno de Mesquita also looked at changing numbers of factions and found that contrary to the media’s portrayal, total violence does not always increase when more factions get involved. In fact, every additional faction further divides the market, meaning there’s less profit to be earned – and less incentive to invest heavily in conflict. “When you’ve got a bazillion little factions, every territory is not so valuable,” says Bueno de Mesquita. The resulting violence is more frequent but less intense.

The author didn’t expect his findings to differ so radically from the conventional wisdom. “I changed my intuitions a surprising amount in the exercise of doing this model,” he admits, “which is not always the case.” 

Bueno de Mesquita hopes the paper will have the same effect on fellow scholars. For instance, his model shows that shocks to the economy of a territory spill over into neighboring territories, which presents a problem to empirical researchers, who typically draw region-to-region comparisons. “That’s the hope with this paper, that one of the things it’ll do – besides giving intuitions about how things like gangs work – is say to empirical researchers, ‘Let’s use this theoretical model to help us think about how to avoid the next set of problems.’”

These results offer a fresh perspective on today’s geopolitical landscape, and have already helped the author to make sense of otherwise confusing current events. “I started working on this paper at the same time that ISIS had its first territorial expansion,” he recalls. “When ISIS first started taking big cities in Iraq, when people thought the Iraqi army would head them off, you had these amazing cases where the opposition simply melted away. ISIS showed up ready to fight, and there was nobody there to fight. They completely walked away from the cities and ceded the territory.”

“We think, ‘No, everybody always stands and fights for things that are valuable!’” he says, “but that’s not right.”