More than a decade ago, political scientist Mark Crescenzi set out to foster peace by studying conflict. He wanted to use science and math to break down conflicts into their basic components.

Crescenzi learned that explaining how conflict happens is much easier than getting people to stop it. In spite of this rude awakening, he maintains that conflict research is valuable. “You have to understand the mechanisms of conflict before you know exactly where to apply pressure to stop it,” he says.

Traditional historians study conflict by examining the details of individual events. Political scientists such as Crescenzi take a broader approach: they compile events from several decades or centuries, identifying the patterns that lead to confrontation. Crescenzi’s strategy reveals trends that can easily be missed when events are studied one by one, he says. “We tend to get too focused on the unique aspects of historical events, which makes it harder to understand how history tends to repeat itself.”

Click to read photo caption.

Consider the Cold War. From the Korean War to the space race, it appeared outwardly that the United States and the Soviet Union were always at odds. There seemed to be no limit to what each country would do to undermine the other. But by looking at the countries’ relationship quantitatively, Crescenzi found that they got along better than expected. In fact, his calculations show that despite shocks such as the Cuban Missile Crisis, the relationship steadily improved over the course of the Cold War.

Quantifying a relationship between two people is hard enough. So when studying the link between two countries for conflict research, political scientists use mathematical models to help characterize how countries interact. In a way, the models are like psychics, who use the answers to a few carefully selected questions to predict their clients’ futures. But a psychic who asks the wrong questions could misinterpret the past, leading to inaccurate predictions. So when researchers are developing a model, they want to ask the best questions they can.

Crescenzi says most of the models that have been used to define conflict potential between states were not accurate because they were static and only focused on a fixed period of history. Relationships between countries are never frozen in time, he says; they’re constantly evolving for better or for worse.

Crescenzi wanted to develop a dynamic conflict model that takes into account changes in countries’ affairs. “I was just a young grad student when I came up with the idea,” he says. Creating the model itself was relatively simple. Designing the research project was much harder. In order to get a complete picture of how countries connect, a good model would have to cover decades, if not centuries, of history.

The main challenge was what Crescenzi calls the “no lab” problem. Unlike scientists who generate their own data in laboratories, Crescenzi and his collaborator Andrew Enterline couldn’t directly observe the historical events they used in their research. They could only incorporate events found in public records. Despite having to use incomplete data, they published their first version of the Interstate Interaction Model in 2001.

Crescenzi and Enterline were never quite happy with that first model. One problem, they realized, was that the model dealt only with militarized conflict. To predict conflict more accurately, the model would also have to take into account cooperative efforts or even periods of inactivity.

Crescenzi and Enterline had gathered data for the conflict portion of the model from a publicly available dataset of statistics on all violent interstate events that occurred worldwide between 1816 and 2001. It was tough to find a comparable source of data about cooperation. So Crescenzi and Enterline, along with a third collaborator, Stephen B. Long, settled on an imperfect but valuable measure: they decided to use membership in intergovernmental organizations such as the United Nations or NAFTA. Joining an intergovernmental body, they reasoned, is one of the strongest signs of international cooperation, and shows that a nation shares interests with the other members of the group. The size of a group also indicates cooperation intensity. For example, a country would be demonstrating more intense cooperation by joining the African Union than by joining a larger organization such as the United Nations.

The researchers also added a decay component to the model to better characterize the ebb and flow of interstate relations. Countries typically go through periods where they interact little, making their relationship more neutral. This holds true even for the United States and Russia, who spent the latter half of the twentieth century aiming nuclear missiles at each other. Within this tension there were periods of calm, such as the détente period of the 1970s. By incorporating cooperation and decay, Crescenzi’s model reflects a relationship that was more amicable than anticipated.

In the case of North Korea, on the other hand, there were no surprises. U.S. cooperation with the communist state has been consistently low for decades. Crescenzi’s calculations show that the relationship has been volatile since the Korean War and will likely remain that way for years to come. A similar situation exists with Iran. While there was some initial cooperation between the United States and Iran, those events have been overshadowed by an increasing tension. Crescenzi’s model suggests that if a crisis were to occur with either North Korea or Iran, it would be nearly impossible to avoid militarized conflict. “As the environment becomes more hostile as a result of history, it becomes harder to negotiate your way out of a crisis,” he says.

To help predict future conflict, Crescenzi’s model gives interstate relationships a score between -1 and 1. A score of -1 indicates that two states are worst enemies; a 1 represents best friends. The lower the number, the more likely conflict will arise.

Crescenzi believes that his Interstate Interaction Model could be useful for improving the relations between countries. Conflict models may not be able to solve crises directly, but they give officials a better understanding of how two countries relate. Cues from the patterns generated by the model can help shape policy. Mathematical models similar to Crescenzi’s are often employed by the Pentagon and other government agencies to gauge how the United States should proceed in foreign policy matters. Will Crescenzi’s model make it to the Pentagon? It’s a long shot, he says. But he’s satisfied with the model’s accuracy. “Understanding the processes that generate events, both in the past and the present,” he says, “is what we do as social scientists.”