'Politics Is Predictable': PW Talks with Bruce Bueno de Mesquita

In The Predictioneer's Game (Reviews, May 25), Bueno de Mesquita illustrates, with a mathematical model that quantifies self-interest, how we can use game theory to predict—and influence—future events.

When did you become aware of game theory's potential?

In my first semester of graduate school, I read one of the first important mathematical treatments of political science by William Reichert. Reichert claimed to prove results, and the proof of one of the results was wrong. I discovered the error and became enormously excited that for the first time in my experience in political science, I could make the statement “This is wrong”: not “I think this is wrong” or “It's my opinion that it's wrong” or “I don't agree.” It was just wrong.

Tell me about the development of your prediction model.

It's been in the making my entire career. In my dissertation, I studied strategies of opposition parties in coalition governments, and I started to think about how people form coalitions and what makes some cohere and others fall apart. Alliances are just another form of coalitions. I began to formulate the first formal, mathematical model of alliance formation that could be tested empirically and could predict a very basic thing: would a country choose to stay out of a war between two other parties or would it join a side, and if so, which side would it join? Forecasting elements came in punctuated parts; I saw how to translate the basic logic of the war model into something that could be applied to policy forecasting. I realized that I could estimate the probability of one side or another winning in any head-to-head contest.

Are you surprised at how adaptable your model is proving?

I was surprised. It was a revelation to me that wars, corporate mergers, political party decisions are all the same problem. And I'm surprised by how well the model works because I know all of its warts! Of course, the model doesn't apply to arenas where individuals are making choices without any need for negotiating, such as the market. It is restricted to arenas where people have to either coordinate with each other or coerce each other.

You're very good at obviating criticism. Has your approach been a hard sell?

It's polarizing. There are people who hold me in high regard, and there are people who wish I would go away. They think this is the wrong way to study politics. My view is: dare to be embarrassed. If you have a better way—great; I'd be happy to throw this stuff in the garbage. I gave a talk in February where I predicted that Ayatollah Khamenei and Ahmadinejad's power was going to decline around now. Given recent events, my predictions haven't embarrassed me, but they could have. And I urge others to take the same risk.