The Predictioneer’s Game

Wednesday, April 21, 2010

Michael Krasny, host of KQED Public Radio’s Forum, challenges Hoover senior fellow Bruce Bueno de Mesquita over his use of game theory in predicting world events. Highlights of that interview:

Michael Krasny: The whole idea of doing things for strategic reasons and for self-interest sort of bypasses the irrationality of people . . . and they do act irrationally and unpredictably.

Bruce Bueno de Mesquita: Well, I would like an example of irrational as opposed to unpredictable; those are different things. To me, rationality is doing what you believe is in your self-interest, and I can really think of only two groups of people who in general don’t behave that way. I don’t make predictions about them, and they don’t [behave rationally] because they’re not able to form stable preferences: two-year-olds—so you offer a two-year-old vanilla ice cream and they say they want chocolate. You put chocolate in front of them. No, no, no, they want vanilla. They don’t know what they want, so it’s very hard to predict their behavior, and it’s very hard for them to act in what we would think of as a rational way. And schizophrenics, who have difficulty forming stable preferences as well. Other than that, I really can’t think of anybody who is irrational.

Krasny: Is it self-interest for somebody to get on a plane and simply blow himself up?

Bueno de Mesquita: Sadly, it is. In fact, in The Predictioneer’s Game I have a lengthy discussion of why suicide bombers are rational and why, at the other extreme, Mother Teresa and people like that are rational. So if we take . . .

Krasny: In the tradition of Christopher Hitchens, taking on Mother Teresa here . . .

Bueno de Mesquita: [laughter] It was kind of fun in The Predictioneer’s Game, picking on her a little bit. If you think about suicide bombers, they come in a variety of categories. Most suicide bombers don’t fit the profile of this Nigerian [the man implicated in the Christmas Day terrorist attempt], and for most suicide bombers—I’m thinking particularly in the Palestine/Israel conflict—it is the case that their families are substantially compensated for killing themselves, for martyring themselves. When the price was low (for example, Saddam Hussein used to pay $10,000 to families), after a while there was a shortage of people willing to blow themselves up. He increased the price to $25,000, and then there were plenty of people available.

Nothing terribly unusual about that. You go back to the Crusades in the twelfth and thirteenth centuries. When the pope called for people to come and fight for the church, hardly anybody showed up. Then he said, “Well, you’re guaranteed to go to heaven,” and some people showed up . . . not a lot. Then he said, “By the way, your families’ debts will be forgiven if you die in the Crusade”—and he had droves of people show up. Suicide bombing follows much the same pattern. This Nigerian fellow I don’t know much about. He clearly came from a well-to-do family, so there was not a financial incentive. One has to believe that among his values are a set of attachments to martyrdom and his notion of an afterlife, but I don’t know that.

Krasny: Well, since you mentioned the Middle East, you talk about the potential for peace . . . a good example of self-interest in that if revenue were shared, let’s say, tourist revenue by the Palestinians and the Israelis, it would make a great deal of difference in terms of the possibilities for peace.

“In The Predictioneer’s Game I have a lengthy discussion of why suicide bombers are rational and why, at the other extreme, Mother Teresa and people like that are rational.”

Bueno de Mesquita: Let me make that a little bit more subtle. It is tourist tax revenue shared by the government, so the idea is that tourism, unlike most economic activities, is extremely responsive to violence. A single Palestinian or Israeli—and it’s pretty symmetrical statistically—having been killed by the other side, results on the Israeli side (we don’t have good data on the Palestinian side) in approximately 1,300 fewer tourists, 2,500 fewer tourist hotel nights. Well, the tax revenue that the governments collect is pretty easy to monitor. If divided by population, roughly speaking 60 percent to the Israelis, 40 percent to the Palestinians, if the tax revenue were just automatically divided that way from tourism by an international agency, then each side for its own reasons would have an interest in policing itself. There would be no requirement of trust, no requirement of cooperation or collaboration or anything like that. Under conservative assumptions, if this program were adopted and there were a substantial decline in violence resulting from it, if on the Palestinian side there were a huge incentive to police themselves . . . the Palestinian gross domestic product is approximately $5 billion, and under this program it would grow to $6 billion. That is a 20 percent growth rate. We are not talking chump change here; we’re talking about transformative growth in the Palestinian economy. And on the Israeli side, under conservative assumptions, it would be a wash between the average level of violence now and the revenue it gets, and giving away 40 percent of the revenue but having a much larger base of tourists. So it’s a self-enforcing strategic approach to moving things forward.

Krasny: You’re a partner in Mesquita and Roundell, a consulting firm that uses game-theory models developed to predict political and foreign policy events. And these are downloadable?

Bueno de Mesquita: There is an online version of the game that I call an apprentice version—good for prediction, not so good for engineering—on my book’s website at www.predictioneersgame.com. Anybody can access it. There’s a training manual so that you can learn how to put a data set together and play with it.

Krasny: So let’s talk about some of your predictions. You’ve said, for example, that by 2012—and this would be good news by many people’s lights—there will be a more secular regime in Iran; I assume that’s based on the idea that it has a strong, young population very much in opposition to the present regime. How’s this calculated?

Bueno de Mesquita: That’s part of it. That analysis is based on a study that I did with my undergraduate students at NYU.

Krasny: Were these the same undergraduates you asked, “If you could give up your cell phone to ameliorate poverty, would you do that?” and most of them, or almost all of them, said no?

“For most suicide bombers—I’m thinking particularly in the Palestine/ Israel conflict—it is the case that their families are substantially compensated for killing themselves, for martyring themselves.”

Bueno de Mesquita: The very ones, yeah. They are self-interested. But if I may just follow up on that for a moment, I recently was in Abu Dhabi giving a talk on this work. Abu Dhabi—most people probably don’t know this—is about the greenest place on earth. They’re building an entirely carbon-free city. I had an audience of a few hundred. I had predicted in The Predictioneer’s Game that [the Copenhagen Climate Change Conference] would fail, and one of the questions I was asked was, what could they do to make things better? I asked for a show of hands: how many people had a cell phone? They all did. And how many would be willing to give up their cell phone if that money went to support green energy sources in the Third World, where people couldn’t afford them? One hand stayed up.

On Iran, my analysis looked at almost ninety players—eighty-seven I think was the exact number—a mix of individual Iranian elites, political elites, various Iranian citizen groups, business groups, student groups, dissident groups, the full panoply of interests as well as outside interests. And the analysis found that . . . [Mohammad Ali] Jafari, the head of the Revolutionary Guard, was rising in power, as were the bunyads, an economic interest group; that they both were going to become much more pragmatic to hold power; and that [Ayatollah Ali] Khamenei’s declining influence and [President Mahmoud] Ahmadinejad’s declining influence would result in a regime shift to a more secular—not a nice regime, not a democratic regime, but a more secular regime with more concern for economic performance than for stimulating problems outside of Iran.

Krasny: And then we’ve got you saying that if you double aid to Pakistan—to what, $1.5 billion this year?—it would mean a more civil and more secular government, but also a government that would be politically, shall we say, not so accommodating to terrorists.

“What the [Iran] analysis found was . . . a regime shift to a more secular— not a nice regime, not a democratic regime, but a more secular regime with more concern for economic performance than stimulating problems outside of Iran.”

Bueno de Mesquita: Yes, in fact, you phrased that very well, because the prediction was that they would make more of an effort to crack down on the terrorists but would not go after the top level of terrorists. They would not try to end terrorism in Pakistan. That’s exactly what the government in Pakistan has been doing. They have been making more efforts to go after low-level terrorist groups, extremist groups, operating in Pakistan. But they explicitly declined the request of the United States to go after the leader of the Taliban in Pakistan. And, you might ask, well, why would they do that? Think about that $1.5 billion, which is basically buying, from the U.S. point of view, policy compliance from the Pakistanis. If they actually got rid of all of the terrorists, would we continue to give them the money? The answer is obvious: we wouldn’t! So they have an incentive to do enough to keep the money flowing but not enough to get rid of the incentive to give them money.

If you think about the Camp David peace agreements between Israel and Egypt back in 1979, as part of that deal, the United States gives a tremendous amount of economic aid to Egypt, making it the second-largest aid recipient (at least it was for many years, Israel being first). Many people wonder why it is that the Egyptian government continues in its schools to teach a dislike of Israel. Why doesn’t it build a more accommodating, friendlier view? It’s because the money that they get from the United States is a substantial part of their economy, and we wouldn’t give them money if they didn’t have an incentive to be nasty toward the Israelis. It is politically good for American leaders in terms of their constituents to be friendly to Israel, and it is good for the Egyptians to teach hostility.

“Think about that $1.5 billion, which is basically buying from the U.S. point of view policy compliance from the Pakistanis. If they actually got rid of all of the terrorists, would we continue to give them the money?”

Krasny: You mentioned Copenhagen earlier, and I wanted to just touch on that because you are and have been fairly pessimistic about anything coming out of Copenhagen that would do much to regulate carbon emissions on a vast scale. You’ve also said some things about how, if this global warming crisis that we’re facing is going to be solved, it’s going to be solved with wind- and rain- and solar-type energies. But for the most part, you’re on record as saying that these kinds of agreements are meaningless.

Bueno de Mesquita: That’s right. Universal treaties are generally feel-good agreements. They ask people to do one of two things: either they ask them not to change their behavior at all, in which case people are happy to comply, or they ask them to undertake big changes in behavior, in which case they don’t then have monitoring or punishment strategies in place to make sure that people who violate the agreement pay for it. So if you think of Kyoto, 175 countries signed Kyoto, the United States not one of them. Out of that 175, 137 are in complete compliance just by reporting what they’ve done in the previous year—which may be nothing—and so they report “we’ve done nothing” and they’re in compliance. The remainder were expected to cut emissions substantially. Instead of cutting emissions, they pretty much, in lockstep, got up and announced, “Oh, we can’t meet the standards that we promised to meet.” No consequences for them in terms of punishments.

That’s the nature of universal agreements, unless they’re about something that everybody can see in the immediate future they will benefit from—for example, agreeing on what time it is in different parts of the world. So my view on Copenhagen from the onset was it would fail for this reason and that if we really were committed to doing something about greenhouse gas emissions, we would start more vigorously at home, unilaterally, instead of saying, “Well, we can’t do anything and let’s everybody agree.” It’s just an excuse.