The Predictioneer’s Game: Using the Logic of Brazen Self-Interest to See and Shape the Future

by Bruce Bueno De Mesquita

  The Germans and Austrians probably would have taken more seriously the prospect that Britain meant business. As we will see, the model indicates just that. Additionally, a naval mobilization of this sort has none of the grave risks associated with the Russian mobilization of ground forces. Russia could move troops quickly to and across the German frontier. Understandably, that made the Germans more than a little jittery. British ships filled with soldiers would have taken a long time to get into position. Finally, the ships would not have been directly in the path on which initial fighting was expected. Thus the shiploads of troops would very much have been a signal of what was to come without precipitating immediate military action. In reality, British ships under French command headed for the Adriatic a few days after war had been declared: too little too late.

  When I simulated the prewar 1914 crisis with the British at 90 as their “flexibility” variable, the model indicated that the Austrians and Germans were as uncertain about Britain’s true intentions as they could be. But when I place Britain’s score on this one factor at 50, the model shows that the Germans and the Austrians are convinced that the British will fight. More tellingly, Austria-Hungary’s and Britain’s pattern of interactions change. When Britain shows little resolve, Austria-Hungary anticipates coercing the British into accepting their position. When Britain shows stronger, but not extreme, resolve (50), the Austrians seek a negotiated compromise with the United Kingdom even as they perceive that if Britain is allowed to move first it will mean war. Not, mind you, the little war that Austria sought with Serbia, but the big war that nobody wanted.

  Have a look at figure 9.2. Here we see the scenario in which Britain shows more resolve (50 instead of 90). By simulating a tougher British signal, we uncover the sort of intelligence about an adversary’s thinking that would be a lifetime coup for a real-world spy. We discover that with British ships heading for the Adriatic shortly after the start of the crisis, well before war is declared, Austria-Hungary, Germany, and the U.K. perceive the possibility of settling their differences quickly. The Austrians and Germans saw no such prospect (or reason for it) in the simulation of the actual situation. But with the U.K. showing more resolve, the strategic environment is greatly altered. The Austrians and Germans believe that they can and should make a deal with the U.K. in a matter of days after the start of the crisis.

  The model says that the Austrians and Germans recognize that they should give up their demand for utter Serbian capitulation. They see an opportunity to persuade the Triple Entente to agree that Austria-Hungary should have real, but not controlling, influence over Serbia’s foreign policy.

  Of course, change begets change. The members of the Triple Entente don’t immediately embrace the new offer that (according to the model’s logic) would have been put on the table if the British had sent navy ships to the Adriatic right away. While not caving in to this simulated proposal, the Triple Entente’s diplomats certainly sit up and take notice. The diplomats remain in charge, keeping the generals on the sidelines. While the negotiators hold out for a few weeks, thinking they can extract more concessions from Austria-Hungary, by early August they realize (in the model’s logic—remember, none of this was actually done in 1914) that the Austrians and Germans are highly reluctant to give more. So they take a deal located at about 75 on the 100-point issue scale. That would have given the Austrians more than was put on the table in the real 1914 crisis, but a lot less than the surrender of Serbian sovereignty. It would have more or less split the difference between the British position and the Aus-tro-Hungarian demand. The simulation under these conditions shows that the French and the Russians would have quickly acceded to this compromise. The war to end all wars would have been avoided.

  FIG. 9.2. Engineering Successful Negotiations During the 1914 Crisis

  Had there but been a thousand mathematicians crunching numbers in London in June 1914, we might not need to ask the next and final question of this chapter: Could World War II have been prevented by the judicious use of a predictioneer’s skills?

  NO MORE MISTER HITLER?

  The rise to power of Adolf Hitler is a strange and horrendous tale that could have been nipped in the bud sometime between November 1932 and March 1933, if not sooner. It is a tale worth heeding. Whatever else can be said of Hitler, it must be admitted that he was honest and open about his intentions. Not only did his biography, Mein Kampf, published in 1925, announce his ravings to the world; so did he, in campaign speech after campaign speech in 1932. Having failed to come to power in 1923 through an attempted coup d’état (the Beer Hall Putsch, as it is known) in which both German police and Nazi insurgents were killed (and Goering seriously wounded), Hitler was now determined to rise to dictatorial control through legal means. The ballot box would replace the bullet for the moment.

  In one campaign stop after another—several per day, as Hitler was the first German politician to take advantage of air travel to cover vast amounts of ground—Hitler declared his intention to ban political parties and suspend the Reichstag, Germany’s parliament, if he came to power. Now, when a campaigner promises peace and prosperity, motherhood and apple pie, we don’t really learn anything about what they plan to do. Being for hope or change or a thousand points of light says nothing. When a politician promises to overturn democracy, that’s a different matter. You don’t lose votes by promising peace and prosperity. I suppose being denied freedom of choice appeals to some people, but certainly not all and probably not many. So when a politician makes such outrageous declarations, we must ask: What does he have in mind? The answer must be that he means what he says. Hitler certainly did.

  Of course, we have little need to pay attention to every fringe movement that makes outrageous declarations. But by 1932 the Nazi Party was no fringe movement. The 1930 German election—they ran elections with incredible frequency—gave it 107 out of 577 seats in the Reichstag. In the July 1932 election, the Nazis became the single largest party in the Reichstag, with 230 seats. By then no prudent person could treat Hitler’s campaign promises lightly. Anyone listening should have understood, given the nature of what he was saying, that he meant what he said and he said what he meant. Hitler was a dictator at heart, one hundred percent.

  Hitler’s party lost some seats in the November 1932 election but still remained the single largest contingent in the German parliament. The election gave the Nazis 196 seats. By January 1933, Paul von Hindenburg, the beloved and elderly German war hero and president of the Weimar Republic, acquiesced under pressure to make Hitler chancellor. Now the door was wide open to his dictatorial ambitions. In early March 1933 another election was held, almost immediately after the Reichstag fire on March 3. Hitler was quick to blame the Communists for the fire, using that as a pretext to ban them from the Reichstag. He wanted all Communist leaders executed that night—no more Mister Nice Guy—but Hindenburg refused to go along.

  The March election was a mixed success for the Nazi Party. On the plus side (from its perspective), the party increased its seat share in the Reichstag from 196 to 288. On the minus side, the Nazis failed to gain a majority. That meant Hitler still had to make deals with other parties; he was not yet completely in control. He was still vulnerable to defeat if a strong enough coalition of parties in the Reichstag joined together to oppose him. The tragedy is that they didn’t.

  Shortly after the March 5 election, on March 23, 1933, he negotiated his way to a two-thirds vote in the Reichstag to change the constitution to comply with the terms of the Enabling Act, a piece of legislation to make him dictator. The Enabling Act gave Hitler as chancellor essentially all of the constitutional authority also granted to the Reichstag so that he no longer needed legislative approval for policy changes he wished to put in place. The Enabling Act made Hitler Germany’s dictator and ended the need for future elections. He was well on the way to doing everything he had promised to do during the previous election campaigns.

  Once the Enabling Act was approved, probably nothing sh
ort of a military uprising or foreign military intervention could have stopped Hitler on his destructive course. What about between November and March? As I said, Hitler’s intentions were no secret. Could he have been thwarted before the Reichstag fire resulted in the one hundred Communist Party members in parliament being banned, making it much easier for Hitler to put together a two-thirds vote? Hitler was stuck operating within the legal system, at least more or less, before the Enabling Act’s passage. It was far from a sure thing that he could muster a two-thirds vote. The key to his success—or failure—was the Catholic Center Party (BVP).

  Allow me for a moment to frame the game as it was set up. There were four principal parties in the Reichstag at this time: the Nazis, the Catholic Center Party, the Social Democrats, and the Communists. The Social Democrats and the Communists generally opposed Hitler, and would resist his Enabling Act (the Communists, of course, irrelevant in the actual voting on account of being barred because of their supposed role in the fire). The Catholic party, however, was divided over whether to support the Enabling Act. Their leader, Ludwig Kaas, was no fan of Hitler’s. Nevertheless, Kaas, a priest, negotiated with Hitler for assurances that Catholic interests—inside and outside government—would be protected if Hitler were given their support. Kaas may also have sought assurances of a concordat with the Vatican. Hitler agreed to his terms, understanding the temporary need for compromise.

  Hitler’s deal with the Catholic party was the crucial decision that gave the Nazis the two-thirds majority they needed for the Enabling Act. With the Communists barred, the Social Democrats were the only party to vote against the act, and of course Hitler won the game as a result. But it remains the case that without the Catholic Center Party, the Enabling Act would have been defeated, Hitler would not have been made dictator, and, who knows, perhaps the course of world history would have been completely altered.

  Beating Hitler was no easy task. We must admit that he played even the parliamentary game the best out of all the parties. His opponents either underestimated him (with fatal consequences) or were simply incapable of outplaying him. That, however, is not to say that there weren’t winning moves open to them.

  What or who could have stopped the Catholic Center Party from going along with Hitler? How could they have made someone—anyone outside the Nazi Party—Germany’s new, democratic boss? Can you see a possible strategy?

  I applied my forecasting model to this question, constructing a data set in which the stakeholders are the political parties represented in the Reichstag plus Hindenburg. Their power is proportional to the number of seats they controlled, except, of course, Hindenburg, who had no seats in the Reichstag. His personal prestige and popularity, however, gave him great weight, even more than Hitler and his Nazis. Remember, Hitler could not execute the Communist leadership without Hindenburg’s (withheld) approval. I accordingly assigned Hindenburg 67 percent more clout than the Nazis. Positions on the Enabling Act are and were well known at the time. The Communists and the Social Democrats were utterly opposed, the Nazis utterly in favor; Hindenburg leaned favorably toward the Act and the Catholic Center Party did too, but ever so slightly. The rest were committed to the Act.

  Right after the November 1932 election, while Hitler was angling to become chancellor, the Social Democrats and the Communists could (according to my model) have struck a deal with the Catholic Center Party, depriving Hitler of the two-thirds majority he needed. To do so, however, they would have had to move meaningfully in the direction of the Catholic Center Party’s policy desires, perhaps so much so that a member of the Catholic Center Party would have become the chancellor instead of Hitler. That is, the Social Democrats and the Communists needed to provide at least as good an assurance of Catholic interests as Hitler astutely provided months later. The model indicates that they did not believe they had this opportunity. They did not think that the Catholic Center leadership would listen to them or make a deal with them, and so, fearing rejection, they didn’t really try (or at least they didn’t try hard enough). The model says they were wrong. Too bad we can’t go back in time to test the waters and see if the deal could have been made.

  We do know that Ludwig Kaas, the Center Party’s leader, had in the 1920s developed good relations with the Social Democrats’ then leader, Friedrich Ebert. It is difficult to imagine that under the circumstances Kaas would not have responded to the Social Democrats, especially if they were prepared to support him as the next chancellor (let’s remember game theory’s view of human nature and what it means for individuals seeking to obtain or maintain power and influence). The model says Kaas would have reached a deal with the Social Democrats and the Communists. Of course, for the Communist Party, atheists that they were, bitter adversaries of both the Social Democrats and the Catholic Center Party, this would have been a bitter pill to swallow. But surely it was better than the easily foreseen fate they met after the Enabling Act was passed. Many Communists were murdered, others sent to concentration camps.

  Even after the Reichstag fire, the Social Democrats still had the chance to cut a deal with the Catholics, but didn’t (although whether Hitler could still have been stopped by that stage is questionable). Yes, he could have been deprived of a two-thirds majority, but just barely, once the Communists were out of the picture, and he was, as I noted earlier, playing his strategic cards effectively. In November, December, and maybe through January, however, there is a good chance that a defeated Nazi Party would have been relegated to the dustbin of history. Had Hitler attempted a coup at that stage, it probably would have failed. The German security apparatus would have rallied behind a Catholic Center—led government. The German general staff had no love for Adolf Hitler and his brownshirts, and he was a long way from being the undisputed representative of the people. Remember, the Nazis lost seats in 1932. German popular will was not yet decisively behind Hitler or his party.

  Without Hitler, the capitalist, democratic world and the communist, totalitarian world led by Joseph Stalin might have butted heads in the 1930s instead of waiting for the cold war. Perhaps even as bloody a war as World War II would have been fought, and perhaps not. Stalin’s Soviet Union might have successfully defended its frontiers, but it probably had nothing like the capabilities of Hitler’s rearmed Germany to wage an aggressive war beyond its borders.

  All this is speculation, of course, but there seems little reason to doubt the model’s accuracy in this or the other cases looked at in this chapter, given its track record over thousands of applications. And if we can replay the past accurately and find ways to improve it, as we just did with World Wars I and II, there is no reason to doubt that we can fast-forward the present and work out ways to make it turn out better. That is the whole purpose of this forecasting and engineering enterprise.

  In the next chapter, we will play with some of the big issues of our time. I will use my newest model to make live predictions whose accuracy you will be able to check for yourself.

  10

  DARE TO BE EMBARRASSED!

  IT’S ALWAYS NEAT to think about what-ifs, to rewrite the past with the idea that we could have worked things out for the better. But thinking about how to rewrite history is one thing; thinking about how to write a script for the future is quite another. It’s so easy to get the past right when you already know what happened. And while examining alternative pasts is fun and informative, we still never get to find out whether we really could have derailed Sparta’s defeat or stopped Hitler in his tracks. Ultimately, solving a seventy- or eighty-year-old problem is fascinating, but not terribly useful outside of what it teaches us about the gaming process. Working out how to solve today’s problems, like stopping al-Qaeda dead in its tracks—now that would be useful. That’s why any predictioneer worth his or her salt must be willing to risk the embarrassment that comes from being wrong.

  In this chapter we will look ahead a year or two from when I am writing this (in April 2009 for the case of Iran-Iraq relations and June 2008 for the P
akistan case). Here is where the rubber really meets the road. We will look at what the United States government could do to diminish the threat of terrorism or insurgency in Pakistan, and the likely relations to develop between Iran and Iraq if President Obama fully withdraws U.S. forces from Iraq or leaves fifty thousand in that country well beyond August 2010.

  Back in the spring of 2008 and again in 2009 I taught an undergraduate seminar at NYU in which twenty terrific students in each class used my new forecasting model. This was a great opportunity for me (as well as, I hope, for my students) to find out how hard or easy it is to teach people with no prior experience how to become effective political engineers. Fortunately, my students were willing guinea pigs, and they did a great job.

  The main idea behind this course, sponsored by NYU’s Alexander Hamilton Center for Political Economy, was to search for solutions to pressing policy problems based only on logic and evidence. That is the Center’s mission. It leaves no room for partisanship, ideology, opinion, anecdotes, or personal wishes when it comes to crafting solutions. Game-theory models, however, are a way to fulfill the mission. With that in mind, I asked my students to pick any foreign policy problem that intrigued them. They clustered themselves into groups and set to work on Pakistan, the Israeli-Palestinian dispute, global warming, nuclear proliferation, relations between Cuba and the United States, relations between Russia and the Ukraine, and many other critical policy concerns.

  Each student studied a problem that he or she really cared about. They took the class knowing that they would use game theory to work out likely future developments and to write a script about how to improve the future from the perspective of any one of the players in the game. They had almost no prior experience with any of the material or models. They had limited access to experts, so they relied on the Internet and major news outlets to put their data together. I mention this to be clear that any hardworking, motivated person can replicate what they did. All this being said, my students used my new model, and I certainly reviewed their work—so any misses are the model’s and mine.