Game theory does not concern itself with good and evil. It seeks to predict not which strategies are just, but which are most effective. John von Neumann, a Hungarian-born polymath with a sideline in predicting the blast radius of an atomic bomb, co-authored the discipline’s seminal work, Theory of Games and Economic Behavior, in 1944. He then began thinking about war-game scenarios that weighed the likelihood of an exchange of nuclear weapons. “Early in nuclear negotiations,” says Steven J. Brams, a New York University professor of politics who worked under former Defense Secretary Robert McNamara during the 1960s, “we didn’t know that they were not meant to be used. It took a few years before strategists digested this.” At the height of the Cold War, the application of game theory convinced leaders of the two nuclear-armed superpowers “that if we’re thinking of using them, we’re in deep s–t.”
This summer’s negotiations over raising the debt ceiling seemed to present Democrats and Republicans with a similar dilemma. Failure to reach a deal threatened to bring on the economic equivalent of a nuclear winter. The leaders of the two parties, Barack Obama and Speaker of the House John Boehner (R-Ohio), appeared to grasp this, but a vocal band of “Tea Party hobbits,” as their fellow Republican, John McCain of Arizona, dubbed them, refused to go along. They made it clear they were not only willing to bear the catastrophic consequences of a U.S. default, but that they might actually welcome it. Trapped in a classic game of “chicken”—a term game theorists use, too—in which both players entertain the option of killing everyone, the President did what game theory suggests a rational actor would do. He recognized his potential maximum losses were greater than his opponent’s. He caved.
Obama’s decision to agree to a deal that calls for $2.1 trillion in spending cuts over 10 years, with no increases in revenue, was greeted with scorn among his own supporters and disdain from global markets, which plunged at news of the accord. Meanwhile, Tea Party Republicans complained that the agreement didn’t go far enough. The outcome of the debt ceiling talks left everyone in a foul mood, not least the President himself, who signed the final legislation on Aug. 2 in the Oval Office, alone and grim-faced.
And yet for all the collective self-loathing that attended the debt ceiling talks, it’s important to remember that, like just about everything in human behavior, it was still reducible to a game. Looked at through the prism of game theory, it’s hard to see how the outcome could have turned out any other way.
Biologists have adopted game theory to describe successful adaptations. Labor arbitrators have used it to ease negotiations. And Brams, who has many books on game theory to his credit, has drawn a decision tree for God’s last discussion with Cain in which Cain can choose to admit, deny, or defend his crime and God, in turn, can choose to kill or punish him. Looking back over the Summer of Debt, Brams can’t find a single move by any party that’s inconsistent with predictions from his discipline. Crucially, game theory assumes that no one is crazy, and it’s true in life that almost no one ever is. There’s also a pragmatic reason to treat your opponents as sane: You can’t make predictions about their behavior unless you do.
Almost everyone—even members of the Tea Party—knows a hawk from a handsaw, and the best strategy is to assume the other player has a rational goal and try to figure out what it is. People act crazy, but they’re at their craziest when they want something. All you can do in response is to make your most honest estimate of what the crazies actually want, and respond as if they are methodically pursuing it. There is no advantage to be gained, for example, in pointing out that Kim Jong Il is a potbellied nut job in a bad suit. Everything he’s done during his reign as North Korea’s leader suggests he’s an amoral, but sophisticated, negotiator. Unpredictability, says Brams, can be a smart strategy.
Game theorists distinguish between “cooperative” and “noncooperative” games. A cooperative game looks to divide a pie in a way that leaves both sides with trust in the process. Binding arbitration, where two sides are obliged by law to submit to a judge’s decision, is a cooperative game.
The two parties in Washington pretended to be playing a cooperative game this summer. At one point, the President and Boehner simultaneously urged the other to get “serious” and behave like an “adult.” The object of the game, as each leader described it, was about how best to divide the pain of closing the deficit, in the same way a family sits down to a pile of bills on the kitchen table.
The President’s bipartisan commission on deficit reduction, set up late last year and chaired by Democrat Erskine Bowles and Republican Alan Simpson, also played a cooperative game. The rules were clear and the game was closed, designed only for one iteration: a single budget, to be sent to Congress if 14 of its 18 commission members approved it. The commission produced the kind of document you’d expect from a well-designed game, a series of compromises and innovations designed to distribute pain and create trust in the result. But it failed, ultimately, because it couldn’t draw enough votes to be turned into legislation. Six of the seven who voted against it were sitting legislators. This is not a coincidence. Washington, as a game theorist would describe it, is noncooperative.
A noncooperative game lacks a higher authority to impose agreements on both sides. In Washington, no politician is bound to reach a compromise to solve any long-term problem. Everyone, however, is playing a game called “election,” and the only possible goal in that game is to win the next one. If you hear someone in Congress say, “Senator X is just playing politics,” a perfectly legitimate response is, “She has to. Those are the rules of the Constitution.” If we grumble, as voters, that we need to throw the bums out, all we’re doing is subjecting a new set of bums to the same game. Anyone who promises to fix or change Washington is merely attempting to impose a cooperative game on a town that, by design, can’t play one.
Obama and the House Republicans, says Steven Brams, were playing chicken this summer, a noncooperative, non-zero-sum game in which both players can lose. A compromise outcome is difficult to achieve in chicken, because it’s not stable. Brams says that each player has an incentive to dissemble, because he will achieve a better outcome for himself if he does.
A game theorist would say that the President is trying to play a cooperative game in a town that can’t play along with him. The trouble for the White House is that the Republicans aren’t playing a game called “fix the budget deficit.” They’re necessarily playing one called “defeat Barack Obama.” A reasonable offer seldom works in a divorce; there’s no reason to expect it would in Congress.
Obama and the Democrats should recognize that anger, too, is a tactic. Brams argues that there’s no value in trying to determine whether anger is real or feigned; it has the same effect either way. Anger comes of frustration, he says, which in turn comes of a sense of powerlessness. But frustration can actually turn a noncooperative game cooperative. In the Aristophanes play Lysistrata, the women of Athens and Sparta withhold sex until their men sue for peace. Intermittent hostility among Greek city-states, a noncooperative game, was bound into the cooperative game of a negotiation, brokered by thwarted lust.
The Tea Party, in this sense, has succeeded by adopting a rational frustration strategy. Like the women in Lysistrata, the Tea Partiers, in effect, have withheld their affections from both the Democrats and the Republican leadership. That has forced Congress into what may prove to be a cooperative game: the next round of deficit negotiations, with a super committee composed of members from both parties charged with finding consensus—enforced by triggers that would cause pain to both sides if they don’t. You can find fault with the Tea Party’s prescription for balancing the budget—most economists do—but if they hadn’t come to Washington last year, Congress would have waited for a real bond crisis, five or 10 years from now, to create its super committee.
In a new book, Game Theory and the Humanities, Brams offers another case study in frustration strategy: Shakespeare’s Lady Macbeth, who is desperate to see her husband crowned, but fears he is too kind. And so she pours her spirits into her husband’s ear, demanding that he murder King Duncan. We will know, at the close of the next round of negotiations, which game the Tea Party has been playing: Balance the Budget or Kill the King.