Let the Game Theory Begin

Let the Game Theory Begin

Let the Game Theory Begin

Science, evolution, and politics explained.
Sept. 21 2001 8:30 PM

Let the Game Theory Begin

Click here to read Will Saletan's response to this column.

Advertisement

My previous column, which warned against uncritically obeying the retributive impulse at this moment in history, drew the following e-mail from one Slate reader: "Dear Mr. Big Zero: Please read another Slate article, TRUTH OR CONSEQUENCES by William Saletan. While just today, I made a vow to boycott your drivel, I do have a morbid curiousity about how you would respond to this article …"

Upon following this guidance, I discovered that my friend and colleague Will Saletan does indeed discuss America's response to last week's terrorist attack in a more militant tone than mine. On the positive side, he takes a clinical, quasi-game-theoretic view, which means we have a lingua franca. So let me critique his model. Maybe he'll reply and we'll together make progress in figuring out what the United States should do.

Will criticizes people who stress how America's past actions fomented the anti-American sentiment underlying the attack. (He mentions David Corn of The Nation, Susan Sontag of The New Yorker, and Gary Kamiya of Salon.) He argues that heeding their counsel—examining our past behavior with an eye toward reform—amounts to letting the terrorists control our conduct. He insists that we focus more on modifying their behavior by dishing out negative reinforcement (something I certainly don't oppose in principle).

This is the standard dilemma in dealing with terrorists. Yes, their terror is motivated by grievances (even if Osama Bin Laden's grievances are more amorphous than average). And in theory you might try to end the terrorism by addressing the grievances. But doing so rewards their behavior, encouraging them to repeat it.

Advertisement

I think this dilemma seems slightly starker than it really is because Will casts the situation as a simple two-player game involving America on one side and "the terrorists" on the other. Let's drop that phrase for now and instead ask: Whose future behavior do we want to shape? If we leave aside the terrorist foot soldiers—who after all tend to welcome death and thus aren't amenable to conventional incentives—we're left with two groups.

First, there are terrorism's elites: Osama Bin Laden and other financiers of terrorism and heads of states that support terrorism.

Second, there are the millions of Muslims who are discontented and blame the United States at least partly for their plight, whether that plight is poverty, or cultural dislocation caused by globalization, or corrupt and repressive political leadership, or whatever. These are the people who, depending on how their futures unfold, could become tomorrow's terrorist foot soldiers. They could also destabilize and even overthrow governments that try to fight the good fight against terrorism.

My key contention: America's relationship to the first group—Osama Bin Laden, et al.—is by and large zero-sum. That is, what's good for them is bad for us and vice versa. When someone is implacably devoted to ending global capitalism and American civilization, trying to find common ground with him is not time well spent. But America's long-run relationship to the second group—the discontented Muslim masses—is non-zero-sum. That is, our fate is positively correlated with their fate. If they get poorer and angrier and more consumed by hatred of America that's bad for us, because it will mean more terrorist foot soldiers—and the possible creation, through revolt, of whole new state regimes sympathetic to terrorists. If these people get more contented—less poor, less repressed, less psychologically threatened by globalization, less consumed by rage—that's good for us, as it will deprive terrorist elites of a political base and future troops.

Advertisement

In sum: We wish Osama Bin Laden et al., ill, and we wish the great bulk of the world's currently discontented Muslims well. The big and difficult question: Are there policies that would let us have it both ways?

Let's start with an example of something that, I contend, wouldn't: massive American air strikes against Bin Laden's suspected hideouts in Afghanistan and government buildings there. Even if this had the desired effect of delivering the ultimate negative reinforcement to Bin Laden and the Taliban leadership—which it probably wouldn't, given the abundance of mountains and caves in Afghanistan—it would deeply antagonize millions of Muslims. It would reinforce the already too-common image of America: a) brutally indifferent to the suffering of the world's poor (a single instance of "collateral damage" captured on videotape would create God-knows-how-many eager terrorist troops); b) particularly hostile toward Muslims; c) cowardly (i.e., favoring smart weapons over brave soldiers).

What's more, when an American military assault involves assistance from an Islamic nation, we risk inciting fundamentalists within its borders, as is already happening in Pakistan. An unnamed administration official was recently quoted worrying that Pakistan's government might be overthrown. This could sure be spooky—nuclear weapons added to the jihad's arsenal. But even if the regime survives, the fundamentalist backlash is bad long-term news for the United States, as it is already energizing tomorrow's terrorist pawns.

Just as it's hard to give Bin Laden negative reinforcement without antagonizing some of the world's Muslims, it's hard to address the discontent of those Muslims without giving bin Laden et al., positive reinforcement. Thus, pulling American troops out of Saudi Arabia might over time chill anti-Americanism among fundamentalists in that country. But it would also send a strong message to terrorism's elites that terrorism pays, since the U.S. presence in Saudi Arabia is probably the most distinct addressable grievance that Bin Laden has.

Advertisement

In sum, the challenge is to a) deliver negative reinforcement to terrorism's elites in ways that minimize the resulting growth of anti-Americanism among the world's Muslim masses and; b) address the causes of these Muslims' anti-Americanism in ways that minimize the positive reinforcement given to terrorism's elites.

OK, if I'm so smart, what's an example of how you do either of these things? Actually, I find examples hard to come by—which is why I suggest we think about this whole mess long and hard before lurching into a war in Afghanistan. But the point is that the challenge we face, however dicey, isn't logically impossible, as is the challenge implied by Will's model: address a given party's grievance without giving that party positive reinforcement.

Before trying to suggest some policies that follow from this new, three-player model, I'll pause and give Will a chance to chime in if he chooses. For starters, does he agree that:

  1. It makes sense to distinguish between these two different groups—terrorism's elites on the one hand and the world's millions of discontented Muslims on the other?
  2. Our relationship with the former is essentially zero-sum and with the latter is essentially non-zero-sum?
  3. The object of the game is to hurt the former and help the latter while minimizing the extent to which pursuing each of these goals compromises the other goal?

P.S.: Footnote on methodological oversimplification—I've included leaders of terrorism-sponsoring states among "terrorism's elites" because their relationship to the United States is, like Bin Laden's, currently zero-sum. However, unlike Bin Laden, some of them have the realistic option of changing policies in a way that would pivot that relationship toward the non-zero-sum.