Monday, July 27, 2009

Obama, Game Theory, Right, Wrong and Justice

It seems the method to Obama's conciliatory style is basic Game Theory. According to Game Theory, compromises often provide the best net outcome for both parties even though the compromised outcome is never the optimal one for either. One example is the game of chicken, i.e., where you drive your car toward your opponent's car, and whoever swerves loses. Best case scenario, you go straight, your opponent swerves. Second best, you both swerve. Third best, you swerve first. Worst, neither of you swerve, and you both die in the crash.

To achieve the optimal outcome, you cannot swerve. Of course, by doing so, you risk the worst one. Obama's tack is to persuade people to settle for the second-best case scenario. By contrast the Bush administration seemed to want to step on the gas and convince its opponents that the steering wheel was broken, and swerving was not an option. Therefore its opponents would be left only with option three or option four. Of course, that only works when the opponent can be easily cowed like the congressional democrats. When dealing with Islamic militants, many of whom would prefer a cathartic crash, we were assured the worst - or second worst if we realized they were more committed than us and backed down in time.

But Obama will necessarily disappoint people who hoped for their optimal outcome. Overall, we can see why this is a good strategy because the outcome on average will likely be better than ones where everyone is dead set on winning. But what makes Obama's tack hard to take is not its outcomes but that compromise seems to proceed irrespective of right, wrong or justice. In other words, Obama will choose option 2, even if his side is absolutely right, and the other side is absolutely wrong. A good (if trivial) example is the Henry Gates incident with the Cambridge police officer.

When Gates, a black Harvard professor, was having trouble getting into his upscale home, a neighbor called the police to report a possible break-in. When the white police officer, Sgt. Crowley, arrived and questioned him, Gates became angry and insulted the officer. The officer ascertained Gates was indeed the owner of the house, but after Gates verbally got after him some more, Crowley cuffed Gates and brought him to the police station.

When asked about the incident, Obama said he didn't know all the details but that the Cambridge police acted stupidly in arresting a man in his own house for yelling at the cop who questioned him. This caused a major uproar, but what Obama said was true - irrespective of the other facts, if a cop arrests a man who's not a danger in his own home, that's stupid.

Due to the uproar, Obama later conceded that "stupid" was an unfortunate choice of words and called the officer to invite him for a beer at the White House. While settling for outcome 2 - both Obama and the Cambridge police swerve - is better than Obama and the police "crashing" into each other, it's unsettling that Obama made this compromise despite characterizing the interaction correctly.

This incident is so trivial that how he handled it does not trouble me too much, and the compromise can easily be filed under "you have to pick your battles." But it's indicative of a larger pattern whereby Obama will shoot for the best average outcome irrespective of what side is right or wrong. The concern here is the loss of justice. To take a far less trivial example, the Obama administration has so far refused to push for an investigation of Bush officials for possible war crimes even though there's ample evidence they violated the Geneva conventions and took part in torture. Again, letting that slide, "looking forward not back" might be less risky and damaging in the short run, but can we really settle for the second best option - no prosecutions, no political fallout - when the best option - justice, accountability, rule of law - is the right thing to do? In other words, at what point does Obama's conciliatory Game Theory optimization yield to bigger, universal truths?

The other side of the coin is this - and I've experienced this in my personal life with various girlfriends and even friends and colleagues: (1) just because you think you're right (or even are sure you're right) doesn't mean your adversary doesn't believe he's just as right; and (2) even if you are in fact right, it's often worth conceding to keep the peace, i.e., just because one is right doesn't mean insisting on one's rightness is the optimal course of conduct.

The hope is Obama's Game Theory tack operates only as a means to achieve optimal ends and not as an end in itself. In other words, one might have to settle for a suboptimal outcome in many battles, but only to ensure ultimately that justice triumphs in the overall war. Perhaps this is what he means when he quotes Martin Luther King, saying "the arc of the universe is long, but it bends toward justice." If not, then settling for the second-best outcome is not a sustainable strategy for achieving our most important priorities.