Bush and Counterfactual Confidence

By Shankar Vedantam
Monday, July 30, 2007

In the face of mounting public and political opposition to the war in Iraq, recent reports from the White House suggest that President Bush remains serenely confident.

Bush's confidants report that the president believes he will be vindicated by history. He keeps Churchill and Lincoln close at hand. No matter how tough the situation in Iraq, Bush remains confident about his decision to go to war because he believes that things would have been much worse otherwise.

"Obviously, it was a difficult decision for me to make -- to send our brave troops, along with coalition troops, into Iraq," Bush said at a recent press briefing about the Iraq situation, where he faced a barrage of questions about flagging support for the war. "I firmly believe the world is better off without Saddam Hussein in power."

Bush's argument is based on something known as a counterfactual. In his mind, the president has run an alternate view of history -- one that imagines Saddam Hussein still in power -- and has come to the conclusion that deposing the Iraqi leader was better.

Bush is not alone in using counterfactual thinking. Coming up with what-if scenarios is how people make sense of the world. When we make a financial decision that turns out poorly, we imagine going back in time and not investing in that stock or buying that house. That scenario looks rosier -- it is an upward counterfactual. But let us say we make a good financial decision. When we imagine not buying that stock or that house, we contrast the money we have made with the money we might have lost had we not made the investment -- producing a downward counterfactual.

But what is dangerous about counterfactuals is that while they may seem reasonable, they easily become a way for us to confirm what we already feel. Bush might not conclude that the war was the right decision because he has reached for a downward counterfactual; he might have reached for a downward counterfactual because he feels the war in Iraq is right.

The basic problem with counterfactual reasoning is there is no way to test your theory. Bush can't actually go back in time and not invade Iraq and see whether things would actually be worse than they are now. Because the arrow of time runs in only one direction, counterfactuals cannot be disproved. (Indeed, this may be why they are so attractive in political reasoning.)

Philip Tetlock, a professor of organizational behavior and political science at the University of California, has found that the careless use of counterfactuals is one reason politicians and experts are often wrong in their predictions.

"History does not give us control groups," he said. With counterfactuals, "the control groups are all being run in the imaginations of the analysts."

Tetlock's large study found that politicians and pundits were rarely better than non-experts in predicting the course of historical events. But he found that experts who were more cautious about using counterfactuals -- who explicitly reminded themselves that they were coming up with scenarios that could not be verified -- were more accurate on average than those who used counterfactuals blithely.

In his book "Expert Political Judgment: How Good Is It? How Can We Know?" Tetlock divided experts into groups that he called foxes and hedgehogs. Both groups used counterfactuals, but the foxes used them warily whereas the hedgehogs used them confidently. Not surprisingly, hedgehogs tended to be far more partisan than foxes.

"How much restraint do you exercise on your counterfactual imagination?" asked Tetlock. "If you allow your partisan imagination to dominate completely, that is a danger sign in how you think about the future. If you exercise restraint, you are willing to acknowledge dissonant possibilities."

Until recently, the Bush administration's Iraq plans have been mostly the work of hedgehogs. The pragmatic recommendations of the Iraq Study Group, led by James Baker and Lee Hamilton, however, are quite clearly the work of foxes.

Bush's heroes, Lincoln and Churchill, offer a study in contrasts. Lincoln leaned toward fox, Churchill toward hedgehog. Lincoln was open to dissent, even within his own Cabinet, and was alert to nuance. Churchill allowed few doubts. Each man was perfectly designed for his historical moment.

Churchill's single-mindedness helped Britain overcome the existential threat of Nazi Germany during the darkest days of the Battle of Britain. But his stubbornness also blinded him to his mistakes. If Churchill was far ahead of the curve in recognizing the menace of Hitler, he was far behind the curve in recognizing that Britain's colonial empire was headed for history's dustbin.

Tetlock's study is not about proving that foxes were better than hedgehogs or the other way around. Rather, it is about the consequences of different styles of thinking. Tetlock did find, for example, that when they make the right calls, hedgehogs are far more likely than foxes to be spectacularly right.

Unfortunately, hedgehogs are also far more likely than foxes to be spectacularly wrong.

© 2007 The Washington Post Company