I KNEW THOSE CLASSES WOULD COME IN HANDY

A brief e-mail exchange with Palmer Haas about the intellectual conflicts one faces in thinking about the impending war with Iraq brought to mind the issue of error in research design, specifically Type I and Type II error. Basically, Type I errors occur when you fail to reject a hypothesis which is not in fact true. Type II errors are the opposite–you reject a hypothesis when it was in fact true, either through faulty methodology or too high an evaluation threshhold. A quick Google search found this very good, and short, primer, applying it to the criminal justice system.

With respect to the current Iraqi crisis, it works like this:

    Alternative Hypothesis: Saddam is building weapons of mass destruction (WMD) which will be used to harm the US or our interests.

    Null Hypotheses: Saddam is not building WMD and/or they will not be used to harm the US or our interests.

    Type I Error: We go to war with Saddam to stop him from doing something bad that he wasn’t going to do anyway, either because he couldn’t or had no intention to.

    Type II Error: We fail to go to war and Saddam kills lots of Americans, our our allies, or sells them to terrorists who kill lots of Americans or our allies

If France, et.al. are right and we go to war anyway, we’ll commit a Type I error. The cost? We spend billions of dollars, kill an unspecified number of Iraqis, and get some American soldiers killed. For no good reason. That’s bad. REALLY, REALLY bad. On the other hand, if we decide to defer to the Security Council, give inspections more time, and Saddam builds and causes to be deployed WMD against us, we commit a Type II error. The cost? An unknown number of Americans and/or our allies get killed. When we could have prevented it. That’s also REALLY, REALLY bad. Indeed, it’s worse than the first error.

Another thing that makes this tough is that we’ll never know if we commit a Type I error. Why? Because you can’t prove a negative here. If we topple the Iraqi regime and no WMD are deployed against the US, we’ll never be able to prove that they WOULD have been used against the US.

In addition to the consequentialist analysis–which type of error would we most like to avoid (and, as the article linked above points out, you can’t totally eliminate both)–there is the probablistic argument. Which of the hypotheses are more likely to be true? Are you willing to risk a Type II error here based on your confidence that Saddam either lacks the ability to get WMD or the will to deploy them? I am not.

FILED UNDER: Best of OTB, Iraq War, Political Theory, , , , ,
James Joyner
About James Joyner
James Joyner is Professor and Department Head of Security Studies at Marine Corps University's Command and Staff College. He's a former Army officer and Desert Storm veteran. Views expressed here are his own. Follow James on Twitter @DrJJoyner.