Thursday, January 29, 2004

In the social sciences (and in medicine and other fields that use statistical analysis), students are taught about Type I and Type II errors. In Type I errors, the results offer a false positive--something that seems to be, really isn't. In Type II errors, something that doesn't seem to be, really is. Imagine a test for cancer. A Type I error on the test would indicate that you have cancer, when you really didn't. This, of course, would cause all sorts of anxiety and worry, but further tests and procedures would ultimately show that you did not have cancer. In contrast, Type II errors are far more worrisome, since they would indicate that you did not cancer when you really did. Patients would remain ignorant of their peril until, perhaps, it was too late. As such, when devising any sort of test like this, you usually want to err on the side of caution by having Type I rather than Type II errors. Better a false positive than a false negative.

Why this primer on statistics? It seems to me that the concept of Type I and Type II errors is highly relevant to current discussions about the absence of WMD in Iraq. Any intelligence estimate regarding WMD is susceptible to Type I and Type II errors. In a Type I error, we would predict that a country has WMD when in fact they do not. A Type II error would predict no WMD when in fact a country did have them.

From all accounts, the Bush administration (and the Blair government, and the French, and the Germans, and the UN, etc.) committed Type I errors in their assesment of Iraq's WMD capability. They thought something was there that really was not. As with the cancer tests, Type I errors cause problems, but they are far better than Type II errors. The Bush administration decided to err on the side of caution in assessing Iraq's WMD capacity. Better to run the risk of finding out there was no Iraqi WMD threat, than to find out too late that such a threat really did exist.

In the aftermath of 9/11 I can't fault the Bush adminstration for this. 9/11 showed that the prospect of a Type II error in intelligence assessment was potentially catastrophic. Yes, Type I errors are not to be dismissed as irrelevant. Going to war has real costs and risks, but in this case, the predicted risk of going to war was far less than the risk of not going to war. Add to that the nature of Saddam's regime and the cost of not going to war becomes intolerable. As Tony Blair put it in his speech to Congress this summer:

Can we be sure that terrorism and weapons of mass destruction will join together? Let us say one thing: If we are wrong, we will have destroyed a threat that at its least is responsible for inhuman carnage and suffering. That is something I am confident history will forgive.

But if our critics are wrong, if we are right, as I believe with every fiber of instinct and conviction I have that we are, and we do not act, then we will have hesitated in the face of this menace when we should have given leadership. That is something history will not forgive.

No comments: