Thanks for the interested and interesting comments. Let me add a few reactions to the discussion.
I agree that terrorist attacks sometimes (but not always) take time. I presented only a very mild statement that maybe the War reduced the probability of attack since there has been none for five years. Surely, if attacks had come, as in Spain-where it took hardly more than one year to generate a very deadly attack- the War would have been blamed, at least in part. The Spanish attack helped to defeat the government that sent troops to Iraq, and hastened a withdrawal afterwards by the newly elected government. Why Spain and not here?
I do not see how anyone can claim that Iraqis are no better off, despite the continuing number of horrible Iraqi deaths. It can hardly be doubted that the vast majority of Iraqis supported the overthrow of Saddam. I suspect that a similar majority would oppose the quick withdrawal of U.S. troops. That should not be the reason for keeping them, but we should be clear on where Iraqis stand.
William Nordhaus is an outstanding economist, and I have learned a lot from his many original writings. But an estimate of the war's cost of between $100 billion and $2 trillion is hardly a sharply defined basis for decision-making.
The $7 million estimate for a statistical value of life is for young persons taking various risky decisions. Of course, soldiers are volunteers, but so too are men and women who take risky construction jobs, or join the police force. Estimates of the statistical value of life are trying to measure the amount of compensation people require in order to induce them to take on additional life-threatening risks. I should add that different studies come up with different number, and I consider $7 million on the high side.
Posner and I have no NBER paper on the cost of the war- one commentator is confused on this. I assume he means the paper by Davis, Murphy, and Topel that we refer to. They are the ones who consider continuing containment as one alternative to going to war-they also discuss other possibilities. Perhaps as claimed in the comment, they overstate the cost of containment, but surely considering the cost of alternatives is better than concluding about the desirability of the War without discussing any alternatives?
There can be only one Muppy, my former student. The suggestion of insurance on a war is original, but I do not see how a country can take out insurance on fighting a war that runs into hundreds of billions of dollars. There would have to be a large and reliable market where people can bet in aggregate large sums against the government. But various internet betting markets could have wagers on the costs in any year, the number of casualties, whether any WMD would be found, etc. That may be a bit macabre, but it does encourage many alternatives to official estimates of the costs and benefits of fighting a war. Recall that there was a betting market on whether Larry Summers would resign as President of Harvard. Unfortunately, for Harvard and other universities, those betting that he would resign won.