Thus, the next president must find ways to convince Iran that America possesses the capability to carry out a successful strike. If Iran could be persuaded that the operation would be successful, the United States would not have to carry out the strike at all, because Iran would come to the negotiating table rather than run the risk of military action.

I quote from a sophomore essay in Princeton's undergraduate foreign affairs magazine, in part because it presents this mind-set so clearly. The  claim that Iran will not come to the negotiating table is dubious; it is not Iran but the current administration which is unwilling to negotiate. However, since it is doubtful Iran will concede anything, the point is less than it might be.

But what interests me is the confident assumpttion: we can rely on the Iranian government, or indeed a statesman in general, not acting contrary to the immediate material interest of his state.  From that we go into this application of game theory, and the art of making offers Iran can't refuse.

I don't blame a sophomore for believing this; it is a novel and illuminating picture of the world, and I'm sure some of the eminent realists Woody Woo has been hiring lately present it with great vigor; the only problem with it is that it does not describe reality; and there are all too many, no longer sophomores, who believe it.

Statesmen do act in ways harmful to the immediate net interest of their states; five empires willingly engaged in war in 1914; three of them were destroyed, and the gains to the survivors scarcely made up for the loss of capital and lives in the fighting.  It would be interesting to make up a history of the world made up of such cases; it might, with enough cynicism, include almost all of human history. But I mean more than this; all parties in WWI vastly underestimated the costs, and overestimated the chances and profits of victory.

Statesmen have chosen to go to war (and do other things, but war is the most obvious) when it was obvious when they did so that it would cost more than the terms of peace they were being offered. The current Gaza disaster may be an example,  both sides of it; but there is an even clearer instance from our own history:

James K. Polk offered to buy California and New Mexico before the Mexican War; if he got his deal, he wouldn't go to war. He made the same mistake our sophomore does; his diary notes that he did not expect to have to fight, because the Mexicans would settle. They did not; they lost; they had to give up the same land, for less money. And it was predictable this would happen; the British minister in  Mexico City, and the former Minister of Foreign Affairs, both warned them, and recommended they give in.

Realists may say that all concerned had no choice: if they ducked the war, they would fall from power anyway. I'm not sure that's true; there's an argument that Ehud Barak and Labor are doing best from this war, for example, not Kadima. But this does not save the argument we began with; Ahmedinejad may be similarly forced to fight rather than surrender the national pride.