Satyam Khanna finds Fred Kagan arguing that "Well, there’s nothing we can do short of an attack to force Iran to give up its nuclear program. … At the end of the day, the only way that you can make for sure that doesn’t happen is with an attack."
It can't be emphasized enough that this is dead wrong. If we bomb Iran we will, presumably, destroy some stuff. But we'll have know way of knowing exactly how much we destroyed or what we left undestroyed. Most likely we'll do no damage whatsoever to Iranian know-how. We may delay the Iranian program by some extent. But if Iran responds by boosting funding for their nuclear program, or if third-party countries respond by decreasing their level of cooperation with the United States we may speed their nuclear program. It's hard to say in advance except you can say for certain that bombing won't "make sure" of anything. The only way to make sure that a nuclear arms race in the Middle East is avoided is to give other countries (Israel, Turkey, Saudi Arabia, etc.) certainty about where Iran stands.
That means verifiable disarmament. That means inspections, etc. And that means Iran needs to agree to verifiable disarmament. That's not to say that we can't wield sticks along with carrots in seeking an agreement with Iran, but still an agreement is absolutely vital. A bombing raid won't accomplish anything, and anyone who says otherwise is being ignorant.
Matthew Yglesias is a former writer and editor at The Atlantic.