Paul Pillar argues for remembering how deterrence works and why it sometimes fails:

A principle repeatedly ignored in American discourse is that in attempting to influence an adversary’s behavior, getting him to believe he will not be punished if he behaves as we wish is just as important as getting him to believe that he will be punished if he does not so behave [bold mine-DL]. This is true not only in situations of true deterrence, in which we want to prevent something from happening, but also in situations, for which Schelling created the term compellence, in which we want the other side to take action it is not currently taking.

As Pillar goes on to say, one reason for Iranian reluctance in making concessions on the nuclear issue is the fear that Iran will continue to be punished no matter how many concessions it makes. In the Iranian case, as in others, the other government does not doubt Washington’s willingness to impose sanctions, since the U.S. has proven time after time that it will do this, but they very much doubt our readiness to lift them in exchange for meeting specific demands. There is good reason to doubt this when it is common for many Westerners to insist that the lifting of sanctions is a “reward” that should almost never be granted to rival and adversarial regimes. This problem is made worse by the tendency of many hawks to view sanctions not only as a means to force a change of behavior on a particular issue, but also see them as a weapon aimed at destabilizing and eventually toppling the government. That is one reason why Iran’s leadership has sometimes dismissed Western complaints about the nuclear program as the cover for a policy that is really aimed at regime change. Once a regime doesn’t have to fear that its survival is at stake, it is likely to be more conciliatory, but will understandably dig in its heels if it thinks that its existence is being threatened. Perversely, those that claim to be the most concerned to stop undesirable behavior from foreign regimes are the most likely to insist on inflexible, hard-line policies that make it more difficult to persuade the other government to act as ours would prefer, and they are the least likely to be willing to offer the inducements and compromises essential to reaching a deal that satisfies the main concerns of both parties. The readiness to bludgeon and coerce other states frequently makes a satisfactory resolution of crises and disputed issues more difficult, because it gives them no positive incentives for cooperation and little reason to trust assurances that further punitive measures won’t be forthcoming if they agree to cooperate.