On the Current Confrontation with Iran
Most obviously, humility is in order. Those of us of a certain age can remember when many thought that the 1972 mining of Hanoi and Haiphong would lead to something worse than the Cuban missile crisis. In the mid-1980s, few analysts thought the Cold War would soon end. Many journalists and not a few scholars claim deep knowledge of the Middle East and the ability to predict how everyone will react, but we should recognize that the layer of regional expertise in the United States is, in reality, thin and that this knowledge, even combined with however much social science, takes us only so far.
Most of our generalizations are probabilistic, and we are in a situation of strategic interaction where multiple actors are trying to anticipate what others will do, knowing that others are doing likewise. As Erik Gartzke explained in his important article years ago, to the extent that the people scholars study come to believe these theories, they will become self-denying prophecies as people behave in ways to avoid the undesired consequences. My guess is that neither President Donald Trump nor the Iranians know what they will do next (and what they think they will do may be different from what they will do when the time comes). While it is possible for analysts to do a better job of predicting how events will unfold than the decision-makers themselves, we should not count on this. The fact that no one I know predicted that Trump would order Iranian Gen. Qassem Soleimani’s assassination should give us further pause about our ability to estimate what will come next.
Indeed, I don’t think that we are yet able to do a good job of explaining why Trump acted as he did. The media reports have been thin, if interesting. Part of the story seems to be the militia attacks in the past month or so, and if these were well-covered in the American media I missed the stories, and so we — or at least I — know little about the calculations of Iran and its aligned militias or how they were analyzed by U.S. leaders. My guess, however, is that Trump’s decision was impulsive, and I doubt if there was much analysis of the likely consequences. When the documents are declassified 20 or 30 years from now, it will be interesting (to younger colleagues) to see if the intelligence community was asked to produce an estimate of the consequences and, if it was, what was said. My years of study and experience as the chair of the CIA’s Historical Declassification Advisory Panel give me the sense that even if senior military leaders did not push for the strike (and the media stories say that they did not), it was not deeply opposed because of the hatred for Soleimani for his purported role in killing American soldiers during the Iraq War. Contrary to the conventional wisdom, the U.S. military usually favors restraint, but the bitterness against Soleimani on the part of those who fought in Iraq should not be underestimated.
For Trump, the fact that an American died in the militia attack on Dec. 27 seems to have been significant. The death of another American similarly triggered his withdrawal from an agreement with the Taliban that apparently was ready to be signed last fall. Now, as then, he has said that any attack that kills another American will be met with a violent response (although there was no reaction to the al-Shabab attack in Kenya that killed three Americans.) This is giving hostages to fate, because while someone’s decision is needed to launch an attack on an American base or embassy, once the rockets or missiles are launched, it is a matter of luck whether an American gets hit. So if Trump calculated, he must have been confident either that the other side — or really sides — would be deterred, or that there is so much empty space on the bases that an attack would not be likely to hit anyone. The Iranian missile strikes indeed did not inflict casualties, but this was not thanks to anything Trump’s administration did.
The point should be generalized. The success of coercion depends on the adversary’s choices. In making a bold move, as Trump did, he may think that he has taken the initiative and is in control. But this is not the case. Iran may choose acquiescence as the path most in its interest, but this indeed is a choice. Although the United States has great influence over how things develop, American well-being and Trump’s reelection prospects are now in Iranian hands. American and Iranian futures are interdependent.
It now seems likely that the Iranians will be satisfied with the barrage of missiles against two bases that host American forces, strikes that produced little damage and so allowed Trump to refrain from continuing the cycle. Even if Iran fired in the middle of the night to try to minimize the chance of casualties, a non-trivial degree of luck was involved. It also shows that even strife at the brink of war involves a degree of theater. We call states and leaders “actors” not only as a handy figure of speech but because they need to perform for various audiences and may implicitly work with adversaries to generate desired impressions.
Of course, these may not be the last moves. Regardless of how things turn out, the ostensible reason for the assassination of Soleimani makes little sense. If attacks were “imminent,” then they were ready to go without further intervention by Soleimani. To think they would be called off would require us to believe not only that Iran would be deterred, but that it has full control over all the local actors. Although Democrats’ demands for the intelligence behind the claims of an imminent attack make political sense, they really are a distraction.
The argument that the killings will greatly degrade the abilities of Iran and its “proxies” also is likely misguided. Here we actually have research — Jenna Jordan, Austin Long, and others have shown that decapitation does not lead to degradation in well-institutionalized organizations.
Rather than either of these rationales having been driving, I think what we have here is an instance of what organization theorists inelegantly call the “garbage can model.” This refers to cases in which a problem (here, the assault on the U.S. embassy in Baghdad) triggers not a search for effective solutions, but rather the reaching for a familiar proposal that has been developed for other purposes. It is at hand, convenient, and doesn’t call for rethinking even if it should.
Of course, the more general rationale for the strike is deterrence. By giving Iran a “bloody nose,” the United States has shown that it will no longer acquiesce to attacks and that Iran (and others) ought to expect strong reprisals to any further adventures. This could yet turn out to be true, but I doubt it. First, while Iran’s leaders presumably were surprised by the assassination, I think they always realized that something like this might occur. In other words, the American move will not radically change their estimate of how the United States is likely to respond in the future. Second, even if it has increased the perceived likelihood of a violent response, and even if this is an outcome that they and others want to avoid, there may be counterbalancing changes that more than offset this. Iran may think that the United States is now set on regime change and that refraining from provocations will do no good. Leaders of various factions in Iran or Iraq may believe that even if a violent American response is bad for their country, it could be good for them and their factions because of aroused public opinion. Third and related, local leaders or terrorists could want to precipitate more conflict between Iran and the United States and so, far from being deterred, may take this as an opportunity.
Iran’s announcement that it will no longer be bound by the restrictions of the multilateral nuclear deal (although it was careful not to say exactly what it would do) raises the question of whether Iranian leaders have come to believe that only nuclear weapons can protect them. This was a lesson previously taught by the American attacks on Libya and Iraq and may be reinforced by the fact that we have not attacked high-level North Korean officials or tried to subvert that regime.
In terms of escalation, the situation resembles an asymmetric game of chicken. Very high levels of violence are the worst outcomes for both the United States and Iran, but as Trump never tires of telling us, it would be even worse for Iran than for the United States. Washington presumably gains some bargaining advantage from the asymmetry, but only a limited amount. Both sides know that all-out violence is to be avoided, and this brings up what students of nuclear weapons call the stability-instability paradox, as each side can feel free to engage in violence at lower levels, knowing that the other side does not want to escalate to the top of the ladder. But presumably even if they have not read the relevant scholarship, both sides also know that things can get out of control and that the consequences of any further violence cannot be predicted with complete certainty.
If Iran responds with a cyber operation, we could move into unprecedented territory of prolonged and destructive cyber campaigns and reprisals. Many of us have been studying the dynamics of cyber escalation, and while it would be good for our research to have new empirical material, even without this I would endorse the conventional wisdom that there are lots of ways for such a conflict to go badly. Of course, there have been many cyber operations in the region, even destructive ones, but this could be more extensive and intense, with multiple opportunities for misunderstandings and miscalculations. Not only do the targets of attack often have trouble estimating what the other side thought it was doing, but those who launch the attack may be unable to predict its effects. In the long run, such a conflict might lead to restraining norms if not formal agreements, but getting there would not be pretty. (I should note that public and even secret knowledge of Iran’s capabilities here is limited. The talking heads say how well positioned Iran is, but it may be that Iran had to resort to kinetic strikes against the Saudi processing facilities on Sep. 14 because it had lost a good deal of its cyber capabilities.)
At least as important as the question of escalation is what the assassination has done to the region. Politics in both Iraq and Iran are characterized by internal struggles for power, and the factions and their calculations are far beyond my knowledge. It does seem as though the protests within Iran, or at least in their anti-regime aspects, have now ended, contrary to the objective of American policy (misguided as it may have been). It also appears that Iraqi nationalism, which over the past months had increasingly been directed against Iran, has turned against the United States. Although no one can be certain how deep and lasting this change will be, and the Iranian attack on Iraqi territory could cause resentment, it is more likely that one of the main consequences of killing Soleimani will be a great increase in Iranian influence over Iraq.
Another consequence may be to give the self-proclaimed Islamic State another lease on life. Increased Iranian influence in Iraq (and Syria) will presumably further alienate the Sunni populations. In addition, even if cooperation between the United States and Iran on the battlefield had ended when the last Islamic State strongholds were destroyed, the continuing effort required close cooperation between U.S. and Iraqi security forces, something that now is impossible. U.S. forces in Syria presumably also now have to prioritize force protection and can’t do much else. Trump’s speech on Jan. 8 noted that destroying the Islamic State is an interest the United States shares with Iran but did not acknowledge that cooperative progress here will be difficult at best.
One of the other ironies is that Trump and Iran’s leaders (and many others) want to get the United States out of the Middle East. This makes the militia’s attack on the American base that triggered this round of violence a bit hard to understand, and also highlights that the possible outcome of Trump’s reaction will be to get the United States more deeply mired in the region (and with even less local support than before). For those in the government who believe we are now in “a new era of great-power competition” (whether this includes Trump is hard to tell), the results are clearly discomforting. They will undoubtedly urge Trump to take from this episode the lesson that even small deployments can be dangers and that, his strong actions having shown his power and resolve, he should follow his instincts and withdraw. At this point I doubt if he knows what he will do.
Four second-order consequences come to mind. First, the increased conflict may make it much harder for a Democratic president to pursue a rapprochement with Tehran a year from now (on the assumption that a Democrat is elected). Perhaps the Iranians would believe that the United States has turned over a new leaf and that it was imperative to put relations in a good enough place so that when the wheel turns again in American domestic politics Iran will not again be in the crosshairs, but more likely is that the obstacles to better relations will have been increased.
Second, what does Kim Jong Un think? Does he see this as an opportunity to resume nuclear or intercontinental missile testing because the United States has its hands full? Or does he believe that Trump is too unpredictable and dangerous to be messed with? Or does he think that he might as well be quiet because it will be impossible to get Trump’s attention for a while? Or perhaps he is so fixated on his own situation that he doesn’t take much notice.
Third, Iraq’s prime minister, Adil Abdul-Mahdi, said that Soleimani was on his way to give him a message for the Saudi leaders as part of the recent effort of Iran and Saudi Arabia to reach out to each other. Whether true or not, this story is likely to be widely believed, and many will also assume that this could not have escaped the attention of the all-seeing U.S. intelligence community. The obvious explanation for the strike will then be that it was an American attempt to sabotage these efforts in order to bolster the policy of “maximum pressure” on Iran. The fact that most Americans would doubt that the United States would be so skillful is irrelevant — it is what the regional players believe that matters.
Fourth, just as the Cuban missile crisis was followed by a warming of U.S.-Soviet relations, the United States and Iran could learn from the current episode that they are running unacceptably high risks and that if not a rapprochement, then at least an informal understanding is needed to avoid disasters. In his remarks yesterday, Trump called for heightened sanctions against Iran, but also declared that the two countries “should work together on … shared priorities.”
World politics rarely follows straight paths.
Robert Jervis is Adlai E. Stevenson Professor of International Politics at Columbia University and author most recently of How Statesmen Think.