I am delighted to welcome a guest post by Matthew Fuhrmann (Texas A&M) and Sarah Kreps (Cornell). This post is based on articles they published on the causes of military strikes against nuclear facilities in the Journal of Conflict Resolution (non-gated version) and the consequences of those strikes in the Journal of Strategic Studies (non-gated version).
The crisis over Iran’s nuclear program is heating up, and some in the United States and Israel have indicated that military action is on the table if sanctions and diplomatic efforts fail. Is Israel or the United States likely to bomb Iran’s nuclear facilities? Would the use of military force significantly affect Iran’s ability to build nuclear weapons?
To find answers to these questions, we decided back in 2007—following Israel’s bombing of a nuclear plant in Syria—to systematically examine the historical record of attacks against nuclear programs.
To begin, we assembled a dataset of all instances since 1941 in which states undertook strikes against nuclear facilities. Those cases include both wartime cases—allied strikes against German facilities in WWII, strikes against Iran’s facilities during the Iran-Iraq War, and attacks against Iraq’s facilities in the 1991 Gulf War—as well as “bolt from the blue” attacks including Israel’s 1981 strike against Iraq’s nuclear reactor known as Osiraq. We also identified all known instances in which countries seriously considered attacking proliferators—but ultimately chose not to strike.
Using this dataset, we analyzed the factors that are most likely to trigger attacks and the circumstances under which attacks are likely to succeed. Some important “lessons” for the Iranian case emerged from our analysis.
What Causes Countries to Attack?
In some respects, Iran fits the profile of a proliferator that is likely to prompt talk of military action. Our statistical analysis showed, for example, that countries are much more likely to consider attacking authoritarian proliferators (such as Iran) than democracies with nuclear weapons programs.
What explains this finding? Nondemocratic leaders are less accountable to domestic audiences (such as legislatures, courts, and the public). This raises fears that authoritarian states will be more likely to engage in belligerent actions such as using (or threatening to use) nuclear weapons against their adversaries. Concerns such as this tend to be overblown, but policymakers often argue that authoritarian states will act as “irresponsible” nuclear powers. For example, President Kennedy worried that China’s regime type (and ideological fervor) would compel Beijing to behave more aggressively if it built the bomb.
It is therefore not surprising that Israel and the United States are considering striking the Iranian nuclear program. Yet, our analysis suggests that an attack (as opposed to serious consideration of the military option) is less likely than mainstream media coverage often implies.
A history of violent conflict between two states is typically necessary for one state to attack another’s suspected nuclear facilities. The reason is that previous violent conflict heightens the fear that the bomb, if acquired, would be used to resolve disputes that had caused the previous conflict. In the absence of conflict, states are unlikely to feel sufficiently threatened by proliferation to take on the risks that can accompany the targeting of nuclear facilities. This explains, in part, why most attacks occurred during ongoing wars.
Only two peacetime strikes have occurred since 1941: Israel’s raids against Iraq and Syria. Multiple wars preceded both of these attacks. For example, Iraq and Syria were part of the Arab coalitions that fought Israel in 1967 and 1973. These conflicts contributed to the belief in Israel that an Iraqi or Syrian bomb represented an existential threat.
Perhaps it is this fact—that Israel and Iran do not have a history of direct interstate war—that has caused the last two Mossad chiefs to suggest that a nuclear-armed Iran does not pose an existential threat, implying that an attack against Iran would therefore be unwarranted. Relations between Iran and Israel (and between the United States and Iran) have certainly been rocky since 1979, in part because of Iran’s support for groups such as Hezbollah. Yet the lack of a history of overt war between the two substantially decreases the likelihood that either state will attack Iran—although it does not imply, of course, that a raid will definitely not occur.
Would an Attack Work?
There is another reason why an attack is less likely than speculation in the media might suggest: attacking Iran is unlikely to “work.”
In our research, we found that some past attacks hindered the target state’s ability to build nuclear weapons. However, a number of the conditions that led to some success in the past are not present in the case of Iran.
First, Israel was able to take out Iraq’s Osiraq facility and Syria’s nuclear plant at al-Kibar because both targets had concentrated their facilities in one location. Iran, however, has likely learned from those cases. Its nuclear program is far more diffuse, meaning that multiple facilities would need to be destroyed in order to significantly affect its capacity to build the bomb.
Second, Iran may be more like Iraq in the early 1990s than many people acknowledge. One of the reasons the United States and its allies had limited success in destroying Iraq’s nuclear plants during the 1991 Gulf War is that they simply did not know which facilities existed and where they were located. It is possible and perhaps likely that Iran possesses clandestine nuclear facilities. After all, back in 2009 Iran revealed a previously secret uranium enrichment plant near the city of Qom, raising the question of whether there are other facilities that it has not disclosed.
Borrowing Rumsfeldian language, we might say there are known unknowns about the program; we know additional facilities may exist but do not know their locations. Needless to say, it would be impossible to include clandestine facilities on a target list, which would naturally limit operational success if those facilities did in fact exist.
Third, attacking states have sometimes reaped indirect benefits when they bombed a proliferator’s nuclear facilities but here too potential attackers should not count on similar benefits when it comes to Iran.
For instance, strikes sometimes “stigmatized” a nuclear program, making foreign suppliers less willing to provide nuclear assistance. France—a critical supplier for Iraq prior to 1981—refused to aid Baghdad’s nuclear program after the Osiraq raid, to cite one example. Although Iran received critical nuclear assistance in the past, it does not depend on external support to the same degree as countries such as Iraq and Syria did at the time they were attacked. And unlike its Arab neighbors, Iran has the human capital to rebuild any facilities that are destroyed in relatively short order.
Thus, there are good reasons to be skeptical that military strikes would work. And if the likely benefits are limited, the discussion about the expected costs of attacking—for example, whether strikes would strengthen the regime, unleash Iranian proxies against the West, lead to regional war, etc.—becomes moot.