Editor’s Note: Back in February I riffed on a post by Erik Voeten in which Erik discussed two articles in International Organization (IO). One, by our colleague Matt Kroenig, argued that nuclear superiority gives states advantages in crisis bargaining (PDF). Another, by Todd Sechser and Matthew Fuhrmann, rejects this claim (PDF).
After the two posts sparked some interesting discussion–both on- and offline–I approached all three about doing a mini-symposium at the Duck of Minerva. They agreed. Earlier we ran Kroenig’s piece. In this post, Sechser and Fuhrmann critique the claims he made in his IO article. Both sides will have an opportunity to respond.
Iran’s nuclear program has been a source of international concern for a long time. Some observers in Israel and the United States are now pushing for war, arguing that a nuclear Iran would brandish its capability like a club, waving it around recklessly and bullying neighbors and rivals into submission with nuclear threats.
This fear stems from a common belief that nuclear weapons are more than just weapons of self-defense and deterrence – they are offensive diplomatic tools as well. But is this view correct? Are nuclear weapons useful for coercion and intimidation?
Nuclear weapons have little impact on the effectiveness of coercive threats.
We recently conducted a study that found a surprising answer. Our study, published in the journal International Organization, investigated whether nuclear states enjoy more coercive success than other states. We found that they do not: nuclear weapons have little impact on the effectiveness of coercive threats. (Note that we use the term “coercive” to refer to attempts to persuade an adversary to change its behavior or give up something valuable. This is distinct from deterrence, where the goal is to preserve the status quo, not change it.)
Our conclusion challenges conventional thinking about nuclear weapons, which holds that nuclear weapons are useful for coercion – and not just deterrence – simply because they are so destructive. This view argues that nuclear-armed states can more easily compel others to make concessions in international crises – and that they can do so without actually going to war. But the conventional view fails to fully appreciate two important limitations of nuclear weapons.
First, nuclear weapons are not very good for seizing disputed objects, like territory. It would make little sense, for example, for Pakistan to try to take Kashmir by launching an offensive nuclear attack against Indian forces there. Doing so would kill ethno-religious kin and could render portions of the land uninhabitable.
Instead, Pakistan would be more likely to say (or insinuate) something like “Give us Kashmir or we will attack New Delhi.” But this brings us to a second point: carrying out a coercive nuclear threat would be tremendously costly. A state that launched a nuclear attack to achieve a coercive objective – that is, to obtain something it didn’t have already – would provoke an enormous international backlash. The consequences could include economic sanctions, diplomatic isolation, efforts by neighbors to acquire their own nuclear weapons, and even military interventions against the coercer.
Of course, the costs of enacting a threat are only one part of the equation, and these costs might be worth paying if one’s survival were at risk. But states making coercive threats rarely face such dire circumstances. Indeed, a key feature of coercion is that the coercer is seeking something that it has already been living without – such as a piece of disputed territory, monetary reparations, or reversal of an unfavorable policy. While these stakes are hardly trivial, in most cases they are not so vital to the coercer’s security as to outweigh the inevitable backlash that would follow an offensive nuclear strike. Pakistan would desperately like to acquire Kashmir, but having lived without it for decades, Pakistan’s survival hardly depends on possessing it. The very nature of coercion thus works against the credibility of nuclear coercive threats.
Nuclear powers should be no more successful, on average, when trying to coerce their adversaries.
A simple prediction follows from our argument: nuclear powers should be no more successful, on average, when trying to coerce their adversaries. Our study tested this proposition by examining compellent threats in international crises. (A compellent threat, as originally conceived by Thomas Schelling, is a coercive demand designed to compel an adversary to change the status quo.) Specifically, we set out to determine whether nuclear states make more effective compellent threats than nonnuclear states.
After looking at more than 200 compellent threats, we found that nuclear states issue successful threats just 20 percent of the time, while nonnuclear coercers succeed more than 30 percent of the time. These results also hold when we apply more advanced statistical techniques and control for confounding variables. This is not the pattern we would see if nuclear weapons are useful tools of coercion.
The recent experience of the United States bears this logic out well. Consider how many American coercive threats have failed in the last quarter-century: Panama (1989), Iraq (1990-91 and 2003), Afghanistan (1998 and 2001), and Serbia (1999), for starters. One might also include implicit threats against North Korea and Iran on this list. By contrast, U.S. coercive successes during this period are few and far between, including just Haiti (1994) and Iraq (temporarily, in 1997). In all of these cases, the United States possessed a diverse nuclear arsenal and nuclear superiority over its opponent. Yet most of these threats were unsuccessful.
A recent article by Matthew Kroenig, in the same issue of International Organization, appears to challenge our findings. Kroenig argues that nuclear weapons do convey advantages in crisis bargaining, but only when a state has nuclear superiority – that is, when it has more nuclear weapons than its opponent. After analyzing 20 crises, he reports “a powerful relationship between nuclear superiority and victory in nuclear crises” (p. 143). Our study of 200 threats, by contrast, found no such relationship.
A closer look at Kroenig’s evidence, however, reveals that it is actually quite supportive of our argument. First, according to the International Crisis Behavior project, the nuclear-superior side actually “won” just 10 of the 20 crises that Kroenig evaluates, losing or settling for a stalemate in the other 10. (See the table below.) In other words, the effect of nuclear superiority basically amounts to a coin flip: on average, the side with nuclear superiority fares about as well as the side without it.
| Crisis |
Start Year | Winning Side |
Victory for Nuclear Superiority? |
| Korean War |
1950 | None |
No |
| Suez crisis |
1956 | United States, Soviet Union |
Yes |
| Berlin deadline |
1958 | None |
No |
| Berlin Wall |
1961 | Soviet Union |
No |
| Cuban missile crisis |
1962 | United States |
Yes |
| Congo crisis |
1964 | United States |
Yes |
| Six-Day War |
1967 | United States, Israel |
Yes |
| Sino-Soviet border war |
1969 | None |
No* |
| War of attrition |
1970 | None |
No |
| Cienfuegos submarine base |
1970 | United States |
Yes |
| Yom Kippur War |
1973 | United States |
Yes |
| War in Angola |
1975 | Soviet Union |
No |
| Afghanistan invasion |
1979 | Soviet Union |
Yes |
| Able Archer exercise |
1983 | None |
No |
| Nicaragua, MIG-21s |
1984 | None |
No |
| Kashmir |
1990 | None |
No |
| Taiwan Strait crisis |
1995 | United States |
Yes |
| India/Pakistan nuclear tests |
1998 | None |
No |
| Kargil crisis |
1999 | India |
Yes |
| Indian Parliament attack |
2001 | India |
Yes |
*Kroenig recodes this case to be a Soviet victory and therefore a victory for nuclear superiority, but the creators of this dataset coded the outcome as a stalemate, not a Soviet victory.
Second, many of the “victories” for nuclear-superior states in Kroenig’s data are successes for deterrence, not coercion. For example, Kroenig counts the 1995-96 Taiwan Strait crisis as a “victory” for the United States over China. During this crisis, China conducted military exercises in the East China Sea and issued hostile statements towards the government in Taipei, mainly to prevent pro-independence moves by the Taiwanese government. To protect Taiwan, the United States deployed an aircraft carrier and other ships to the Strait. Kroenig counts this as a U.S. victory, presumably because China did not attack Taiwan. But even if American nuclear superiority contributed to this outcome – which we doubt – the effect of the U.S. nuclear arsenal was to deter, not coerce.
Overall, Kroenig’s study provides little reason to doubt our conclusion that nuclear (and nuclear-superior) states have no particular advantage in coercive bargaining.
All of this implies that fears about Iranian nuclear blackmail are at least somewhat overblown. A nuclear Iran might indeed attempt to use its nuclear arsenal for coercive purposes, but history suggests that it is unlikely to succeed. Coercive nuclear threats have rarely worked in the past, and we see no reason to expect that pattern to change in the future.
Though their “offense is the best defense” strategy is well executed, this reader wishes that Fuhrmann and Sechser had addressed what is arguably Kroenig’s most biting set of criticisms:
“According to MCT [Sechser’s dataset], not only are nuclear weapons not correlated with compellent success, nothing is correlated with compellent success. In Table 1, we see that nuclear weapons of the challenger, nuclear weapons of the target, the conventional balance of power, and the stakes in the crisis, all fail to
reach statistical significance. (I should note that the tests also exclude many obvious controls, like regime type). The only variable that consistently reaches statistical significance in Table 1 is a variable they call “resolve,” which actually measures whether a state engages in a show of force during the crisis. This failure to present almost any meaningful results in the published article is not very helpful.
To see what is really going on in the data, therefore, let’s turn to the tables published in the online appendix. According to Tables 2 and 3 in the online appendix, MCT tells us the following about interstate coercion: states without nuclear weapons are more likely to issue successful compellent threats than states with nuclear weapons; compellent threats are more likely to succeed against states with nuclear weapons than against states that lack nuclear weapons; and, the less conventional military power a state possesses relative to its opponent, the more likely it is to issue successful compellent threats. These findings suggest that military weakness is a key asset in coercive diplomacy! This flies in the face of 2,000 years of international relations research. The weak do what they will and the strong suffer what they must? If they found that nuclear possession was statistically insignificant and the rest of the findings made sense, then that would be one thing. But, it is impossible to tell a plausible story to account for these findings. There is clearly a major problem with the way the MCT is coded that biases the results in favor of weak states.”
I want to clarify that F&S are not responding to K’s post. They are responding to his article, just as K is responding to the F&S research note. They have an opportunity to rebut each other’s *posts* if they so desire.
Indeed, I moved their piece forward precisely to reduce the likelihood that readers would think that Part II is a rebuttal of Part I.
However you dice it, slice it, nuclear weapons are good for deterrence only, not coercion. Somebody’s law: “Nations with nuclear weapons don’t get attacked.” Except for some terrorism, maybe. I think that’s the governing rationale in Teheran.