Nobody Loves Deterrence, But We’ll Keep Doing It Anyway

October 9, 2017

For special access to experts and other members of the national security community, check out the new War on the Rocks membership.

Editor’s Note: This is the third installment of “The Brush Pass,” a new column by Joshua Rovner (@joshrovner1) on intelligence, strategy, and statecraft.

On Oct. 1, State Department spokesperson Heather Nauert declared that North Korea “will not obtain a nuclear capability,” even though it achieved a nuclear capability a long time ago. North Korea obtained fissile material in the 1990s, conducted its first test in 2006, and has spent the last two years publicly flaunting its growing missile capability. Puzzled observers mocked the administration for boldly announcing it will never allow a country to own something it already owns. Others worried that the administration’s casual ignorance, alongside the president’s bluster, were injecting dangerous instability in the ongoing nuclear standoff.

While the Trump administration’s language is characteristically outlandish, the underlying sentiment is hardly new. Indeed, policymakers have aggressively committed to nuclear nonproliferation for decades, and are rarely willing to acknowledge nonproliferation failures. When friends like Israel and sometime partners like South Africa went nuclear, the United States allowed them to retain the pretense of ambiguity, neither admitting nor fully denying their new capability. When rivals like the Soviet Union and China were on the nuclear threshold, Washington thought seriously about preventive military strikes.

The one thing the United States almost never does is embrace deterrence. Policymakers are loath to accept the fact that rivals possess the capability to wreak mass destruction on the United States. They prefer to remove that capability, rather than deterring rivals from using it. Counter-proliferation – cajoling a state to eliminate its nuclear capability, or using military strikes to destroy nuclear facilities – holds out the prospect of restoring a safer time. Deterrence, on the other hand, is a holding action, a precarious effort to reduce the risk of terrible violence. And the holding action may not last long. Adversaries are fully capable of expanding and improving their capabilities, which can upset what one analyst famously called the delicate balance of terror.

One reason policymakers are reluctant to engage in deterrence is because they doubt all states can be deterred. Some leaders are irrational and unreceptive to deterrent threats. Others might be rational, but unreliable custodians of nuclear power who do not fully control site security and launch procedures. Relying on deterrence in these cases is insufficient, according to skeptical policymakers, because it does nothing to reduce the danger of loose nukes and accidental or unauthorized launches.

Leaders also hate the fact that nuclear adversaries restrict U.S. freedom of action. The United States maintains extraordinary military capabilities and usually enjoys a favorable conventional balance when it intervenes in other regions. This allows it to pursue a variety of interests and to extend its influence in distant locations. U.S. leaders fear that proliferation by other countries undermines these advantages and constrains their actions. They don’t like the prospect of being deterred.

Finally, leaders have moral qualms about deterrence. The perverse logic of mutually assured destruction is that nuclear-armed states are safer if they are equally vulnerable. If all sides realize that aggression is suicidal, they will not act. This logic has never been appealing to policymakers, who struggle against the idea that it is morally acceptable to leave their citizens open to annihilation and to hold millions of foreigners hostage. Ronald Reagan viewed deterrence as a “sad commentary on the human condition” and pleaded for alternatives like missile defense: “Wouldn’t it be better to save lives than avenge them?” Instead of accepting the logic of deterrence, presidents from both parties invested in offensive tools for preempting enemy capabilities, and defensive tools for protecting against an attack.

Policymakers aren’t the only ones who find deterrence distasteful. Military officers are often uncomfortable with the idea. Their professional norms stress the importance of distinguishing civilians from legitimate military targets. Nuclear weapons make this difficult or impossible, given their vast destructive power. And they would always prefer to fight at the conventional level, where the United States possesses durable advantages.

Scholars and defense analysts also struggle with deterrence. Even those who are receptive to its logic face the basic problem of making inferences about non-events. It would be much easier to test deterrence theory if there were clear variation in outcomes, but this has not been the case: No one has used nuclear weapons in anger since 1945. So while scholars often talk about when and how deterrence works – I’m no exception – they rely more on deductive logic than empirical tests. As Austin Long put it to me recently, “What is needed are novel methods of collecting data without blowing up the world.” Until that happens, academics and commentators will remain in the unsettling position of writing about immensely important issues, and sometimes making important policy recommendations, on a very thin empirical foundation. Policymakers can be forgiven for taking their advice with a grain of salt.

Everyone has a different reason for disliking deterrence. Policymakers don’t like the strategic risks involved, and officers don’t like the implications for conventional operations. Neither group is comfortable with the fundamental moral problems of nuclear weapons. And scholars, whose research ought to be useful, are inherently hamstrung by a lack of data.

Despite all these problems, however, deterrence remains the least risky option for dealing with emerging nuclear rivals. The United States usually goes through a period of consternation when a hostile state acquires nuclear weapons, but it soon learns that deterrence is the safest route to security. While U.S. leaders continue to invest in technologies they might use for preventive attacks, they are unlikely to take the risk. Meanwhile, they shore up U.S. retaliatory capabilities, and issue periodic reminders of their willingness to use them. They are not enthusiastic about deterrence, but they do it anyway.

The U.S. response to Soviet and Chinese nuclear acquisition is instructive. Years before the Soviet Union conducted its first test, high-profile hawks advocated preventive military strikes. These calls resurfaced during the Berlin airlift and the Korean War; the idea of taking out the handful of Soviet nuclear installations was alluring. But the president ultimately rejected these calls as too risky. Similarly, the Kennedy administration toyed with preventive attacks as China marched towards the bomb in the early 1960s. It worried that Mao Zedong might be so irrational that he was impossible to deter. The Johnson administration feared this as well, and it even reached out to Moscow to see about coordinating diplomatic actions to dissuade China from nuclear tests. Moscow demurred, and Johnson ultimately chose caution.

The same pattern may be at work today. President Donald Trump and his advisors stress that military options are available, and the president hasn’t been shy about threatening to use them. But these options carry enormous risks. Some observers warn that North Korea is likely to escalate quickly because its doctrine depends on striking first. Even if it doesn’t cross the nuclear threshold, it can credibly threaten South Korean with conventional and chemical weapons. In addition, military action against North Korea threatens to escalate into a wider conflict with China. The United States has good reasons to show restraint, much as it showed restraint after the Soviet Union and China went nuclear.

Deterrence is not emotionally satisfying, especially when dealing with morally odious regimes. It offers no conclusive solution to the problem of nuclear proliferation. What it does is buy time. This is significant because emerging nuclear powers tend to mellow as years go by. The moment of acquisition is fraught with danger, especially when a state demonstrates the ability to mate warheads with ballistic missiles, as is now the case with North Korea. Yet the danger steadily attenuates as nuclear powers discover the limits of what they can accomplish with their arsenals, and as other states adjust to new facts on the ground. In this way deterrence offers short and long-term benefits. It reduces the immediate risk of war, and it lays the groundwork for better relations in the future. Deterrence may be unloved, but it is irreplaceable.

Joshua Rovner is Associate Professor in the School of International Service at American University. He is the author of Fixing the Facts: National Security and the Politics of Intelligence (Cornell, 2011), and writes widely about intelligence and strategy.  

Image: U.S. Army/Staff Sgt. Keith Anderson

We have retired our comments section, but if you want to talk to other members of the natsec community about War on the Rocks articles, the War Hall is the place for you. Check out our membership at!