War Stories

Mattis Goes Nuclear

Trump’s secretary of defense has recently adopted some dubious and dangerous ideas about nuclear strategy.

Defense Secretary Jim Mattis walks past a protester holding a sign that says, "BAN THE BOMB. CODE PINK."
Defense Secretary Jim Mattis walks past a protester as he arrives at a hearing on the Nuclear Posture Review on Capitol Hill on Tuesday. Mark Wilson/Getty Images

James Mattis has fully joined the nuclear tribesmen.

As recently as 2015, Mattis urged Congress to reassess the need for the triad, the long-standing practice of placing nuclear weapons on three types of platforms—land-based intercontinental ballistic missiles, submarine-launched missiles, and long-range aircraft. He even advocated dismantling the land-based missiles. He argued that their hair-trigger posture increased the chance of nuclear war.

His views at the time reflected his status as a just-retired Marine four-star general—the Marines being the only branch of the U.S. military that has never possessed nuclear weapons—and his friendship, as a fellow at Stanford University, with prominent arms-control advocates on campus, notably former secretary of defense William Perry.

But now Mattis himself has been secretary of defense for a year, and, as he acknowledged at a House hearing Tuesday morning, his views have evolved. He came in to the Pentagon, he testified, “wanting to challenge just about everything,” but after speaking with experts of all stripes, he realized that the land-based ICBMs are “a stabilizing element” and that other weapons he once looked at askance also have their merits.

“We must look reality in the eye and see the world as it is, not as we wish it to be,” Mattis wrote in the introduction of the Nuclear Posture Review, a 75-page document that the Pentagon released last week. In a Washington Post article on Monday called “How Mattis Changed His Mind on Nuclear Weapons,” Thomas Karako, a think-tank nuclear strategist, is quoted as saying that Mattis, like many incoming officials over the years, was “mugged by reality.”

As someone who has studied the nuclear world for 40 years and has interviewed hundreds of its denizens for a book and many articles, I have learned this: When it comes to nuclear strategy, there is no reality. The weapons are real, and their destructive power is cataclysmic. But the countless attempts to harness this destruction into an elaborate war-fighting strategy are excursions into metaphysics, not the hard-boiled realism that its purveyors like to believe.

Mattis owns a famously vast library, and he has said that, in coming to his new views on nuclear matters, he read many books on the subject. This kind of scholarly rigor has served him well in the past, but when Mattis commanded troops in Iraq, he saw connections between the battles he was fighting and the campaigns he’d studied in the classic volumes on strategy—because those volumes were based on what has actually worked and not worked across centuries of real warfare. If Mattis ever finds himself staring down the abyss of a nuclear war, he’s likely to find the books he’s consumed nearly useless; the scenarios they spin, however tightly, have no grounding in reality, as no one has ever fought this kind of war.

In a fundamental sense, Mattis seems to know this. At the hearing, he quoted a line from the Nuclear Posture Review stressing that the United States would use these weapons only “in the most extreme circumstances.” He also noted that he never says “nuclear strategy” but rather “nuclear deterrence strategy,” suggesting that the main goal is to deter nuclear war, not to fight one. This is assuring, but he also said that a deterrent isn’t credible unless your opponent believes that you would actually use it—in other words, that you would retaliate to a nuclear strike with a nuclear counterstrike—and that, to foster this impression, you need plans and weapons that enable you to do this.

The nuclear debate that embroiled hawks and doves during the Cold War, and that seems primed for a revival now, was in good part over the question of what you need for deterrence: How far do you need to go toward planning to fight a nuclear war to convince your opponent that you will do so if necessary, and at what point will these preparations have the effect of increasing, rather than lowering, the chance that a nuclear war breaks out?

In the course of immersing himself in this debate, Mattis, it seems, has gone native. He has bought in to the premises of the more hawkish faction of this debate, which once dominated this discussion and, in the post–Cold War years, has continued it—all but unnoticed by the wider public—in specialized military branches and think tanks.

Let’s take a close look at his conversion on the question of land-based ICBMs. The argument for dismantling them is straightforward. These missiles are the most lethal weapons in the nuclear arsenal—accurate and powerful enough to destroy Russia’s ICBMs, which sit in underground silos. Yet, by the same token, they are also the most vulnerable weapons—the Russians’ ICBMs can destroy ours, just as ours can destroy theirs. Both sides’ missiles can reach their targets in less than a half-hour.

In other words, the very existence of these weapons creates a hair-trigger situation. If a crisis is building, one side might feel compelled to launch a first strike on the other side’s ICBMs—if just to pre-empt the other side from launching a first strike on its ICBMs. Dismantling the ICBMs removes this hair-trigger pressure. This was the argument—long held by many arms-control advocates—that Mattis recited back in 2015.

In recent years, the more hawkish advocates in this debate have come up with a new rationale for keeping the land-based ICBMs. We have 400 of these missiles scattered around remote areas of the United States. To have a high confidence of destroying them, Russia (or some other enemy) would have to launch 800 warheads, two at each target. That’s a complicated attack—and the radioactive fallout would kill tens of millions of Americans, ensuring that we would retaliate. In other words, these advocates conclude, the ICBMs strengthen deterrence. This is the argument that Mattis has now bought.

There are three things wrong with this argument. First, back during the Cold War, these same advocates warned that the Soviet Union might fire 2,000 warheads at the 1,000 ICBMs that we had at the time, then threaten to launch its remaining nukes at American cities if we didn’t surrender—and that, faced with this dilemma, the American president would surrender and not retaliate. It was a crazy argument at the time, but, more to the point, it was made to rationalize a new generation of ICBMs—just as the new theory is made, chiefly, to rationalize retaining the ICBMs.

Second, the U.S. also currently has 280 submarine-launched ballistic missiles (on 14 submarines) and 60 nuclear-capable bombers—loaded, all told, with 2,768 nuclear bombs and warheads. That’s more than five times the number of nukes that the ICBMs carry. Half the submarines roam underwater, undetectable, at any one time. (In the event of a crisis, the other half could be put out to sea, as well.) And the bombers—the third component of the triad—could take off from their bases quickly too. In other words, the weapons on submarines and bombers should discourage any foreign leader from contemplating a nuclear attack on American soil—at least as much as a few hundred ICBMs do, if not more so.

Third, let’s say there’s something to the theory that ICBMs are needed to soak up Russian warheads and that three nuclear platforms (air, land, and sea), as opposed to just two, make an enemy attack even a little more complicated—and, therefore, a little more unlikely. Do we need 440 ICBMs? Would, say, 100 be enough—or a dozen? It seems unlikely that the number needed for a sufficiently absorbent sponge just happens to be the number we now have. And do we need to replace them with a new generation of these sponges, at great cost, as Mattis is now proposing? The current ICBMs, called Minutemen, are old, but they’ve been refurbished many times, from top to bottom. No one asked these questions at Mattis’ House hearing

The other controversial program that Mattis once opposed, but now endorses, is a new low-yield warhead for at least some of the Trident II missiles on U.S. submarines. (Low-yield means warheads with the explosive power of 5–10 kilotons, far smaller than their current yield of roughly 100 kilotons. Even so, it’s worth noting, the bomb that destroyed Hiroshima was “only” 14 kilotons.)

Mattis’ argument, as laid out in the Nuclear Posture Review, is that the Russians have low-yield weapons and that they’ve simulated using them in military exercises that simulate a war with the NATO nations in Europe. Therefore, we have to match this capability, so the Russians don’t think that they can lob some mininukes at Western Europe, as a pressure tactic, leaving us no choice but to surrender or to fire back with much more powerful weapons, which would prompt the Russians to fire back at us.

There are two problems with this argument. First, the United States already has about 1,000 low-yield nuclear weapons. Rep. Joe Wilson, Republican of South Carolina, did raise this at the House hearing on Tuesday. Mattis replied that those weapons are on bomber aircraft, which could be shot down by Russian air-defense weapons. However,  528 of those weapons are on air-launched cruise missiles, which fly at low altitudes after being fired by aircraft from far outside the battle zone—in other words, they are much harder to shoot down.

Second, deterrence requires the ability to respond in kind to an attack, but it’s a bit baroque to insist that we need to respond with exactly the same kinds of weapons. The trick to deterrence is convincing adversaries and allies that we will respond. This trick involves a combination of military firepower and diplomatic signaling. It doesn’t require copying whatever the adversary does.

At Tuesday’s hearing, Mattis committed this fallacy in another, particularly crude manner. Several times, he pointed to a chart, which he had on display, listing the number of new nuclear weapons fielded in the past decade by Russia (14), China (nine), and the United States (one). Since they are building new nukes, we need to build new nukes, too.

But the obvious question, which no legislator asked, is, why? As long as we have enough weapons to deter an adversary from attacking us, and as long as this stance seems persuasive, the rest is secondary. (It should be noted that Russia isn’t building more nuclear weapons. The United States and Russia signed the New START agreement in 2010, reducing the number of weapons in both sides’ arsenals, and both sides are in compliance with that accord.) The relative number of production lines has never been a measure of strength in the balance of power. If Russia and China want to waste their money, it doesn’t mean we have to waste ours too.

By all accounts, Mattis is a smart man who likes a lively debate. Why did he change his mind on nuclear weapons? One answer is that people who assume high posts of power almost always change their minds on nuclear weapons. Barack Obama—who came into office wanting to abolish nuclear weapons (eventually) and whose Nuclear Posture Review called for reducing the role that nuclear weapons play in military policy—nonetheless upheld the sanctity of the nuclear triad (including land-based ICBMs) and declined to adopt a “no-first-use” nuclear policy. Though he canceled George W. Bush’s plans to build a low-yield Trident II warhead, he approved every other new missile, bomber, and submarine that the Trump administration is now continuing. Trump’s $1.3 trillion plan for nuclear modernization over the next 30 years is very similar to Obama’s plan.

How does this conversion take place? Nuclear weapons are, properly, frightening to behold. They cross a line—of destruction, chaos, and inhumanity—that very few presidents or politicians want to get near. (The fact that Trump has spoken of them, at times so casually and at other times so aggressively, is one of the biggest reasons why his presidency is so frightening.) Nobody wants to take chances. If an expert or a general argues that some weapon or another enhances deterrence even a little bit, it’s hard for a president, or secretary of defense, to refuse it. Other experts may argue the contrary, but what if they’re wrong? If the Russians seem to believe the nonsense they’re spouting, if they do believe they can lob a mininuke at Europe and make us back off, is it enough simply to wave them away and insist that we have all the firepower we need to deter them? What if it isn’t?

Another problem is that there hasn’t been a public debate on nuclear weapons for the past 35 years. The experts Mattis consulted are members of the community that has pondered nuclear weapons for all this time, even after most of us moved on to other things. To varying degrees, they have accepted the logical strands of nuclear strategy as first principles—when, in fact, they are built on flimsy premises. There once was a debate in which the premises were challenged, then rejected or modified or embraced but with eyes open. Not enough eyes are staring at the nuclear abyss with an analytic squint. If we’re now witnessing the second coming of the nuclear age, it’s time for more people to start looking, studying, and talking back.

Correction, Feb. 7, 2018: The original version of the article cited numbers of American nuclear weapons as of more than a year ago. There are currently 400 land-based ICBMS, not 440; 280 submarine-launched ballistic missiles, not 288; and 60 nuclear-capable bombers, not 113. These submarines and bombers are capable of holding 2,768 nuclear bombs, not 2,070.