I thank Crispin Rovere for his thoughtful response to my original post on US declaratory nuclear policy. But I must take issue with the guts of his argument in relation to the sole purpose issue. In particular, I want to touch upon his two specific cases: Russian conventional aggression against the Baltic states; and, some form of US–China conflict close to the Chinese mainland.

Let’s start with the Baltic. Crispin says the US simply won’t use nuclear weapons to defend the Baltic states against Russian aggression, even though those states are members of NATO. After all, he notes, the Baltic territories were earlier occupied by the USSR for fifty years and no-one contemplated fighting a nuclear war over them. There’s a logic to that position. But two points need to be made. First, President Obama has specifically stated that there’s no difference between new NATO allies and old ones. And NATO’s 2012 Deterrence and Defence Posture Review states clearly that ‘the supreme guarantee of the security of the Allies is provided by the strategic nuclear forces of the Alliance, particularly those of the United States’. If that assurance turns out not to be true, it’ll raise questions about the credibility of US extended nuclear assurances elsewhere, and not just in the rest of NATO.

Second, even if the US is pretty unlikely to use nukes to defend Lithuania (for example), does it really want to stand up in advance and promise Putin that, no matter what level of conventional force he uses to seize Lithuania, the US won’t cross the threshold? That’s what a sole purpose declaration does. Such a declaration says unequivocally that nuclear weapons exist only to deter other countries’ use of nuclear weapons. Wouldn’t we want a potentially pushy Putin to have even a small doubt in his mind about whether the US extended nuclear assurance to Lithuania might actually turn out to be real? I don’t see how a sole-purpose declaration helps.

What then of the second case, the hypothetical US–China conflict? Here Crispin says the US has a clear incentive to use nuclear weapons first against China, because of its nuclear advantage, especially once American ballistic missile defences are factored in. The implication is that a sole purpose declaration would make China less anxious, and so less inclined to pursue an expansion of its nuclear capabilities in ways that would make Asian neighbours nervous. Mate, China already is expanding its arsenal. But it’s not the growth of China’s nuclear arsenal that most worries neighbours; it’s the expansion of its conventional power—which it’s not shy of using.

Exactly the same argument applies here as in the Russian case; why would the US want to tell China that regardless of how it uses its conventional power, and regardless of the costs it imposes on US forces or those of its allies, such actions won’t carry any risk of nuclear escalation? How, exactly, would that make US allies and partners in Asia feel safer?

At a time when hard power is increasingly pushing its way back to centre stage, it doesn’t make sense for the US to declare that its nuclear arsenal is reserved solely for the deterrence of nuclear coercion. It’s like owning a car, but declaring you’ll only use it when other people use theirs.

Rod Lyon is a fellow at ASPI and executive editor of The Strategist.