Crispin Rovere and Kalman Robertson have responded to my earlier comments on their ‘low-yield nuclear weapon treaty’ proposal, but I’m not convinced by their argument. They claim that low-yield weapons ‘are a unique class of non-strategic nuclear weapon (NSNW) that carry special risks which greatly outweigh any marginal benefit’ and go on to suggest that ‘prohibiting the development and retention of nuclear weapons with yields below 5 kilotons therefore strengthens security overall.’ But from whose strategic vantage point do they make these claims? Viewed from Washington, Paris or London, you could argue that a perceived absence of an overwhelming threat, together with clamorous calls for nuclear abolition, makes consideration of a low-yield NSNW ban a logical step towards a world of ‘global zero’. The wisdom of rushing to achieve that ultimate end-state can be strongly challenged: advocates of abolition consistently don’t address the hard reality that states have nuclear weapons for valid strategic purposes—most prominently deterrence—and eliminating low-yield NSNWs increases risks for some states that deterrence will fail.

Nuclear deterrence isn’t about bluffing. A credible nuclear deterrent demands both the ability and willingness to undertake war fighting if vital national interests are challenged. It also demands a clear requirement to communicate such a posture to any adversary. This was the basis for the success of nuclear deterrence throughout the Cold War, and remains so today. During the Cold War, NATO willingness to threaten a limited use of tactical nuclear weapons provided a nuclear ‘firebreak’ in which both sides could choose to step back from the brink. Intra-war deterrence was not a myth. However, the alternative of immediately ‘nuking Moscow’ in response to conventional attacks along the inner-German border would have rushed both sides needlessly into a series of strategic exchanges. This could have only resulted in mutually assured destruction (MAD).

The question is ‘why escalate’? It’s much more logical to practise intra-war deterrence and bring hostilities to an early end, rather than mindlessly rush to embrace mutual destruction. Rovere and Robertson’s argument forces states down the path to MAD. I see utility in low-yield NSNWs because they offer nuclear-armed states’ national command authorities the flexibility of a more credible nuclear deterrent (backed up by an ability to fight nuclear war at a level considerably lower than mutual annihilation) and time to bring hostilities to an end. Removing the ‘nuclear firebreak’ provided by low-yield NSNWs doesn’t increase security, but instead increases risk because it leads to an all-or-nothing choice for national leaderships that undermines the credibility of deterrence.

This is a strategic dilemma relevant not only to new nuclear weapons states like Pakistan and India, but also traditional P5 states such as Russia, which continues to rely on NSNWs to deter conventionally superior potential opponents, including a rising China. China’s arsenal and posture remain relatively opaque. But like Russia, China can’t merely rely on strategic nuclear forces. China’s strategic outlook isn’t only shaped by the maritime space of East Asia, but also by the potential for Sino-Indian tensions to lead to miscalculation and longer-term risks associated with resource competition in Central Asia and Far-Eastern Siberia. Planning for future PLA ‘Air-Land’ operations means low-yield NSNWs retain utility as a deterrent to the nuclear forces not only of the United States, but also potentially, India, Russia and other nuclear weapon states that could emerge in East Asia.

Rovere and Robertson sidestep the issue of nuclear-armed states employing selectable yield warheads to cheat on any agreement by asking ‘why they will cheat’? This rhetorical question ignores the practical issue that must be at the foundation of any such agreement. A minimum-yield threshold treaty must be fully verified and monitored through intrusive inspections for all signatories otherwise it’s of little use. The fact is that unless signatories with selectable-yield warheads are prepared to eliminate those weapons and agree to intrusive monitoring to ensure they don’t produce low-yield NSNWs, states can cheat. Rod Lyon’s comments on the US B-61 bomb are well put. Relying on trust and legal mechanisms is an insufficient basis for maintaining deterrence stability.

Finally, Rovere and Robertson challenge my scepticism that such a treaty can be made to ‘actually happen’. Their response notes that the key challenge states which I highlighted—North Korea, Israel, India, Pakistan and potentially Iran—are either outside or not in compliance with the 1967 Nuclear Non-Proliferation Treaty (NPT). Once again they sidestep the challenges of negotiating a minimum-yield threshold treaty with an entirely separate point about the desirability of the NPT. This point isn’t quite clear to me. If it’s okay for these states to maintain low-yield NSNWs, but desirable for the P5 states to eliminate theirs, that doesn’t eliminate low-yield NSNW. Rather, it highlights a more fundamental challenge that the nuclear abolitionists are yet to satisfactorily address—how to make nuclear disarmament actually happen!

Malcolm Davis is assistant professor in International Relations and post-doctoral research fellow in China-Western Relations at Bond University. Image courtesy of Flickr user Nicolas Raymond.