The Logic of Latent Nuclear Deterrence

Nuclear deterrence is central to international relations theory and practice. Most people assume that countries must possess nuclear weapons in order to reap deterrence benefits from their nuclear programs. This article shows, however, that latent nuclear powers – nonnuclear states that possess the capacity to make weapons – can deter aggression, despite their lack of assembled warheads. Latent nuclear deterrence works because states that possess the technology needed to produce bombs can threaten to initiate or accelerate nuclear weapons programs if they are attacked. A fixed effects regression analysis that includes 170 countries from 1946 to 2010, using data compiled by the author on the global spread of sensitive nuclear technology, provides evidence consistent with three of the theory’s testable predictions. First, switching from non-latency to latency reduces the probability of being targeted in a violent military dispute in a given year by 3.32 percentage points. Second, having nuclear latency does not deter less serious, nonviolent disputes. Third, the development of non-sensitive nuclear technology that does not provide states with latent nuclear capacity is not associated with a lower likelihood of being attacked. A qualitative analysis of Iran’s nuclear activities from 2002 to 2015 illustrate these statistical findings. This evidence has lessons for the debate about nuclear disarmament: most scholars and policymakers are skeptical that the prospect of nuclear rearmament in a disarmed world could deter serious international disputes, but the case for latent nuclear deterrence is stronger than critics would lead us to believe. Much ink has been spilled on the subject of nuclear deterrence.1 Most scholarship assumes, at least implicitly, that nuclear technology has little value for deterrence unless countries can immediately retaliate with nuclear weapons. In this view, a nuclear arsenal provides an effective deterrent because it can swiftly destroy an opponent’s dearest possessions, such as cities. Without providing the capacity to launch a nuclear strike, nuclear technology cannot deter military conflict and may actually invite it by providing adversaries with incentives to launch preventive wars. The conventional view of nuclear deterrence may understate the strategic benefits of nuclear technology. Nuclear weapons are indeed valuable deterrents, but the mere capacity to produce a nuclear bomb could limit military aggression as well. Countries with nuclear latency – nonnuclear states that have developed the capacity to make weapons – might be able to deter attacks even though they are incapable of immediate nuclear retaliation. How is nuclear deterrence without bombs possible? States with dual-use nuclear technology possess the potential foundation for a nuclear weapons program. When states build sensitive nuclear plants they send a warning to others: “we have come this far and we can go further, if necessary.” The development of nuclear latency, then, resembles a shot across the bow. If a latent nuclear power perceives that its interests are threatened, it might begin a nuclear weapons program or accelerate an existing one. The prospect of fomenting nuclear proliferation may induce caution among the latent state’s potential adversaries, leading to less military conflict. I call this means of influence latent nuclear deterrence. Some practitioners appear to embrace the notion that nuclear latency bolsters deterrence. In 2011, Shigeru Ishiba, a former Japanese defense minister, put it this way: “I don’t think Japan needs to possess nuclear weapons, but it’s important to maintain our commercial reactors because it would allow us to produce a nuclear warhead in a short amount of time.” Having nuclear latency, he added, serves as “a tacit nuclear deterrent.”2 Pakistani officials, particularly those in the Army, also hold an “enduring belief” that the country’s latent nuclear capacity deterred military conflict in the 1980s – before it assembled a warhead or carried out a nuclear test.3 Does latent nuclear deterrence work? If so, under what conditions? We know surprisingly little about the efficacy of deterrence without bombs. This is unfortunate given the practical significance of this issue. More than 30 states – three times as many as built nuclear weapons – have developed sensitive dual-use nuclear technology over the last 70 years. There are 1See, for example, Brodie (1959), Kahn (1960), Schelling (1966), Miller (1984), Jervis (1989), Powell (1990), Gartzke & Jo (2009), Sagan & Waltz (2012), Narang (2009, 2014), and Bell & Miller (2015). 2Quoted in Dawson (2011). See also Hoey (2016, 485). 3Fair (2014, 223).