Julian Schofield and Micah Zenko
The normative wisdom is that states should not share nuclear weapons
technology, and consequently do so rarely. In the 1968/1995 Non-Proliferation Treaty,
nuclear weapon states promise not to assist non-nuclear weapon states to obtain nuclear
weapons, who in turn promise not to receive nuclear weapons.
The text of the NPT gives
the impression that there is little trade in nuclear weapons, both normatively and actually.
However, nuclear weapons have been shared and our goal is to explain the conditions
under which they are shared.
Those that argue that deterrence is easy to achieve explain
nuclear sharing as a function of whether a state seeks to have a target state possessing
effective deterrence or not, and this usually depends on the state’s interest in preserving
the systemic status quo. Those that argue that nuclear deterrence is difficult to achieve
see nuclear weapons as akin to conventional weapons, and therefore see few incentives in
sharing such powerful technology.
We argue that nuclear sharing, because nuclear weapons are so exceptionally
powerful instruments, is never conducted with allies, and only with geographically
distant and dependent proxies opposing a powerful and common adversary. In effect, a
theory incorporating the underlying logic of the balance of power environment does a
superior job of explaining when states share nuclear assets than any theory grounded in
the special qualities of nuclear weapons themselves. In all six of our cases of nuclear
sharing, the principal incentive to transfer nuclear resources was the setting-up of a
nuclear proxy to distract the attention of another power.
Conventional Wisdom
The conventional wisdom on the sharing of nuclear weapons is captured by
Egyptian President Nasser’s observation that “…nations do not generally sell atomic
bombs. At least not if they can help it.”
One version argues that before states come to
adhere to the nonproliferation regime, they all make one critical mistake and help another
state obtain nuclear weapons.
While true in the case of France, the U.S. and Soviets
were initially very reluctant to share nuclear weapons technology, and tended to do so
only after the initial temptation. Nor does this late sharing necessarily require that the
recipient state demonstrate national viability, as in the case of Israel’s assistance to South
Africa or China’s assistance to Pakistan.
The first case of conscious nuclear sharing
involved the transfer of uranium oxide from wartime Germany to Japan (via U-boat). The
Japanese believed, correctly, that the Germans would not tolerate nuclear competition,
and so hid their real purpose until very late and after protest.
This first opportunity is a
poor case for any test because the Germans were only dimly aware of the long-term
political implications of such a transfer, which were done in the last days of the Nazi
regime. However, it does fit with our general proposition here that states trade nuclear
secrets in order to distract common adversaries, in this case the Allies operating in the
Pacific theatre.
The dynamic of nuclear weapons proliferation seems logically linked to the
purposes intended by the nuclear proliferator.
Thus the two models of nuclear sharing
are actually based on perceptions of nuclear weapons utility.
We propose two models:
the easy deterrence and elusive deterrence sharing models.
In the easy deterrence sharing model nuclear weapons are seen as robust
According to Robert Jervis, nuclear weapons will reduce disputes
and war and make preserving the status quo relatively easy.
relatively easy to achieve.
Therefore a state’s willingness to share nuclear weapons is
2
Convention