What Drives Nuclear Proliferation?

What drives governments to develop nuclear weapons? For those concerned in slowing, halting, or reversing the spread of nuclear weapons, the answer to this questions matter immensely. It helps identify what policies are most likely to convince states to abandon the quest for nuclear weapons, to not pursue them in the first place, or even to give up their existing stockpiles. But international-studies analysis provides a host of different, and sometimes contradictory answers. The same is true for the growing body of statistical literature on nuclear weapons. As Mark Bell argues in his International Studies Quarterly article, it even “offers many more distinctive explanations for proliferation than there are cases of proliferation in the historical record.” Bell conducts a series tests using “extreme bounds analysis, cross-validation, and random forests to evaluate 31 variables identified as significant determinants of proliferation.” He finds that, “While some variables—particularly, the pursuit and possession of other weapons of mass destruction, receipt of sensitive nuclear assistance, and some measures of threat—perform better than others, the overall results should give us pause. The majority of variables identified as significant determinants of proliferation fail to provide robust explanations for existing patterns of proliferation. They also offer little predictive ability beyond what we can achieve with an extremely simple model. The quantitative literature on proliferation has, for now, produced more tentative findings than scholars typically understand.”

This is, to put it mildly, an important set of conclusions. Thus, we asked a number of scholars of nuclear-weapons policy and proliferation for their reactions to Bell’s research note. They do so in this International Studies Quarterly Online symposium. Philipp C. Bleek finds Bell’s analysis mostly persuasive, and writes that “the fact that so few variables perform well is both surprising and depressing.” Rupal N. Mehta generally agrees, and stresses that Bell’s analysis “suggests that academic training should more heavily incorporate a more intuitive and complete understanding of the value - and limitations- of statistical analysis.” On Tuesday, Matthew Fuhrmann and Todd S. Sechser separately weigh in. Fuhrmann engages in a replication analysis of Bell’s note. He “extended his extreme bounds analysis in two ways, both of which suggest that some variables may be better at explaining proliferation than his analysis implies.” Sechser discusses some limitations of Bell’s analysis, and concludes that “His research note is not an indictment of quantitative methods; it is an endorsement of them.” On Wednesday, Etel Solingen and Joshua Malnight evaluate the substantive, methodological, and academic-policy gap issues raised by Bell’s research note, and Bell responds to his interlocutors.

As always, each piece will also be accessible from the callout as we publish them.

Discuss this Article
There are currently no comments, be the first to post one.
Start the Discussion...
Only registered users may post comments.
ISQ On Twitter