This is one of a series of weekly review papers I had to write during my “Introduction to International Relations” course. It discusses Scott Sagan’s viewpoint on nuclear proliferation as espoused in The Spread of Nuclear Weapons.

In The Spread of Nuclear Weapons: A Debate Renewed, Scott Sagan and Kenneth Waltz debate the effects of nuclear weapons proliferating among new state actors: whether these weapons make states safer and whether proliferation presents intolerable risks of terrorists acquiring nuclear weapons. Waltz makes a marvelous argument that nuclear weapons increase safety in a world of sufficiently rational actors, but fails to persuade me that our world is sufficiently rational; his argument that terrorists won’t use nuclear weapons is sadly perplexing. Sagan refutes Waltz via organization theory, arguing convincingly that nuclear weapons make the world a more dangerous place because people are imperfect and it takes only one mistake in the wrong place to spark a war. It is for these reasons that I believe Sagan’s conclusion is the correct one: the spread of nuclear weapons creates a more dangerous world, and we should work to prevent proliferation.

Both Sagan and Waltz agree that there are three necessary behaviors if nuclear weapons are to make the world safer: states must not engage in war to prevent other states from acquiring nuclear weapons, all states with nuclear weapons must be deterred from fighting each other by the certainty of massive retaliation (and so nuclear forces must be invulnerable), and nuclear weapons must be protected from accidental or unauthorized use. The beginning portion of Sagan’s argument is devoted to debunking rational deterrence theory’s applicability to our world; he instead argues that our models of behavior should be based on organization theory. He backs this by demonstrating that professional militaries tend to make choices that cause “deterrence failure” and war, then arguing that future nuclear states will not have enough civilian control over the military to counteract these tendencies.

Sagan gives three reasons that the military is likely to cause nuclear war even when civilian leadership doesn’t. First, military officers advocate preventive war at a much higher frequency than civilian leaders do. Through both self-selection and training, they view war as inevitable more often than civilians do; they focus on military victory at the expense of political, moral, and monetary costs; and they plan incrementally. For all these reasons, and citing evidence from the Cold War, Sagan claims that military leadership will preemptively strike at nuclear at times when civilian leaders would recognize the dangers and avoid conflict.

Second, Sagan argues that professional militaries are incapable of developing reliable second-strike nuclear weapons. First, states are notorious for overestimating the number of nuclear weapons required for deterrence. Second, the military is organizationally resistant to creating new and expensive methods of protecting their nuclear weapons for a number of reasons, including because doing so removes money from other areas and requires acknowledging that they may not be the first to attack. Again Sagan points to the Cold War, to China and the Soviet Union in addition to the United States, showing how it took each country decades to develop second-strike capabilities and how the military resisted each new method.

Third, Sagan argues that militaries as organizations will eventually have nuclear accidents. He claims that organization theory guarantees accidents will eventually occur in any complex system, and that bureaucracy’s layers and blame-passing ensures that organizations often don’t learn from accidents. Sagan also provides evidence for this view from historical US incidents.

Sagan’s arguments are rooted in sound organizational theory (itself rooted in individual self-interest) and explain why nuclear weapons are made dangerous by the systems that govern them; Waltz’s claims seem weak and illogical in comparison and don’t have the historical accuracy that Sagan’s do. Because there is no convincing theory disputing the argument outlined above, I conclude that the spread of nuclear weapons is dangerous and should be discouraged as much as possible.