Do nuclear weapons provide countries with advantages in international bargaining? If so, under what conditions? Scholars and policymakers have debated these questions for decades. Remarkably, nearly seventy years into the nuclear age, we still lack consensus about the coercive value of nuclear weapons. Our goal in this chapter is to add greater clarity to the nuclear blackmail debate.

We start by describing the basic complexion of coercion in international politics. Next, we develop a generalized framework of coercion that yields several conclusions about the conditions that favor coercive success, and then ask whether nuclear weapons help bring about – or bolster – these conditions. In the end, we conclude that they do not. Nuclear weapons may be useful for deterrence and self-defense, but they are not useful for coercion.

Coercion: An Introduction

In its broadest sense, coercion involves using threats – either explicit or implied – to motivate someone to act. At its core, then, coercion is about behavior modification. A coercer aims to persuade a victim to alter its behavior by taking actions that serve the coercer's interests. The coercer's objective is to change the target's behavior without actually having to execute the threat. Executing threats can be costly not only for the target, but also for the challenger. Coercers therefore would prefer that their words be sufficient. As Clausewitz wrote: “The aggressor is always peace-loving… he would prefer to take over our country unopposed.” Coercion, then, is at its most effective when no punishment is ever imposed.

In our lives, as in international politics, coercion is inextricably woven into the daily rhythms of human interaction. A parent threatening to withhold a toy from a misbehaving child, a boss warning an insubordinate employee, or a homeowner threatening to sue a builder for breach of contract are all engaging in coercion. In each case, the coercer holds out the possibility of some unpleasant consequence unless the target behaves to the coercer's liking. Consider the following (hypothetical) scenarios: