Nuclear bargaining : using carrots and sticks in nuclear counter-proliferation

Abstract

This dissertation explores how states can use positive inducements and negative sanctions to successfully bargain with nuclear proliferators and prevent the spread of nuclear weapons. It seeks to answer the following question: How effective are offers of inducements and threats of sanctions - i.e., 'sticks' and 'carrots' - in efforts to prevent or to roll back the proliferation of nuclear weapons? I pay particular attention to the use of positive inducements, asking: How effective and efficient is the use of inducements relative to negative sanctions? Under what circumstances is each most likely to be useful? I propose an issue-linkage theory with which the use of negative sanctions and positive inducements might be understood. In the theory, negative sanctions and positive inducements are conceptualized as bargaining proposals that link punishments or rewards to particular demands made to the target state. Negative sanctions and positive inducements are effective when they enhance the power and interests of domestic political factions in the target state that support compliance with the sender's nuclear demands. I argue that positive inducements are typically a more effective tool of foreign policy than negative sanctions. Also, the difference in the effectiveness between the two is more pronounced when dealing with adversaries than it is with allies. With allies, both sanctions and inducements can be effective, but inducements are more likely to secure long-term cooperation. With adversaries, negative sanctions are not only much less effective than inducements, but run the risk of triggering escalation. The theory is tested against three counter-proliferation cases: North Korea, South Korea, and Libya. I find that negative sanctions were successful only in the South Korean case, while threats and sanctions triggered escalation in both cases involving US adversaries. On the other hand, positive inducements were used successfully in all three cases. A detailed reading of the three cases lends support to the theory, and suggests that a broader test is warranted.