美國在台海嚇阻穩定性的評估：以陳水扁時期

Abstract

本研究評估有關美國嚇阻台海軍事衝突之穩定性，尤其以陳水扁前總統執政時期(2000年至2008年)為關注之焦點。本研究發展一個簡單的分析框架以檢測嚇阻模式 (deterrence model) 以及螺旋模式 (spiral model) 兩個戰爭爆發的模式在台海的相對效應。理解哪一種模式在台海具有主導性可以幫助我們了解嚇阻不穩定的主因，與美國應該採取的嚇阻策略。如果嚇阻模式具有主導性，中國對台灣的修正主義主要是導因於可能的獲利與實力的最大化。嚇阻最有可能失敗的原因是因為美國嚇阻的可信度不足。在此種情形下，美國的嚇阻策略應該強調提升其懲罰中國對台灣動武的可信度。如果螺旋模式具有主導性，中國對台武力威脅主要以消除威脅與避免損失為動機。在此種情形下中國認為沒有作為的代價比戰爭的代價還高。美國的嚇阻策略必須強調對北京的再保證，以表明美國對台灣的安全承諾並無協助台灣變更現狀之意。 本論文先運用此一框架分析1954至1955年、1958年以及1995至1996年之三大台海危機，然後用以分析陳水扁前總統執政時期之嚇阻動態。本論文之分析顯示在三大台海危機以及陳水扁前總統執政時期之八年期間，螺旋模式的效應強於嚇阻模式。雖然上述之台海危機常被認為是成功嚇阻的範例，但是此三大台海危機更應該被歸類為「再保證的失敗」或者「雙重嚇阻的失敗」。在陳水扁前總統執政時期，中國軍事威脅也主要是以避免損失為動機。然而，在此一時期得以避免軍事危機的部分原因是因為美國的政策強調雙重嚇阻。本研究之結論為，在可預見的未來，螺旋模式仍將具有主導性。近期對於嚇阻穩定性最大的威脅來自於中國相信美國將支持台灣獨立，而非美國對於台灣的安全承諾之可信度不夠。只要美國繼續實施雙重嚇阻，嚇阻保持穩定的可能性相當大。長期而言，當中國的軍事實力持續增強，嚇阻模式的主導性可能亦將隨之增加。

This study assesses the stability of American deterrence of a military conflict in the Taiwan Strait, focusing in particular on the period of Chen Shui-bian’s presidency in Taiwan from 2000 to 2008. A simple analytical framework is developed to test the interaction in the Taiwan Strait of two models of the outbreak of war: the deterrence model and spiral model. Understanding which of these models is dominant in the Taiwan Strait can help to identify the main source of deterrence instability and the appropriate US deterrence policy. If the deterrence model is dominant, China’s revisionism toward Taiwan is primarily motivated by gains and power-maximizing. Deterrence is most likely to fail because the credibility of US deterrence is insufficient. In this case, US deterrence policy should emphasize increasing the credibility of Washington’s threat to punish PRC aggression against Taiwan. If the spiral model is dominant, Beijing’s threat to use force against Taiwan is primarily motivated by threats and losses. In this case, deterrence is most likely to fail because China believes the costs of inaction are greater than the costs of war. In this case, American deterrence policy must emphasize reassurances to China that its security commitment to Taiwan is not intended to help Taipei change the status quo.

The framework is first applied to examine the three major Taiwan Strait crises in 1954-55, 1958, and 1995-96 and then to analyze deterrence dynamics during the Chen Shui-bian period. During the three Taiwan Strait crises as well as during the eight years of Chen Shui-bian’s administration, spiral model dynamics were stronger than deterrence model dynamics. Though they are sometimes considered examples of successful deterrence, the three Taiwan Strait crises would be better classified as “reassurance failures” or “dual deterrence failures.” During the Chen Shui-bian period, China’s military threat was also primarily motivated by threats and losses. However, during this period, a military crisis was avoided in part because US policy emphasized dual deterrence. This study concludes that in the near-term, spiral model dynamics will remain dominant. The greatest threat to deterrence stability in the near future is the perception in China that the US supports Taiwan independence, not that the US security commitment to Taiwan lacks credibility. If Washington continues to practice dual deterrence, a militarized crisis in the near-term is unlikely. In the long-term, deterrence model dynamics may strengthen as China’s military power grows.