防衛合作協定之形構：以四國印太戰略與歐盟共同防衛政策為例 - 政大學術集成

全文

(1)國立政治大學國際事務學院外交學系 Department of Diplomacy, College of International Affairs, National Chengchi University. 碩士論文 Master’s Thesis. 政治大. 立防衛合作協定之形構：. ‧ 國. 學. 以四國印太戰略與歐盟共同防衛政策為例. ‧. The Origins of Defense Cooperation Agreements:. Nat. io. sit. y. The Quadrilateral Indo-Pacific Strategy and. n. er. the European Common Defense Policy al v i n Ch engchi U. Student：Fang, Chi Advisor：Lien, Hong-Yi 中華民國 109 年 1 月 January 2020. DOI:10.6814/NCCU202000065.

(2) 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. DOI:10.6814/NCCU202000065.

(3) Acknowledgement. Acknowledgement. Six years. Twenty-four seasons have passed ever since my very first connection with the Department of Diplomacy. Saying that a long dream has finally come to an end is definitely nothing exaggerated. I fortunately received the most integrated education and experience in the last few years. Let me indulge myself with several lines to express my sincere gratitude.. 政治大. First and foremost, I owe a great deal to my parents' unconditional support for my pursuit. 立. of knowledge on the campus. Despite increasingly spare gathering with them, they have been. ‧ 國. 學. my motivation to go further on this career to honor their love.. ‧. I also thank all the faculty who taught me in the National Chengchi University, particularly. y. Nat. al. er. io. sit. those in our department. Without them, it would be hardly to imagine a place like this that. v. n. enables me to learn and enjoy. In particular, I would like to thank Professor Hong-yi Lien,. Ch. engchi. i n U. Professor Ping-kuei Chen, and Professor Teng-chi Chang. Their guidance is indispensable for me to complete this thesis. Lien's course on strategic studies and Chen's courses on IR and FP theories definitely enlightened me that I decided to challenge myself with a rather theoretical theme. Of course, having warm companion of the colleagues is a true fortune that I would cherish and recall in the future. Both of those in the research room and the drinking gang surely makes me go much further than I used to anticipate. I also thank my instructors and colleagues in the i. DOI:10.6814/NCCU202000065.

(4) The Origins of DCA. Department of Political Science at UC Berkeley and Institute of Political Science of Academia Sinica. Their brilliant thinking and careful instruction let me know what a good research should look like and what can be done by myself. Most importantly, I finally know what I might be interested in after this long period of time passed. This thesis is definitely imperfect. However, it surely concludes and includes my experience and thought on theories and policies so far. As committed to the advance of. 政治大. knowledge, I remain a loyal skeptic against "isms" or any absolute belief in a single approach,. 立. be it formal models or qualitative studies. This spirit, without doubts, is the greatest treasure I. ‧ 國. 學. have found after this long journey.. ‧. January 2020, Taipei. n. al. er. io. sit. y. Nat. Chi Fang. Ch. engchi. i n U. v. ii. DOI:10.6814/NCCU202000065.

(5) Abstract. Abstract. In the past decades, we are witnessing a rapid growth of defense cooperation agreements (DCAs) in contrast to the decreasing amount of newly formed alliances. I examine the formation of DCAs in the lens of contextual change and costly signals. In terms of contextual change, I construct the environment of the strategic interaction by specifying both structural. 政治大. and institutional context. Within these preceding contexts, I contend that DCAs are the product. 立. of two efforts by states. First, they are states’ adjustments toward unexpected contextual change,. ‧ 國. 學. e.g.: global and regional structural change, institutional bargaining, or the combination of both.. ‧. Second, they are states commitment to the existing international institutions as well as broader. Nat. io. sit. y. regional and international order. In both cases, states attempt to reduce the uncertainty in the. er. international society. Accordingly, I conduct two case studies: the Quadrilateral Security. al. n. v i n Strategy in the Indo-Pacific and theCEuropean Defense Policy. By conducting both U h e n gCommon i h c in-depth case studies and controlled comparisons, I conclude that given the rapid change of security challenges, states opt to strengthen existing regimes rather than engaging outright confrontations. Consolidation as such is achieved by signing DCAs to bolster the security and defense networks.. Keywords: DCA, Costly Signals, Commitment, Cooperation, Indo-Pacific, European Defense. iii. DOI:10.6814/NCCU202000065.


(7) Contents. Contents. Acknowledgement ...................................................................................................................... i Abstract .................................................................................................................................... iii Contents ..................................................................................................................................... v List of Tables and Figures........................................................................................................vii Chapter I:. Introduction ........................................................................................................ 1. Research Goal ..................................................................................................................... 2. 政治大 Locating DCAs in typologies .................................................................................... 10 立 Structural Pressure..................................................................................................... 15. Tracing DCAs in Literature: Structure, Institution, and Strategic Bargaining ................... 5. ‧ 國. 學. Institutional Dynamic ................................................................................................ 21 Rational Choice ......................................................................................................... 26. ‧. Scopes and Limitations ..................................................................................................... 31. y. A Contextual Approach of DCAs .................................................................... 37. sit. Chapter II:. Nat. Conclusion ........................................................................................................................ 33. er. io. Pathway of Bargaining: Structural and Institutional Context........................................... 38. al. v i n Ch Institutional Pathway ................................................................................................. 42 U i e h n g c Bargaining............................................ 46 Arming through Aligning: Specifying Strategic n. Structural Constraints ................................................................................................ 39. From Exogenous Shock to Institutional Evolution: Credibility through Consolidation ............................................................................................................ 48 DCAs as Costly Signals: Making Credible Commitment ......................................... 50 Layered Contexts: Restructuring Strategic Interactions ................................................... 54 Research Methods and Design ......................................................................................... 56 In-Depth Case Studies ............................................................................................... 57 Cross-case Analyses .................................................................................................. 62 Observable Implications ............................................................................................ 65 Case Selection ........................................................................................................... 66 Chapter III:. Case Study – Indo Pacific ................................................................................ 69. v. DOI:10.6814/NCCU202000065.

(8) The Origins of DCA. The Quadrilateral Indo-Pacific Strategy ........................................................................... 69 Unsettling the Hub .................................................................................................... 70 Uncertainty within Strategic Interests and Options................................................... 78 Network Defense Cooperation: From Bilateralism to Minilateralism ...................... 94 Conclusion ...................................................................................................................... 103 Chapter IV:. Case Study – EU CSDP ................................................................................. 107. The EU Common Security and Defense Policy ............................................................. 107 The Evolution of CSDP........................................................................................... 108 Challenge and Threat to CSDP: Competing Interests and Options ........................ 113 Defense Cooperation in the Era of EUGS ............................................................... 134 Conclusion ...................................................................................................................... 138. 政治大 The Origins of DCA: Structure or Institution? ............................................................... 141 立 Discussion and Implications ........................................................................................... 145 Conclusion ..................................................................................................... 141. 學. ‧ 國. Chapter V:. Global and Regional Politics ................................................................................... 146 Convergence and Diffusion ..................................................................................... 148. ‧. Costs and Commitment ........................................................................................... 150 Research and Policy Prospects ....................................................................................... 152. y. Nat. n. al. er. io. sit. Bibliography .......................................................................................................................... 157. Ch. engchi. i n U. v. vi. DOI:10.6814/NCCU202000065.

(9) List of Tables and Figures. List of Tables and Figures Tables Table II.1: Determinants of External Constraints .................................................................... 45 Table II.2: Tests of Process Tracing for Causal Inference....................................................... 59. Figures Figure I.1: Growth Trend of Alliance and DCA ........................................................................ 6. 政治大 Figure II.2: Causal Sequence of立 DCA Creation....................................................................... 55 Figure II.1: Contextual Specification of Strategic Interaction ................................................. 55. ‧ 國. 學. Figure III.1: Power Distribution and Growth in the Indo-Pacific, 2007-2018 ........................ 73 Figure III.2: Context of the Indo-Pacific Defense Cooperation ............................................ 105 Figure IV.1: Distribution of Power among Main EU Players, 2007-2018 ............................ 117. ‧. Figure IV.2: Military Expenditure and Military Expenditure as GDP Share in Europe, 2007-. y. Nat. 2018 ...................................................................................................................... 122. io. sit. Figure IV.3: Context of CSDP’s Defense Cooperation ......................................................... 139. n. al. er. Figure V.1: Causal Sequence of DCA Creation .................................................................... 145. Ch. engchi. i n U. v. vii. DOI:10.6814/NCCU202000065.


(11) Introduction. Chapter I:. Introduction. Ever since the inauguration of the Obama administration, the US foreign policy in AsiaPacific has made considerable progress. Despite the fluctuation under the Trump administration, the US role within the region has remained crucial. From the earlier “Rebalance to Asia,” to a more recent “Free and Open Indo-Pacific Strategy,” they all anchored on the. 政治大. long-standing hub-and-spoke alliance system.. 立. Beyond that, however, is a qualitative change. While the system still hinges on a US-. ‧ 國. 學. centered mechanism, it has also been decentered from her. Instead, we are seeing a trend of. ‧. growing autonomy, which is especially obvious in terms of defense cooperation. For instance,. Nat. io. sit. y. Japan and Australia, despite no formal alliance relationship, have strengthened their. er. cooperation in defense and security. Identically, India has also pursued a greater engagement. al. n. v i n in the West-Pacific. The efforts areC particularly in its strategic relationship with Japan. U h e n gclear i h c This so-called “spoke-and-spoke” approach reveals the complexity of regional cooperation. On the other edge of Eurasia, a similar tendency also emerged along with a rising discussion of closer security cooperation between the European members (the EU). Though the EU has stumbled for decades when military issues are involved, its recent progress should not be ignored. Since the earlier "European Security and Defense Policy (ESDP)”, which is later known as “Common Security and Defense Policy (CSDP),” the EU has been consistently. 1. DOI:10.6814/NCCU202000065.

(12) The Origins of DCA. calling for better security policy in response to diversified challenges around the globe. As a result, in the past five years a series of initiatives has been enforced in more concrete ways, including “European Global Strategy (EUGS)”, “Permanent Structured Cooperation (PESCO)”, “Coordinated Annual Review on Defence (CARD),” and “European Defence Fund.” The cases above are not exemplary. Instead, they are part of a bigger picture that there is, in fact, a tendency showing that defense cooperation is becoming even more common in. 政治大. recent years. While there is considerable literature discussing policy and geopolitical trends,. 立. the theoretical and academia counterparts are relative of shortage. Defense cooperation. ‧ 國. 學. particularly is receiving limited attention nowadays. Security and defense, conventionally. ‧. attributed as “high politics” deserve more inquiries for their rich leverage to explore the state’s. Nat. io. sit. y. behavior rationale. Furthermore, to go beyond ad-hoc discussion and oversimplified. n. al. er. assumption, this discussion should benefit from a more inclusive way of theory-building that. Ch. engchi. could relax the tension of policy relevance.. i n U. v. Research Goal The main goal of this thesis is to investigate the formation of defense cooperation agreements. Three threads of discussion are worth examining. Altogether, they crystallize the formation of defense cooperation agreements at different levels and aspects. First and foremost, the realist explanations provided that the formation of defense cooperation agreements roughly reflects the dynamic of international structure. From the 2. DOI:10.6814/NCCU202000065.

(13) Introduction. earlier Waltzian proposition to a more recent argument put by Mearsheimer, this thread has analyzed the security affairs through the lens of power and geopolitics. A sense of strategic contingency has featured in both theoretical and policy discussions in this line. Secondly, neoliberal institutionalists argued for a more introspective view of institution formation. While the realist accounts committed themselves to a more structural explanation, the institutionalists theorized states “demand” for international institutions, or more broadly, international. 政治大. cooperation. This line of explanation offered a specific framework of how institutions might. 立. be formed in the long run. Final but not least, the deterrence theories and alliance theories. ‧ 國. 學. delivered a more “relational” account of the interaction between states. Formal and informal. ‧. literature relocate their aims on “interaction” per se and mapped out states’ rational. In terms. Nat. sit er. io. threats or niche for interests.. y. of these theories, it is plausible to conjecture that defense cooperation is an instant response to. al. n. v i n Using this literature, some ofC the hpreliminary i U should be answered. Why states e n g c hquestions. sign defense cooperation agreements? What do they expect to respond, threats or opportunities? Specifically, in the presence of higher-level structures like alliances, why do they choose a certain form of defense cooperation rather than others? It is apparent that theories above provide considerable answers for why states cooperate. Yet so far there is limited discussion about defense cooperation agreements and why this specific form is chosen. Answering these questions should enable us to understand the formation of defense cooperation agreements,. 3. DOI:10.6814/NCCU202000065.

(14) The Origins of DCA. which is becoming pervasive in the realm of international security. Furthermore, the insights derived from the literature will help us establish a concrete context in which states act upon. This point would be particularly important given the turbulent international situation nowadays. This thesis is structured as follows. First, I will review the literature relating to defense cooperation agreement. Particularly, I should develop a feasible typology that should facilitate theory-building later on. Second, I will wrap up different schools of theories to look into the. 政治大. theoretical account for the establishment of cooperation. As for the scope, I mostly. 立. concentrated on the materialist explanations since an idealist one would require an almost. ‧ 國. 學. different ontology. Here, the most important task is to structure hypotheses accounting for the. ‧. formation of defense cooperation agreements. Next, I will layout the research design. In this. Nat. io. sit. y. thesis, I employed the framework of Alexander George’s “structured and focus comparison.”. er. To fortify the causal inference, I also use process tracing to test the hypotheses derived from. al. n. v i n C htwo cases. The first the literature. Finally, I will examine e n g c h i U one is the security and defense cooperation in the Indo-Pacific region which has gradually replace Asia-Pacific as a new space for strategic and geopolitical engagement in the past decade. The other is the European Common Defense Policy. Despite its long criticized stumbles, its enormous progress in the past few years should not be ignored. Furthermore, progress does not come from nowhere. Instead, it should be deemed as an outcome stemming from early experience.. 4. DOI:10.6814/NCCU202000065.

(15) Introduction. In sum, this thesis aims not only for theoretical synthesizes but also for a more coherent way to inquire about the development of international security. After all, to my own belief, a discussion regarding the international relations theory should not isolate itself from the dynamic development of the world, but rather, engage.. Tracing DCAs in Literature: Structure, Institution, and Strategic Bargaining. 政治大 to alliances. Nonetheless, it is立 rather unreasonable if we look into the trend in the past four Defense cooperation agreements (DCA) received fewer inquiries, especially in contrast. ‧ 國. 學. decades Using the data from Correlates of War, it can be soon discovered that there has been. ‧. much less newly formed alliance than DCAs (Kinne 2019). A quick comparison could be. sit. y. Nat. observed in Figure I.1. While the blue line represents DCAs and the red counterpart stands for. er. io. alliances, the horizontal axis marked the starting year of given observations. It is clear that,. al. n. v i n after the end of the Cold War, international to face a surge of DCAs. Compared C h society came U engchi to the almost dormant growth of alliances, DCAs continue to increase at a rapid rate. Merely in terms of the sample size, DCAs carry great potential for us to explore states’ security policy.1. 1. The Defense Cooperation Agreement Dataset (Kinne 2019) spans from 1980 through 2010, which covers years during the Cold War and the post-Cold War. Nonetheless, the data should be interpreted with caution. The face value of the dataset implies a great deal for analyzing security cooperation after 1980. Yet we should not easily conclude that DCA flourishes only after 1980. On the contrary, it is highly likely that DCA has already been popular immediately after the Cold War, no matter as a part of alliances or as a more informal exchange between states. 5. DOI:10.6814/NCCU202000065.

(16) The Origins of DCA. Figure I.1: Growth Trend of Alliance and DCA. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 6. DOI:10.6814/NCCU202000065.

(17) Introduction. Despite its pervasiveness, DCAs have no uniform definition. Besides, the data used above suffered from relatively short temporal span which ranges from 1980 to 2010, during which the paradigm or first case of DCAs would be hard to be located. However, there is some reference that could enable us to commence the inquiries. To my knowledge, there is only one general definition in academic society trying to locate DCAs. It roughly sketches DCAs by defining them as “formal bilateral agreements that. 政治大. establish institutional frameworks for routine defense cooperation. DCAs typically involve. 立. relatively symmetric, long-term commitments for both sides, with an emphasis on coordinating. ‧ 國. 學. core areas of defense policy and encouraging interpersonal contacts” (Kinne 2018). This. ‧. definition and its following studies effectively pointed out that DCAs are important tools to. Nat. io. sit. y. solve the information asymmetry and distribution problems, both of which are a prominent. er. hindrance against cooperation. Nonetheless, as I will point in the following paragraphs, the. al. n. v i n Cthe definition is too limited to probe into h edefense h i U between states despite its nicely n g ccooperation framed operationalization. Governmental agencies have also attempted to delineate the scope of cooperation. Despite the frequent usage of DCA in naming bilateral documents, security cooperation appears to be more common in general documents. For instance, the US Defense Security Assistance Agency (DSAA), which is now known as the Defense Security Cooperation Agency. 7. DOI:10.6814/NCCU202000065.

(18) The Origins of DCA. (DSCA), coined the term security cooperation. In the most recent manual, it was defined as follows (Defense Security Cooperation Agency 2019): All DoD interactions with foreign defense establishments to build defense relationships that promote specific U.S. security interests, develop allied and partner nation military and security capabilities for self-defense and multinational operations, and provide U.S. forces with peacetime and contingency access to allied and partner nations. This includes. 政治大. DoD-administered security assistance programs.. 立. Yet it should be noticed that the mandate is not exclusive to the Department of Defense. ‧ 國. 學. (DoD). In fact, the clause includes a broader reference to the Department of State (DoS) which. ‧. possesses general control but not necessarily actual executing. 2 More specifically, those. Nat. io. sit. y. administering “any program or interaction of DoD with the security establishment of a foreign. er. country to build capabilities, to provide access or to build relationships” are mostly referred as. al. n. v i n security cooperation. For example,Cthe provides a comprehensive U h emanual n g cjusth imentioned. categorization. Other than funding, arms sales, and peacekeeping operation (PKO), security cooperation puts more emphasis on personnel exchange, security education, and military-tomilitary or military-to-civilian contacts. The activities above are all policies to enhance engagements between the US military and its partners. Furthermore, they are integral parts to. To clarify, it is the official definition that distinguish the so-called “security assistance”, which refers to those solidify the military capability through multiple financial means. See Defense Institute of Security Cooperation Studies (2018); Defense Security Cooperation Agency (2019). 2. 8. DOI:10.6814/NCCU202000065.

(19) Introduction. support the national strategy for US that each of them receives comprehensive assessment (Office of the Under Secretary of Defense for Policy 2016). DCAs function is specifically illuminated by the following address given by the US ambassador to Slovakia, Bridget Brink, before the 15th anniversary of NATO (U.S. Embassy Bratislava 2019), “...a Defense Cooperation Agreement that, once completed, will offer a legal and practical mechanism for increasing cooperation across all facets of our defense. 政治大. relationship. It will also provide the legal framework for the United States to extend. 立. approximately $105 million dollars in European Deterrence Initiative funding for needed. ‧ 國. 學. infrastructure improvements at Malacky and Sliac Airbases…”. ‧. These types of agreements, as mentioned after the quotation above, are all bilateral in nature.. Nat. io. sit. y. Again, the agreements are aimed at upgrading defense capabilities. Put in another way, if the. n. al. er. national strategy is the bones, then defense cooperation plays the role of connecting meats and bones.. Ch. engchi. i n U. v. It is immediately clear that security cooperation includes a wide array of instruments and realms. While it provides a riveting starting point for us to bolster our understanding of interstate cooperation, it nonetheless posed a theoretical hindrance. That is, how should we locate it in the current literature? While there is much literature on alliance politics, the DCA counterparts are relatively absent. But this is not the end of the world. Literature regarding alliance politics, in fact, offers the very starting point from which we might locate DCAs.. 9. DOI:10.6814/NCCU202000065.

(20) The Origins of DCA. Accordingly, it would be vital to set a theoretical stage in order to determine where we should stand to elaborate the formation of DCAs.. Locating DCAs in typologies. Some might doubt the plausibility to juxtapose DCAs with alliances. Since the formation of DCAs has just caught attention, there are currently no consensual definition and. 治政 position for DCAs so far. Moreover, some have argued that 大DCAs are analytically different 立 ‧ 國. 學. from the alliance for they usually include no mutual defense or nonaggression obligation (Kinne 2018). My goal here is not to refute the conceptualization. Instead, I believe that. ‧. alliances and DCAs are highly compatible.. sit. y. Nat. io. n. al. er. The connection is cogent in that DCAs have been used as a tool to advance the national. i n U. v. interests at least since the end of the Cold War. Despite different nomenclature, the aims remain.. Ch. engchi. For instance, William Clements (1972). claimed,. Had we not become involved and, for more than two decades, supported and encouraged the efforts of allied and friendly countries to protect themselves against threats to their territorial integrity and internal security, the complexion of the globe might be dangerously different today, and the international climate far more hostile.. 10. DOI:10.6814/NCCU202000065.

(21) Introduction. Clearly, DCAs have been a major instrument of both security and foreign policy through coordination and collaboration with other countries. The problem, then, would be the way we use the word, alliance. While the word has been used in a quite uniform fashion in the recent literature, it is not always so. For instance, Stephen Walt in 1987 used alliance and alignment in a rather ambiguous way. It is clearly not the case in other places. For more recent literature, alignments could refer to extensive forms of groupings while the more formal ramifications would tend to be known as an alliance.. 立. 政治大. DCAs without doubts is a form of alignments since they would require states to hold. ‧ 國. 學. share interests to behave in a coordinated, or at least not conflictual fashion. After all, alignment. ‧. is a rather comprehensive concept capturing a pattern in which states cooperated in any form. Nat. io. sit. y. as long as they have common interests on certain issues (Wilkins 2012). In terms of DCAs,. er. states have to have some extent of shared views on security and defense issues before entering. al. n. v i n C hwhy DCAs are signed into such agreements. Part of the reason e n g c h i U in practice is that signing parties. are aware of the need to normalize military activities even during peacetime (Murphy 1991). It is clear that whether DCAs are alignments tells not much about this form of cooperation. The unclear position of DCAs arguably leads us to further instigate whether DCAs are a form of alliances. Thus, I start the discussion with the definition of the alliance. Alliance, as just described, has been endowed with multiple meanings. Stephen Walt, for example, defines alliance as “a formal or informal relationship of security cooperation between two or more. 11. DOI:10.6814/NCCU202000065.

(22) The Origins of DCA. sovereign states” (Walt 1987). This definition provides little insight for its almost infinite coverage of interstate security cooperation and hence not of much use here. In contrast, Glenn Snyder provided a more concise definition: “alliances are formal association of states for the use (or non-use) of military force, in specified circumstances, against states outside their own membership” (G. H. Snyder 1997). In this sense, the scope becomes usefully narrower for it clearly pointed out that formalization and presumed threats are the central features of alliances.. 政治大. Similarly, the Alliance Treaty Obligations and Provisions (ATOP), a dataset focusing. 立. on the military alliance, makes a stricter requirement. Unless parties in alliances are obligated. ‧ 國. 學. to cooperate in the face of military conflicts, even formal security cooperation that is legally. ‧. binding would not suffice to be “alliances” (Ashley Leeds et al. 2002). Clearly, several. Nat. io. sit. y. necessary conditions could be listed to identify alliances. First, they are formal documents. er. signed between independent states. Second, they required mutual cooperation during wars or. al. n. v i n C and conflicts. That says, sharing facilities do not automatically qualify despite h ejoint n gexercises chi U apparently close relationships. Finally, they tend to assume a certain state of crisis to be dealt with.3 Thus far, DCAs are hardly equivalent to alliances given the lack of obligation to mutual defense or threat-oriented guidelines. Even in terms of defense pacts, DCAs remained outside 3. The analytical context of ATOP should be taken into consideration. Ashley Leeds and her colleagues (2002) focus on the contingency amidst warfare, which is a significant departure from DCA. DCA is mostly applied in peacetime cooperation rather than warring scenarios. Such a difference, however, does not undermine the comparability between alliance and DCA. As I elaborated in the previous and following sections, both alliance and DCA are on the exactly same continuum of alignment, especially in the realm of security cooperation. Hence, even though the doubt of comparability is legitimate, it does not erode my arguments here. 12. DOI:10.6814/NCCU202000065.

(23) Introduction. of the typology. A defense pact is a type of alliance that required allies to provide wartime assistance. For instance, the Austro-Germany Alliance of 1879 concentrated mostly on the coordination in case of crises or specified circumstances.4 DCAs, on the other hand, do not comprise such demands. As we have seen above, DCAs are rather concerning training and exchanges. In other words, they are primarily measures of improvements in capabilities, strategic coordination, or logistics support. Viewing DCAs as alliances would certainly be. 政治大. misleading since they could be signed, logically and realistically, between non-allied states.5. 立. This is exactly the case of the 2015 Framework for the U.S.-India Defense Relationship, which. ‧ 國. 學. extends the content of the previous New New Framework for the India-U.S. Defense. ‧. Relationship and mentioned no obligation of mutual defense (United States Department of. io. sit. y. Nat. Defense 2015a).. er. Departures as such, nonetheless, do not mean that DCAs and alliances are mutually. al. n. v i n C h even supplementary exclusive. Rather, they are compatible, e n g c h i U to each other, in terms of their functions and aims. Taking a deeper look at the security treaty between the US and its allies,. treaties as such provided mostly a broad range of obligation in concise phrases and clauses. For. 4. The alliance bred itself out of the souring relationship within the Three Emperors League, an alliance between Germany, Russia, and Austro-Hungary Empires. The alliance was first established in 1873 and soon disbanded in 1878 given the earlier tension between Germany and Russia. Because of Russia’s involvement in holding back Germany’s deterrence against France and Germany’s counterefforts to displace Russia’s payoff in the Russo-Turkish War of 1877-1888. The Austro-German Alliance of 1879, hence, was mainly an effort to deter attacks from Russia as well as accommodating Austria. In this case, the alliance is mainly one of a defensive pact rather than wartime coalition. For detailed assessment, see Snyder (1997, 84–93). 5 However, it should be noted that while there are fewer newly formed alliances, there are indeed some other forms of formal relationships that might essentially equivalent to alliances. 13. DOI:10.6814/NCCU202000065.

(24) The Origins of DCA. example, Australia, New Zealand, United States Security Treaty (ANZUS) outlined the circumstances requiring consultation and mutual-aid. Yet the treaty does not offer any details for establishing frameworks or mechanisms that might facilitate defense cooperation between signing parties. In turn, it is exactly DCAs that beef up the cooperation between allies. Building on this, further observation could be made such that DCAs and the other forms of security cooperation establish an intense network that bolsters the existing security. 政治大. architecture. Some DCAs might be established given the existence of other institutional. 立. frameworks, while the rest turn out to be independent and parallel to their nested counterparts.. ‧ 國. 學. One of the most prominent cases for nested DCAs appears within North Atlantic Treaty. ‧. Organization (NATO). While NATO provides a set of foundational norms that support the. Nat. io. sit. y. operation of the US and its European allies, it nonetheless leaves the legal base of military. n. al. er. operations and the other general affairs to separate agreements between parties (The Federal Republic of Germany 1981).. Ch. engchi. i n U. v. Before concluding this section, one final critique should be proposed. The definition provided by the existing literature is too limited. By focusing only on formal documents that mandate military to military cooperation, existing operationalization actually overlooks a large part of defense cooperation. Take the US defense cooperation with other countries for example, they do not have to establish an agreement or treaty in order to cooperate. What they need, instead, is merely executive admission. Furthermore, defense cooperation does not have to be. 14. DOI:10.6814/NCCU202000065.

(25) Introduction. bilateral. They are sometimes, or usually, multilateral. An instant example would be the initiatives established by the ASEAN Defense Minister Meeting, which is multilateral in nature. In this sense, we should further question the motivation of formalizing defense cooperation instead of their occurrence. Thus far, we come to realize that DCAs are in fact distinct options that entail more concrete cooperation between states. It would then be appropriate to inquire about the source. 政治大. of these agreements. Why and how do they occur? Why are they even formalized? Despite the. 立. spare literature on DCAs, the literature on alliance politics might lend some leverage to reveal. ‧ 國. 學. the formation of DCAs. I would argue that this direction would be legitimate given the fact that. ‧. DCAs and alliances, in fact, share several concerns and features as discussed above. After all,. Nat. io. sit. y. they are both a type of alignments that aim to secure national interests. In the following sections,. n. al. er. an overview of the literature on the formation of DCAs is carried out through the lens of alliances.. Ch. engchi. i n U. v. Structural Pressure. First, DCAs, as the nomenclature suggests, focused on the security and defense cooperation between states. In the realist sense, it would be reasonable to view these acts as a way to achieve maximize their security. Furthermore, since DCAs also play the role of supplementing higher strategic goals, they to some extent reflect the strategic need of given. 15. DOI:10.6814/NCCU202000065.

(26) The Origins of DCA. states. That is, why do states align with other states through entering into these formal agreements at certain points and international environment? The very first approach commences from the neorealist explanations. Why states align or ally with each other has always been the central concern of international relations. One of the most prominent explanations comes from Kenneth Waltz. Based on his theories, we could instantly infer that weaker states cooperate with each other to balance against the stronger.. 政治大. Balancing in this sense is deterministically a natural outcome of structural pressure, which is. 立. determined by international distribution of power. Such phenomena, in fact, reflect states’. ‧ 國. 學. concerns for its position in the international structure (Waltz 1979). Accordingly, states have. ‧. two ways to enhance their odds of survival, external and internal balancing. That is, states could. Nat. io. sit. y. either arm themselves or ally with others. Later arguments furthered the theories by adding. er. variables such as offense-defense balance and distinguishability. For example, Christensen and. al. n. v i n C h by developing Snyder extend Waltz’s and Jervis’ argument e n g c h i U a theory of chain-gang and buckpassing (Christensen & Snyder, 1990). Still others, like James Morrows and Randall Schweller, explored how domestic factors could alter the ought-to-be rationality under certain international structures (Morrow 1993). However, these approaches are flawed for their presumed universality. It would be inconceivable to reduce the real world into a billiard table in which the same mechanism applied to all states. Instead, each state is bind by territoriality derived from its unique. 16. DOI:10.6814/NCCU202000065.

(27) Introduction. environment. This inherent problem was noted by Steven Walt. By disputing the applicability of Waltzian theory in the Middle East, he further proposed that states would balance or bandwagon given the perceived threats. He adds an additional constraint, states’ geographic position.6 Such a theory largely explained why some pattern of regional alignments has been emerging that they sometimes deviate from the expected behavior under global structure. Similar arguments were also put out by John Mearsheimer. By elevating the prominence of. 政治大. geographic constraints in his offensive realism, he pointed out that certain balance could be the. 立. the rise of another regional hegemon (Mearsheimer 2001).. 學. ‧ 國. outcome of offshore balancer’s manipulation such that a regional hegemon prudently prevents. ‧. In sum, structural explanations above are derived from a firm belief in the necessity of. Nat. io. sit. y. depicting the world and structural constraints as they are (Walt 2018). These theories alone,. er. nevertheless, fell prey to their own doctrine in that they fail to illuminate how states evaluate. al. n. v i n their environment and accordinglyC their For example, according to Walt and h erealn gpolicies. chi U. Mearsheimer, we should see a distant state reluctant to participate in disputes beyond its reach or outside the region she belongs to. Yet as I would elaborate in the later chapters, Australia turns out to be one of the most active states to participate in defense cooperation beyond the South Pacific. Similarly, it would be implausible for France and the United Kingdom to send their navy to Asia-Pacific. None of the states mentioned is hegemon that enjoys latitude to do. 6. The perception is determined, according to his theory, by four variables: aggregate power, geographical proximity, offensive power, and aggressive intention. See Walt (1987). 17. DOI:10.6814/NCCU202000065.

(28) The Origins of DCA. offshore balancing. How can we mitigate the departures? Geopolitics provides a consistent framework to enrich analyses of international relations. Geopolitics is, in fact, a distinct discipline from international relations. However, they are not incompatible but mutual supplementing. Rudolf Kjellen for example regarded geopolitics as a science inquiring the relationship between states and space. This view was concurred by later scholars such as Karl Haushofer and Zbigniew Brzezinski, both supporting that politics. 政治大. might be constraint by geography. The importance of specifying the essence of this discipline. 立. that it derives states acting logic from the geographical conditions instead of assumed logic.. ‧ 國. 學. Such a rationale actually provides a rather foundational roadmap of each state.. ‧. Despite the objectivity at the first glimpse, there are still multiple vantage points that hold. Nat. io. sit. y. various views in terms of the best way to specify the priorities. First and foremost, Halford. er. Mackinder provided one of the most important systemic geopolitical publication in which he. al. n. v i n C h Embracing the U segregated the world into several areas. e n g c h i potential of land transportation and human power, he later eloquently argued for the importance of Eurasia, which he termed as. “heartland” (Mackinder 1904). Extending the argument, Brzezinski held the historical view that, since the US is the only superpower outside Eurasia, it should be able to identify the geopolitical players and pivots to manage the global strategy (Brzezinski 1997, 10–24). Contrary to this view, earlier scholars proposed a quite different prospect. Alfred Thayer Mahan, the American naval officer, and historian, in his tremendous volumes, underscored the. 18. DOI:10.6814/NCCU202000065.

(29) Introduction. vigorous nature of sea. Flexible government, progressive people, and free, open sea lanes altogether contributed to the prosperity of human society. Hence, sea power should be the utmost concern for a country desiring dominance (Mahan 1918). Still another admonished against the singular focus on the heartland, but instead on areas engaging both continents and sea. The so-called rimland proposed by Nicholas Spykman appeared to be the main theatre of conflicts more often than continents and sea. Furthermore, they demonstrated viable options to. 政治大. extend the US influence toward the globe. This view is widely believed to inform the US. 立. containment during the Cold War.. ‧ 國. 學. Geopolitics thus far supplements the structural arguments in international relations. ‧. theories. They to some extent concur with the realist view of the balance of power that assumed. Nat. io. sit. y. a zero-sum game in which expansion must be encountered by groups of the weak (Mackinder. er. 1943). The thoughts above provided a general sketch of how structural theories could make. al. n. v i n C h politics (Simón international e n g c h i U 2013,. concrete the “context” of. 26–33). While multiple. publications fell short of clarifying geopolitics theories in their arguments, they, in fact, inherited some of the most important thoughts in the geopolitics. East Asia, for instance, has been seen as an inherently bifurcate area engendered by continent and sea. Because of this environment setting, states in this area have derived either maritime or continental interests that determine their strategic orientations and alliance choices (Ross 1999). Yet this does not imply that geopolitics hold the absolute role in alliance politics. Instead, situational and historical. 19. DOI:10.6814/NCCU202000065.

(30) The Origins of DCA. distrust could exacerbate the security dilemma which would further influence the outcome alliance dynamic (Christensen 1999). The structural pressure and geopolitical calculation both inform states’ strategy. For example, defense diplomacy has emerged to be a prominent way to engage allies and partners to foster a friendly environment. Australia, for example, has been proactively pursuing direct engagement with foreign military to address some specific issues. In a report, the Department. 政治大. of Defense stated the objectives of the Defence Cooperation Program as “working with allies,. 立. regional partners and others to shape the global and regional environment in a way favourable. ‧ 國. 學. to Australia and the ADF” (Department of Defence of Australia 2006). Others alike saw. ‧. defense diplomacy as a tailored way to achieve security goals that might conducive to a broader. Nat. io. sit. y. national goal (Baldino andCarr 2016). In the following years, Australia set out to be one of the. er. most vigorous states to reinforce defense cooperation with other countries. This progress was. al. n. v i n C hassertiveness. TheUgrowing threats destabilizing the highly motivated by China’s growing engchi region has clearly helped crystallize deeper defense cooperation, say between Australia and. Japan, for holding common region-wide concerns in especially maritime Asia (Nikkei Asian Review 2018; Matsumodo 2018; Hall andHeazle 2018).. 20. DOI:10.6814/NCCU202000065.

(31) Introduction. Institutional Dynamic. DCAs, on the other hand, are not likely to be created without mutual trust derived from previous positive feedback. A quick glimpse over multiple existing defense cooperation agreements would reveal the fact that quite a few DCAs are preceded by a security framework at higher levels. Do these phenomena imply that a higher level of cooperation is necessary before creating DCAs? Or otherwise, is it possible that DCAs start from a rather lower level of. 治政 cooperation? While the literature above has informed us of大 the formation of DCAs with some 立 see. the. ‧ 國. to. 學. possible explanations through the lens of alliance formation, it would be even more interesting interaction. between. DCAs. and. the. institutional. landscape. they. ‧. derived. Conventionally, alliance and other “high politics” have been a topic largely ascribed. sit. y. Nat. io. n. al. er. to realism international theories. However, as Keohane (1988) argued, an alliance is arguably. i n U. v. a type of institution that indeed conforms to most definitions. 7 For the proximity between. Ch. engchi. DCAs and alliances, it would be plausible also to examine how DCAs, as a member of international institutions, formed in the current institutional context. Quite contrary to the deterministic arguments put forward by structural and geopolitical theories, institutional arguments illuminated different mechanisms. Briefly say, the structural. There are several definitions for institutions. For example, Ruggie (1975) defined regime as “a set of mutual expectations, rules and regulations, plans, organizational energies and financial commitments, which have been accepted by a group of states.” Krasner (1982) alike defined that, “[r]egimes can be defined as sets of implicit or explicit principles, norms, rules, and decision-making procedures around which actors' expectations converge in a given area of international relations.” For more detailed definitions, see Krasner (1983). 7. 21. DOI:10.6814/NCCU202000065.

(32) The Origins of DCA. arguments follow the rationale that states' behavior is a rational reaction to given structural constraints and geographical conditions. This materialistic inference nonetheless omitted the possibility that states pursue cooperation out of their proactive demands instead of passive response. Recognizing individualist argument would shed light on an alternative approach that requires us to focus on DCAs themselves. Neoliberal institutionalists, for example, while acknowledging the anarchic assumption as. 政治大. neorealism does, delivered a different approach to examining why cooperation could be. 立. sustained with the facilitation of international institutions. In the absence of the leadership. ‧ 國. 學. provided by the hegemony, states would still compromise some sovereignty in exchange for. ‧. greater coordination and the subsequent positive feedback derived from institutional. Nat. io. sit. y. cooperation. More specifically, the existence of international institutions solved problems such. er. as uncertainty and transaction costs and that might further extend the prospect of future. al. n. v i n cooperation (Oye 1985; Axelrod andCKeohane Moreover, the establishment of a regime h e n 1985). gchi U enables a more effective and costless way of communication under which a pattern of coordination is created among multiple actors (North and Denzau 1994). Similarly, sunk costs that are unable to be recovered propel states to sustain compliance with international institutions. As long as the benefit derived from cooperation outweighs the profit gained from altering policies, states are likely to maintain institutions.8. 8. This argument actually underscored the persisting value of international institutions in that they make alternative courses less tempting since changing course entails greater costs. In other words, “path dependence” 22. DOI:10.6814/NCCU202000065.

(33) Introduction. However, this argument simply tells us why each DCA is created instead of its arrangements among states and relationships with other institutions, such as the relationship between alliances and other DCAs. If we go deeper into the strand of cooperation, we could legitimately doubt the origin of certain regimes. Do they appear ex nihilo? The answer would probably be no! Regimes do not come from nowhere. In addition to shared interests and necessity to coordinate with others, regimes are mostly created in an existing pattern that has. 政治大. been sustained (Keohane 1984, 79). In this aspect, scholars have been argued for a nested. 立. structure of institutions that accommodate the bargaining and issue linkage between states.. ‧ 國. 學. To specify the so-called pattern, studies have indicated that the relationship between the. ‧. World Trade Organization and Asia-Pacific Economic Cooperation best described the nested. Nat. io. sit. y. structure of international institutions. The bargaining on the latter to be “open” or “close”. er. demonstrated the disputed nature of inter-institutional bargaining. Particularly, this bargaining. al. n. v i n process is motivated by an externalCimpetus creates fluctuation of provisions of goods. h e nthat gchi U Then, the following strategic interaction would be defined by actors’ decision and perception on the nature of goods (or externalities), nature of change (e.g.: modify or create), institutional scope (e.g.: bilateral or multilateral), bargaining route (e.g.: bilateral or multilateral), and the. is exactly in place to ensure the continuity of international institutions. Such an underlying logic composes the essence of so-called institutional explanation that concentrates, besides structural limitations, man-made constraints creating feedbacks eliminating the other potential developments. While they might be seemingly similar, it could be misleading to take them as the same. To qualify as an institutional explanation, it has to show that institutional concerns surpass their structural counterparts. See Keohane 1984: 46, 100, 178; Parsons 2007; Wallander 2000 23. DOI:10.6814/NCCU202000065.

(34) The Origins of DCA. type of issue linkage as well as institutional reconciliation.9 Through examining this strand of bargaining could we illuminate the institutional outcomes. Essentially, say two institutions, how reconciliation might be achieved would really base on the types of linkage, given that they might be nested or parallel, substantive or tactical (Aggarwal 1998b). In addition, this framework, as some studies labeled themselves, could be deemed as realist institutionalists. Despite general focus on institutional matters, they further substantiate. 政治大. their arguments with distributional effects within the international institutions. This effect is. 立. especially clear in the contentions within the North Atlantic Treaty Organization (NATO).. ‧ 國. 學. Multilateralism has continued to serve as a major norm that binds all concerning parties despite. ‧. ongoing discords (Crawford 1998; Krebs 1999). We could accordingly argue that formal. Nat. io. sit. y. defense relationship as such might be gradually altered for members’ desire to change. n. al. er. institutional context. Yet to a greater extent, they are still constrained by or locked into the institutional pattern.. Ch. engchi. i n U. v. This tentative conclusion, however, falls short of describing the regional concentration of certain arrangements. It failed to appreciate variation across regions and the role of history and ideology that might facilitate alliance formation. This concern undoubtedly justifies a necessity. 9. Ernst Haas (1980) discussed the issue linkage of different types. In addition, recent scholarships have argued for a different mechanism to evidence the issue linkage of DCAs. For instance, Kinne and Bunte (2018) pointed out that DCAs are widely complemented by governments’ lending pattern. What distinguished their studies from the other is their explicit reference the relationship between bilateral interests and “network” (instead of multilateral) interests, for the former can be compatible within the latter. Furthermore, one network’s change might be reciprocated by the other. In such cases, the relationship between institutions can actually more heterogeneous than simply overlapping. 24. DOI:10.6814/NCCU202000065.

(35) Introduction. to supplement the rational institutionalist approach with historical institutionalism in that the latter reveals a greater space to conjoin external-to-institutions factors with institutional explanations.10 In terms of alliance formation, scholars have attributed the variation to different extent of how the US exerted its power. In the case of NATO, European countries’ higher likeliness to recover quicker and suffer from imminent threats from the Soviet Union has also permitted a. 政治大. multilateral form of the alliance under the US power. In contrast, Asian states were mostly. 立. newly decolonized states that seemed unlikely to recover soon. It was hence less desirable for. ‧ 國. 學. the US to bind itself in a multilateral setting. Bilateral arrangements thus become the dominant. ‧. form. Yet the US imperium is not the only dictating factor. Instead, regional powers such as. Nat. io. sit. y. Japan, Germany, China have all pursued their own preference through responding to. er. institutional structure, globally and regionally. Feedbacks as such are indeed products resided. al. n. v i n whichCishsignificantly influenced e n g c h i U by. in the regional interaction. geographical proximity. (Katzenstein 2005). Accordingly, we are presented with a plural network of states’ mutualreference and dynamic security concerns. The regional confinement created increasing pressure on states to create alternative and overlapping institutions to tackle regional affairs. The. 10. The virtue of historical institutionalism lies in its explanatory power that utilizes historical perspectives to contextually analyze institutions. By leveraging sequence, path dependence, and even conjuncture effects derived from multiple events, historical institutionalism largely explains why some institutions but not the others are more enduring. Through analyzing expectations and consequences, scholars are able to inquire why some institutions survive resiliently comparing to actors and political environment. Additionally, actors within the context also assume important roles of adjusting institutions that directly amend the course of institutional developments. For more, see March & Olsen (1998); Pierson & Skocpol (2002). 25. DOI:10.6814/NCCU202000065.

(36) The Origins of DCA. complexity alongside the previous institution arrangements will methodologically prompt us to identify the choice of certain institutional paths (Alter and Meunier 2009). Thus far, it could be observed that institutional explanation provides several insights in explaining the formation of DCAs. First, they are demanded by states to facilitate coordination. Second, they are a product of bargaining within the current institutional context. Finally, they emanate from a broader and deeper historical context which is endowed by institutional. 政治大. features. Each of them alongside structural claims, however, lacks the concrete micro-level. 立. calculations that informed the bargaining process. The next sections will fill this gap.. ‧. ‧ 國. 學. Rational Choice. Nat. io. sit. y. Finally, while DCAs are mostly driven by the need to reinforce existing cooperative. er. relationships or to enhance current capabilities, the feature also implies the concern behind. al. n. v i n Care DCAs. After all, defense and security of the same coin. Accordingly, it would h ethentwog sides chi U be tempting to question which threats are to be addressed through DCAs. Each observation above actually provides different cut to inquire how DCAs are formed, or more specifically, their context. Yet none of them properly addressed the rational calculation within certain periods. In the rest of the section, I will review the literature of rational choice to see why states decide to align with others and probe the reason why DCAs are chosen instead of making a new alliance.. 26. DOI:10.6814/NCCU202000065.

(37) Introduction. One of the most telling literature regarding rational choice comes from Thomas Schelling’s studies on deterrence and compellence. That is, the question is mostly hinged on how a state could effectively dissuade others from taking certain actions. A crude form of such dissuasion is deterrence. By incurring decisive costs on enemies, states are able to deter others from taking unwanted actions. This rationale features popularly in deterrence studies, which becomes even more complicated in the presence of nuclear weapons and the consideration of. 政治大. offense-defense balance. Alliance, in terms of deterrence, plays a vital role since it provides. 立. extended deterrence. That is, superpowers will come to allies’ aid to deter attacks from rivals.. ‧ 國. 學. While the realists see the phenomena as a way to maintain balance of power, the deterrence. ‧. theory rather focuses on how it increases costs of fighting to prevent wars (Weede 1983; Huth. Nat. io. sit. y. and Russett 1988; Fearon 2018).. er. This literature largely informs the later studies in that game theory is widely adopted to. al. n. v i n C h In other words,Uthis approach uses insights from entail the rationale within an interaction. engchi. dynamics within relationships than actors or structures. However, this is not to say that actors are not important, they are instead of the highest importance since their actions and options entail the micro-foundation of history. With these insights, scholars have been urging for the so-called “strategic choice approach.”11. 11. Strategic choice approach is largely informed by game theory. In this approach, the problem of level of analysis is downplayed. Instead, it maintains that all actors, no matter which level they reside, share the same strategic logic under rational assumption. Basically, this approach contains two factors, environment and actors. The former is made up by actions and information while the latter consists preference and belief about others’ preferences. This approach is conducive to breaking down the wall between paradigms for that it largely 27. DOI:10.6814/NCCU202000065.

(38) The Origins of DCA. Unlike previous literature that explain the formations by specifying structural pressure and path dependence, rational choice elicits the intricate balance between actors’ best strategy. That is, despite the general environment that formulates actors’ roadmap, they are drawn into strategic interactions embodying particular preferences (G. H.Snyder 1984). Thus, an actor’s concerns are taken into consideration by one another, so is otherwise. For example, in the face of potential conflicts, alliance formation could be the outcome of complex cost-benefit. 政治大. assessment that calculates expected payoffs. While an aggressor might be bolstered with. 立. offensive alliances, a victim might also be reassured with defensive pacts. In either case,. ‧ 國. 學. alliance formation does not rely on realist’s positionality but rationalist’s atomic and purposive. ‧. rationale (Smith 1995).. Nat. io. sit. y. How does rational choice, in terms of strategic choice approach, help understand. er. alliance or DCAs formations? What are alliances’ effects? Why do states ever enter formal. al. n. v i n C himportant functionsUof alliances is its ability to convey agreements as such? One of the most engchi information. Credible information is usually accompanied by sufficient costs that are either unable to recover or tying members’ hands. Through entering alliances, states credibly signaled that they will devote themselves to some extent of contingency through enhancing peacetime coordination which requires additional resources and costs. Hence, states in a defensive. investigated the similarity between different schools. For instance, the structural and institutional explanations above share a lot, if not every, common assumptions, or even the explanation of outcomes. It will be strategic approach’s top priority to theorize the commonalities in order to generate a more coherent yet inclusive explanations. See Lake &Powell (1999). 28. DOI:10.6814/NCCU202000065.

(39) Introduction. alliance might be more capable to deter aggressors, while those in an offensive alliance could be prone to initiate conflicts. The tendency is further strengthened given alliances’ formality. By formalizing obligations, the information conveyed by alliances is credible enough to affect enemies’ calculations. This forms the very first issue of strategic choice, signaling. The signals released by formal agreements can affect rivals’ prediction of a state’s next move (Morrow 1994, 2000; Leeds 2003).. 政治大. While signals are products of moments, alliance commonly endures for years, so are. 立. commitments aligning with them. Commitments are essentially dynamic given that they are. ‧ 國. 學. promises made by actors, who vary and change over time or across different types (Morrow. ‧. 1999). For example, some commitments, when made public, are credible since they produce. Nat. io. sit. y. hand-tying costs. Yet others might fail since they are not backed by enough power (Fuhrmann. n. al. er. & Sechser, 2014). The variance of these commitments’ reliability conveys different messages.. Ch. engchi. On the other hand, commitments could be unstable. iv n over U time,. a credible and reliable. commitment today can be babbling and slippery tomorrow. A strong commitment made by forming alliances, according to the logic above, will encourage a stronger stance held by member states. However, if the commitment change in that allies no longer honor their commitments. Those who stand firm or fight aggressively would be forced to fight a war that is not so preferable (G. H.Snyder 1997, 186–92).12. These dilemmas are respectively labeled “abandonment” and “entrapment.” Abandonment refers to the situation that a state might consider the alliance she belongs to credible yet get abandoned by allies when being 12. 29. DOI:10.6814/NCCU202000065.

(40) The Origins of DCA. The same problem could also stem from the internal change of member states. Say, the ruling party changes in member states so that the government becomes responsible for different interest groups. Will this new government continue keeping its promise? Studies have pointed out that democratic states are more likely to maintain their commitments than nondemocratic states since the former a constantly checked by the domestic institutions (Bueno de Mesquita et al. 1999; B. A.Leeds, Mattes, and Vogel 2009). On the other hand, the power change within. 政治大. alliances might also incentivize potential aggressors to attack. Either the stronger getting. 立. stronger or the weaker being weaker would reduce the credibility of alliances. While the former. ‧ 國. 學. now perceives intervention less beneficial, the latter might simply make no change.. The. ‧. change in power in both cases thus entailed the unreliability of alliances (Johnson and Joiner. io. sit. y. Nat. 2019).. er. How then does rational choice help understand the formation of DCAs? If we borrow. al. n. v i n C h on alliances, weUcould render DCAs as an outcome the conclusions derived from the literature engchi of strategic interaction. For those deciding to enter DCAs, the formality and peacetime coordination and collaboration commit signing parties to a more durable security relationship.. attacked by enemies. Entrapment, on the other hand implies that states might be forced to fight a war she does not want. The reluctance could be the outcome of a binding alliance that she committed heavily. In a similar logic, states might deliberately avoid the burden of standing firm against menace while expecting other states to fight back. Still others might consider the costs to fight low enough to launch an immediate strike. In either cases, structural constraints still matter such that they condition cost-benefit calculation in different fashions. It should nevertheless be noted that they act in a strategic context. See G. H. Snyder (1984), Christensen and Snyder (1990). 30. DOI:10.6814/NCCU202000065.

(41) Introduction. Furthermore, by enhancing the capabilities of signing parties, DCAs also helpfully increased the costs for those attempting to challenge them and hence signals a stronger resolve.. Scopes and Limitations One of the inherent hindrances of my research is the so-called “too many variables” problem. Indeed, small-N studies employing within- and cross-case studies can be largely restrained given their goals to exploring multiple variables in a small number of cases. For this. 政治大 problem, as I have argued above, this problem could be mistaken without considering the 立. ‧ 國. 學. causal expectation in this study. That is, my aim is to reveal conjunctural causality that induced different degrees of dependent variables through varying sequences of explanatory variables.. ‧. sit. y. Nat. In such cases, it is much more important to reveal how and to what extent a combination of. n. al. er. io. explanations can explain the cases. Further, to reduce indeterminacy, I rely on the method of. Ch. i n U. v. structured, focused comparison which allows me to avoid wandering in sought of causality.. engchi. Yet another problem emerges, the amount of case is still small, in this study, two. This problem, nevertheless, can be alleviated by disaggregating each case. Through conducting intracase comparisons such as longitudinal and cross-sectional analysis, the leverage increases in terms of the observations. For instance, Jack Snyder (1991) investigated states’ overexpansion by looking into the dynamics within cases: such as how certain roles change. 31. DOI:10.6814/NCCU202000065.

(42) The Origins of DCA. over time; how multiple roles within states induce overexpansion; and most importantly, how different states develop unique patterns of expansionist policy.13 There are multiple limitations to my research. The first limit is one of the time spans. As some could already notice, the structural and institutional features mapped into states’ preference come with a long period of time, decades if not exaggerated. For instance, Nordic states’ strategic interests cannot be viewed in isolation from their history with Russia (or the. 政治大. Soviet Union). However, attempts to encompass all these factors will incur unnecessary costs. 立. in my research given the limit of time as well as the length of this thesis. Accordingly, I restrict. ‧ 國. 學. myself to the past decade: roughly since the Global Financial Crisis. Ever since the disaster,. ‧. the international landscape changed considerably. The US, for instance, was struck heavily as. Nat. io. sit. y. the source of the financial breakdown. Also, the US global strategy was bolstered with the. er. Rebalance to Asia instead of further entanglement in Middle East. A constant narrative in this. al. n. v i n C from period of time will reduce the noise However, I do not totally discard the h e nmygfocus. chi U. information before the time span. If certain events have lingering or critical effects, then they might be included to facilitate the inference.. 13. Maximizing leverage in case studies and historical research has been a major task to scholars conducting small-N research. For example, a coherent qualitative research scheme can be enriched by increasing covariation across cases, say countries and organizations. To make it even more pluralized, these cases can be extended by examining time-variation within cases that demonstrated the dynamic development of single cases. Disaggregating cases comes as another approach to increase observable implications. In such cases, we can go through different levels in each case to simply increase sample size. For more, see King, Keohane, and Verba (1994, 5–6) and Snyder (1991, 60–65). 32. DOI:10.6814/NCCU202000065.

(43) Introduction. The second limit will be one of the issue areas. To sketch states or regions’ strategy architecture will, in fact, require overwhelming understandings across different issues. For instance, Obama’s Rebalance to Asia can hardly exclude the negotiation of the Trans-Pacific Partnership (TPP). However, issue-linkages should not impede my research. Instead, they could appear as conditional variables that ease the rigidity of some causal claims. The main focus of this research will adhere to security and defense issues. In other words, I recognize the. 政治大. possibility of interaction between issue areas. Yet I simply left the broader issues space out of. 立. the current research since they are not of interest here.. ‧ 國. 學. In sum, a crucial aim of this study is to develop a middle-range theory that explains the. ‧. development of regional defense cooperation. In other words, this study could be seen as one. Nat. io. sit. y. of theory generating rather than theory testing. By opting for this route, I am in fact aim at. er. striking the balance between academic progress as well as policy implications. Hence, an. al. n. v i n eclectic point of view should betterC serve of this thesis. In the next paragraph, I hold h ethenaim gchi U. this goal in mind and proceed to explain the defense cooperation in Indo-Pacific and the European Union.. Conclusion This chapter illuminates the importance of DCAs in contemporary international politics. There are simply more newly formed DCAs then alliances. Furthermore, I have already pointed out that DCAs are not alliances. They actually represent a wider population of interstate and 33. DOI:10.6814/NCCU202000065.

(44) The Origins of DCA. inter-military cooperation, which is still growing. We have no excuse to ignore these exciting trends despite the sparseness of systemic inquiries on this topic. In this chapter, I explore the possible explanations accounting for the formation of DCAs. First, the structural explanations indicate that the distribution of power and geographical environments have both shaped and constrained states' decisions to align or ally with others. Despite assumed rationality, this approach focuses mostly on the material structure. 政治大. that dictates states’ behavior. Second, the institutional approach turns to the interactive effects. 立. generated by DCAs. Institutions created by states, with constant feedback and response, locked. ‧ 國. 學. states into a pattern that states are willing to stay in and act according to them. States might. ‧. pursue the change of institutional contexts; they nonetheless behave as institutions predict.. Nat. io. sit. y. Third, DCAs could also be a deliberate outcome of certain interactions. States purposively sign. er. these agreements in the hope to maximize their security by either deterring enemies or assuring. al. n. v i n C interesting allies. Each of them, indeed, provides regarding the formation of DCAs. h e n g insights chi U. These theories, despite tremendous contribution to our understanding of how DCAs might derive, they fail to address the following problems. Why, from the very beginning, do states choose to adopt this form? This question is particularly interesting since DCAs coexist with many alliances. The reason why states decide to separate them could tell us how states step up to the mutual defense. Subsequently, what are the externalities of DCAs? While DCAs are task-specific, it might be tempting to explore the extra effects generated by DCAs. For. 34. DOI:10.6814/NCCU202000065.