Full Text

1 Introduction Regionally based intergovernmental organizations have become more prominent in response to the rapidly changing opportunities and challenges of the post-Cold War period. Since 1990, a number of new regional organizations have been founded – for example, the Asian Regional Forum (ARF) – and existing organizations have ‘widened and deepened’ by extending their membership and mandate, further formalizing their institutions and taking on new tasks, such as civilian and military peacekeeping. The European Union (EU) is the most striking example for these developments, but they can also be observed for other regional organizations, e.g. the Organization of American States (OAS) and the African Union (AU). In contrast, Asia has experienced a relatively weak and late development of regional security organizations (Acharya, 2009; Webber, 2010). The underlying rationale of ‘new regionalism’ (Acharya, 2007; Acharya and Johnston, 2007) emphasizes the importance of economic integration and democratization for regional prosperity and peace. Regional security organizations (RSOs) explicitly seek a prominent role in regional security governance by broadening the economic agenda of customs unions and economic communities or the narrow security guarantees offered by military alliances, such as the North Atlantic Treaty Organization (Kirchner and Dominguez, 2011 p. 1). In effect, RSOs can be viewed as the main institutions for constructing security communities (Deutch et al., 1957; Adler and Barnett, 1998). Under what conditions RSOs indeed become relevant institutions able to contribute to regional security remains, however, a largely unanswered empirical question. Accordingly, our central research questions are (i) under what conditions RSOs take up a broader agenda or scope in security governance and (ii) whether a broader scope improves regional security? We aim to answer 118 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from these questions first of all by means of a quantitative comparative analysis of the scope and impact of 25 RSOs over the period 1990–2009.2 The statistical analysis is followed by a more in-depth case study of Asian RSOs. East Asia is notable because relevant RSOs have only limited scope. Does this mean that there is indeed an East Asian ‘exceptionalism’? Based on the literature review, we identify similarity and concentration (or hegemony) as the two central conditions for increased scope. The empirical evidence, however, suggests only that RSOs with member states that are more similar in their capacities and political systems are more likely to have a broader scope. Our research further points at the importance of institutional features; agencies (rather than arrangements) that have been around longer and encompass more members have been more successful in expanding their security agenda. Next, we include scope of RSOs into basic models of interstate and intrastate conflict. The importance of international organizations is one of the elements of the Liberal Peace, together with democracy and economic interdependence (Russett and Oneal, 2001). Further, Hegre and Sambanis (2006) identify economic development, poverty, and population size as robust factors contributing to civil wars (see also Fearon and Laitin, 2003). However, there is only limited empirical support for the pacifying effect of RSOs on regional security. The results of the statistical analysis are useful to reexamine whether Asia is indeed an outlier. Is the scope of Asian RSOs limited because of factors that are unique to Asia? Furthermore, are Asian RSOs unable to play a pacifying role because of their limited scope? Among others, Acharya (2009) has argued that great power interference, rivalry, and mistrust, involving the, China, and Japan and, to some extent, India, have led to a unique trajectory for Asian RSOs. Focusing on our key dependent variables (scope and regional security), Asia appears indeed to be an outlier; Asian regional security organizations have limited scope and their members experience relatively high levels of intrastate and interstate con- flict. However, we show that current explanations tend to focus too much on the inhibiting role of (regional) superpowers. Instead, the key factors in the general models, namely whether members have similar capabilities and 2 Statistical analyses of regional security organizations are still extremely rare, whereas there are some quantitative studies on the impact of regional economic organizations on regional security; in particular by Bearce (2003), Bearce and Omori (2005) and Haftel (2007, 2012). Haftel (2013) analyzes the extent to which regional economic organizations engage in security cooperation. The scope and impact of regional security organizations 119 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from political systems as well as since how long regional organizations have been developing their security mandate, seem to apply similarly both within and outside Asia. Further, we find that – particularly in Asia – the legacy of conflict appears to inhibit the development of a regional security structure. 2 Scope and impact of regional security organizations In their comparative case studies, Tavares (2009) and Kirchner and Dominguez (2011) consider the characteristics of RSO members as well as the scope of RSO activities to evaluate their impact on regional security. Analytically, however, scope and impact are distinct, albeit related, issues. In itself it is interesting to know when RSOs are able to establish themselves as actors with a broad security mandate and operational capacity. Since not all RSOs are similar, it is important to control for scope when analyzing their impact, if any, on security. Moreover, the same factors that may make con- flict more likely in a particular region may also limit the scope of RSOs (Buzan and Wæver, 2003; Colaresi et al., 2007). Thus without proper controls, the pacifying effect of RSOs can also easily be overstated. 2.1 Hegemony and similarity as determinants of scope Political leadership is often seen as crucial for creating and sustaining international organizations. Hegemonic stability theory (Kindleberger, 1973; Krasner, 1976; McKeown, 1983) argues that periods of political stability and economic prosperity result from the ability and willingness of powerful ‘hegemonic’ states to regulate interactions in the international system. Hegemons establish and sustain international organizations in order to extend their political influence and to legitimize their normsetting behavior.3 In terms of Olson’s (1965) logic of collective action, a hegemon has private incentives to provide security as a public good. Classic examples are Great Britain ‘holding the balance of power’ in the 19th century Concert of Europe, or the United States founding the United Nations (UN) after the Second World War. A regional power may operate 3 Keohane (1984) argues that the institutions, norms and regulated behavior may well persist after the hegemon has lost its dominant position. 120 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from similarly at the sub-systemic level, for example, the importance of France and Germany for the creation of the European Community. International organizations also provide a framework for member states to agree on appropriate burden sharing in the provision of security as a collective or public good (Sandler, 1992). On the basis of instruments and functions of security governance, Kirchner and Sperling (2007) differentiate between protection, prevention, assurance and compellence, each category with its distinct joint production function (Hirschleifer, 1983). Dorussen et al. (2009) argue that the contributions of member states to assurance and prevention policies weigh equally, whereas the minimal contribution sets the collective level of protection, and the maximum contribution the collective level of compellence. It follows that free riding is most likely for compellence policies and least likely for protection, and thus, the importance of hegemony for security governance should be particularly pronounced for compellence policies. Koremenos et al. (2001) similarly conjecture that scope increases with the severity of distribution and enforcement problems. Applying Kirchner and Sperling’s (2007) distinction between assurance and compellence, Haftel (2013) finds that regional hegemony (Buzan and Wæver, 2003) increases security cooperation, whereas rivalry between major powers in the region (Colaresi et al., 2007) limits it. Major powers can also impede the development of regional organizations. Since they are best able to protect their state sovereignty, they have less need for international organizations (Gowa, 1989). They are also likely to veto any external attempts to limit their freedom to act in defense of perceived national interest. Major power support for RSOs is often conditional on their ability to determine collective policies and does not imply any willingness to delegate sovereignty. Since hegemons face fewer enforcement problems, it follows from Koremenos et al. (2001) that hegemons may also reduce the scope of RSOs. Major powers often end up limiting the autonomy or freedom of action of regional organizations – even when they were instrumental in setting them up. In summary, hegemons may prefer weakly institutionalized arrangements with limited scope. Colaresi et al. (2007) argue moreover that a strategic rivalry involving the regional hegemon will further undermine regional security cooperation. As Acharya and Johnston (2007) observe, developing countries often favor regional cooperation as a means to safeguard their state sovereignty. Individually, these states consider themselves to be in a weak position relative to major powers or even international organizations such as the UN The scope and impact of regional security organizations 121 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from or the EU. Accordingly, similarity of capacity and interests should promote the formation of regional organizations. Moreover, since similar states individually will find it more difficult to enforce adherence to collective agreement, they should be more willing to increase the scope of RSOs (Koremenos et al., 2001). As the case of the Association of Southeast Asian Nations (ASEAN) illustrates, RSOs can be seen as opposing an emerging dominant power. Moreover, neo-functionalism (Haas, 1964) holds that regular interaction between policy-making elites to settle technical or economic issues leads via sectoral/functional spillovers to the recognition of overlapping interests and similarity of ‘policy cultures’ and thus promotes further regional cooperation, including areas of ‘high’security governance. The EU is the classic example for the latter where security (apart from police matters) only became part of its remit nearly forty years after the foundation of the European Coal and Steel Community. In engaging with regional security problems, shared local understanding of cultures and interests provides regional organizations with a comparative advantage over global organizations (e.g. UN peacekeeping operations in Africa) or organizations operating outside their main area of interest (e.g. NATO’s International Security Assistance Force in Afghanistan) (Tavares, 2009, p. 4). Perceived shortcomings of the UN to effectively address important security problems further stimulate regional cooperation (e.g. the International Conference in the African Great Lakes area or the Economic Community of West African States (ECOWAS) in West Africa). During the ‘third wave’ of democratization, regional security cooperation has been of particular importance for newly democratized states (Huntington, 1991); most notably, democratic institutions were a precondition for joining the European Community/Union and have contributed to the consolidation of democracy in Southern and Eastern Europe. Mansfield and Pevehouse (2008) find that membership in an international organization can enhance the credibility of a political leader’s commitment to democratic reform. Accordingly, democratizing countries are more likely to join international organizations. The compatibility and shared norms of democracies are seen as conducive to peaceful international interaction and functioning of RSOs. Donno (2010), however, also observes that any RSO response to a flawed election is less likely as a country’s geopolitical size and importance increases demonstrating that power inequality can undermine RSO effectiveness (see also Haftel, 2013). 122 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from In summary, the literature offers two opposing explanations for the creation of regional organizations and the expansion of security tasks – or their scope. Hegemonic stability theory emphasizes the importance of a regionally dominant state or power inequality. Alternatively, similarity of capacities and interests promotes regional cooperation in opposition to external dominant states or threats. Similar political, in particular democratic, institutions are also seen as favorable for regional cooperation. These explanations, moreover, only partially contradict each other, since it is possible that a small group of large states leads integration (e.g. the significance of Germany, France and the United Kingdom for European integration), or for states to share similar democratic institutions while differing significantly in their capabilities (e.g. cooperation between democratic states with significant variation in population size and economic prowess as in the ‘Southern Common Market’, Mercosur). Finally, RSOs are bureaucratic organizations that may have a momentum of their own. RSOs representing a larger number of (richer) member states are more likely to have the resources needed to develop new initiatives (Koremenos et al., 2001). A (semi-)independent secretariat is also more likely to explore new areas of operations (Boehmer et al., 2004). Organizations that have been in existence for longer periods are more likely to initiate new policies in response to challenges.4 Finally, member states are also more likely to turn to existing, successful organizations as instruments for additional security governance tasks, e.g. the redefinition of NATO’s role after the end of the Cold War. 2.2 Security organizations and regional security Russett and Oneal (2001)include international organizations as one of the core elements of the Liberal Peace, together with democracy and economic interdependence. International organizations provide forums for interaction between political leaders, allowing them to communicate their concerns and find shared interests. International organizations can also mediate directly to solve conflicts and or provide mechanisms such as peacekeeping or monitoring missions to help states implement agreements. Empirically, however, the pacifying effect of international organizations is seen as the weakest element of the Liberal Peace (Dorussen and Ward, 2008). Since 4 Historical institutionalism emphasizes the path dependencies and lock-in effects of existing institutions that were built in the shadow of the Cold War (Katzenstein, 2005). The scope and impact of regional security organizations 123 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from international organizations differ markedly in their mandate, level of institutionalization (Boehmer et al., 2004) as well as membership characteristics (Pevehouse and Russett, 2006), it may be illusionary to expect a uniform effect. Thus, it is helpful to focus on regional security organizations and to explicitly control for their actual, instead of stated, involvement in security government. By definition regional security, organizations have regional peace and security as their core mandate, making this the ultimate criterion to judge their effectiveness.5 In the post-Cold War period, the most serious security threats for many states resulted from state failure and civil wars. Following the withdrawal of external support of the global superpowers, many states (some of them newly established) faced serious internal challenges to their authority. The spread of civil wars with associated flows of refugees and internally displaced people further increased regional instability. The presence of failed states in a region providing ‘safe-havens’ for terrorism, drug-trafficking and internationally organized crime have also led to (regional) insecurity. Hegre and Sambanis (2006, p. 531) identify economic development, poverty and population size as robust factors contributing to civil wars: We confirm that a large population and low per capita income increase the risk of civil war, and this is consistent with many studies of civil war. We have found other robust relationships: civil wars are more likely to occur in countries with recent political instability and inconsistent democratic institutions; countries with small militaries and rough terrain; countries located in war-prone, undemocratic regions; and countries with low rates of economic growth. The typology developed by Kirchner and Sperling (2007) underscores the broad set of challenges of security governance where other states often play only a minor role as protagonists and, secondly, threats against the state are often indirect rather than direct. The main functions of regional security governance encompass institution building as well as conflict resolution employing both persuasive and coercive instruments. The better RSOs are able to engage with the complete set of security governance 5 Bearce (2003), Bearce and Omori (2005) and Haftel (2007, 2012) argue that regional economic organizations increase regional security. It is also important to recognize that the boundaries between regional economic and security organizations can be blurred. 124 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from tasks – protection, prevention, assurance and compellence – the better able they should be to promote regional stability. 3 Comparative analysis of regional security organizations It is not immediately obvious which international organizations qualify as regional security organizations. Tavares (2009) identifies a total of 35 RSOs whereas Kirchner and Dominguez (2011) limit their attention to 10. The boundaries between regional economic and regional security organizations are also poorly demarcated. Haftel (2012) identifies 25 regional economic organizations (REOs) but 11 of them are also classified as RSOs by Tavares.6 Since it is most comprehensive, we start with the list compiled by Tavares (2009, pp. 10–11) and also use his classification of RSOs as either agencies or arrangements. Organizations that, upon further research, appear to be exclusively concerned with economic or cultural matters were excluded as well as completely inactive organizations. Table 1 provides a survey of the 25 RSOs included with information on their formation date (using the date that the organization attained its security mandate). All data were coded for five-year periods: 1990–2004, 1995–99, 2000–04, and 2005–09, resulting in a data set with a maximum of 87 observations.7 The main sources of information on RSOs are the comparative case studies of Tavares (2009) and Kirchner and Dominguez (2011). Further basic information on membership (whenever possible) was taken from the Correlates of War IGO data (version 2.3; Pevehouse et al., 2004).8 We also added information using the organizations’ websites, Wikipedia, as well as further case studies.9 6 Haftel (2013) compiles a list of 34 REOs, of which we classify 15 as regional security organizations. Ten of these 15 RSOs are coded by Haftel as having assurance policies, whereas five (CSTO, ECCAS, EU, GCC and ECOWAS, seeTable 1) include compellence. 7 We collected data for starting year of any particular period as well as averaging over the five year period (which had, however, no impact on our results). To avoid losing observations because of lagging, we also collected information on conflict for 1985–89. Finally, we only counted years that countries/organizations were actually in existence. 8 Note that RSO membership as well as membership of the international system is not necessarily constant over the five year periods. We adjusted the data accordingly. For some organizations, particularly those with many small island states, official membership varies notably from the membership size as measured in our data. 9 Jácome et al. (2005) was particularly useful for Latin American RSOs. The scope and impact of regional security organizations 125 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from Table 1 Regional security organizations Acronym Name Since RSO mandate Agency/ arrangement Region Scope (mean) ARF Asian Regional Forum 1993 1993 Agency Asia 0.25 ASEAN Association of Southeast Asian Nations 1967 1967 Agency Asia 0.75 AU African Union 2002 2002 Agency Africa 5 CAN Andean Community 1967 1967 Agency America 2 CARICOM Caribbean Community 1973 1973 Agency America 3 CICA Conference of Interaction and Confidence Building Measures 1999 1999 Agency Asia 0 CSTO Collective Security Treaty Organization 2002 2002 Agency Former SU 1.5 ECCAS Economic Community of Central African States 1985 1985 Agency Africa 0 ECOWAS Economic Community of West African States 1975 1975 Agency Africa 3 EU European Union 1957 1957 Agency Europe 4.75 GCC Gulf Cooperation Council 1981 1981 Agency Middle East 0 GRIO Rio Group 1986 1986 Arrangement America 0 GUAM Organization for Democratic and Economic Development 2001 2001 Agency Former SU 0 IC/GLR International Conference of the Great Lakes Region 2000 2000 Arrangement Africa 0.5 IGAD Intergovernmental Authority for Development 1986 1986 Agency Africa 1 LAS League of Arab States 1945 1945 Agency Middle East 2 MERCOSUR Mercado Comun del Sur 1991 1991 Agency America 0.5 MRU Mano River Union 1973 1973 Agency Africa 0 NATO North Atlantic Treaty Organization 1949 1949 Agency Europe 2 OAS Organization American States 1948 1948 Agency America 5 OECS Organization of East Caribbean States 1981 1981 Agency America 0 OSCE Organization for Security and Co-operation in Europe 1973 1995 Arrangement Europe 4.25 SADC Southern African Development Community 1992 1992 Agency Africa 0.5 SCO Shanghai Cooperation Organization 1996 2001 Agency Asia 0.67 UNASUR Union of South American Nations 2008 2008 Arrangement America 0 126 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from The main dependent variables are RSO scope and regional insecurity. Scope was assessed using Kirchner and Sperling’s (2007) classification of security governance. The scope of an RSO includes assurance if it fields civil peacekeeping missions or engaged in mediation. Prevention was coded if an RSO (or affiliated regional development bank) was listed as providing multilateral aid where we mainly relied on Aiddata (www.aiddata.org, 20 November 2013, date last accessed). Protection is coded when we could find evidence for collaboration on police, justice and corruption matters; in particular, we found instances of collaboration to fight international cross-border crime and border protection. Finally, three indicators are used to assess compellence. Heldt and Wallensteen (2007, pp. 46–55) list non-UN peacekeeping missions with a military component. The International Institute for Economics (http://www.iie.com/research/topics/sanctions/sanctions-timeline.cfm, 20 November 2013, date last accessed) is the main source for (economic) sanctions. Finally, we coded if members were suspended from an RSO. All indicators were coded as dummy variables if a particular policy was present during a particular period. The resulting scope variable is an eight-point scale (0–7), but 6 is the actual maximum value (attained by the AU and the EU). Nearly half of our observation score zero on scope with some organizations scoring zero for all four periods, e.g. the Mano River Union and the RIO group. Regional (in)security is measured using the Uppsala conflict data (http:// www.pcr.uu.se/research/ucdp/datasets/, 20 November 2013, date last accessed), which include both interstate and intrastate conflicts with more than 25 casualties annually. Regional insecurity aggregates all conflict years for all RSO member states where countries may have more than one civil war and, by construction, any interstate war between two RSO members is counted twice. The Penn World Tables (Heston et al., 2012) 10 provide the data on economic development (in purchasing power parity), openness and population. Hegemony is measured as the proportion of total GDP controlled by the ‘largest’ member of the RSO.11 Similarity in capacity is measured as the standard deviation of total GDP. Polity IV (Marshall et al., 2012) and 10 PWT 7 from http://pwt.econ.upenn.edu/php_site/pwt_index.php, 20 November 2013, date last accessed. 11 In his cross-sectional analysis, Haftel (2013) uses a dichotomy of hegemony where an REO scores high on hegemony when half or more of all resources are held by one member. Seven of the 15 organizations that are included in our list of RSOs have high hegemony. Using Colaresi et al. (2007), Haftel (2013) identifies 8 REOs with major rivalry, but only one of them (CAN, seeTable 1) is included in our list of RSOs. The scope and impact of regional security organizations 127 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from Freedom House (http://www.freedomhouse.org/, 20 November 2013, date last accessed) were used to assess level of democracy. The Polity data give the average level of democracy (/autocracy) of the members of the RSO. The standard deviation of the democracy and autocracy scores were used to assess degree of political similarity of political system of the member states. As an alternative to the Polity data, we also include the Freedom House data. Acknowledging that the Freedom House data are not necessarily compatible over time, we only use their classification of political systems as free, not free or partially free. The proportion of member-state years coded as free, partially or not free measures similarity of political system. Further, we identify the region where the RSO is mainly relevant (see Table 1).12 To avoid bias in the estimation of any effect of scope on security, it is important to control for the main causes of (regional) insecurity. Following existing literature on the Liberal Peace, we control for democracy, and trade (by means of openness). Based on the civil wars literature, we include population and economic development. Several factors that are considered to support regional stability directly could also do so indirectly via their effect on scope; for example, hegemonic stability theory suggests that a hegemon has an incentive to provide regional stability directly as well as via creating and sustaining international organizations. The democratic peace suggests that regions with predominantly democratic states are more peaceful than those where democratic and autocratic states co-exist. However, democracies are also considered to be more supportive of regional security cooperation. Including member-state characteristics into the model is thus also important to avoid possible spurious correlation between RSO scope and regional (in)security. The logit and Poisson models are appropriate for limited dependent variables and, where needed, control for temporal effects and over-dispersion of the error terms. 4 Explaining the scope of regional security organizations Table 2 provides the results for the Poisson regression with scope as the dependent variable and formation of the organization, hegemony, and 12 The grouping is based mainly on the geographical location of the majority of the member states, but admittedly the global superpowers (in particular the USA) tend to be member of RSOs commonly associated with different regions. The allocation of RSOs to specific regions was compared with and largely followsBuzan and Wæver (2003). 128 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from similarity of capabilities, democracy, and conflict as the main independent variables. The models further control for the EU and Asian based RSOs.13 RSOs that have held a mandate in security governance for a longer period, with a semi-independent secretariat, or that have a larger Table 2 Scope of activities of regional security organizations, 1990–2009, Poisson random effects (1) (2) (3) Years since RSO formation 0.017 (0.006)a 0.022 (0.006)a 0.018 (0.006)a Number of member states 0.053 (0.014)a 0.044 (0.012)a 0.067 (0.013)a Agency 1.509 (0.425)a 1.070 (0.387)a 1.437 (0.455)a EU 0.663 (0.262)b 1.010 (0.292)a 0.983 (0.296)a Asian RSO −0.744 (0.520) −0.650 (0.510) −0.171 (0.455) Hegemony −0.818 (0.526) −0.926 (0.552)c −0.590 (0.591) Divergence capacity (SD_Total GDP, ln) −0.111 (0.034)a −0.109 (0.030)a −0.129 (0.038)a Polity2 (means) 0.056 (0.025)b Divergence political system (SD_Polity2) 0.101 (0.068) 0.082 (0.068) Democracy (mean, Polity2) −0.092 (0.046)b Autocracy (mean, Polity2) −0.269 (0.066)a Partial free (proportion, Freedom House) 2.247 (0.703)a Free (proportion, Freedom House) 0.882 (0.450)b UCDP conflict (lagged) −0.014 (0.012) −0.005 (0.010) −0.018 (0.012) Constant −0.382 (0.678) 0.978 (0.661) −1.011 (0.606)c Observations 83 83 87 Wald χ 2 150.49a 159.82a 168.49a PseudoR 2 0.37 0.40 0.39 a Significant at 1%, b significant at 5%, c significant at 10%. Robust standard errors in parentheses. 13 The EU is generally recognized to have a special status; excluding the EU the main findings remain robust but using xtpoisson (random effects) rather than Poisson regression becomes necessary. The scope and impact of regional security organizations 129 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from membership are significantly more likely to take on a large set of tasks. These findings are the most robust and strongly indicate the significance of autonomous development of regional organizations. The models further confirm the special status of the EU. Contrary to expectations from hegemonic stability theory, scope does not increase if the share of capacities of the largest state increases. Divergence of capacity among RSO members decreases scope (or conversely similarity of capacity increases it). The relation between democracy and scope deserves more attention. Model 1, using the average Polity score as well as the standard deviation of Polity score, finds that an RSO with on average more democratic members is more likely to have a broader scope. Model 2 suggests that the effect of democracy is in effect nonlinear; RSOs with on average more democratic as well as autocratic members are less likely to have a broader scope. In both models, however, we find no effect that similarity of political system across members is significant. Model 3 uses the Freedom House scores and considers the proportion of RSO members that are free or partially free rather than not free. Increasing the proportion of free (rather than not free) member states and, even more pronounced, partially free member states increases scope. These findings give strong empirical support to the observation that, after the Cold War, democratizing states have attempted to use RSOs to consolidate their transitions (Mansfield and Pevehouse, 2008). In summary: RSOs that are more similar in their capabilities and that have a larger proportion of democratic (or democratizing) member states are more likely to have a broader scope. Somewhat surprisingly, the effect of previous regional conflict on RSO scope is insignificant. Possibly the effect of conflict is indeterminate: conflict increases the need for effective RSOs but also makes it more difficult for states to cooperate. Finally, the statistical analysis confirms the status of EU as an outlier, but does not provide any immediate support for the idea that the scope of Asian RSOs is exceptional. We discuss Asian RSOs in more detail later. Figure 1 gives the marginal effects of the independent variables (of Model 3) on the scope of RSOs to illustrate their substantive impact. The impact of membership size is particularly noteworthy. Further, Fig. 1 shows that the substantive effect of Asian RSOs is low, just like their statistical significance. The remaining statistically significant variables – years since RSO formation, similarity of capacity and political system and agency – have similar substantive impact that is, however, relatively small (about one point on an eight-point scale). 130 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from 5 The limited impact of regional security organizations on security Does a broader agenda help RSOs to effectively address regional security concerns? The findings of the previous section further suggest that it is relevant to control for membership characteristics as well as‘how dangerous the neighborhood is’. Table 3 presents the main models for total conflict, Figure 1 Scope of regional security organizations, marginal effects. Figure 2 Increase of regional conflict, marginal effects. The scope and impact of regional security organizations 131 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from increase, and decrease of regional conflict, respectively. Figure 2 presents the marginal effects on the likelihood of increased conflict (Model 5). The first important observation is that the overall fit of the conflict models is fairly weak. Secondly, the ‘standard’ explanations for conflict are only significant in Model 4 (for total amount of regional conflict). Here, we find that as expected wealth (measured as per capita GDP) decreases, and population size increases the amount of regional conflict (Hegre and Sambanis, 2006). Larger regions (encompassing more members) are also more prone to conflict.Moreover, confirming the expectations of the Liberal Peace, trade interdependence and democracy appear to be pacifying. If the proportion of not free or partially free member states in the RSO increases, the total amount of regional conflict increases (with a significant effect for Table 3 Regional security organizations and conflict, 1990–2009 (4) Sum (5) Increase (6) Decline Conflict (lagged, total) 0.007 (0.007) −0.088 (0.031)a 0.110 (0.037)a Size of membership 0.039 (0.012)a 0.140 (0.052)a −0.051 (0.043) Asia 0.110 (0.494) 0.976 (1.182) −2.832 (1.333)b GDP pc (ln, mean) −0.256 (0.154)c −0.164 (0.346) −0.351 (0.358) Openness (ln, mean) −1.064 (0.402)a −1.795 (1.705) 3.340 (1.585)b Openness (standard deviation) 0.013 (0.007)c Population (ln, mean) 0.416 (0.173)b −0.057 (0.279) 0.544 (0.266)b RSO scope 0.017 (0.050) −0.510 (0.253)b 0.093 (0.218) RSO agency 0.215 (0.202) 0.657 (0.793) 0.049 (0.797) Not free (proportion, Freedom House) 1.235 (0.518)b 1.900 (1.243) −1.317 (1.182) Partial free (proportion, Freedom House) 0.638 (0.558) 1.684 (2.232) −2.215 (1.953) Constant 2.824 (2.457) 6.686 (7.569) −16.189 (7.060)b xtpoisson logit logit Observations 87 87 87 α 0.36 Wald χ 2 62.14a 13.22 16.98c Pseudo R 2 0.12 0.16 a Significant at 1%, b Significant at 5%, c significant at 10%. Number of RSOs = 25; robust standard errors in parentheses. 132 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from the proportion of autocracies). Trade openness decreases the probability of conflict, whereas variation in trade openness increases it.14 Focusing on RSO characteristics, the findings are disappointing. Both scope and agency status are generally insignificant. RSOs with a semiindependent secretariat and a broader mandate are not more or less likely to affect the amount of regional conflict. Only in Model 5, there is some evidence for the pacifying effect of scope: RSOs with a broader mandate make it less likely that the amount of regional conflict increases, which suggests that RSOs may have a deterring effect. As shown in Fig. 2, the substantive impact of membership size and previous conflict is about half a point (on a one point scale from zero to one), whereas the impact of scope is a third of a point; specifically for our empirical data, this means that an increase in RSO scope by three points makes it
33% less likely that the amount of conflict in the region will increase. Model 6 shows that the effect of scope on decline of conflict is, however, again insignificant. Overall, very few variables reach accepted levels of significance in the Models 5 and 6. More conflict in the previous period makes it less likely that conflict further increases and more likely that there is a decrease of conflict. This may reflect war weariness or the overall downward trend in conflict between 1995 and 2010 (after a sharp increase in the period around the end of the Cold War). Conflict is more likely to increase in larger regions. As shown in Fig. 2, if RSOs have 25 more member countries, the probability of having more (rather than less) conflict increases by 50%. Finally, Asian RSOs generally do not differ significantly; the only exception is that conflict appears less likely to decline in Asia. 6 Are regional security organizations in Asia ‘exceptional’? Explanations abound on why Asia has experienced a relatively low and late development of regional security organizations.15 Prolonged interstate 14 Dorussen and Ward (2010) and Gartzke (2007) argue that trade openness is a more appropriate measure for trade interdependence that bilateral trade, since it includes both direct and indirect (via third parties) trade flows. This should apply particularly for any assessment of regional interdependence. 15 Part of the late development are failed attempts by Asian regional organizations such as the Indian-sponsored Asian Relations Organization, the US-led Southeast Asia Treaty Organization, the Japanese and Australian backed Asia Pacific Council, the Southeast Asia Treaty Organization (SEATO) and the Malaysia–Philippines–Indonesia grouping The scope and impact of regional security organizations 133 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from and intrastate conflicts in the region, the persistence of several autocratic or conservative regimes, the absence of a shared liberal democratic culture, and the relatively low degree of intra-Asia trade are among the main explanations provided for this phenomenon (Mattli, 1999, pp. 169–170; Chang-Fawn, 2009;Webber, 2010; Jones et al., 2013, p. 146). Similar characteristics have plagued security organizations in Africa without, however, having inhibited organizational development, as witnessed by the relative strength of the AU, ECOWAS, and the South African Development Community. What sets developments in Asia apart from those in Africa seems to be great power interference, rivalry, and mistrust, involving the United States, China, and Japan and, to some extent, India (Acharya, 2009). Unlike in Europe, where the United States supported NATO and the EU, in Asia, the United States focused on a construction of a ‘hub-and-spokes’ network of bilateral military alliances, in which their treaties with Japan and South Korea were the centerpieces (Hemmer and Katzenstein, 2002). This ‘hub-and-spokes’ approach has enabled the United States to play the role of an external balancer for half a century and to offset the potentially destabilizing rivalries of the region’s major powers: Japan and China (Jones et al., 2013, p. 113). The organization with the longest endurance in Asia is ASEAN, founded at the time of the Vietnam War in 1967 by a group of non-Communist South Asian states concerned to overcome their own rivalries and conflicts in the interests of national economic development and to bolster their capacity to resist the threat of communist expansion in their neighborhood (Webber, 2013, p. 97). The transnational challenges arising after the end of the Cold War – terrorism, the proliferation of weapons of mass destruction, maritime security, greater uncertainty regarding the stance of the Great Powers in the region, and the fear of isolationism – required ASEAN to reassess its existing security provisions. This resulted in the formation of the ARF in 1996, which sought to create the conditions for regional stability and to ensure a distribution of power (MAPHILINO). It was only in the mid-1960 that a more permanent regionally based organization came into being in the shape of ASEAN, and it took until the mid-1990s before China and Japan were willing to join a regional organization, in the newly formed ARF. In 2001, China became a founding member of the Shanghai Five grouping in 1996, which was renamed in 2001 to the Shanghai Cooperation Organization (SCO). For further details on why China and Japan changed their initial reluctance towards regional integration, see Lanteigne (2005) and Chang-Gun Park (2006). 134 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from among the three major Asian-Pacific states: China, Japan, and the United States. The ARF remains, however, based on ASEAN-style diplomacy; it relies on noninterference in internal affairs, nonuse of force, pacific settlements of disputes, consensus decision-making, and ‘good neighborliness’ as main principles of action. It has a strong preference for non-binding and non-legalistic approaches and displays minimal institutionalization, thus setting itself apart from European security structures (Webber, 2013, p. 346). By seeking to promote lasting peace through confidence-building measures, the ARF is about ‘identity-building’ and its members hope that ‘dialogue should lead to socialisation which, in turn, will lead to the dissipation of conflicts of interests’(Garofano, 1999, p. 78). In common with other regional security organizations in Africa, Europe, and the Western Hemisphere, Asian regional security organizations have put forward progressive ideas, aims, and measures to deal with internal and external security threats.16 However, Asian RSOs remain slow in implementing these stated aims and/or in delegating appropriate tasks to central institutions. The created institutions are often little more than consultative forums. Governments have seldom been willing to accept even the most modest of constraints on their autonomy in policy-making as the price of constructing institutions. There is hence a failure to agree on clearly specified and enforceable obligations or on effective dispute settlements (Chang-Fawn, 2009, p. 235). Moreover, while the ARF has the advantage of regrouping most of the Asian states, its effectiveness is diminished by the fact that it includes important outside actors, such as the EU, Russia, and the United States, which impede its organizational effectiveness. Equally, the appearance of sub-groupings or extensions of ASEAN in the shape of the ASEAN Plus Three (China, Japan, and South Korea), the East Asia Community, and the East Asia Summit (ASEAN 10, plus China, Japan, South Korea, India, Australia, and New Zealand) seems to foster greater competition rather than a strengthening of, for example, the ARF, ASEAN, or the Shanghai Cooperation Organization. 16 Examples comprise the proclaimed Zone of Peace, Freedom and Neutrality (ZOPHAN), the Treaty of Amity and Cooperation in Southeast Asia (TAC), the Southeast Asia-NuclearWeapon-Free Zone (SEANWFZ), the ASEAN Political and Security Community, and the ASEAN Charter. The scope and impact of regional security organizations 135 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from Our data largely confirm this characterization of Asian RSOs. The scope of activities of RSOs based in Asia is significantly smaller than of RSOs based outside Asia.17 Regional organizations attained their security mandate significantly later in Asia compared with other regions. The member states of Asian RSOs have significantly more varied political systems with fewer democratic and more autocratic member states. However, the distribution of capacities (hegemony and similarity) does not differ significantly between Asian and non-Asian RSOs. Table 4 compares the effect of key independent variables on scope for Asia-based RSOs in contrast to those based outside Asia.18The marginal effects are illustrated in Fig. 3. Comparing the effect for RSOs inside and outside Asia, the coef- ficients tend to be quite similar, which indicates that Asia does not structurally deviate. The finding that the coefficients for Asia are less often significant is unsurprising, given the relatively few observations for Asian RSOs. Figure 3 clearly shows the large uncertainty in the estimates for the impact of agency, hegemony, and similarity for Asian RSO. In contrast, whereas years since RSO formation and the membership size have a positive impact on RSO scope outside of Asia, the impact is smaller and statistically insignificant for Asian RSOs. However, the most notable difference is the effect of regional conflict (Fig. 3); the coefficient for conflict is only significant for Asia, indicating that the legacy of conflict in Asia tends to inhibit in particular the ability of Asian RSOs to increase their scope of activities. Seen from the perspective of ongoing conflict between India and Pakistan over Kashmir, the Afghanistan problem, and territorial disputes in the East and South China Sea, Asian regional security organizations would appear to have little impact on peace and stability in Asia. However, beyond this prima facie evidence, the structural arrangement of the ARF and the inclusion of the great powers appear indirectly to contribute to peace and stability in the region. In particular, ASEAN has been seen as having provided rapprochement of Vietnam and Laos and other ASEAN 17 The average scope of RSO in Asia equals 0.4 whereas it is 1.7 for those outside Asia. Using the t-test for a comparison of means, the null hypothesis that the scope of Asia RSOs equals the scope of non-Asia RSOs can be rejected withP< 0.05. 18 Model 3 (in Table 2) is the basic underlying model to explain RSO scope, and the analysis focuses on the interaction between Asian RSOs with date of security mandate, membership size, agency, hegemony, divergence, and proportion of non-free and free states as defined by Freedom House. 136 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from members (Chang-Fawn, 2009, p. 23) and as having been instrumental in concluding the conflict between Cambodia and Vietnam at the end of the 1980s (Ravenhill, 2009, p. 221). Our data confirm the initial impression that conflict is relatively more common in the member states of Asian RSOs compared with those outside Asia; we find on average 25 conflict years per five-year period for Asia RSOs compared with 10 conflict years outside Asia.19 Asia is, however, not distinct in the increase (or decrease) of conflict since the post-Cold War period. Table 5 compares the effect of key features – scope, membership size, and agency – of Asia-based RSOs on regional security with those based outside Asia. Comparing the effects of RSO features, the coefficients are once more generally quite similar, indicating that Asia does not structurally deviate. Interesting marginal effects are given in Fig. 4, which first of all shows that the impact of membership on probability of increased conflict and scope on decline of conflict are nearly identical for Asian and Non-Asian RSOs. The impact of scope on the probability of increased conflict is also largely the same for Asian and Non-Asian RSOs, Table 4 Scope of RSO activities, comparison of Asian and non-Asian RSOs, 1990–2009, Poisson random effects Scope Asia Non-Asia Years since RSO formation 0.012 (0.017) 0.017 (0.005)a Size of membership 0.017 (0.035) 0.061 (0.012)a Agency 1.263 (0.734)c 1.487 (0.450)a Hegemony −0.410 (1.267) −0.474 (0.601) Divergence (SD_Total GDP, ln) −0.102 (0.029)a −0.086 (0.027)a Non free (proportion, Freedom House) −0.725 (0.837) −1.655 (0.573)a Free (proportion, Freedom House) −5.494 (3.189)c −0.707 (0.643) UCDP conflict (lagged) −0.036 (0.016)b −0.011 (0.012) a Significant at 1%, b significant at 5%, c significant at 10%. Model 3 (Table 2) is underlying model, only coefficients reported are interactions with Asia and Non-Asia regions; N= 87; robust standard errors in parentheses. 19 The null hypothesis that the sum of conflict years for Asian RSOs is similar to non-Asian RSO can be rejected withP< 0.01. The scope and impact of regional security organizations 137 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from Figure 3 Scope of regional security organizations, comparison of Asian and non-Asian RSOs, marginal effects. 138 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from although it is only (weakly) significant outside Asia. The only notable observation applies to membership size; especially in Asia, more inclusive regional security organizations are less likely to see a decline Figure 3 Continued. Table 5 Impact of RSO features on regional conflict, comparison of Asian and non-Asian RSOs, 1990–2009 Sum Increase Decline Asia Non-Asia Asia Non-Asia Asia Non-Asia Scope 0.021 (0.117) 0.016 (0.053) −1.132 (1.523) −0.492 (0.265)c 0.737 (0.834) 0.120 (0.207) Members 0.044 (0.024)c 0.034 (0.011)a 0.276 (0.102)a 0.131 (0.046)a −0.233 (0.080)a −0.024 (0.039) Agency 0.228 (0.208) 0.072 (0.321) 0.957 (1.143) 0.532 (0.815) −0.418 (1.202) 0.245 (0.812) Model xtpoisson logit logit a Significant at 1%, b significant at 5%; c significant at 10%. Models 4, 5, and 6 (Table 3) are underlying models, only coefficients reported are interactions with Asia and Non-Asia regions; N= 87; number of RSOs = 25, (robust) standard errors in parentheses. The scope and impact of regional security organizations 139 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from Figure 4 Conflict and regional security organizations, comparison of Asian and non-Asian RSOs marginal effects. 140 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from of conflict indicating the persistence of certain conflicts (e.g. Kashmir and Afghanistan) during the post-Cold War period. 7 Conclusions and discussion Somewhat dependent on one’s definitions, there are currently between 20 and 35 regional security organizations active. A significant number of these were formed (or fundamentally reorganized) in the immediate aftermath of the Cold War, but others have been active for most of the post-Second World War period and have been recognized as a significant contributors to global security governance – as shown for example in UN Chapter VIII status. There are, however, very few attempts to use quantitative analysis to study the design and impact of RSOs. Further, the scope of security governance has widened in the post-Cold War period. Political (and scholarly) attention is no longer exclusively focused on state sovereignty and interstate conflict, but security governance encompasses prevention, threats from non-state actors, human rights concerns, and even environmental risks. The ultimate concern is to identify the conditions under which RSOs indeed become relevant institutions able to contribute to regional security. Similarity of political system and capacity across member states increases RSO scope. Further, organizations appear to follow their own organizational dynamics. Interestingly, there is very little evidence that hegemony contributes to the expansion of collective security governance. Notably, the evidence to support the idea that more independent RSOs with a broader agenda contribute significantly to regional peace is very weak. What may explain these, somewhat contradictory, findings? First of all, some general doubt exists regarding the significance of regional security organizations in terms of their contribution to regional security. The effectiveness of the EU – clearly politically the most developed RSO–or NATO– indubitably with the most military power – is often questioned. Consequently, it may be heroic to expect strong results for such weakly institutionalized RSOs as the Mano River Union. Secondly, although we include a relatively large number of RSOs with considerable variation in mandate and scope, some (inevitable) selection bias remains. Some conflict-prone regions, such as South Asia and East Africa, are underrepresented. Thirdly, the current research design focuses on intraregional conflict (as defined by RSO membership) and thus excludes the The scope and impact of regional security organizations 141 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from possibility that RSOs may promote security in the wider neighborhood. Afinal,andpossiblymostintriguing,possibilityisthatexistingstudiesonRSOs tend to overestimate their impact because they link the security agenda of the RSO directly to regional (in)security. In fact, there may well be some common factors underlying both variables: regions with states that are willing to support regional integration (or cooperation) are also more likely to be peaceful. Our study provides some evidence for this conjecture. The more detailed comparison of regional security organizations based in and outside Asia addresses some of these issues. Despite the relative longevity of the ASEAN secretariat, no significant delegation of powers to either ASEAN or the ARF in the provision of security has occurred. This is often attributed to the fundamental principle of‘no outside interference’ guiding security cooperation in Asia. However, there are indications that scope expansion has taken place and is continuing. It is argued that ASEAN economic motivations, once clearly central, have now become secondary to conflict management and resolution (Amer, 1999; Forbes, 2003; Acharya, 2009). Moreover, it is suggested that the Treaty of Amity and Cooperation in Southeast Asia (which China joined in 2003) has contributed to ‘confidence among states that were previously in conflict with one another, thereby generating a sense of community’ (and a consequent willingness to cooperate) (Ravenhill, 2009, p. 220). Our analysis indeed suggests that the limited scope of Asian RSOs is not exceptional; to the contrary, most variables that are useful to explain RSO scope outside Asia also apply to Asia. There is one important exception: our analysis suggests that the legacy of conflict in the region still limits the scope of activities taken up by Asian RSOs. A pivotal concern of Asian regional security structure seems to be to entice China away from balance of power practices and to adopt communitarian practices (Adler and Greve, 2009, p. 76; Goh, 2007/2008, p. 139). This type of reason implies the possibility of placing China in a ‘golden cage’, or expecting China to develop a form of what Ikenberry calls‘institutional self-binding’ (Ikenberry, 2001, p. 238), whereby a greater power adopts a structure of restraint towards its weaker neighbors in exchange for the latter’s collective recognition of its own economic and security interests and leadership. Jones et al. (2013) deviate from this view, arguing that the balance of power remains critical for South East Asia security relations with the United States, China, Japan, and North Korea. This view chimes also with the suggestion that China sees US policies of 142 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from engagement through the ARF as a form of ‘soft containment’ (Acharya, 2009, p. 217). Acharya (2009, p. 214) further argues that ‘(m)ultilateralism in the manner of the ARF mutes and moderates balancing tendencies and helps, pending the transformation of balancing behaviour, preservation of regional stability.’ Our analysis largely agrees with Acharya’s assessment that the prospects for development of regional security organizations in Asia are rather modest. However, this is not necessarily indicative of an Asian exceptionalism; rather it reflects the diversity of political structures and capacities across Asia as well as the legacy of regional conflicts. Acknowledgements We thank Erik Gartzke, Ben Goldsmith, Koji Kagotani, Barry O’Neill, Arthur Stein, Atsushi Tago, Jeff Weber, and other participants of the East Asian Security Workshop, 18–19 April 2013, for their helpful comments. Funding The East Asian Security Workshop was funded by the Suntory Foundation. References Acharya, A. (2007) The Quest for Identity: International Relations in Southeast Asia. Singapore: University of Oxford Press. Acharya, A. (2009) Constructing a Security Community in Southeast Asia: ASEAN and the Problem of Regional Order. Abingdon: Routledge. Acharya, A. and Johnston, I.A. (eds) (2007) Crafting Cooperation: Regional International Institutions in Comparative Perspective. Cambridge: Cambridge University Press. Adler, E. and Barnett, M.N. (eds) (1998) Security Communities in Comparative Perspective. Cambridge: Cambridge University Press. Adler, E. and Greve, P. (2009)‘When security community meets balance of power: overlapping regional mechanisms of security governance’, Review of International Studies, 35, 59–84. Amer, R. (1999) ‘Conflict management and constructive engagement in ASEAN’s expansion’,Third World Quarterly, 20, 449–461. Bearce, D.H. (2003) ‘Grasping the commercial institutional peace’, International Studies Quarterly, 47, 347–370. Bearce, D.H. and Omori, S. (2005) ‘How do commercial institutions promote peace?’, Journal of Peace Research, 42, 659–678. The scope and impact of regional security organizations 143 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from Boehmer, C., Gatzke, E. and Nordstrom, T. (2004) ‘Do intergovernmental organizations promote peace?’,World Politics, 57, 1–38. Buzan, B. and Wæver, O. (2003) Regions and Powers: The Structure of International Security. Cambridge: Cambridge University Press. Chang-Fawn, R. (2009) ‘Regions and their study: wherefrom, what for and whereto?’, Review of International Studies, 35, 5–34. Colaresi, M.P., Rasler, K. and Thompson, W.R. (2007) Strategic Rivalries in World Politics: Position, Space and Conflict Escalation. Cambridge: Cambridge University Press. Deutch, K.W., Burrell, S.A., Kann, R.A., Lee, M., Jr, Lichterman, M., Lindgren, R.E., Loewenheim, F.L. and van Wagenen, R.W. (1957) Political Communities in the North Atlantic Area. Princeton, N.J: Princeton University Press. Donno, D. (2010) ‘Who is punished? Regional intergovernmental organizations and the enforcement of democratic norms’, International Organization, 64(3), 593–625. Dorussen, H., Kirchner, E.J. and Sperling, J. (2009)‘Sharing the burden of collective security in the European Union’,International Organization, 63, 789–810. Dorussen, H. and Ward, H. (2008) ‘Intergovernmental organizations and the Kantian peace’, Journal of Conflict Resolution, 52, 189–212. Dorussen, H. and Ward, H. (2010) ‘Trade networks and the Kantian peace’, Journal of Peace Research, 47(1), 29–42. Fearon, J.D. and Laitin, D.D. (2003) ‘Ethnicity, insurgency, and civil war’, American Political Science Review, 97, 75–90. Forbes, V.L. (2003) ‘Geopolitical change: direction and continuing issues’, in Chia Lin Sien (eds.), Southeast Asia Transformed: A Geography of Change. Singapore: Institute of Southeast Asia Studies. Garofano, J. (1999) ‘Flexibility or irrelevance: ways forward for the ARF.’, Contemporary Southeast Asia, 21(1), 74–94. Gartzke, E. (2007)‘The capitalist peace’, American Journal of Political Science, 51 (1), 166–191. Goh, E. (2007/2008) ‘Great powers and hierarchical order in Southeast Asia: analyzing regional security strategies’,International Security, 32, 113–157. Gowa, J. (1989) ‘Rational hegemons, excludable goods, and small groups: an epitaph for hegemonic stability theory?’,World Politics, 41(3), 307–324. Haas, E.B. (1964) Beyond the Nation-State. Stanford, CA: Stanford University Press. Haftel, Y.Z. (2007) ‘Designing for peace: regional integration arrangements, institutional variation, and militarized inter-state disputes’, International Organization, 61, 217–237. 144 Han Dorussen and Emil J. Kirchner at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from Haftel, Y.Z. (2012) Regional Economic Institutions and Conflict Mitigation: Design, Implementation, and the Promise of Peace. Ann Arbor, MI: University of Michigan Press. Haftel, Y.Z. (2013) Regional economic organizations and security institutions in a comparative perspective, paper presented at the International Studies Annual Convention, San Francisco, 3–6 April 2013. Hegre, H. and Sambanis, N. (2006) ‘Sensitivity analysis of empirical results on civil war onset’, Journal of Conflict Resolution, 50(4), 508–535. Heldt, B. and Wallensteen, P. (2007) Peacekeeping Operations: Global Patterns of Intervention and Success, 1948–2004. Sandöverken, Sweden: Folke Bernadotte Academy Publications. Hemmer, C. and Katzenstein, P. (2002) ‘Why is there no NATO in Asia? Collective identity, regionalism, and the origins of multilateralism’, International Organization, 56, 575–607. Heston, A., Summers, R. and Aten, B. (2012) Penn world table version 7.0, Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania. Hirschleifer, J. (1983) ‘From weakest-link to best shot: The voluntary provision of public goods’,Public Choice, 41(3), 371–386. Huntington, S.P. (1991) The Third Wave: Democratization in the Late Twentieth Century. Norman, OK: University of Oklahoma Press. Ikenberry, G.J. (2001) After Victory: Institutions, Strategic Restraint, and the Rebuilding Order after Major Wars. Princeton: Princeton University Press. Jácome, F., Milet, P. and Serbin, A. (2005) Conflict prevention, civil society and international organizations: the difficult path for peace building in Latin America and the Caribbean, paper prepared for FOCAL, January 2005. Jones, D.M., Khoo, N. and Smith, M. (2013) Asian Security and the Rise of China: International Relations in an Age of Volatility. Cheltenham: Edward Elgar. Katzenstein, P.J. (2005) A World of Regions: Asia and Europe in the American Imperium. Ithaca, NY: Cornell University Press. Keohane, R.O. (1984) After Hegemony: Cooperation and Discord in the World Political Economy. Princeton, NJ: Princeton University Press. Kindleberger, C.P. (1973) The World in Depression, 1929–1939. Berkeley, CA: University of California Press. Kirchner, E.J. and Dominguez, R. (eds.) (2011) The Security Governance of Regional Organizations. London and New York: Routledge. Kirchner, E.J. and Sperling, J. (2007) EU Security Governance. Manchester: Manchester University Press. Koremenos, B., Lipson, B. and Snidal, D. (2001) ‘The rational design of international institutions’,International Organization, 55, 761–799. The scope and impact of regional security organizations 145 at Columbia University Health Sciences Library on January 30, 2014 http://irap.oxfordjournals.org/ Downloaded from Krasner, S.D. (1976) ‘State power and the structure of international trade’, World Politics, 28(3), 317–347. Lanteigne, M. (2005) China and International Institutions: Alternative Paths to Global Power. London: Routledge. Mansfield, E.D. and Pevehouse, J.C. (2008) ‘Democratization and the varieties of international organizations’, Journal of Conflict Resolution, 52(2) 269–294. Marshall, M.G., Jaggers, K. and Gurr, T.R. (2012) Polity IV project: political regime characteristics and transitions, 1800–2010. Dataset Users’ Manual, available at http://www.systemicpeace.org/polity/polity4.htm (20 November 2013, date last accessed). Mattli, W. (1999) The Logic of Regional Integration: Europe and Beyond. Cambridge: Cambridge University Press. McKeown, T.J. (1983) ‘Hegemonic stability theory and 19th century tariff levels in Europe’,International Organization, 37(1), 73–91. Olson, M. (1965) The Logic of Collective Action. Public Goods and the Theory of Groups. Cambridge, MA: Harvard University Press. Park, C.-G. (2006) ‘Japan’s policy stance on East Asian neo-regionalism: from being a reluctant to becoming a proactive state’, Global Economic Review, 35, 285–301. Pevehouse, J.C. and Russett, B. (2006) ‘Democratic international governmental organizations promote peace’,International Organization, 60(4), 969–1000. Pevehouse, J.C., Nordstrom, T. and Warnke, K. (2004) ‘The Correlates of War 2 international governmental organizations data version 2.0’, Conflict Management and Peace Science, 21(2), 101–119. Ravenhill, J. (2009) ‘East Asian regionalism: much ado about nothing?’, Review of International Studies, 35, 215–236. Russett, B. and Oneal, J. (2001)Triangulating Peace: Democracy, Interdependence, and International Organizations. New York: Norton. Sandler, T. (1992) Collective Action: Theory and Application. Ann Arbor, MI: University of Michigan Press. Tavares, R. (2009) Regional Security: The Capacity of International Organizations. London and New York: Routledge. Webber, D. (2010) ‘Two funerals and a wedding? The ups and downs of regionalism in East Asia and Asia Pacific after the Asian crisis?’, The Pacific Rim Review, 14, 339–372. Webber, D. (2013) ‘Regionalism and EU-Asia Relations: patterns, trends and determinants’, in T. Christiansen, E.J. Kirchner and P. Murray (eds.), The Palgrave Handbook of EU-Asia Relations. Basingstoke: Palgrave.