European Journal of International Relations http://ejt.sagepub.com/
Rule tensions and the dynamics of institutional change: From 'to the victor go the spoils' to the Stimson Doctrine Joseph O'Mahoney European Journal of International Relations published online 24 July 2013 DOI: 10.1177/1354066113483781 The online version of this article can be found at: http://ejt.sagepub.com/content/early/2013/07/17/1354066113483781
Published by: http://www.sagepublications.com
On behalf of:
Standing Group on International Relations of the ECPR
Additional services and information for European Journal of International Relations can be found at: Email Alerts: http://ejt.sagepub.com/cgi/alerts Subscriptions: http://ejt.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Jul 24, 2013 What is This?
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
483781
EJT0010.1177/1354066113483781European Journal of International RelationsO’Mahoney
EJIR
Article
Rule tensions and the dynamics of institutional change: From ‘to the victor go the spoils’ to the Stimson Doctrine
European Journal of International Relations 0(0) 1–24 © The Author(s) 2013 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1354066113483781 ejt.sagepub.com
Joseph O’Mahoney Brown University, USA
Abstract When and under what conditions do norms and rules change? Dominant conceptions of institutional change in International Relations theory are based on the idea that it is the result of a shift in power: new actors become able to impose their vision on the world. However, the source of change need not be the power or preferences of actors in society, but could come from the internal dynamics of the rule system governing these actors. This article develops recent research in this area by linking Sandholtz’s model of norm change to recent dynamic institutionalist work and exploring and specifying particular mechanisms, or types of tensions, in rule systems that produce change. Institutions and complexes of rules exhibit rule tensions: inconsistencies, ambiguities, and inadequacies that can lead to disputes over the application of the rules. Actors then have to solve problems or disputes over rule interpretation. Change can thus occur without the introduction of new actors or a shift in the power of existing actors. I apply these ideas to a significant change in the rules of war in the early 20th century: the shift from the rule ‘to the victor go the spoils’ to the Stimson Doctrine, or the rule that states should not profit from aggression. Keywords Discourse, institutional change, institutions, Jus ad bellum, norms, rules
Introduction When and under what conditions do norms and rules change? Some rules of international behavior persist over time while others mutate or are abandoned altogether. The Corresponding author: Joseph O’Mahoney, Department of Political Science, Brown University, Providence, RI 02912, USA. Email:
[email protected]
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
2
European Journal of International Relations 0(0)
dominant attitude in International Relations (IR) theory towards rule change is that it is the result of a shift in power: new actors become able to impose their vision on the world. However, scholars have recently begun to explore other possibilities. One model of norm change holds that the source of change need not be the power or preferences of actors in society, but instead comes from the internal dynamics of the rule system governing these actors. This article develops this model by linking it to recent dynamic or discursive institutionalist work and specifying particular mechanisms, or types of tensions, in rule systems that produce change. I apply these ideas to a significant change in the rules of war in the early 20th century: the shift from the rule ‘to the victor go the spoils’ to the rule that states’ gains from aggression will not be recognized, also known as the Stimson Doctrine.1 In the 19th century, concessions imposed on a defeated state were considered to be valid by the international community. In 1871, the victorious German Empire demanded, among other concessions, most of Alsace-Lorraine and five billion francs from France. Other powers treated these concessions as a normal and acceptable part of war-making. Now, however, states accept the principle that they cannot legally obtain territory or other advantages through the use or threat of force. There is now a rule of nonrecognition of aggressive gain. Jus ad bellum rules had been developing before and during the 19th century and changed substantially after World War I. However, the turning point of this particular change did not come at the end of either of the world wars. Instead, it was the product of a contingent response by US Secretary of State Henry Stimson and the League of Nations to a dispute over Japan’s actions in Manchuria in the early 1930s. Conventional IR theories, such as much existing realist or rationalist scholarship (e.g. Hawkins et al., 2006; Ikenberry, 2001; Koremenos et al., 2001), would have difficulty explaining changes that occur in the absence of a shift in power. Likewise, even most constructivist or ideational work on norm change (e.g. Widmaier et al., 2007) would struggle to explain the timing and form of much incremental rule change that occurs outside of large-scale crises. Building on Sandholtz’s model of norm change (2008), I argue that the concept of rule tensions is necessary to understand much of the rule change that we see in the international system. The international system, or society, is not just made up of states with preferences. There are also sets of rules and norms governing the behavior of those states. These intersubjective agreements about appropriate ways to behave have internal tensions, contradictions, and weaknesses that may interact with actual state behavior and create political space for disputes over their interpretation and application. In an attempt to resolve these disputes, states creatively put existing rules to new purposes, combine existing principles and practices in new ways, or even create new rules to deal with these new problems. Some change thus occurs not as a direct result of a shift in power or an overall crisis in legitimacy. While states or their representatives do the arguing and resolving of those disputes over the application and interpretation of the rules, the opportunity for these disputes to arise and for their resolution to make sense comes from the dynamics of the rule system. In this article, I first explain the concept of rule tension, how it relies on a dynamic conception of institutions, and how it succeeds in explaining cases of change where power-shift explanations are unsatisfactory. I contribute to existing work in this literature
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
3
O’Mahoney
by formally specifying precise types of tension: ambiguity, inconsistency, and inadequacy. These are mechanisms that potentially create disputes and opportunities for change. Second, I ask why the rules changed from ‘to the victor go the spoils’ to nonrecognition of aggressive gain. I explore the creation of the Stimson Doctrine in the Manchurian Crisis and examine its origins in the complex of international rules dealing with war and aggression during the early interwar period. This demonstrates how formal specification of the rule tensions can improve empirical work. Third and finally, I discuss ways of generalizing from the Stimson Doctrine case and the implications it has for theories of institutional change in IR.
Theorizing institutions and institutional change Rules, norms, and institutions The literature on rules, norms, and institutions is bedeviled by inconsistent terminology.2 Ordinary use of the term ‘international institution’ suggests an organization peopled by a staff, like the United Nations or the World Bank. However, institutions do not have to be organizations. A useful starting definition in IR theory is that an institution is a bundle of rules and practices that govern relations between occupants of recognized roles (March and Olsen, 2006). But the nature of the rules that make up an institution is where the primary conceptual disagreement lies. The core property of the phenomena referred to as rules, norms, institutions, and laws is a set of beliefs. Bicchieri defines a norm in terms of ‘a sufficient number of people believing that it exists and pertains to a given type of situation, and expecting that enough other people are following it in those kinds of situations’ (2006: 2). Her precise formal definition begins with a necessary condition for a social norm to exist. This condition says that for each individual i in a population, i knows that a rule R exists and applies to situations of type S (Bicchieri, 2006: 11).3 So, i has to both believe that a rule exists and believe that others believe that it exists. This collective belief in or awareness of a rule is constitutive of the existence of the rule. This minimal property leaves open the questions of whether people conform to the rule or hold the rule to be legitimate, or the extent to which a rule is instantiated or the nature of its instantiations. One particular implication of this definition is that a rule does not have to be written down in order to exist. Everyday social norms and rules, such as those governing queueing or gift-giving, are part of unwritten, informal institutions. There are many informal international institutions, in addition to those formalized in treaties and other instruments of formal international law. International law is not synonymous with the norms of international behavior. International law constitutes a realm conceptually and practically separate from other shared understandings between states, or that guide the decisions of state representatives. Brunnee and Toope argue that international law should be conceived of as specific legal norms, or criteria of legality, such as ‘generality, promulgation, non-retroactivity, clarity, non-contradiction, not asking the impossible, constancy, and congruence between rules and official action’ (2010: 6). Without these criteria, shared understandings between states are non-legal. In this article, I use the term ‘institution’ to mean a bundle of (one or more) rules, the term ‘norm’ to mean a subtype of rule that has normative content, and reserve ‘law’ for legal rules.
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
4
European Journal of International Relations 0(0)
If rules are constituted by beliefs and beliefs about others’ beliefs, what form do these beliefs take? Crawford and Ostrom (1995), in pursuit of a common terminological standard, propose that regardless of how institutional statements are expressed in natural language, all institutions be defined according to a common grammar. This grammar includes an attribute, a deontic statement, an aim, a condition, and an ‘or else’ or sanction. Deontic statements include the words ‘must’ or ‘should’ or ‘may’ and evoke a feeling of ought-ness. Aims are the thing that the actors with the specified attribute (like ‘all states’) should, must, or can do.4 Using this grammar, we can precisely define the change under consideration. The rule of ‘to the victor go the spoils’ could be defined as: All militarily victorious states may use their new-found dominance to impose political changes at the expense of the defeated state(s).
In contrast, the rule that states should not profit from aggression can be defined as: All militarily victorious states must not use their new-found dominance to impose political changes at the expense of the defeated state(s), or else the gains resulting from that use will not be recognized.
The rule here is not one that is written down in precisely this form. This brings us to the issue of instantiation. How do we see these institutions in the world? One manifestation of an institution is written statements of the rules and practices that are consciously and officially agreed to. Good international examples are treaties, agreements, covenants, and resolutions. However, these do not constitute the entirety of institutions. Many aspects of institutions are informal rules and norms. These can be seen when actors make reference to them and orient their action around them. An informal institutional rule exists to the extent that actors refer to a rule when considering action, when justifying or legitimating action, and when interpreting action, for example, as being allowed or forbidden by a rule. Finally, there are the practices that are part of the institution and that also instantiate the rules. These include actions prescribed or proscribed by the rules, as well as sanctions against violators. In what way are the principles of ‘to the victor go the spoils’ and ‘nonrecognition of aggressive gain’ rules of international behavior? How are they part of an institution? There was no explicit written document to which all states had formally acquiesced, the most intuitive sense of a rule, laying out the rights of the winners of wars. However, prior to the 20th century, not only would people act as if conquest brought rights to the conqueror, but they would actively justify their actions and various outcomes in terms of the rightness and legitimacy of a victor disposing of the fruits of victory (Korman, 1996). After World War I, the League of Nations Covenant was drawn up as a means to try to limit the outbreak of war. However, the Covenant did not include the idea of nonrecognition of the results of aggression. The Kellogg–Briand Pact in 1928 extended the promise of all signatories not to use force to resolve disputes between them, but also did not include nonrecognition of the results of force as a sanction against violators of the pact. After the Manchurian Crisis and the nonrecognition of Manchukuo, several treaties and declarations explicitly included the idea that nonrecognition was a sanction against
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
5
O’Mahoney
aggression.5 Further, the acceptance or nonrecognition of situations resulting from conquest or invasion became a major part of the discourse surrounding the use of force. For example, one of the crucial questions regarding Mussolini’s conquest of Ethiopia in 1935 was whether it would be accorded de jure recognition. How should we analyze institutions? Schmidt (2008) argues that the standard three new institutionalisms — rational, historical, and sociological — tend to view institutions as constraints on action. Rational institutionalists usually view an institution as an equilibrium, that is, as behavior (e.g. Calvert, 1995). Katznelson and Weingast (2005) argue that there are many points of intersection between historical and rational-choice institutionalism. Some of these points of intersection are that institutions are treated as static, institutions are treated as constraints on action, and change occurs only through changes in the bargaining resources of actors. Historical institutionalist work in IR has followed this pattern (see Fioretos, 2011). In contrast to this view, some recent work in IR has turned to a more dynamic view of institutions (e.g. Hay, 2006; Krook and True, 2012; Sandholtz, 2008; Van Kersbergen and Verbeek, 2007). What Schmidt calls ‘discursive institutionalism’ starts from the idea that institutions are the context of meaning in which actors think, speak, and act and that are the result of those thoughts, words, and actions (2008: 314). Action in institutions is not just rule-following, but is the process whereby institutions are created and maintained. Krook and True call this view ‘norms as processes.’ One implication of this view is that instead of being fixed, norms have an ‘ongoing potential for contestation’ (Krook and True, 2012: 109). A growing body of work in political science is adopting this approach to institutions. Actors use their background ideational abilities to make sense of the meaning context of the institution and then use their foreground discursive abilities to change or maintain the institution they are in (Schmidt, 2008: 314). One important point of difference with the historical-rationalist conception is that there is no assumption of equilibrium. Instead, actors are working inside an institutional setting that never reaches equilibrium and instead is continually and cumulatively reinterpreted (Carstensen, 2011: 160).
Rule tensions and institutional change In this section, I outline a mechanism of change: the resolution of disputes arising from rule tensions. Later, I contrast this with the mechanism of change underlying dominant approaches in the field: rule manipulation by newly powerful actors. A system-centered approach to rule change, rather than starting with actors and their pursuit of gain, starts with features of the rules and institutions themselves. Institutions and complexes of rules exhibit rule tensions that can lead to disputes over the application of the rules. Pragmatic actors have to solve problems or disputes over rule interpretation and answer to a variety of constituencies. These pragmatic actors combine elements of ideas, principles, and practices in order to resolve the disputes and also to make policy options acceptable to those various constituencies, that is, to legitimate one policy over another. The combination or recombination of elements is constrained by what already exists, in that some solutions are more immediately and easily conceivable. Combinations are also constrained by what the constituencies will find acceptable. A change in the rules comes about either through novel combinations of existing ideas or the introduction of a new
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
6
European Journal of International Relations 0(0)
idea or practice into one of those combinations. Change can thus occur without the introduction of new actors or a shift in the power of existing actors. One key feature of social and legal rules that necessitates interpretation is vagueness or imprecision; rules have an ‘open texture.’ The extent of this imprecision can vary but cannot be eliminated (Hart, 1994: 123). Law refers primarily to classes of persons and classes of acts, things, and circumstances and as such there are ‘uncertainties as to the form of behavior required by them … in particular concrete cases’ (Hart, 1994: 124). Farrell (2009) points out that institutions, as social rules, have similar properties. This abstract, ambiguous nature of rules leaves open the possibility that the material situation is not precisely covered by the rules, or that there are multiple conflicting rules. This means not only that it is difficult to predict how other actors will react to the material situation, but that actors can exploit these sites of ambiguity. Mahoney and Thelen argue that ‘institutional change often occurs precisely when problems of rule interpretation and enforcement open up space for actors to implement existing rules in new ways’ (2010: 4). However, they do not incorporate this insight into their model of institutional change. Sandholtz’s (2008) cycle model of norm change is based on the tensions inherent in normative systems that lead to disputes about the interpretation and application of the rules. For Sandholtz, disputes arise because there are conflicts between the generality of rules and the specificity of concrete experience, or conflicts between different sets of rules. Actions can generate disputes over which norms apply to the action and what the relevant norms require or permit. Sandholtz emphasizes the way in which arguments modify international norms. A process of argumentation takes place, during and after which norms under dispute are altered. Such alteration could be strengthening, weakening, change in substantive content, or change in an other dimension. The key mechanisms of change here are embedded in the argumentation process. He draws on theories of legal reasoning and proposes that arguments are successful due to consistency, analogy, and precedent. Successful arguments change norms. Sandholtz and Stone Sweet make the collective contestation over the interpretation of existing rules the centerpiece of their theory of international normative change: rule structures are at the heart of any dispute that might interest social scientists, for two reasons. First, the inevitable gap between general rules and specific actions means that the application of rules is always subject to interpretation and contestation. Second, because no complex rule system provides comprehensive solutions to conflicts among all of its constituent components, tensions and contradictions among norms are also commonplace, and likewise fuel debate. When normative disputes of these kinds arise, actors bring to bear both normative and material powers, and thus the distinction between law and politics vanishes. (2004: 242)
Relatedly, Diehl et al. (2003) view international law as comprising both an operating system, which provides mechanisms for cross-border interactions, and a normative system, which shapes the values and goals of those interactions. They argue that these two systems are often not parallel and sometimes develop autonomously of each other, with one lagging behind the other. When the normative system moves ahead of the operating system, change in the operating system is driven by its insufficiency, its incompatibility with the normative system, or the ineffectiveness of the existing rules.
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
7
O’Mahoney
So, for Diehl et al., properties of the complex of rule systems generate change. A rule tension view of change expects changes to occur when there is some dispute over the application of the rules. A Sandholtzian framework for norm change is illuminating and fruitful. However, there is room for more precision in the mid-level concepts used to delineate types of tension in the rules. Sandholtz identifies two varieties of tension: incompleteness and internal contradictions. However, he does not formally specify what is meant by these terms. When only using these highly general definitions of rule tension, it is hard to know exactly what to look for empirically. It also leaves open what the resolution of the dispute might look like. If we have more precise specifications of tension, we can narrow down the range of arguments that might be successful in resolving the dispute. When scholars have applied Sandholtz’s framework (Sandholtz and Stiles, 2009), they have identified a wide variety of tensions and a wide variety of responses to those tensions. While this is a good demonstration of the applicability and utility of Sandholtz’s framework, such variety lacks precision. When describing the outcomes, the changes in the rules, Sandholtz is intentionally expansive. Change does not simply mean that rule A is replaced with rule B, or a change in content. ‘A norm may become weaker or stronger, more clear or less, more specific or less, more qualified by exceptions or less’ (Sandholtz, 2008: 110). This is consistent with the latest work in the historical institutionalist literature on different types of gradual change (Mahoney and Thelen, 2010). However, more precision at delineating types of tension can provide a guide to the types of outcome we might expect to see. If a particular tension generated the dispute, we should see a change oriented around resolving that dispute.
Stasis vs dynamism How can we reconcile something as seemingly static as a constant tension in the rules, a tension that is the same at time t1 as at time t2, with a dynamic view of institutions? The key is in the way that tensions lead to disputes. Rather than some sort of mechanical idea that tensions automatically create disputes, it is the uncertainty surrounding how to interpret or understand action in terms of rules that means that disputes are possible. If acts were self-evidently defined, allowed, or forbidden, then disputes about rule application would be impossible. Instead, actors have the option (whether chosen consciously or not) to draw attention to the particular rule tension and push for a resolution of that tension. Given the amount of latent tensions in any rule system (as Hart (1994) and Franck (1988), make clear), there are rule tensions that could have been but are not exploited all the time. If this is true, if rule tensions do not always lead to rule change, in what sense do they cause change? Rule tensions allow actors to dispute the collective interpretation of action. They are a permissive condition. How do they allow disputes? A rule tension means that there are at least two arguments that people might find persuasive. Imagine that a rule says that all acts of type A are forbidden. An action, α1, is performed. A dispute is possible if someone decides to exploit the fact that α1 could be an act of type A and thus forbidden. The only way that this matters is if some people care that α1 might be forbidden by the rules and that there is some uncertainty whether it is.6 That is, the argument that α1 is of type A has some force. However, this is still not sufficient for a dispute. Only
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
8
European Journal of International Relations 0(0)
if someone else decides to contest the claim that α1 is of type A, presumably in this case the actor who performed α1, is there a dispute. Again, if this is not true, if, say, everyone in a community unproblematically views an act as a clear instance of A, then there is no room for a dispute. But a dispute only occurs if an actor decides to exploit the tension. As Sandholtz points out, the issue is not ‘abstract arguments,’ but ‘practical disputes arising out of specific actions’ (2008: 102). The rule tension view of change is thus only consistent with a dynamic view of institutions and rules as contested, reinterpretable, resources to be deployed. Rule tensions cannot work in this way if rules and norms are either behavioral equilibria or fixed and coherent ideas. Once a dispute occurs, what exactly does it mean to ‘resolve a dispute’? Resolution does not mean that the tension leading to the dispute has been settled beyond the prospect of future contestation. The very nature of rules as involving relations between classes of acts and classes of circumstances, as well as the multiplicity of rules, means that rule tensions are an inevitable feature of rule systems. The resolution of a dispute over rule application means that a certain interpretation of a course of action in the current situation is accepted, or at least acquiesced to, by other states. Of course, it is possible that a dispute is not resolved and instead there is continuing contestation over the application of the rules, or states might even resort to force. Another question is why some disputes result in rule changes and others do not. Sandholtz argues that all disputes result in some change (although not necessarily in the formal written instantiations of the rules) and the issue instead is the scope and nature of change (2008: 110). Part of the contribution of specifying precise types of rule tension is that it is a step towards a better understanding of why particular rule changes occur.
Ambiguity, inconsistency, inadequacy I now formalize three distinct types of rule tension that can lead to disputes over the application and interpretation of the rules. These are not exhaustive of all the different types of tension, but they do cover some important classes of tension. Rule ambiguity is when an existing rule covers a set of acts or situations and does not distinguish between subsets of those acts or situations. This is a necessary aspect of all rule systems as rules by nature cover general classes of acts, actors, or situations. An abstract example of a rule is the following: R1: All acts of type A are forbidden except in extenuating circumstance C. There are inevitably going to be some possible actions that could be interpreted as instances of A but that are not clear-cut or canonical instances. And the circumstances surrounding actions of type A may be related to C but again not uncontroversially so. In domestic law, all instances of murder are forbidden. Some types of killing are excused, such as those in self-defense or those done under the heat and passion of the moment. But whether a particular killing is in self-defense, and hence allowed rather than forbidden, can often be argued either way. Social rules and norms have the same property of ambiguity. If rule ambiguity generates a dispute, one outcome we might expect is that a resolution of the dispute changes the precision of the rule. This may involve more or less
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
9
O’Mahoney
precision, but because the dispute was made possible by ambiguity, a resolution of the dispute should be oriented around that tension. Rule inconsistency is a situation arising when two rules conflict with each other. An act may be allowed under one set of rules, but forbidden under another. Imagine that in addition to R1 above, we have: R2: Actors of type B have the right to perform acts of type D. Further suppose that the set of acts D intersects or overlaps with the set of acts A. If an actor performs an act that could be construed as a member of the set D ∩ A, then there is a conflict over whether the act is allowed or forbidden. This sort of conflict would trigger a dispute over the status of the act. These sorts of conflicts occur frequently in domestic politics. An example is the conflict between the injunction on unjust discrimination and the right to free association. Setting limits on who may be admitted to a club may fall under both rules. A prominent example from international relations is the conflict between sovereignty and the responsibility to protect. If rule inconsistency generates a dispute over whether an action falls under one or another rule system, then we should expect the outcome to involve specifying which system governs actions of that type. So, if the dispute is over whether an act is of type A or type D, we should expect the outcome to be oriented around the classification of that act as either A or D or some other type. The concept of rule inadequacy concerns the ineffectiveness of sanctions, which may include the absence of a sanction. The rules may not only be ambiguous over which acts are covered or the steps to be taken in the event of a violation, they also may not stipulate what is to be done. For example, in a rule system including R1 above, what is to be done if an actor does perform an act of type A and not in circumstance C? R1 does not stipulate what the sanction is to be. This inadequacy could trigger a dispute over what exactly the sanction should be. Even if there is a sanction specified, it may be ineffective or unsuited to violations of the rule (Diehl et al., 2003: 56). Take the following rule: R3: All acts of type A are forbidden except in extenuating circumstance C. If an actor performs an act of type A then sanction S will be applied. Now at least there is a specified sanction. However, sanction S may be ineffective at securing compliance, or may be too costly or inconvenient to administer. If rule inadequacy is the case, an act of type A may trigger a dispute over the nature of the sanction to be applied. If so, we should expect the outcome of the dispute to involve either the provision of a sanction or the re-evaluation of an existing sanction. An explanation for the emergence of the rule of nonrecognition of aggressive gain based on a process of the resolution of disputes about the application of the rule is:
Rule tension explanation Nonrecognition became part of the rules of international behavior because actors resolved disputes generated by the ambiguity, inconsistency, or inadequacy of the rules.
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
10
European Journal of International Relations 0(0)
If this is the process by which nonrecognition arose, the first thing we should see is an identifiable tension in the rules governing the management of aggression and peacemaking. Then we should see a dispute break out over the application or interpretation of those rules. Actors should explicitly identify some uncertainty about the application of the rules when discussing how to react to the dispute.
Alternative explanation: Power-shifts and change The primary purpose of the case study is to demonstrate how the precision of the rule tension types is helpful in guiding empirical work. However, it is useful to have an explicit comparison with an alternative explanation. This is especially true when there is a dominant type of explanation. In this case, power-shifts are the baseline explanation for institutional change. A common explanation for international arrangements of rules is that they are constructed in the interests of a dominant state. Hegemonic imposition relies upon the clear decision of a strong state to impose an institutional form on the international community. Gilpin argues that ‘the principal mechanism of change throughout history has been war or ... hegemonic war’ (1983: 15). Ikenberry (2001) explains the form of post-war orders as a function of the ability of the dominant state to establish restraints on their power and to credibly commit not to use their power to exploit the other states in the system. From a more constructivist direction, Clark (2005) argues that change in the legitimacy of international orders is embodied in the peace settlements at the end of major wars. More general than hegemonic imposition, rational design theory is based on the premise that institutions are the result of bargaining between powerful states (Koremenos et al., 2001). There are several common features of mainstream rationalist theories of institutional change. The first is the idea that institutions are designed or chosen from amongst alternatives ‘by a conscious or deliberate process of calculation’ (Wendt, 2001: 1036) in order to achieve some specified goal. Other common features concern the purposes that institutions serve. Rationalist theories of institutions are fundamentally based on the assumption that states create international institutions because they solve coordination problems and achieve efficiency gains from specialization (e.g. Hawkins et al., 2006). The main mechanism for institutional creation is thus that they help newly powerful self-interested states realize material goals. More specifically, states create institutions because they help states to credibly commit to future courses of action. Why does change happen for these theories? It happens because newly powerful actors shape the rules for their benefit. An interest group or coalition is gaining some benefit from the current arrangement, such as rents or a beneficial tax regime. This group supports or maintains the rules because they are clear about how they are currently benefiting and wish to continue benefiting. Any change in the rules, even a gradual one like the addition of a new rule (layering) or the conversion of an existing rule to a different purpose (Mahoney and Thelen, 2010), can only occur because another group newly has the power to effect this change in pursuit of some benefit that they then want to receive:
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
11
O’Mahoney Table 1. Observable implications of alternative explanations.
Permissive condition Change event Justification of need for institutional innovation Justification for form of institutional innovation
Power-shift
Rule tension
Shift in power Choice of powerful actors Actors refer to benefits they will get from new rule over old rule Actors refer to novelty or amount of benefits
Tension in rules Breakout of dispute Actors refer to problem with existing rules Actors refer to previous practices/principles
Power-shift explanation The rules changed because the new rule(s) served distributional purposes for actors who constituted a dominant coalition based on their power position.
The observable implications of this view include: an identifiable group of powerful actors; an identifiable benefit that they are getting from the new institutional arrangement; and, crucially, the shift in power that enabled new actors to renegotiate that arrangement. If this is the mechanism of change, we should see actors planning and discussing the new use of the rules in terms of the benefits that this new use will bring them. Some of this planning will be done in private, away from public view, but the new use of the rules requires coordination and bargaining among the members of the coalition in power that stand to benefit. So, there should be public discussion between members of the powerful coalition coordinating their new use of the rules. There are two possible explanations, or hypotheses, for a change in the rules here. Table 1 presents a summary of the evidence that could in principle decide which is most helpful in understanding the move towards the nonrecognition of aggressive gain. Thus far in this article, I have explained the mechanism of rule tension, specified particular types of rule tension, and stated an alternative explanation. I will now apply these ideas and explore the change in international rules resulting in the rule of nonrecognition of aggressive gain. The primary purpose of this case study is to show how the precision of the formal specification of rule tensions can guide empirical work. It will also provide a clear instance of international institutional change that cannot be adequately explained by power-shift conceptions of change and demonstrate the utility of theories of rule tensions in explaining international change.
The Stimson Doctrine: Nonrecognition of aggressive gain Nonrecognition of aggressive gain is an international practice that became regularized and institutionalized into formal treaties and agreements. There was a change in the bundle of rules, or institution, governing the acceptable use of force. No longer is ‘to the victor go the spoils’ a valid position. Instead, states now frequently do not recognize aggressive gain. High-profile examples include Israel’s territorial gains in the 1967 war,
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
12
European Journal of International Relations 0(0)
the establishment of a Turkish–Cypriot state after Turkey’s invasion of Cyprus in 1974, Morocco’s seizure of the Spanish Sahara in 1975, Indonesia’s annexation of East Timor in 1975, South Africa’s occupation of Namibia from 1966 to 1989, and the Republic of South Ossetia, declared after Russian military victory over Georgia in 2008. The policy or doctrine that I term ‘nonrecognition’ involves the individual or joint refusal by states to recognize, that is, admit the legality or legitimacy of, certain acts or situations. Nonrecognition is not merely an absence of action; it is a highly visible, provocative act. In the same way that saying ‘I promise to do X’ does not describe a promise but is a promise, a declaration of nonrecognition, or recognition, creates what Searle (2010) calls an ‘institutional fact.’ States throughout history have chosen to not recognize various situations for various reasons, some systematic, some relatively idiosyncratic. In this article, I analyze the nonrecognition of gains made by states via the threat or use of aggressive force. Nonrecognition of aggressive gain was not a feature of interstate behavior during the 19th century. Some states adopted it during the 1930s, but after World War II, it became an accepted part of the rules of international behavior and was formalized in international law. Nonrecognition of aggressive gain is a diplomatic sanction against aggression. Sometimes, the object of nonrecognition is territorial gain; sometimes, the creation of a new state; and sometimes, a peace treaty. Nonrecognition can be an explicit declaration or an absence of recognition when recognition is requested or expected. As part of the process of reacting to aggression, states also often make threats or declarations of intent to not recognize some claim. The ‘nonrecognition of aggressive gain’ is referred to throughout simply as ‘nonrecognition.’ This is often referred to in both the international law and diplomatic history literature as the Stimson Doctrine, after US Secretary of State Henry Stimson: [The Stimson Doctrine] has since 1932 acquired the force of customary international and treaty law, through its adoption by the League of Nations and, in the post-World War II period, the Charter and practice of the United Nations. In many of the worst crises that have confronted the international community in recent years, a statement refusing to recognize the legality of any consequences of the aggressor party’s actions has often been one of the first reactions of the UN Security Council. (Turns, 2003: 107)
The Manchurian Crisis and the genesis of nonrecognition of aggressive gain This section analyzes the Manchurian Crisis in light of the theoretical considerations presented above. This analysis will provide answers to a significant part of the puzzle over why the rule of ‘to the victor go the spoils’ was superseded by the rule of ‘nonrecognition of aggressive gain.’ The Manchurian Crisis was a major international diplomatic incident arising from Japan’s use of force in Manchuria in Northern China over the period 1931–1933. It was the setting for the first use of nonrecognition as a sanction against the illegitimate use of force. Numerous commentaries identify the Manchurian Crisis as a seminal event in the development of the laws of war and the genesis of the rule of nonrecognition (Korman, 1996; Langer, 1947; Malawer, 1977).
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
13
O’Mahoney
Historical overview After the Japanese Kwantung army in Manchuria used an explosion on the railway at Mukden (Shenyang) on 18 September 1931 as an excuse for the preconceived seizure of towns in the area, the Nationalist Chinese government brought the issue to the attention of the Council of the League of Nations, and appealed to the United States under the terms of the Kellogg–Briand Pact. A League Council resolution on 30 September 1931 merely noted ‘the importance of the Japanese Government’s statement that it has no territorial designs in Manchuria’ and urged both sides to settle their differences peacefully. However, the Japanese military, effectively autonomous from the civilian government, continued its activities, including large-scale military operations. On 24 October, the Council issued another resolution calling for the Japanese troops to be withdrawn into the railway zone allocated to Japan by treaty. However, the Japanese army persisted. The League appointed a five-man commission, on 10 December, chaired by the British Lord Lytton to investigate who was responsible for the Mukden incident and whether either side was acting in self-defense. On 7 January 1932, US Secretary of State Henry L. Stimson sent a Note by telegram to both the Chinese and Japanese governments, which included a statement that the US ‘does not intend to recognize any situation, treaty or agreement which may be brought about by means contrary to the covenants and obligations of the pact of Paris of August 27 1928’ (US Department of State, 1948: III, 8). Despite this, the Japanese military continued its operations. On 16 February, the League Council appealed to Japan, including stating that: ‘no infringement of the territorial integrity and no change in the political independence of any Member of the League brought about in disregard of this article ought to be valid and effectual by Members of the League’ (League of Nations, 1932: 383–384). The Assembly then passed a resolution on 11 March 1932 which said that it was ‘incumbent upon the Members of the League of Nations not to recognize any situation, treaty or agreement which may be brought about contrary to the Covenant of the League of Nations or to the Pact of Paris’ (League of Nations, 1932). These League actions took the threat of nonrecognition from a unilateral statement by the US to a collective act by the international community. On 9 March, the former Emperor of China, Henry Pu-Yi, declared an independent state of Manchukuo in Manchuria. Japan formally recognized Manchukuo on 15 September, defying the League and the US. The Lytton Commission’s report was discussed by the League Assembly in December 1932, and in February 1933, a League Assembly resolution adopted the first eight chapters of the Lytton Commission report, which insisted on the withdrawal of Japanese troops and Japanese recognition of Chinese sovereignty over Manchuria. The adoption of the Lytton Report thus represented a collective act of nonrecognition. Rather than agree to these terms, Japan formally withdrew from the League on 27 March 1933.
Rule tensions and institutional change The opportunity for Stimson’s institutional innovation and its adoption by the international community as a rule of international behavior came about due to disputes about the
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
14
European Journal of International Relations 0(0)
interpretation and application of the existing rules. Such disputes are brought about by tensions in the systems of rules in which international actors are enmeshed. Ambiguity over the definition of aggression made the application of sanctions less clear-cut. Inconsistency between standards of acceptable conduct in China, Japan’s existing rights in China, and new rules against war and the use of force created uncertainty about which rules were to be applied. Inadequacy of the enforcement provisions of the international peace structure led to a need for innovation. The basic problem was that there was no provision for what to do in the event that a state resorted to war and other sanctions were not enacted. It is important to distinguish the process by which rule tensions produce change via allowing disputes over practical action from another process in which tensions in the rule system are addressed by interpretive debate leading to revision or amendment. In this latter type of process, logical arguments about ambiguity, inconsistency, and inadequacy in the rules might very well prove persuasive to legislators or jurists. In this way, some rules might develop almost purely as a corollary of existing rules. At the international level in the interwar period, the League of Nations held debates and discussions about amending or clarifying the articles of the Covenant without being prompted by exigencies outside the Council Chamber or committee room. Some tensions were indeed addressed in this way. For example, in 1920, the ambiguity of Article 18 of the Covenant was addressed by the Council of the League of Nations. Article 18 stated that ‘Every treaty or international engagement entered into hereafter by any Member of the League’ should be registered with the League and published. In a memorandum of 19 May 1920, the Council addressed what was seen as a vagueness in this article and further stipulated that this ‘comprises not only every formal Treaty of whatsoever character and every International Convention, but also any other International Engagement or Act’ (Ferrell, 1968 [1952]: 55). However, nonrecognition was repeatedly rejected in this type of abstract debate throughout the 1920s. In 1921, a League Committee rejected a proposed amendment to the Covenant stating that all members would ‘consider null and void’ treaties granting spoils to an aggressor (Malawer, 1977: 50). In 1928, one representative argued to the Arbitration and Security Committee that ‘We may, indeed we must, infer from [Article 10 concerning resort to war] that Members of the League cannot recognize as final or admissible a situation resulting from an act incompatible with the provisions of the article itself’ (League of Nations, 1928: 75). However, the Committee rejected this suggestion. Even after the adoption of the Kellogg–Briand Pact, during discussions in 1929 over how to incorporate the Pact into the League Covenant, a proposed amendment to the Covenant explicitly forbidding the registration of treaties ‘imposed by force as a consequence of war’ was considered and rejected (League of Nations, 1930: 353, 360, 363). So, even though nonrecognition of the results of war might seem a logical consequence of a prohibition on war, it was explicitly rejected as part of the peacemaking apparatus up until 1932. Some, notably Quincy Wright (1919, 1935), argued that nonrecognition was implied by the League Covenant. However, John Fischer Williams, in an article titled ‘The new doctrine of “recognition,”’ argued forcefully that prior to Stimson and the League’s actions in 1932, recognition was not at issue in the event of war and its results (Williams, 1932). Similarly, McNair refers to nonrecognition as a ‘new doctrine’ that had been recently ‘discovered’ by Stimson and the League (1933:
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
15
O’Mahoney
66, 74). Stimson himself, while certainly arguing that nonrecognition was consistent with the ‘new viewpoint’ and ‘new covenants’ about war, viewed his Note and the subsequent League resolutions as a novel development of the laws of war and the Pact of Paris in particular (see e.g. Stimson, 1932). It was not until there was the pressing issue of having to resolve a very immediate and practical problem — the establishment of Manchukuo as a result of the successful use of force by the Japanese army — that nonrecognition was seen as a solution. The following sections lay out the nature of the tensions leading to the adoption of nonrecognition. Each section demonstrates how the particular type of tension opened up space for a dispute over application of the rules. Discussion of the alternative explanation follows.
Ambiguity The ambiguity built into the League Covenant (‘resort to war,’ ‘external aggression’) and the Pact of Paris (‘recourse to war’) as to what acts count or do not count as violations of the rules against war or aggression opened up space for institutional innovation. The abstract form of the rule would be: R4: All acts of aggression are forbidden. In the Manchurian Crisis, Japan’s actions were not clearly and unambiguously classified as war or aggression. For example, there was no declaration of war by either Japan or China. The ambiguity over the nature of the situation was so severe that the League dispatched the Lytton Commission ‘to study on the spot and report to the Council on any circumstance which, affecting international relations, threatens to disturb peace between China and Japan, or the good understanding between them, upon which peace depends’ (Council resolution of 10 December 1931). However, the extent of the ambiguity was such that the final report of the Lytton Commission was not able or willing to resolve it definitively. In the section on ‘Principles and Conditions of Settlement’ (Chapter IX), the Lytton Report admits that the issues are ‘exceedingly complicated.’ Part of this complexity is that: This is not a case in which one country has declared war on another country without previously exhausting the opportunities for conciliation provided in the Covenant of the League of Nations. Neither is it a simple case of the violation of the frontier of one country by the armed forces of a neighbouring country, because in Manchuria there are many features without an exact parallel in other parts of the world. (Quoted in Willoughby, 1935: 400)
That said, the report continues and states that: It is a fact that without a declaration of war a large area of what was indisputably Chinese territory has been forcibly seized and occupied by the armed forces of Japan, and has in consequence of this operation been separated from and declared independent of the rest of China. (Quoted in Willoughby, 1935: 400)
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
16
European Journal of International Relations 0(0)
If Japan had declared war, or perhaps even if the Japanese military had had to openly cross a clearly delimited and defended international border to engage opposition forces, then there would have been less ambiguity over whether its actions constituted a norm violation. Under Article 16 of the Covenant, the actions to be taken if a member resorts to war illegally (i.e. without first submitting to arbitration or complying with a Council report) are clear: members ‘hereby undertake immediately to subject [the violator] to the severance of all trade or financial relations, the prohibition of all intercourse between their nationals and the nationals of the covenant-breaking State, and the prevention of all financial, commercial or personal intercourse between the nationals of the covenantbreaking State and the nationals of any other State, whether a Member of the League or not.’ The ambiguity over the definition of aggression meant that a possibility existed that other members of the international community, especially the Great Powers, would not object to Japan’s actions. When planning the military action in Manchuria and when defending those actions publicly and in the League, Japanese decision-makers were conscious of and attempted to exploit this ambiguity in order to reduce opposition to their actions (Ogata, 1984; Yoshihashi, 1963). The League and the US wanted to make it clear to Japan, and to each other, that actions of this type were not acceptable and did in fact fall under the formally agreed concept of external aggression. This was one of the motivations appealed to by prominent decision-makers involved in the nonrecognition of Manchukuo. For example, Henry Stimson’s Note of nonrecognition of 7 January 1932 seems to have been partly intended to let the Japanese government know that Stimson considered the occupation of Jinzhou an act of aggression. Prior to that occupation, Stimson was undecided as to whether Japanese military action was or was not aggression and therefore a serious norm violation. When discussing matters with the Japanese Ambassador before Christmas 1931, Stimson made it clear that if the Japanese army stopped where it was, their actions were not definitively aggression but that ‘an attack if it came against the regular forces, would necessarily be in the nature of an aggression’ (Stimson, 1931–1932: 23 December 1931). What was the role of this ambiguity? The ambiguity over the status of the action meant that there was extensive discussion over whether the actions were acceptable or prohibited. This was part of the dispute over the international rules governing the use of force that resulted in declarations of nonrecognition of Manchukuo.
Inconsistency The existing rules of international behavior conflicted in ways that allowed the various parties to the crisis to claim that their position conformed to accepted and legitimate ways of acting. These inconsistencies meant that there was space for a dispute over which of the conflicting rules applied. During the interwar period, the new rules governing the use of force, represented by the Covenant and the Kellogg–Briand Peace Pact, were casting doubt on the old ways. For some practices, it was unclear whether the new rules were applicable and the actions were unacceptable, or whether the actions fell under the old understandings and so were allowed and did not merit condemnation. One of the main sources of uncertainty was the conflict between what was considered standard colonial behavior and the new prohibition on aggression. Especially inconsistent
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
17
O’Mahoney
with the new injunctions on war and aggression were the seeming prerogatives of colonial powers in China and East Asia. The formal agreements between the Great Powers and what appeared to be standard practice allowed behavior that looked very much like what Japanese troops were doing during the Manchurian Crisis: R5: Great powers have the right to use force against the indigenous population to protect their colonial possessions (particularly in China). It is easily understandable that there might have been genuine confusion over which sets of rules were to be applied. Was this an acceptable instance of colonial intervention or was it aggressive war? This inconsistency helped to generate a dispute over whether actions like those of the Japanese in 1931–1933 were acceptable. Throughout the 19th and into the 20th century, European Great Powers, the United States, and Japan used force in China and extracted a variety of concessions. These included territorial cessions, most notably Hong Kong, but also other national settlements in cities such as Hankou, Tianjin, and Shanghai, and extensive rights such as extraterritoriality and the right to station police and military forces. Japan acquired Taiwan (1895) and Korea (1905–1910) outright through the use and threat of force. After the end of World War I when Japan issued a list of 21 Demands, which would have made China a protectorate, the reaction was limited and not framed in terms of opposition to the use of force to gain advantage. In the period leading up to the Manchurian Crisis, powers other than Japan exercised a broad freedom of action regarding their possessions and rights in China. Examples are the British, US, and other powers’ actions in reaction to the threats to their concessions in Hankou, Jiujiang, and Nanjing. In 1927, during a military campaign by the Guomindang called the Northern Expedition, intended to both eliminate the Communists and unify the country, Chinese troops attacked and looted foreign concessions. While damage was light in Hankou and Jiujiang (Munro-Faure, 1989 [1927]), the situation in Nanjing, known as the Nanjing Incident, was more serious. Several foreigners were killed, including British, American, French, Italian, and Japanese nationals. In response, British and American gunboats shelled the city, and naval forces of other states were also involved in protecting their nationals. During this period, Britain had 20,000 troops stationed in Shanghai and the US had 5000 there and in the Beijing-Tianjin region (Thorne, 1972: 48–51). Japan was urged to increase her military presence in Shanghai for the mutual defense of the settlement. Also, in May 1928, Chinese troops in Jinan (in northern China) came into conflict with the Japanese army and killed Japanese civilians. The Japanese mounted a large-scale operation and drove the Chinese troops out of the Jinan area. These actions met with no international condemnation (Shinkichi Eto, 1986: 112). When the League Council decided to invoke the Pact of Paris, on 17 October 1931, and notify China and Japan, the Japanese reply took two interpretive gambits. First, they argued that they were acting in self-defense: ‘Japanese railway guards, in taking military measures in Manchuria since the night of 18 September last, have been actuated solely by the necessity of defending themselves as well as of protecting the South Manchuria Railway and the lives and property of Japanese subjects, against wanton attacks by Chinese troops and armed bands’ (League of Nations, 1931: 2343). The Japanese troops
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
18
European Journal of International Relations 0(0)
were stationed in Manchuria under a system of treaty rights similar to those enjoyed by many powers under instruments like the Boxer Protocol of 1901. Further, Japan charged that Chinese actions, specifically an anti-Japanese economic boycott and acts of violence against Japanese residents, were not ‘in harmony with the letter or the spirit of the stipulations contained in article two of the pact of Paris’ (League of Nations, 1931: 2343). The League Assembly Resolution of March 1932 began with a declaration of the three principles that were felt to be relevant, which included ‘the principle of a scrupulous regard for treaties.’ The Japanese representative responded: the Japanese Government is particularly glad that attention has been called to the principle of scrupulous respect for all treaty obligations. The Japanese Government attaches the greatest importance to that principle. One of the most important factors in the present situation in the Far East is undoubtedly the systematic attacks on legitimate rights and interests guaranteed by treaty, and China’s policy of unilaterally denouncing the most solemn undertakings. (League of Nations, 1932)
The inconsistency between the colonial behavior rules agreed upon, both explicitly and implicitly, and the new rules, especially after the adoption of the Pact of Paris, created the possibility of a dispute over which set of rules applied to an actual occurrence of actions. Japan’s use of force in Manchuria starting in 1931 generated just such a dispute. Another important inconsistency was that between the strict rule of pacta sunt servanda and the new rule of non-aggression, or of the protection of territorial integrity and existing political independence against external aggression. The tension was that if treaties were to be respected and if aggression against territory or political independence was illegitimate, what was the status of a treaty cementing a change in territory or political independence that was the result of aggression? There could have been numerous different responses to this tension. One might have been to focus purely on the avoidance of conflict and encourage the cession of territory or power prior to the outbreak of violence. This was in fact the solution taken when parts of Czechoslovakia were annexed by Germany in 1938. Another response might have been to come up with a reason why this case was an exception to the general rule, such as that the use of force was justified by the goal of national liberation or self-determination. However, the solution to this inconsistency was to declare that some treaties, or in this case situations, were in fact not to be respected. Stimson’s Note of 1932 was the first attempt to assert that Japan’s actions did not fall under the rules of colonial behavior but instead fell under the rules of non-aggression and the subsequent League statements on the issue gave this assertion a collective character.
Inadequacy Willoughby argues that the Sino-Japanese dispute made ‘more evident than had previously been perceived deficiencies in the existing peace structure of the world’ (1935: 580). Part of these deficiencies was the absence of any sanction that could be put on a
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
19
O’Mahoney
rule violator to demonstrate the continuance of the rule, to reaffirm the illegitimacy of aggression, without imposing prohibitive costs on the international community: R6: If a state commits aggression, other states will enact military and impose economic sanctions on that state. Under this abstract definition (Ostrom’s ‘or else’ component of a rule), there is no provision for what should be done if economic or military sanctions were too costly or unattractive for some reason. In this sense, then, there was a rule inadequacy. During the decision-making in the UK, the US, and the League of Nations, there was a sense that the peace machinery was unprepared for this eventuality. Japan’s actions, once deemed unacceptable, a norm violation, required a response or the newly forged consensus that war and aggression was illegitimate would be under threat. Nonrecognition constituted a means of collective condemnation, reasserting the joint disapproval of aggression of the members of the international community. Much of the early discussions by decision-makers revolved around the heavy and undesirable costs of economic or military sanctions. In the US, President Herbert Hoover was adamant that any economic pressure on Japan, such as an embargo, would ‘lead to war’ and thus was to be avoided (Stimson, 1931–1932: 7 November 1931). In the UK, the Foreign Office was generally agreed that Britain’s strategic weakness in the Far East meant that economic sanctions were off the table. The view was generally held that sanctions against Japan ‘were likely to lead to an armed clash with that country’ (Thorne, 1970: 1619–1620). Threats of war, or economic sanctions potentially leading to war, might have been credible when dealing with smaller powers but not against the strongest power in Asia. Austen Chamberlain famously said, ‘the League could make a contribution when sparrows were quarreling but was of little avail when eagles were fighting’ (Lloyd, 1995: 162). So, economic sanctions were out, but this by itself did not necessitate an innovation. It was the desire to do something, in the words of UK Foreign Secretary John Simon, to ‘reassert … the conditions under which every Member of the League is pledged to conduct relations with every other Member’ (League of Nations, 1932: 63). An absence of action by the League would have looked like approval or acceptance of Japan’s actions. Nonrecognition was suggested as an novel option (initially by Hoover (Current, 1954)) because there was no existing provision for a relatively cheap sanction that would nevertheless create common knowledge of the collective valuation of the rule of non-aggression (see O’Mahoney, 2011). As Stimson later pointed out, if the threat of nonrecognition did not deter the Japanese, ‘it would lay a firm foundation of principle upon which the Western nations and China could stand in a later reckoning’ (Stimson and Bundy, 1947: 258; see also 1947: 235). After the Manchurian Crisis and the adoption of nonrecognition as the position of the League, the nonrecognition principle was incorporated into numerous other international instruments and was practiced in many cases of the illegitimate use of force. The International Law Association (ILA) at the 1934 Budapest Meeting issued a (non-binding) resolution on the interpretation of the Paris Pact (Kellogg–Briand Pact). This included Article 5, which stated: ‘The signatory states are not entitled to recognize as acquired de jure any territorial or other advantages acquired de facto by means of a
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
20
European Journal of International Relations 0(0)
violation of the Pact’ (Willoughby, 1935: 594). This ILA resolution is good evidence that there was an awareness of an inadequacy in the specification of the formal institutions that could be resolved by the explicit elaboration of a duty of nonrecognition.
Why not power-shift? There are many ways in which the dominant conception of institutional change in IR is not useful in understanding the rise of the nonrecognition of aggressive gain. The mere fact that powerful states like the US and the UK were involved in the political process is not evidence in favor of the power-shift hypothesis (see Table 1 for a specification of relevant evidence). First, there was no identifiable shift in power that allowed actors to change the rules to their benefit. World War I had resulted in some material power changes, but these preceded both a rejection of a rule of nonrecognition and the adoption of the practice of nonrecognition. The second issue is whether the new enactment of the rules was the result of a conscious choice by the ruling coalition. The overwhelming impression from the Manchurian Crisis is that the adoption of nonrecognition was an ad hoc reaction to circumstances, that is, the dispute over the rules, rather than a premeditated decision. Third, there was almost no consideration of potential benefits from the rule. Few, either publicly or privately, thought that nonrecognition would do much at all, let alone allow the states of the League and the US to reap the rewards. Stanley Hornbeck, the Chief of the Far Eastern Affairs Division at the State Department, viewed flexibility as the only advantage of nonrecognition, that is, the US could ignore it in the future (Doenecke, 1981: 84–85). The discussion in the League Council and Assembly revolved around protecting the League and the institution of non-aggression from unraveling rather than anything more directly beneficial to individual states.
Conclusion The disconnect between the delegitimization of war as a dispute-resolution mechanism and the institutional features available to enforce this new rule generated problems. Several previous disputes had been satisfactorily dealt with through the existing institutional resources in the League of Nations Covenant. However, the Manchurian Crisis proved insoluble via any of the existing procedures and so required a different reaction. The tensions in the rule system were central in generating the dispute. The Chinese and their supporters claimed that Japan was violating the League Covenant and the Kellogg– Briand Pact. Japan insisted that all of its actions were compliant with the existing rules and was supported in this position by sizeable constituencies in the US and League states. The rules were unclear on whether the current situation was a violation. Both the Covenant and the Kellogg–Briand Pact referred to ‘war,’ but neither Japan nor China declared war. Japan also did not invade China in any obvious sense, as there were many Japanese troops already legally stationed in China. The existing rules also did not specify the action to be taken in the event that Japan refused to cease military action or acknowledge Chinese sovereignty over Manchuria. US Secretary of State Henry Stimson was the first to combine the practice of nonrecognition, the principle that
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
21
O’Mahoney
illegal acts cannot create rights, and the illegality of aggression and threatened to ‘not recognize any situation, treaty or agreement which may be brought about by means contrary to the covenants and obligations’ of the League Covenant or the Kellogg– Briand Pact (US Department of State, 1948: III, 8). Once this position was adopted by the League, it constituted a change in the rules of international behavior. What are the implications of these findings for the way that IR theory treats institutional change? There are limitations to the idea that the only ways for change to occur are through a replacement of actors with different preferences or through a change in the bargaining power of already powerful actors. The addition of the practice of the nonrecognition of aggressive gain cannot be explained in these terms, and it seems plausible that this is also true for a large amount of other institutional changes. This further means that predictions about when change will happen need to take into account the existing tensions in the rule system as well as the power of actors to effect change in that system. Further, analysis based on a dynamic view of institutions should take seriously the concept of rule tensions. There is more work to be done specifying different types of tensions and the ways that the disputes generated by these tensions are resolved. There may be recurrent patterns in the link between tensions and the nature of change. For example, one especially interesting question for future research is under what conditions the absence of a specified sanction for norm violation, an inadequacy in the rule system, leads to a rejection of the rule rather than the specification of such a sanction. This kind of fruitfulness of the rule tensions concept suggests that it is a valuable a part of the conversation surrounding norm and institutional change in IR. Acknowledgments I would like to thank Robert Adcock, Davy Banks, Henry Farrell, Martha Finnemore, Daniel Nexon, Susan Sell, and participants in the IR History panel at International Studies AssociationNorth East 2011 for comments.
Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Notes 1
2 3
4
This is despite the fact that the Stimson Doctrine is a unilateral political statement of the US government, whereas the phenomenon under consideration is a rule of international behavior. This rule is one example of the more general idea that breaches of law should not be recognized as lawful, or ex injuria jus non oritur. Another concept, regimes, has been largely superseded by the concept of institutions in IR. While unanimity is simpler conceptually, Bicchieri also allows that a norm exists if a ‘sufficiently large subset’ of conditional followers exists, and discusses in detail what ‘sufficiently large’ might mean. One example used by Crawford and Ostrom is ‘All male US citizens 18 years of age and older (attribute), must (deontic) register with the Selective Service (aim) by filling out a form at the US Post Office (condition) or else face arrest for evading registration (or else).’
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
22 5
6
European Journal of International Relations 0(0)
An important early example is the Anti-war Treaty of Non-aggression and Conciliation, known as the Saavedra Lamas Treaty, signed by a variety of American nations including the US in October 1933 (see: http://avalon.law.yale.edu/20th_century/intam01.asp, accessed 5 January 2013). A more recent example is the UN ‘Vienna’ Convention on the Law of Treaties’ Article 52, which states, ‘A treaty is void if its conclusion has been procured by the threat or use of force in violation of the principles of international law embodied in the Charter of the United Nations’ (see: http://untreaty.un.org/ilc/texts/instruments/english/conventions/1_1_1969.pdf, accessed 5 January 2013). These are crucial assumptions and ones that are usually not made in rationalist institutionalist accounts.
References Bicchieri C (2006) The Grammar of Society: The Nature and Dynamics of Social Norms. Cambridge: Cambridge University Press. Brunnee J and Toope SJ (2010) Legitimacy and Legality in International Law: An Interactional Account. Cambridge: Cambridge University Press. Calvert R (1995) Rational actors, equilibrium, and social institutions. In: Knight J and Sened I (eds) Explaining Social Institutions. Ann Arbor, MI: University of Michigan Press, 57–95. Carstensen MB (2011) Paradigm man vs. the bricoleur: Bricolage as an alternative vision of agency in ideational change. European Political Science Review 3(1): 147–167. Clark I (2005) Legitimacy in International Society. New York: Oxford University Press. Crawford SE and Ostrom E (1995) A grammar of institutions. American Political Science Review 89(3): 582–600. Current RN (1954) The Stimson Doctrine and the Hoover Doctrine. American Historical Review 59(3): 513–542. Diehl PF, Ku C and Zamora D (2003) The dynamics of international law: The interaction of normative and operating systems. International Organization 57(4): 43–75. Doenecke JD (1981) The Diplomacy of Frustration: The Manchurian Crisis of 1931–1933 as Revealed in the Papers of Stanley K. Hornbeck. Stanford, CA: Hoover Institution Press. Eto S (1986) China’s International Relations 1911 -31. In: Fairbank JK and Feuerwerker A (eds) The Cambridge History of China, Volume 13, Republican China 1912 -49, Part 2. Cambridge: Cambridge University Press, 74–115. Farrell H (2009) The Political Economy of Trust. Cambridge: Cambridge University Press. Ferrell RH (1968 [1952]) Peace in their Time: The Origins of the Kellogg–Briand Pact. Hamden, CT: Archon Books. Fioretos O (2011) Historical institutionalism in international relations. International Organization 65(2): 367–399. Franck TM (1988) Legitimacy in the international system. American Journal of International Law 82(4): 705–759. Gilpin R (1983) War and Change in World Politics. Cambridge: Cambridge University Press. Hart HLA (1994) The Concept of Law. Oxford: Oxford University Press. Hawkins DG, Lake DA and Nielson DL (2006) Delegation and Agency in International Organizations. Cambridge: Cambridge University Press. Hay C (2006) Constructivist institutionalism. In: Rhodes RAW, Binder S and Rockman B (eds) The Oxford Handbook of Political Institutions. Oxford: Oxford University Press, 56–74. Ikenberry GJ (2001) After Victory: Institutions, Strategic Restraint, and the Rebuilding of Order after Major Wars. Princeton, NJ: Princeton University Press.
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
23
O’Mahoney
Katznelson I and Weingast BR (2005) Preferences and Situations: Points of Intersection between Historical and Rational Choice Institutionalism. New York: Russell Sage. Koremenos B, Lipson C and Snidal D (2001) The rational design of international institutions. International Organization 55(4): 761–799. Korman S (1996) The Right of Conquest: The Acquisition of Territory by Force in International Law and Practice. New York: Oxford University Press. Krook ML and True J (2012) Rethinking the life cycles of international norms: The United Nations and the global promotion of gender equality. European Journal of International Relations 18(1): 103–127. Langer R (1947) Seizure of Territory: The Stimson Doctrine and Related Principles in Legal Theory and Diplomatic Practice. Princeton, NJ: Princeton University Press. League of Nations (1928) Official Journal Special Supplement 64. League of Nations (1930) Official Journal 1930. League of Nations (1931) Official Journal 1931. League of Nations (1932) Official Journal Special Supplement 101. Lloyd L (1995) The League of Nations and the settlement of disputes. World Affairs 157(4): 160–174. McNair A (1933) The Stimson Doctrine of nonrecognition: A note on its legal aspects. British Yearbook of International Law 14: 65–74. Mahoney J and Thelen K (2010) A theory of gradual institutional change. In: Mahoney J and Thelen K (eds) Explaining Institutional Change: Ambiguity, Agency, and Power. Cambridge: Cambridge University Press. Malawer SS (1977) Imposed Treaties and International Law. Buffalo, NY: William S. Hein & Co. March JG and Olsen JP (2006) Elaborating the ‘new institutionalism’. In: Rhodes RAW, Binder SA and Rockman BA (eds) The Oxford Handbook of Political Institutions. Oxford: Oxford University Press, 3–20. Munro-Faure PH (1989 [1927]) The Kiukiang incident of 1927. Journal of the Royal Asiatic Society Hong Kong Branch 29: 61–76. O’Mahoney J (2011) Proclaiming principles: Rule maintenance and symbolic sanctions. Paper presented at International Studies Association Annual Conference. Ogata SN (1984) Defiance in Manchuria: The Making of Japanese Foreign Policy, 1931–1932. Newport, CT: Greenwood Press. Sandholtz W (2008) Dynamics of international norm change: Rule against wartime plunder. European Journal of International Relations 14(1): 101–131. Sandholtz W and Stiles K (eds) (2009) International Norms and Cycles of Change. Oxford: Oxford University Press. Sandholtz W and Stone Sweet A (2004) Law, politics, and international governance. In: ReusSmit C (ed.) The Politics of International Law. Cambridge: Cambridge University Press, 238–271. Schmidt VA (2008) Discursive institutionalism: The explanatory power of ideas and discourse. Annual Review of Political Science 11: 303–326. Searle J (2010) Making the Social World. Oxford: Oxford University Press. Stimson HL (1931–1932) Diaries, 1931–2, Library of Congress. Stimson HL (1932) The Pact of Paris: Three years of development. Foreign Affairs 11(1, Special Supplement): i–ix. Stimson HL and Bundy M (1947) On Active Service in Peace and War. New York: Harper and Brothers. Thorne CG (1970) The Shanghai crisis of 1932: The basis of British policy. American Historical Review 75(6): 1616–1639.
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013
24
European Journal of International Relations 0(0)
Thorne CG (1972) The Limits of Foreign Policy: The West, the League and the Far Eastern Crisis of 1931–1933. London: Hamish Hamilton. Turns D (2003) The Stimson Doctrine of non-recognition: Its historical genesis and influence on contemporary international law. Chinese Journal of International Law 2(1): 105–143. US Department of State (1948) Foreign Relations of the United States, 1932, vol. III. Washington, DC: GPO. Van Kersbergen K and Verbeek B (2007) The politics of international norms. European Journal of International Relations 13(2): 217–238. Wendt A (2001) Driving with the rear-view mirror: On the rational science of institutional design. International Organization 55(4): 1019–1049. Widmaier WW, Blyth M and Seabrooke L (2007) Exogenous shocks or endogenous constructions? The meanings of wars and crises. International Studies Quarterly 51(4): 747–759. Williams JF (1932) The new doctrine of ‘recognition’. Transactions of the Grotius Society 18: 109–129. Willoughby WW (1935) The Sino-Japanese Controversy and the League of Nations. Baltimore, MD: Johns Hopkins University Press. Wright Q (1919) Effects of the League of Nations Covenant. American Political Science Review 13(4): 556–576. Wright Q (1935) The legal foundation of the Stimson Doctrine. Pacific Affairs 8(4): 439–446. Yoshihashi T (1963) Conspiracy at Mukden: The Rise of the Japanese Military. New Haven, CT and London: Yale University Press.
Author biography Joseph O’Mahoney is currently teaching at Brown University, USA. He has a PhD from the George Washington University. His primary research interests are international norm dynamics and the rules governing the use of force, and especially how norms shape the politics of peace settlements.
Downloaded from ejt.sagepub.com at BROWN UNIVERSITY on July 28, 2013