Conceptualizing interstate cooperation

Abstract

There seems to exist a general consensus on how to conceptualize cooperation in the field of international relations (IR). We argue that this impression is deceptive. In practice, scholars working on the causes of international cooperation have come to implicitly employ various understandings of what cooperation is. Yet, an explicit debate about the discipline's conceptual foundations never materialized, and whatever discussion occurred did so only latently and without much dialog across theoretical traditions. In this paper, we develop an updated conceptual framework by exploring the nature of these differing understandings and situating them within broader theoretical conversations about the role of cooperation in IR. Drawing on an array of studies in IR and philosophy, our framework distinguishes between three distinct types of cooperative state interactions – cooperation through tacit policy coordination (‘minimal’ cooperation), cooperation through explicit policy coordination (‘thin’ cooperation), and cooperation based on joint action (‘thick’ cooperation). The framework contributes to better theorization about cooperation in two main ways: it allows scholars across theoretical traditions to identify important sources of disagreement and previously unnoticed theoretical common ground; and the conceptual disaggregation it provides grants scholars crucial theoretical leverage by enabling type-specific causal theorization.

Introduction

Cooperation is one of the most central concepts in international relations (IR), and the study of its causes has always been at the heart of IR research. Early thinkers such as Thucydides, Polybius, and Niccolò Machiavelli, founding figures of modern IR theory such as E. H. Carr and Hans Morgenthau, and influential contemporary scholars such as Alexander Wendt, Charles Glaser, and John Ikenberry all discuss questions such as: when do states cooperate and when do they not? Can sustainable cooperation between states be fostered? If so, how? And if not, what renders long-term cooperation impossible to achieve?¹

It was only in the late 20^th century, however, that a prolific group of theorists who tried to understand the role of regimes and institutions in international politics first began to address the foundational question of what interstate cooperation is and how IR theorists should define it.² This literature – heavily influenced specifically by the path-breaking work of Robert Keohane – argued that interstate cooperation occurs whenever actors adjust their policies in such a way that their implementation creates benefits for all sides involved.³ Soon after the account was developed and initial scrutiny judged it to be the most promising approach available at the time, it was widely adopted across the discipline and quickly became the conventional account of interstate cooperation.⁴ Today, little explicit debate about how to conceptualize cooperation takes place in the field of IR.⁵ The vast majority of scholars working on issues of interstate cooperation across a variety of issue areas either adopt the conventional account as it was originally put forth, propose seemingly minor alterations to it in passing, or refrain from defining cooperation altogether.⁶ At first sight, there thus seems to exist a general consensus on how to conceptualize interstate cooperation in the discipline.

In this paper, we argue that this impression is deceptive. Although the conceptual literature has long stagnated, a survey of the contemporary IR literature on the causes of interstate cooperation reveals that scholars have come to implicitly employ various understandings of what cooperation is and accordingly identify a large variety of different kinds of state interactions as cooperation. By exploring the nature of these differing understandings and situating them within broader theoretical conversations about the role of cooperation in IR, we develop an updated conceptual framework that captures and orders this evolving variety in how scholars understand interstate cooperation. We first suggest that any plausible understanding of interstate cooperation should exclude the possibility of ‘accidental cooperation’ in which the actors involved do not take each other into account in the policy-making process. Put differently, we show that cooperation always involves some basic intentionality. Based on this theoretical groundwork, we then identify and define three distinct types of interstate cooperation implied by the existing IR literature: cooperation through tacit policy coordination (or ‘minimal’ cooperation), cooperation through explicit policy coordination (or ‘thin’ cooperation), and cooperation based on joint action (or ‘thick’ cooperation). Finally, we illustrate that these three types of interstate cooperation can be integrated into a common conceptual framework that distinguishes between the three types by reference to two criteria: first, whether they involve a tacit or explicit process of policy coordination; and second, whether they involve the pursuit of individual or shared goals.

As we discuss in detail below, this conceptual framework contributes to better theorizing about cooperation in two main ways. First, it facilitates theoretical exchange between different branches of IR literature that have largely developed independently of each other. By providing them with a common, more specific conceptual language vis-á-vis interstate cooperation, the updated conceptual toolkit allows scholars to identify important sources of disagreement as well as previously unnoticed theoretical common ground. Second, our framework enables IR scholars to disaggregate the often-employed dependent variable ‘cooperation’ into more specific subtypes and thus develop theories that are specifically targeted at the form of cooperation they are most interested in. This promises to grant scholars considerable theoretical leverage. Because it employs the conventional, very extensive conceptualization, current causal theorizing about interstate cooperation targets a very large population of cases. Yet, said population is likely causally heterogenous: ‘minimal’, ‘thin’, and ‘thick’ cooperation each represent a unique kind of state interaction, and it is very well possible that theories apt at explaining one type of cooperation might struggle to explain instances of the other two. Our framework provides the kind of conceptual disaggregation necessary to engage in more targeted, type-specific causal theorization.

The remainder of this paper proceeds as follows. First, we provide an overview of the conventional account's conceptualization of interstate cooperation and discuss its theoretical implications, highlighting that it classifies a broad range of interactions as cooperation. Second, in order to lay the groundwork for our own conceptual framework, we demonstrate that the conventional account technically allows for what would amount to truly ‘accidental cooperation’ and suggest that any plausible understanding of interstate cooperation should include a requirement on the intentions states hold in the process of adjusting their policies. Third, turning to our main task of identifying and defining different types of interstate cooperation, we show that many IR scholars have, in fact, come to apply a more restrictive understanding of cooperation than originally implied by the conventional account and point out that their adaptations suggest a distinction between two types of interstate cooperation: ‘minimal’ and ‘thin’ cooperation. Fourth, we note that this distinction still leaves unrecognized a third type of interstate cooperation, namely ‘thick’ cooperation, which is the subject of much debate especially within the social constructivist literature. Fifth, we provide a synthesizing framework that spells out the relationship between these three types of cooperation, before explaining how such a framework can crucially contribute to productive theorizing about IR. Finally, our conclusion highlights promising avenues for future research.

The conventional account of interstate cooperation

As noted above, the conventional approach to defining interstate cooperation originated in the literature on regimes in international politics. The understanding of cooperation implied there finds its most explicit and prominent definition in Keohane's seminal treatise After Hegemony, which today is widely considered to offer the ‘now standard way [to define] … cooperation’ in IR.⁷ In a nutshell, Keohane posits that actors cooperate with each other ‘when [they] adjust their behavior to the actual or anticipated preferences of others, through a process of policy coordination’ and thereby realize mutual benefits, that is, these adjustments must bring about a situation in which ‘the policies actually followed by one government are regarded by its partners as facilitating realization of their own objectives’.⁸ Cooperation between a set of actors, therefore, entails that all of the actors realize gains, but the ‘joint gains in such a situation need not be equal’.⁹ Especially, cases of multilateral cooperation – a phenomenon that has become increasingly common in the international arena – almost always necessitate some kind of compromise between the different actors, meaning that the amount of gains each of them realizes will often vary.¹⁰

Crucially, the account's definition of cooperation specifies that states' policies not only must lead to the realization of mutual benefits (what we also refer to as ‘the outcome’), but also that the outcome must be brought about through a particular process, namely one involving policy coordination.¹¹ Indeed, the frequency with which Keohane refers to policy coordination as a restriction on the process through which cooperation might be produced alone suggests that it constitutes a crucial substantive component of the conceptual framework.¹² However, a close read reveals that the account ultimately places no meaningful substantive restriction on the process producing the outcome. Effectively, in the conventional account, policy coordination merely refers to the emergence of coordinated policies, not to a particular kind of coordinating process. Keohane lists a plethora of processes through which cooperation might emerge. For instance, cooperation could involve a negotiation process in which the actors come together and explicitly bargain over potential policy adjustments. Yet, he notes that such bargaining could also take place tacitly, a well-known example of which is cooperation emerging in an iterated Prisoner's Dilemma without communication between the two actors.¹³ Another possibility is that an adjustment of policies is imposed by a more powerful actor.¹⁴ Keohane goes even further, however, and highlights that even unilateral behavior in which ‘one country may shift its policy in the direction of another's preferences without regard for the effect of its action on the other state’ also qualifies as policy coordination.¹⁵

Given the variety of processes and interactions the conventional account considers to adequately fulfill the policy coordination restriction, it is ultimately unclear what kind of processes would not be considered producing cooperation when they lead to mutually beneficial policy adjustments. Put differently, because the conventional account does not specify what kind of processes would prevent mutually beneficial policy adjustments from being categorized as instances of cooperation, it ends up defining cooperation merely as the emergence of coordinated policies, that is, a reduction in policy incompatibility between actors that entails mutual benefits.

Importantly, many scholars have applied such a permissive reading of the conventional account in their substantive works on interstate cooperation. For example, classic works in the security studies literature on cooperation and competition suggest that cooperation between states can result from various kinds of behaviors, including ‘decisions to make concessions during a crisis … decisions to forgo launching a war … formal and informal reciprocated restraint’, or the implementation of other kinds of ‘unilateral defensive policies’.¹⁶ As long as the actors' policy adjustments create benefits for all sides, these scholars conceive of states as cooperating with one another. This becomes particularly conspicuous in the literature on the security dilemma, where scholars imply that if two great powers decide to forgo military armament instead of engaging in an arms race, such ‘reciprocated restraint’ constitutes an important instance of international cooperation.¹⁷ In this context, the observation that states cooperate entails no claim about the motivations or process behind either side's policy choice; rather, the exclusive focus lies on the emergence of a particular set of mutually beneficial policies.¹⁸ Relatedly, work on international regimes suggests that ‘cooperative outcomes’ can be produced in a myriad of ways. Lisa Martin, who has done much to advance the study of international cooperation, notes that ‘states can reach the [relevant policy] decisions’ that produce cooperation ‘through genuinely multilateral discussion, a series of bilateral agreements, or the imposition of decisions on a unilateral basis’.¹⁹ Finally, along similar lines, work by Michaela Mattes and Mariana Rodríguez on the role of regime type as a cause of cooperation also remains agnostic about whether a set of mutually beneficial policies emerges ‘in the form of a treaty or an international agreement whereby the parties simultaneously mutually adjust their policies’ or whether it ‘emerge[s] informally as a result of individual unilateral steps that taken together constitute mutual policy adjustment’.²⁰

Note that since the conventional account – and thus large parts of the literature on international cooperation – focuses on the production of mutual benefits through policy adjustments, a large variety of decision situations are viewed as presenting opportunities for interstate cooperation. One important implication of this is that most IR scholars do not consider situations that can be modeled as a Prisoner's Dilemma to be the only kind of cooperation problem. As an example, Martin notes that there are several ‘ideal types of cooperation problems’, including ‘coordination, suasion, and assurance’ games that in parts differ profoundly from the classic Prisoner's Dilemma.²¹ For instance, in coordination games (such as the Battle of the Sexes) or assurance games (such as the Stag Hunt), mutual defection is not the only Nash equilibrium.²² Yet, despite the fact that the strategic context, and thus expected outcomes, are fundamentally different across these games, the common understanding in the IR literature is that they all capture important problems of cooperation.²³

The implausibility of ‘accidental cooperation’

The conventional approach to defining interstate cooperation is thus, at a conceptual level, extraordinarily permissive, in that it allows for a variety of processes to be involved in the emergence of cooperation. Any instance in which mutually adjusted policies produce mutual benefits is classified as an instance of cooperation.

Before beginning our main task of distinguishing between different types of interstate cooperation, we argue in this section that any plausible understanding of cooperation needs to place at least some basic restriction on the process through which the outcome emerges. Specifically, we show that the conventional account in its widely accepted formulation would technically classify instances in which two states accidentally realize mutual benefits through their unilateral policy adjustments as cooperation. To denote such scenarios as some sort of ‘accidental cooperation’, however, appears implausible. In response, we suggest that any plausible definition of cooperation must at least include what we call the intention constraint. This base requirement constitutes an important theoretical device for demarcating plausible from implausible instances of cooperation, thereby laying the ground for our endeavor in the subsequent sections, namely, to develop a schematic overview of different plausible types of interstate cooperation.

To see how a radically permissive approach to defining interstate cooperation allows for what would be instances of ‘accidental cooperation’, consider the following thought experiment. Imagine two states that have virtually no domestic environmental protection laws in place because they seek to maintain the competitiveness of their domestic manufacturing sector. The unregulated production of manufacturing goods creates serious negative environmental spillovers, both for geographically proximate states, for example, through ‘the pollution of international waterways’, as well as for more distant states, for example, through their negative impact on the global commons.²⁴ The two states' lack of an environmental protection policy, as an unintended by-product, imposes considerable negative externalities on other states, which hinders them from pursuing some of their central goals, including securing the health of their population as well as their economic performance. Now, in an effort to modernize their economies and thus increase their overall competitiveness in world markets, both states unilaterally stop the subsidization of their manufacturing sectors and enact far-reaching environmental protection laws within their respective territories. As a consequence of this pair of decisions which is driven by an individual striving for competitiveness, each state significantly reduces the harm it inflicts on the global commons and the degree to which it exposes neighboring states to the costs of border-traversing pollution.

According to the conventional account, this scenario describes an instance of interstate cooperation. At first, the two states' lack of environmental regulations imposes tremendous costs on the other and thus hinders its economic performance and ability to maintain the health of its citizens. Their policy adjustments in the form of both unilaterally enacting far-reaching environmental protection laws eradicate these hindrances, meaning that their environmental policies are now coordinated in the sense that they realize mutual benefits.

However, it strikes us as implausible that this scenario, in fact, describes an instance of interstate cooperation. In our light, this intuitive oddness stems from the fact that the actors completely disregard one another in the process of adjusting their policies. Cooperation seems to crucially presuppose that each actor takes the other actors and their goals into account and shows regard for them in their decision-making process.²⁵ Conversely, when such mutual regard is entirely missing and increased policy harmony is nothing more than accidental, it seems incorrect to attest cooperation.²⁶

We, therefore, posit that, in any plausible case of cooperation, the actors must mutually regard each other in the sense that they hold an intention to at least not obstruct the other actors' goals. Tellingly, philosophers working on issues related to cooperation tend to agree that such an intention must underlie any plausible understanding of cooperation. Take, for instance, the path-breaking work of Raimo Tuomela. He argues that ‘“coordination-achieving” cooperation based on private goals’ only qualifies as such if each of the actors involved has ‘the intention of satisfying one's goal by means-actions which do not conflict with others’ attempts to achieve their goal’, which ensures the presence of some basic ‘cooperativeness’.²⁷ Although Tuomela focuses on interpersonal cooperation, it stands to reason that a requirement along similar lines should hold in the case of interstate cooperation. The preceding discussion therefore suggests the following constraint on the emergence of interstate cooperation:

Intention Constraint. Interstate cooperation can emerge between actors only if states adjust their policies intending not to obstruct the other actors' goal pursuit, or to even positively further it.

Importantly, having such an intention does not imply that the actors necessarily harbor an intrinsic concern for each other's interests. The intention constraint also allows for mere instrumental concern: actors might intend not to obstruct each other's goals solely because they are aware that this will ultimately help fulfill their own goals.

Requiring fulfillment of the intention constraint as part of any definition of cooperation remedies the problem of ‘accidental cooperation’ as characterized above. To see this, imagine a variant of the pollution case in which the intention constraint is, in fact, satisfied, that is, the actors' policy adjustments are now guided by an intention to not obstruct the achievement of the other actor's goals. In this case, the actors still each work toward achieving their own individual goals but, at the same time, consciously try to do so in a way that is compatible with the other actor's goal pursuit, thereby taking each other into account in the decision-making process. Due to precisely this kind of mutual regard displayed by the actors, it seems appropriate to attest cooperation in this modified case.²⁸

To sum up, including the intention constraint sidesteps the danger of potentially recognizing truly ‘accidental’ instances of cooperation and makes clear that cooperation always involves some basic form of intentionality. Whenever the intention constraint is not fulfilled – for example, when states implement more restrained military policies without any regard for their effect on other actors, or when the emergence of mutually beneficial environmental policies is mere happenstance that involves no intention vis-à-vis other actors at all – IR scholars should not attest cooperation.

The analysis conducted in this section is highly relevant for our main endeavor of advancing a comprehensive conceptualization of different types of interstate cooperation. By equipping us with a formal device for distinguishing plausible from implausible instances of cooperation, the preceding analysis not only fills an important gap within the existing conceptual literature but also prepares the ground for developing a comprehensive typology of different plausible instances of interstate cooperation.

Distinguishing ‘minimal’ and ‘thin’ cooperation

We now turn to our main task of distinguishing between – and formally conceptualizing – different types of interstate cooperation. In a first step toward developing an integrative typology of interstate cooperation, we show in this section that a significant portion of scholars have, in practice, come to modify the conventional account in important ways. We demonstrate that, through their adaptations, they have more or less explicitly come to suggest a distinction between two types of interstate cooperation: ‘minimal’ cooperation and ‘thin’ cooperation. To establish this distinction, we discuss several examples of how scholars have turned toward a more restrictive interpretation of the conventional account, propose a formal demarcation criterion between the aforementioned types of cooperation based on their works, and illustrate the distinction by providing real-world examples of both ‘minimal’ and ‘thin’ cooperation.

Reconsidering the policy coordination process

As noted above, the conventional definition of cooperation focuses on the emergence of coordinated policies, that is, mutually beneficial policy adjustments. In consequence, the account – like many scholars employing it in their works – is explicitly and deliberately agnostic about the kind of state interaction involved in the production of cooperation.

Over the years, however, a significant number of IR scholars have come to adopt a much more restrictive approach to defining interstate cooperation than originally suggested by the conventional account. Rather than being agnostic about how mutually beneficial policy adjustments emerge, they have come to understand interstate cooperation as implying that a particular process of policy coordination takes place between the actors involved.

A prime example can be found in Eric Grynaviski's work on the origins of international cooperation. Generally basing his conceptualization of international cooperation on the conventional account, he deviates from its original formulation by positing that cooperation entails ‘the explicit coordination of policies’ and ‘usually tak[es] the form of an international agreement’. He argues that ‘cooperation’ requires ‘the explicit formulation of policies that intend to generate outcomes that improve the condition of both partners’. As such, Grynaviski moves away from an understanding of cooperation that merely implies the implementation of a particular set of policies to one that conceives of cooperation as necessarily involving explicit negotiation and bargaining in the process of reaching mutually beneficial policy adjustments.²⁹

Many other theorists employ a similarly restrictive definition of cooperation in their works, although some of them are much less explicit about it than Grynaviski. For instance, Michael Webb suggests in his discussion of cooperation in the international political economy that it entails ‘negotiated mutual adjustment that causes states to pursue different policies than they would have chosen had policy-making been unilateral’. He continues: ‘[i]t is important to note that according to this definition, coordination involves both negotiation and policy adjustment’.³⁰ Along similar lines, Andrew Moravcsik conceives of ‘international cooperation as an effort to arrange mutually beneficial policy coordination among countries whose domestic policies have an impact on another’. For him, too, cooperation implies that states coordinate their policies through ‘interstate bargaining’ which ultimately produces ‘substantive agreements’.³¹

To reiterate, what these scholars have in common is that, to them, interstate cooperation entails a claim about a particular process of explicit coordination taking place. In doing so, these theorists advance a markedly different conceptualization of interstate cooperation than those who, as previously discussed, focus on the emergence of a certain set of policy decisions and largely disregard what kind of process led to their realization. At first sight, especially given that these scholars frequently refer to the conventional account approvingly, this might appear to be nothing more than a semantic issue.³² Yet, to the contrary, the turn toward an understanding of cooperating as necessitating an explicit coordination process has far-reaching conceptual implications.

To see what exactly those are, recall that the conventional account originated out of an effort to understand how international regimes might influence interactions of states under anarchy. Scholars working in this tradition paid close attention to situations involving ‘different configurations of state interests, i.e., … different types of “cooperation problems”’.³³ Conceiving of a plethora of scenarios as game-theoretic problems of strategic interaction, they sought to understand under what conditions states implement policies that benefit all parties involved and, conversely, eschew strategies that would leave everybody worse off. Against this backdrop, it is evident why many IR theorists working on international cooperation avoided placing significant limitations on the process that precedes policy adjustments. After all, these scholars thought that many state interactions could be modeled as, for instance, iterated Prisoner's Dilemmas in which the actors' ability to communicate with each other was drastically limited. Whenever cooperation – again, understood as the implementation of mutually beneficial policies – emerged in these models, it was the result of a tacit adjustment process, that is, of unilateral, ‘independent policy-making that takes international factors into account’.³⁴ The conventional account's permissive conceptualization of cooperation thus reflects a desire to capture such tacit forms of policy coordination in addition to those instances in which policy coordination involves an explicit negotiation process.

Yet, those scholars that have turned away from this permissive conceptualization have done so precisely because they do not want to diagnose cooperation in instances where coordinated policies emerge without direct communication and bargaining between the actors. For instance, Webb directly opposes defining cooperation in a way that ‘does not distinguish independent policy-making that takes international factors into account from internationally negotiated adjustments to national policy’.³⁵ Similarly, Grynaviski highlights that, according to his understanding of cooperation, ‘cooperating in the Prisoner's Dilemma is not cooperating; the prisoners in separate cells do not have the ability to explicitly coordinate their actions’.³⁶

Structuring the debate: ‘minimal’ vs. ‘thin’ cooperation

We argue that this substantial – albeit rarely recognized and thus not yet formalized – disagreement within the IR literature is best understood as suggesting a distinction between two types of interstate cooperation: ‘minimal’ and ‘thin’ cooperation. Both types of cooperation focus on mutual policy adjustments that create benefits for all sides. The crucial difference between the two lies in how these adjustments come about. In cases of ‘thin’ cooperation, the actors involved explicitly coordinate their policies, seeking to adjust them in such a way that mutual benefits are realized. What makes the coordination process explicit is that it involves direct communication and negotiation that aims at producing beneficial policy adjustments. In cases of ‘minimal’ cooperation, on the other hand, such policy adjustments emerge out of a tacit coordination process, that is, they are the result of each side's unilateral policy adjustments that are dictated by their individual calculus in a context of strategic interdependence. Instead of explicitly debating possible changes in policy, each of the actors involved is independently ‘acting in pursuit of those things she wants or values in part in light of what she believes the other is doing, and where the other's actions depend in part on what the other thinks she will do’.³⁷ In contrast to instances of ‘thin’ cooperation, when states cooperate in the ‘minimal’ sense, there is thus no joint ‘design of cooperative arrangements and no direct communication’ about the possible realization of mutual benefits through policy adjustments.³⁸ Rather, mutually beneficial policies emerge out of a tacit process of mutual adjustment in which the actors strategize about the possible effects of their own actions on other actors and how the latter might respond to them.

Formally, we can thus define two distinct types of interstate cooperation implied by the IR literature:

Minimal Cooperation. Minimal interstate cooperation obtains if, and only if, states realize mutual benefits through policy adjustments that are the result of a tacit process of policy coordination.

Thin Cooperation. Thin interstate cooperation obtains if, and only if, states realize mutual benefits through policy adjustments that are the result of an explicit process of policy coordination.

Importantly, the processes involved in ‘minimal’ and ‘thin’ cooperation – tacit and explicit policy coordination, respectively – should be understood as entailing satisfaction of the intention constraint.³⁹ This is to say that, for instance, when states cooperate in the ‘minimal’ sense, they need to hold an intention to at least not obstruct the other actors' goal pursuit during the process of tacit policy coordination. This is crucial because, as we have seen in our discussion of ‘accidental cooperation’ above, in the absence of such basic cooperative intentions our conceptualization of interstate cooperation would run the danger of producing conceptual implications that contradict our intuitions about what cooperation is.

Furthermore, note that both processes, tacit and explicit policy coordination, will in many cases take place over an extended period of time. The kind of communication and bargaining characteristic of explicit policy coordination will usually unfold over time, as state representatives discuss, negotiate, and ultimately agree on mutual policy adjustments. Similarly, tacit policy coordination will often occur in a gradual, temporally extended process, as states observe each other's behavior, adjust their policies in response, observe these policy changes, adjust to them, and so on and so forth. It is in this sense that our conceptualization of ‘minimal’ and ‘thin’ cooperation reflects one of the key insights in the literature on international cooperation, namely that many, if not most, instances of cooperation result from an iterative process in which states engage each other repeatedly over time.⁴⁰

‘Minimal’ and ‘thin’ cooperation are mutually exclusive types of cooperation: either the actors engaged in cooperation coordinate their policies explicitly or mutual policy adjustments occur through a tacit process of policy coordination. To be sure, in practice, any group of two or more states will often engage in both ‘minimal’ and ‘thin’ instances of cooperation: over time, and across different issue areas, states frequently engage in various kinds of cooperative activities. However, as we will now illustrate by discussing real-world examples of both ‘minimal’ and ‘thin’ cooperation, one can usually classify particular sets of interactions on a given issue as closely approximating either the ‘minimal’ or ‘thin’ type of interstate cooperation.

Real-world examples of ‘minimal’ and ‘thin’ cooperation

A turn toward real-world examples helps illustrate what this theoretical distinction looks like in practice. Examples of ‘thin’ cooperation are easily discerned. Consider, for instance, the Anti-Ballistic Missile Treaty concluded between the United States and the Soviet Union in 1972, which marked the end of negotiations designed to bring about a coordination in policies that would leave both sides better off.⁴¹ Similarly, international coordination on trade policies that was formally codified in the General Agreement on Tariffs and Trade was the result of explicitly negotiated policy adjustments.⁴² It is equally easy to identify attempts at ‘thin’ cooperation that ultimately failed. For instance, Axel Marschik points at the early 2003 UN Security Council negotiations regarding Iraq as an important example of great powers deliberately seeking to reach an agreement and coordinate their policy positions on a contentious issue that ultimately ended in ‘a breakdown of negotiations, an acrimonious blame game and recourse to force outside the system of the UN Charter’.⁴³

Cases of successful ‘minimal’ cooperation, on the other hand, look markedly different. As an illustrative example, consider the informal restraint the two Cold War superpowers came to exercise after first experimenting with anti-satellite weapons (ASATs) during the 1960s and 1970s. As Nancy Gallagher explains, neither the USA nor the USSR ‘made a sustained effort to develop and deploy ASATs’ although they ‘were technically feasible and legally permissible’. No treaty specifically addressing the issue of ASATs was negotiated and, in fact, ‘it does not appear that the superpowers exchanged views about military space activities writ large’. However, over time, both sides ultimately came to demonstrate ‘reciprocated restraint’ with regard to ASATs: both anticipated that if one of them were to deploy them, the other would quickly follow suit and that, conversely, if one exercised restraint the other might reciprocate and thus enable the two sides to avoid a costly arms race that was not in the interest of either party.⁴⁴

Along similar lines, Benjamin Miller points at US and Soviet policies in the Middle East after the Suez Crisis as yet another example of cooperation being the ‘outcome of unilateral moves’, that is, a tacit coordination process in which communication was limited to unilateral ‘show of force signals’. Despite the fact that both sides ‘fundamentally rejected the idea of spheres of influence’, in practice they came to tacitly coordinate their policies such that each power's sphere would be respected by the other. Such policy coordination, which helped the two superpowers avoid a direct confrontation, occurred ‘from necessity [rather] than by choice’, in that it resulted from the actors' unilateral decisions taken in a situation of strategic interdependence. Miller summarizes aptly: the two great powers ‘tacitly … recognized the stakes of the rival superpower, its sphere of influence, and its equal status’ as, over time, ‘it became easier for each superpower … both to identify and respect the interests and spheres of influence of their rivals in crisis situations’.⁴⁵

Crucially, although no explicit policy coordination took place in these cases of ‘minimal’ cooperation, the emergence of coordinated policies was far from accidental: considerations of the effects their actions would have on the other side featured prominently in both superpowers' decision-making processes, and in adjusting their policies both states came to intend not to obstruct the other on the given issue, albeit for solely egoistic reasons. It is in this sense that even though policy coordination took place only tacitly, it still involved some form of basic intentionality as demanded by the intention constraint.

Widening the conceptual scope: we-mode cooperation

The distinction between ‘minimal’ and ‘thin’ cooperation captures two key forms of cooperative interactions between states in the international arena. We now turn to show, however, that focusing on these two types alone would produce an incomplete typology of interstate cooperation. In this section, we argue that we need to expand our conceptual toolbox to incorporate and account for what IR scholars have referred to as more ‘sophisticated forms of cooperation’ that go distinctly beyond ‘cooperation based on overlapping individual preferences and in the pursuit of self-interested goals’ and cannot be subsumed under either ‘minimal’ or ‘thin’ cooperation.⁴⁶ Although the discussion of such instances of cooperation that entail a particularly involved mode of collaboration has not gained much conceptual attention within IR, research within philosophy on different forms of cooperative endeavors offers a promising starting point for exploring their conceptual foundations. Based on our review of these literatures and how they relate to recent discussions of cooperation in IR, we develop a definition for a third type of interstate cooperation, ‘thick’ cooperation, which refers to the joint pursuit of a shared goal.

We proceed as follows. First, we provide a brief survey of work in philosophy that establishes cooperation in which actors work together toward the achievement of a shared goal as a distinct mode of social interaction. Next, we illustrate that IR scholars across the discipline – but, in particular, those working in the social constructivist tradition – frequently allude to similar interactions taking place between states in the international arena. Based on these observations, we then provide our definition of ‘thick’ cooperation as a third, distinct type of interstate cooperation, before finally providing a series of real-world examples.

Philosophical research on sophisticated forms of cooperation

Over the past 30 years, philosophers have increasingly called attention to and examined phenomena that fall under the heading ‘joint action’. In these sorts of activities that are presumed to represent a vital aspect of human sociality, a set of two or more actors jointly pursues a common – or shared – goal.⁴⁷ Philosophers highlight that paradigmatic instances of joint action occur in relatively mundane situations, for instance when actors paint a house together, lift a table, or sing a song together.⁴⁸ Crucially, in these kinds of cases it is not ‘enough to say that [the actors] each act independently, where the sum of those individual actions involves the realization of a certain effect. In each case there is something that [the actors] together do; apart from [their] individual actions, there is an action that [they] together perform’.⁴⁹

An important class of joint action refers to a distinctive mode of cooperation that philosophers occasionally refer to as ‘we-mode cooperation’ or ‘shared cooperative activity’.⁵⁰ In such cooperative endeavors, the actors' goal pursuit is driven by a commitment to attain their shared goal together, that is, as members of a group with a joint responsibility for achieving that end. The joint pursuit of a shared goal so understood requires an intimate kind of collaboration. For instance, the actors involved need to engage in an ongoing process of determining and coordinating their respective part-actions necessary for achieving the envisaged, shared goal. This process can be labeled ‘we-reasoning’ and, inter alia, includes answering questions such as ‘What should we do in this situation?’ and ‘What should I do as a group member as my part of our group's action?’ instead of merely ‘What should I do in this situation?’⁵¹ As such, cooperation on the basis of joint action clearly involves more than a set of actors ‘taking each other as features of the environment in acting to achieve [their individual aims]’.⁵²

In order to clearly distinguish this kind of cooperative interaction from less elaborate forms of cooperation, philosophers have sought to identify what features capture its unique phenomenology. For example, Bratman characterizes what he calls ‘shared cooperative activity’ (SCA) through reference to three distinct features. First, the actors engaged in an SCA need to exhibit a ‘mutual responsiveness’ toward each other. This means that ‘each participating actor attempts to be responsive to the intentions and actions of the other’. Of course, this feature also seems to figure in ‘minimal’ and ‘thin’ cooperation and is therefore insufficient to distinguish SCAs from these less-involved types of cooperation.⁵³ Yet, Bratman notes that SCAs exhibit two additional features. The actors in an SCA, second, must also each exhibit a ‘commitment to the joint activity’, and ‘their mutual responsiveness’ needs to be ‘in pursuit of this commitment’. Third, and finally, the actors engaged in an SCA each need to have a ‘commitment to mutual support’, meaning that they are ‘committed to supporting the efforts of the other to play [their] role in the joint activity’.⁵⁴

Note that a major philosophical debate surrounds the question of how to best analyze the ontology underlying these cooperative interactions.⁵⁵ Specifically, philosophers controversially discuss the question of whether their analysis requires a reference to irreducible collective intentional states, or even group-level agency, or whether these phenomena can be understood merely by reference to individual actors and their intentional states.⁵⁶ But, although philosophers ultimately continue to disagree on what is the most convincing ontological framework for joint action and the corresponding form of cooperation, it is almost orthodoxy among them that these phenomena are genuine and should be taken seriously.

Before demonstrating the relevance of cooperation based on joint action for contemporary IR theory, it is worth stressing that the phenomenon has already been fruitfully integrated into social scientific debates. As Raul Hakli and his co-authors note, this sense of cooperation is already a salient issue in disciplines such as economics and decision theory, as is evidenced by the highly regarded works of Robert Sugden and Michael Bacharach.⁵⁷ Hakli et al. themselves illustrate how an appeal to cooperation as involving ‘we-reasoning’ leads to more faithful decision-theoretic models of actors' behaviors. In particular, they argue that appeals to cooperation in the we-mode have the potential to remedy well-known problems for standard non-cooperative game theory such as its predictions in games like the Prisoner's Dilemma or Hi-Lo that ‘seem to clash with either empirical results or intuitive rationality judgements’.⁵⁸ Drawing on both economic and philosophical work on joint action, they argue that these decision-making problems can be solved if we suppose that the actors frame the particular problem as one to be solved by ‘we-reasoning’ and not solely by individualistic strategic reasoning. As Hakli et al. summarize, each ‘agent conceives the situation – not as a decision-making problem for individual agents – but as a decision-making problem for the group conceived as an agent’.⁵⁹ Adopting such a collective stance toward decision-making problems promises to more adequately model actors' behavior as well as provide guidance for decision-making procedures.⁶⁰

Sophisticated forms of cooperation in IR theory

The relevant literatures in philosophy, economics, and decision theory primarily focus on instances of joint action as they occur in interpersonal relationships. However, given that large sections of IR scholarship treat states as unitary actors – or even persons – and find this assumption to be compatible with their commitment to scientific realism, it is unsurprising that many scholars contend that cooperation based on joint action also occurs between states in the international arena and thus should be recognized as a distinct phenomenon in need of theorization.⁶¹ In fact, Stacie Goddard and Daniel Nexon have gone as far as noting that ‘beneath their disparate analytical frameworks, social ontologies, and even conceptions of power, many of the different “paradigms”, “turns”, and “isms” found in contemporary … international relations scholarship … share a focus on … mechanisms and processes by which actors engage in, seek to achieve, or orient joint action’.⁶²

It is especially work in the social constructivist tradition, however, that first comes to mind as invoking cooperation based on joint action as an essential phenomenon in IR. Most notably, in her work on global governance, Jennifer Mitzen makes a strong case for viewing ‘concerted action’ or ‘doing something together’ as an important, often overlooked mode of state interaction that ‘is not merely cooperation as IR scholarship has [traditionally] understood it’. When engaging in global governance, she posits, states intentionally design order at the macro-level, ‘concerting their power for common public interest’. Rather than tacitly or explicitly coordinating their individual policies as is the case in what we introduced as ‘minimal’ and ‘thin’ cooperation, Mitzen observes states as at times ‘commit[ting] to address particular problems together’ and thus ‘constitut[ing] a public power beyond the state with the capacity to steer international political outcomes’.⁶³

Along these lines, John Ruggie has famously called for greater attention to ‘collective intentionality’ as an important phenomenon in IR that he considers unduly overlooked by large parts of traditional IR theory. Ruggie specifically notes that states can hold ‘intersubjective beliefs’ of the form ‘we intend’ and ‘I intend only as part of our intending’. He, thereby, suggests that – at least at times, if not regularly – a group of states can closely resemble a group of individuals that jointly pursue a shared goal and work toward its achievement as a collective.⁶⁴

Similarly, Wendt criticizes work ‘which takes self-interested actors as constant and exogenously given and focuses on the selective incentives that might induce them to cooperate’ for ignoring the possibility that ‘states might form collective … interests’. He notes that states often exhibit ‘an empathetic rather than instrumental or situational interdependence between self and other’. The formation of a ‘collective identity’ blurs the distinction between self and other and, in turn, ‘is a basis for feelings of solidarity, community, and loyalty and thus collective definitions of interest. Having such interests does not mean that actors … no longer calculate costs and benefits but, rather, that they do so on a higher level of social aggregation’. For Wendt, it is precisely the emergence of such collective identities and interests that ‘provides an important foundation for’ certain forms of joint action in which states commit to managing a particular policy issue as a group.⁶⁵

Finally, the literature on security communities has pushed for an introduction of the ‘language of community to understand international politics’ properly.⁶⁶ They build on the seminal work of Karl Deutsch who famously defined security communities as ‘a group that has become integrated, where integration is defined as the attainment of a sense of community, accompanied by formal or informal institutions or practices, sufficiently strong and widespread to assure peaceful change among members of a group’.⁶⁷ Communication between members of such groups, Deutsch and his co-authors note, allows them ‘to think together, to see together, and to act together’.⁶⁸ Along these lines, constructivist scholars integrating this line of scholarship into modern IR theory have noted that groups of states sometimes exhibit ‘a compatibility of core values derived from common institutions and mutual responsiveness – a matter of mutual identity and loyalty, a sense of “we-ness”’.⁶⁹

Social constructivist work alone, thus, suggests the need to expand our conceptual toolkit in such a way that it is able to accommodate cooperation based on joint action as a distinct type of interstate cooperation. Indeed, although Goddard and Nexon's aforementioned reference to the centrality of joint action to all traditions of IR theory rightly implies its relevance for the discipline at large, explicit discussions of the phenomenon have been mostly limited to work within this theoretical tradition: it is social constructivists that have put debates about collective intentionality, community, and joint action in IR center stage, and it is them who have been most successful at integrating these discussions into their overarching theoretical frameworks.

However, it is worth noting that work within other branches of IR scholarship, too, hints at the possibility of conceiving of certain cooperative state interactions as involving joint action rather than ‘minimal’ or ‘thin’ cooperation as defined here. For instance, realist work on economic and political integration has more or less explicitly recognized that interstate cooperation at times goes beyond mere ‘minimal’ or ‘thin’ cooperation. As an example, Sebastian Rosato notes that, under certain circumstances, ‘states agree to joint control over their militaries or economies’ and decide to ‘make policy jointly’. In these cases, ‘rather than resting in any one state's hands [policymaking] is determined by the member states acting together’.⁷⁰ When engaging in this specific ‘form of centralized cooperation’, states interact with each other in the kind of we-mode characteristic of joint action: ‘if they are to achieve their common goal’, Rosato posits, ‘they need to know what issues are covered, what is expected of them, what they can expect from their partners, how decisions are going to be made, and so on’.⁷¹

Work in normative IR theory has also analyzed the forms of cooperation that go distinctly beyond the two types of cooperation – ‘minimal’ and ‘thin’ – that have received most attention in mainstream IR theory. As an example, Toni Erskine highlights that states often concert their actions ‘in a way that cannot be simply reduced to the actions and intentions of individuals acting in isolation’. Specifically, she notes that when they ‘share a common purpose and have developed’ at least ‘an informal organizational structure and deliberative capacity in order to pursue it’, states can sometimes engage in what she terms ‘joint purposive action’, that is, ‘sophisticated forms of cooperation’ that emerge when actors ‘come together to participate in a common project’, are ‘united in pursuit of a common purpose’, and engage in ‘deliberation and special coordination’.⁷²

Overall, important branches of IR scholarship, thus, suggest that states at times engage in more sophisticated forms of cooperation than those involved in ‘minimal’ or ‘thin’ cooperation. To be clear, not all of them explicitly discuss these interactions as involving collective intentionality and joint action; in fact, few outside of the social constructivist and normative IR literature do. Yet, what the above examples illustrate is that other scholarly traditions at times portray state interactions along similar lines and at least latently acknowledge the existence of such more sophisticated forms of cooperation. Hence, any typological framework of interstate cooperation seeking to supply IR scholars with the conceptual tools to guide their substantive works should be able to accommodate such cooperation based on joint action.

Defining ‘thick’ interstate cooperation

Scholars within philosophy, economics, and IR, thus, all identify a distinct sense of cooperation that does not merely refer to actors (tacitly or explicitly) coordinating their policies while pursuing their individual goals, but instead involves actors who jointly work toward the attainment of a shared goal. In Margaret Gilbert's words, the kind of cooperation at stake is coupled with a unique sense of ‘acting together’ in which ‘[t]wo or more people … collectively espouse a certain goal, and each one is acting in a way appropriate to the achievement of that goal, where each one is doing this in light of the fact that the goal is their collective goal’.⁷³ Based on the foregoing analysis, we propose the following definition of a distinct type of cooperation that we term ‘thick’ interstate cooperation:

Thick Cooperation. Thick interstate cooperation obtains if, and only if, states engage in the joint pursuit of a shared goal.

A couple of clarifications about the reasoning behind and implications of this conceptualization are in order. First, note that ‘thick’ cooperation – like ‘thin’ cooperation – entails an explicit process of coordination. To see why this is the case, note that for ‘thick’ cooperation to take place the actors need to intentionally create a shared goal that they subsequently pursue jointly. It is hardly conceivable that such shared goal creation and pursuit would ever occur without any direct communication and negotiation between states: at some point, they will need to explicitly discuss – whether in a formal or informal setting – how to begin and organize the joint endeavor. Against this backdrop, it is unsurprising that Erskine decidedly excludes groups that lack the capacity to directly communicate with one another from what she calls ‘joint purposive action’. As she explains, ‘members of a collective must have … the capacity to deliberate (however informally) in order to coordinate their actions (even imperfectly) in circumstances in which the required collective action is not obvious but, rather, open to disagreement’.⁷⁴ This means that although, strictly speaking, it is logically conceivable that actors could engage in ‘tacit we-mode cooperation’ and jointly pursue a shared goal without any direct communication between them, it is unlikely that this kind of cooperative interaction would ever be observed in the real world.⁷⁵ Accordingly, we do not further distinguish between two types of ‘we-mode’ cooperation – one involving explicit and one involving tacit coordination – as we did in the case of ‘I-mode’ cooperation. Rather, we recognize only one distinct type of ‘we-mode’ interstate cooperation, ‘thick’ cooperation, which assumes an explicit process of policy coordination to take place.

Second, ‘thick’ interstate cooperation paradigmatically satisfies the intention constraint. Applied to ‘thick’ cooperation, the intention constraint requires that the actors hold an intention to not obstruct the other actors' efforts in achieving the respective shared goal.⁷⁶ Since both actors share the respective goal, they each have an interest in its attainment and will therefore not intend to obstruct the other actors' pursuit of it.⁷⁷

Finally, ‘thick’ cooperation is conceptually distinct from, and cannot be reduced to, the types of interstate cooperation discussed previously, that is, ‘minimal’ and ‘thin’ cooperation. Recall that both these types of cooperation characterize cooperation as essentially being a matter of mutual policy adjustments conducive to the actors' individual goals. They are, therefore, based on what can be called merely individualistic conceptualizations of cooperation. The main reason why any individualistic conceptualization will fail to elucidate ‘thick’ cooperation is that its defining feature – the joint pursuit of a shared goal – does not necessitate that any of the actors' individual goals are facilitated during their ‘thick’ cooperative endeavors. This independence allows for two conceivable cases. In the first, the joint pursuit of a shared goal does not entail any benefits for the actors' individual goals.⁷⁸ Here, merely individualistic conceptualizations would plainly disregard relevant instances of cooperation. In the second, pursuing a shared goal coincidentally does lead to mutual benefits for the actors' individual goals. Although individualistic conceptualizations might, in this case, formally attest cooperation, they would do so for the wrong reason. After all, the reason for which cooperation would emerge would be the coincidental promotion of the actors' individual goals, but not their joint pursuit of a shared goal. Therefore, the possibility of ‘thick’ cooperation ultimately presents merely individualistic conceptualizations, such as those underlying ‘minimal’ and ‘thin’ cooperation, with a dilemma: either they cannot identify instances of ‘thick’ cooperation to begin with, or – in cases where they may identify those – these conceptualizations would provide the wrong kind of explanation.

Real-world examples of ‘thick’ cooperation

Let us now turn to several real-world examples of ‘thick’ interstate cooperation, that is, instances in which states jointly worked toward the attainment of a shared goal. A first set of illustrative examples of ‘thick’ cooperation can be found in recent instances of so-called ‘coalitions of the willing’ jointly working toward a shared goal. The term, Erskine explains, ‘connote[s] a temporary, purpose-driven, self-selected collection of states’ that ‘come together to respond to a specific crisis’. Coalitions of the willing are an increasingly ‘common phenomenon in world politics’, and are most readily recognized when they ‘are summoned and established in cases of military intervention.’ The term first gained prominence when a coalition of the willing under US leadership responded to Iraq's 1990 invasion of Kuwait and fought together in the Gulf War. Other examples involving such a group of states working together for the achievement of a shared goal include the 1999 humanitarian intervention in Kosovo, or the 1999 intervention in East Timor. In each of these cases, the respective states working together formed a group that ‘seem[ed] to possess something that at least resembles a capacity for purposive action’. The mode of collaboration in the coalitions went beyond mere ‘minimal’ or ‘thin’ cooperation. In Erskine's words, any plausible ‘description[…] of actions and intentions’ of the states taking part in these joint endeavors differs ‘in important ways from accounts of individuals acting independently or in isolation’.⁷⁹

In instances in which joint efforts take place in a more formalized and institutionalized setting than that present in the context of coalitions of the willing, ‘thick’ cooperation is even more easily discerned to take place. Consider, for instance, joint Anglo-American efforts to defeat Nazi Germany in World War II. Instead of merely coordinating and pursuing their individual policies in ‘I-mode’, the US and UK leaderships created the Combined Chiefs of Staff which integrated the Western Allies' decision-making procedures. Working within this formal framework, the two states were engaged in far-reaching cooperation, as they collectively worked toward the achievement of their shared goal.⁸⁰ Similarly, in the wake of World War II, the minor Western and Central European powers' joint pursuit of protection from the Soviet threat led to the creation of an institutional framework that ‘provide[d] members with enhanced strategic, operational, and tactical coordination’ and thus paved the way for engagement in ‘we-mode’ cooperation.⁸¹

The arguably most emblematic example of ‘thick’ cooperation at work, however, is the Concert of Europe. Following Napoleon's defeat, at the Congress of Vienna (1814–15) the five European great powers – Austria, Britain, Prussia, Russia, and France – reached the Vienna Settlement which ‘established the great powers’ intention to govern together’. Over the subsequent years, they managed European security affairs together, closely collaborating to avoid another devastating great power war. Importantly, none of the states involved abdicated ‘their independent capacity and authority to act for their own population’. However, ‘the great powers interacted in a markedly different manner from how they had acted in the eighteenth century or before’. Mitzen aptly summarizes what made this ‘experiment in international cooperation’ so unique: ‘the concerting of nineteenth-century Europe can be interpreted as not “just” cooperation based on overlapping individual preferences and in the pursuit of self-interested goals’. In other words, the Concert was not just an instance of ‘thin’ cooperation, that is, of states explicitly coordinating their individual policies for mutual benefits. Rather, it represented ‘a purposive political project articulated and pursued together’. The great powers, forming a collective intention ‘to manage Europe together’, acted together in the joint pursuit of a shared goal.⁸²

A typology of interstate cooperation and its benefits

As a last step in our efforts to provide a comprehensive conceptual toolkit for use in the IR literature on international cooperation, we now present a synthesizing framework of the three different types of interstate cooperation developed above that specifies along which dimensions ‘minimal’, ‘thin’, and ‘thick’ cooperation differ from one another. The remainder of this paper is devoted to explaining this framework as well as highlighting its contributions to IR theory.

Three types of interstate cooperation: a schematic overview

The conceptual framework developed in the preceding sections can be summarized in a matrix that distinguishes between different types of interstate cooperation through reference to the following two dimensions (see Figure 1).

Fig. 1. Types of interstate cooperation.

The first dimension, represented by the rows, distinguishes the different kinds of goals that actors pursue in their cooperative endeavors. Both ‘minimal’ and ‘thin’ cooperation are centrally defined in terms of the actors' individual goal pursuit and therefore located in the upper row. In contrast, ‘thick’ cooperation is essentially a matter of the actors' shared goal pursuit, thus located in the bottom row, and in this sense conceptually distinct from both ‘minimal’ and ‘thin’ cooperation. Put differently, ‘minimal’ and ‘thin’ cooperation take place in I-mode, which is to say that the actors seek to realize their own individual goals, whereas ‘thick’ cooperation entails a we-mode of interaction.

The second dimension, represented by the columns, distinguishes the kind of coordination process involved in the actors' goal pursuit. ‘Minimal’ and ‘thin’ cooperation are primarily distinct from one another in that ‘minimal’ cooperation involves a tacit process of policy coordination, whereas ‘thin’ cooperation entails an explicit process that includes procedures such as explicit negotiation or bargaining. As discussed above, ‘thick’ interstate cooperation also involves an explicit process of policy coordination since it is hardly conceivable that the kind of ‘we-mode’ cooperation it entails would ever take place through a tacit coordination process alone. Accordingly, the bottom-left cell in our two-by-two matrix is left empty.⁸³

Note that our integrative schema highlights that there is a clear sense in which all of the included types can be considered united, that is, instantiating the same general phenomenon of cooperation. At bottom, they all characterize cooperation as deliberative attempts by states to facilitate their (individual or shared) goals through policy coordination. Although this clearly seems reminiscent of the conventional account as originally put forth, our typology goes beyond it by vividly illustrating that there exists a variety of distinct senses of cooperation that differ in important, theoretically relevant ways from one another. As we will now show, it thus provides IR scholars with considerable theoretical leverage.

The typology's contribution to IR theory

Political methodologists have long recognized that descriptive typologies can crucially contribute to theorization in the social sciences.⁸⁴ In fact, within the discipline of IR, similar conceptual frameworks on other foundational phenomena have led to important theoretical advancements and have done much to improve our understanding of international politics. Notable examples include John Vasquez’ disaggregation of the concept of war into several different subtypes as well as Thomas Wilkins' distinction between different types of interstate alignment.⁸⁵ Similar to these studies, our three-way typology of interstate cooperation aids theorizing about international politics in two main ways: it facilitates conversation across traditionally rigid theoretical boundaries, and it allows for better theorization of the causes of cooperation through disaggregation of the overarching concept.

First, the conceptual tools developed here facilitate conversation across different branches of IR literature and allow scholars to explore fruitful ‘links [between] previously unrelated theories and strands of research’.⁸⁶ The study of international politics today often takes place within strictly delineated theoretical camps and without much interaction across them. Indeed, at times it appears as if scholars treat different theoretical approaches to the study of IR as incommensurable. Consequently, in Thomas Kuhn's words, ‘the proponents of competing paradigms’ have ‘fail[ed] to make complete contact with each other's viewpoints’.⁸⁷

Against this backdrop, our typology of interstate cooperation facilitates productive conversation between different bodies of IR theory by providing them with a common, more specific conceptual language vis-á-vis interstate cooperation. Specifically, our typology allows scholars to contrast their differing viewpoints on important instances of interstate cooperation more precisely and effectively. Take, for example, the literature on the Concert of Europe. Much of the theoretical disagreement there finds its roots in competing judgments of what kind of interaction took place between the European great powers in this period. For many balance-of-power theorists, the Concert was merely an instance of ‘thin’ cooperation in which the great powers, for a brief period of time, explicitly coordinated their policies in such a way as to facilitate the attainment of their overlapping individual goals.⁸⁸ Social constructivist theorists, on the other hand, employ a vastly different outlook on the Concert that aligns with the interpretation we presented above, noting that ‘phenomenologically, it [the balance of power story] overlooks the fact that the leaders intended to manage Europe together’.⁸⁹ In other words, like us, they observe an instance of ‘thick’ cooperation. The fine-tuned conceptual toolkit provided here, thus, allows theorists to locate important sources of disagreement and specifies along which dimensions their perspectives on the world differ from one another. A more generalist conceptualization of interstate cooperation would struggle to lay out the fault lines of disagreement in a similarly specific manner.

Conversely, our typology also helps scholars uncover previously unrecognized theoretical common ground. As discussed above, both social constructivists working on global governance and rationalist scholars working on economic and political integration at times allude to cooperation in the ‘we-mode’ as an important type of state interaction that differs in kind from other, less involved forms of cooperation.⁹⁰ Yet, since the existing conceptual literature on interstate cooperation lacked the tools to delineate ‘thick’ cooperation as a distinct phenomenon in international politics, this common theoretical interest has not been recognized so far, and consequently these bodies of IR theory have continued to develop largely independently of each other.

Second, by disaggregating the concept of ‘cooperation’ into three subtypes, our framework allows for better theorization about the causes of interstate cooperation. As Derek Beach and Rasmus Pedersen explain, the ‘population[s] of cases’ social scientists are interested in can often be ‘very causally heterogeneous’. This is to say that ‘a given cause might have many different effects across different cases, in terms of either the same cause producing a different outcome or the same cause being linked to the same outcome through different causal mechanisms’. Scholars run the danger of focusing on such causally heterogeneous populations of cases, Beach and Pedersen warn, when ‘ignoring causally important differences between cases by lumping them together using excessively broad definitions of concepts’. This issue can be addressed by disaggregating the overarching concept into several sub-types and engaging in ‘typological theorization’.⁹¹ Notably, Vasquez's well-known study of the causes of war rests on this very idea and utilizes it with great success. As he notes, a review of the historical and scholarly record suggests that ‘there are different kinds of wars and … a theoretical typology of wars … would be conceptually useful in explaining findings’.⁹²

Given that the different forms of cooperation highlighted here significantly differ in character, it appears likely that conceptual disaggregation will provide similarly important theoretical leverage in the debate about the causes of interstate cooperation. After all, considering that ‘minimal’, ‘thin’, and ‘thick’ cooperation each entail a unique kind of interaction between states, it is very well possible that causal theories which, for example, are apt at explaining the emergence of ‘thin’ cooperation might struggle to explain instances of ‘thick’ cooperation, or vice versa. Our framework enables the kind of nuanced comparison of ‘concrete cases, both cross-sectionally and over time’, that is necessary for engaging in such type-specific causal theorization.⁹³ As an example, it allows scholars to discern that interstate cooperation in Western and Central Europe has vastly varied in kind over time: at times only sporadically coordinating their actions in tacit ways, the European states at other times collaborated in a much deeper sense and even managed their security environment together.

Conclusion

It is understandable that IR theorists have been content to draw on the conventional account when conceptualizing interstate cooperation. After all, it represents a well-designed framework capable of identifying many instances of cooperation in the international arena. Yet, although the substantive literature on the causes of international cooperation has evolved over time and scholars have come to implicitly employ various understandings of cooperation in their works, an explicit debate over its meaning never materialized. Whatever discussion occurred about the discipline's conceptual foundations did so only latently and without much dialog across theoretical traditions. In consequence, scholars missed out on an opportunity to sharpen their conceptualizations of one of the most important outcomes in international politics and gain crucial theoretical leverage through conceptual refinement.

In face of this status of the literature, the present paper has begun to reconstruct and systematize the conceptual treatment of interstate cooperation in IR research. As an important preliminary step, we first demonstrated that any plausible understanding of cooperation must evoke a basic form of intentionality on the part of the states involved. In particular, we proposed the intention constraint as a theoretical tool that allows us to exclude what would amount to truly ‘accidental cooperation’ from our conceptual scope. On this basis, we then argued that the IR literature on interstate cooperation has implicitly come to identify three distinct, plausible senses of interstate cooperation – ‘minimal’, ‘thin’, and ‘thick’ cooperation – for all three of which we provided an explicit definition and illustrated their unique character through reference to a series of real-world examples. Finally, to supply scholars with a unified framework that integrates this apparent conceptual diversity, we developed a typological schema that specifies the relationship between the three types of cooperation and highlighted its multifaceted contributions to better theorization in IR.

However, many questions remain unanswered, and much work remains to be done. There appears to exist ample opportunity for a productive dialog across theoretical boundaries on a host of further research questions. First, does the present typology capture the main understandings latently employed in the substantive IR literature, or are there other types of interstate cooperation not identified in our framework? Second, are additional restrictions needed besides the intention constraint to effectively exclude implausible instances of cooperation from our conceptualization? Third, would IR theory benefit from even further, more fine-grained conceptual disaggregation? For instance, are there several unique forms of ‘thick’ cooperation? Finally, are there promising ways to conceptualize interstate cooperation that employ a fundamentally different theoretical approach, that is, that do not take mutual policy adjustments for the facilitation of individual or shared goals as a conceptual starting point? Ultimately, we hope that this paper encourages IR theorists to pursue some of these and countless other avenues for future research on this foundational issue.