The three schools of thought regarding international cooperation [regimes] - interest-based theories, power-based theories, and knowledge-based theories - provide numerous insights from which it is possible to draw some general findings about cooperation. The basic question the authors examine is how different schools of thought analyze and explain "What accounts for the instances of rule-based cooperation in the international system?"

The following table is presented by the authors (p6):

A regime exists whenever agents cooperate. Formally a regime is the "implicit or explicit principles, norms, rules, and decision-making procedures around which actors’ expectations converge." Regimes can be both formal and informal, and can include institutions as well as organizations.

Interest-based Theory

Interest-based theories of cooperation focus on the ability of self-interested rational agents to overcome collective action dilemmas, i.e. situations where cooperation avoids suboptimal outcomes for the cooperators. Agents are considered to be rational utility-maximizers with given preferences. Attention is given to the role of regimes/institutions in shaping preferences and facilitating cooperation.

Interest-based theories note the spillover effects of cooperation (functionalism). Because of the costs of creating and maintaining institutions, establishing cooperation in one issue-area can result in solutions that can then be reused in other issue-areas. Cooperation is a result of institutional bargaining (contractualism) and results in negotiated agreements and commitments. Compliance with or defection from negotiated contractual agreements has reputational effects.

There are two primary approaches to interest-based cooperation:

• situation-structural - focuses on the game-theoretic properties of the problem.
• problem-structural - focuses on the issue-area, or topic, of the problem.

Situation-Structural: The situation-structural approach involves "interpreting different kinds of regimes as collective responses to the functional requirements of different kinds of collective action problems" (p 45). Models from game theory - such as Prisoner’s Dilemma, Coordination Game, and Assurance Game - are commonly used. Collaboration requires sanctions and compliance, whereas coordination merely requires agreement. Coordination - such as deciding which side of the road to drive on or allocating the frequency spectrum - is self-enforcing because participants have no incentive to defect. Also, various games have different "second order" dilemmas regarding costs of implementation and enforcement: collaboration is the most costly and the Assurance Game is the least costly.

Problem-Structural: The problem-structural approach involves observing the nature of the issue-area of the problem. In this analysis there are two modes of conflict with different likelihoods of cooperation in each:

• Dissensual conflict
• Conflict about means: when goals are agreed upon but methods are not. The possibility of cooperation is medium.
• Conflict about values: when parties desire different outcomes. The possibility of cooperation is low.
• Consensual conflict
• Relative gains: when your benefit is relative to the benefits of others. The possibility of cooperation is low.
• Absolute gains: when everyone benefits from a solution. The possibility of cooperation is high.

Some factors can aid in the possibility of cooperation. When agents operate under a "veil of uncertainty" regarding benefits and costs, they will often cooperate more readily because there are no known distributive issues to argue over. Exogenous shocks and public crises/outcry can spur cooperation on an issue. The key factor is that the issue-area regarding a given problem must be amenable to a contractual solution in the first place.

Power-based Theories

Power-based theories of cooperation focus on the importance of relative gains and security concerns to otherwise rational agents. The distribution of power and the presence of anarchy (the absence of an authority to enforce contractual obligations) are paramount. Because these concerns never change and are external to the agents involved, power-based theories are predominantly static and positivist.

There are three power-based theories of international cooperation:

• Hegemonic Stability Theory
• Power-based Research Programme
• Realist Theory of Cooperation

Hegemonic Stability Theory

A hegemon is a powerful agent who provides public goods because it has the self-interest and the capacity to supply them. This provision generates free riders. According to hegemonic theory the weak exploit the strong. Hegemony can be coercive (imperialist) or benevolent (leadership).

Hegemons are necessary to shoulder the costs of rulemaking and enforcement (second-order cooperation dilemmas). In return, they generally set the rules and others adjust. Mancur Olson and Duncan Snidal have noted that small groups can provide public goods by cooperating and sharing costs, instead of relying on a single hegemon. In addition, hegemons can vary according to issue-area (the environment, nuclear weapons, etc.)

Power-based Research Programme

According to power-based theories cooperation does not result in mutual adjustment at all but instead requires the less powerful to adjust to the more powerful. In addition, power differences shape the following:

• Who gets to play the game?
• What are the rules?
• What are the payoffs?

Power-based analysis suggests that the Prisoner’s Dilemma is not the best game by which to study cooperation when power is a factor. In the Prisoner’s Dilemma there is one "best" solution and the challenge is for the various agents to arrive at it. By contrast, in Battle of the Sexes, the optimal outcome is different for each player, thus there is fundamental disagreement over what constitutes the "best" solution (mathematically there is no one best solution).

As a result, cooperation and institutions merely serve the interests of the powerful. Powerful players extend their power through these means. Because differences in the distribution of costs and benefits always exist, even under conditions of absolute gains not everyone gains equally. Oran Young has questioned the reliability of assuming that structural power is translatable into bargaining power regarding outcomes.

Realist Theory of Cooperation

The Realist theory of cooperation attempts to explain cooperation given states’ overwhelming concern with security, independence, and autonomy. It is not merely relative gains that are a concern but a systemic intolerance for relative losses. All acts could result in the destruction of the agent, so power asymmetries trump all other concerns. In this scenario, absolute gains just do not exist. There is always the concern over "who will gain more?" The result is "defensive positionalism," or reluctant cooperation, wherein agents will cooperate only if they feel it is absolutely necessary. Rationality, in this case, is constrained by fear of destruction and the presence of anarchy.

For Realists, institutions matter but only because they facilitate the necessary stabilizing exertion of power: payoffs to other agents, sanctions, and norms of reciprocity (that make accepting relative gains losses in the now or on a particular issue easier in expectation of compensation on other issues or in the future). With power, cooperation is rare at best, but without power it is impossible.

Knowledge-based Theories

Interest-based theories of cooperation focus on the ability of self-interested rational agents to overcome collective action dilemmas, i.e. situations where cooperation avoids suboptimal outcomes for the cooperators. Agents are considered to be rational utility-maximizers with given preferences. Attention is given to the role of regimes/institutions in shaping preferences and facilitating cooperation.

Knowledge-based theories (cognitivism) focus on the way in which knowledge - and in particular inter-subjectively shared knowledge and beliefs - shape agents’ behavior and identities. Norms are of major interest to knowledge-based theories.

There are two cognitivist variants:

• "Weak cognitivism" is concerned with the origins of rational actors’ behavior.
• "Strong cognitivism" is concerned with the origins of actors’ understandings of Self and Other.

Weak Cognitivism

Weak cognitivism assumes rational actors but instead of taking preferences as given, problematizes preferences and investigates the origins of agents’ interests, as well as the impact of norms on preference formation.

Weak cognitivism sees itself as complementary to other approaches. The role of knowledge is central, including ideas and learning. Because knowledge is filtered by interpretation, preferences become fluid as knowledge changes. Because knowledge is primary, knowledge-shapers are powerful influences. Epistemic communities inform policy-makers about currently accepted shared understandings, from intersubjectively held ideas to problem definitions and concerns.

Thus the idea landscape acts as a "road map" from which agents choose their routes. Learning is possible and subsequent course-correction as well. Furthermore, the institutionalization of knowledge shapes agents’ preferences. Institutions contribute to consensus through knowledge and information sharing.

Strong Cognitivism

Strong cognitivism, on the other hand, dispenses with rational actors in favor of a sociological model of behavior. Agent’s perceptions of their own and others’ identities and roles are central objects of study. Agents are role-players, not utility maximizers.

Strong cognitivism positions itself as an alternative to other approaches. This "sociological turn" investigates how knowledge and beliefs constitute agents and make possible both power and cooperation. Agents’ very identities exist only by virtue of shared understandings. Groups and institutions define who we are and what behaviors are possible and meaningful. Strong cognitivism stands in opposition to atomistic and positivist attempts to understand cooperation.

Because there is no behavior without prior socialization, strong cognitivism suggests that agents act according to a "logic of appropriateness" rather than a utilitarian "logic of consequences."

For strong cognitivists, sanctions and cost/benefit analysis, i.e. self-interest, is insufficient to explain cooperation. Norms both regulate behavior as well as constitute agent’s identities. This means that norms constrain but also affix meaning to certain actions. An example is a game of chess, wherein the rules must be coherent and accepted before any meaningful moves can be made. As a result, compliance is not the only indicator of cooperation. Strong cognitivists also point to the justifications used by a defecting agent as well as the response of other agents when rules and norms are violated. Generalized norms of cooperation and reciprocity create a web of meaning within which behaviors are contextualized, interpreted, and evaluated.

There are four schools of thought regarding cooperation:

• The power of legitimacy - studies the society of states and its rules.
• The power of arguments - studies communicative rationality.
• The power of identity - studies role-specific understandings of self and other.
• The power of history - studies stabilizing v. critiquing the ‘world order.’

For strong cognitivists, neither human agency nor social structures should be given ontological priority; they share a "codetermined irreducibility."

The power of legitimacy

• States comply with rules seen as legitimate even when it seems not to be in their self-interest.
• For agents in a system, insuring the validity and stability of the system is the first priority.
• Reputation and trust among stakeholders maintain the system.

The power of arguments

• Understanding discourse is essential to understanding cooperation.
• "Strategic action" attempts to control others, whereas "communicative action" attempts to convince them.
• Common understandings are necessary to agree on goals.
• Common understandings provide starting points for discussion.

The power of identity

• Constructivism is the study of socially constructed identities in political science.
• Understanding cooperation among egoists is insufficient to understanding cooperation in general.
• The evolution of cooperation leads to evolution of community.
• Diffuse reciprocity creates cooperation and cost-sharing without direct incentives.
• Cooperation is self-stabilizing when agents who cooperate for selfish interests come to identify themselves as "cooperators."
• Neither social structures nor social identities exist independent of interaction and reproduction (practices).

The power of history

• The current world order is a product of western history and ideology.
• Identities and norms are constructed, but by powerful elites and historical forces.
• These forces are responsible for the dissemination and control of ideas.
• Certain societal actors benefit from prevailing modes of production, accumulation, and dominance.
• ‘Hegemony’ is a set of ideas promulgated by a powerful hegemon in the interest of various elites.
• Cooperation is really collusion designed to reinforce or stabilize the existing world order.
• Elites must compromise to maintain consensus and the stability of the world order.
• Stability requires the marginalization of any radical conceptualizations or alternatives.
• Seeming alternatives take "the existing order as given, as something to be made to work more smoothly, not as something to be criticized and changed."
• Knowledge is always an ideology that works in favor of some groups and not others.

Conclusion

Cooperation is problematic. Whether cooperation is desirable or not, and why, as well as how it can be promoted or prevented depend on various assumption about the nature of agents and their interactions. Nonetheless, it is possible to uncover the processes that foster cooperation within each school of thought.