REFINING “COOPERATION” IN THEORY
By Amien Kacou
According to Stephen Walt, international relations theory has become—due in large part to its growing professionalization—increasingly irrelevant to international relations policy. And, judging from his remarks elsewhere, the problem seems to be even more pronounced in the case of formal international relations models. However, perhaps when it comes to international cooperation policies in particular, the most prominent strategic international relations models have been especially irrelevant, due not only to the growing professionalization of the field but also to their inherent (and unjustified) “competition bias.”
This essay provides a brief description of that bias and suggests that its persistence is due not only to the tradition of realism in international relations but crucially also to a failure by prominent non-realist international relations theorists to treat trade on a par with security and greed as a fundamental motive of states. Charles Glaser’s recent rationalist account of international relations provides a case in point, even though he concludes, contrary to the traditional accounts of structural realists and offensive realists, that international anarchy needs not automatically generate a tendency toward competition between rational security seeking states. Glaser’s conclusion is based on the argument that variations in material variables (between offense and defense) and variations in information about state motives (between security and greed) can either avoid or eliminate security dilemmas (which are the only source of conflict between security seekers: situations in which their pursuit of security threatens that of others).
More specifically, on the issue of material variables, Glaser agrees with defensive realists that states can sometimes distinguish and assess the “offense-defense balance” in military capabilities—as their opportunities to attack, or defend against, one another will vary, not only with the nature of their military items but also with the circumstances of their use. In particular, when the “balance” swings toward defense, then security seeking states are more likely able to reach their basic objective without having to threaten, or otherwise compete against, one another.
We could add further that in the ideal case when effective purely-defensive items (e.g., pure “denial” items or pure shields) and effective purely-offensive items (e.g., pure “punishment” items or pure weapons) are equally available, then we should expect rational security seekers to prefer the former. This is mainly because, unlike pure punishment items for instance, pure denial items can provide protection without sending the wrong message and triggering a security dilemma. And it is also because, as I would argue, pure denial items are more directly and robustly protective than pure punishment items, considering that only the former can actually mitigate the damage of attacks (i.e., reduce the impact of attacks in the present) whereas either type of capability can deter attacks (i.e., reduce the likelihood of attacks in the future). Besides, the credibility—and thus the effectiveness—of the threat of punishment as a rational deterrent may well be problematic.
Likewise, on the issue of information about state motives, Glaser adds that, even assuming that some states are greedy, security-seeking states that are informed that some other states are also security seekers are further more likely able to reach their basic objective without having to compete, depending in some degree on their likelihood of interacting with the latter. Hence, for example, in the ideal case when interacting security seeking states are certain of each other’s type (assuming further that no material scarcity is creating a contingent conflict of interests), then competition is extremely unlikely (and increases in the amount or offensive potential of military capabilities are less likely to change this situation). And, conversely, when interacting security seeking states are uncertain of each other’s type (i.e., when they believe that some states might be greedy), then, Glaser argues, they should balance between competitive and non-competitive strategies (presumably, in some consistent proportion to their likelihood of encountering a greedy type, on the one hand, or a security seeker, on the other).
Moreover, Glaser seems to suggest that some inclination toward security is more likely to be shared across states than any inclination toward greed, considering his argument that some minimal baseline of security is required in the pursuit of other motives—as most greedy states are likely to value what they already possess.
Nevertheless, while this account of information about motives might seem (perhaps naively) to suggest that non-competitive strategies are as likely as, or even dominant over, competitive strategies for rational security seekers with equal odds of encountering a state of either type (i.e., greedy or security-seeking), that is, in fact, the opposite of what the most common and basic formal models of international relations seem to indicate. This is due not only to the fact that, as Glaser accurately observes, prominent traditional formal models tend to assume that states are greedy—for example, applications of the Prisoner’s Dilemma or the Game of Chicken assume that states would prefer dominating one another to cooperating with one another. Instead, the point is that merely reversing this preference (i.e., assuming that states prefer cooperation over domination—as in applications of the Stag Hunt) does not seem sufficient to completely reverse the rational dominance of competitive strategies (even though experimental subjects in repeated Stag Hunt simulations have been shown to cooperate on the very first round of play). Surely, the Stag Hunt, in which both mutual cooperation and mutual competition are pure Nash equilibriums (as defined below) while mutual cooperation offers the highest actual payoff (by definition) is less competitive than the Prisoner’s Dilemma, in which mutual competition is the only pure Nash equilibrium. (Nash Equilibriums occur when both players in an interaction play their best response to one another’s given strategy—and thus when neither player stands to gain by unilaterally switching his or her strategy.) However, even in the Stag Hunt, mutual competition seemingly continues to be the rationally-dominant strategy because competition is the least risky (or most robust) option—whereas the expected payoff of competition (i.e., the sum of potential payoffs for competing, given the entire set of possible interactions) is at least equal to that of cooperation.
As a result, assuming for our purposes that the offense-defense balance were impossible to distinguish or assess, it might continue to look as if, more in accordance with traditional realist accounts, the very nature or structure of international interaction introduced just enough risk to create a rational competition bias. This would be the case irrespective of the fact that cooperation equilibriums could still emerge in repetitions of the Stag Hunt, through adjustments of either information variables or preferences based on past interaction and context. Such adjustments could occur either when a favorable change in information about state motives takes place among security seekers (as Glaser would expect), or, more in accordance with structural constructivist accounts, when collective identities or similar social structures take form—structures capable of constraining states or even changing their motives to the point of introducing trust in Stag Hunt dilemmas in groups of states.
However, it is also possible to build two-player models of international relations in a manner that makes competition even less attractive from the beginning—in particular, by refining the concept of cooperation in order to allow that rational states might get an even higher payoff from cooperation equilibriums than they do in the Stag Hunt. In other words, it is possible to build a model that specifically negates (or at least comes closer to negating) Kenneth Waltz’s assumption that states are “more dangerous than useful to one another.”
Such a model would require that we sacrifice neither empirical relevance nor the assumption that states are egoists (i.e., self-interested). In fact, it would seem both analytically and historically sound to consider that selfish states have among their fundamental motives not just greed (whose ideal might be expansionism or unilateral aggression) or security (whose ideal might be isolationism, “live and let live,” or mutual indifference) but also trade (i.e., the selfish ideal of mutual gain via comparative advantage, as formulated by David Ricardo). In other words, whereas the application of Stag Hunt models to international relations seems in effect to systematically reduce the concept of cooperation to a coarse concept of noncompetition, it seems both possible and justifiable to consider the presence or absence of insecurity and the benefit of cooperation as three distinct but equally possible outcomes of the fundamental structure of international interaction.
Accordingly, on the one hand, we would describe security seeking states interacting in a Stag Hunt application in accordance with the following order of preferences: mutual indifference (with a payoff equal to 3) is preferred to competing militarily against states that are indifferent (with a payoff equal to 2), which is preferred to mutual military competition (with a payoff equal to 1), which is preferred to being indifferent to militarily aggressive states (with a payoff equal to 0). (Here, I have simply substituted “indifference” for “cooperation.”)
Likewise, on the other hand, we could build on this basis to describe states capable of successfully interacting through trade in accordance with the following order of preferences: mutual gain in resources from trade (with a payoff equal to 4) is preferred to mutual indifference or to being indifferent to states that make costly trade overtures (with a payoff equal to 3), which is preferred to making unrequited costly trade overtures or wasting any military capabilities on competition against benign states, whether these are indifferent or making trade overtures (with a payoff equal to 2), which is preferred to mutual military competition (with a payoff equal to 1), which is preferred to being indifferent to militarily aggressive states (with a payoff equal to 0), which is preferred to making costly trade overtures to militarily aggressive states (with a payoff equal to -1). Here, for interacting trade seekers as for interacting security seekers, mutual competition would appear to remain the most robust pure Nash equilibrium. However, unlike interactions between security seekers, interactions between trade seekers involve three pure Nash equilibriums (instead of two), which are (in order of payoff magnitude): trade; mutual indifference; and mutual competition. And, moreover, also unlike the case of security seekers, in the case of trade seekers, the expected payoff of competition over the entire set of possible interactions, or (2+2+1)/3 = 5/3, is equal to that of trade overtures, or (4+2–1)/3 = 5/3, but, crucially, inferior to that of indifference to other states, or (3+3+0)/3 = 6/3 = 2.
This very simply confirms that, as we might expect by intuition, states that can mutually benefit from trade should lose even more from the selection of the equilibrium of mutual competition than is typically assumed in the application of the Stag Hunt as a model of international relations—and that, therefore, all fundamental strategic theories of international relations that take into account as “structural” not just the motives of security and greed but also that of trade should be more optimistic (or at least generally agnostic) on the general prospect for international cooperation.
Considering, then, that, as discussed above, refining the concept of cooperation in order to distinguish between the motives of trade and security is both analytically and empirically sound, then the question becomes: why do rationalist theorists like Glaser continue to rely on the coarser dichotomy between competition and cooperation?
|
| Comments in Chronological order (2 total comments) |
|
|
