Political pluralism: neoinstitutionalism redefining the link between resources and outcomes

The rebirth of power analysis in IR/IPE has to do with what Baldwin called the ‘paradox of unrealised power’. In a regional conflict, the major power of the world was apparently unable to lay down the rules and had to accept a humiliating military and political defeat against the Vietcong. Some scholars tried to explain this paradox away by identifying the lack of ‘will’ on the side of the US to use these resources, i.e. so-called ‘conversion failures’. An explanation based on alleged conversion failures implies that the war did not show the relative weakness of the US (in spite of its military capabilities), simply its unrealised strength. But such an explanation can re-arrange ex post any outcome to suit any power distribution. In other words, such an explanation has the scholarly implication that the significance of power cannot be empirically assessed at all. As so often, the trouble with this type of power analysis is not that it is wrong, but that there is no way it can go wrong.

Neoinstitutionalist analysis has opted for another approach. Not having a single international power structure and no lump concept of power at hand, this analysis starts from a conception of international politics which is segmented into issue-areas and/or more contingent. This understanding of politics leads to two different types of power analyses. As I will argue, these two very valuable attempts have problems of their own. They can stick to a *behaviouralist explanations in which, however, power plays no longer a central role, or propose a causal, but situation-specific concept of power which is however inconsistent with behaviouralist explanations, since the power variable is no longer embedded in a  general theory.

One conceptual way out consisted in accepting the apparent lesson of the Vietnam War. Consequently, control over resources, even issue-area specific ones, does not necessarily translate into control over outcomes. Power does no longer function as a determining cause.  While discussing *hegemonic stability theory, Robert Keohane shows empirical anomalies both of the ‘crude’ and purely systemic version, which derives the existence of regimes solely from shifts in the distribution of power, and of a  more sophisticated version in which leadership, an aspect of the unit-level, figures prominently. In Keohane’s analysis, the determination in the explanation shifts from interests defined in terms of the distribution of power, to rational choice made on the basis of given interests defined in terms of power, expectations, values, and conventions. Hence, only predictions of a very limited kind are possible – with a secondary role for power: power is assumed to be one of the two primary motivations for rational action and the distribution of power is part of the actor’s definition of interests. Hence, power no longer serves as main determinant for outcomes.

Another solution to the paradox of unrealised power has been proposed by David Baldwin who has taken the issue of power fungibility most seriously. His approach keeps a strong causal role for power by further specifying the relational and situational context that defines which policy instruments can count as actual power resources in the first place. Hence, any assessment of power independent of such situational factors is erroneous (and here goes *realism); any generalisation beyond such cases, is contingent and to be individually established (and here goes *behaviouralism).

Baldwin’s conception is shaped by his relational understanding of power. In Keohane and Nye, this relational aspect is present to the extent that power derives from relations of vulnerability interdependence defined by the high cost to be paid for substituting these resources. Baldwin, on the other hand, begins with individual, not collective actors, with relations, not resources. Baldwin’s preferred example for the relational character of power might serve as illustration. If a suicide candidate is threatened with a gun to choose between his money and death, he might not feel threatened at all. The gun-bearer has no power over the suicide candidate in this relation.

In other words, power comes not (only) out of the utility attached to resources, but exists through the actual value systems of human beings in their relations with each other. The major difference to utilitarian action theories is that personal value systems cannot be simply assumed in the empirical power analysis. Instead, the researcher has first to analyse the value systems of the interacting parties in order to establish whether there are power resources in the first place. For this reason, Baldwin insists that one can only study power, if understood as a causal variable, in well circumscribed ‘policy-contingency frameworks’. The context then specifies the scope and domain of power, as well as the norms and values within which interaction takes place. Once circumscribed, power can be defined as a causal antecedent to an outcome.

The price for this, however, is that power analysis must potentially become very narrowly circumscribed to particular instances, where no prediction is possible. Whereas Keohane’s institutionalist move kept, however limited, predictive capacity of a theory, based on rational choice and not on power, this second move saves a central causal role for power at the price of predictability in IR/IPE. Keohane, by keeping contextual analysis to a fairly general level, can make probabilistic predictions on behavioural outputs, but not outcomes, leaving power as a secondary variable. Baldwin’s attempt to keep power as a central causal determinant forces him to include all the necessary information into a contextual analysis, thereby making clear predictions on the basis of power impossible.

Besides the relational concept of power, the lack of fungibility, and the potential need to have a very elaborate contextual analysis prior to the assessment of power, it is the unavoidably counterfactual character of power which makes Baldwin’s approach of power both causal and yet little generalisable. Counterfactuals are no easy matter in empirical sciences. Since power relations involve getting somebody else to do what he or she would not have done otherwise, one can assess power ex ante on the basis of what this foregone action would have been. With that move power analysis stops to be a central element of building middle-range theories, but instead appends itself to an action theory to be able to make meaningful expectations about ‘normal behaviour’.

Unlike in realism, power is no longer the motivational basis and essence of international action for neo-institutionalists. It is a component in a wider action theory. Keohane’s solution points into the direction of a rationalist neo-institutionalism, Baldwin’s less generalisable, contingent and situational solution to historical institutionalism.

Steffano Guzzini, Power