Resources As Measuring Power

Probability and Opportunity Costs

The main move in Harsanyi's articles was away from the assumption we can measure power in [Page 572]terms of probability to the idea that we can measure the strength of an actor's power in terms of opportunity costs. Robert Dahl takes up the former position. He thinks power can be measured in terms of the number of people one has the ability to influence (extension), what they can be influenced to do (scope), and the probability of success (amount). Probability is the important variable here because the scope and extension of an action will also be subject to a probabilistic calculation. For instance, the extension and scope of a media outlet's power will rest entirely on how their information is interpreted and received, yet the reaction from the public or their readership can never be anticipated perfectly. The greater the probability that their publication's scope and extension will be broad, the greater their power. The Shapley—Shubik Index, for example, attempts to measure the power of a voter by the probability that that voter will be the pivotal voter in an election outcome. (The pivotal voter can be defined as the last member counted of a winning coalition.) The greater the probability the voter will be pivotal (i.e., the smaller the group of voters, assuming each vote is equally weighted), the more power each voter will have to affect the electoral issue.

It is important to note that Dahl does in fact include an actor's resources (power base) and the means of power (specific actions like threats and offers) as constituents of the power relation. In the Shapley—Shubik Index, one's vote can be seen as a resource. Nevertheless, these constituents do not directly factor into measurement. Each vote is counted as a variable in a probability calculation of power, not as an indicator of power itself. It can be inferred from the means and resources an actor has that that actor will probably be able to influence a lot of people in a lot of ways. However, there is no way these resources should measure power directly. It seems here that there was—and still is—a concern with the vehicle fallacy, identified by Peter Morriss, which is to equate power with its vehicle. The vehicles of power are the means and resources required to wield it.

However, if one wants to quantify power in a satisfactory way, Harsanyi believes more variables are required for its measurement. Dahl's account seems to leave two big holes, one descriptive and one explanatory. Descriptively, it does not account for a situation where an actor could technically perform an action but refrains because the costs attached to the action are too great. Intuitively this should indicate a lack of power, but it does not according to Dahl's formal definition: the scope, extension, and amount remain the same. Leaders of great armies have vast power in extension, scope, and amount. Suppose, however, that they are captured by enemy forces and held at gunpoint with the ultimatum to radio their troops to pull back to a strategically undesirable position or take a bullet to the head. It would be misleading to say they have just as much power now as before being captured. We must factor opportunity costs in to power analysis to describe this situation properly. The commander does not have the power to refuse because the costs of noncompliance are too high.

...