The Causal Structure of Power

This paper attempts a useful formalization of a central concept in political theory and the social sciences – power. I interpret power causally, and so the formal analysis of power is developed using the frameworks of causal structural equation modeling and graphical causal modeling. Power relations can be analysed at various levels of abstraction, depending on the form of power one is interested in. This determines the appropriate selection of variables for the causal model, as well as the appropriate causal links between variables. Power relations at a certain level of abstraction can be detected if a causal signature characteristic of one party having power over another is visible in the causal model appropriate to that level. I use the neo-Republican conception of domination as a basis for articulating the causal signature of power, and develop formal definitions of one party having direct power over another (i.e. being able to directly affect their ability to act in certain ways) as well as one party having indirect power over another (i.e. being able to ensure that their actions do not result in the desired outcomes). I also examine the concept of structural power, where power is supposed to inhere in the network structure of the causal system itself, rather than in the causal relations between agents. I develop a formal account of how one might evaluate structural power by comparing a causal model to other “nearby” causal models, in terms of the range of desired outcomes achievable by an agent in these models. Causal systems that restrict an agent’s range of action in comparison to nearby alternatives can plausibly be understood as embodying structures of power or domination.

---

Tarun Menon is an Associate Professor in the School of Arts and Science, Azim Premji University, Bangalore. His work focuses on the nature of explanation in the natural and social sciences, particularly causal and probabilistic explanation. He also works on the role of values in scientific inquiry, and on the interaction between science and public policy.