Manipulability, Causal Inference and Interpretive Methods in Political Science

• In Event: Foundational Concepts and New Directions in Qualitative Research

Abstract

An interventionist manipulability notion of causality underpins standard ways of engaging with the key task of causal inference in political science. It provides the basis for the Neyman-Rubin causal model, informing much experimental and statistical work in the discipline. The key argument of this paper is that there are different and rarely acknowledged ways in which the manipulability notion of causality can be applied. The paper demonstrates that, contrary to most existing research, manipulability perspectives may also supply a robust basis for approaches that apply qualitative and interpretive methods to establish causal accounts of events and processes. It does so by first unpacking the notion of manipulability tied to the view on human beings as engaging causally with the material and social world. It then proceeds to identify the implications of this notion for perspectives on causal inference that use qualitative and interpretive methods, providing concrete examples of applications for political science. The paper thus opens new ways of thinking about methods for causal inference that rely on a manipulability perspective and offers a shared basis on which both interpretive causal research and experimental research may rely.

Author

• Ludvig Norman, Stockholm University