NTU Philosophical Review

Issue 26, March 2003

Szu-Ting Chen (陳思、廷)
Pages 53-90

The Distinction between Causation and Invariance and Its Implications for the Philosophical Discussion of Economic Theorizing

Recently, certain philosophers have argued that an explanatory relation is a causal relation that is fundamentally about the invariance of a relation between variables of interest under intervention-i.e., about a manipulable invariant relation. This manipulative theory tends to reduce a causal relation to a manipulable invariant relation. By explicating a case from contemporary econometrics, this paper argues that a manipulable invariant relation can be obtained only when the causal chain or causal structure of the targeted relation is free from disturbing influences. In other words, a manipulable invariant relation can be regarded only as a special kind of causal relation, and so the notion of invariance can never replace the idea of causation. This paper also shows that the distinction between causation and invariance has methodological import concerning the philosophical discussion of economic theorizing and of economic theory development. 1. Introduction 2. Manipulation, Invariance, Superexogeneity, and Causal Structure 2.1 The Manipulability Theory of Causation 2.2 The Idea ofWeak Exogeneity 2.3 The Idea of Invariance and Its Relation to the Idea of Superexogeneity 2.4 Can We Equate a Causal Relation with an Invariant Relation? 3. The Methodological Import of the Distinction between Causation and Invariance 4. A Causal Structuralist Account of Economic Theorizing and Economic Theory Development 5. Conclusion