Breaking the Paradigm: From Correlation to Causation

Abstract

In this editorial, I explore the shift from correlation-based research to causal inference in the field of Accounting, examining the limitations of traditional econometric methods in establishing causality. It critiques the over-reliance on statistical associations, often neglecting endogeneity concerns, and advocates for the use of quasi-experimental techniques such as Difference-in-Differences, Instrumental Variables, and Regression Discontinuity. Drawing from corporate governance research, I show how these methods can be employed to identify causal effects, including the role of gender diversity in board performance. I stress the necessity of robust econometric foundations and proper contextual understanding to ensure that empirical approaches align with theoretical frameworks, warning against separating causal evidence from its underlying mechanisms. The editorial provides key references and offers suggestions for further reading.