The paper by Barnett, Barrett, and Seth explores the relationship between Granger causality and transfer entropy, two statistical measures used to infer causal relationships in data. Granger causality, originally developed in econometrics, assesses causal influence through vector autoregression, while transfer entropy is an information-theoretic measure of directed information transfer between processes. The authors show that for Gaussian variables, these two measures are equivalent up to a factor of 2. This equivalence bridges the gap between autoregressive and information-theoretic approaches, providing a unified framework for causal inference. The result has implications for both theoretical and practical applications, such as spectral implementations of transfer entropy and the identification of nonlinear autoregressive models. However, the authors also note that the Gaussian assumptions underlying these measures may need to be validated in empirical studies, especially for highly multivariate datasets.The paper by Barnett, Barrett, and Seth explores the relationship between Granger causality and transfer entropy, two statistical measures used to infer causal relationships in data. Granger causality, originally developed in econometrics, assesses causal influence through vector autoregression, while transfer entropy is an information-theoretic measure of directed information transfer between processes. The authors show that for Gaussian variables, these two measures are equivalent up to a factor of 2. This equivalence bridges the gap between autoregressive and information-theoretic approaches, providing a unified framework for causal inference. The result has implications for both theoretical and practical applications, such as spectral implementations of transfer entropy and the identification of nonlinear autoregressive models. However, the authors also note that the Gaussian assumptions underlying these measures may need to be validated in empirical studies, especially for highly multivariate datasets.