Causal Relations via Econometrics

Abstract:     
Applied econometric work takes a superficial approach to causality. Understanding economic affairs, making good policy decisions, and progress in the economic discipline depend on our ability to infer causal relations from data. We review the dominant approaches to causality in econometrics, and suggest why they fail to give good results. We feel the problem cannot be solved by traditional tools, and requires some out-of-the-box thinking. Potentially promising approaches to solutions are discussed.

Number of Pages in PDF File: 31

Keywords: causality, regression, exogeneity, Hendry methodology, natural experiments

JEL Classification: C5, C59

Accepted Paper Series