• Krieger, Nancy
• Smith, George Davey

Abstract

“Causal inference,” in 21st c CE epidemiology, has notably come to stand for a specific approach, one focused primarily on counterfactual and potential outcome reasoning, and using particular representations, such as directed acyclic graphs (DAGs) and Bayesian causal nets. In this essay, we suggest that, in epidemiology, no one causal approach should drive the questions asked or delimit what counts as useful evidence. Robust causal inference instead comprises a complex narrative, created by scientists appraising, from diverse perspectives, different strands of evidence produced by myriad methods. DAGs can of course be useful, but should not alone wag the causal tale. To make our case, we first address key conceptual issues, after which we offer several concrete examples illustrating how the newly favored methods, despite their strengths, can also: (a) limit who and what may be deemed a “cause,” thereby narrowing the scope of the field, and (b) lead to erroneous causal inference, especially if key biological and social assumptions about parameters are poorly conceived, thereby potentially causing harm. As an alternative, we propose that the field of epidemiology consider judicious use of the broad and flexible framework of “inference to the best explanation,” an approach perhaps best developed by Peter Lipton, a philosopher of science who frequently employed epidemiologically-relevant examples. This stance requires not only that we be open to being pluralists about both causation and evidence but that we also rise to the challenge of forging explanations that, in Lipton’s words, aspire to “scope, precision, mechanism, unification and simplicity.”

Topics

• strength
• forging