The confounding problem of the counterfactual in economic explanation*

Abstract

Economists who have emphasized uncertainty have tended to draw a sharp epistemic distinction between an ascertainable past and an unknowable future. But in one critical respect – in extracting causal relationships – the epistemic distinction is unsustainable. Many types of causal arguments in economics depend on counterfactual reasoning. Counterfactualizing entails the construction of fictions about what would have happened or would happen in the world absent an event that is taken to be causal. But that alternative world is foreclosed the moment the causal event occurs. Complicating matters, there is no dependable method for ascertaining the uniquely true counterfactual. Distinct research methods, and distinct economic paradigms, generate alternative plausible counterfactual accounts. This implies that causal claims in economics, too, are irreducibly fictitious – regardless of whether the subject matter concerns the past, or the future.

KEYWORDS:

• Causation
• irreparable ignorance
• counterfactual reasoning

JEL Codes:

• A11: role of economists
• B41: economic methodology

Acknowledgements

Katie Aldrich, Abby Brown, Lauren Craig, Joe Downes, Sophia Gonzalez-Mayagoitia, Quentin Good, Amanda Hayden, Tasia Poinsatte, and especially Holden Fitzgerald, provided wonderful research assistance on a project from which this talk is drawn.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Not all future events are equally unknowable, of course, as the Knight quotation indicates. Here I pass over the difficult question of specifying just which kinds of events fall into the category of probabilistic risk and which fall instead into the category of true uncertainty.

2 Though I will not pursue the matter here, more complex accounts of causation (as represented by complexity models, for instance), amplify the epistemic difficulties associated with specifying the correct counterfactual that I explore below.

3 The range of plausible counterfactuals expands exponentially if we ground economic causation in complexity or ‘quantum’ models that feature emergent properties and irreducible uncertainty. See Orrell (2018).

Additional information

Notes on contributors

George F. DeMartino

George F. DeMartino is Professor of economics in the Josef Korbel School of International Studies, University of Denver where he co-directs the MA in Global Finance, Trade and Economic Integration. His most recent books include The Economist's Oath (Oxford University Press), and the Oxford Handbook of Professional Economic Ethics, which he co-edited with Deirdre McCloskey. He is now completing The Tragedy of Economics: The Harm Economists Do as They Aspire to Do Good (University of Chicago Press).