Epistemic Causality and Hard Uncertainty: A Keynesian Approach

The interplay of epistemic and empiric conditions of human behaviour plays a crucial role in economic causality but it is not satisfactorily analysed by the existing approaches to economic causality, including the most influential of them: Granger causality. In order to find a more satisfactory approach to economic causality, this paper draws inspiration from the contributions of Keynes. In particular it is argued that the systematic use of probabilistic causality in the General Theory is deeply rooted into the theory of probabilistic causality outlined in the Treatise on Probability which is based on the crucial distinction between epistemic and empiric causality. In this view, since epistemic causality is conceived as probabilistic, also empiric causality has to be conceived as probabilistic in economics since in this case, differently than in natural sciences, the observed events are mediated in a crucial way by epistemic links, such as beliefs and expectations. In particular, the awareness of the relevant ignorance, represented by what is called by Keynes the "weight of argument", affects in a crucial way the behaviour of economic agents. The Keynesian approach is extended by observing that a change in the weight of causal arguments represents a sort of "second-order epistemic causality" that may become a possible independent source of empiric causality.