Working Paper

Irreducible Uncertainty and its Implications: A Narrative Action Theory for Economics.

Download paper


At the heart of economics is a theory of action. It reflects views about how human beings make economic decisions and leads to an analysis of aggregate consequences.

As we know, despite the insights of pioneering figures like Adam Smith, Alfred Marshall and Maynard Keynes, the story of economic analysis, until very recently, is in essence that economic action rests on the conclusions of deductive logic machines not human judgment. However the history of AI (despite successes) shows that logic machines don’t do well with irreducible uncertainty. The present state of human knowledge and its achievements, on the other hand, shows that humans are biological, sentient, reflexive, social and imaginative beings who as a species are highly specialised and successful at adapting to uncertainty.

Today I want to argue that the difference I have drawn between how machines and humans handle uncertainty matters a lot. In fact, in reality all the big decisions in economic life – investment in innovation, capital goods and future projects, all finance and all large scale economic and organisational management, for instance – require action under that condition and it is highly consequential.