Strategic Decisions under Uncertainty: A Role for Requisite Risk Models?

This talk reflects on several projects - past, present, and planned - where there is a commonality in the methodological modelling challenge although diversity across the real-world problems addressed. The problems include making strategic choices about novel technological concept design to be robust to future uncertainties; making strategic decisions to manage risk and enhance resilience in Arctic search and rescue; and informing effective strategies to be able to respond to future malicious attacks in urban areas. All problems can be characterized by different types of uncertainty, decisions to be taken in a context of multiple stakeholders and choices to be made that have long-term implications for a socio-technical system. Framing the problems as one of making choices under uncertainty to manage risk, we examine an analytical approach that that allows us to create requisite models. That is, models whose "form and content that are sufficient to solve a particular problem" (Phillips, 1984). In turn, we argue such models can enable us to create solutions that, according to the economist John Maynard Keynes are "roughly right rather than precisely wrong".

A particular challenge is how we appropriately mix relevant methods to create a defensible methodology. We have faced/are facing this challenge in the context of the motivating problems where we mix methods ranging from scenario planning (Cairns and Wright, 2017) to stochastic modelling (Aven and Jensen, 2013). On the face of it these methods are more usually associated with different types of uncertainties, supporting analysis of risk to inform decisions at different organizational levels and planning horizons. However, by structuring models at an appropriate level of the unit of analysis for the problem it can be feasible to provide a meaningful bridge between the deep uncertainties surfaced through foresighting with those uncertainties grounded in experience. We share examples of how we have achieved this and discuss the practice benefits achieved as well as the scientific challenges remaining.

Keywords: Scenario analysis, Decision making under uncertainty.

Download PDF