Less precision, more truth: uncertainty in climate economics and macroprudential policy

Abstract:

The climate and financial system share various features—they are both “complex systems”—but they also have important differences. This chapter explores what lessons might be learned from climate system modelling for financial system modelling and macroprudential policy. First, we argue that the primary lesson from climate science is that systematic data collection at a number of scales is a vital ingredient for properly understanding and modelling systemic risk. Yet, there is no equivalent for the financial system to the comprehensive data collection used by humanity to monitor Earth’s climate. Second, we consider models for the economic consequences of climate change, which we refer to as climate economics. We argue that, unfortunately, the main lessons from climate economics concern how not to model complex systems. As with estimates of physical parameters throughout history, the treatment of uncertainty in climate economic models remains inadequate, and outputs are expressed with far too much confidence. Climate economic models often omit significant variables, inadequately account for feedbacks, non-linearities, heterogeneity and non-rational behaviour. Suggested cures to these problems include: stating conclusions that are less precise but more truthful; employing scenarios to explore the ranges of possible system behaviours; placing greater focus on resilience and not just efficiency; and making policy with an eye on precautionary principles.