Prescriptivism, risk aversion, and intertemporal substitution in climate economics

Working paper; comments welcome!


The question of how to discount the distant future has long been at the core of climate economics. It has also divided economists. Some argue for prescriptivist approaches to discounting, often calling for social discount rates of as low as 1% per year. Others argue strongly for descriptivist approaches and rates as high as 5% or more. A look to financial economics has since added another wrinkle, by pointing to the need to separate risk aversion from intertemporal substitution to calibrate real-world behavior, at times lowering effective, ‘descriptivist’ rates close to prescriptivist ones.

We attempt to reconcile some of these methodological differences by identifying three types of prescriptivism. Economists are frequently uncomfortable with what we term parameter prescriptivism, while being comfortable with both axiom and policy prescriptivism. That faces theoretical challenges. We argue that if a priori moral reasoning is not allowed to influence parameter values, then the results of one’s analysis should not be framed as a prescriptive policy ‘recommendation’. While descriptivist analysis is relevant to policy, we must be clear that it can only inform policy choices, not determine them. As a case study, we use our framework to evaluate arguments for replacing the standard isoelastic utility function with Epstein-Zin preferences in economic analyses of climate change.

JEL codes: D6, H43, Q5

Full text: “Prescriptivism, risk aversion, and intertemporal substitution in climate economics” (version: 11 February 2018)

Kelleher, J. Paul and Gernot Wagner. “Prescriptivism, risk aversion, and intertemporal substitution in climate economics.” SSRN working paper (11 February 2018).

Up Next

Ramsey discounting calls for subtracting climate damages from economic growth rates

More Academics

Keep in touch.