Inferring the rate of pure time preference under uncertainty

This paper studies how to infer the rate of pure time preference (ρ) from the Ramsey Rule when multiple asset returns exist due to uncertainty. Using a Generalized Uncertainty Ramsey Rule derived from a model that separates intertemporal substitution and risk aversion, we find that the U.S. historical data on consumption growth and asset returns imply that (i) for the reciprocal of the elasticity of intertemporal substitution less than or equal to one, ρ lies within ±1% from zero for a plausible range of the coefficient of relative risk aversion; and (ii) for the larger reciprocal of the elasticity of intertemporal substitution, ρ tends to be negative. These results contradict the widely-held belief in the environmental economics literature that the inferred ρ must be significantly larger than zero and suggest that it is appropriate to use ρ=0 as a benchmark for economic analysis of environmental policies.