The Consequences of Uncertainty and of Irreversibility on Optimal Intertemporal Co2 Emission Policies

Abstract

This paper investigates how uncertainty affects the optimal intertemporal accumulation of a stock externality using CO2 emissions and the greenhouse effect as a topical pars pro toto. More precisely, the evolution of the future temperature is assumed to follow an Ito-process with the drift provided by current carbon-dioxide emissions. In this context, the paper emphasizes two different kinds of irreversibilities. First, CO2 once dissolved in the air cannot be collected later which rules out quadratic value functions even in linear quadratic applications. Second, stopping CO2 emissions (i.e. a shut down of the entire fossil fuel industry) is irreversible; this introduces the technical complication that 'smooth pasting' is not applicable. Hence, this paper applies techniques that contrast both the deterministic set ups and the stochastic applications so far in the literature. The supposition that stopping emissions is reversible (i.e. a costless re-start of emissions is possible) leads to termination at a temperature level where the marginal benefit of emission still exceeds the net present value of marginal external costs. Although irreversibility is expected to mitigate the usual consequences of uncertainty (e.g. the conservative stopping rule above), it actually further lowers the threshold temperature at which the fossil fuel industry should be shut down forever.

JEL Classification: Q48, D81, C61