Optimal climate policy under exogenous and endogenous technical change: making sense of the different approaches

Integrated assessment models (IAMs) provide key inputs to decision-makers on economically efficient climate policies, and technical change is one of the key assumptions in any IAM that estimates mitigation costs. We conduct a systematic survey of how technical change is currently represented in the main IAMs and find that a diversity of approaches continues to exist. This makes it important to conduct an up-to-date assessment of what difference technical change makes to IAM results. Here we attempt such an assessment, using an analytical IAM with a reduced-form representation of technical change, which we can calibrate on the relationship between abatement costs and the timing of abatement in 109 IAM scenarios from two major databases. We first show in theory how a range of technical-change mechanisms can be adequately captured in a reduced-form model, in which the key difference is whether technical change is a function of time, i.e., exogenous, or cumulative past emissions abatement, i.e., endogenous. We then derive analytical and quantitative results on the effect of technical change on optimal climate policy, for both cost-benefit and cost-effectiveness policy problems. Under cost-benefit analysis, technical change has a quantitatively large, negative effect on long-run emissions and temperatures. The effect on carbon prices differs markedly depending on whether technical change is exogenous or endogenous, and whether clean technology deployment is incentivised by carbon prices or a dedicated deployment subsidy. Under cost-effectiveness analysis, technical change has a small effect on transient emissions and temperatures, but it has a large, negative effect on carbon prices almost irrespective of the policy instruments available. We make several practical recommendations for how IAMs can better incorporate TC, particularly when facing computational constraints.