When using climate models to calculate optimal greenhouse gas emission scenarios, there are two common approaches: cost-benefit analysis and cost-effectiveness analysis. Cost-benefit analysis minimises the cost of emission reductions and the anticipated costs of damage caused by climate change over the coming centuries; at the same time, the model maximises human welfare. Cost-effectiveness analysis minimises only the cost of emission reductions but adds a constraint: to stay below a given temperature target, such as 1.5 or 2°C. Alternatively, instead of a temperature constraint, the models can use a constraint on the atmospheric CO2 concentration or cumulative emissions. Constraints are often applied only after the year 2100, allowing the model to ‘overshoot’ the target before 2100. For the same temperature outcome in 2100, emissions reductions in cost-effectiveness models tend to happen later than in cost-benefit analysis. This paper evaluates the welfare cost of this delay, using several types of cost-effectiveness constraints and comparing results from cost-benefit and cost-effectiveness models.

In their (welfare-maximising) benchmark cost-benefit model, the authors find that a 1.5°C target in 2100 requires very rapid emission reduction in the coming decades. There are almost no net negative emissions at the end of the century (i.e. relying on negative emissions in the future is never optimal because abatement happens too late to avoid near-term damages). Turning to cost-effectiveness models, the authors find that a constraint on cumulative emissions produces the second-best welfare outcomes, and a temperature constraint that allows temperature to overshoot the 1.5°C target before 2100 is third-best. By contrast, a constraint on CO2 concentration (with overshoot allowed) results in insufficient early abatement, leading to a substantial welfare loss of US$29 trillion, spread out over the century. Repeating these analyses with a 2°C target, all cost-effectiveness models lead to emissions abatement that happens too late to be optimal, but the welfare impact of this inefficiency is milder. Again, a CO2 concentration constraint with target overshoot produces the worst results, compared with the welfare-maximising cost-benefit benchmark.

Key points for decision-makers

  • The authors analyse the impacts of applying different constraints to climate model cost-effectiveness analysis on optimal carbon prices, emissions and human welfare.
  • They ‘fit’ a model with abatement costs, capital repurposing costs (associated with stranded assets such as coal mines) and technological change onto the climate scenarios produced by the Intergovernmental Panel on Climate Change (IPCC) and the Network for Greening the Financial System (NGFS).
  • They investigate the three most popular constraints in cost-effectiveness analysis: a temperature constraint (e.g. less than 2°C of warming), a cumulative emissions constraint and a CO2 concentration constraint.
  • The authors assess the implications of these three forms of constraint (with and without overshooting the different targets) and compare them against a cost-benefit approach. This enables them to rank the welfare performance of different types of constraints.
  • Cost-effectiveness analysis is popular because it avoids the need for a damage function, which is notoriously difficult to estimate, and because international climate agreements such as the Paris Agreement set a maximum temperature. However, the authors show that cost-effectiveness analysis tends to abate too late because it is insensitive to the timing of climate change impacts. They propose technical solutions for modellers who cannot apply cost-benefit analysis due to its computational complexity.
Keep in touch with the Grantham Research Institute at LSE
Sign up to our newsletters and get the latest analysis, research, commentary and details of upcoming events.