Cumulative carbon emissions and economic policy: in search of general principles


In this research Dietz and Venmans have built a model of economically-efficient carbon dioxide emissions using recent advances in climate science that show that the warming response of the atmosphere to cumulative CO2 emissions is effectively linear and nearly instantaneous. They combine these features with representations of climate damages and the costs of CO2 emissions abatement, which capture the stylised facts of research on each topic.

The model is surprisingly simple and provides solutions for economically optimal peak warming of the planet, optimal emissions along the transition to peak warming and optimal carbon prices, including under a temperature constraint consistent with the Paris Agreement.

Key points for decision-makers

  • Optimal peak warming depends sensitively on several parameters that are highly uncertain. This implies that optimal peak warming is itself highly uncertain.
  • Even if optimal peak warming is high, optimal transient warming over the coming centuries is not. The transition is slow, because of the stock-flow nature of CO2-induced warming. If optimal peak warming is 3.4°C, optimal transient warming one century from now is only 1.7°C.
  • Contrary to some previous work, the optimal carbon price in the model initially grows faster than output per capita, but in the long run it grows at the same rate. This is because damages grow more than proportionally as a function of cumulative emissions, and one of the main reasons for this is the saturation of carbon sinks. For central parameter values, the authors calculate that the optimal carbon price grows 0.5 percentage points faster than the economy initially.
  • The optimal carbon price under a binding temperature constraint like that of the Paris Agreement comprises the social cost of carbon, plus a so-called ‘Hotelling premium’ that reflects the need to allocate the additional costs of the Paris constraint efficiently over time. If we take account the cost of damages from CO2 emissions, then we should abate emissions more quickly than if we simply meet the temperature constraint at the lowest discounted abatement cost. This effect is quantitatively large.
  • When the objective is to minimise abatement costs alone, the optimal carbon price follows the simple Hotelling rule, not various kinds of augmented Hotelling rule, as in previous work. This is because modest thermal inertia and the saturation of carbon sinks more-or-less exactly offset the effect of decay of atmospheric CO2.