Their ev doesn’t assume divestment, which outweighs negligible Green Paradox effects

Fossil fuel market dynamics will have a significant impact on the effectiveness of climate policies1 . Both fossil fuel owners and investors in fossil fuel infrastructure are sensitive to climate policies that threaten their natural resource endowments and production capacities2–4, which will consequently affect their near-term behaviour. Although weak in near-term policy commitments5,6, the Paris Agreement on climate7 signalled strong ambitions in climate change stabilization. Many studies emphasize that the 2 °C target can still be achieved even if strong climate policies are delayed until 20308–10. However, sudden implementation will have severe consequences for fossil fuel markets and beyond and these studies ignore the anticipation effects of owners and investors. Here we use two energy–economy models to study the collective influence of the two central but opposing anticipation arguments, the green paradox11 and the divestment effect12, which have, to date, been discussed only separately. For a wide range of future climate policies, we find that anticipation effects, on balance, reduce CO2 emissions during the implementation lag. This is because of strong divestment in coal power plants starting ten years ahead of policy implementation. The green paradox effect is identified, but is small under reasonable assumptions. Since the eighteenth century, fossil fuel use—and resulting CO2 emissions—has increased steadily. Yet the remaining endowment of coal, oil and gas is abundant, which, if combusted using conventional technologies, will lead to dangerous climate change in the future. The Paris Agreement aims to limit the global mean temperature increase ‘well below’ 2 °C. This implies a very tight global carbon budget for the twenty-first century and, consequently, that policies to reduce CO2 emissions would leave a considerable share of fossil fuels underground1,3,4 . Previous studies have shown that cost-effective emission reduction policies tend to begin as soon as possible and progressively tighten over time, thus allocating the CO2 budget most efficiently over the remainder of the century. The Paris Agreement, however, is weak in near-term policy ambition. The sum of the nationally determined contributions (NDCs), made as part of the Paris Agreement, is around 12–14 GtCO2-eq higher in 2030 than the global costeffective greenhouse gas emission pathways commensurate with the 2 °C target9 . Previous studies assessed that it is possible to delay the implementation of strong climate policies until 2030 and still achieve the 2 °C target. However, this significantly increases the challenge of complying with the 2 °C target because of: (1) higher necessary CO2 prices13; (2) a strong drop in fossil fuel prices because of the rapid reduction in demand1 ; (3) stranded assets of fossil fuelbased infrastructure14; (4) and a strong acceleration in the required ramp-up of low-carbon technologies10. These studies made the simplifying assumption that emitting sectors are forced to follow prescribed emission trajectories until 2030. Some of these baselines include weak climate policies such as NDCs6,10, while others do not8 , but a key limitation is that relevant actors are unable to react to policies that are brought into force post-2030 before they are actually implemented. If actors were given knowledge of future climate policy, it has been argued that they would choose to deviate from these prescribed trajectories, which could therefore lead to large deviations in CO2 emissions. Two arguments, the green paradox and the divestment effect, are central in the debate. Both are responses to a climate policy that has a lag between announcement and implementation. The green paradox15 argument hypothesizes that near-term CO2 emissions would rise above the baseline. Fossil fuel owners frontload supply from their endowments to evade the negative consequences of future drops in fossil fuel prices due to future climate policies. The argument relies on resource owners seeking to maximize rents received from their resources: they seek to avoid, or reduce, the devaluation of fossil fuels that are rendered ‘unburnable’ because of future climate policies. In the near-term, this increases fossil fuel supply at lower prices and thus boosts CO2 emissions11,15–17. Conversely, the divestment effect argues that near-term CO2 emissions would decrease below the baseline. Investors avoid building fossil fuel-based infrastructure whose operation would become un-economic by the future climate policy. They divest away from infrastructure with high emission intensities, high capital costs and long technical lifetimes that could become stranded and search for alternative investment opportunities. The divestment effect intensifies as climate policy ambition increases and the policy implementation date comes closer. The green paradox effect materializes directly after the policy announcement, since fossil fuel owners frontload extraction as long as climate policies are weak or even absent during an implementation lag. The growing body of literature on the green paradox relies on highly idealized theoretical models that agree that near-term emissions tend to exceed baseline levels because of future climate policies11,16,17. Yet these studies ignore the divestment argument since infrastructure inertia is difficult to treat analytically. Conversely, while a few numerical studies have investigated the divestment argument12, the methods employed were unable to take into account the maximization of intertemporal resource rents that are essential for the green paradox effect18,19.