Rupert Way 

Date/Time: 22.04.2026, 12:00 

Location: D2.2.094 

Abstract 

Rapidly decarbonising the global energy system is required to address climate change, but concerns about low carbon technology costs have been a barrier to implementation. Global economy-climate models have for decades been used to attempt to understand the characteristics of different decarbonisation scenarios and guide policy decisions. However, within these models future technology costs are often generated using poorly justified forecasting methods, and in hindsight they have all systematically overestimated the future costs of all the most important decarbonisation technologies, thereby overestimating total decarbonisation costs and reducing incentives for policy makers to act. We discuss technology forecasting methods and how they are implemented, and we use a backtested probabilistic learning-curve cost forecast model to show that rapid clean electrification is likely the cheapest scenario, generating trillions of dollars of net savings compared to continuing with a fossil-fuel based system, even without accounting for climate damages. We discuss energy modellers’ systematic pessimism, the ongoing need to ensure models sensibly reflect observed data and trends, and the subsequent effects on policy making and climate narratives more broadly.

Bio 

Rupert is an Honorary Research Associate at the Smith School of Enterprise and the Environment, and an Associate at the Oxford Institute for New Economic Thinking. He is also an Oxford Martin Fellow. He has a PhD in mathematics, and his current research focuses on energy system modelling, technology forecasting, innovation and technological change, portfolio analysis, decision making under uncertainty, and understanding strategies for accelerating the sustainable energy transition. Broader interests also include  physics, complex systems, political and financial networks, and sustainability.