Conference Agenda

Climate change poses significant risks to the global labour force, with rising temperatures adversely affecting both worker productivity and health. Exposure to extreme heat increases the incidence of heat stress, morbidity, and mortality, thereby reducing labour supply and raising hospitalization costs. This study investigates the compounded effects of temperature extremes on labour productivity, healthcare costs, and the role of increased energy consumption as a form of autonomous adaptation. We integrate this framework into the WITCH Integrated Assessment Model to assess the long-term economic implications of these impacts,

quantifying new energy trends and adaptation costs as increased AC stocks across future climate scenarios.

We developed an hybrid approach in which impacts on energy (supply and demand), health and labour due to temperature extremes are explicitly disaggregated from the impacts on the other sectors, allowing

us to explicitly model different channels of impacts and relative adaptation feedback. Impact functions for energy demand are derived from an econometric study by Colelli and Sue Wing (2025), which estimates the response of energy demand by fuel type and sector on a global scale, due to projected CDD (24) HDD (18) temperature extremes, following the methodology established in Colelli et al. (2022). These are endogenously determined by the climate emulator in the model. Projected energy increase is also used to determine the level of AC stock as increased adaptation investments. The same temperature extremes are

used to determine the climate change shock on the labour/weighted total factor productivity (tfpy), considering only the effect on low exposure sectors which benefit from AC. In this way, a direct link is established to explicitly account for the protective effect of adaptation on labour productivity. This allows to run different scenarios with or without the energy feedback. In this way, the model endogenously determines fuel specific energy demand and production across different scenarios. Additionally, following Adelaide et al. (2022), the impact of extreme heat on morbidity is assessed through an analysis of increased emergency visits, outpatient visits, and hospitalisations linked to extreme heat events, reflecting anticipated regional

disparities in health outcomes. Projected reductions in heat-related hospital admissions are directly linked with the increased adoption of AC stock endogenously determined. Three scenarios evaluated: (1)Baseline

(no climate, no adaptation): In this scenario the output level is the one in which there is no additional impact of temperature on labour productivity, and accordingly, no additional energy shall be used for adaptation. (2) Climate impact scenario: This scenario accounts only for the impact of heat on productivity, without

new energy consumption, implying lower levels of output Y. Increased hospitalization costs are added as well. (3) Adaptation scenario: Both the additional energy consumption and Ac stocks and reduced labour productivity are included. The endogenous response should therefore imply additional energy consumption

respect to the baseline to reduce the impact on labour. Depending on the adaptation effectiveness, the final consumption should roughly match the one in the baseline, net of the increased costs for the additional AC stocks.

Results are robust in showing how the negative effect on labour significantly affect highly exposed countries, but also many countries in European region and the US, leading to significant reduction in the economic output. The energy use for adaptation is assessed with a specific detail on energy sources, allowing

to evaluate also the impact on mitigation pathways and emissions.

Our results indicate that climate-induced labour impacts can lead to considerable economic losses and rising healthcare expenditures even in a moderate climate change scenario. While increased energy use for adaptation—primarily through cooling—proves cost-effective, it only partially offsets the broader negative

consequences for welfare and productivity.