What are We Missing in the Relationship between Climate and Economies?

Modellers widely use damage (or impact) functions to define the climate's feedback on the economy. Many of these functions are based on econometric theories and are empirically estimated using historical data. However, it is unclear which uncertainties to check for and how these interact in determining the impacts. These robustness checks are pivotal since impact functions are also used to provide policy-relevant insights. We reestimate and analyze four relevant damage functions: Burke, Hsiang, and Miguel (2015), Kalkhul and Wenz (2020), Kotz et al. (2024), and Waidelich et al. (2024). These four functions represent not only different structures, but also different philosophies. We consider three sources of uncertainty: estimation, future scenarios, and internal climate variability. We use a newly developed dataset on climate and climate extremes to reestimate them, performing cross-validation to understand the quality of the fit. We then quantify their uncertainty and perform multivariate sensitivity analysis at the global and regional levels to identify country-specific relevant factors. We gain relevant insights for modellers and policymakers, identifying gaps in our current understanding of climate impacts on our societies.