The insurance costs of climate change

The study
SEO Amsterdam Economics quantified the costs of climate change for insurers: the insured losses attributable to climate change worldwide. By combining industry estimates of weather-related insured losses with state-of-the-art climate research, we calculate how much of the insurance bill can be directly linked to anthropogenic climate change (i.e., climate change caused by humans). These estimates differ across regions and by type of natural hazard (such as forest fires, floods, and hurricanes). The results show that climate change has led to large and rising costs for insurers. The report also warns about the growing ‘uninsurability’ of houses in high-risk areas, which can shift costs to households and governments. 

Key findings

• Insured losses attributable to climate change have nearly doubled since 2012. They have grown on average at 6.5% per year, outpacing total insured-loss growth (4.9% per year).  Between 2002 and 2022, climate change cost insurers an estimated USD 475–720 billion — on average about USD 30 billion per year. This is equivalent to 34% of all natural hazard insured losses, or 37% of weather-related losses. 
• Estimates of the ‘Fraction of Attributable Risk’ (FAR) can be used to attribute insurance losses to climate change. This study shown that a FARbased framework, paired with clear reliability flags, offers a transparent, reproducible approach for insurers, supervisors and policymakers to track and communicate the direct insurance costs of climate change. This framework also helps to assess the implications for pricing and (lack of) insurance coverage.  
• The risk of an ‘uninsurability crisis’ warrants the attention of policy makers. Given the rising insurance costs caused by climate change, there is a growing risk that homeowners will no longer be able to insure themselves against extreme weather-related events. To reduce this risk, there is an urgent need for regulation and policy frameworks. 

Methods
For this study, SEO collected and analysed data on reported insured losses from five leading insurance companies (Aon, Gallagher Re, Munich Re, Swiss Re, and Verisk) and combined these with peer-reviewed scientific estimates of the Fraction of Attributable Risk (FAR), reported in Newman & Noy (2023). Triangulating these different sources, we then derived the climate-attributable share of insured losses by event type and region. To support careful use, we also introduced a ‘reliability score’ based on statistical significance and confidence intervals, and we advised using global averages when regional estimates are weak or missing. As a cross-check and to gain deeper insights into the global climate events landscape, we used the EM-DAT comprehensive catastrophe database. 

The interview with Nienke Oomes on BNR Nieuwsradio can be found here.