Quantifying climate mortality risk: How rising temperatures are reshaping life insurance modeling

The insurance industry has spent years building sophisticated models for climate-related property damage. Mature and reliable systems have been developed to assess the impact of natural disasters on homes, businesses and infrastructure. In the world of personal health and life insurance, however, the equivalent work has barely begun. 

David Gómez-Ullate, Professor of Applied Mathematics at IE University and Director of IE Research Datalab, is undertaking groundbreaking research at the intersection of climate change and life insurance modeling. Leading an interdisciplinary team in close collaboration with major European insurer Vienna Insurance Group (VIG), he has developed a novel framework to quantify the impact of climate change, particularly rising temperatures, on human mortality.

The gap in life insurance planning 

Climate risk has been integrated into property and casualty insurance for nearly two decades. These natural catastrophe (NatCat) models convert meteorological inputs, such as storms, floods and wildfires, into loss estimates that feed directly into pricing, capital models and reinsurance structures. Supervisory frameworks such as the Network for Greening the Financial System (NGFS) have reinforced this, pushing insurers to adopt quantifiable climate stress testing as a standard part of their risk governance.

Life insurance doesn’t operate the same way. A term life policy or annuity contract could carry liabilities stretching decades into the future. The assumptions they use about mortality rates, life expectancy and how both will evolve depend on historical mortality data and demographic projections, such as improvements driven by advances in medicine. Climate change doesn’t yet feature in these assumptions in a systematic or quantitative manner.

"Insurers are aware that climate change is affecting human health and longevity," says Gómez-Ullate. "After all, there is a growing body of scientific evidence linking temperature and environmental conditions to mortality and morbidity. But this evidence hasn’t been translated into the life tables, pricing assumptions or technical provisions that actuaries use."

This is starting to attract regulatory attention. Across Europe, regulators are moving from high-level climate scenario narratives toward expectations of quantitative integration into core risk processes, including for long-term liabilities. There’s a steady shift, in other words, away from qualitative climate awareness towards a more robust model comparable to NatCat.

The challenges holding the industry back

Epidemiologists have been studying the relationship between temperature and mortality for decades. They have built detailed relative risk curves that show how death rates shift across age groups as temperatures rise and fall across geographies. Climate modelers have sophisticated scenario frameworks that project how temperature distributions will evolve over the course of the next century. And actuaries have well-developed tools for projecting and pricing mortality risk. But until now, these insights haven’t been combined. 

"Developing a quantitative approach to climate change and life insurance involves incorporating data from many different sources," Gómez-Ullate explains. "You need an interdisciplinary approach that draws on the expertise of epidemiologists, climate modelers, insurance actuaries, public health specialists, computer scientists and mathematicians. These specialists tend to work in silos, and it isn’t easy to pull their collective knowledge together."

But without a quantitative bridge between climate science and actuarial practice, insurance companies can’t stress-test climate exposure, assess its financial impact or build climate-adjusted life tables.

A life table is a statistical tool that maps the probability of death at each age across a population. When an insurance company prices a personal insurance product, the relevant life table offers an indication of the statistical life expectancy for someone of that age and location, which the insurer then uses as the basis for the premium they charge. These life tables are an important aspect of the industry’s work, and given that life insurance liabilities span the same time horizons as climate change, they need to incorporate its effects. 

These effects, of course, are vast and varied; narrowing them down to focus on temperature was a practical methodological decision.

"The scope of climate change’s impact on mortality is very broad and complex, and it can be challenging to quantify," Gómez-Ullate explains. "The impacts of climate change extend to land and energy use, inter-country cooperation and rivalry, and major demographic shifts, all of which affect mortality. Together, they make the overall picture difficult to isolate and measure. But if you look at temperature alone as a driving factor, then you can develop a specific and quantitative assessment. As far as we’re aware, no one else has undertaken this in a quantitative way in Europe before."

Turning climate data into mortality risk

The collaboration between VIG and IE University has produced one of the first scenario-based frameworks to incorporate temperature-dependent mortality into life tables. The multidisciplinary team of experts built a framework with direct financial applications for life insurance products, one that insurers can use to update the mortality assumptions that underpin their pricing, reserving and solvency calculations. 

The framework operates as a three-stage pipeline. Climate projections drawn from different models and emissions scenarios are combined with epidemiological evidence on how temperature affects mortality across different ages and locations. These mortality adjustments are then applied as a multiplicative factor to existing life tables, producing climate-adjusted versions that are ready for actuaries to use. 

"Our work draws on a lot of research in public health modeling and epidemiology, including valuable work undertaken by the London School of Hygiene and Tropical Medicine and ISGlobal in Barcelona," Gómez-Ullate adds. "These groups have statistical models that correlate the number of deaths recorded every day in a given city with the temperature a few days before (there’s always a lag). This enables them to see whether more people die when temperatures rise or drop above or below a certain threshold. There’s a sweet spot at a certain temperature where people die the least. They then use this data to create relative risk curves, which we translate into the insurance industry’s life tables."

The framework is also designed to fit in with existing industry practices. Rather than requiring companies to rebuild their actuarial models, it delivers climate adjustment factors that are specific to city, age group, time horizon and emissions scenario and that can be applied directly to current life tables. "Our model doesn’t need to know the inner workings of a particular company, or how they’re pricing their products,” explains Gómez-Ullate. “It simply introduces a modification that factors in temperature. This makes it easy to apply and universally applicable."

The framework’s early results are instructive. Climate change affects heat and cold mortality in opposing directions. Rising temperatures increase heat-related deaths, particularly among older age groups, while reducing cold-related ones. These effects partially offset each other, so the near-term financial impact is moderate. However, the effect increases over time, and it varies considerably by region. 

Under a high-emissions scenario, for example, the framework projects that by 2075, the mortality for a 60-year-old in Bucharest could be around 2% higher than its historical baseline due to the effects of temperature alone. When translated into financial terms, a rise in mortality increases the expected present value of life insurance benefits but reduces the value of annuity payments. In affected markets, this means that the expected present value of a life insurance contract could rise by roughly 1% or more, while annuity values decline slightly. 

Scaling the model’s impact

With the first phase of research complete, the next step is to develop a climate-mortality dashboard. This interactive platform will translate the underlying framework into a practical tool insurers can use to support their decision-making. It will allow users to explore climate-adjusted mortality projections across countries, cities, emissions scenarios and product portfolios, and to run consistent scenario analyses for regulatory purposes. It will also be possible to incorporate new geographies, climate channels and risk metrics as the science and regulatory landscape evolves.

The research agenda is also expanding in two directions. First, it will include additional climate factors such as exposure to fine particulate matter, which is linked to cardiovascular and respiratory mortality. Second, it will extend the framework to include both mortality and morbidity, helping assess how a warming planet affects health, sick leave and hospitalization rates.

"Gathering data from death is simple because most deaths are registered, but measuring the health effects of climate change is much more complex because you can be ill in so many ways," Gómez-Ullate says. "We will need to be able to quantify all these channels as we strive to develop a solution that assists insurers with both their life and health insurance plans."

Driving change in the insurance industry

NatCat models, now standard in property insurance, were once a major innovation that connected meteorological data and financial risk, representing a genuine milestone for the industry. The VIG–IE research aims to make a similar leap in personal life insurance. Gómez-Ullate and his team argue that climate risk is quantifiable, can be integrated into life tables, and can produce systematic and forward-looking mortality adjustments for life insurance.

"All the work we've undertaken so far has made use of public data," concludes Gómez-Ullate. "We intend to continue in this vein, and our idea is to share everything publicly, even the code. It will only exclude information that is proprietary to VIG. We want people to build on the work that we've done so that this field of research can evolve.” As the effects of climate change become more acute and regulatory expectations tighten, this open approach could prove invaluable for the insurance sector and beyond.

This article is part of a series on incorporating climate change into mortality assessments. If you’re interested in learning more about this topic, you can find other related articles here.