his is important because those models underpin decisions about public health, pensions and insurance. How can we improve them so that we are better prepared in the future? 

 That question was the focus of new research by Frank van Berkum, Michel Vellekoop (both from the Amsterdam School of Economics, UvA) and Bertrand Melenberg (Tilburg University). In collaboration with CBS Statistics Netherlands, they have developed a new framework to better understand and predict excess mortality during pandemics. The study was commissioned by the Dutch Ministry of Health, Welfare and Sport and funding organisation ZonMw.

Better predictions and better scenarios 

By combining historical mortality data with detailed weekly data from the COVID-19 period, the researchers were able to see exactly how the pandemic affected different age groups. This approach allows policymakers and health authorities to make much more accurate estimates of excess mortality and age distribution. It also allows them to better evaluate whether measures and interventions are effective enough.  

The research also contributes to short-term and long-term predictions made by actuaries and demographers. Based on the outcomes, they are able to make more accurate statements about the impact of a pandemic on life expectancy. This also results in more accurate scenarios about the financial impact of pandemics. Armed with those insights, policymakers can take increasingly well-informed decisions. As Van Berkum concludes: 

'By using detailed data and robust statistical models, we can paint a clearer picture of the true impact of pandemics on mortality, supporting better decisions for society.'

Combining large data sets 

The researchers used existing predictive models as a basis, but adapted them to the unique circumstances of a pandemic. First, they identified general, normal trends by analysing decades of mortality data. They then combined this data with the much more detailed, weekly records kept during the pandemic. As a result, they were able to monitor changes closely and clearly visualise differences between age groups. By combining long-term trends and weekly, updated changes, the researchers succeeded in developing a demographic model that accurately shows the effects of the pandemic. Moreover, this model can be used to explore scenarios for the future.  

Development of adaptive models

The researchers recommend that future research should include not only the results of the pandemic, but also data relating to the subsequent period. In other words, the period in which we are currently living. One as-yet-unanswered question, for example, is whether mortality rates are returning to the trends visible in the pre-pandemic period, or whether the pandemic has permanently affected the mortality rates. In that context, the researchers also state their intention of developing adaptive models that can respond flexibly to new data from the Netherlands and other countries. Furthermore, future research should ideally pay more attention to individual causes of death in order to obtain an even better and more complete picture of demographic trends. By connecting knowledge and data, these researchers are not only ensuring that we can understand the past, they are also giving us tools to prepare increasingly effectively for future pandemics.