This paper consults on our proposed methods and Core parameters for the next version of the CMI Mortality Projections Model, CMI_2026, which we aim to release in March 2027.
We propose to use the same method and parameters for CMI_2026 as we used for CMI_2025, other than:
- a routine update to the data used to calibrate the Model – calibrating to data for 1986-2026 and reflecting the latest ONS population estimates; and
- possible changes to methods for high-age population estimates – if our ongoing review can reach a conclusion in good time for CMI_2026.
The paper includes reviews of our approach to three existing methods and parameters:
- the half-life assumption for the fitted overlay;
- projection assumptions; and
- high-age population estimates.
We also include research into two new areas:
- the impact of using non-calendar year periods rather than calendar years; and
- how “mortality selection” may have affected recent mortality rates and improvements.
The paper seeks views of users of the Model on our proposal for CMI_2026. We would like to receive comments by 5 November 2026.
We have also released a new “parameters paper”, which summarises the rationale for the parameters for the current version of the CMI Model, CMI_2025. This is available alongside CMI Working Paper 211 together with other outputs relating to CMI_2025. We intend to release an updated version annually, alongside each new version of the CMI Model.
Documents:
This paper and the spreadsheets are available to Authorised Users only.
Events calendar
No results found.