ONS Makes Excess Deaths Disappear

By John Blundell

Below is a response from ONS regulators based on my article ONS just made excess deaths disappear.

I have received this email in response to my observation that ONS have made ONS excess deaths disappear using an experimental methodology that is inappropriate, non-credible or believable.

There is no answer other than to use a 5 year average, but that includes abnormal years 2021-2024 as is used by EuroMomo, World Data, and UN.

In which case the abnormal years 2021-2023 excess all causes of death are absorbed into a 5 year average producing misleading invalid statistics.

As I have shown in my Substack appraisal, there are perfectly valid and more believable methodologies that use stable normal sampling regression trend methods or to simply to include population increase % to a known normal 5 year average 2015-2019.

Their “Experimental” methodology now elevated to “Developmental”. However all excess deaths now officially made to disappear as official statistics in development!

Reply Email Sent on behalf of Ed Humpherson, Director General for Regulation 16/04/2024 16:40

Dear John,

Thank you for your recent correspondence addressed to Sir Robert Chote as Chair of the UK Statistics Authority. Your email was passed to the Office for Statistics Regulation (OSR) so that we could investigate your concerns regarding the recently adopted methodology for calculating excess deaths in the UK.

In your emails, you raised a concern that the Office for National Statistics’ (ONS) new excess deaths methodology artificially inflates the level of expected deaths, and in doing so eliminates all excess deaths. You requested that the new methodology be withdrawn from use and its data be declared to be invalid. 

It is worth noting that there is no definitive measure of excess deaths. As you will be aware from your own work on this topic, excess deaths are the difference between the actual number of deaths and the expected number of deaths. Unlike actual deaths, expected deaths cannot be directly measured. As you observed in your correspondence, they are estimated using different approaches, which yield different results. Each approach has its own strengths and limitations; none is perfect. 

The new methodology was introduced after careful consideration of the limitations of the 5-year average (5YA) method, compared to the new methodology, which ONS outlined in its methodology paper in February 2024. A summary of the reasoning behind the change in approach is given below:

The previous 5YA methodology did not take account of population growth and ageing. As the UK’s population is growing and ageing, ONS and its UK partners expect the number of deaths to increase. Furthermore, most deaths occur among the oldest age groups, so if there are a greater number of people in the oldest age groups this will also lead to an increase in deaths. By not taking these factors in account, the former 5YA method tended to underestimate the number of expected deaths and therefore overestimate excess deaths.

The previous 5YA methodology did not take account of trends in mortality. The historical data on excess deaths, using the previous 5YA methodology, shows negative excess deaths for most of the long-term period prior to the pandemic. This is because mortality rates were reducing over time and calculating an expected value based on the previous five-year average led to an artificially high expected value and subsequently underestimated excess deaths. Until around 2015 the new methodology results in higher estimates of excess deaths. 

The previous methodology excluded the full year of 2020 when calculating expected deaths due to the very high levels of mortality at points during that year. The new methodology only excludes specific weeks and months (in spring 2020 and in winter 2020/2021) rather than the full calendar year. 

When the new excess deaths methodology was introduced in February 2024, OSR recognise that ONS had communicated the changes to users effectively and that there was good coordination among the relevant producers across the UK.

In your correspondence, you requested that the data produced by the new methodology be declared to be invalid as National Statistics, and you made several suggestions for changes to how estimated deaths in the UK are calculated. OSR has been assured that ONS and its partners will regularly review estimates produced by the new excess deaths methodology, with a commitment to make further refinements to the approach being undertaken if necessary. Potential future work to further refine the methodology has been outlined in the ONS methodology paper. You will therefore be interested to know that, as a result, estimates of excess deaths produced by the new methodology will be labelled as official statistics in development while further review, testing and development work is undertaken. User feedback will form part of this review, and you can contact ONS directly if you have further suggestions for this work. OSR does not have any plans to review the methodology further while this developmental work is ongoing.

  I want to assure you that OSR recognises the importance of these datasets and will continue to monitor them for trustworthiness, quality, and transparency.

Kind regards,

Sent on behalf of Ed Humpherson, Director General for Regulation

Chris Ferrier

Subscribe to the Freedom Magazine

Insightful articles, entertaining columns and much more! Stuff you don’t get in the mainstream media – all printed on paper.

We don’t use social media. Please share this article by email: