Excess Mortality 2013 to 2023

England & Wales

This is not the mortality pattern you would expect from a deadly, respiratory pathogen. And certainly not what you would expect to see as a result of a Safe and Effective™ cure…

Personally, I think this is more symptomatic of what happens when you disrupt the system that has been put in place to keep people alive who have chronic illness, or intervene in some other way that is going to be detrimental to that cohort. I wonder how many of the deceased, across all ages, were in good metabolic health?

Female Excess Mortality Table

Male Excess Mortality Table

---

Female Mortality Distributions

Death Expectation as a Function of Birth Year

Male Mortality Distributions

Death Expectation as a Function of Birth Year

---

Notes

A note on the model for those who have not followed my previous work. In spite of claims that it is not possible to find a single, reliable, excess mortality model, my overarching principle is that the mortality probability distribution actually is the only reliable model.

Indeed, probability distributions are also known as models of “expectation” so it stands to reason! Thereafter, no-one disagrees that excess is simply the difference between actual (observed or “empirical”) deaths and “expected” deaths in any one period.

Using this most intuitive underlying model eliminates all of the issues that plague other naïve models, like changing population sizes but, just as importantly, changing morbidity, over time. In addition, careful consideration has been taken into using occurrence-date data and sex stratification.

Arbitrarily, I took a mortality year to span July to June, in order to avoid seasonal effects. This conveniently isolated “COVID” in 2020 and the jab in 2021.

For both sexes, it is apparent that just two Gompertz functions are required to fully characterise the entire distribution post 21 years of age, for both sexes¹. The models were fitted numerically (i.e. solving iteratively for the three parameters of the Gompertz model), by minimising the squared error between the empirical data and the data implied by the model.

Only data up to 2019 was included in the calibration process. Due to limited availability of empirical data², unknown historical and future data was also derived from adjacent models in a stepwise fashion, e.g. for model M, data at M+t+1, where t is the last known empirical point, was taken from model M+1+t and so on…

Evidently, this method produces consistent and robust results, as can be observed in the time-lapse videos above. However, ultimately, it is still a model and therefore, not perfect (but still much better than cross-sectional models by age!). With limited empirical data and more unknown future points than known for the younger ages (where the model is most sensitive), margin of error should be considered.

1. At least two more distributions are required to accommodate the distinct mortality distributions of the under two’s, adolescents, and especially in the case of males, a pronounced hump in mortality between the ages of 15 and 21. However, they do not impinge on this analysis. Regime 3 remains on the legend, because it affects mortality under the age of 40, which was modelled, but cut off from the charts for better resolution.

2. Daily occurrence deaths by sex, year of birth since 1988 to most current day, people born in England and Wales and people born elsewhere. This ONS dataset only starts in 1988, which is when it was stratified by place of birth. There is similar data between 1970 and 1988 which includes those born elsewhere, which can be manipulated to provide a longer series of empirical data.