The Strange Story of Week 48

11 December 2020. Updated 14 December 2020.

by Guy de la Bédoyère

Dr Mike Yeadon’s fascinating ariticle on the Pseudo Epidemic got me thinking. The Government likes to highlight how bad this year has been by citing only the five-year average of deaths by comparing one week with its counterpart over previous years. So, I decided to play the same game but by looking back over eleven years. What the Government also doesn’t appear to do is consider the fact that the population is rising annually in the UK. I therefore divided the population for each year 2010–2019 into 67 million (which is approximately what it will be at the end of this year) to provide me with a factor I could use to multiply the death figures for a given week in England and Wales (Week 48, this year ending November 27th for which the ONS has supplied a figure for registered deaths of 12,456) to get a truer picture of the impact of Covid. The death figures on the left and the annual amount of the England and Wales population are taken from the ONS.

YearDeaths for Week 48 (England and Wales)Population of England and Wales (millions)Weighted deaths for Week 48 (based on a continuous 2020 England and Wales population of 59.7 million)
2020 (theoretical non-Covid)*11,13159.7 (estimated)11,145
2020 (actual)12,45659.7 (estimated)12,456 (3,040 CV, 9,416 remainder)

*Had there been no Covid then based on existing trends this is what Week 48 in 2020 might have looked like.

Quite clearly, the deaths in Week 48 this year look dramatically higher. But we can also see immediately is that although the increase in the number of deaths is not continuous, the overall trend up is. Had there been 59.7 million people in England and Wales back in 2010 then Week 48 in 2010 would have seen 9,899 deaths and that figure ought to have remained pretty constant in the right hand column had the population not changed.

But that isn’t what happened. Had there been 59.7 million people in the England and Wales in 2019 then Week 48 would have seen 11,013 deaths, by weighting the actual figure. The increase in deaths during Week 48 over those ten years was 11%. In other words, by 2019 in the UK 1.11 times as many people died in Week 48 as a proportion of the population as had done in 2010. In 2010 Week 48 saw 0.016% of the population of England and Wales die. In 2019 it was 0.018% and the trend upwards throughout that period is clear.

It all depends on what you compare with what. Between 2015 and 2019, for example, based on a weighted continuous population of 59.7 million, the overall increase in registered deaths in England and Wales in Week 48 was 8.6%, an annual average increase of 1.72%. Between 2010 and 2014 the average annual rise was only 0.86%. Part of the explanation is an ever-ageing population (those over 65 increased from 15.9 to 18.3% of the total UK population between 1998 and 2018). Since the final death rate is invariably 100% (we are not immortal, whatever the Government likes to think) an ageing population will always exhibit a rise in the overall death rate, which appears to be quickening. This is very much a modern phenomenon and is caused by improved medicine, nutrition and housing, and birth control, in developed nations.

Given that average rate of increase in deaths in Week 48 at 1.2% per annum over a decade we could legitimately argue by using the weighted figures that 2020 in theory would have seen 11,013 (2019) x 1.012 = 11,145 deaths in Week 48 anyway. But bear in mind that all deaths in 2019 and earlier are non-Covid deaths. If we apply the average annual increase in the death rate in Week 48 from 2015–19 which was 1.72% (1.0172) then we’re up to 11,202 expected deaths for Week 48 this year, Covid aside. In the overall period 2010–2019 the number of the England and Wales population dying in Week 48 seems to have risen by 11%.

By weighting the figures, I am only looking at the theoretical rise in deaths as a proportion of the population. If we were to stick with absolute numbers, the effect of a) a rising and ageing population AND b) the rise in the proportion of the population dying in Week 48 since 2010 means that was an increase of 18.9% in absolute numbers of people dying between 2010 and 2019 in Week 48.

Back to this year. Since the ONS helpfully tells us that 3,040 registered deaths were attributed in Week 48 to Covid, then the total 12,456 registered deaths figure for the week must mean that deaths from all other causes were only 9,416.

Anything less than 9,416 in absolute terms has not been seen in a Week 48 since 2013. When weighted for the lower size of population we can see that Week 48 in 2020 has allegedly seen the lowest death rate from all causes other than COVID-19 for over a decade.

The ONS is thus claiming that 24.4% (1:4) of all deaths in England and Wales in Week 48 were attributable to COVID-19 at a time when, by some extraordinary chance, deaths from all other causes dropped so dramatically that had it not been for COVID-19 we would have been exhorted to dance in the streets for joy at the nosedive in UK deaths to the lowest level for more than a decade, if not longer.

The ONS also claims that the deaths in Week 48 were 20.3% higher than the five-year average (at 2,099). This of course conveniently masks the fact that, as I have shown, there has been an apparent steady increase in the rate of death in Week 48 (and doubtless other weeks) for years. And there’s nothing special about Week 48, I can assure you. Looking at one week in one year on its own is of course potentially flawed, but not when one looks at that week over a much longer period of time. The ONS and SAGE make much of the idea that this year’s figures for many weeks are ‘higher’ than the five-year average. But since, as I have shown, the trend was rising anyway then of course this year’s figures will be ‘higher’ than the average. The question is how much of that rise is attributable to Covid and how much to existing trends.

The alternative way of looking at that is to refute this on the grounds that it is impossible. Therefore, we must consider that whatever remains, however improbable, must be true, which is that the figures have been presented in a way that exacerbates the impression of Covid’s impact.

The 3,040 deaths attributed to COVID-19 by the ONS can be reduced immediately to 1,311. This is achieved by taking the 12,456 recorded deaths for Week 48 and subtracting the deaths we might have expected that week anyway, based on the last 10 years and the mean increase as weighted against a rising population. Since that figure, 11,145 deaths, does not take into account 2020’s attested increase in deaths from suicide, untreated conditions and so on we could theoretically reduce the 1,311 further by assuming at least some of those additional deaths are attributable to non-Covid factors. Deductions could also be made for those whose deaths have been attributed to Covid when in fact they died from something else and/or thanks to false positives (although, to be fair, we may have already stripped them out by reducing the number of Covid deaths in Week 48 from a putative 3,040 to a more realistic 1,311). Needless to say, all these are impossible to unravel and take us into the realms of probability and speculation.

Of course, it is still manifestly the case that Week 48 in 2020 has shown a noticeable increase in registered deaths over previous years, much of which is attributable to Covid, though not to the extent the ONS would have us believe. The proof will be in the longer run. As I pointed out the other day on Lockdown Sceptics, basing statistical comparisons on short periods (i.e., a single week in isolation) is not a sound basis for assessing the true impact of unusual conditions, since it is as absurd as assuming any given week in a year is exempt from ever experiencing extreme weather that bucks the trend.

The first two months of 2018 showed a dramatic increase in overall deaths on a weekly basis, attributable to a flu epidemic, and an ageing and growing population. The overall and excess figures matched or exceeded those that have been made so much of for October and November this year including Week 48. They were also worse than previous January and February of recent years but excited no comment at all at the time – or Government intervention. If we really are in the midst of a major epidemic now, why weren’t we in one in early 2018?

In the end you can make of these figures what you will. The absolute truth is as ever elusive. There’s no attempt here to pretend we don’t have a serious public health issue to confront and deal with. But it doesn’t take much scrutiny to see what some statisticians call ‘Number Theatre’ is afoot and as we’ve all discovered the price of a ticket to see the show is exorbitant.

Social media & sharing icons powered by UltimatelySocial