UK’s dodgy data

We seem to have a standard practice now of pushing data back in time just as significant government announcements are made.


On 23rd June, Boris Johnson said restrictions were being reduced and that the death rate was going down. Deaths in the last 24 hours were 171.  This would continue the downward trend at around 3% per day.

But the published data on June 22nd and June 23rd showed a total number of deaths increasing by 280!  This would have been a bad figure for the government to publish on the very last daily briefing as restrictions were being lifted. 
Johnson waffled  that 109 deaths from April, May, and June were added which would account for the discrepancy.  Johnson is not known for his scrupulous honesty.

 If the figures are really going up on 23rd June, this would imply a rise of infections in early June.

From Tracking 7th June
Restrictions have been loosened as of 1st June and any changes in the death rates will not be apparent until mid June.  At the same time demonstrations are taking place in the USA and the UK over racial equality and this will exacerbate the spread. It will also make it impossible to point to any reason for a change in infection rates.

Over the last 14 days 1677 people have died from the disease in England. With a mortality of 1.4% this implies that about 120,000 people currently have the virus in England, or one person in 500.

It easy to understand why most of the scientific community think that it is to early to ease the lockdown to the extent announced.  Currently England is leading the way out of locldown, despite having twice the death rate than Wales, four times the death rate than Scotland and 8 times the death rate than Northern Ireland.

As regards the data, the question is:-

“Will bad days be pushed back again to ease the governments embarrasment?”

So, do I think that the government is manipulating the data so that it looks good on the days of the announcements of changes in the lockdown? 
Well, two announcements and two “adjustments” on the same day, and no adjustments at any other times is not conclusive, and could be co-incidental.
It is a worry though. About both the real situation of the pandemic in the UK and about the trust in official government statistics.

The effects of the latest easing will be seen in deaths – around the time of the next announcement.


Lockdown easing began on June 1st

On 31st May, a total of 38,489 deaths were reported on the government stats. See fig 1.

On 1st June a total of 39,045 deaths were reported. making a rise of 556. BUT only 111 were reported as the death total for 31st May was retrospectivley increased to 38,934  See fig 2.

So what happened to this over 400 extra deaths?

It turns out that most were re-allocated to the 24th May

This was probably (hopefully) simply re-allocating deaths to the day they actually occured.
Before the change the data exhibited a worry rise in death rate from May 31st with the 7 day total up to 2nd June being the highest figure since May 23rd.

This “pushing the data back” enabled Matt Hancock to announce the number of deaths on June 1st to be 111, the lowest number since lockdown.

Using the governments’s own data from the previous day, the figure was 556, the highest number in three weeks.


Figures below shich are snapshots from the governments website show how numbers were altered for the period between 23rd May and 1st June

Interestingly, as this surge in deaths on June 1st would have made the growth rate in those days positive and therefore an R rate of greater than 1.  See fig 6.

Fig 1. Covid 19 Deaths as reported on 31st May
Fig 2. Covid 19 Deaths as reported on 1st June
Fig 3. Covid 19 Deaths as reported on 31st May
Fig 4. Covid 19 Deaths as reported on 1st June
Fig 5 The effect of the changed data on the weekly death rate.
Fig 6 The effect of the changed data on the measured growth rate.

On the left is the current data, on the right is the data had the change not happened.