Why Are UK Covid Cases Declining?

As many of you know, the UK declared July 19 as Freedom Day, when nearly all remaining Covid restrictions were to end. We didn’t say much about it because so many in the Twitterverse and the British press declared it to be a terrible idea, given that only 56% of the population was fully vaccinated (although that does amount to 68% of adults). That’s not enough to reduce R0 to below 1, given how contagious the Delta variant is.

It didn’t get off to a good start. From Reuters:

Prime Minister Boris Johnson’s ‘freedom day’ ending over a year of COVID-19 lockdown restrictions in England was marred on Monday by surging infections, warnings of supermarket shortages and his own forced self-isolation….

But Johnson’s big day was marred by “pingdemic chaos” as a National Health Service app ordered hundreds of thousands of people to self-isolate – prompting warnings supermarket shelves could soon be emptied….

Britain has the seventh highest death toll in the world, 128,708, and is forecast to soon have more new infections each day than it did at the height of a second wave of the virus earlier this year. On Sunday there were 48,161 new cases.

Yet here we are, a bit over a week later, and UK Covid cases are down, despite the announcement of what looked to be superspreader events in the making.

Here are two contrasting takes, the first from the Financial Times, which takes the fall at face value, and then from Richard Murphy, who speculates that it is to a significant degree a data artifact. Of course, this could also be “Covid moves in mysterious ways” (recall the unexpected decline in the US after the year end through third week of January surge, which was anticipated to get even worse) or another example of Boris Johnson again being lucky.

First from the Financial Times:

A sharp fall in the number of people in the UK testing positive for Covid-19 over the past six days has surprised and delighted scientists who feared that cases would rise towards 100,000 a day, as health secretary Sajid Javid had warned in early July.

The latest daily count of 24,950 new cases, released on Monday afternoon, was the sixth successive decline — less than half the peak of 54,674 reached on July 17 and the lowest for three weeks…

But other experts said they had already dismissed scenarios in some epidemiological models that showed positive tests increasing beyond 100,000 a day later in the summer as a consequence of the government removing almost all legal curbs on social interactions from July 19….

The reasons why cases are dropping so fast now — and the prospects for the next few weeks and months — are far from clear, so Downing Street struck a cautious tone on Monday….

“The daily test numbers will only begin to see the effect of the end of lockdown towards the end of this week,” said Professor James Naismith of Oxford university, director of the Rosalind Franklin Institute….

Finn listed some of the factors at play, including infection-induced immunity, vaccine-induced immunity, and behaviour. “We still have enough non-immune people around to reverse this trend if we completely stop trying to avoid spreading the infection,” he said. “But with every passing day another cohort of people, recently immunised, is added to our protection alongside those who have recently had the infection, survived and recovered.”….

Some of the past week’s fall in cases may be due to a decline in the number of people taking Covid-19 tests, perhaps because they want to avoid self-isolation, but experts say the drop is too steep for that to be the main reason.

Another factor may be being past the Euro 2020 football events. Fans would congregate at pubs and homes to watch and cheer.

Hospitalizations, which lag, have not yet moved down.

Now to Richard Murphy. Note that he was highly critical of the “Freedom Day” gambit.

By Richard Murphy, a chartered accountant and a political economist. He has been described by the Guardian newspaper as an “anti-poverty campaigner and tax expert”. He is Professor of Practice in International Political Economy at City University, London and Director of Tax Research UK. He is a non-executive director of Cambridge Econometrics. He is a member of the Progressive Economy Forum. Originally published at Tax Research UK

On 17 July there were 54,674 new Covid cases. There were forecasts that cases could reach well over 100,000 a day. On the basis of reasonable exponential growth expectations that seemed likely. And on July 25 there were 29,173 cases.

What is happening? No one can be sure. Hospital admissions and deaths are, of course, lagged, so they will still rise for a while even if this data is true, which it could be.

And, there again, it might not be. Let’s consider those possibilities.

First, schools in England, and more especially the parents of schoolchildren in England, are not requiring testing anymore as it is summer holiday. So who knows if there are cases there?

Second, people want their summer holidays. Why are they going to get tested, which might prevent that?

Third, they are told there is a ‘pingdemic’ so they delete the NHS app and do not report that they have a positive test, even if they have one.

And fourth, they just don’t care anymore: they are now convinced reporting having Covid does not help, so they are not doing so.

What is the chance that these factors are the dominant strain in the reactions to ‘Freedom Day’? Right now I think they are likely. And as evidence, I note that the so-called Zoe app, which measures case rates voluntarily reported by a rather self-selected cohort, thinks that they are still rising. https://covid.joinzoe.com/data#interactive-map

Let’s be clear: I do not know what is happening. But given all the narratives that are being pumped out by the government and media I think the likelihood that cases are simply not being reported anymore is very high indeed. And that is worrying.

Print Friendly, PDF & Email


  1. vlade

    IMO, testing is a confounding indicator. One has to see hospitalisations, so we’ll have to wait for a bit yet.

  2. PlutoniumKun

    I’ve been banging on this for a year, but I think there are a lot of question marks over UK figures as there has been no consistent match up on either side of the Northern Ireland border (whatever the problems with the Irish response, testing has been reasonably competent). Once again, we’ve recently seen a flare of Delta on those parts of the Irish border where there is a lot of cross-border traffic, with the declared figures in the Republic being higher than in Northern Ireland. It makes very little sense unless you assume that either they are over counting in Ireland (unlikely) or undercounting in the UK (likely).

    I think holidays are a likely reason – people simply don’t want a positive result just when they’ve had their long awaited summer break, so they ignore symptoms. Since most cases are among young people, symptoms are likely to be relatively light and dismissible as a bad cold or allergy (especially as the symptoms for Delta seem quite different from the first wave).

    A further possibility is that unusual weather has kept people from mixing as before.

    But as Yves observes, we’ve seen waves of the virus stop with surprising rapidity for reasons that seem to defy the models, so it may well be that as usual, Johnson has lucked out.

  3. Terry Flynn

    Correction to Murphy re holidays. Schools here in Nottm are NOT YET on summer holiday. My nephews (year 7) are at school TODAY. Yes, the holiday begins any day now, and yes you’re right to question why on earth term didn’t end on Friday but the fact is it didn’t. It ends one day mid-week THIS WEEK.

    Where he IS correct is cancellation of holidays. I am fairly sure my relatives would have done an overseas holiday to a green or amber zone destination. The fact these are changing status daily is mainly what has made them plan “staycations”.

    We don’t have enough data. Thus I am forced to speculate from medical doctor friends and their anecdotes – testing is apparently being reduced and we’ve reached a “lower equilbrium” of key workers (including hauliers and those who restock supermarket shelves). Shelves have NOT been emptied of toilet roll and flour but the amounts are obviously low and perilously low – such that any new “panic” would empty them again just like last year.

  4. Lou Anton

    Also very confused by this. Seeing the case count skyrocket heading into Freedom Day and then plummet immediately is certainly odd. Inclined to give it 2-3 more weeks and see where we are. vlade is right about hospitalizations, but if the decline in infections is legit, then hospitalizations ought to flatline in the next few weeks too.

    For UK folks, was July 17th really that big of an event (like a New Year’s Eve sort of thing?)? Did people make a point to get tested, was there a big push to find out for sure if you were virus free? Right now, I think Richard Murphy’s #4, people don’t care, has the highest chance of being right (though the test frequency data don’t decline as fast as the positives).

    Maybe Boris is the luckiest man alive.

    1. Redlife2017

      Honestly, in my part of London not much changed. People who can still work from home are. Most big companies that did WFH are saying full hybrid working will start 6 September. We still wear masks in most establishments. Schools weren’t officially out until Friday, but so many bubbles burst that many kids and families have been at home over the past fortnight.

      One other thing is that even though people are getting rid of the app, obviously enough aren’t because how else could we have the “pingademic”? I mean having tons of kids and working class people at home over the past 10+ days had to have slowed the spread a lot too, I would have thought…

  5. Boomka

    June was mostly gloomy and rainy but by its end we had 2-3 weeks of lovely sunny weather, a heatwave even, with people crowding the parks and the beaches. Perhaps we are now seeing the results of a quick and massive vitamin D boost in the population.

  6. Gerd

    What about the Euros? Lots of people getting together in pubs and homes to watch the games. The final was on July 11th (England were in the final). Four days later is July 15th.


    Just saw on re-reading that Yves mentioned the Euros in her intro.

    1. Terry Flynn

      Fairly sure this is a factor. No official data (rolls eyes) but anecdotally lots who are now eligible finally took up their jab in run up to final or, if vaccinated, took it as carte blanche to ditch masking. Schools are still petri dishes to spread the virus. Try walking around Nottingham in a quest for a masked kid. Should be the new Pokemon game.

      I believe delta has gone wild – fairly sure both parents and me have had it (all vaccinated early) but by time we suspected, we knew it was too late. Lo and behold the suspected superspreader at dad’s factory (who WAS tested) got a negative but told he “probably came for testing too late” – he was already practically well again. No antibody test.

  7. The Rev Kev

    I regret to say that this may be a case of having to wait to see how it plays out. For all we know, the UK is in the middle of the eye of the hurricane or maybe it is just Boris playing funny buggers with the official statistics. And if it does blow up, then we will be able to pick out the causes of it later using the facts and figures whereas now we are still blind. When the pandemic started I thought that the developed countries would be a good source of information and statistics to understand this virus with. How little I understood. All I know is that as far as the UK is concerned, “It’s Quiet……Too Quiet.”

    1. Brian Beijer

      I know! I too was so naive that I assumed they wouldn’t try to bugger with the statistics. I thought, what motive would they have to mislead people into letting their guard down and causing unnecessary illness and death? How would that be to any country’s advantage? I couldn’t have been more wrong. The only people truly prepared for this crisis were the accountants.

  8. polar donkey

    Back in early spring there was an uptick in cases. A friend who is a nurse at a minor medical clinic said almost everyone who came in with minor symptoms were convinced they had allergies. The nurse had to argue with them to be tested for covid. She took a guilty pleasure in telling the most adamant deniers they had covid.

  9. Ensign 19

    Published case numbers are inherently unreliable. If you want to know the actual state of play, look at hospitalisations (even these involve some discretionary subjectivity), or even better, deaths. Death stats are pretty immune to subjective ambiguities or gaming, and better documented.

    I do understand that because hospitalisations/deaths will lag behind test result numbers, it’s tempting to give test results more weight than they may merit to seem “current”.

    1. amused_in_sf

      But even that approach is flawed, depending on the question you are asking. You need to age-adjust (and comorbidity-adjust) deaths and hospitalizations to get an accurate picture of infection rates, since the virus does not cause severe reactions randomly, and age (and other) cohorts are not vaccinated at the same rates.

      Since I am mostly worried about getting a moderate case through purely discretionary activities (i.e. I’m vaccinated), positivity rate is still the best indicator for me, with the caveat that the error bars are large.

  10. amused_in_sf

    Since the positivity rate is going down as well, and the number of tests doesn’t seem to have fallen massively, there doesn’t seem to be an obvious problem with the data.

    Murphy’s reasoning is a little wacky on a number of counts: a) if you remove children from the testing regime, I would expect positivity numbers to increase (since infected kids are not as likely to be positive as long as adults), unless schools are massive spreaders, which we haven’t seen, and b) even if someone removes the NHS app, surely the testing center would still report the positive result!

    Considering that vaccine uptake skews towards the older population (more so in the US, but still true in the UK), and mass socializing is skewed towards the young, it’s very possible that new infections are happening more in people that will show no or minor symptoms, and thus don’t get tested. Thus, we need to recalibrate the multiplier for test results->actual infections.

    If that multiplier has doubled (which seems within the range of reasonableness to me), then this wave is similar to previous ones and could have passed its peak.

  11. Anonymous 2

    The conflict between the official figures and the Covid Zoe app figures is a puzzle. Generally my understanding is that the British authorities and Covid Zoe work together to try to understand what their figures – collected differently – are showing and adapt their methodologies where it becomes clear that something is going on which makes one or other set of figures unreliable. And yet at present the two sets of figures have been going in opposite directions – the Government showing falls while Covid Zoe show increases (including quite a steep rise in my area – up 25% in the last week). Something is not right but it is unclear at present where the errors are.

  12. Eric

    I’m sorry, but I think you’re wrong on this one. Firstly, to discuss your first argument: The UK has had a fairly steady pace of vaccinations from January through to about May. There’s no big lump of increased vaccine prevalence that could drive a spike like that.


    Secondly: the UK does PCR on nearly all positive tests. That looks for viral DNA being used to produce the spike protein. In the months after a vaccination that would be zero as the vaccine (mRNA or Adenovirus) would no longer be present producing spike protein.

    Lastly: the UK does very good tracking of the prevalence of Delta in the positive samples. Right now Delta is nearly all samples. The vaccines were made based on the original wild type virus so I don’t think that the Variant testing would show that if it was the case.


    What I think is more probable is that with Delta being much more transmissible, and with schools open it established itself in the schools where it had free run of a non-vaccinated and non-masked population. I attribute the drop to the combination of the Euros and the pingdemic working. At one point in July 2 million students were off school because of a positive case in their school bubble.

  13. Peter Dorman

    I’m not an epidemiologist (I always seem to be prefacing my thoughts with disclaimers like this these days), but wave dynamics seem poorly understood with Covid. Not so much the upswing, but the stall and fall. There are speculative reasons offered, which tend to treat the transmission dynamics as if the population were uniform.

    But consider the possibility of population curves as amalgams of smaller wavelets built around transmission networks (family, friendship groups, coworkers, etc.). There’s a lot of transmission within these networks and less between them. In addition, transmission is highly variable: most infected people are not very infectious but a few are highly infectious. If so, it’s not difficult to model a self-limiting process as immunity increases within networks encompassing superspreaders, even though the increase in immunity on a population basis is modest.

    But this is a speculation like everyone else’s.

    1. Cuibono

      actually not so speculative. there are many many papers out there form 50 years on stochastic forces in epidemic curves

  14. Cuibono

    How many people have Covid-19 antibodies in the UK?

    Jul 7, 2021

    While the UK has long abandoned the strategy of achieving herd immunity by allowing the virus to spread throughout the least at risk sections of the population, the availability of vaccines finally opened up the possibility of getting there without having to sacrifice the public’s health. As the latest government figures show though, all of the UK has now reached the 80 percent mark. As of the week beginning 14 June, estimates based on blood test results taken from a randomly selected subsample of the adult population, indicate that Wales had the largest share of people with the relevant SARS-CoV-2 antibodies – 91.8 percent.”

    1. Biologist

      That’s of the adult population, not the whole population. >20% of the population is below 18 years old.

  15. Terry Flynn

    My nightmare scenario has been sketched out already on NC. The Olympics, via its IOC “nobody should have vaccine advantage” ideology, creates a new variant that is worse than delta. It builds on delta (which has gone as far as possible via the spike protein(?)) and produces a variant that is even more contagious, minimally vulnerable to vaccines (especially the new mRNA ones) and which degrades organs fast. My BFF has lived in Japan for 2 decades and is not at all hopeful that even the ultra-efficient Japanese can keep the virus under wraps.

    This would explain the likely temporary benefit us Brits are experiencing. Longer term? Welcome to the post-Black Death world of 1/3 of people gone.

    Anecdotally all my medic friends have started chatting more about “fun with kids” stuff and don’t/won’t discuss covid. These are family physicians, radiologists and ER docs. THey just refuse to engage, like “we know what’s coming and we’re going to enjoy the moment”.

  16. Hayek's Heelbiter

    Bulletin from Brixton:
    A lot of my friends are in their late 20s or early 30s, who avoid the government track-and-trace app with contemptuous disdain.
    Within the past few weeks, about a dozen have caught [suspected] Covid. Only one or two have been officially tested. Most do the lateral flow tests at home, which are less accurate than PCR tests. And more than a few are anecdotally self-diagnosed.
    A conversation I had with a recovering [unvaccinated] friend about half an hour ago, “I spent the weekend with J and A, and later both had Covid. A few days later, I started feeling awful and lost my sense of taste, so I did a home flow test, and sure enough, two bars.”
    Like many of our friends, he decided to tough it out. He did not see any doctors or got the hospital. Friends shopped for him and did his laundry till he felt better. His taste and energy levels are slowly recovering.
    Out of the dozen or so people I know who did get [suspected] Covid, it would not surprise me that only four or five showed up in official statistics.
    Make of it what you will.

  17. R

    I have watched the data here in the UK like a hawk and I think the effect is real.

    The effect is most pronounced in Scotland and Wales. These are anti-Bojo places with no reason to fiddle the figures for him (OK, still reason to fiddle for themselves!). Scotland had a terrible early delta outbreak which has now subsided. Working hypothesis is this was seeded by the Scotland game in the Euros bring delta back from London -rates in young men were six times those of the rest of the population afterwards – and stopped by the early end to the school and university terms in Scotland (that far north, the summer is short and school’s out early to help in the fields).

    England lasted longer in the Euros (right to penalties…) and school terms are a mess, breaking up any time from late June (private schools) to late July (state schools) because of bizarre state insistence on quality not quantity plus private school parents expecting discounts on early holiday fares to offset school fees! Euros now over, schools mostly finished, wave subsiding.

    If you check the University town I live in, May and June had insane case numbers on campus as they seized a few weeks sunshine to socialise again. Then they all went home to infect their parents and cases dropped dramatically. Then the Euro finals wave hit and is now subsiding and the schools have broken up.

    Rather than worry if the drop now is real, worry if it is sustainable. Because in September, the mixing machine if university and school starts up again along with return to work after the summer holiday and the return of indoor weather….

Comments are closed.