Is the lockdown killing people?

I’ve published an article at The Critic, that most excellent new UK magazine. It’s on the issue of whether the excess non-Covid deaths are undiagnosed Covid deaths, or caused by the lockdown and the associated hysteria.

Regular readers may remember I wrote a blog post a while back about this, looking at whether sex differences allowed us to decide, and I concluded that the data was not conclusive on this issue, but since then the week 16 data has come in, and this really does have a large excess of female deaths, which doesn’t support the undiagnosed Covid hypothesis.

The data isn’t conclusive, though, and there may be other explanations. But the issue of whether the lockdown/panic itself is killing people has been little discussed, and there’s been virtually no acknowledgement that the ONS figures show that there are thousands of excess non-Covid deaths, which demands an explanation.

Andrew Lilico did send me this good StatNews article, though:

The same is true for appendicitis and stroke. Clinicians say patients with these life-threatening conditions have also stopped seeking treatment in large numbers. “My worry is some of these people are dying at home because they’re too scared to go to the hospital,” Gulati said.

Others are coming in so late, she added, that some are presenting with massively damaged hearts, including heart muscles that have ruptured. “That was something I’d only seen before in textbooks, to study for exam questions,” she said. “Now we’re seeing those cases because people are putting off care.”

But the really astonishing news is something that no-one else has even noticed. Professor Neil Ferguson himself has admitted that the lockdown is probably causing deaths. In an interview for Unherd he said the following:

18:42: “Some of the excess deaths we’re seeing at the moment, all-cause excess mortality, is probably down to people not being able to access healthcare for other conditions.”


20:04: “I’m sure part of what we see when we look at these enormous spikes in excess deaths in heavily affected countries is not just undetected Coronavirus deaths, it’s people dying of other conditions which otherwise might have been treated.”

He has an excuse, of course:

21:07: “All of those effects would have been much worse had we not adopted lockdown. We would have seen enormous spikes in mortality in a sustained period of several months of not having healthcare capacity for other types of patient.”

He doesn’t mention the fact that his claim that non-Covid deaths would have been even higher without lockdown (due to overwhelmed hospitals) comes from his computer model, and is not supported by any real-world evidence. In fact, in light of Sweden health systems doing fine we can say that his claim is very unlikely to be true.

Share this article on social media:

27 thoughts on “Is the lockdown killing people?

  1. Lovely, no more long wait to load and an attractive appearance! I do hope the link to Idiots will return in due course.

  2. Congratulations! That is an interesting article in the Critic. The data is noisy and subject to potential biases in several directions, which makes it tough to prove any hypothesis conclusively at present.

    Something which does seem to be uncontested is that there are more women than men in the prime At Risk group of the old & medically-compromised. And there are anecdotal reports that hospitals have unloaded many of their “normal” patients (probably many of them in the At Risk category) to care homes, where they have died — whether from catching C-19 or from lack of adequate medical facilities. Combining those factors, could we be seeing an acceleration in the deaths of old ladies?

    If that hypothesis is correct, then we would expect to see weekly ONS Reported Deaths at some point in the future fall below the 5-year average, as people who would have died later in the year depart earlier than usual. Time will tell !

  3. I tend to agree that deaths are likely to slow to below average over the summer and perhaps into next winter. There has been a big clearout of God’s Waiting Room. Interesting is what will happen longer term: a. WuFlu fades out as a factor or, b. COVID-19 mutates fast enough to dodge a vaccine and retains its culling effect so that we see an end or even a retreat in our expectations of ever increasing life expectancy.

  4. What gives you so much confidence that any difference you detect is due to “the lockdown”? At the start of the article you write it as “the lockdown and accompanying panic” but by the end you have dropped the latter. As I have said before, the lockdown has no legal restrictions saying “patient, thou shalt not seek medical help for thy heart attack” nor “doctor, thou shalt not treat thy patients for cancer”. So if you’re blaming the lockdown then you can’t really blame its rules directly, rather because it has certain knock-on effects. But the lockdown isn’t the only thing going on at the moment. There’s a pretty nasty, potentially fatal, pandemic going round and it is concentrated in healthcare facilities. And healthcare services are being reconfigured, partly to make space for COVID patients and partly because certain procedures are currently deemed unsafe.

    The presence of the pandemic itself induces panic and its concentration in healthcare facilities substantially discourages care-seeking behaviour among patients, potentially postponing it until it is too late. There’s also the fact people feel bad about using healthcare resources while the service is overstretched and are constantly being told not to misuse the NHS if they’ve got something minor (for the non-Brits, we have been conditioned by decades of advertising on this issue), which has a nasty side-effect that too many people try to “struggle on and get by”. I accept the argument that lockdown may increase levels of fear and make people feel guiltier for going to hospital but to the extent it makes a difference it is only incremental. I would contend that many people are panicking primarily because there is a bloody nasty pandemic and many of us know someone who is hospitalised or dead as a result of it, or have friends or neighbours whose relatives have died. It is not a straightforward exercise to apportion how much panic is fundamentally pandemic-related versus how much is lockdown-related. If you want to argue that the lockdown is the strongest cause then you need to actively make a case for the way you are disentangling the factors.

    The healthcare reconfiguration is also separate to the lockdown – in my area, well before the lockdown the GP surgery had already shut for face to face appointments and the hospital cancelled elective surgery, outpatients appointments and a lot of screening and scanning. Surgeons everywhere are considering what surgical options are sufficiently safe or urgent when the risk of acquiring infection in hospital is high, and similarly chemotherapy and radiotherapy are being postponed. Easing the lockdown isn’t going to help with this access issue.

    Even if you don’t trust Ferguson’s model you need some way to explore a counterfactual – assuming the UK would look exactly like Sweden is also wrong to some extent due to differences like Sweden having more one-person households (we know a lot of transmission occurs within households), so even if you use Sweden as the model for your counterfactual you still ideally need to make some kind of statistical correction to it. There is a strong intuitive reason to think lockdown reduces new cases, since we know COVID is an infectious disease and we know lockdown has reduced interpersonal contact. If there hadn’t been a lockdown it seems likely on those grounds alone that more infected people would have ended up in hospital, and certainly unlikely that there would have been fewer infections and hospitalisations. It’s therefore hard to see how in “no lockdown” world, the situation with access to healthcare could have been better than today in terms of cancer treatment, surgeries, screening etc, even if there might be floor effects that prevent it being any worse (eg if all treatment of a certain type is cancelled with lockdown levels of COVID patient load, then it couldn’t get any worse even if COVID patient demand was much higher).

    At present much of your argument rests on “anything that is going wrong, beyond direct deaths from COVID, is the fault of the lockdown”. Health service reconfiguration and fear of COVID itself (and especially of COVID hotspots like hospitals) are also very plausible contributors. Do you have any methodology that will allow you to distinguish between them?

  5. I agree with Gavin above, the excess female deaths will be a consequence of a combination of a) lots of elderly female hospital patients being shoved out of hospital back to care homes and them dying there, either of CV-19 or just lack of the correct medical care, and b) the virus getting into care homes and killing the residents, a large proportion of whom are female.

    So not exactly lockdown related, but definitely partly the consequence of the NHS’s reaction to the crisis.

  6. The lockdown was specifically ‘sold’ to the public as something that was required to ‘protect the NHS’. The two are, both in the public’s mind, and the pronouncements of all the public officials, intimately linked. It wasn’t ‘Stay at home otherwise you and your loved ones might die’ it was ‘stay at home to protect the NHS’ That was the very kernel of the lockdown policy. So if that policy is having the effect of making the marginal non CV-19 patient reluctant to call 999 or speak to a doctor because they don’t want to ‘burden the NHS’ then the policy is to blame for that. At no point in all of this have the talking heads ever said – ‘If you are ill you must call a doctor or dial 999 regardless of the CV-19 situation’. Its only now that evidence is piling up that the public are not seeking medical attention despite it being necessary that people are starting to speak out. It wasn’t that hard to predict, so one assumes the rhetoric was specifically chosen from the word go to have the chilling effect on demand for NHS services that it has.

  7. That distinction isn’t clear in your Critic article, which probably ought to be standalone. In ordinary usage I don’t think the word “lockdown” (a specific government policy banning people from leaving houses except for certain purposes) incorporates also “any moronic utterances of Piers Morgan”. These deserve to be treated as separate phenomena because what government policy to restricting civil liberties ought to be in the face of a pandemic is a serious issue in its own right and stands distinct from the question “just how big a moron is Piers Morgan”. Moreover any government change to said legal restrictions is unlikely to stop Morgan, or much of the rest of the media, from bouts of moronicity.

    Even so, let’s accept that by “lockdown” you really mean “panic”, then how should one disentangle the results of the “panic” from the substantial changes to health services?

    And how can one then compare to a counterfactual in which government policy has turned in a desired direction but the media are still full of morons and our hospitals are still, inevitably, hotpots for a nasty infectious disease?

  8. I wonder whether the data don’t actually support the opposite contention – that COVID deaths are being over-reported, and in fact even more people are dying of other conditions under the lockdown than are being acknowldeged. I tried toying around with the Week 16 data by subtracting off a different number but a constant sex ratio for covid deaths. That way, the excess female non covid deaths can be made to largely disappear. It seems unlikely that non covid deaths would be more preferentially female than normal.

  9. Surely ‘lockdown’ here necessarily refers to its repercussions understood concretely, i.e. scaremongering and ensuing panic. Unless he was coming from a purely legal standpoint: the rights and wrongs of its efficacy and/or legal standing of curbs on civil liberties, which is plainly not the case.

  10. I definitely see the case for this. But I still think it’s worthwhile to distinguish the legal restrictions of the lockdown from the messaging that surrounded it. It is possible to agree with the restrictions but not the messaging for example. Your memory may be playing tricks on you as the senior medical bods have been very explicitly telling people who need care to seek it as usual throughout the crisis, though I’m not sure how well reported this was and I’m sure not everyone sat through all the news briefings. See Prof Powis at this news conference from the beginning of April, for example, saying what I think you’d want him to. This messaging didn’t seem to get incorporated into eg the public information advertisements that I’ve seen, sadly.

    Obviously there’s a tendency for people to feel like they don’t want to be a burden (“I don’t really ‘need’ the doctor, I’ll get over it” etc) so even being told to go to hospital when you “need” it is somewhat discouraging, and when that’s combined with all the talk of hospitals full of people dying from a nasty infectious disease it’s going to result in people staying away. Until hospitals are better at keeping “hot” and “cold” patients apart – something they’ve worked hard on, but is tricky to get right without fast-turnaround testing technology – a lot of people, especially the vulnerable, are going to regard hospitals as no-go zones.

  11. “The lockdown” is a very specific government policy response to the crisis – one of many, but it stands out as the greatest restriction of civil liberties this country has ever experienced, so merits special scrutiny. This is why I’m disappointed when I read thoughtful and considered pieces which claim to be discussing the lockdown but are, in fact, treating the word “lockdown” as if it is synonymous with something else. My feeling is this: if you mean the lockdown, say “lockdown” and if you mean panic, say “panic”. While any good critique of the lockdown would necessarily examine its impact in the round – including mass psychological effects, themselves likely unprecedented – identifying the lockdown as essentially synonymous with a wave of panic that swept over the country is wrong. It’s possibe to argue the lockdown was a rushed policy response that actually arose from panic sweeping through government institutions, and it’s possible to argue the lockdown contributed to the tide of panic in the general public. Nobody reasonable would argue they aren’t deeply interlinked. But even if the government had chosen a different set of policy responses that did not include lockdown, there would still be public panic about a lethal infectious disease, panic stoked by a media who know bad news sells better than good news, panic stoked by whatever the government and health services did instead (clearly no government could completely ignore a pandemic and any action or inaction would have some unpalatable effects, clearly health services could not remain open as usual) and panic within the government as the world they knew turns upside-down. For that reason the panic “ensuing” from the lockdown can only be understood as one component among the general panic, in my view, and certainly can’t be treated as the sole or even (without evidence or argument) primary determinant of healthcare-seeking behaviour.

  12. The lockdown has served 2 purposes. Firstly, it has almost certainly “squashed the sombrero”. This has created the time to build NHS capacity and hopefully saved more folk as the health experts get better at treating the bloody thing.

    Second, although it may not look like it, it has reduced the media noise targeted at the government, allowing them to get on with the job of governing during a terrible time.
    No government would have survived UK hospitals looking the Italian ones in late February or a daily death rate of 1000+ without a lockdown in place.

    Only a flipping long time period is going to say if it really worked or not, although Spanish Flu learning would indicate some form of lockdown can work.

    The government has done the easy part. How they unlock is really hard. Good luck to them.

  13. Quick estimate taking account of ages at which excess is concentrated.

    From the age breakdown of week 16 deaths, the effect is almost entirely in the over 90s, M:F of 1,879:3,278, 1,399 more F than M deaths and a total of 5,157 compared to a “normal” week of ~2,400.

    So, ~2,757 “extra” deaths. How many C-19 and how many non-C-19?

    The overall UK population has roughly 1:2, M:F in the over 90 population. C-19 deaths are ~58:42, M:F.

    If you assume that normally 2 over-90 women die for each man (roughly correct from ONS), take account of the sex ratio and C-19 mortality ratio and solve the equations for the number of C-19 and “normal” deaths, you find (I think) that the numbers can be accounted for as follows:

    C-19 deaths, M:F = 376:272
    => total C-19 deaths 648

    Other cause deaths, M:F = 1,503:3,006
    => total other cause deaths = 4,509

    A “normal” week has ~2,400 deaths of over 90s. So it’s roughly an excess from “non-C-19 causes” of death in over 90s of ~2,100, divided (by assumption) 700:1,400, M:F.

    If this is roughly correct, “excess non-C-19 causes of death”, presumably due to lockdown, are more than 3x C-19 deaths among over 90s, 2,100 to 648.

  14. “No government would have survived UK hospitals looking the Italian ones in late February or a daily death rate of 1000+ without a lockdown in place.”

    Perhaps the pandemic response is more about saving the govt than the NHS or lives. They weather images of people on trollies in hospital corridors during the winter, but that coupled with media coverage of daily Covid-19 deaths could be too much for any govt to withstand. After all, Covid-19 deaths are highly visible, whereas all the early deaths caused by the response and economic damage aren’t visible in the same way.

  15. “….. the senior medical bods have been very explicitly telling people who need care to seek it as usual throughout the crisis, though I’m not sure how well reported this was and I’m sure not everyone sat through all the news briefings.”
    This has only really been emphasised this last week. Before that it, was all protect the NHS, protect the NHS, protect the NHS.
    Sorry – protect our fantastic NHS.

  16. “…….. it has reduced the media noise targeted at the government, allowing them to get on with the job of governing during a terrible time……”
    I haven’t seen much evidence of governing being done!

  17. Lockdown has had one, and only one effect, and its negative. It has reduced the number of healthy people infected and elongated the time duration of the epidemic.
    The NHS increased facilities have not been used. Panic over ventilators was misplaced as , in general, they worsen not help the situation.
    We have locked up the healthy and infected the at-risk groups in hospital wards and care homes.
    It has been an unmitigated disaster.

  18. “Hopefully this theme will be faster loading.”

    It’s a blog.

    A web page with not a lot of need-to-be-dynamic stuff. As most blogs are.

    Why – since it appears to be a problem – should such a page take longer than a second (and even that’s too long TBH) to load?

    Seriously – blogs do not need to be shiny and have the latest and greatest JS, ajax, and top-to-bottom multiple calls back and forth to the database, and interactivity to be…


  19. > …

    .. or comment forms (with, no doubt, associated software) that decide that three periods in a row need to be converted to an ellipsis.

    Get rid of the bloat, get a faster website.


  20. Loading speed is now good – though I do have a problem with my internet provider being locally overloaded (small island, low bandwidth, too many users at once). It’s also good to have all the trimmings, too.

  21. I totally agree with you JimW, though it’s had one more effect. The trashing of the economy has made the lives of the economically vulnerable much more precarious.

  22. Myburningears: “At present much of your argument rests on “anything that is going wrong, beyond direct deaths from COVID, is the fault of the lockdown”. Health service reconfiguration and fear of COVID itself (and especially of COVID hotspots like hospitals) are also very plausible contributors. Do you have any methodology that will allow you to distinguish between them?”

    Try this:

  23. The reason the Italian hospitals looked so dreadful is that the media always showed pictures of the hospitals in Bergamo and the worst affected areas. If they had filmed in a hospital in southern Italy it would have just looked like a normal hospital.

Comments are closed.