Do lockdowns work?

Preamble: I wrote this post for another site that was considering myth-busting type posts, which is why it’s written in the third person. However, the myth-busting part of the site never really took off, so I thought I would post it here. There’s no specific reason to do so now. It just seemed unfortunate to waste a post and it allows me to not think about writing another post for a while.

There have been some who have argued that lockdowns don’t work, or that they’re not effective.  Here we want to clarify the current understanding of the impact of lockdowns.  One issue is that the term lockdown hasn’t always been well-defined.  We define it as a set of restrictions that substantially reduces contact between individuals who are not in the same household.

It is very well understood that viruses mostly spread through people coming into close contact with others.  The spread could be through direct contact with another person, touching something that has been touched by someone else, or through aerosol transmission.  Consequently, restrictions that limit contact between people will also limit the spread of the virus.  The stricter the restrictions, the more they will do so.

Consequently, if we implement restrictions that we might describe as a lockdown (stay at home apart from essential travel and some exercise) then we would expect this to substantially reduce the spread of the virus.  Not only is this consistent with our basic understanding of virus transmission, there are also plenty of studies that indicate that this is indeed the case.

An issue, though, is that a very effective lockdown that brings down the number of cases and limits the spread of the virus, will leave a large fraction of the population still susceptible.  Hence, if the restrictions are then lifted while the virus is still circulating in the community, the infection can start spreading again, cases can start rising again, and limiting this would then require implementing new restrictions. 

So, the issue isn’t so much the lockdown itself, but what we do to avoid cases rising again when we exit the lockdown.  Consequently, public health experts would argue that lockdown-like restrictions should be used to bring cases down to the point where we can then implement alternative interventions, such as effective border controls, and test, trace and isolate, to control the spread of the virus in the community once the lockdown restrictions have been relaxed.

Of course, it may be possible to control the spread of the virus without initially implementing lockdown-like restrictions. However, for a highly transmissible virus this would require acting while case numbers are still low so that these alternatives can be effective.  Any delay means that more stringent restrictions would then need to be implemented. 

Also, during the early stages of an epidemic, cases increase exponentially; in the UK cases were doubling every 3-4 days in early March 2020.  Consequently, if less stringent interventions are implemented and they aren’t effective, cases will continue to rise, as will the number of deaths and the incidence of long-term illnesses. 

So, there is a very short period of time over which we can attempt to control the spread of the virus using interventions that are less stringent than a lockdown.  If these don’t work effectively, then we will have missed an opportunity to limit the number of cases, and deaths, and may have to implement lockdown-like restrictions anyway.

This can also lead to the somewhat counter-intuitive result that the outcome in countries that implemented very strict, lockdown-like restrictions can be worse than in countries that implemented less stringent restrictions.  However, this is because stricter interventions become necessary if the initial interventions had little effect, or if there is a delay to the implementation of restrictions. 

Hence, it’s not lockdowns themselves that cause the outcome to be worse, it’s that regions that have failed to act appropriately at an early stage of the epidemic need to then implement much stricter interventions than those regions that took effective action at an early stage of the epidemic.  Taking decisive action early can limit the spread of the virus without having to implement lockdown-like restrictions, or can do so with shorter, more focussed, lockdowns.

The motivation here is not to debate whether or not lockdowns are good, or bad, but to simply highlight that lockdown-like restrictions clearly have an impact on the spread of the virus.  We also want to stress that the implementation of lockdown-like restrictions, and how invasive they are, depends on how promptly action has been taken to limit the spread of the virus.  More stringent restrictions will tend to be associated with worse outcomes because of delays in implementing effective interventions, not because these stringent restrictions led to these worse outcomes. 

This entry was posted in advocacy, Philosophy for Bloggers, Research and tagged , , , , . Bookmark the permalink.

22 Responses to Do lockdowns work?

  1. wmconnolley says:

    > there are also plenty of studies that indicate that this is indeed the case
    Which this margin is too small to contain?

  2. wmconnolley says:

    Nurture: you’re looking at text like “Less disruptive and costly NPIs can be as effective as more intrusive, drastic, ones (for example, a national lockdown)”?

  3. WMC,
    1. I did make that point in the post (Of course, it may be possible to control the spread of the virus without initially implementing lockdown-like restrictions.).

    2. The bit you seemed to want evidence for was the claim that many studies show that lockdowns do limit the spread of the virus, which I think those papers do show (even if they also point out that you maybe didn’t need to implement a lockdown).

    3. I’m not arguing that we needed a lockdown (although I am suggesting that you might if you don’t implement suitable interventions early enough) I’m simply pointing out that lockdown-like restrictions clearly have an impact on the spread of the virus (i.e., they work, if that is the goal).

  4. Oddly, I see lockdowns as being most obviously useful at the two extremes: when cases are soaring and a lockdown is needed to avoid overwhelming hospitals, or when cases are so low that a lockdown could be the difference in eliminating local spread and loses control of the pandemic (e.g., it makes sense for places like Australia or New Zealand to do a serious lockdown when a case slips past the border controls).

    The middle ground is frustrating to me, in that it seems to me like the goal during a pandemic should be to squeeze the numbers down until test & trace can be effective, and once you have effective test & trace you not only can keep squeezing numbers down, but you get a much better understanding of how transmission is happening so you can really target the riskiest behaviors. The US totally failed on both the getting numbers down and the test & trace capabilities. But in theory, when a society is willing to take the drastic kinds of behavior changes that we’ve seen in places like the US & the UK, it seems to me that we should have been able to eliminate COVID in rich countries. But that there just wasn’t the compliance that one would need.

  5. climatemusings,
    Yes, I agree with that. You could impose very stringent restrictions very early on to get things under control and so that there is time to develop other interventions (test, trace, isolate, border controls, etc) that will prevent spread when the lockdown-like restrictions are lifted, or you impose it when things are getting out of control and you need to avoid overwhelming the healthcare system. It’s a bit of a pity than many countries seem to have gone for the latter, rather than the former, especially as those countries who did the former seem to have had better outcomes.

  6. Dave_Geologist says:

    Setting asides studies by people with a Kooninesque track record, not just in publication but in being, per a few posts ago, advocates and joining pressure groups or saying things in the libertarian press or websites that go beyond the papers, one distinction I have noticed is the use of lockdown to describe generic social distancing, vs. stay-at-home-orders in addition to everything else.

    I think there is a valid question as to whether stay-at-home laws, sometimes with certificates to go out, add much more when you’ve already closed schools, non-essential stores and hospitality, implemented mask and distancing rules, installed sanitisers in essential stores, asked those who can to work from home, and banned communal exercising and restricted visits to friends or relatives. Of course once you’ve banned all those other things anyway, there’s not much reason to go out beyond essential shopping and exercise. Does taking your dog for a walk in the Peak District 10 miles from home add much more risk to walking it within five miles of home?

    Of course the usual suspects will take a result like that and conflate it with a claim that none of those measures work and we wrecked the economy for nothing.

    There is a question for social scientists though: how much does the symbolism of stay-at-home orders and the word lockdown get the message home in a way that enhances compliance with all those other measures? “Right guys and gals, this is serious.”

  7. Dave,
    I suspect one issue is that we haven’t consulted behavioural scientists enough. I don’t think that we’ve properly assessed how well people will comply with the restrictions. If we were confident that compliance would be high, we may well have been able to impose less stringent restrictions. Of course, there’s then a risk that compliance is low and the restrictions aren’t then effective.

  8. Joshua says:

    Dave –

    > I think there is a valid question as to whether stay-at-home laws, sometimes with certificates to go out, add much more when you’ve already closed schools, non-essential stores and hospitality, implemented mask and distancing rules, installed sanitisers in essential stores, asked those who can to work from home, and banned communal exercising and restricted visits to friends or relatives.

    There’s another aspect of the SIP orders (at least with respect to the US in particular), which is that there’s likely to be, along with such orders, greater government supports such as extended unemployment benefits and other stimulus. Also, with SIP orders, you’re giving people a way to stay at home from work w/o being fired from an employer for doing so (along with the ability to collect unemployment insurance).

    There’s no particular reason that those ancillary beneficial policies would *have to* only be implemented if SIPs are implemented, of course. But the reality is that it would tend to be harder to implement those other policies if SIPs weren’t issued.

  9. Joshua says:

    Also, related to Anders’ point…there could be some advantages to the uniformity that comes among with SIPs – for example where the same conditions would exist across state lines – that might help to reduce incentives to travel. If course, as always, there’s a balancing act as imposing government mandates has the likelihood of increasing resistence counterproductively, as again we’ve seen so prominently in the US.

  10. Joshua says:

    Anders –

    I’m looking forward to reading your review of Koonin’s new book. I just watched a portion of an interview with him (no end to the benefits of reading Judith’s blog) where he said that the mainstream climate economists’ concensus is that warming of 5 degrees Celsius would be economically beneficial.

  11. Joshua says:

    Interesting article on Vietnam, related to the efficacy of interventions and the importance of differentiating among different interventions rather than just using a facile rhetorical frame of “lockdowns.”

    Interesting take on being open to changing one’s mind in the face of emerging evidence.

  12. Joshua,
    I probably won’t bother reading a book written by something who thinks not only that 5C of global warming would be economically beneficial, or that this is the mainstream economic consensus (I don’t even think R. Tol would claim this – although he might claim that we could easily adapt to this).

  13. Joshua says:

    I need to correct what I wrote (I actually went back and listened more carefully): He said that the consensus view (in the IPCC report) is that a 5 degree centigrade increase is anticipated to have a “minimal” economic impact on the US and global economy, delaying the growth by a couple of years by the end of the century.

    Also, in all fairness, he acknowledges “deep uncertainty” w/r/t evaluating economic impact.

  14. David,
    That’s depressing but not unexpected.

  15. Bob Loblaw says:

    One paragraph in the story says:

    The work involved combing death records for each country for the time covering the pandemic and then comparing those numbers with the average number of deaths over the past several years. Numbers in excess of the average were then assumed attributable to COVID-19 or other factors associated with the pandemic, such as an increased likelihood of dying from cancer due to fear of seeking treatment during a pandemic.

    This is clearly the “excess deaths” analytical approach that may were talking about in the early days of Covid.

    I’m sure that the denialati will try to claim that most of the deaths are due to an over-reaction – caused by things other than Covid that only happened because we diverted attention to Covid. Hopefully the following link will work:

  16. Bob,
    Could also apply to cartoon to climate.

  17. Bob Loblaw says:

    People are still applying it to Y2K….

  18. It seems that noone really disagrees that there are probably more COVID deaths than currents estimates, but the suggestion is that the estimate in the paper linked to above may be too high.

  19. Actually, maybe it’s more that the data is such a mess that it’s hard to make any kind of reliable estimate.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.