Imho this thread (while it provides useful data) is a good example of a common pitfall in covid discourse: extreme utilitarianism / focus on numbers with no regard for what the experience of the individual is. 1/19 https://twitter.com/jburnmurdoch/status/1354158351907778567

One issue is choice. It is disingenuous to describe a fall in accidents during lockdowns as "saving lives", if the reason for it is that people were prevented from taking mundane risks that everyone in 2019 thought were acceptable, in the pursuit of enjoying their lives. 2/

To take an extreme example, if you close playgrounds, and fewer children die falling from swings, you cannot boast about saving children's lives. Sure the numbers look good, but 99% of the children whose deaths were averted and their parents would rather have run the risk. 3/

The other issue is the trolley problem. The people who die under lockdown are not the same as the people whose deaths are averted by lockdowns. It is entirely fair to blame lockdowns for the former, regardless of the latter. 4/

If A is pushed to suicide through isolation, and B and C get much better once they stop going to their toxic workplace, you get -1 excess suicides during LD. This does not let you conclude that "LD does not cause deaths". It does. It killed A. 5/

In fact I would argue that policymakers deserve blame for the death of A, which is a direct, predictable consequence of their draconian policy (which they at least should have attempted to mitigate), while they deserve zero credit for the survival and B and C, 6/

because this is an unintended second-order effect from disrupting the normal order. It's also unknowable in individual cases since we cannot be sure B and C would have killed themselves, except maybe in some implausible scenarios. 7/

Both these issue also apply to covid deaths, to some degree. The trolley problem applies because the people dying of covid under LD, while they may be less numerous, have no reason to be a proper subset of those dying of covid in the less-mitigated scenario. 8/

Worse than that it might be that certain identifiable groups, for instance essential workers or care home residents, end up with a higher attack rate in the end, especially if the restrictions fail to really suppress covid in the medium term (as has happened in Europe and NA). 9/

To be clear I am not an epidemiologist and I do not know how likely it is that this is actually happening in the real world. It does seem that this sort of non-monotonic effects of restrictions can happen in toy models. The vaccines should mitigate it. 10/ https://twitter.com/WesPegden/status/1295752780792115206

The issue of choice also matters. The risk of unmitigated covid has been estimated in various ways but even the worst scenarios "only" double the mortality rate over a year in a developed country. This is obviously terrible but 11/

it is still far from "bring out yer deads", and for a given age group the extra absolute risk is commensurate with the base death rate, so that this is a level of risk people are familiar with. I think it has been short-sighted to just assume that many elderly people 12/

would not have preferred to run some of this risk in exchange for some normalcy.

In particular it should be obvious to anyone who isn't a geriatric doctor or a technocrat that many CH residents would prefer less mitigation. They value normalcy a lot, and their lives little. 13/

Of course the pro-LD reasoning is that high transmission is an externality that affects everyone. My point is that the reduced risk is nothing to boast about if elderly people are prevented from doing things they would be willing to do even under the higher level of risk. 14/

Btw: I am absolutely not claiming anybody does a sophisticated risk calculus before doing things. I am claiming people value normalcy a lot and will tolerate very high risks in exchange, and their behaviour might reflect it if technocrats had not decided they know better. 15/

To make it very concrete, I claim there is a nonzero share of elderly people who would be fine having lunch at their neighbours' even in the middle of the unmitigated covid peak. The more you mitigate, the more the share increases of course.16/

If you mitigate so much they can't do it any more, what's the point?

As an aside: it has been infuriating how technocrats have been enforcing their values on acceptable risks not by trying to persuade the elderly, but by shaming their children. 17/ https://dr-no.co.uk/2020/12/23/the-granny-ultimatum/

OTOH it seems fair to say you're saving lives by preventing hospitals getting overwhelmed, because people are going to need hospital care no matter what. I have no qualms with this. 18/

To conclude, you won't make the ethical case for LDs by looking at excess deaths. I suspect the degree to which this "thinking in numbers" has infested public discourse is probably due to the increasingly technocratic nature of our democracies. 19/19 https://twitter.com/BallouxFrancois/status/1311662363603423234