Who knows what is the death rate when people are not any getting medical help at all? But in any case. Let's say it is less than 3%. What number of people dying do you consider to be acceptable? 1.2 mln deaths is too much, but 600k is ok? What is your authoritative source? Stalin's guide to statistics?
I already discussed CFR at length and it is something I have used a lot in my own career. I don't want to get into it again because it is complicated and too time consuming for me (he says before writing 600 words). However, for every single viral outbreak, the CFR calculation method is always poorly estimated. It's just the nature of the beast. Sometimes it is underestimated, but usually it is overestimated. So, when the ebola outbreak gets underway the CFR will peg it at around 90% because that's a pretty shitty disease. However, when an outbreak ends the percentage rate is typically between 25% and 90% and mostly somewhere in the middle.
In other words, the CFR is almost always wrong and increasingly so when a viral outbreak is just getting underway. For COVID-19, it seems many people either don't have symptoms or don't get sick enough to require medical attention (and obviously then don't die). CFR does work to account for those cases, but it can only ever be an estimate. As I previously wrote, CFR is actually a beautiful methodology which is why I have used it in the past. However, it is never accurate this early in an outbreak and usually skews too high.
Now, if the COVID-19 situation remains the same (i.e., it doesn't mutate or spread in some novel way), then it seems entirely likely that the current CFR ranging from 3-5% depending on source is too high and something like 1% or even lower may be the case. I am not pushing an agenda either, looking at the data from the CDC to the WHO to just about every other institution working on this outbreak shows that they also predict the final mortality rate will be below what it currently is. I don't know why this is controversial for people who don't work in science, what difference does it make?
It is worth pointing out for fairness that the CFR could also have been underestimated because there could be people dying of COVID-19 but being listed as having died from something else. Looking at the data, I believe the probability of an overestimation outweighs an underestimation, but admittedly both scenarios are possible.
Note: IFR is arguably a more accurate measurement for assessing the MR of a viral outbreak. However, due to the uniqueness of COVID-19, it is failry useless for this pandemic... at least for now.
The fatality rate is not important, what's important is how many people get infected. If 100,000 people get infected in Argentina and 1% die, that would be 1,000 deaths. If that happened, the people shouting that more people die from flu (I don't like the comparisons with flu personally) will be correct and seasonal flu would be a bigger killer. However, the way COVID-19 spreads means it will not stop at 100,000 people if left to its own devices. So, it could reach Spanish Flu levels of infection (1/4 of the population). If that happens, whether its 1% or 0.01% mortality, we are looking at a lot of death.
Two more things. I worked with data and statistics in a scientific field for several years. In terms of COVID-19 I am just a reader, but when it comes to CFR I am my own authoritative source. That said, I am also trying to stick to the science of the outbreak and not the news. For more information on why CFR is useful, as I said a beautifal data tool in fact, it is not the be all and end all of the mortality rate (and as such, there is no mortality rate for COVID-19 at the moment), you can try:
Our interactive data visualizations that show the case fatality rate in each country are updated daily.
ourworldindata.org
This is an excellent website for data addicts anyway, and I also reccomend their wider COVID-19 data breakdown.
Country-by-country data and research on the pandemic. Updated daily.
ourworldindata.org
When immediate risks related to the lockdown exceed risks of the coronavirus. When public opinion changes drastically and people decide they can not take it anymore.
I've seen a model where the lockdown is lifted when hospitals occupancy drops below some level and is reinstated when it goes above some level. Sounds reasonable to me.
Then you mostly agree with me and Perry. For what it's worth, I have never said the lockdown itself is a bad idea. I have said it should never have got to the stage where it was needed, a worldwide mistake that started in Janaury. Since the start of March, I have argued that the lockdown is the only measure possible once the situation got to the level it did. However, it cannot go on. Like you say, I think the lockdown can only last until the "immediate risks related to the lockdown exceed risks of the coronavirus". I suspect the only difference is that we have different ideas of when that point may be reached.
