4 ms·
>Australia has 9 doses per 100 compared to Canada's ~30. I'm not sure what you are saying, doses of vaccines? Number of deaths per capita don't take QoL into
by totony 5y ago
>Australia has 9 doses per 100 compared to Canada's ~30.
I'm not sure what you are saying, doses of vaccines?
Number of deaths per capita don't take QoL into account over the period you analyze and the response's effect in the future.
- whimsicalism 5y agoDoses of vaccine, yes. I'm curious, what QoL increase makes up for 400k additional deaths over Canada's number?
- totony 5y ago>Doses of vaccine, yes. I have not claimed doses of vaccine to be a good metric. >I'm curious, what QoL increase makes up for 400k additional deaths over Canada's number? Who has 400k more? You also have to take into account that policy differences do not solely explain the difference in death toll. Saying 400k more death over canada = canada good response is a reductionist view when there are other factors in play.
- whimsicalism 5y agoOkay, so what does explain the difference in death toll? I certainly don't think it is only policy diffferences. > Who has 400k more? If we had Canada's per capita rate, 400k fewer people would have died.
- totony 5y agoI am not familiar with Australia's response enough to give a conjecture on how good its policy is. Although I am familiar with Canada's response and (independently of death toll) think it was very suboptimal. >I certainly don't think it is only policy diffferences. And I agree with you, which is why I think using death toll as a single metric is a bad idea. EDIT: >If we had Canada's per capita rate, 400k fewer people would have died. Even with the same policy as Canada, it is unlikely that the death toll would just transpose (it's a bad assumption, because policy doesn't define the death toll - it affects it)