Part 2: What the heck happened starting in August?

Looking at predicted death reduction vs. actual

After finishing my last post, it occurred to me that if I can get % vaccinated by age, by month, then I can look at monthly expected reduction in deaths vs. actual reduction in deaths from 2021 vs. 2020.

Ok, so same sources as in previous post. Now, I can calculate expected reduction in deaths by month (using the fully vaccinated rate at the beginning of each month):

Note that the way to arrive at the overall expected reduction is to take a weighted average of the age level expected reductions (weighted by 2020 deaths in that age group). So, as expected the reduction in deaths would increase each month as more folks are vaccinated, from an expected 42% in April, to 72% in October.

So, we can now easily compare the actual reduction in deaths vs. those expected by this calculation method:

In April the actual reduction is bigger than expected. I can think of at least 2 reasons this might be the case: 1) VE is actually >90% at that point; 2) I don’t include any benefit of a 1st shot, perhaps that is also helping; 3) April 2020 numbers might be inflated as we didn’t know much about care at that time (overuse of ventilators, etc.).

For May-July, I’m actually pretty shocked at how close the actuals are to what we might predict based on my method. But, then a huge surprise in Aug & I truly don’t know what the H-E-double hockey sticks happened in September.