There's possible confounds, but none can explain vaccinated people being fewer than 6% of deaths (my MLE: about 2% of deaths) and more than 50% of the population. Especially since the vaccinated are, overall, a sicker and older population than the country at large.
The vaccine RCTs were not powered sufficiently to show a reduction in COVID death, given that they had no or basically no COVID death in either group. They did show a huge reduction in illness and severe illness, and now we're observing a commensurate reduction in death in the population.
population normalized death incidence is down by 10x in vaccinated over unvaccinated (which includes recovered individuals in with the unvaccinated, so the number compared to susceptible is probably 20x). with a bit over half the population vaccinated that suggests that at least the 100,000 deaths since june would have been twice as bad without vaccinations. not including any effects of vaccination on reducing infection/transmission (and for all the hype about "waning" the protection against infection is still substantial).
That was the trial that only had 40k participants. Adverse effects are so rare, it had to be administered to tens of millions of people to the public at large before just a few showed up.
Was that 20 more dead from thrombosis than would normally be seen in a similar population of that size? Or 20 total? Because my understanding is that it was only a few more than usual.