Tuesday November 8, 2022 is Election Day in the United States of America. In addition to Congressional "midterms" at the federal level, many state governors and other more local offices are up for grabs. Given how things shook out over Election Day 2020, things could get a little crazy.
...or, perhaps, not! But here's the Megathread for if they do. Talk about your local concerns, your national predictions, your suspicions re: election fraud and interference, how you plan to vote, anything election related is welcome here. Culture War thread rules apply, with the addition of Small-Scale Questions and election-related "Bare Links" allowed in this thread only (unfortunately, there will not be a subthread repository due to current technical limitations).
Jump in the discussion.
No email address required.
Notes -
One percent getting cut to a tenth of a percent is an order of magnitude and that was before we found out that rates of infection may have actually been much higher than previously thought making the disease that much less lethal.
If that article is correct, with about 6.6M deaths and world population of 8 billion, then the IFR is still over 0.1%, within 1 order of magnitude of the original estimates.
A naive rate calculation 6.6 million deaths in a population of 7.8 billion yields an IFR of 0.084%, and that's assuming that 100% of those 6.6 million deaths were actually caused by covid and not "died by other causes while infected with covid". Meanwhile at the height of the lockdown hysteria government officials were speculating that the IFR might be as high as 3%, but that was quietly memory-holed when it came out that that particular estimate was based solely on data from the state of New York where some bright spark had decided that the best place to house patients with a respiratory disease would be in public nursing homes.
Accordingly I stand by my initial statement.
That's not an IFR. Your source claimed about 60% of the world may have been infected, so the IFR would be 6.6 million / (7.8*0.6) = 0.14%.
It's also assuming that there weren't deaths caused by covid which were missed. This blog post, which I thought was posted here or on the subreddit at some point, finds that total excess deaths usually substantially exceed official COVID deaths, although there's no way to know if that's because of missed COVID deaths or because of other factors, such as the spike in traffic fatalities in the US. (Unfortunately the post itself skips this point and just calls the difference a "fudge factor.")
You have compared the absolute highest IFR I've seen for Covid (actually I'm not sure I've ever seen 3% claimed; this paper gives estimates of over 5%, but that's for the case fatality rate, and so is much higher where there are more uncaught cases; the lower end 0.15%, almost identical to the 0.14% I gave above, is probably closer to the IFR, but I can't find any similar papers attempting to calculate IFR directly with early data), which also was not the "initial" estimate since NY didn't have a big wave until at least 6 months after it started in China and which you also seem to agree was an anomaly that was retracted, and comparing to an IFR that is substantially lower than even what your own source would support. Even with the absolute highest gap one can possibly construct, misleadingly so in fact, you can only muster log_10(35) = 1.54 orders of magnitude, and even that includes real changes to the IFR over time (improved treatment, most vulnerable people dying first--according to this paper, IFR might have dropped by around 1/3 in the last 9 months of 2020).
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link