Igor Chudov makes the claim in his recent post, based on a reported 23% drop in birth rate this May (2022) over last May (2021) in Taiwan, that Taiwan has experienced what constitutes a 26-sigma deviation from expectation.
Whatever is going on, there is a big problem with this claim. To make it, Igor uses the Taiwan government’s report, annualizes its result and compares it to results taken from another source (macrotrends.net).
However, if you look at previous reports release from the Taiwanese government, the result, while pretty big, is nowhere near so extreme seeming.
For example, I went back through the reports available here: https://www.ris.gov.tw/app/en/2121?sn=22161405
Going back to just before the pandemic (Jan 2020) and scraped (by hand :(, ugh) the percent month-over-prior-year’s-month changes. And plotted them here
As you can see, there was actually a larger change (positive!) this past January.
In this case the standard deviation was 13.54, and the average was -3.55. So a -23% decrease was not even two “sigma” away from the normal expectation using a crude calculation when the data are sourced consistently.
May be worth looking into, but certainly not worth panicking over.
Thank you VERY much for commenting on this. I always like second opinions and skeptical reviews.
Births are always seasonal, this is typical for any country. Month-to-month curves always look wild.
We need to look at year-to-year comparisons for individual months.
Take a look at Scotland for example, which is also highly seasonal.
https://scotland.shinyapps.io/phs-covid-wider-impact/