Comment by renewiltord

9 hours ago

ACS, the Decennial Census, and the Austin Demographer has the data we need. Let's take a look https://data.census.gov/table/ACSDT1Y2023.B01003?q=B01003:+T... and https://data.census.gov/table/DECENNIALPL2020.P1?q=P1:+TOTAL... (no ACS in 2020 because of COVID, but we got the Census) and https://demographics-austin.hub.arcgis.com/

You'll see that the ACS 1-year was not performed and released in 2020 due to data quality

    2015:  931840

    2016:  947897

    2017:  950714

    2018:  964243

    2019:  979263

    2020:  961855 (Decennial Census), 965,872 (ACS 5-year, not comparable to 1-year), 996369 (Austin City Demographer)

    2021:  964000

    2022:  975335

    2023:  979700

    2025: 1025668

I think there is no single source that shows population decrease post 2020 to a value in 2020 from the same source. But the sources do differ, so likely what has happened is that you have not performed proper data hygiene. You cannot regress across sources without compensating for cross-source differences.

In this case we have the property that 2021 onwards shows increase in ACS and that Austin Demographics shows increase from 2020 onwards but that there is difference between 2020 numbers between sources.

We also see that Austin Demographer has higher number in 2019 than ACS sample https://demographics-austin.hub.arcgis.com/documents/cf1519a...

This means it's probably not good idea to compare 2023 ACS number to 2020 Austin Demographer number. Fortunately, we have Austin Demographer number from 2019, 2020, and 2025. From all this, we can conclude that cratering hypothesis is likely false. We must conclude that following statement of yours is unlikely: