Here is a screenshot from yesterday's Census Bureau Webinar on new specifications for the 2020 census. The table shows the crazy inconsistencies in the block-level data, comparing the version they adopted April 28 with the new version just announced./1
The demonstration data released in April was terrible, as we and others explained.
We were expecting the new version to be more accurate than the previous one, but for blocks it turned out even worse./2
users.pop.umn.edu/~ruggles/Artic…
The Census Bureau is deliberately introducing the errors because they claim it is necessary to protect privacy. I dispute that claim in this working paper. /3
assets.ipums.org/_files/mpc/wp2…
Under the new specifications, the number of blocks with no people but with occupied housing units almost doubled, to 149,000. There are zero blocks like that in the real data./2
The number of blocks with no people but with occupied housing units almost doubled, to 149,000. There are zero blocks like that in the real data./3
As I mentioned yesterday, the Lord-of-the-Flies blocks with all children and no adults went from 91,000 to 164,000 in the new "production" version of the data. No such blocks exist in the real data./4
This graph from the Webinar shows the mean error in the population census blocks under the old system and the new one. /5
The previous version got the population of Liberty Island wrong by a factor of 24. In the new data, it is likely that the population of small blocks is off by at least an order of magnitude. Data of that quality is not worth producing./end
Share this Scrolly Tale with your friends.
A Scrolly Tale is a new way to read Twitter threads with a more visually immersive experience.
Discover more beautiful Scrolly Tales like this.