TODAY, November 30th, the IPUMS team updated their Online Data Analysis System to include
the single-year 2021 ACS PUMS data.
https://usa.ipums.org/usa/sda/
This is great: we can now look at single-year ACS PUMS for all years up until 2021,
include the 2019 AND 2020 years with Experimental Weights. The IPUMS folks updated the
2019 to include the experimental weights earlier this month.
I strongly recommend using the IPUMS for analyzing PUMS data. The web-based software is
LIGHTNING fast, and is fairly easy to use / learn / apply.
As a backup, I would recommend using the R package TIDYCENSUS to acquire/analyze PUMS
data. Dr Walker has great tutorials on how to use TIDYCENSUS for the ACS PUMS, including
proper use of the “replicate weights” used to analyze PUMS standard errors.
For example, if you needed a table showing the distribution of households by number of
adults (16+) (1, 2, 3, 4+) by the number of vehicles in the household (0, 1, 2, 3, 4+)
then you would need to use TIDYCENSUS, not IPUMS. (If you needed just household size by
household vehicles, then use IPUMS).
If you need something like the "median age of transit commuters in the United States,
1960 to 2021", then you would use IPUMS (without question!)
Another amazing thing about IPUMS: If you have LARGE counties (100,000+) where the PUMAs
nest directly into your LARGE COUNTY, then the IPUMS has county codes for the user to work
with!!
IPUMS folks are rockstars when it comes to making Census data so readily available!!
cheers,
Chuck