By: Mallory Bateman
New data, with caveats
The Census Bureau today released the 2020 1-Year American Community Survey (ACS) Experimental Estimates in 54 summary tables for the nation and states and the accompanying Public Use Microdata Sample (PUMS). Due to data collection issues and nonresponse bias created during COVID-19, the quality of the data available did not meet the Census Bureau’s standards for a full release. Respondents in 2020 typically had higher educational attainment, higher incomes, and were more likely to own their own homes than the total population. The Census Bureau used administrative data to help inform weighting in order to address nonresponse bias at the state level.
The summary tables are not available at data.census.gov or through the Census API but are available to download here. Due to the unique process used to create this product, data users should avoid comparisons with previous ACS estimates or decennial data and this “experimental” data set. Additionally, data users working with the PUMS should use caution as the Census Bureau optimized sampling for states and the nation, not Public Use Microdata Areas (PUMAs). Additional guidance from the Census Bureau on working with this data is