The U.S. Census Bureau today released experimental 1-year estimates from the American Community Survey (ACS), one of the nation’s most comprehensive sources of population and housing information about the United States. The experimental data cover a limited number of topics for the nation, all 50 states and the District of Columbia. A technical report, blog and 1-Year Public Use Microdata Sample (PUMS) experimental data were also released.
The Census Bureau typically releases three different sets of ACS data estimates each year in the form of 1-year and 5-year period data sets, as well as 1-year supplemental estimates. Our release is different this year due to the impacts of the COVID-19 pandemic, which disrupted ACS data collection in 2020. Given the limitations, we were unable to collect information from certain segments of the population, as explained in the previously released blog and analytical report, An Assessment of the COVID-19 Pandemic’s Impact on the 2020 ACS 1-Year Data.
The people who did respond to the survey had significantly different social, economic and housing characteristics from those who didn’t, resulting in nonresponse bias in the data. The Census Bureau determined the standard 2020 ACS 1-year estimates did not meet statistical quality standards and decided to ultimately release estimates as an experimental product.
The working paper, Addressing Nonresponse Bias in the American Community Survey During the Pandemic Using Administrative Data, details modifications to the ACS weighting procedures for the 2020 experimental data. Weighting can mitigate the effects of nonresponse bias, which occurs when the characteristics of people who complete the survey (respondents) differ from those of people who do not complete the survey (nonrespondents). The blog provides an overview of the report and experimental methodology.
The 2020 1-year PUMS files include experimental weights and data on approximately 1% of the U.S. population while protecting the confidentiality of survey respondents. The files provide population and housing characteristics down to Public Use Microdata Areas (PUMAs), which are special nonoverlapping areas that partition each state into contiguous geographic units. They are drawn after each decennial census so that they contain no fewer than 100,000 people.
The Census Bureau does not recommend comparing the 2020 ACS 1-year experimental estimates with our standard ACS estimates or the decennial census, or comparing the 2020 1-year PUMS data with standard pre-tabulated products or PUMS-based estimates from previous years. In addition, because the experimental weighting procedure was designed primarily to produce experimental estimates for states, estimates for PUMAs should be used with caution as the experimental weights are not optimized to produce estimates for these areas.
Recognizing the difficulty that the lack of standard 1-year ACS estimates will have on data user communities, the Census Bureau previously provided a list of resourcesoffering technical data assistance. Additional resources are available in the press kit. More information on changes to the 2020 ACS 1-year release is available at www.census.gov/programs-surveys/acs/data/experimental-data.html.
The Census Bureau announced in early November that the 2016–2020 ACS 5-year data release originally targeted for December 2021 had been delayed. Additional time is needed to continue refining our methodology to minimize the impact of nonresponse bias on the 5-year estimates.