The U.S. Census Bureau today announced that it will not release its standard 1-year estimates from the 2020 American Community Survey (ACS) because of the impacts of the COVID-19 pandemic on data collection. The Census Bureau will release experimental estimates developed from 2020 ACS 1-year data.
The standard 2020 ACS 1-year estimates do not meet the Census Bureau’s Statistical Data Quality Standards designed to ensure the utility, objectivity and integrity of the statistical information. Unlike the ACS, the 2020 Census was able to postpone their Nonresponse Followup to a time when they could carry out the full operation, limiting the impact of the pandemic on data quality in ways the ACS could not.
The COVID-19 pandemic posed numerous challenges to collecting ACS data in 2020, as described in our recent Adapting the American Community Survey Amid COVID-19 blog. As a result, the ACS collected only two-thirds of the responses it usually collects in a survey year and the people who did respond to the survey had significantly different social, economic and housing characteristics from those who did not. This is called “nonresponse bias.”
Specifically, Census Bureau staff found high nonresponse from people with lower income, lower educational attainment, and who were less likely to own their home. Nonresponse bias is a natural part of sample surveys, and often statisticians can adjust for nonresponse bias by giving more weight to responses from underrepresented groups. However, Census Bureau staff found that standard nonresponse adjustments to the ACS 1-year estimates could not fully address the differences in a way that meets Census Bureau quality standards.
“The Census Bureau is committed to providing high quality data. The 1-year estimates for the 2020 ACS don’t meet our standards, so we can’t release them,” Census Bureau Acting Director Ron Jarmin said. “The ACS is one of the most comprehensive sources of information about the U.S. population, and it is relied upon by the government, business community and individual Americans. Because of that, it is essential that ACS data truly represent our communities, and the standard 1-year products from the 2020 ACS don’t do that.”
Data Collection in 2020
The ACS surveys a sample of 290,000 households each month on a variety of demographic, social, economic and housing characteristics. Each year, the Census Bureau typically combines the monthly responses into a set of 1-year estimates for the nation, states and communities with populations of 65,000 or more. The Census Bureau also produces 5-year estimates for geographies down to the “tract” or neighborhood level.
Response rates for March through September 2020 — half the months that would make up the 2020 ACS 1-year estimates — were severely impacted by the pandemic. To protect the safety of employees and the people who respond to the survey, the Census Bureau was forced to suspend many ACS data collection operations including mailing survey materials, following up in person with households that didn’t respond, and collecting data from nursing homes, college dorms, prisons and other group quarters.
These challenges in collecting responses significantly impaired the quality of the resulting estimates, which were often inconsistent with benchmarks or administrative data or changed in unexpected magnitudes. The issues Census Bureau staff observed during data review were widespread across social, economic and housing characteristics, signaling a serious quality issue in the data.
The Census Bureau plans to release a report detailing the issues once it finishes its analysis.
Experimental 1-Year Estimates
The Census Bureau will release a series of “experimental” estimates from the 1-year data in November. Because of the underlying quality concerns, the Census Bureau urges caution in using the experimental estimates as a replacement for standard 2020 ACS 1-year estimates. Users should evaluate the estimates and alternatives to determine if they are suited for their needs. To create these estimates, the Census Bureau will apply an alternative set of weights to the 2020 ACS data to attempt to adjust for some of the nonresponse bias. Data from other surveys and administrative records will inform the weights.
The Census Bureau plans to release the experimental estimates in the form of a limited number of data tables for limited geographies. It will also release at the same time a research paper detailing the methodology for the experimental weights.
In December, the Census Bureau plans to release a 2020 ACS 1-year Public Use Microdata Sample (PUMS) file containing experimental weights.
We do not plan to release 2020 ACS 1-year experimental products on data.census.gov.
Technical Data Assistance
Recognizing the difficulty that the lack of standard 1-year ACS estimates will have on data user communities, the Census Bureau is providing our Resource Hub and data flow chart for more information.
Customer Help Center
The Customer Help Center can answer a broad spectrum of questions from Census Bureau customers. They are available by phone at 1-800-923-8282 or email at <ask.census.gov>.
Public Information Office (PIO)
https://www.census.gov/newsroom/about.html
PIO assists journalists in using Census Bureau data. PIO is available by phone at 301-763-3030 or email at <pio@census.gov>.
Data Dissemination Specialists (DDS)
https://www.census.gov/data/academy/request.html
DDS assist in bringing the world of data to you by providing trainings, workshops and presentations to you and your organization at no cost. DDS is available by phone at 1-844-275-3282 or email at <census.askdata@census.gov>.
Office of Congressional & Intergovernmental Affairs (OCIA)
https://www.census.gov/about/cong-gov-affairs.html
OCIA advises Congress and tribal and intergovernmental associations on Census Bureau programs and policies. They are responsible for responding to Census data inquiries from Congress. OCIA is available by phone at 301-763-6100 or email at <cao@census.gov>.
State Data Centers (SDC)
https://www.census.gov/about/partners/sdc.html
The SDC program is a partnership between the 50 states, the District of Columbia, Puerto Rico, and island areas to make data available locally to the public through a network of state agencies, universities, libraries, and regional and local government.
Census Information Centers (CIC)
https://www.census.gov/about/partners/cic.html
The CIC program provides local access, education and technical assistance on census data for planning and decision-making to underserved communities.
2016-2020 ACS 5-Year Estimates
The Census Bureau is still reviewing the quality of the 2016-2020 ACS 5-year estimates against our Statistical Quality Standards and tentatively plans to release them in December. More details will be announced this fall.