OCT. 19, 2020 — As a federal statistical agency, the U.S. Census Bureau welcomes open discussions about our work and the methods used to produce our statistics.
As of the end of data collection operations, well over 99.9% of addresses nationwide have been accounted for in the 2020 Census, with 67.0% accounted for through self-response online, by phone or by mail, and 32.9% accounted for through our Nonresponse Followup (NRFU) operation.
The final self-response rate for the 2010 Census was 66.5%. This is important, as better data comes from self-response.
The majority of occupied housing units in the NRFU operation were completed by a census taker interviewing a member of the household. The balance was completed by interviews with proxy respondents or through the use of high quality administrative records.
As of October 15, approximately 5.6% of addresses nationwide, have been resolved using high-quality administrative records, which is 13.9% of the NRFU workload. See our FAQs for more on the use of high-quality administrative records.
As of October 15, approximately 24.1% of occupied housing units in the NRFU workload have been enumerated by proxy response—which is similar to the 2010 rate. See our FAQs for more on enumeration by proxy.
Each census, we produce coverage estimates and conduct extensive assessments that we share with the public. We plan to do the same again following the 2020 Census—the completion rates are just early indicators.
There are two main ways of evaluating the quality of a census—comparing the results to other population totals and evaluating how well the process of conducting the census worked. As we do after every recent decennial census, the Census Bureau will share information from each method as it becomes available over the coming months and years. In terms of comparing the results to other population totals, the Census Bureau will produce estimates from Demographic Analysis in December 2020.
The Census Bureau is also currently conducting a Post-Enumeration Survey, which will give estimates of the undercount by demographic groups and geography. The first results from this survey are expected in November 2021, with additional results available in February 2022. In terms of evaluating the process, the Census Bureau conducts extensive evaluations of how the census was conducted. These evaluations will include data, such as the final proxy rate, that in many cases can be compared to prior decades.
We will release these evaluations as they become available; many won’t be available until after the census response data has been processed.