So far in 2020 Census processing, 27 of the 33 anomalies we’ve found are of this type. Let me give a couple of examples.
Miscalculating age for missing birthdays. We found that our system was miscalculating ages for people who included their year of birth but left their birthday and month blank. We fixed this with a simple code correction. Making sure ages calculate correctly helps us with other data processing steps for matching and removing duplicate responses.
Incorrectly sorting out self-responses from group quarters residents. The 2020 Census allowed people to respond online or by phone without using the pre-assigned Census ID that links their response to their address. As a result, some people who live in group quarters facilities, such as nursing homes, were able to respond on their own even though they were also counted through the separate Group Quarters Enumeration operation. This also makes their address show up as a duplicate — as both a group quarters facility and a housing unit. Our business rules sort out these duplicate responses and addresses by accepting the response coming from the group quarters operation and removing the response and address appearing as a housing unit. We found an error in how this rule was being carried out. The code was correctly removing the duplicate address but wasn’t removing the duplicate response. We fixed this with another code correction, which enables us to avoid overcounting these residents.
Author(s): MICHAEL THIEME, ASSISTANT DIRECTOR FOR DECENNIAL CENSUS PROGRAMS, SYSTEMS AND CONTRACTS
Publication Date: 9 March 2021
Publication Site: U.S. Census Bureau