Census: Early analysis shows falsifying data was rare

FILE - This Sunday, April 5, 2020, file photo shows an envelope containing a 2020 census letter mailed to a U.S. resident in Detroit. On Tuesday, Oct. 13, 2020, the U.S. Supreme Court stopped the once-a-decade head count of every U.S. resident from continuing through the end of October. (AP Photo/Paul Sancya, File)
FILE - This Sunday, April 5, 2020, file photo shows an envelope containing a 2020 census letter mailed to a U.S. resident in Detroit. On Tuesday, Oct. 13, 2020, the U.S. Supreme Court stopped the once-a-decade head count of every U.S. resident from continuing through the end of October. (AP Photo/Paul Sancya, File) (Copyright 2020 The Associated Press. All rights reserved)

Responding to criticism that a shortened schedule jeopardized data quality, the U.S. Census Bureau on Thursday said less than a half percent of census takers interviewing households for the 2020 head count may have falsified their work, suggesting such problems were few and far between.

The statistical agency said in a statement that a preliminary look at the data suggests 0.4% of the hundreds of thousands of census takers, also known as enumerators, may have either falsified data or performed their jobs unsuccessfully.

“Therefore, enumerators who may have falsified data or performed poor quality work were very rare," the statement said.

The Census Bureau issued its statement after a report from its watchdog agency Wednesday that expressed concerns over lapses in quality control checks on the data used for deciding how many congressional seats each state gets and how $1.5 trillion in federal funding is distributed each year. The lapses raised concerns about the quality of the census data, according to the report by the Office of Inspector General.

The report said the Census Bureau failed to complete 355,000 reinterviews of households to verify their information was accurate. Reinterviews also were not conducted with more than a third of the census takers who completed a household interview, and 70,000 cases that were red-flagged for reinterviews were given a pass even though a census clerk was unable to determine if the original interview data was correct, the report said.

About a third of the nation's 130 million households required visits from census takers, while residents in the remaining two-thirds of households self-responded either online, by phone or by mail.

Because of the failure to conduct the reinterviews, the Census Bureau can't provide a full picture of the falsification that may have taken place, said Rob Santos, president of the American Statistical Association.

“Just like with COVID testing, you won’t find it if you don’t look for it," Santos said Thursday in an email.