Updated September 14, 2021 at 12:22 PM ET
An independent panel of researchers said it has found no major irregularities in the 2020 census results that were used to reallocate congressional seats and Electoral College votes for the next decade.In a report released Tuesday, the American Statistical Association's task force of census experts also noted that despite concerns about interference by former President Donald Trump's administration, their review uncovered "no evidence of anything other than an independent and professional enumeration process" by the Census Bureau's career civil servants who took "corrective actions as necessary."Their assessment comes about five months after the April release of state population totals from a national head count that was upended by last-minute schedule changes because of the coronavirus pandemic and Trump officials."We may have come to a different conclusion had the election come out differently and they had to put out numbers Dec. 31," said Nancy Potok, a former deputy Census Bureau director who co-chaired the task force, during a press briefing before the report's release in reference to Trump officials' failed efforts to move up the delivery of state population counts.Instead, the bureau ultimately pushed that back to run more quality checks, which Potok said gave the task force "a lot of confidence" in the agency's work.
The report did not conclude the census results had no errors
The task force stopped short of declaring the first major set of 2020 census results problem-free."We're statisticians, don't forget, so we believe that the absence of evidence is insufficient to state conclusively that there were no errors," added Potok, who volunteered last year as a member of President Biden's transitionagency review team for the Commerce Department, which oversees the bureau.More evaluations are needed, the task force warns, because its assessment was limited to reviewing how the bureau's career staffers prepared the congressional apportionment count. The task force's researchers also inspected certain internal data files about the agency's operations at the state level, which does not provide a comprehensive picture given that census participation can vary greatly by neighborhood.The bureau, the task force said, received a copy of the report before its release. In a statement, the bureau said it will "carefully review" this analysis to prepare for the 2030 census.Two original members of the task force, which was formed last year, have been named by the Biden administration to join the bureau. Robert Santos, Biden's nominee to be the bureau's next director, and Kenneth Prewitt, a former bureau director who is now a senior adviser at the agency, resigned in April and were not involved with the final report, the task force says.
There's another independent evaluation of census results
The bureau's acting director, Ron Jarmin, has said during news conferences that while "no census is perfect," the figures released so far "meet our high data quality standards."Still, at the bureau's request, the National Academy of Sciences Committee on National Statistics put together a separate independent panel that is also evaluating last year's census.The American Statistical Association's task force says it's no longer planning to issue a report on the quality of the detailed demographic data used for redistricting that the bureau put out in August.Instead, it's urging the National Academy of Sciences' panel to take a close look at the high rate of households not answering the race, Hispanic origin and other demographic questions on census forms, as well as last-minute changes to how government records were used to account for unresponsive households due to the Trump administration cutting short door-knocking efforts. The task force is also recommending that plans for the 2030 census include making quality indicators about different operations for the head count available before the first set of results is released, instead of months later. Copyright 2021 NPR. To see more, visit https://www.npr.org.