President Trump should not be surprised that his anti-immigrant apportionment gambit is on the brink of collapse because of logistical issues with the census. The risks of rushing the census so the data could be produced before he left office were laid out by the bureau itself earlier this year.
The Census Bureau last week identified another round of snags — known within the bureau as “anomalies” — that are likely to delay the release of the census apportionment data until after Trump leaves office.
Without those numbers, Trump will be unable to implement an unprecedented policy of excluding undocumented immigrants from that count, which determines the number of House seats and Electoral College votes each state gets.
A few weeks ago, as TPM initially reported, the Census Bureau was already warning that anomalies that had been identified were threatening plans to release the data in early January. But the administration’s Trump-appointed leaders have continued to spin that sticking to close to that timeline remained possible.
After a Sunday evening meeting among career experts to discuss the scope of the new anomalies, it was not clear how the latest round of anomalies would affect the timeline for delivering the apportionment data, according to a source within the bureau. But the anomalies are not going to be an easy fix, the source said, and it is expected to take some time to investigate and resolve the issues.
Every census, the bureau identifies these kinds of snags. The survey’s processing period after the data collection ends — a period that usually lasts around five months —is designed so that the bureau can work through these issues.
However, because the administration demanded that the bureau trim four months off the coronavirus-related extended timeline it had laid out for dealing with the pandemic’s disruptions to the count, that left only two and half months for processing.
The move to truncate that schedule is widely believed to be motivated by the administration’s desire to guarantee that Trump is still in office to implement the anti-immigrant policy, as the extended timeline would have given the bureau until late March 2021 to finish processing the apportionment data. The bureau was informed of the need to speed up the count just days after Trump announced the policy.
Not surprisingly, several of the anomalies identified in recent days are related to aspects of the count most disrupted by COVID-19, going to show why the bureau had requested the four extra months to overcome these challenges. Ironically, at least some of the anomalies would have been caught much earlier in the process, had the bureau not been forced, with the administration expedition order, to cut a specific review step that helps vet the data in question.
“This day by day urgency, with some level of desperation, to complete data processing on a rushed timeline simply highlights the folly of this administration’s effort to force the bureau to produce numbers for congressional apportionment before the bureau is able to do so,” Terri Ann Lowenthal, a former House staffer who now consults on census issues, told TPM. She spoke to TPM on Friday afternoon, as buzz about the new anomalies began traveling through the community of outside bureau observers.
The Census Bureau’s press shop has not yet weighed in on the development.
Several of the anomalies the bureau has identified are related to what’s known as “group quarters,” i.e. nursing homes, colleges, prisons and other settings where large groups of people live together. Even before the pandemic, some aspects of how the 2020 survey would count group quarters were untested; because of budget constraints, the bureau had to rework the vetting it had planned for these operations in its end-to-end test in 2018.
But the pandemic presented unique challenges to counting group quarters populations. Nursing homes weren’t eager to let census enumerators in to do their data collection. College students were fleeing campuses right as the census was getting started.
The risks these dynamics posed for the survey’s accuracy were compounded when the administration began pressuring the bureau to deliver the data by the end of the year — the statutory deadline — even though earlier in the pandemic, President Trump supported Congress granting the four-month extension.
The bureau has insisted that the steps it would take to speed up its schedule wouldn’t come at a cost to the accuracy. One of the adjustments it did make, however, severely limited the opportunity the bureau usually gives state demographers to do a preliminary review of the group quarters count. The step, which was supposed to take place in September, could have helped avert the hiccups the bureau is running into now, according to a source within the bureau.
Typically, the demographers review the entire list of group quarters that were counted, to help identify for the Census Bureau which settings they might have missed while there was still time to get back in the field to enumerate those populations. For instance, they might find that the bureau has skipped over counting a prison that could amount to a fifth of a county’s population, Robert Rhatigan, a New Mexico state demographer, told TPM.
Because that review was scaled back, he said “there was no independent verification that all group quarter facilities were actually counted.”
Cutting that step was mentioned in a presentation the bureau’s experts gave Secretary of Commerce Wilbur Ross in August, on how it could speed up its schedule to meet the Dec. 31 deadline.
The presentation warned that the truncating the overall review phase that is now under way “creates risk for errors being present in the data.”
It turned out, the bureau knew what it was talking about when it gave Ross that warning.