PROVINCETOWN — After a slow start that saw all four Outer Cape towns lagging behind in 2020 census self-response rates, the U.S. Census Bureau now claims that every single household in Barnstable County has been accounted for. But independent observers have raised questions about the bureau’s methods and are concerned that results will be less than accurate.
Until Aug. 1, the Census Bureau relied on self-responses by mail, by phone, or online to record data for the Outer Cape and across the nation. Starting on Aug. 1, the Bureau began to send census takers, or enumerators, door to door to record information manually for households that had not yet responded on their own. At this point, although the average self-response rate for the four Outer Cape towns is still only 33.5 percent, enumerators have recorded information about all the remaining households. That’s how they arrived at a 100 percent response rate.
That 100-percent figure, however, can be misleading.
“Overall enumeration rates tell only part of the story,” wrote D’Vera Cohn, a senior writer and editor at the Pew Research Institute specializing in demographics, in an email to the Independent. “The overall rate is based on enumerating households using a variety of techniques, some more accurate than others. Some census-watchers are concerned that the Census Bureau will rely too much on the less accurate methods.”
Census employees on the Outer Cape faced a number of challenges specific to the area.
“Our seasonal areas typically take the longest amount of time to complete, because it’s hard to find someone who can verify information,” New York Regional Census Bureau Director Jeff Behler told the Independent. To shorten the timeframe here, he said, “we brought in extra people so we could blitz areas all at once.”
An anonymous enumerator, one of several hundred hired in August to go door to door, told the Independent that getting data in Wellfleet was challenging, both because there are so many seasonal residences and because internet connections were spotty, which prevented enumerators from recording data via their census devices onsite at many homes.
“It often took multiple attempts to obtain a proxy,” said this enumerator. “You would go and there wouldn’t be anybody there, or the people that were there wouldn’t have any information [about the owners] because they were renters.”
According to Cohn, beyond self-responses, other means of recording data “include census-worker interviews with proxies; using ‘high-quality’ government records from agencies such as the IRS or Social Security Administration; or classifying a housing unit as vacant based on census-taker observation. If the bureau cannot enumerate a household, it will use a statistical technique called imputation to count it.
“The Census Bureau’s own research shows that the most accurate responses are self-responses — what people tell the bureau about themselves on their own,” continued Cohn.
In both Barnstable County and Massachusetts as a whole, self-response rates are slightly higher than they were for the 2010 Census. Behler said that, at the state level, self-response rates have risen by just 0.3 percent since the last census, but in Barnstable County, they rose by 5.2 percent. The number of households whose data was collected by imputation “is not something that’s tracked,” he added.
In response to pressure from the White House, the Census Bureau announced on Aug. 3 that it would cease counting operations on Sept. 30, a month earlier than planned. Although many feared this would result in an inaccurate count, Behler said that the Census Bureau was able to reach a complete count by that point.
A Sept. 24 district court ruling allowed the Census Bureau to continue counting. But on Tuesday the Supreme Court overturned the district court and permitted the Trump administration to end the count early after all.
Behler said that did not make a difference in our region’s count. On the Outer Cape, he said, there are zero cases left. “Certainly, in those four towns, we couldn’t do any better,” he said.
That might very well be true: Cohn pointed out that a Census Bureau official testified in court recently that the national rate of proxy interviews in 2020 so far is about what it was in 2010. But it will take a careful analysis of the Census Bureau’s final report when it is released in a few months to determine whether we could, in fact, have done any better.
Residents who think they may not have been counted should check their status at 2020census.gov.