The U.S. Census Bureau revealed Thursday that 2020’s national headcount was a bust. Significant overcounts or undercounts in 14 states will not only have profound implications for how Americans are represented in government but will also affect the amount of money their states get from Washington.
It’s a scandal, albeit one for which no individual or institution deserves any blame.
Owing to the unforeseen pressures of the pandemic, hurricanes in the South and unreliable response rates, the 2020 census was marred by inaccuracies, the bureau’s regular Post-Enumeration Survey revealed. The states that will suffer the most as a result of these errors are, by and large, majority Republican. It’s a scandal, albeit one for which no individual or institution deserves any blame. Republicans are certain to make the most of it, but Democrats cannot ignore the errors either. Their own rhetoric around reapportionment won’t let them.
Overcounts helped Hawaii, Delaware, Rhode Island, Minnesota and Massachusetts boost their standing in the reapportionment process. The overestimations range from a little more than 100,000 residents to nearly 1 million people who don’t exist but will enjoy representation in Congress. Minnesota, for example, barely managed to hang on to its eight seats in the House. But the survey’s conclusion that Minnesota’s population was overestimated by up to 310,000 residents means the state almost certainly should have lost a seat.
By contrast, most of the states that were undercounted are led by Republicans: Texas, Florida, Mississippi, Tennessee and Arkansas were undercounted by two to five points on average. The Natural State is, perhaps, the biggest loser of the 2020 census. The survey estimates that 5 percent of the population was missed, or about 160,000 residents. For comparison, consider that the state’s biggest city, Little Rock, is home to 198,000 people.
The errors don’t all cut in one partisan direction. Ohio, a Republican-led state that is shedding population and lost a seat in the House because of the 2020 census, was overcounted. Illinois, a Democratic stronghold, lost a seat in Congress (some analysts expected that it would lose two) when, according to the Post-Enumeration Survey, the state actually added roughly a quarter million residents in the past decade. New York lost a seat in the House because of population decline, coming up short in the official tally by only 89 people. That was, at the time, controversial. But we know that the bureau assumed anywhere from 382,000 to 1 million more New Yorkers than there are in reality. But the exceptions prove the rule: The populations of GOP-led states were more often the victims of the census’s errors than those in states dominated by Democrats.
Most gratingly, and as is too often the case, the census likely undercounted Black people and Hispanic people while overestimating white people and Asians. The bureau is unable to put a finer point on the matter than that because the survey is not broad enough to measure subgroups with the requisite specificity. And yet, particularly in the South, the underrepresentation of Black people and Hispanic people is measurable.
As is too often the case, the census likely undercounted Black people and Hispanic people while overestimating white people and Asians.
“The post-enumeration results will be what they are,” Census Bureau Director Robert Santos told NPR in February. “No census is perfect.” True enough, and the bureau does better in some decades than others. But Santos assured his interlocutor that the “quality” data his agency accumulated was “fit for the purposes of reapportionment and redistricting.” Not so much, it turns out. Had the count been more accurate, “there definitely would have been changes in reapportionment,” Queens College professor and demographer Andrew Beveridge told The New York Times.
There will be no subsequent changes to the congressional maps of the 2020s as a result of this survey. In 1999, the Supreme Court barred the use of statistical sampling to produce census data for the purpose of reapportionment. The Post-Enumeration Survey can, however, be used where federal aid is apportioned to the states based on their populations: funding for highways, for example. Congress is sure to take a keen interest in providing reparative relief to the states wronged by the census undercount, particularly if Republicans retake one or both chambers of Congress.
The census’s errors, however understandable due to the exigencies of the demonic year 2020, are outrageous. Democrats are obligated to treat this development as a scandal if we are to take their heated rhetoric over a court-drawn congressional map in the state of New York at face value.








