Will census errors in 14 states, where more than 3.5 million people were miscounted, lead to surprising results?
By Joe Barnett
Tomorrow’s election is the first nationwide vote after Congress was reapportioned following the 2020 Census, but errors in the decennial count could lead to some surprising outcomes in 14 states.
It appears that more than 2.1 million individuals in six states weren’t counted (called the undercount), and more than 1.5 million nonexistent residents in eight states were counted (the overcount), according to data in the Census Bureau’s Post Enumeration Survey (PES) released in May.
As a result, some of the states miscounted did not receive additional, or as many additional, representatives as they were entitled to, and other states retained congressional seats they should have lost. The Census errors also affect the distribution of Electoral College votes, and the allocation of billions of dollars in federal funds based on population.
The Census can’t tell us—at least yet—who was miscounted or where they live (or were alleged to live) within each state, so we do not know if the miscount will affect the outcome of any close congressional races in those states. It depends on whether the undercount or overcount is peculiar to some localities, concentrated among some demographic groups, or uniform and statewide.
The Census block-level counts of residents are used to draw district lines for state legislatures and local elective offices. The boundaries of these districts are usually drawn to balance the number of residents and often to ensure representation of racial and ethnic minorities. Inaccuracies in the Census could affect the outcome of close elections in small constituencies, where a handful of votes can be decisive.
Of course, Americans move so frequently the population figures of April 1, 2020 are already out of date. But while voters in 36 states (and the District of Columbia) can have a fair degree of confidence that they were accurately counted, people in 14 states—no so much.
The Census Bureau conducts the PES to determine the accuracy of the headcount.
The overall error rate for the 2010 Census survey of households—a slightly smaller subset of the total population—was 0.01 percent, which was determined to be statistically insignificant, compared to the PES-derived error rate of 0.24 percent for 2020.
The 2020 error rates range from a 6.79 percent overcount in Hawaii to a 5.04 percent undercount in Arkansas. The PES determined the total U.S. population on April 1, 2020 was accurately recorded by the Census as 331,449,281. However, in terms of sheer numbers, the errors in 14 states are substantive.
Applying the household survey error rate to the total population of each state, the PES data indicates 2,167,919 people in six states weren’t counted, whereas in the eight states with overcounts, 1,541,608 phantom people were counted in state-level census reports. The census error rates and the author’s calculation of the number of individuals involved, are:
Undercount: Arkansas (-5.04%, 151,893), Florida (-3.48%, 750,654), Illinois (-1.97%, 252,608), Mississippi (-4.11%, 121,817), Tennessee (-4.78%, 330,628) and Texas (-1.92%, 560,319). (Total undercount: -2,167,919)
Overcount: Delaware (+5.45%, 54,001), Hawaii (+6.79%, 99,143), Massachusetts (+2.24%, 157,550), Minnesota (+3.84%, 219,254), New York (+3.44%, 695,422), Ohio (+1.49%, 175,952), Rhode Island (+5.05%, 55,457) and Utah (+2.59%, 84,829). (Total overcount: +1,541,608)
Surprisingly, more people were not counted in Florida (750,654) than in Texas (560,319), which has nearly one-third greater population. Observers say Florida would have received two additional congressional seats, if the Sunshine State’s residents had been counted accurately.
Miscounts ‘Regional in Nature’
U.S. Rep. Troy Nehls (R-TX) demanded answers regarding errors in the state-level population counts in a letter to Director of the Census Robert Santos on October 21.
“It has come to my attention that the U.S. Census Bureau significantly undercounted the populations of six states, which primarily vote Republican, and overcounted the populations of eight states, which primarily vote Democrat,” Nehls wrote. “As a result, these blue states will now have more representation in Congress, more votes in the Electoral College, and may well receive more federal funds than they should.”
Nehls’ letter to Santos notes Texas was just 189,000 residents short of receiving a third additional member of Congress.
Election Data Services, Inc. (EDS) said the results of the Census were unexpected, in a press release titled “Final Census Apportionment Counts Surprises Many Observers; Raising Questions of Why?” on April 28, 2021, after the Census Bureau released its state-level population counts.
“[T]he results shifted the number of seats that were projected to change in six different states from the 2019 population estimates released by the Bureau just five months ago,” states EDS. “This change appears to be regional in nature, with the southern states of Arizona, Texas and Florida not gaining or not gaining as many seats as expected. On the flip side, the northeastern states of New York and Rhode Island, and the upper Midwest state of Minnesota kept seats that they were expected to lose.”
Nehls says the errors could indicate political bias in the Census.
“[W]hile Texas likely was cheated out of an additional member of Congress, states like Minnesota and Rhode Island kept seats they should have lost if not for overcounting,” wrote Nehls.
Nehls asked Santos for answers to a number of questions about how the census was conducted, including: “Was the U.S. Census Bureau instructed by anyone in the Executive Branch or otherwise to take steps differently than in 2010 that would lead to this inaccuracy?”
“I do not believe this was a coincidence, as the Bureau has failed to provide any explanation for this severe undercounting and overcounting,” wrote Nehls. “This is deeply concerning for the legitimacy of our Democracy.”
Who Are They? Where Are They?
Who was under/overcounted? Where do they live? In its PES report, the Census Bureau states, “Results cannot reliably be broken down by demographic characteristics or geographic areas within states.”
The demographic characteristics of the miscounted are relevant because racial and ethnic categories are used in drawing electoral districts, as required by the Voting Rights Act. If the miscounts occurred in particular areas of the 14 states, redistricting put too many, or too few, residents in the boundaries of some districts.
The Census estimated undercounts/overcounts nationally by demographic group, in a report released in March. Historically, some groups are consistently undercounted or overcounted, states the Bureau:
“The Black or African American alone or in combination population had a statistically significant undercount of 3.30%. This is not statistically different from the 2.06% undercount in 2010. The Hispanic or Latino population had a statistically significant undercount rate of 4.99%. This is statistically different from a 1.54% undercount in 2010. … The non-Hispanic White alone population had a statistically significant overcount rate of 1.64%. This is statistically different from an overcount of 0.83% in 2010. … The Asian alone or in combination population had an overcount rate of 2.62%. This is statistically different from 0.00% in 2010.”
In 2023, the Census Bureau will release a report with more specific information on the demographic characteristics and geographic location of the miscounted. So far, the Bureau has pointed to the difficulties they faced, given the COVID-19 pandemic.
EDS President Kimball W. Brace, according to the EDS press release, “speculated that it’s possible the southern state changes, with their large and growing Hispanic populations, have been caused by the Trump Administration’s efforts to keep non-citizens from being counted in the Census. It is also reported that these three states [Arizona, Florida, and Texas] failed to have an effective state-sponsored outreach program to promote the Census.”
Possibly, given the flow of migrants into these states. However, that would not explain how some states’ residents were overcounted. Rhode Island, for example, kept its second congressional seat “by a margin of only 19,127 people to spare,” EDS calculates, but was expected to “lose its second seat by 14,529 people.” The 55,000 overcount saved Rhode Island’s second congressional seat.
Regardless of the cause(s), the Census errors are likely to be the subject of congressional scrutiny after the new Congress is seated in January 2023, says Nehls.
“The swamp in Washington has an agenda,” said Nehls, in a press release. “They want Democrats in power and won’t let anything get in their way. We must get to the bottom of what happened. When Republicans take back the House majority, we will use our oversight authority to investigate the Census Bureau and determine how and why these significant errors happened to ensure this doesn’t happen again.”
Elections are decided by citizens who choose to register and vote, but if the size and composition of the population pool in a state’s districts are off by 5 percent—or more—there could be electoral surprises in the 14 states where more than 1 percent of the total U.S. population was miscounted.
Joe Barnett is a former editor of Budget & Tax News and a senior editor at The Heartland Institute. For more Budget & Tax News.