The highly anticipated 2020 census that dictates political representation and federal funding across the country was either overcounted or undercounted in at least 14 states.
That’s according to a report released Thursday from the U.S. Census Bureau that details an overcount as high as 6.8% in Hawaii, and a 5% undercount in Arkansas, with 12 states in between. The other 36 states did not have a statistically significant miscount.
Utah, the report found, was overcounted by 2.6%. Based on the Census Bureau’s 2020 state population count of 3,271,616, that represents over 85,000 people, although the margin of error makes that more of an estimate than fact.
The discrepancy was found during the standard post-enumeration survey, which operates in 10,000 household sample blocks, independent of the census and after it’s completed, to create a population estimate that is then cross-referenced with census data. In total, 161,000 housing units were surveyed.
Demographic characteristics and geographical data from the survey — information about who was overcounted and where — will not be released.
In Utah, most of the errors came from marking people down in the wrong state, duplications or “other reasons.” The bureau says it will use the survey estimates to plan for the 2030 census.
If accurate, it’s likely some of the discrepancies in Thursday’s announcement could have changed the allocation of congressional seats — but the margin of error for several states is also high enough to throw uncertainty into any decisions made based on the survey. Utah’s standard error, for instance, is 1.23%.
Regardless of accuracy, a Supreme Court ruling prevents survey results from divvying up House seats.
Utah is one of 24 states where groups are suing over newly drawn voting districts — some, like the Beehive State, are accused of partisan gerrymandering, while others say the maps were racially discriminatory.
A spokesperson for the lawsuit in Utah, which includes plaintiffs like the League of Women Voters and Mormon Women for Ethical Government, said it was too early to know whether the bureau’s report will have any bearing on the litigation.
The survey will also not change the official 2020 census results, which put the U.S. population at 331,459,281 people.
“It is important to remember that the quality of the 2020 census total population count is robust and consistent with that of recent censuses,” Census Bureau Director Robert L. Santos said in a news release. “However, we know there is still more work to do in planning future censuses to ensure equitable coverage across the United States, and we are working to overcome any and all obstacles to achieve that goal.”
By comparison, no states had a statistically significant miscount in the 2010 census, based off the post-enumeration survey. However, in 2000, the bureau found 21 states had an overcount.
Five of the six states that were undercounted are in the Deep South, while most of the states that were overcounted were to the north or northeast, excluding Utah and Hawaii. The report’s findings are as follows:
- Hawaii: +6.8%.
- Delaware: +5.5%.
- Rhode Island: +5.1%.
- Minnesota: +3.8%.
- New York: +3.4%.
- Utah: +2.6%.
- Massachusetts: +2.2%.
- Ohio: +1.5%.
- Texas: -1.9%.
- Illinois: -2%.
- Florida: -3.5%.
- Mississippi: -4.1%.
- Tennessee: -4.8%.
- Arkansas: -5%.