Assessing Miscounts in the 2020 Census

June 4, 2019

Urban Institute

T he decennial census, which aims to count every US resident each decade, is critical to our democracy. It affects congressional seats and funding decisions at every level of government. But the 2020 Census faces unprecedented challenges and threats to its accuracy. To understand how these factors could affect census counts, the Urban Institute published two new resources:

1) The interactive data tool shows projections of potential miscounts based on three scenarios reflecting the miscount risk as low, medium, or high. Results can be filtered by state or by demographics. Depending on the risk scenario, Georgia's overall population could be undercounted by 69,200 to 177,000 people.

2) The report highlights findings from the Urban Institute's 2020 Census count assessments nationally as well as for different population subgroups and states under low-, medium-, and high-risk scenarios. Key findings include the following:
  • The overall accuracy of the national population count in 2020 could range from an undercount of 0.27 percent in the low-risk scenario to an undercount of 1.22 percent in the high-risk scenario. While these percentages may seem small, considering the overall US population, between nearly 900,000 and over 4 million people could be missed.
  • If the 2020 performance of the census mirrored that of 2010, the national population could be undercounted by 0.27 percent because of demographic changes over the last decade.
  • If the 2020 Census performs as the US Census Bureau expects, the national population could be undercounted by 0.84 percent.
  • Some states may be more at risk for miscounts. For example, California has projected 2020 undercounts that could range from 0.95 to 1.49 to 1.98 percent by risk scenario (low, medium, and high risk, respectively).
  • The miscounts may disproportionately affect some groups more than others. Black and Hispanic/Latinx-identified people in the high-risk scenario could be undercounted nationally by 3.68 and 3.57 percent, respectively. White, non-Hispanic/Latinx people are at risk of being overcounted nationally by 0.03 percent in the high-risk scenario.
  • Historically undercounted, children under age 5 are again at risk of being undercounted by up to 6.31 percent in the 2020 Census in the high-risk scenario.

Back to Listing