STANDARD DEVIATIONS: 2020 Census State Level Accuracy

By: Dr. William P. O’Hare, President, O’Hare Data and Demographic Services LLC

Standard Deviations blog posts represent the views of the author/organization, but not necessarily those of the Census Project.

This new short paper focuses on states ranked by net coverage and omission rates in the 2020 Census and offers ideas for further analysis which would take advantage of state variation on census accuracy measures.

In the release from the Census Bureau on May 19, the Bureau noted that six states had statistically significant net undercounts and eight states had statistically significant net overcounts in the 2020 Census (U.S. Census Bureau 2022b). This is a sharp contrast to the 2010 Census when no state had a statistically significant net overcount or net undercount (U.S. Census Bureau, 2012, Table 5). The fact that accuracy differences among states are more pronounced in 2020 compared to 2010 is consistent with the fact that net coverage errors for Non-Hispanic White Alone, Blacks, Hispanics, Asians, and American Indians/Alaskan Natives living on Indian Reservations were larger in 2020 than in 2010 (U. S. Census Bureau 2022a Table 4). Net undercount rates for the Blacks+, Hispanics, and American Indians/Alaskan Natives living in Indian reservations increased from 2010 to 2020 and the net overcount rates of Non-Hispanic White Alone and Asians increased between 2010 and 2020. Across groups and across states, differences in Census accuracy were more pronounced in 2020 than in 2010.

Net undercounts and overcounts are important measures of census accuracy, but they are not the only measures of census accuracy that are informative. Omissions rates for states were also released on May 19. For more information on the distinction between net coverage rates and omissions rates see O’Hare (2019a and 2020b). For many stakeholders, it is the people missed in the census that is the key element of quality.  Most of the work by stakeholder groups in promoting the 2020 Census were efforts to make sure people were not missed in the Census.

Moreover, net coverage is a balance between people missed and people counted twice or counted erroneously. If there were a lot of people missed and also lot of people counted twice, the net coverage error would appear low.

Omissions are defined by the Census Bureau (2022b, page 3) as, “people who should have been correctly counted in the census but were not.” However, this definition is not as straightforward as it may seem. U.S. Census Bureau (2022c) describes two groups of people who are considered omissions.

  • People who were not included in the census count.
  • People who were included in the census count but do not meet the definition of a correct enumeration (e.g., people in “population count only” households are omissions even though they were included in the census count).”

In other words, some of the people in the omissions category were included in the Census count.  


Dr. O’Hare, a member of The Census Project Advisory Committee, is a social demographer who has spent forty years using data to increase public understanding of disadvantaged groups. For the past 25 years, he has been involved in the KIDS COUNT project at the Annie E. Casey Foundation. Bill has a Bachelor of Science, Master of Arts, and a Ph.D. from Michigan State University.