Keep Factually independent
Whether you agree or disagree with our analysis, these conversations matter for democracy. We don't take money from political groups - even a $5 donation helps us keep it that way.
Fact check: Which states showed net overcounts or undercounts in the 2020 census according to official reports (2020 or 2021)?
Executive Summary
The official 2020 Post-Enumeration Survey (PES) and subsequent analyses identify 14 states with statistically significant net coverage errors: six with net undercounts (Arkansas, Florida, Illinois, Mississippi, Tennessee, Texas) and eight with net overcounts (Delaware, Hawaii, Massachusetts, Minnesota, New York, Ohio, Rhode Island, Utah). These findings come from the Census Bureau’s 2022 PES release and are reiterated in later GAO and Census summaries that place the PES results in broader demographic context, showing persistent patterns of undercount for Black, Hispanic, young children and renters, and overcount for non‑Hispanic Whites, older adults and homeowners [1] [2] [3] [4] [5].
1. What the official numbers say—and where they came from that changed the public record
The Census Bureau published estimated undercount and overcount rates from the 2020 Post‑Enumeration Survey in May 2022, providing state‑level net coverage error estimates used to identify which states had statistically significant miscounts. The PES compared the 2020 Census roster to an independent survey and produced state‑by‑state net coverage error rates; the Bureau’s release explicitly flagged 14 states as having statistically significant net coverage errors, a result summarized in public fact sheets and news accounts in mid‑2022 [3] [2]. The PES is the official technical product for assessing coverage error; its results have been cited repeatedly in later analyses and GAO reporting as the authoritative source documenting where the 2020 Census produced measurable net undercounts and overcounts [1] [5].
2. The list: which states were undercounted and which were overcounted
The PES and accompanying Census reporting list six states with significant net undercounts—Arkansas, Florida, Illinois, Mississippi, Tennessee, and Texas—and eight states with significant net overcounts—Delaware, Hawaii, Massachusetts, Minnesota, New York, Ohio, Rhode Island, and Utah. These state lists are explicitly reported in summaries and news explainers that drew on the Census Bureau’s May 19, 2022 release and subsequent fact checks in June 2022 [1] [2] [3]. The PES measures household population coverage; the identification of a state as “undercounted” or “overcounted” reflects a statistically significant net difference between the Census enumeration and the PES estimate, not necessarily proof of deliberate error or fraud.
3. Who was missed—and who was counted more than expected
Beyond state totals, both the Census Bureau and GAO analyses highlight systematic demographic patterns underpinning the state results: Black or African American persons, Hispanic or Latino persons, American Indian/Alaska Native persons on reservations, young children, and renters were disproportionately undercounted, while non‑Hispanic White persons, adults 50 and older, and homeowners tended to be overcounted. The demographic skew helps explain why certain states—especially those with larger shares of renters, children, or minority populations—showed net undercounts, while states with different demographic structures showed net overcounts [4] [5]. These patterns were emphasized in GAO’s 2024 coverage report and in Census Bureau materials that contextualize the PES findings for 2030 planning.
4. How other reports interpret or challenge the PES numbers
Some later commentaries and analyses focus on other technical issues that can complicate interpretation, notably the application of differential privacy to published 2020 data and debate about the absolute magnitude of net error. Research on differential privacy shows the privacy algorithm alters published counts at small geographic scales, affecting county and subgroup comparisons and complicating migration or age‑by‑origin estimates; this does not change the PES methodology but influences how researchers use published 2020 counts [6]. Independent assessments, including GAO, corroborate the PES conclusion of 14 states with significant net errors while placing those errors in context of historical patterns and Census operational challenges [5].
5. What to watch next—interpretation, uses, and policy implications
The PES state results are not a rerun of apportionment but rather an analytic tool to inform future operations and planning for the 2030 Census; they are widely cited to improve outreach to undercounted groups and to assess methodological changes. Policymakers and researchers must weigh PES state flags alongside demographic patterns and the distortionary effects of differential privacy on published tabulations; using raw 2020 counts for fine‑grained policy decisions without regard to PES findings or DP adjustments risks misinterpretation [3] [6] [5]. The combination of PES state lists and demographic breakdowns forms the official basis for acknowledging where major vulnerabilities in enumeration occurred in 2020 and guiding corrective steps for 2030.
6. Sources, competing narratives, and why the official list matters
The identification of the 14 states with significant net coverage errors is drawn from the Census Bureau’s PES release (May 19, 2022) and reinforced in GAO and Census follow‑ups; political commentary since has ranged from claims that the census was broadly accurate to critiques emphasizing operational failures and differential impacts on communities of color. The PES‑based state list is the clearest official statement on where net undercounts and overcounts occurred; users should combine that list with the demographic breakdowns and DP caveats to form a full picture when using 2020 data for research or policy [3] [2] [4] [5].