Every ten years, the federal government undertakes one of its most ambitious tasks: counting every person living in the United States. The goal is to count everyone only once and in the right place. This count determines the federal funding — and often state funding — that communities receive for schools, hospitals, roads, housing and other essential services for the next decade. The preparation process is long and arduous, but it’s critical to get the count right. The U.S. Census Bureau, however, is taking steps to scale back and reduce Census preparations in ways that could reduce the accuracy of the count in your community.
Why Does the Census Matter for Cities?
For cities, towns and villages, accurate census data means fair funding and representation. In fiscal year 2023 alone, census data determined how over $2.24 trillion in federal dollars were distributed. When a city or state is undercounted, that means that their community’s taxpayer dollars are not properly returned to them, and it can have devastating consequences on a community’s ability to adequately fund public safety, transportation, housing and more. For example, in Texas a 1.92 percent undercount cost the state more than $25 billion (PDF) in lost federal funding over the decade.
There has never been a perfect census, but the 2020 decennial had structural issues that were exacerbated by the COVID-19 pandemic. Mid-decade tests are essential for the Bureau to test new strategies, implement new technologies and root out failure points before they become costly errors. The 2026 test is the only planned test this decade, and right now it’s being scaled back in ways that will hurt communities in both the short and long term.
What Was Supposed to Happen During the 2026 Census Test?
To help avoid the issues faced in 2020, the U.S. Census Bureau planned one test this decade. The test was to take place in six strategically chosen areas that reflect the challenges of historically undercounted communities. The test was also expected to include a nationally representative self-response testing component. (You can also see an overview of the originally planned test.)
What Changed?
1. There are Now Two Sites Instead of the Planned Six
Instead of six proposed sites that span diverse community types, populations and census challenges, the census will only focus on two areas – Spartanburg, S.C., and Huntsville, Ala., which are relatively similar geographic areas.
The test is no longer representative of the country or focusing on the areas and populations that were missed in 2020.
Troublingly, this will be the second decade in a row where no rural areas are tested.
2. The Scope Has Changed
The 2026 test was originally designed to test a wide range of census challenges. Instead, the Bureau has narrowed the test (PDF) to focus on whether postal workers can serve as enumerators. The Bureau has attempted to test postal workers before without success. If the Bureau is serious about testing the efficacy of using postal workers, they must be tested in geographically diverse areas of the country, including rural areas. Additionally, the proposed test does not address the issues that led to undercounts in 2020.
3. Group Quarters are No Longer Being Tested
Colleges and universities, nursing homes, correctional facilities and military barracks are considered “Group Quarters” and are some of the most complex and consequential units to count. These facilities were a major source of undercounting errors in the 2020 Census, and the Bureau originally selected Spartanburg and Huntsville as test sites in part because of their complex Group Quarters populations. Removing this from the test scope means the Bureau will enter 2030 without having piloted strategies to address one of its most persistent and costly challenges. This gap in preparation could directly translate into lost residents — and lost funding — in the final count.
4. English-Only Testing Materials
The 2016 Census Test was available in English, Spanish, Chinese (Simplified) and Korean, and the 2020 Census was available in English, Spanish, Chinese (Simplified), Vietnamese, Korean, Russian, Arabic, Tagalog, Polish, French, Haitian Creole, Portuguese and Japanese. The 2026 census test will only be conducted in English. If the Bureau is serious about its goal of counting everyone, then the census test must be conducted in the languages that community members feel most comfortable speaking. For communities with large immigrant populations, English-only materials could mean entire neighborhoods go uncounted and unfunded.
5. National Sample Has Been Eliminated
While initially planned, the current proposal does not include any national self-response testing component, which is critical to informing 2030 Census messaging, mailing strategies and materials. Communities are asked to give their own time, energy and money toward ensuring an accurate decennial count. The national sample is critical for both the Bureau and communities to be prepared for 2030. Eliminating it now will likely lead to higher costs for both communities and taxpayers as we get closer to 2030.
6. Using the Wrong Questionnaire
Census tests historically use the previous decade’s questionnaire to mimic real conditions and produce results that can be measured against past performance. The Census Bureau has instead decided to use the American Community Survey (ACS) methodology panel. The ACS is important — but it’s not the decennial. The ACS questionnaire takes four times longer to fill out and includes information not relevant to the decennial. Additionally, by fundamentally changing the methodology, it ensures that this test cannot be compared to either the 2020 census or previous census tests. Without a comparable baseline, the Bureau will have no reliable way to measure whether 2030 preparations are actually improving on 2020 outcomes.
What Can You Do?
Each of these changes, taken together, significantly weakens the Bureau’s ability to prepare for an accurate 2030 count. The Census Bureau is accepting comments on the proposed changes to the 2026 census test until March 5, 2026. Communities can make their voices and concerns heard by reaching out to the Bureau directly or using NLC’s template letter.
If you have any questions, email the NLC Census team with questions.
Take Action
The Census Bureau is accepting comments on the proposed changes to the 2026 census test until March 5, 2026.
About the Author
Three Things to Know Video Series: February 2026
CitiesSpeak Podcast: Sitting Down for America’s Potluck
Member Winter Pulse Survey
From the Event: Young Municipal Leaders Virtual Meeting
What a Dept. of Homeland Security Shutdown Could Mean for Cities, Towns and Villages
- Yucel Ors
