Our next once-a-decade count by the Census Bureau is under imminent threat. On September 10, 2026, the Census Bureau published a proposed rule that would remove entire categories of Minnesota residents from the next count and stop the Census from asking about race and ethnicity at all. The public has 30 days to respond.

The decennial census helps us better know ourselves and our communities. We use information from the count to allocate political representation, determine federal funding flows, understand educational enrollment, make business decisions, and identify community needs.

This is the third time in two years we have written about threats to the data Minnesotans depend on: in February 2025 the data disappeared, in March 2026 the systems behind it were dismantled, and this month the rules themselves are being rewritten.

What the rule would do

The proposal is not a memo or an executive order. It is a regulation, written into the Code of Federal Regulations, where it would outlast any single administration. And it upends rules going back to the very first census in 1790. Four changes matter most.

  • Redefine where you live. Instead of using the concept of "usual residence" – the place where you live and sleep most of the time as of Census Day – your census residence would become the place where you "lawfully spent the greatest number of days" between January 3 and April 1 of the census year, consistent with tax records.
  • Change who counts toward representation. Apportionment is the process that divides the 435 U.S. House seats among the states. The rule would count only U.S. citizens and lawful permanent residents. Everyone else would be left out, including:
    • Students on visas
    • Workers on employment visas
    • Refugees
    • People granted asylum
    • People with temporary protected status
  • Stop the Census from asking about race and ethnicity. Questions about race, ethnicity, and sexual orientation would be prohibited on the census questionnaire. Sex, date of birth, and household relationship would stay for administrative purposes.

  • Limit what gets published. The Bureau could not release demographic data when doing so would require adding statistical uncertainty to state or national population counts.

Here is what stands out to me as someone who works with Census data often: The rule does nothing to restrict what the government can know. The American Community Survey (ACS), the large annual survey that fills the years between censuses, can keep asking about race and ethnicity. The Bureau explicitly preserves its ability to pull demographic information from other agencies' records. Nothing in this rule makes the federal government know any less about you. It makes the rest of us know less about ourselves and one another.

 

Why does this matter for Minnesotans?

Representation and federal funding depend on who gets counted. According to State Demographer Susan Brower, if just 26 fewer Minnesotans participated in the 2020 Census, Minnesota would have lost our eighth congressional seat. We held the seat because Minnesotans showed up. This state posted the highest census self-response rate in the country, with over 75% of households responding on their own by phone, mail, or online. Minnesota is home to roughly 5.8 million residents, some of whom are lawfully present but are not citizens or permanent residents. Census data also shape how more than $36 billion in federal funding reaches the state each year.

A rule written around winter falls hardest on states that have one. The Bureau would set your residence by where you spent the most days between January 3 and April 1. For most of the country that window is unremarkable. In Minnesota it is winter. Snowbirds who leave for warmer climates during the coldest months would be counted in another state, and so would their share of representation and federal funding. Snowbirds tend to be older, so the services that follow them south are the ones older adults depend on. They would still vote, pay property taxes, and see their doctors here. 

The count shapes what we know about ourselves for the next ten years. We only do a decennial count once every ten years, and then rely on that count for a decade. Over the following ten years, decennial counts are used as a base for other data programs like the American Community Survey (ACS), informing who will be surveyed, whether results are representative of the population, and how results might need to be adjusted to more accurately reflect our communities. This means that an incomplete or faulty decennial count detracts from what we know about ourselves, not just in 2030, but for years beyond the official count. 

The census is the only source of small-area race and ethnicity data. Not the ACS. Not administrative records. The census. Block-level detail is what draws legislative districts, supports countering efforts to undermine the Voting Rights Act, and serves as the denominator for nearly every rate any of us calculates: infant mortality by race, uninsured rates by community, third-grade reading proficiency measured against the population it should represent.

The ACS is valuable, but it is not a substitute. In fact, accurate ACS data rely on the population base that we obtain from the decennial census. As a sample of the population, the ACS gathers information from a far smaller group of residents. Its margins of error for small geographies and communities are sometimes wide enough to swallow the very differences a program is trying to see. A rural county, a small city, or a single neighborhood cannot be described from a survey sample the way it can from a full count. For Minnesota’s cultural communities, the loss would be sharpest exactly where the populations are smallest.

This touches the Compass pages you use most. Our race and ethnicity data, cultural community profiles, small area data profiles for counties, cities, and neighborhoods, and Data for Equity work all rest on decennial data. If the census stops asking, we would not be able to report what was never collected, and 2030 would not be comparable to 2020 or 2010.

 

What makes this moment different

In February 2025, more than 8,000 pages were already gone from federal websites, thousands of them from the Census Bureau and the CDC, taking with them health guidance, local disaster risk tools, and community demographics. During the 43-day shutdown that ran from October 1 to November 12, 2025, the household survey behind the national unemployment rate never went into the field, and neither did the price collection behind October's inflation figure. That data does not exist and never will.

This time, the rule is only proposed. It is not final, there is a formal comment period, and comments become part of the legal record that agencies must consider.

 

What you can do about it

  • Comment on the rule. This is the action that matters most. Go to regulations.gov and search docket USBC-2026-0628. Comments are due October 13, 2026 at 11:59 PM EDT. Anonymous comments are accepted, and it takes about ten minutes.

Agencies must consider substantive comments, and the most useful ones are specific and grounded in your own work. Name the program, the grant, the report, the population. Tell them what you use census race and ethnicity data for, at what geography, and what breaks without it. If you have tried to substitute survey estimates for a small community and watched the margin of error exceed the estimate, describe exactly that. A concrete account from one Minnesota organization carries weight that a hundred form letters do not.

Every comment counts! Even if you're not with an organization that uses Census data, the information affects your representation, services, and federal funding flows to our state and communities. It's your right to comment on federal rule changes that impact you. Be as specific as you can.

Also worth doing

  • Contact your delegation. Tell your members of Congress how your organization uses census race and ethnicity data and ask them to exercise oversight of the proposal.
  • Get ahead of the next threat. Follow Minnesotans for the ACS, The Census Project, the Association of Public Data Users, and NPR’s Hansi Lo Wang, who has covered census policy longer and better than almost anyone.
  • Raise it with your board and your funders. If your 2030 planning assumes decennial race and ethnicity data at the tract or block level, that assumption is now uncertain. Better to surface it now than to discover it in 2031.
  • Send it to one person. Share this article with a colleague who builds programs on this data, and follow Minnesota Compass on LinkedIn for updates as the comment period closes.

Where we stand

We have written this twice before, and we will write it again. Facts matter. Data matter. Wilder Research and Minnesota Compass remain committed to gathering and sharing trustworthy information about Minnesotans, for Minnesotans.

We are watching a pattern. First the data disappeared from websites. Then the leadership and the funding behind the data were dismantled. Now the rules are being rewritten, so that some of us are not counted, and the questions that let us see one another are not asked at all.

Minnesotans have answered this call-to-action before. In 2020, in the middle of a pandemic, this state had the highest census self-report response rate in the country, and a congressional seat came down to a couple dozen people. This past winter we showed up again, in the cold, on short notice, with and for neighbors who are lawfully here and not citizens. Those are the same neighbors this proposed rule would not count.

Meet the Author