Action Needed: Sweeping Census 2030 Proposed Changes

Census 2030 has been in the news lately. The administration has proposed changes that will decrease the quality of data available for decision-making in Connecticut until at least 2040.  

The short summary: In alignment with priorities laid out in Project 2025, the Department of Commerce has issued a proposed 23-page regulation that makes drastic changes to census questions and collection, upending 250 years of history.  

The major changes, which are explained in more detail below, are:

  1. Redefining the word “reside” and adding a citizenship question. Rather than using the Constitutional requirement to count “every person once, only once, and in the right place,” the proposed rule adds 3 different requirements to determine where to be counted. The stated reasoning in the rule is to eliminate non-citizens from apportionment counts.

  2. Eliminate questions related to race, ethnicity, and sex. The decennial census serves as the denominator for all race, ethnicity, and sex population counts and the basis for every other survey in the U.S., so this will reduce the usability of these data until they are collected again in the decennial census.

In addition, the department has given the public an uncharacteristically short 33 days to review and submit comments when these regulations typically require 90 days.

It is crucial that all data users in Connecticut who use population or demographic data, respond to the ruling. As of September 18, over 75% of the comments in the Federal Register are in favor of the change. We must change this!

At the end of this post, we provide sample text for you to comment and a link to where to you submit your comments. It should take no more than 5 minutes. CTData will be submitting a comment, but the more individuals and organizations that do, the better.

For more detailed explanation and on how to take action, continue reading.

Part 60: Proposed Changes to Residency Criteria

There are several changes being proposed to residency.

One of our biggest concerns: we won’t have a complete and accurate count of the Connecticut population

The proposed rule states that citizens of foreign countries who are “also either citizens or lawful permanent residents of the U.S.” will be counted. However, all other citizens of foreign countries would not be counted for apportionment. This includes: international students, work visa holders, recognized refugees, and asylees. This rule is in direct conflict with the Constitutional mandate to “count all persons.”

In 2018, testing by the Census Bureau showed that citizenship questions reduce self-response in certain locations. The fear of ICE and this administration’s abuse of the public’s information through unlawful and unethical data sharing across agencies will have a negative impact on response rates. Further, non-response follow-up for Census 2030 has been reduced to only two follow-ups, meaning that by making this change we can be certain that there will be an undercount.

Instead of Congress debating the question of whether or not non-citizens should be considered in apportionment, which is within their Article 1, Section 2 powers, the Executive branch is changing the way apportionment will be determined with this rule. And this sweeping, untested change, will damage the quality of federal population data until at least 2040.

Adding confusion on where to be counted

The proposed change also extends the enumeration from a specific date to a time period – January 3 to April 1st during the decennial census year. For each individual to consider the “residency” under which they respond in the census, they need to do so from the place where they have spent the most number of days between January 3 and April 1. People who have more than one home, or who winter in a different state, will be counted at their winter address, rather than the place they spend most of their time, pay taxes, and vote. This would have major implications for Connecticut and other northern states.

The proposed rule adds further confusion with language referring to the address of their tax record. However, the address of a tax record is not always connected to a “usual residence.” Additionally, any residence where a person “spent at least one hour physically and lawfully present there that day” counts toward determining our “usual residence.” So people who reside in more than one place are going to have to start a calendar to tally where they should respond from. What could go wrong?

The table at the end of this post, created with the assistance of Claude, goes through greater detail about the proposed changes, the reasoning provided in the rule, and the concerns raised by census data users.

Part 70: Proposed Regulations for Demographic Questions

The proposed rule eliminates the collection of race and ethnicity questions from the enumeration. Race and ethnicity questions have been on the Census since it started in 1790.

According to dataindex.us, because the decennial census sets the standard that other surveys follow, removing race there is expected to erode race data system-wide and set a precedent for further demographic-data cuts. Even if collection of race and ethnicity continues in the American Community Survey, the denominators are based in the decennial data, and so will be inaccurate.

The implications are broad for removing race and ethnicity data. Public health interventions will be flying blind. Because the census supplies the population denominators used to calculate rates, dropping race data would make health disparities invisible in the data. In areas such as maternal mortality, diabetes, and overdose deaths, gaps between racial and ethnic groups are large and will be more difficult to address without data (dataindex.us). 

As with Part 60, this change is not based on a desire for improvement of the quality of the data, but on political ideology. Project 2025, pages 680-681, makes the claim that the data “could be skewed to bolster progressive political agendas.” Maternal mortality, diabetes, and overdose deaths have nothing to do with political ideology. These changes seek to obscure reality, hide the truth, and promote a partisan agenda.

There are other proposed changes in the Federal Register Notice that are concerning. Of note:

  • Prohibition against including questions on sexual orientation, which is currently not collected on the decennial form.

  • Incarcerated people will be ‘counted at the facility’ instead of at their residence.

The table at the end of this post, created with the assistance of Claude, goes through greater detail about the the proposed changes, the reasoning provided in the rule, and the concerns raised by census data users.

Take Action

The most impactful way you can take action is to write a simple comment. If you need accurate population data for planning or writing grants, or use any race and ethnicity data in any way to improve the services you provide or to obtain funding, specific examples of how it has benefited you will be most impactful. You can use the example below to craft your own comment or look at this quick overview of what to include.

When you are ready to comment, go here to the Federal Register and provide your comment. The more comments the better. Respond as an individual or an organization.

I respectfully request that the untested changes in Document ID USBC-2026-0628-0001, which degrades and eliminates data necessary for effectively serving my community, does not go into effect. I am a [data analyst] at [your organization]. My work relies on [ACS 5-year estimates at the town level]. [Eliminating questions on race and ethnicity in the decennial census will degrade the quality of ACS demographic data since decennial data serves as the base count. Our organization will not know whether our program is reaching the population that need our services most and whether we need to change our program or outreach strategy in order to do so. This means we will be unable to meet our mission to … ].

Race has been asked in some form on every census since 1790. The existing demographic questions have been tested and shown they do not impose a burden on respondents. Additionally, the change in the definition of usual residence will cause confusion, further degrading the quality of the data I use on a daily basis. I respectfully request that these untested changes not go into effect.

Essential Reading

While I refer to some of these documents, these articles are worth reading to understand the broad list of potential impacts on data quality for the next decade and reasons for the changes.

For More Information

As the Connecticut Census State Data Center, we will continue to stay on top of census news and share it as it becomes available. You can bookmark our blog or get updates by subscribing to our newsletter and following us on Facebook, and LinkedIn, and Instagram. To learn more about the Census and resources provided by CTData, head to our Census Data Portal. Explore other data sets and analyses at data by topic and data projects. Join us for our conference to connect with other data users or learn some new data skills at our workshops.

Part 60 — Residence Criteria

Who gets counted, and where.

Italicized text marks interpretation or contested claims, not the text of the rule itself.

Proposed changeReasoning given by the Census BureauConcerns raised
Exclude illegal aliens and non-citizens with less-durable status from apportionment; count only U.S. citizens and lawful permanent residents. “Persons”/“inhabitants” means members of the “body politic”; “usual residence” implies allegiance and durable domicile. Framed as the best reading of the Constitution and, alternatively, as agency discretion under 13 U.S.C. 141 and Franklin v. Massachusetts. The census has always excluded tourists and business travelers but counted residents. Under the proposal, residents such as international students, work and H-1B visa holders, refugees, and asylees would be excluded from the apportionment count for the first time. Whether excluding residents is unconstitutional, and whether the Secretary can act without Congress, is contested and expected to be litigated.
Define “usual residence” as where a person lawfully spent the most days, evidenced by tax records. Ties residence to verifiable records and the “lawful presence” concept. Builds a lawful-presence test into the basic definition of residence. Tax records often will not match where a person actually lives. It is unclear whether self-reported residence or an address derived from tax records would govern, how record-linkage errors would be handled, or how people absent from tax records would be counted (Anand, Jarosz & Dick, dataindex.us).
Fixed enumeration period of January 3 – April 1. Provides a defined window; January 3 is the congressional term start, April 1 is the statutory Census Day. Critics argue a fixed January–April period conflicts with Title 13, which ties enumeration to a single census date rather than a period, breaking with every prior census. Where to respond is already a point of confusion for some groups, this change increases confusion (Anand, Jarosz & Dick, dataindex.us).
Use administrative records to determine legal status and residence (IRS, DHS, State, HHS, “Department of War,” USPS, state/local/tribal, commercial data). The Bureau claims far broader data access and better data linkage than in 2020, citing expanded DHS and State Department agreements. The notice does contain the language about broad data access and advanced data linkage. Critics read this as positioning the Bureau as a data-access tool for the administration and as a threat to public trust in Title 13 confidentiality.
Invites comment on adding a legal-status question to the short form. Cites Dept. of Commerce v. New York (2019) that a citizenship-type question is permissible. A citizenship/legal-status question has never been asked of the entire population. This undercuts any “restoring past practice” framing. (Anand, Jarosz & Dick, dataindex.us). A legal-status question might also escape Part 70’s limits via §70.2(d), which exempts residence-related questions.
No dedicated college-student criterion in Part 60’s section list. Students would therefore default to the general rule. Not stated; the section list simply omits a student provision. Could change where students are counted compared with the 2020 “count at school” approach. The rule also says students should be counted at their separate usual residence even though they may file taxes at a parent’s address, which conflicts with the tax-record test (Anand, Jarosz & Dick, dataindex.us).
Only lawful permanent residents are explicitly included; the Bureau seeks comment on statuses “materially similar” to LPR. Seeks input on whether other statuses resemble LPR status. Refugees, asylees, TPS holders, parolees, and visa holders are not addressed. As written, they would fall into the excluded category — a large population whose treatment is left to the comment process (Anand, Jarosz & Dick, dataindex.us).

Sources: U.S. Census Bureau, proposed rule, 91 FR 57524 (Sept. 10, 2026), Federal Register. Concerns drawn in part from Anand, Jarosz & Dick, “Why Seemingly Arcane Process Changes to the Decennial Census Would be Catastrophic for All Americans,” dataindex.us (Sept. 15, 2026).

Part 70 — Demographic Questions

What the questionnaire may and may not ask.

Italicized text marks interpretation or contested claims, not the text of the rule itself.

Proposed changeReasoning given by the Census BureauConcerns raised
Prohibit race, ethnicity, and sexual orientation questions on the enumeration questionnaire. These do not advance the core apportionment purpose; are “(potentially) sensitive”; and are of “reduced” need. Cites “colorblind Constitution” case law and Executive Order 14151 (ending federal DEI). In addition to degrading the quality of the data we have about the population, race has been collected in every census since 1790. Surveys and polls will no longer be able to assess their representativeness to the population. We will be unable to compare changes in race and ethnicity over time. It will also harm redistricting and Voting Rights Act enforcement and hamper the ability of public health officials to effectively address public health concerns.
Limit mandatory questions to biological sex, date of birth, and household relationships; make other demographic questions optional. More questions lower response rates; responding is legally required (13 U.S.C. 221), so removing sensitive items reduces burden. Making response voluntary for nearly all demographic questions is argued to be inconsistent with 13 U.S.C. 221, which makes answering census questions mandatory (Anand, Jarosz & Dick, dataindex.us). Footnote 11 is also read by some as a hint that tribal membership could still be asked, or that American Indian/Alaska Native status might be imputed from tribal records.
Bar publishing individual demographic data if doing so requires adding uncertainty to population counts. Prioritizes apportionment/redistricting accuracy and reduces disclosure risk under 13 U.S.C. 9. Critics call the burden and disclosure rationales pretextual: the response burden for these questions is minimal, and the disclosure risk is described as self-inflicted because a June 2026 Commerce action barred the modern privacy protections previously used for race and ethnicity, which is now cited to justify this rule (Anand, Jarosz & Dick, dataindex.us). The bar may also limit publication of the permitted variables (sex, date of birth) at fine geographic levels.
Preserve demographic questions on the ACS and other non-enumeration surveys. Restrictions are scoped only to the enumeration questionnaire. The ACS is a sample, not a complete count. And because the decennial census, the population estimates, and the ACS are methodologically intertwined, removing race from the decennial would degrade the weighting of the ACS and other surveys anyway — so keeping the ACS questions would not preserve reliable race and ethnicity data (Le & Klein, dataindex.us).
Severability clauses in both parts. So the rest survives if one provision is struck. Read as a signal the Bureau expects parts of the rule to be litigated.

Note: This is a proposed rule, not final. The public comment period closes October 13, 2026. Comments may be submitted through the docket (USBC-2026-0628) on Regulations.gov.

Census & ACSSarah Eisele-Dyrlicensus, equity in data, immigration