V. Census Content and the Future of Community Participation

The questions included in the census determine what the nation will be able to know about its population for the next decade. Decisions about race and ethnicity, relationships, sex, citizenship, and other characteristics affect redistricting, civil rights enforcement, program administration, research, and the visibility of particular communities in public data.

But census content also affects participation. People are more likely to respond when they believe the census recognizes their communities, collects information that serves a legitimate public purpose, and will produce data that are useful to them. They may be less willing to participate when questions are confusing, exclude important identities, appear politically motivated, or raise fears that the information could be used against them.

These effects extend beyond individual respondents. The Census Bureau depends heavily on community organizations, local governments, service providers, advocates, and other trusted messengers to explain the census and encourage participation. These partners do not simply deliver a federal message. They make their own judgments about whether the census is fair, safe, and valuable to the people they serve.

If Census weakens or removes information that communities have spent years working to improve, some organizations may have less reason to invest in Get Out the Count efforts. Others may find it difficult to reassure people that participation will benefit their communities. Changes to census content can therefore affect both the quality of the resulting data and the network of trusted partners on which direct response depends.

This section examines the relationship between census content, public legitimacy, and community participation. It considers how changes to major demographic questions could affect data quality, whether the questionnaire continues to reflect tested standards and community input, and how content decisions may strengthen or weaken the willingness of organizations and residents to support the count. The issue is not simply which boxes appear on the questionnaire. It is whether the 2030 Census will continue to provide communities with a meaningful reason to participate.

A. The Operational Plan Does Not Settle Questionnaire Content

The 2030 Census Operational Plan should not be read as a final decision about what questions will appear on the census form. The plan reflects the laws, statistical standards, policy assumptions, and questionnaire concepts in place when the document was prepared. It describes how Census expected to collect and process information as of June 30, 2025, but it does not prevent later political or administrative decisions from changing that content.

This distinction is especially important because questionnaire content can change outside the ordinary operational-planning process. The President, the Secretary of Commerce, Congress, the Office of Management and Budget, or Census leadership may seek changes involving citizenship, race and ethnicity, sex, household relationships, or other characteristics. Some changes could occur through new legal requirements or revised government-wide standards. Others could emerge through Commerce Department directives, questionnaire-development decisions, or changes in how Census processes and publishes the responses it receives.

The Operational Plan generally assumes that Census will continue collecting the information needed to carry out the design described in the document. For example, plans for the Person Characteristic Frame, response processing, quality review, and data products depend in part on assumptions about which characteristics Census will collect directly and how those characteristics will be defined. If the questionnaire changes, those operational systems may also need to change.

The plan therefore provides a useful baseline, but not a complete account of the content risks facing the 2030 Census. It cannot resolve proposals developed after its publication, and silence in the plan should not be treated as evidence that a politically contested question or category is secure.

This uncertainty applies to both the addition and removal of content. A citizenship question or another politically sensitive item could be added even though it is not part of the baseline design. Race and ethnicity questions could be weakened or revised despite the plan’s apparent reliance on the 2024 federal standards. Sex and relationship categories could also be changed in ways that affect the identification of same-sex couples, transgender people, or less conventional families.

Changes may also occur through processing rather than through the visible questionnaire. Census could preserve a question but alter how responses are categorized, edited, combined, or released. A household may therefore see familiar language on the form while the resulting data provide less detail or treat particular responses differently.

Advocates should consequently track questionnaire content as a separate policy process rather than assuming it was settled by the Operational Plan. The relevant questions include not only what Census currently intends to ask, but also who retains authority to change those questions, what testing and public review would precede a change, and how late in the decade Census could revise the content without undermining operations or data quality.

Current commitment

The Operational Plan reflects the questionnaire concepts and statistical standards that were in place when the document was prepared.

Open design question

Census has not established that those assumptions will remain unchanged through 2030, particularly for politically contested characteristics.

Credible risk

Later policy decisions could add, remove, weaken, or reinterpret census content after operational systems and community-engagement plans have already been built around a different questionnaire.

Why advocates should care

Questionnaire changes can affect more than the variables available in the final data. They can alter public trust, community participation, partner engagement, processing systems, and the ability of Census to produce accurate and comparable information.

B. Race and Ethnicity Questions

Race and ethnicity are among the most consequential and politically vulnerable subjects on the decennial census. These questions produce the data used to understand where racial and ethnic communities live, whether they have equal access to political representation and public programs, and how the nation’s population is changing. They also affect whether communities see themselves accurately represented in the census and whether trusted organizations are willing to encourage participation.

The current baseline is the 2024 revision to Statistical Policy Directive No. 15 (SPD 15). The revision requires federal collections to use a combined race and ethnicity question, adds Middle Eastern or North African as a co-equal minimum category, allows respondents to select multiple identities, and generally requires the collection of detailed race and ethnicity information. The standards resulted from years of Census Bureau research, federal expert review, public comment, and engagement with racial and ethnic communities. Census concluded that the combined question would reduce missing and “Some Other Race” responses, particularly among Hispanic and Latino respondents, and that a separate MENA category would better reflect how many people in those communities identify.

As of August 2026, the 2024 standards remain formally in effect. OMB has not rescinded them, and its official website continues to describe them as the governing federal standards. OMB has, however, extended the implementation timeline, including the deadline for bringing federal information collections into compliance. Delay does not necessarily mean reversal, but it creates additional time for OMB or political leadership to reconsider the standards before Census finalizes the 2030 questionnaire.

There is therefore a substantial risk that the Administration will roll the standards back. The most plausible rollback would be a return to the 1997 standards, which governed the 2000, 2010, and 2020 Censuses. Those standards require separate Hispanic-origin and race questions, allow respondents to select more than one race, and classify people with Middle Eastern or North African origins within the White category. They permit detailed data collection but do not require it in the way the 2024 standards generally do.

A return to the 1997 standards would not eliminate race and ethnicity data, and Census could choose to preserve some of the detailed response options used in 2020. But it would abandon the combined question and MENA category before they are fully implemented. It would also remove the government-wide requirement to collect detailed information in most circumstances. The result would likely be less accurate reporting among Hispanic and Latino respondents, continued classification of MENA communities as White, and greater uncertainty about whether detailed groups will remain visible in federal data.

A more severe rollback to the 1977 standards appears less likely but may be possible. Those standards recognized only four minimum race categories, combining Asian and Pacific Islander populations, and treated Hispanic origin as a separate concept. They predated the 1997 decision to allow people to select more than one race. Returning to that framework could eliminate multiracial reporting as a federal requirement and collapse communities that have been separately represented for decades.

There is also a non-zero possibility that political leadership will attempt to remove race and ethnicity questions from the decennial census altogether. SPD 15 establishes how federal agencies must collect and present race and ethnicity data when they conduct such a collection; it does not itself require every agency or program to collect the information. Rescinding or weakening SPD 15 is therefore not the only possible route to a change in the race and ethnicity questions for the 2030 census. Commerce or Census leadership could argue that the decennial census no longer needs to collect the information directly.

No such decision has been formally announced, and removing the questions would face significant legal, operational, and political objections. But several current developments make the possibility more credible than it would have appeared in prior census cycles.

First, the Administration has directed agencies to dismantle disparate-impact rules and enforcement. Executive Order 14281 instructs the Attorney General to initiate the repeal or amendment of agency Title VI regulations that contemplate disparate-impact liability and directs agencies to deprioritize enforcement based on that theory. Race and ethnicity data remain important for identifying intentional discrimination and administering many federal programs, but reducing the government’s role in addressing unequal outcomes also reduces one of the institutional demands for collecting consistent demographic data.

Second, federal agencies are beginning to question the collection of data on race and ethnicity. In July 2026, the Equal Employment Opportunity Commission proposed rescinding longstanding requirements for employers and other covered entities to report workforce data by race and sex. The Commission’s stated rationale was not merely that the reports were burdensome. It argued that routinely categorizing workers by race and sex may conflict with a “colorblind” interpretation of civil rights law and raise constitutional concerns. That proposal applies to employment reporting, not the census, but it demonstrates that the Administration may view the collection of data on race and ethnicity as legally or politically suspect.

Third, the Supreme Court’s decision in Louisiana v. Callais narrowed the role of the Voting Rights Act in race-conscious redistricting. The Court held that Section 2 liability requires evidence supporting a strong inference of intentional racial discrimination, placed new limits on the use of race in illustrative maps, and gave less weight to historical discrimination and continuing disparities. The decision does not prohibit Census from collecting race and ethnicity data. Those data remain relevant to identifying intentional discrimination, evaluating voting patterns, and conducting redistricting. But the decision weakens one of the clearest statutory uses of detailed decennial race and ethnicity data and may give political officials an opening to argue that the information is no longer as necessary as it once was. That conclusion is an inference from the decision, not a holding of the Court.

Redistricting and Voting Rights

Race and ethnicity data remain essential to the redistricting process. Census’s Public Law 94-171 products provide population data at very small geographic levels, including race, Hispanic origin, and voting-age population. States use these files when drawing congressional, legislative, and local districts, while courts, advocates, and government agencies use them to evaluate whether districting decisions unlawfully diminish the political opportunities of racial or ethnic communities.

Callais changes the legal analysis, but it does not make data on race and ethnicity irrelevant. Determining whether officials intentionally discriminated requires evidence about which communities were affected and whether race influenced the resulting map. Without accurate small-area race and ethnicity data, both enforcement and defense of voting-rights claims would become more difficult.

Removing the questions could also create a practical contradiction. States would still have to draw districts with equal populations, and federal voting-rights protections would still prohibit intentional racial discrimination. But the government would have weakened the principal nationwide source for understanding where racial and ethnic populations live, making it much more difficult to identify where intentional discrimination has occurred.

Civil Rights Enforcement and Program Administration

Race and ethnicity data support civil rights enforcement well beyond voting. Agencies use them to identify possible discrimination in employment, education, housing, health care, lending, environmental protection, and federally funded programs. The data do not prove that discrimination occurred, but they can show patterns that warrant investigation and provide a baseline for determining whether policies affect communities differently.

Federal, state, tribal, and local programs also use race and ethnicity data for planning, evaluation, outreach, and the distribution of services. Census itself explains that these data support antidiscrimination enforcement and help governments plan and fund programs serving particular populations. Not every funding formula contains a direct race variable, but many programs depend on population estimates, geographic definitions, survey weights, and measures of community need that are built from or evaluated using decennial census data.

The Administration’s rejection of disparate-impact liability may narrow some of these uses, but it does not eliminate them. Laws prohibiting intentional discrimination remain in force. Agencies still need to understand whom their programs serve, whether eligible communities can access services, and whether enforcement actions are supported by evidence. Removing the data would make those responsibilities harder to carry out, not make the underlying differences disappear.

Detailed Community Data

Broad categories cannot describe the experiences of every community included within them. Statistics about the Asian population as a whole can conceal major differences among Chinese, Filipino, Hmong, Asian Indian, and other groups. The same problem arises among Black, Hispanic or Latino, White, American Indian or Alaska Native, Pacific Islander, and MENA communities.

The 2020 Census ultimately produced data for nearly 1,500 detailed racial and ethnic groups, tribes, and villages. These data allow communities to understand local populations that may be nearly invisible within broad national categories. They support language planning, health research, community services, civil rights work, and public understanding of demographic change.

The 2024 standards strengthen this work by making detailed collection the general expectation rather than an optional enhancement. Rolling back to 1997 would not necessarily end every detailed tabulation, but it would remove that expectation. Returning to 1977 or eliminating the questions could make many communities impossible to identify in decennial data at all.

Trust Built Through Research and Engagement

The 2024 standards were not developed through a quick change in political preference. They followed more than a decade of Census testing and a government-wide review that included public comments, listening sessions, expert analysis, and extensive engagement with racial and ethnic communities. Census publicly told those communities that the new approach would produce more accurate and representative data.

Reversing the standards before implementation would therefore do more than change a questionnaire. It would signal that years of scientific research and public participation can be discarded before the resulting improvements reach the public. Communities that advocated for a MENA category, a combined question, or better detailed data may reasonably question whether future engagement with Census will influence final decisions.

That loss of trust could affect enumeration itself. Community organizations are more likely to invest in Get Out the Count work when they believe the resulting data will recognize and serve their communities. If race and ethnicity questions are weakened or removed, some organizations may reduce their participation or struggle to explain why residents should respond. Lower trust and weaker community engagement could then reduce direct response in the same populations whose administrative records are least likely to provide a complete substitute.

The content decision therefore cannot be separated from census operations. Weakening the race and ethnicity questions could reduce the quality of demographic data, diminish community participation, and increase reliance on administrative records and field methods that may represent those communities less accurately.

Current commitment

The 2024 SPD 15 standards remain in effect and establish a combined race and ethnicity question, a MENA category, multiple-response reporting, and generally required detailed data. Census has previously stated that it intends to implement the standards for the 2030 Census.

Open policy question

OMB and the Department of Commerce have not guaranteed that the 2024 standards or Census’s implementation plans will remain unchanged through 2030.

Credible risk

The Administration returns to the 1997 standards, restoring separate race and Hispanic-origin questions, classifying MENA respondents as White, and eliminating the general requirement for detailed data.

More severe risk

The Administration attempts to return to the 1977 framework or remove race and ethnicity questions from the decennial census altogether, based on the view that government should not routinely classify people by race.

Why advocates should care

Race and ethnicity content affects redistricting, civil rights enforcement, program administration, detailed community data, and the willingness of trusted organizations to support the census. Weakening the questions could make communities less visible in the final data while also making them less likely to participate in the count.

Where to look: For the current baseline, see OMB’s 2024 revision to Statistical Policy Directive No. 15 and the Census Bureau’s race and ethnicity standards page, which describes plans to implement the combined question, MENA category, and detailed data collection in the 2030 Census. Census materials on its question-design research explain why the combined format improves reporting, particularly among Hispanic and Latino respondents.

For the uses of the data, see the Census Bureau’s P.L. 94-171 Redistricting Data Program and its detailed demographic data products. These materials show the role of race and Hispanic-origin data in redistricting and the much greater community detail available beyond the broad minimum categories.

For emerging policy risks, watch future OMB and Commerce decisions, Federal Register notices, and Census questionnaire and data-product memoranda. Relevant indicators of the Administration’s broader direction include Executive Order 14281 on disparate-impact liability, the EEOC proposal to eliminate annual race and sex reporting, and the Supreme Court’s decision in Louisiana v. Callais. These sources do not establish what Census will do, but they help explain why continued implementation of the 2024 standards should not be assumed.

C. Citizenship or Other Politically Sensitive Additions

A citizenship question is the clearest example of a questionnaire addition that could affect the census far beyond the information collected by the question itself. Other additions could create similar concerns if the public understands them as connected to immigration enforcement, voting eligibility, benefit eligibility, or another politically contested government purpose.

As of August 2026, Census has not announced that a citizenship question will appear on the 2030 decennial questionnaire. Citizenship is included in the 2026 Census Test, but that test uses an American Community Survey-style questionnaire containing many subjects that would not ordinarily appear on the short decennial form. Its inclusion should therefore be treated as an issue to watch, not as proof that Census has selected citizenship content for 2030.

The possibility is nevertheless credible. The current Administration has made citizenship verification a major priority in other contexts, including through executive actions directing increased use and sharing of federal citizenship data for election administration. Those policies do not require a census question, but they increase the likelihood that political officials will view citizenship information as important to the decennial census or to the population data produced from it.

A citizenship question would not be categorically unlawful. In Department of Commerce v. New York, the Supreme Court concluded that the Enumeration Clause and Census Act did not themselves prohibit asking about citizenship. But the Court rejected the Commerce Department’s decision because the evidence did not match the Secretary’s stated explanation for adding the question. A future Administration could try again with a different rationale and administrative record, making the operational and evidentiary questions at least as important as the abstract legal authority.

Effects on Self-Response

The most immediate concern is whether a politically sensitive question would reduce the number of households that respond on their own. Self-response generally produces better information at lower cost than field follow-up, and declining self-response would increase reliance on enumerators, proxies, administrative records, and imputation.

Census research does not produce one simple estimate of the effect of a citizenship question on self-response rates. An analysis comparing the 2010 Census and American Community Survey predicted a substantial reduction in response among households that might contain noncitizens and an overall decline in self-response. The randomized 2019 Census Test, however, found no statistically significant difference in overall self-response between forms with and without citizenship, although it found lower response for some subgroups and places. A later Census analysis found considerable variation within the test results, with the largest declines among households containing noncitizens who were ineligible for Social Security numbers and households containing immigrants born in Latin America.

The most defensible conclusion is therefore not that a citizenship question would depress response equally across the country. It is that national averages may conceal meaningful declines in response rates among the households most likely to fear how the information could be used. A small national effect could still produce serious undercounts in particular communities.

The political context will also matter. A question asked during a period of aggressive immigration enforcement, extensive government data sharing, or public statements linking population data to political representation may produce a different reaction from the same question asked in a less threatening environment. Testing the wording alone cannot fully reproduce the conditions under which the census will occur.

Census should therefore evaluate both overall self-response and differential response. Research should examine effects by citizenship, nativity, race and ethnicity, language, mixed-status household composition, and geography. It should also test whether trusted communications and strong confidentiality messages can mitigate the effects rather than assuming that the experience will be uniform.

Household Omissions

The risk is not limited to households refusing to respond to the census altogether. A person may respond but leave a particular household member off the roster, avoid answering the sensitive question for one household member, or provide information they believe is safer rather than information they believe is accurate.

Mixed-status households may be especially vulnerable. A citizen may complete the census for the household but omit a noncitizen relative. A parent may include U.S.-born children but hesitate to identify another adult. A respondent may be uncertain whether supplying citizenship information for someone else could create legal consequences, even when Census promises that the response is confidential.

Census research on citizenship data has found that some noncitizen respondents skip the question, report themselves as citizens, or fail to supply citizenship information for other household members. Administrative records also have weaknesses, including delayed information about naturalizations. These findings show why neither a questionnaire answer nor an outside record should be assumed to provide a complete and error-free solution.

The 2019 Census Test did not include the full Nonresponse Followup operation and therefore could not determine the question’s ultimate effect on the completeness and accuracy of the population count. It also could not fully measure whether people who responded left particular household members out. A conclusion that the question produced little overall change in test response should not be treated as proof that it would cause no differential omissions in production.

Those omissions would be particularly difficult to correct through the administrative-record systems discussed in Part Two. The people most likely to be withheld from a response because of immigration-related fear may also have incomplete, inconsistent, or outdated government records. Adding a sensitive question could therefore reduce the direct information available while increasing reliance on a substitute that is weakest for some of the same households.

Partner Participation

Questionnaire content also affects the organizations that help Census reach the public. Community groups, service providers, immigrant-rights organizations, faith institutions, local governments, and other trusted messengers often explain why the census matters and reassure residents that participating is safe.

Census’s research for 2020 found that fear of repercussions, concerns about privacy, and general distrust of government were important barriers to participation. It also found that trusted community voices could be particularly valuable in reaching people with the greatest distrust. Information about confidentiality was important but was not, by itself, sufficient to resolve those concerns.

The addition of a politically sensitive question could place partner organizations in a difficult position. An organization serving immigrants may strongly support a complete count while also believing that the question creates a real or perceived danger for the people it serves. Some partners may continue their Get Out the Count work but refuse to promote the questionnaire without qualification. Others may redirect their efforts toward explaining legal rights, warning residents about possible risks, or advocating for removal of the question.

Partner withdrawal would matter even where the question’s direct effect on response appears modest. Census cannot easily replace the relationships, language skills, and community credibility those organizations provide. Losing trusted messengers could weaken participation across entire communities, including among U.S. citizens who are themselves not the subject of the question.

Census should engage partners before making a content decision, not after the questionnaire is finalized. It should ask whether organizations will remain willing to promote the census, what information they would need to do so responsibly, and whether the addition would change the communities or messages their outreach can reach effectively.

Communications Strategy

A politically sensitive question would require major changes to the communications campaign. Census would need to explain why the question is being asked, how the information will be used, whether every person must answer, and how Title 13 limits access by immigration, law-enforcement, public benefit systems, and other agencies.

The campaign would also need to address questions Census may not be able to answer through a simple slogan. People may ask why citizenship is needed on the decennial census when the American Community Survey and administrative records already collect related information. They may ask whether the data will be used to alter apportionment, redistricting, or immigration policy. They may also distinguish between the Census Bureau’s legal promises and their trust in the broader federal government.

Census’s 2020 communications research found that privacy statements alone did not eliminate concerns about repercussions and government distrust. Messaging was more effective when delivered through trusted voices and connected to concrete benefits for the community.

The communications problem would become harder if political officials and career Census staff offer different explanations. A campaign focused on confidentiality and universal participation could be undermined by statements elsewhere in the Administration that connect citizenship data to immigration enforcement, voting, or the allocation of political power. Census should assess the full public environment, not merely test whether its own advertisements are understandable.

A late content change would make this work even more difficult. Audience research, translations, partner materials, advertising, enumerator scripts, and public explanations would all need to be revised. Messages developed around one questionnaire cannot simply be reused if a newly added question changes what communities believe is at stake.

Fieldworker Safety

Lower self-response would require Census to conduct more in-person follow-up in communities where residents may have serious and well-founded concerns about how the government could use their information. If a questionnaire includes citizenship or another question associated with immigration enforcement, some residents may understand a knock at the door not as a routine census visit, but as a potentially consequential interaction with the federal government.

That fear could make fieldwork harder and more stressful for everyone involved. Residents may be reluctant to open the door, identify everyone in the household, or continue an interview once they learn what information is being requested. Some may worry about exposing relatives, jeopardizing immigration status, or creating a record that could later be used by another agency. These reactions should not be treated as hostility or unwillingness to cooperate. They may reflect the real experiences of families and communities that have seen government information used for punitive purposes.

Enumerators would be placed in the difficult position of asking people to trust protections they did not create and cannot personally guarantee. Residents may want to know whether a noncitizen must be included, whether citizenship information can be shared, or whether another agency could identify the household. A short script about confidentiality may not resolve those concerns, particularly when statements and policies from elsewhere in the federal government appear to point in a different direction.

Many enumerators will also live in or come from the communities they are counting. They may share residents’ concerns, speak the same languages, or have family members affected by immigration policy. Census should not place them in the position of defending a politically contested question without honest information, strong training, and meaningful institutional support.

The Bureau would need to prepare enumerators to respond with empathy and accuracy, recognize when a resident is afraid rather than simply refusing to respond, and avoid escalating an already sensitive interaction. It should consult fieldworkers, immigrant-serving organizations, and local partners about appropriate scripts, training, identification practices, and support systems. It should also be clear that meeting production goals does not require enumerators to pressure residents who believe answering could place their families at risk.

This is another reason the effects of a sensitive question cannot be measured only through online self-response rates. Even a modest decline in self-response could create a much larger field operation concentrated in communities experiencing heightened fear and uncertainty. The result could be more difficult interactions, greater emotional strain on residents and enumerators, and less complete information from the households Census most needs to reach directly.

Litigation and Schedule Risk

A citizenship question or another politically sensitive question addition would almost certainly receive close legal scrutiny. Department of Commerce v. New York establishes that Commerce has broad authority over census content, but it must provide a genuine explanation supported by the administrative record. A decision based on an implausible rationale, disregard of Census’s own evidence, or inadequate consideration of accuracy and operational consequences would be vulnerable under the Administrative Procedure Act.

Litigation could address additional issues depending on the question and stated purpose, including equal protection, discriminatory intent, statutory authority, privacy, and compliance with required procedures. Even if the government ultimately prevails, the case could consume senior leadership attention and leave Census uncertain about the questionnaire while systems and materials must be finalized.

Federal law requires the Secretary to report proposed census subjects to Congress at least three years before Census Day and proposed questions at least two years before Census Day. For 2030, those deadlines fall in 2027 and 2028. They provide important markers, but they do not eliminate the possibility of later changes or litigation over the Secretary’s final decision.

Census’s own assessment of the 2020 content operation recommends finalizing content before the dress rehearsal and beginning comprehensive testing early. That lesson is particularly important for a question likely to affect response behavior, communications, translations, processing, fieldwork, and litigation.

A late addition could force Census to change the questionnaire after important testing has occurred. It could require revisions to electronic instruments, paper forms, processing rules, training, translated materials, administrative-record systems, and communications research. Litigation could then make it unclear which version Census should prepare to use.

The 2020 experience demonstrates the schedule problem. The Supreme Court did not hold that a citizenship question was inherently impermissible, but it rejected the rationale offered for adding it near the point when Census needed to finalize its forms. The question was ultimately omitted because there was not enough time to complete a lawful new decision and remain on schedule.

The risk for 2030 is not simply that litigation could delay printing. A more technologically integrated census has dependencies across questionnaire programming, person-frame matching, near-real-time processing, outreach, and administrative-data use. A late judicial ruling could require changes across many systems at once.

Current status

Census has not announced a citizenship question or another politically sensitive addition to the 2030 decennial questionnaire. The citizenship question included in the 2026 Census Test is part of a broader ACS-style test instrument and does not settle the final 2030 content.

Open policy question

Political leadership retains substantial authority to propose questionnaire content, subject to statutory reporting requirements, administrative-law review, and the operational need for testing.

Credible risk

A citizenship question or similar addition reduces response and increases omissions among particular communities, weakens partner participation, complicates communications, and sends more fieldworkers into households with heightened distrust.

Worst-case scenario

A controversial question is added late, after systems and outreach plans have been built around a different questionnaire. Litigation remains unresolved as the dress rehearsal or production deadlines approach, forcing Census either to use an inadequately tested question or redesign major parts of the operation on an emergency schedule.

Why advocates should care

A question can affect the accuracy of the census even when most households answer it. Concentrated reductions in participation, household omissions, partner withdrawal, and litigation-driven disruption could shift the greatest burden onto the communities already most likely to be undercounted.

Where to look: For current indications about 2030 content, see the 2026 Census Test materials. The test includes citizenship because it uses the same questions as the American Community Survey, so its inclusion does not establish that citizenship will appear on the 2030 decennial questionnaire. The most consequential formal disclosures will be the Secretary of Commerce’s reports to Congress identifying proposed census subjects in 2027 and proposed questions in 2028, as required by 13 U.S.C. § 141(f).

For evidence about participation and operational effects, see the 2019 Census Test Report and the 2020 Census Barriers, Attitudes, and Motivators Study. The 2019 test found no statistically significant overall change in self-response, but later research found significant differences in sensitivity to the question by race and citizenship status. The barriers research documents concerns about repercussions, confidentiality, and government distrust, as well as the importance of trusted community messengers.

For the legal and schedule risks, see Department of Commerce v. New York and the 2020 Census Content and Forms Design Operational Assessment. The Supreme Court did not prohibit a citizenship question but rejected the rationale offered for the 2020 addition. The operational assessment recommends finalizing census content before the dress rehearsal and beginning comprehensive testing early in the decade.

D. GOTC Infrastructure as Part of Census Operations

Get Out the Count, or GOTC, activity is sometimes described as community outreach that happens alongside the official census operation. That framing understates its importance. Community participation is part of the practical infrastructure Census relies on to reach people who may not respond to federal mail, advertising, or a knock from an unfamiliar government employee.

The Census Bureau cannot build direct relationships with every community in the country during a short enumeration period. It depends on local governments, tribal nations, civil rights organizations, service providers, faith institutions, libraries, schools, unions, neighborhood groups, and other trusted messengers to help explain what the census is, why it matters, and how people can participate safely.

These organizations provide more than publicity. They often understand which messages will be credible, which languages and communication channels residents use, and which concerns Census must address before people will respond. They may know that mail is not reaching a particular building, that residents of an informal housing unit are afraid to identify themselves, or that a federal advertisement is being misunderstood in the local political context.

Partners can also help Census recognize problems that are difficult to see in national data. A response-rate map may show that participation is low in a neighborhood. A local organization may be able to explain whether the cause is unreliable addresses, limited internet access, fear of immigration enforcement, inaccessible response options, or a rumor spreading through the community. That information can help Census choose a more effective operational response.

GOTC infrastructure is particularly important for populations that have strong reasons not to trust government communications. A statement from Census that responses are confidential may be legally accurate but insufficient on its own. Residents may be more willing to consider that assurance when it is explained by an organization with a record of standing with the community and providing accurate information during other government interactions.

The value of trusted messengers does not mean that community organizations can substitute for Census’s own responsibilities. Partners cannot compensate fully for an inaccessible questionnaire, inadequate language services, weak field procedures, or policies that make residents fear how their information will be used. Census must provide the legal protections, response options, staffing, translated materials, and operational follow-through that make a partner’s message credible.

Nor should Census treat community participation as free labor. Effective GOTC efforts require staff time, planning, training, materials, data, and often direct financial support. Organizations serving historically undercounted communities may already operate with limited resources while responding to urgent needs unrelated to the census. Their participation should be planned and supported as a core operational investment rather than assumed to arise automatically from goodwill.

The content of the census can strengthen or weaken this infrastructure. Organizations are more likely to invest in GOTC work when they believe the questionnaire will produce useful data, recognize the communities they serve, and protect respondents from harm. They may be less willing to promote a census that removes important demographic information, adds a question they believe will frighten residents, or appears connected to punitive government policies.

This does not mean every partner will withdraw in response to a controversial decision. Some may intensify their work because they believe an accurate count is even more important under difficult conditions. But they may change the message they deliver. An organization may encourage residents to participate while also warning them about a question, explaining their rights, or criticizing Census’s policy choices. Others may decide that they cannot responsibly offer the unqualified reassurance Census seeks.

These choices can have operational consequences. If trusted organizations reduce their participation, Census may lose access to community events, local media, language expertise, neighborhood networks, and messengers who can counter misinformation. The Bureau may then face lower self-response, more difficult fieldwork, and greater reliance on proxies or administrative records.

A weakened partnership network can also make emerging problems harder to detect. Community organizations often provide early warnings about fraud, confusing mailings, inaccessible systems, language failures, or fears affecting participation. Without those channels, Census may not understand why response is falling until there is little time to adjust.

Census should therefore evaluate major policy and content decisions partly through their likely effects on partner participation. Before adding a sensitive question, reducing language services, or changing demographic content, the Bureau should ask whether trusted organizations will still be willing and able to promote the census and what support they would need to do so credibly.

Partnership planning should also begin well before peak census operations. Census should maintain relationships between decennial cycles, consult community organizations while major design decisions remain open, and involve them in testing messages and response tools. Asking partners to support a decision only after it has been finalized is not meaningful engagement.

Accountability should run in both directions. Census should explain how partner feedback affected its decisions and report where partnership and outreach resources were deployed. Community organizations should be able to identify emerging problems without being treated as responsible for solving operational failures that belong to the Bureau.

The central point is that GOTC work is not external decoration added to an otherwise complete census design. It is one of the ways the census reaches people whom formal government systems do not reach effectively on their own. A design that weakens community trust or partner participation may therefore undermine the count even if the questionnaire, technology, and field systems function exactly as planned.

Current commitment

Census plans to work with governments, community organizations, and other trusted partners to encourage participation and reach populations that may not respond through standard federal communications.

Open policy question

Census has not yet shown how questionnaire changes, language policies, confidentiality concerns, or other politically sensitive decisions will affect partner willingness and capacity to support the 2030 count.

Credible risk

Census treats GOTC participation as a communications supplement rather than an operational dependency and makes major policy decisions without considering whether trusted organizations can continue to promote the census credibly.

Why advocates should care

When community participation weakens, Census loses more than promotional support. It loses local knowledge, trusted messengers, language capacity, early warnings about operational problems, and one of its most effective ways to increase direct response among historically undercounted populations.

E. Worst-Case Scenario

The worst-case scenario is not simply that one questionnaire change reduces data quality or another reduces participation. It is that Census simultaneously weakens the public value of the count and increases the perceived risk of responding.

In this scenario, the Administration removes the race and ethnicity questions or rolls them back so substantially that the resulting data no longer meet the needs of many communities. Detailed identities disappear, the 2024 SPD 15 improvements are abandoned, or the available categories become too broad and outdated to support meaningful civil rights, redistricting, research, and community-planning work.

Racial justice organizations and other community groups then face a difficult choice. They may still believe that every person should be counted, but the practical reason for devoting substantial staff time and resources to Get Out the Count work becomes weaker. Organizations that spent years advocating for better race and ethnicity data may conclude that the census will no longer produce information capable of serving the communities they represent.

Some organizations may withdraw from GOTC work altogether. Others may continue participating but with fewer resources or a less enthusiastic message. Instead of telling residents that the census will make their communities visible and support political representation, services, and civil rights, partners may be left explaining why participation still matters even though the government has chosen not to collect or publish the information those communities fought to improve.

At the same time, Census adds a politically charged question, such as citizenship, or adopts another policy that people associate with immigration enforcement, voting eligibility, or punitive government action. The addition increases fear among immigrants, mixed-status families, and others who worry that information supplied to the federal government could be used against them or their relatives.

These two developments reinforce one another. People perceive more risk in responding while seeing less benefit from the data the census will produce. The trusted organizations that would ordinarily help explain confidentiality protections and encourage participation have less capacity or willingness to do so. Census loses community infrastructure at precisely the moment it needs that infrastructure most.

The effects would likely extend beyond the groups directly affected by the questionnaire changes. Fear can spread through families and social networks. A U.S. citizen may hesitate to respond because another household member is undocumented. A racial justice organization may reduce its GOTC work even though the census still collects a basic population count. Confusion about one question may undermine trust in the entire questionnaire.

Lower self-response would then place greater pressure on field operations. Census would need to send more enumerators into communities where residents are already anxious about government contact. Fieldworkers would be asked to reassure households about policies they did not create, while trusted local partners may no longer be available to reinforce those assurances.

If direct follow-up does not recover the missing responses, Census would rely more heavily on proxies, administrative records, and in-office enumeration. Yet those sources may be least complete for recent immigrants, highly mobile families, young children, informal households, and others most affected by the decline in participation. A content decision that discourages response could therefore increase reliance on the very methods least capable of replacing the information residents would have provided themselves.

The result could be a census that both undercounts communities and describes them less accurately. Some residents would be absent from the population count. Others might be counted through outside records but lose their detailed race and ethnicity, current household relationships, or correct address. Because community organizations have stepped back, Census may also receive fewer early warnings about where participation and data quality are deteriorating.

This scenario could create a lasting cycle of distrust. Communities would see that years of research and engagement did not protect important questionnaire improvements. They would also see the census adopt content associated with government enforcement or political conflict. A resulting undercount could then be blamed on residents’ unwillingness to participate rather than on the policy choices that made participation appear both less valuable and more dangerous.

The damage would not end with the 2030 Census. Organizations asked to rebuild trust for later federal surveys may be less willing to do so. Residents may be more skeptical of assurances that their participation will influence government decisions or that collected information will remain limited to statistical purposes. Weakening one census could therefore make future public data collections harder as well.

This outcome is not inevitable. Census can preserve useful and scientifically tested demographic content, avoid additions likely to undermine participation, consult partners before major decisions are final, and invest in the community infrastructure needed to reach historically undercounted populations. But it should not assume that GOTC organizations will continue to mobilize regardless of what the census asks or what data it produces.

Worst-case scenario

Census removes or substantially weakens race and ethnicity content while adding a politically charged question that increases fear among immigrants and other communities. Trusted organizations no longer believe the census produces data worth mobilizing around or cannot credibly reassure residents that responding is safe. Self-response declines, community feedback weakens, and Census becomes more dependent on fieldwork and administrative records that represent the affected populations less accurately.

Why advocates should care

Census depends on communities believing both that participation is safe and that the resulting data are valuable. A design that undermines both beliefs at once could weaken the count more profoundly than either content change would on its own.

F. Research and Advocacy

The most important content decisions for the 2030 Census may not be final for several years, but research and advocacy cannot wait until the questionnaire is formally announced. By that point, political positions may be entrenched, testing may be complete, and the time available to revise instruments, communications, and field procedures may be limited.

Advocates should prepare for several possible outcomes at once. The 2024 race and ethnicity standards may remain in place, be rolled back to an earlier framework, or be weakened through processing and data-product decisions. Citizenship or another politically sensitive question could be proposed. Some changes may be announced publicly, while others may emerge gradually through testing documents, OMB decisions, Commerce directives, or revised questionnaire specifications.

The work should therefore combine research, coalition development, policy advocacy, legal preparation, and communications planning.

Model How Questionnaire Changes May Affect Self-Response

Census and independent researchers should evaluate how proposed content changes may affect participation before those changes are adopted. Research should examine not only whether a question produces a statistically significant change in national self-response, but whether it changes response among particular communities, household types, and geographic areas.

This analysis should consider several possible effects:

• households that do not respond at all;

• households that respond but omit particular residents;

• respondents who skip or misreport a sensitive item;

• changes in the willingness of trusted organizations to promote the census; and

• the resulting increase in fieldwork, proxy responses, and administrative enumeration.

Research should also test combinations of content changes. Removing useful race and ethnicity detail may reduce the perceived benefit of participation, while adding a politically charged question may increase the perceived risk. Studying each change separately could miss the larger effect of weakening the reason to participate at the same time the questionnaire becomes more threatening.

The political environment must be part of the analysis. A survey experiment conducted without the surrounding rhetoric, enforcement activity, or data-sharing policies of the production census may underestimate how respondents will react. Researchers should therefore combine quantitative testing with interviews, focus groups, community consultations, and evidence from other federal data collections.

Any analysis should report differential effects clearly. A small national change should not be treated as harmless if it produces much larger declines among mixed-status families, particular racial or ethnic communities, or places with existing undercount risks.

Document How Community Partnerships Affected 2020 Participation

The federal government and philanthropy invested heavily in GOTC infrastructure for the 2020 Census, but much of its value remains difficult to see in conventional operational reports. Census and independent researchers should document how community partnerships affected participation, corrected misinformation, identified operational failures, and helped residents navigate the questionnaire.

The research should examine more than whether a county with an active Complete Count Committee achieved a higher response rate. It should investigate what partners actually did and which functions were most important. Community organizations may have translated official information, identified missing housing units, helped residents access online response, or explained confidentiality protections. They may also have alerted Census to scams, inaccessible systems, or fear spreading through a particular community.

Researchers should pay particular attention to communities where participation might otherwise have been substantially lower. The contribution of a trusted organization may not appear as a dramatic increase over a national benchmark. Its effect may be that response remained stable despite the pandemic, political conflict, or fear created by the proposed citizenship question.

Documentation should also identify the conditions that made partnership effective. Organizations need timely information, usable data, translated materials, access to Census staff, and resources to support their work. Research should distinguish genuine partnership from arrangements in which Census asks organizations to distribute a predetermined message without giving them meaningful influence over the policies affecting their communities.

This evidence would help Census treat partnership infrastructure as an operational investment rather than an optional communications activity. It would also help funders and advocates determine which relationships must be maintained between census cycles rather than rebuilt shortly before enumeration.

Develop a Coalition Position on Minimum Census Content

Advocates should develop a shared position on the minimum content necessary for a useful and legitimate decennial census. Waiting until a specific rollback is proposed may force organizations to negotiate basic principles under severe time pressure.

A coalition position should address both the questions asked and the data ultimately produced. At a minimum, it could establish expectations that Census:

• continue collecting race and ethnicity directly;

• implement scientifically tested standards that allow multiple and detailed identities;

• preserve information needed to identify same-sex couples and varied household relationships;

• avoid politically motivated additions that are likely to impair participation; and

• publish enough detail for civil rights, redistricting, program administration, and community use.

The coalition will need to decide where it can accept variation and where a change would cross a red line. For example, organizations may agree that returning to separate race and ethnicity questions would be harmful while differing about whether it should trigger withdrawal from GOTC work. Developing those positions in advance would allow partners to respond more coherently if OMB or Commerce announces a change.

The coalition should include groups that use the data and groups that conduct GOTC work. Researchers and civil rights lawyers may focus on the quality of the resulting datasets. Community organizations may be better positioned to explain how a content decision will affect trust and participation. Tribal governments, racial and ethnic communities, immigrant-serving organizations, LGBTQ organizations, and groups representing people with disabilities or complex household arrangements should have meaningful roles in defining the position.

A coalition statement should not imply that every organization must adopt the same communications strategy. Some groups may continue encouraging participation under almost any questionnaire because they view the population count as essential. Others may place greater weight on the risks to their communities. The shared position can define policy goals while allowing organizations to make their own judgments about how to engage residents.

Seek Early Transparency on Content Testing and Approval

Advocates should seek information about questionnaire development before the statutory reports to Congress in 2027 and 2028. By the time proposed subjects and questions are formally transmitted, major testing and approval processes may already be well advanced.

Census and Commerce should disclose:

• which content changes are under consideration;

• who requested or directed the change;

• what legal and statistical purpose it is intended to serve;

• what testing has been conducted;

• what Census subject-matter experts recommend;

• how the change would affect participation and operations; and

• what role OMB and political leadership will play in final approval.

Testing materials should include the exact question wording, response options, instructions, and processing rules. A question cannot be evaluated fully without understanding how Census will edit, code, combine, and publish the responses.

Transparency should also extend to proposals that Census rejects. If political officials request a question that subject-matter experts conclude would harm accuracy, the public should know that the proposal was considered and why it was not adopted. Similarly, if leadership overrides technical recommendations, the competing analyses should be preserved in the administrative record.

Advocates should monitor sources beyond the main Census planning documents. Changes may appear in OMB standards, Federal Register notices, information-collection submissions, test questionnaires, procurement materials, budget documents, or Commerce directives. Congressional oversight can help require earlier and more complete disclosure where the agencies do not provide it voluntarily.

Prepare Legal, Congressional, and Communications Strategies Early

A controversial content decision could move quickly and require several forms of response at once. Advocates should prepare legal, congressional, and public communications strategies before the final questionnaire is announced.

Legal preparation should identify the claims and evidence that might be relevant under the Enumeration Clause, Census Act, Administrative Procedure Act, equal protection principles, and other applicable laws. The strongest challenge may depend less on whether Census has abstract authority to ask a question than on whether the decision was supported by a genuine rationale, appropriate testing, and adequate consideration of the effects on accuracy.

Organizations should also consider standing, potential plaintiffs, venue, timing, and the remedy they would seek. A lawsuit filed too late could leave the court with few practical options, while a premature challenge may be dismissed because the final decision has not been made. Preserving technical evidence and the decision-making record will be essential.

Congressional strategies should include oversight of content development, requests for underlying research, appropriations restrictions, and questions for Commerce and Census leadership. Members of Congress can press for explanations before a decision is final and help ensure that career experts are able to provide their analysis without political interference.

Communications planning should be equally proactive. Advocates will need to explain why questionnaire content affects the accuracy of the entire census, not merely the availability of one statistic. Messages should connect content decisions to self-response, household omissions, partner participation, field operations, and reliance on administrative records.

Communications should also avoid placing responsibility on frightened communities. If residents hesitate to respond because the government adds a question associated with enforcement, the resulting undercount should not be described simply as a failure of public cooperation. Advocates should explain how government choices shape whether participation appears safe, worthwhile, and credible.

Coalitions should prepare different messages for different outcomes. The response to a return to the 1997 race and ethnicity standards may differ from the response to eliminating those questions altogether. A citizenship proposal may require both national litigation messaging and highly practical information for local organizations deciding what to tell people in their communities.

The goal is not to assume the worst outcome or treat every planning signal as a final decision. It is to avoid beginning the response only after Census has invested years in a design and statutory deadlines are approaching.

Research and advocacy should ultimately reinforce one another. Research can identify which content changes create the greatest risks and which protections reduce them. Community organizations can explain effects that formal testing may miss. Legal and congressional work can create pressure for transparency, while clear communications can help the public understand why these decisions matter.

The central advocacy principle is that census content cannot be separated from census accuracy. Questions and processing rules determine what information is produced, but they also shape whether residents trust the census and whether community organizations are willing to support it. Advocates should evaluate every proposed change against both standards: whether the content produces useful and scientifically valid data, and whether it strengthens or weakens the public participation on which the count depends.

****