X. What Happens to Field Enumeration

In 2020, Nonresponse Follow-up was the main mechanism for completing the count for households that did not respond on their own. Census enumerators visited addresses that had not responded, determined whether the housing unit was occupied, and attempted to interview a resident. Separate operations delivered census materials in areas without ordinary mail delivery and enumerated some remote communities.

For 2030, Census has combined much of this work into a single In-Field Enumeration operational area. The plan integrates the functions previously performed through Nonresponse Follow-up, Update Leave, and Update Enumerate. It also contemplates cross-trained workers who can handle housing units, group quarters, and transitory locations, along with real-time case assignment and address updating.

This integration could make fieldwork more flexible and allow workers to resolve a problem when they encounter it. But Census is also explicitly seeking to reduce the number of visits and the size of field workloads. The design of the contact strategy, workforce, offices, and training program will determine whether those efficiencies eliminate unnecessary work or eliminate meaningful opportunities for direct response.

A. Contact Strategy

The field contact strategy will determine how many visits a nonresponding household receives, when those visits occur, and what happens after an attempt fails. The Operational Plan says the strategy will establish the number and type of visits and seek to resolve cases using an “optimal amount of fieldwork.” Census also states that reducing contacts can produce smaller workloads and lower costs.

The meaning of “optimal” is therefore central. A strategy optimized primarily for cost may look very different from one designed to maximize direct responses or reduce differential undercounts.

A household may not respond to the first visit for many reasons. No one may be home during working hours. The resident who answers may not know enough about the household. A bilingual worker may be needed. A person may want to verify the worker’s identity before participating. A building may require several attempts before a fieldworker gains access.

The number of attempts alone will not show whether Census made a meaningful effort to achieve a census response. Timing and form matter. Three weekday visits at similar hours may provide less opportunity than one evening visit, one weekend visit, and a telephone or paper option. A notice left at the door may encourage self-response but is not equivalent to speaking with a resident.

The current plan contains two formulations that require clarification. The In-Field Enumeration section says administrative and supplemental data may remove a case from further follow-up after at least one in-person visit. The In-Office Enumeration section says administrative enumeration will occur after “multiple contact attempts.” The plan does not yet explain what counts as an attempt, whether all households receive an in-person opportunity to respond, or how many attempts must occur before Census considers data collection complete.

These rules will determine whether in-office enumeration is truly a last resort. If several mailings, a telephone contact, and well-timed field visits precede the use of administrative records, the system may preserve a strong preference for direct response. If one unsuccessful visit satisfies the field requirement and the other “attempts” consist primarily of automated mailings, administrative enumeration could become a routine substitute for household contact.

Models will also help determine when a worker visits, when a proxy such as a neighbor is contacted, and when available records are considered sufficient. Census plans to use information about response propensity, vacancy, household characteristics, and administrative-record quality to shape those decisions.

Tailoring can improve fieldwork. A household likely to include young children may warrant additional direct attempts before a proxy is used. An address with conflicting occupancy information may need an early visit. A household likely to respond in the evening should not receive every visit during the day.

But tailoring can also create unequal levels of effort. Communities predicted to be difficult or expensive to enumerate could receive fewer visits, earlier proxy contacts, or greater reliance on administrative records. Models trained on prior response patterns may treat the effects of language barriers, housing instability, or distrust as reasons to reduce contact rather than reasons to improve it.

Census should publish the basic field contact rules before the 2028 Dress Rehearsal. Those rules should explain:

  • the minimum opportunities for resident contact;

  • which contacts count toward administrative enumeration thresholds;

  • how visits will be distributed across days and times;

  • when fieldworkers may seek a proxy;

  • when a case may be removed based on administrative records; and

  • how the strategy changes for households likely to contain people at higher risk of omission.

Census should evaluate the strategy by enumeration method and population, not only by cost or case-completion speed. Important measures include the share of occupied households completed through a resident interview, a proxy, or administrative records; the number and timing of attempts preceding each outcome; and the completeness of household rosters and characteristics.

Open design question

What amount of personal contact will Census require before a household can be counted through a proxy or administrative records?

Why advocates should care

In-office enumeration can be described as a last resort while still becoming common if the contact strategy defines one brief or poorly timed visit as a sufficient field effort.

B. Combined and Cross-Trained Field Operations

The 2030 plan combines several kinds of household fieldwork and contemplates “multi-operational” enumerators. Some workers may be trained to enumerate housing units, group quarters, and transitory locations. A worker who discovers that an address has been misclassified could move it into the correct workload and use the appropriate questionnaire instead of leaving the problem for another operation.

This could resolve mistakes more quickly. A worker sent to a listed house might discover several apartments, a residential program, or people living in recreational vehicles behind the property. If the worker can update the address, classify the living quarters correctly, and continue the enumeration, the household may not need to wait for a separate team.

Cross-training could also give Census greater flexibility. Workers could be shifted among workloads as local needs change rather than remaining idle while another operation faces shortages. In rural or remote areas, a single worker who understands several operations may be more practical than deploying separate specialists over long distances.

The Full Listing and Enumeration Integration research project similarly examines whether address listing and enumeration can be performed through the same system and training. Census is explicitly studying the benefits and challenges of training field staff to perform both functions.

But breadth can come at the expense of depth. Enumerating an ordinary household is different from working in a college residence hall, correctional facility, shelter, marina, or outdoor location. Each setting has distinct residence rules, privacy concerns, contact procedures, and sources of information.

A broadly trained worker may know how to open the correct questionnaire without understanding the circumstances of the population. A fieldworker who recognizes that a building is a group home may still need specialized guidance on whether residents should respond individually, whether an administrator may provide records, and how to protect privacy.

The risk is greatest if cross-training is used to justify shorter training or fewer specialized teams. Operational integration should not mean that every worker is expected to master every population through one general online course.

Census should use a layered model:

  • all fieldworkers receive a strong foundation in residence rules, confidentiality, address correction, disability access, and language assistance;

  • workers receive additional modules for the operations they are likely to encounter; and

  • particularly sensitive settings retain specialized staff, supervisors, or rapid access to expert support.

Assignment systems should take skills into account. A worker who speaks the relevant language or understands local addressing conventions should be prioritized where those skills matter. Group quarters, shelters, tribal areas, and remote communities should not receive whichever generalist happens to be geographically closest.

Census should also publish performance measures separately for each type of work. A combined operation can appear efficient overall even if results deteriorate for one small but important population. Completion rates, rework, classification errors, missing characteristics, and proxy use should be reported by living-quarters type and field activity.

Potential benefit

Cross-trained workers could correct misclassification, add previously unknown locations, and resolve cases without waiting for another operation or team.

Credible risk

Training becomes so broad that workers lack the specialized knowledge needed for group quarters, shelters, transitory locations, tribal areas, and other complex settings.

Why advocates should care

Simplifying the organizational chart does not necessarily simplify the work. Census should not use operational integration to obscure reduced field effort or weaker attention to specialized populations.

C. USPS Personnel as Enumerators

Census is testing whether employees of the U.S. Postal Service can help collect responses from households that do not respond on their own. The 2026 Census Test uses two different models.

In Huntsville, postal workers are hired separately as Census Bureau employees and conduct census work outside their postal hours. They identify themselves as Census workers rather than as postal employees. In Spartanburg, postal carriers perform census duties while working their ordinary routes as USPS employees. They may wear postal uniforms but are specially sworn to protect Census information. Census says participating workers receive background checks and census-specific training.

Legal Status and Title 13

Title 13 allows the Secretary of Commerce to use temporary staff, including employees of other federal agencies, to perform Census work. Those individuals must be sworn to observe the confidentiality requirements of section 9.

Special Sworn Status therefore provides a legal mechanism for postal employees to access protected information. Once sworn, they are subject to the same obligation to protect identifiable census responses and the same penalties for unauthorized disclosure.

That authority does not resolve every legal and operational question. An earlier postal-carrier enumeration pilot planned for the 2018 Census Test was canceled after Census and USPS identified conflicts among Title 13, postal law, and agency regulations that they could not reconcile. The 2026 arrangements appear designed differently, but Census should explain publicly how the earlier barriers were resolved and what agreements govern supervision, data access, equipment, and the separation of postal and census records.

Special Sworn Status should be a starting point, not the entire confidentiality analysis. Census should establish that:

  • response information remains only in Census-controlled systems;

  • postal route records do not indicate who responded or what was reported;

  • USPS supervisors cannot access household answers;

  • devices and cases are controlled through Census security rules; and

  • confidentiality duties continue after the pilot or employment ends.

Public Understanding and Trust

Postal workers may bring valuable local knowledge. Carriers often understand which buildings contain several units, where addresses are difficult to locate, and when residents are usually present. Some households may be more comfortable speaking with a familiar carrier than with a temporary worker they have never seen.

But the model may also confuse the public. Residents ordinarily understand mail delivery and census interviewing as different government functions. A carrier who arrives in uniform and begins asking for names, ages, relationships, race, and ethnicity may appear to be collecting information for USPS rather than for Census.

Census should explain the arrangement well before visits begin. Materials should tell residents when a postal worker may conduct an interview, what identification the worker will carry, how to verify the worker, and whether the information is accessible to USPS. The worker should give a short, consistent explanation before requesting any response.

The two test models may produce different reactions. In Huntsville, a postal employee working as a separate Census employee may not be recognizable as the household’s carrier. In Spartanburg, the carrier may be familiar but the overlap between mail delivery and enumeration is more visible. Census should evaluate trust, response quality, refusal, and respondent understanding separately for each model.

Familiarity will not reassure everyone. Some residents avoid contact with federal personnel because of immigration concerns, experiences with law enforcement, fear of benefit consequences, or distrust of government data practices. A uniformed carrier may be seen as more trustworthy by one household and as evidence of greater government coordination by another.

Testing should therefore examine differential effects, not only overall productivity. Census should ask whether postal enumeration changes participation among immigrants, people with disabilities, rural households, people living in informal units, and others who may interact differently with postal workers.

Group Quarters and Shelters

The current 2026 pilot exclusively includes visits to nonresponding housing units. Census has not publicly proposed using postal personnel to enumerate group quarters, shelters, outdoor locations, or other special populations.

Any expansion beyond ordinary households would require a separate review. Postal familiarity with an address does not provide the training needed to enumerate a correctional institution, college residence hall, intimate partner violence shelter, or outdoor encampment. Assignments could also reveal the location or status of confidential facilities.

Census should clearly affirm that confidential shelter information will not enter ordinary postal workloads and that postal employees will not be used in settings requiring specialized group quarters or homelessness protocols unless a distinct, publicly reviewed design establishes that the arrangement is safe and appropriate.

Training and Supervision

Postal workers performing census duties need more than a basic explanation of the questionnaire. They must understand residence rules, complex households, hidden housing units, confidentiality, language and disability access, proxy procedures, and how to respond when a resident is fearful or suspicious.

The two employment models also raise different supervision questions. A carrier working during a postal route remains embedded in USPS’s daily management structure, while Census must retain authority over enumeration quality and confidentiality. Census should identify who reviews cases, provides corrective training, investigates complaints, and decides whether a worker may continue census duties.

The pilot should be evaluated against ordinary Census fieldworkers on:

  • resident and proxy interview rates;

  • roster and characteristic completeness;

  • refusals and break-offs;

  • address and unit discoveries;

  • confidentiality or security incidents;

  • time required per completed case; and

  • respondent understanding of the worker’s role.

Potential benefit

Postal workers may provide local knowledge, an existing workforce, and trusted relationships that improve access to some nonresponding households.

Credible risk

Combining postal and census roles confuses residents, weakens perceived confidentiality, or places complex enumeration work under a workforce and supervision structure not designed for it.

Advocacy priority

Census should publish the legal agreements, privacy safeguards, training standards, evaluation measures, and results of both 2026 pilot models before considering nationwide implementation.

D. Workforce and Office Footprint

A successful census requires a large temporary workforce distributed across the country. Workers must be recruited, screened, trained, supervised, paid, equipped, and supported during a short and demanding collection period.

Census plans to reduce its physical field-office footprint for 2030. The Operational Plan anticipates greater use of virtual offices, centralized administrative and technical support centers, work-at-home options for regional staff, and direct shipment of equipment to fieldworkers. Physical offices would include regional census centers, centralized Census Administrative Technical Support centers, and additional field space where needed.

Virtual and centralized support can offer real advantages. Census may need fewer leases, workers may avoid long trips to an office, and specialized payroll or technical staff may serve several regions. Remote work could also broaden the pool of administrative employees and allow Census to adjust staffing more quickly.

But field offices serve functions beyond processing forms and storing equipment. They create a local center of responsibility. Workers can meet supervisors, replace failed devices, obtain supplies, practice difficult procedures, and discuss unusual cases. Community partners, elected officials, and congressional offices know where to raise an urgent local problem.

A smaller footprint could make the field structure less visible and less accessible. A community organization that discovers an unlisted shelter or a serious language problem may have no local manager to contact. A fieldworker struggling with an unusual household may rely on a distant service desk that does not understand local geography or housing practices.

Census should publish the proposed office structure, the territory and workloads assigned to each office, and the methods communities will use to reach responsible managers. Every area should have a clearly identified escalation pathway, even where the office itself is virtual.

Recruitment and Retention

Fieldwork depends on recruiting people who are available where and when the work occurs. In 2020, Census sought to hire workers from their own neighborhoods and communities, including bilingual workers, so they could reach residents at appropriate hours and draw on local knowledge. The Bureau ultimately hired more than half a million temporary employees.

That local recruitment principle should remain central. A national applicant pool does not ensure that Census has workers in remote areas, tribal lands, neighborhoods with limited English proficiency, or places where travel is difficult.

Pay, mileage reimbursement, reliable scheduling, timely hiring, and responsive supervision will affect retention. A worker who waits months for a background check or encounters payroll problems may accept other employment before fieldwork begins. Workers who receive inconsistent assignments or cannot reach a supervisor may leave during the operation.

Virtual hiring and onboarding can speed the process, but they can also exclude otherwise qualified applicants who have limited broadband, lack a suitable device, or need help completing an online process. Census should retain telephone and in-person assistance and work with community partners to recruit locally.

Language Skills and Local Knowledge

Language capacity should be part of workforce planning, not an informal benefit Census hopes to obtain from the available applicant pool. Staffing models should estimate the need for bilingual workers by geography and operation, and Census should publish whether it met those goals.

Hiring a worker who speaks a language is not enough. Census must assess whether the person can explain census concepts and residence rules accurately in that language. At the same time, testing requirements should not become so rigid or English-centered that they exclude people with valuable community and language skills.

Local knowledge matters beyond language. Workers may understand informal address descriptions, seasonal movement, shared housing, tribal geography, or the trusted way to approach a particular community. Real-time assignment systems should use that knowledge rather than repeatedly sending workers into unfamiliar areas because an algorithm finds the route more efficient.

Public and Congressional Access

Local offices have also provided an identifiable point of contact for congressional staff, local governments, and advocates. A reduced office footprint should not eliminate that access.

Census should designate public-facing managers for every state and major field area. Their contact information, responsibilities, and escalation procedures should be available before field operations begin. Virtual offices should be operationally virtual, not publicly unreachable.

Potential benefit

Virtual offices and centralized support could reduce administrative costs, expand remote staffing, and allow Census to move resources more flexibly.

Credible risk

Fewer local offices weaken recruitment, supervision, troubleshooting, community relationships, and access to managers precisely when complex field cases require more local judgment.

Why advocates should care

A national field system still needs local accountability. Communities should know who is responsible when the census is not working on the ground.

E. Training, Badges, and Background Checks

Census plans to streamline hiring and onboarding for 2030. The Operational Plan says the Bureau intends to reduce the quantity and types of background checks and badges required. It also describes virtual training, automated remedial training, and greater use of integrated personnel systems.

These changes affect both access to census jobs and the quality and credibility of fieldwork.

Training

Virtual training can provide consistent content across the country and allow workers to review materials at their own pace. It can also reduce travel and make onboarding easier for people who live far from an office.

But field enumeration is not learned entirely through recorded modules. Workers must practice introducing themselves, addressing privacy concerns, applying residence rules, identifying hidden units, using an interpreter, and responding to confusion or fear. Supervisors need to observe whether a worker can perform these tasks before sending them into the field.

All-virtual training may be particularly difficult for workers with weak broadband, limited digital experience, disabilities, or little prior experience with survey interviewing. It may also reduce the informal exchange of local knowledge that occurs when workers and supervisors train together.

Census should combine standardized online instruction with live practice, proficiency checks, and operation-specific exercises. Workers assigned to group quarters, shelters, remote areas, tribal lands, or multilingual communities should receive additional training designed for those settings.

The 2030 Census Research Project Explorer includes a Cultural Competency Training project examining specialized content, local flexibility, and the relative value of in-person, live virtual, and asynchronous delivery. Those results should shape the operational training plan rather than remain a general workforce initiative.

Training should also continue after deployment. Automated systems can identify errors or unusual patterns, but remedial training should help workers improve rather than simply pressuring them to increase production. Supervisors should distinguish between a worker who needs instruction and a case that is inherently difficult to resolve.

Badges and Public Verification

An official badge is one of the clearest ways residents identify a legitimate census taker. In 2020 and the 2026 Test, Census told residents to look for a photo badge and Census-branded materials.

Simplifying badges may reduce administrative burden, but Census should preserve a nationally consistent form of identification that is difficult to counterfeit and easy to explain. Every fieldworker, including postal personnel and contractors where applicable, should carry identification that clearly states the capacity in which the person is conducting census work.

Residents should also have an independent way to verify a worker by telephone or through an official website. Verification should be available in multiple languages and should not require residents to provide personal household information.

The need is especially important if different kinds of workers appear in different clothing. A postal carrier wearing a USPS uniform, a Census employee carrying a Census bag, and a specialized group quarters worker should not each rely on a different verification system that the public must somehow understand.

Background Checks

Some form of screening is appropriate for Census employees who enter residential buildings and receive access to confidential cases and response data. But background checks are an imperfect tool. They identify certain events that appear in official records, not every person who might misuse their position. They can reduce some known risks, but they cannot establish that an applicant is trustworthy, capable, or likely to treat respondents appropriately.

The screening process can also create its own problems. Applicants in rural or remote areas may have to travel long distances for fingerprinting. Processing delays can cause Census to lose qualified candidates. Overly broad consideration of old, minor, or unrelated records may exclude people with valuable language skills and local knowledge without producing a meaningful safety benefit. Because arrests and convictions are also shaped by unequal enforcement, record-based screening may impose especially heavy barriers on some communities.

The 2020 workforce assessment recommended providing investigative staff and resources earlier to reduce background-check backlogs. The 2030 research agenda is considering shorter applications, reducing the time required for checks, fingerprinting and badging rules, and special hiring issues for workers with language skills and workers in remote areas.

Census should explain what it means to reduce the “quantity and types” of background checks. Eliminating duplicative requirements or tailoring reviews to the responsibilities of a particular position is different from abandoning reasonable screening. The Bureau should identify the specific risk each requirement is intended to address and assess whether the information collected is actually useful for that purpose.

Reviews should focus on conduct genuinely relevant to the job rather than treating any record as evidence that a person is unsuitable. Applicants should have a timely opportunity to correct inaccurate information, explain individual circumstances, and seek reconsideration of an adverse decision.

Background checks should also be understood as only one part of a broader system of protection. A person may pass every screening requirement and still mishandle confidential information, misunderstand residence rules, pressure respondents, or behave inappropriately. Strong training, limited access to data, effective supervision, clear complaince procedures, and prompt investigation of misconduct are more directly connected to how workers perform once they are in the field.

The goal should be a screening system that addresses identifiable risks without creating unnecessary barriers or providing false reassurance. Census should judge the system by whether it supports a capable, representative, and accountable workforce, not by how many applicants it excludes.

Potential benefit

Streamlined hiring, virtual training, and simpler onboarding could reduce delays and make census employment more accessible.

Credible risk

Census compresses training or weakens identification and screening at the same time that workers receive broader duties and more difficult cases.

Advocacy priority

Census should publish its training hours and formats, proficiency standards, specialized modules, badge design, verification system, and revised background-check requirements before peak hiring begins.

F. Worst-Case Scenario

Census reduces the number of physical field offices and centralizes administrative support. Hiring and training take place almost entirely online. Background checks and badges are simplified, equipment is shipped directly to workers, and supervisors manage large territories remotely.

At the same time, Census broadens fieldworker responsibilities. Generalist enumerators are expected to handle ordinary housing units, correct addresses, recognize group quarters and transitory locations, apply several different questionnaires, and resolve cases selected by automated systems.

The cases reaching them are more difficult than the average 2020 workload. Straightforward households have already responded or been removed through in-office processes. The remaining cases involve hidden housing units, complex families, weak administrative records, language barriers, unusual residence situations, and residents who need time and reassurance before participating.

Training covers each responsibility briefly but gives workers little opportunity to practice. Local offices are too distant or virtual, supervisors are difficult to reach, and specialized staff are unavailable when a worker encounters a shelter, informal subdivision, or complicated household.

The contact strategy emphasizes completion speed and fewer visits. Workers receive assignments generated in real time but have limited discretion to return at a better hour, request a bilingual colleague, or spend additional time resolving a case. After one unsuccessful visit, some households move quickly toward a proxy interview or administrative enumeration.

Recruitment and retention problems deepen the pressure. Workers leave because assignments are inconsistent, technical support is slow, or the job is more complex than the training suggested. The remaining staff cover larger territories and receive increasingly broad workloads.

National measures still show high productivity. Cases are closed, the office footprint costs less, and automated systems supply information where interviews are incomplete. But fewer people provide direct responses, fieldworkers miss unlisted living quarters, and difficult households are disproportionately counted through proxies or records.

Worst-case scenario

Cost-saving measures reduce local infrastructure, compress training, broaden job duties, and limit contact attempts at the same time that automation concentrates the most complex cases in the field. Census closes the workload but weakens the personal effort needed to count households that do not fit conventional systems.

Why advocates should care

Field enumeration is the final opportunity to correct errors created earlier in the census. If the workforce lacks time, training, local knowledge, and support, the people most poorly represented in automated systems will also receive the weakest human follow-up.

Where to Look

See the 2030 Census Operational Plan, particularly section 3.2.2, “In-Field Enumeration,” pp. 30–32, for the field contact strategy, integrated field operations, cross-trained workers, early and remote enumeration, address updates, and the use of administrative data to reduce follow-up. Section 3.2.3, “In-Office Enumeration,” pp. 32–34, explains how models and administrative data may inform contact attempts and determine when data collection is complete. Section 4.3.1, “Field Support,” pp. 77–79, covers recruiting, training, background checks, badges, virtual offices, centralized support, and the reduced physical footprint. Appendix B, pp. 86–90, discusses field confidentiality, Special Sworn Status, physical security, and protection of data on field devices.

The 2030 Census Research Project Explorer includes several projects that should shape this chapter. Case Assignment and Contact Strategy examines field-assignment rules, contact strategies, manual discretion, and training. Full Listing and Enumeration Integration, or FLEXI, studies whether listing and enumeration can be performed through one system and training program. Reimagining Regional Census Centers and Area Census Offices considers full-service offices, depots, virtual offices, telework, and a smaller physical footprint. Human Resources and Onboarding Policy Decisionscovers recruitment, background checks, fingerprinting, badging, training, language-worker waivers, retention incentives, and virtual hiring. Cultural Competency Training examines specialized content and the appropriate mix of in-person and virtual instruction.

The 2026 Census Test is the primary place to watch the USPS proposal. Census is testing two distinct models in Huntsville and Spartanburg, including postal workers hired separately by Census and USPS employees performing census duties during their routes under Special Sworn Status. Advocates should seek publication of the test design, interagency agreements, privacy and security procedures, training materials, respondent research, cost comparisons, quality results, and any incidents or operational problems. The earlier report on the canceled postal-carrier pilot for the 2018 Test provides important background on the legal and administrative issues Census and USPS previously encountered.

For the 2020 evidence base, see the operational assessments for Nonresponse Followup, Field Infrastructure: Recruiting, Onboarding, and Training, and Field Infrastructure: Field Office Administration and Payroll. These reports document the scale and distribution of the workforce, local hiring, training methods, background-check delays, field-office functions, payroll and administrative support, and recommendations for 2030. The assessments for In-Field Address Canvassing, Group Quarters Enumeration, and Enumeration at Transitory Locations provide additional evidence about the difficulty of combining or coordinating work across living-quarters types.

Future versions of the Operational Plan should be reviewed for the final minimum-contact rules, proxy thresholds, fieldworker discretion, office map, staffing model, training requirements, badge and verification procedures, and division of responsibility between Census and USPS. The 2028 Dress Rehearsal will be the most important opportunity to determine whether these components work together under conditions approaching a full census.