XX. Funding, Contracting, and Field Capacity

The 2030 Census is being designed to cost less than a census that simply repeated older methods at larger scale. Census plans to use more administrative data, automation, enterprise systems, virtual offices, and shared contracts while reducing some labor-intensive field activities.

There are good reasons for that strategy. The 2020 Census demonstrated that automation can improve field productivity as well as reduce costs, and reusable systems can avoid rebuilding the same capabilities every decade. GAO concluded that several 2020 innovations improved implementation and helped slow the long-term growth in census cost.

But cost savings and count quality are not interchangeable measures of success. Congress and advocates should be able to see where Census expects savings to come from and what evidence shows that each reduction will preserve or improve the count.

A. Cost Savings Versus Count Quality

Several major elements of the 2030 design have both an efficiency rationale and a potential quality consequence.

Less fieldwork. Expanded administrative enumeration could substantially reduce the number of households that Census visits. GAO has noted that the planned increase in administrative-record use could greatly reduce the field workload.

That can be a genuine improvement when high-quality records resolve households that otherwise would require repeated unsuccessful visits or low-quality proxy responses. It becomes more concerning if field effort is reduced primarily to meet a cost target and administrative records become the cheaper substitute for obtaining information directly.

Census should therefore report savings from In-Office Enumeration alongside quality measures: how many visits were avoided, how accurately records identified household members and their location, which characteristics could be obtained reliably, and how performance varied among populations.

Fewer physical offices. Census says it intends to continue the reduction in field offices begun in 2020 by centralizing administrative functions and replacing many physical offices with virtual operations. The Bureau expects this to reduce costs and increase flexibility.

That may make sense for payroll, administrative support, and other functions that no longer require workers to gather in a physical office. But Census should separately evaluate whether fewer local management locations affect recruitment, supervision, knowledge of local conditions, equipment support, or the ability to respond quickly when a field operation is struggling.

More automation. Automation will touch almost every part of the census, including workload assignment, matching, quality monitoring, recruiting, training, and telephone assistance. For example, the Operational Plan says Census intends to use Interactive Voice Response technology to answer routine telephone questions and thereby reduce the number of call-center staff required. It also plans a smaller physical call-center footprint, potentially supported by employees working from home.

Those changes may make operations substantially more efficient. The quality question is what happens when automation fails to resolve the difficult cases. Census should preserve enough human capacity for unusual households, language needs, accessibility requests, complex group quarters, and other situations that do not follow the standard automated path.

More enterprise contracting and shared systems. The 2030 Acquisition Strategy explicitly calls for using Census Bureau and Commerce-wide contract vehicles instead of decennial-specific contracts where possible. The current CenTAM acquisition, for example, is designed as a Bureau-wide application-modernization framework that includes both enterprise and decennial components.

Shared systems can lower development costs and create capabilities that improve between censuses. They can also make it harder for an outside observer to determine exactly what Census is buying for the decennial census, how much it costs, and whether the decennial program receives the capacity it needs on its schedule.

Smaller or more centralized support workforces. Automated recruiting, payroll, training, call-center functions, and administrative support may reduce the number of people needed for some activities. The Operational Plan describes the overall goal as an “appropriately sized” workforce rather than simply a smaller one.

That distinction should be maintained. The correct staffing level is the one needed to conduct an accurate census under realistic conditions, including conditions in which self-response is lower or fieldwork is harder than expected.

Central question

When Census identifies an “efficiency,” does it remove unnecessary work, or does it reduce the effort available for cases that are expensive precisely because they are difficult to count?

B. Procurement Schedule

Major census contracts must be in place early enough to support development, integration, the 2028 Dress Rehearsal, and final production. The 2030 Acquisition Strategy specifically identifies early acquisition planning and sufficient procurement lead time as lessons from 2020.

A comparison with the 2020 cycle provides some context:

Capability	2030 status as of August 2026	Approximate 2020 timing
Census Questionnaire Assistance	Market research and industry engagement underway; latest industry day held in May 2026	Major CQA contract awarded July 2016
Application modernization / integration	CenTAM draft solicitation materials released in April 2026; industry day held July 2026	2020 Technical Integration contract awarded August 2016
Recruiting, hiring, and training	DARHTS RFQ issued June 2026	C-SHaRPS recruiting/HR system awarded November 2016
Mobile devices and field IT	D-X market engagement underway beginning in 2025–26	Decennial Device as a Service procurement moved through solicitation in late 2016 and was awarded June 2017
Integrated communications	No dedicated communications acquisition currently appears on the public major-procurements page	2020 Integrated Communications Contract awarded August 2016

Current 2030 statuses come from Census’s procurement page; the 2020 milestones are documented in Census’s contemporaneous program-management materials.

The comparison does not show that 2030 procurement is generally behind schedule. Several acquisitions appear to be moving at roughly comparable points in the decade, and the D-X effort began market engagement earlier than the eventual 2020 device contract award.

It is also not an exact comparison. The 2030 strategy intentionally relies more heavily on enterprise contract vehicles and reusable Census capabilities, so a single 2020 contract may be divided among several 2030 acquisitions or incorporated into a broader enterprise system.

The communications procurement remains the most obvious item to monitor, as discussed in the previous chapter. More broadly, Census should publish enough milestone information for Congress and the public to identify which acquisitions are on the critical path for the 2028 Dress Rehearsal and what contingency exists if an award is delayed by a protest or other procurement problem.

Issue to monitor

The important question is not whether every 2030 contract was awarded on exactly the same date as its 2020 counterpart. It is whether each capability will be available early enough for realistic integration and testing before production.

C. Enterprise Contracts

The growing use of enterprise systems creates a funding question that is easy to miss in a decennial budget.

Census plans to rely on capabilities that serve multiple programs across the Bureau. Its acquisition strategy specifically encourages use of enterprise and department-level contract vehicles, and the Operational Plan calls for memoranda of agreement to define the products and services that other Census components will provide to the 2030 program.

That can be efficient. If Census needs a system for data collection, storage, matching, or application development throughout the decade, there is little reason to build a second version solely for 2030.

The use of an enterprise contract vehicle does not, by itself, mean that decennial appropriations are paying for unrelated Census programs. A Bureau-wide vehicle can contain separately funded task orders, including work dedicated specifically to the decennial census.

But as the distinction between “decennial infrastructure” and “Census enterprise infrastructure” becomes less clear, cost allocation becomes an important oversight issue.

Congress should be able to determine:

• how much decennial funding is being spent on shared systems;

• what portion reflects capabilities specifically required by the 2030 Census;

• which other Census programs contribute to development and operating costs;

• what service levels the decennial program is guaranteed in return;

• who pays for long-term maintenance after peak census operations; and

• whether cost savings attributed to 2030 result partly from costs appearing elsewhere in the Census budget.

The reverse concern also matters. Congress should know whether a reduction in funding for a broader enterprise modernization program could leave the decennial census without a capability it assumes will be available.

Census already plans to use internal memoranda of agreement to ensure that intra-agency services meet 2030 requirements. Those agreements, together with the Life Cycle Cost Estimate, should make the financial and operational dependencies visible enough for meaningful oversight. The Operational Plan says the Bureau will continuously update the official Life Cycle Cost Estimate and use it to formulate and execute annual budgets.

Advocacy priority

Enterprise reuse should produce genuine savings, not make decennial costs and dependencies harder to see.

D. Workforce Assumptions

Even with substantially more automation and administrative data, Census will still need a very large temporary workforce.

The scale of that challenge should not be underestimated. For 2020, Census received more than 3 million applications and ultimately hired more than 523,000 employees across temporary field and support positions. The Bureau’s 2020 assessment recommended further integration of recruiting, hiring, payroll, training, and background-check systems for 2030.

The 2030 Operational Plan says staffing models will determine workforce and recruiting needs based on expected workloads for In-Field Enumeration, Group Quarters Enumeration, Coverage Estimation, and other operations. It also explicitly recognizes that recruiting and training qualified field workers affects census quality.

Headcount alone, however, is not enough. Census needs people with the right capabilities in the places where they are needed.

Local Knowledge

Local workers can recognize unusual address conventions, informal housing, locally known group quarters, changing neighborhoods, and places that may be difficult for an unfamiliar enumerator to find.

As Census reduces its physical office footprint and centralizes management, it should ensure that it does not also lose local knowledge. Recruiting goals should therefore be evaluated geographically, not simply nationally.

A national surplus of applicants does little to solve a shortage in a remote rural county, tribal area, high-cost metropolitan community, or neighborhood where Census has difficulty recruiting.

Language Skills

Census needs enough bilingual workers to conduct field interviews and enough multilingual staff to support telephone and specialized operations.

Recruiting those workers cannot be assumed. During preparation for the 2018 Census Test, for example, Census reported difficulty filling some non-Spanish language CQA positions, including Korean, Vietnamese, Mandarin, and Cantonese.

Workforce planning should therefore identify language needs early, set language-specific recruiting goals, and track whether those goals are being met before fieldwork begins.

If the 2030 Language Program is narrowed because of the Commerce English-language policy, Census should not treat that policy change as evidence that the underlying need for multilingual workers has disappeared.

Specialized Group-Quarters Experience

Group quarters operations cover institutions and living arrangements that require very different forms of expertise, including correctional facilities, college housing, nursing facilities, shelters, and outdoor locations used by people experiencing homelessness.

The 2020 Census counted approximately 8.4 million people through Group Quarters Enumeration and Service-Based Enumeration across more than 184,000 group quarters.

Census should identify which operations require specialized training, experienced supervisors, or staff who understand the institutions and populations involved. General field enumeration training should not automatically be considered adequate for every living situation simply because the 2030 workforce is increasingly cross-trained.

Technical Capacity

The field workforce will also operate in a more automated census. Workers must use mobile devices, follow dynamically assigned workloads, interpret system-generated instructions, complete electronic training, and respond appropriately when the automated process does not fit the situation in front of them.

Technology can reduce administrative burdens, but Census should test whether training prepares workers to understand the underlying rules rather than simply follow prompts. An enumerator needs to recognize when a system-generated result appears wrong and know how to escalate the problem.

Community Trust

Finally, Census needs workers whom residents will allow through the door.

Local residence does not guarantee trust, and a technically skilled worker is not automatically an effective interviewer. Recruitment and training should consider whether workers can communicate respectfully with the communities they serve, explain confidentiality and response requirements accurately, and recognize when distrust or fear requires a different approach.

This is particularly important if immigration enforcement, data-use controversies, language policies, or other government actions make some households more reluctant to engage with federal workers.

Central workforce question

Census should report not only whether it has enough applicants and employees nationally, but whether it has enough qualified workers with the right language, local, technical, and specialized skills in the places where fieldwork will be hardest.

E. Congressional Oversight

Many of these issues will be substantially harder to fix after the 2028 Dress Rehearsal. Congressional oversight should therefore focus on the assumptions underneath the budget while important design and acquisition decisions remain open.

Appropriations and oversight committees should ask:

• What does the census actually cost? Is the Life Cycle Cost Estimate current, independently reviewable, and explicit about uncertainty and contingency? GAO reported in 2025 that Census was still developing the first 2030 life-cycle estimate and emphasized budget uncertainty, cost, schedule, risk, data quality, privacy, and cybersecurity as continuing oversight issues.

• Where do projected savings come from? How much is attributed to fewer field visits, fewer offices, administrative records, automation, shared IT, staffing reductions, or other changes, and what quality evidence supports each assumption?

• What happens if the assumptions are wrong? Does Census have enough contingency funding and contractual flexibility if self-response is lower, administrative data perform worse, recruiting is harder, or a major system requires redesign?

• Are critical acquisitions on schedule? Which contracts are necessary for the 2028 Dress Rehearsal, when must they be awarded, and what contingency exists for protests, contractor failure, or schedule slippage?

• How are enterprise costs allocated? Which broader Census systems depend on decennial funding, what share of their cost is charged to the 2030 Census, and what service commitments protect the decennial program?

• Is field capacity being preserved? What workforce assumptions underlie the budget, and do staffing models separately account for areas requiring language skills, local knowledge, specialized enumeration, or unusually intensive field effort?

• What functions are vulnerable to annual funding shortfalls? Which research, testing, partnerships, communications, language services, or infrastructure investments would be delayed first if appropriations fall below the requested level?

Congress should also ask Census to identify decisions that become expensive or impossible to reverse in each upcoming fiscal year. That would allow appropriators to see when a short-term funding reduction could create a much larger operational problem later.

GAO’s experience with the 2020 Census provides an important warning. The Bureau initially expected substantial savings from its redesigned census, but its projected life-cycle cost increased from $12.5 billion in 2015 to $15.6 billion in 2017 as design and risk assumptions matured. GAO repeatedly emphasized the need for reliable, well-documented cost estimates so Congress could make informed funding decisions.

That history does not mean 2030 will experience the same cost growth. It does mean that Congress should be skeptical of false choices between economy and accuracy. Underfunding preparation can produce higher costs later if Census must rush procurements, abandon testing, rebuild systems, or compensate for weak self-response with emergency fieldwork.

Advocacy priority

Congressional oversight should connect dollars to operational consequences. When Census proposes a savings or Congress proposes a reduction, the relevant question is what changes in the census design, capacity, or quality as a result.

Why advocates should care

Budget and contracting decisions can quietly settle substantive census policy. A field visit that is never funded, a bilingual worker who is never recruited, or a system capability omitted from a contract cannot easily be restored in the final months before Census Day.

Where to Look: See section 4.1.1, “Program Management,” pp. 68–71, and section 4.3.1, “Field Support,” pp. 77–79, of the 2030 Census Operational Plan for cost estimation, budget management, workforce planning, field offices, recruiting, and training. Section 3.2.11 on Census Questionnaire Assistance describes planned call-center efficiencies.

The 2030 Census Acquisition Strategy and 2030 Census Procurements webpage are the best sources for contracting strategy and current acquisition activity, including CenTAM, CQA, DARHTS, D-X, and background-check services.

For the prior-decade baseline, see the 2020 Field Infrastructure Operational Assessments and GAO’s ongoing 2030 Census oversight reports, which provide useful comparisons on workforce scale, cost, acquisition, testing, and program risk.

****