2020 Census: Actions Needed to Improve Census Bureau's Process for Working with Governments to Build Address List
Fast Facts
The Census Bureau needs an accurate address list to count everyone once, only once, and in the right place. To that end, the Bureau allows tribal, state, and local governments to review and offer updates to its master address list.
The Bureau received 5.1 million updates—more than it expected. As a result, it only reviewed a fraction of them in the office. This means that Census workers will have to visit more addresses in person—which could mean millions of dollars of additional fieldwork.
For the future 2030 Census, we recommended reviewing more of these updates in the office.
Census worker approaching a house
Highlights
What GAO Found
The Census Bureau generally followed the operational design for its Local Update of Census Addresses (LUCA) program, which is intended to give tribal, state, and local governments the ability to review and offer modifications to the Bureau's Master Address File (MAF). The Bureau met milestones, apart from extending the participation window for natural disaster-stricken areas, and generally followed plans for outreach, training, and participation options.
However, some decisions created additional fieldwork. The Bureau received more updates from participants than it expected, so it only reviewed roughly 860,000 of the 5.1 million updates that did not match to the MAF (see figure below). The rest will be added to potential fieldwork. Had more addresses been reviewed in-office, many may have been rejected, based on the rejection rate for reviewed addresses. Avoiding this unnecessary fieldwork could have saved the Bureau millions of dollars when following up with non-responding households.
Most LUCA Updates That Did Not Match the Census Bureau's (Bureau) Data Were Added to the Bureau's Address List for Address Canvassing and Enumeration
The Bureau has not reexamined LUCA with respect to the cost, quality, and public perception of the census since the program was authorized in 1994. Yet much has changed since then, from the tools the Bureau uses in building its address list to the provision of publicly accessible address data. As the Bureau turns to its strategic planning process for 2030, it will have several issues to address regarding the future of LUCA, including:
whether LUCA should continue to have a role in building the address list given the advent of other address-building initiatives;
how often to have governments review the MAF for the census, in light of the costs and benefits of administering such a program more frequently;
whether statutory nondisclosure protection of census address data is still needed given that address data sources and services are more prevalent.
Why GAO Did This Study
A complete address list is a cornerstone of the Bureau's effort to conduct an accurate census. LUCA is one of several operations the Bureau uses to produce its address list. It gives tribal, state, and local governments the opportunity to review the address list for their areas and provide the Bureau with any updates before the census.
GAO was asked to review the status of LUCA, including its effect on other operations, as well as LUCA's overall effectiveness and necessity. This report examines (1) LUCA's status and its likely effects on 2020 field operations, and (2) what considerations the Bureau and other stakeholders could use to reexamine LUCA for 2030. GAO reviewed Bureau plans, analyzed data from LUCA participation and the Bureau's review of submissions, and held 9 discussions on a possible reexamination of LUCA with relevant Bureau officials, a council representing participating governments, and census data subject matter specialists.
Recommendations
GAO is making eight recommendations to the Department of Commerce, including that the Bureau ensure more LUCA submissions are reviewed and reexamine LUCA to address the related issues GAO identified as part of the Bureau's strategic planning process for the 2030 Census. The Department of Commerce agreed with our findings and recommendations and described several cost savings and efficiency gains—which we have not audited—from their related address list-building efforts. The Census Bureau, Office of Management and Budget, and U.S. Department of Transportation each also provided us with technical comments, which we incorporated as appropriate.
Recommendations for Executive Action
| Agency Affected | Recommendation | Status |
|---|---|---|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau identifies metrics on the extent to which governments participating in LUCA overlap in their coverage of residents, as well as the characteristics of participants such as type of government and geographic area, and reports on such metrics. (Recommendation 1) |
Commerce agreed with this recommendation. In June 2022, the Bureau described in its 2020 Census Local Update of Census Addresses (LUCA) Operational Assessment Report how it measured the overlap between submissions from different related government entities, such as from a county as well as from its state. Having such metrics about the extent to which governments participating in LUCA overlap in their coverage of residents that the Bureau can use in near-real time during future LUCA-like efforts, could help the Bureau assure itself of getting desired coverage through additional outreach to other overlapping levels of government to address gaps in coverage or address-data quality.
|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau takes steps to conduct in-office reviews of a greater share of addresses submitted by governments before the addresses are added to the Bureau's address list for potential field work. (Recommendation 2) |
Commerce agreed with this recommendation. In its April 2020 action plan, the Bureau indicated that it would continue to identify improvements to address list-sharing programs so that more addresses submitted by governments are reviewed using in-office methods. In December 2022, Bureau officials briefed us on how they expect their evolving plans for more continuous intercensal geographic updates later in the decade to result in increased reliance on in-office reviews. In February 2023, officials informed us that their current thinking was that every address submitted by governments that does not match to the census lists will go through an in-office review process before being included in potential field work. In January 2024, officials provided an approved project plan for the 2030 Local Update of Census Addresses operation that established that all address submissions that do not automatically match the Bureau's list will be reviewed using in-office validation methods. By subjecting all submissions to these methods, instead of the sample that was done for 2020, the Bureau is likely to further reduce fieldwork during address canvassing.
|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau, as part of the Bureau's assessment of LUCA for 2020, consults with OMB to report on the factors that led to enumeration outcomes of addresses reinstated to the Bureau's master address list by the LUCA appeals process. (Recommendation 3) |
Commerce agreed with this recommendation. In June 2022, the Bureau described in its 2020 Census Local Update of Census Addresses (LUCA) Operational Assessment Report several factors contributing to the enumeration results of appealed addresses that were reinstated. Related factors include the level of government submitting the addresses, the relative number of addresses being submitted, how old the imagery was that the Bureau was relying on, and the extent that addresses had been submitted by non-governmental partners. The assessment was based in part on OMB's own data and operational close-out efforts. Knowing these factors can help establish lessons to help lower the amount of fieldwork in the future, and thus costs.
|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau identifies and tracks specific costs for related address list development efforts. (Recommendation 4) |
Commerce agreed with this recommendation. In its April 2020 action plan, the Bureau indicated that it would use existing systems to identify and report the costs of individual address list update-related activities. The Bureau also indicated that it would develop the means for capturing the cost of machine-based methods of updating the address list. In December 2022, Bureau officials said that they have not yet developed a methodology to identify and monitor costs and variances for address list development projects, but they were going to follow up with their Decennial Budget Office regarding recent cost tracking initiatives that might be leveraged moving forward. In March 2024, they stated that they expect to have more information by December 2024. In order to fully implement this recommendation, the Bureau will need to identify and track costs with sufficient detail and breadth to compare the relative cost-effectiveness of alternative efforts it considers and uses to build its address list.
|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau improves the use of LUCA results to inform procedures of other decennial operations, such as sharing information on address update quality to inform NRFU planning or administrative records modeling. (Recommendation 5) |
Commerce agreed with this recommendation. In its April 2020 action plan, the Bureau indicated that it would investigate how best to improve the flow of address data from governments into other census activities, such as research on administrative records or address canvassing. In December 2022, Bureau officials said that they are documenting various ways this may be implemented. Officials expect that later in the decade the process diagrams that will control the flows of geographic information to and from the Census Data Lake will likely demonstrate the use of address-update data by other decennial activity, such as by administrative-record modeling staff. As of March 2024, officials reported that they had determined the steps required to add LUCA data to the Enterprise Data Lake but were waiting until other research or operational uses identified their needs for the data before adding LUCA data to the data lake. In September 2024, the Bureau uploaded its 2020 LUCA data to its data lake for broad researcher access and documented and committed to the steps needed to do so similarly with the 2030 LUCA data when it is available. These steps will help ensure that the data on addresses collected from governments and on their quality will help inform the planning of other census activities that rely on address data. Improved planning and integration of such projects can help improve the efficiency of decennial operations contributing to improved data quality.
|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau realigns the schedule of LUCA-related programs to provide participants with more time to review addresses. (Recommendation 6) |
Commerce agreed with this recommendation. In June 2022, the Bureau described in its 2020 Census Local Update of Census Addresses (LUCA) Operational Assessment Report various benefits of providing more time for the LUCA process and recommended starting LUCA in the 2030 Census earlier in order to achieve them. In December 2022, Bureau officials described their forthcoming high-level schedule of LUCA activity in support of the 2030 Census that would demonstrate their plan to provide participants with more time to review addresses. In February 2023, they informed us that while the full schedule for LUCA activities would be completed only much later in the decade, they expected the milestone schedule to be available sometime in 2023. In January 2024, officials provided an approved LUCA project plan with a milestone schedule that provides 180 days for participating governments to review the address list (up from 120 days for the 2020 cycle). Allowing more time for governments to participate will benefit the LUCA program by making it likely that more governments will participate and that their submissions will be of higher quality than would otherwise be the case.
|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau uses the Bureau's data on hard-to-count areas to inform geographic activities such as: targeting LUCA outreach to tribal, state, and local governments; planning additional rounds of in-office address canvassing; and providing feedback to tribal, state, and local governments on gaps in their respective address data. (Recommendation 7) |
Commerce agreed with this recommendation. In its April 2020 action plan, the Bureau indicated that it would give participants access to the Bureau's data on hard-to-count areas so that participants could prioritize their address list review efforts. In December 2022, Bureau officials described a series of tools they were working on that would enable the Bureau and participants to better see the geography where address improvement activity may be most needed. They also described a research plan they would work on later in the decade for locating and enumerating historically undercounted populations, which they expected would help address this recommendation. In February 2023, they informed us that they target the related data analysis to be completed by September 2024; and in March 2024, confirmed that they will use the hard-to-count information in their forthcoming data-driven coverage-based approach for LUCA. In order to fully implement this recommendation, the Bureau will need to demonstrate later in the decennial cycle how it is using its data on hard-to-count areas to improve targeting of outreach to governments, planning other address-improvement activity, and providing feedback to governments.
|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau, as part of the Bureau's strategic planning process for geographic programs, reexamines LUCA in conjunction with stakeholders, other federal agencies as appropriate, and Congress to address the issues we have identified, including but not limited to:
|
Commerce agreed with this recommendation. In its April 2020 action plan, the Bureau indicated that it would reach out to stakeholders in the Summer of 2022 to discuss reexamining LUCA and other address frame initiatives. In March 2022, Bureau officials described the status of that ongoing outreach and the wide range of issues it is raising. They said they anticipate developing some earlier proposals for administrative changes that work within current legislative authorities before considering more broadly what legislative proposals might be needed. In March 2024, Bureau officials said they were continuing this reexamination. In order to fully implement this recommendation, the Bureau will need to carry out a reexamination with stakeholders on the issues identified in our report as well as identify and report on alternatives and legislative proposals as may be appropriate.
|