2020 Census: A More Complete Lessons Learned Process for Cost and Schedule Would Help the Next Decennial
Fast Facts
By the time work ends in 2024 on the 2020 Census, it will have cost $13.7 billion—which is below the Census Bureau's original estimate of $15.6 billion. This was partly due to increased productivity during data collection—such as by using laptops instead of paper and pen to collect census data.
The Bureau has collected dozens of cost and schedule lessons from its 2020 Census experience. It also has a process in place for tracking how it implemented the resulting recommendations. However, it hasn't evaluated whether this process has been effective.
We recommended that it evaluate this process.
What 2020 Census Spending Bought, as of September 2022
Highlights
What GAO Found
The 2020 Census will have cost roughly $13.7 billion by the time its activity ends in 2024. This falls below the Census Bureau's October 2017 estimate of $15.6 billion. The Bureau allocated more than 80 percent of its 2020 Census spending to enumeration operations, infrastructure, and information technology. The largest area of spending was for enumeration projects such as following up with households that did not return their census forms and counting people who live in group quarters, such as skilled nursing and correctional facilities. The second largest spending area was for infrastructure to support various operations, such as hiring field staff and leasing office space. The third largest spending area was for census survey and engineering, which was mostly spent on IT.
The actual cost of some census operations was higher than planned. Other operations cost less. For example, the Bureau reported it used technology to increase the productivity of field data collection above expectations. This, in turn, resulted in spending less than planned on activities such as following up with non-respondents. Conversely, the Bureau spent more than planned on temporary office space, which it used longer than expected because of the COVID-19 pandemic.
The Bureau delayed or extended census activity prior to and during the COVID-19 pandemic. For example, prior to the COVID-19 outbreak, Bureau hiring was delayed due to problems processing temporary staff background checks. The Bureau also had delays in integrating IT systems within their operations. Additionally, the Bureau paused, extended, or delayed several of its 2020 operations in response to the COVID-19 pandemic.
The Bureau is learning lessons from the 2020 decennial to be used for its early 2030 planning. It is using seven of the eight steps GAO has previously identified for a lessons-learned process. For example, in addition to collecting internal lessons, the Bureau stores and archives resulting recommendations, along with those from external oversight bodies. The Bureau's recommendation management plan describes many of these steps. However, neither the plan nor other decennial guidance refers to periodic evaluation of the process. Documenting and carrying out this process step for the 2030 Census can improve the Bureau's ability to build on prior successes and address future challenges.
Eight Steps of a Lessons-Learned Process
Why GAO Did This Study
Conducting the census is an enormous, expensive, and complex undertaking. The Bureau spends years planning for it. Documenting program cost estimates and implementing good schedule management are essential to conducting a cost-effective, high-quality census within statutory time frames.
GAO was asked to review the Bureau's implementation of the 2020 Census. This report analyzes how (1) the Bureau's actual 2020 Census spending differed from plans, (2) the Bureau's actual schedule differed from plans, and (3) the Bureau is using lessons from the 2020 Census to inform 2030 planning. GAO analyzed the Bureau's 2020 cost and schedule data and reviewed prior related GAO and Bureau reporting. GAO reviewed documentation and interviewed Bureau officials about the lessons the Bureau learned from the 2020 Census.
Recommendations
GAO is making two recommendations to the Department of Commerce, including that the Bureau take steps during the 2030 Census to document and evaluate its lessons-learned process. The Department of Commerce agreed with GAO's findings and recommendations.
Recommendations for Executive Action
| Agency Affected | Recommendation | Status |
|---|---|---|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau document and take steps during the 2030 Census to evaluate the Bureau's comprehensive lessons-learned process. (Recommendation 1) |
Commerce agreed with this recommendation. In August 2023, the Bureau provided us with its action plan describing steps it will take, culminating in an audit report of related actions it plans to complete in September 2025. In March 2024, Bureau officials provided us with and briefed us on an updated quality assurance management plan documenting procedures that help address this recommendation, such as the regular internal reporting to managers on the status of prior lessons learned. The implementation of those procedures as documented in the management plan partially implement this recommendation. In order to fully close this recommendation, the Bureau will need to follow through on other steps it described, completing its evaluation of the lessons learned process.
|
| Department of Commerce | The Secretary of Commerce should ensure that the Director of the Census Bureau include steps in its 2030 schedule management plans for learning lessons from systematic ex-post evaluation of the Bureau's extensive decennial and related schedule data. (Recommendation 2) |
Commerce agreed with this recommendation. In August 2023, the Bureau provided us with its action plan describing steps it will take. The plan refers to providing 2030 Census planners with schedule data and lessons learned from the 2020 Census--steps we described in our report--but the plan makes no reference to either a systematic ex-post evaluation of the schedule data for the 2020 Census or plans to do so for the 2030 Census, as recommended. In March 2024, the Bureau briefed us on steps it is taking with preparation of its 2030 Census schedule and provided us with a copy of its latest schedule management plan, which also makes no reference to systematic ex-post evaluation of 2030 Census schedule data. They also referred to their formal process of documenting schedule lessons learned drawn from the experiences of decennial staff with the 2020 Census scheduling activity, which we had reported on. In order to fully address this recommendation, the Bureau will need to demonstrate how it plans to leverage the extensive quantitative data it accumulates with its decennial scheduling activity to look for and document possible additional lessons learned.
|