In any given year, data users may see ACS estimates that look different from what might be expected based on recent trends. There are several factors that may contribute to variation from past trends in the 2024 ACS 1-year estimates:

• The sample. The ACS is a sample survey, meaning any individual or household represents many more people. Who is ultimately in the sample directly affects what estimates of the characteristics look like for the community.
• Changes in the population. The nation’s population is dynamic, and the ACS often picks up on changes from year to year.
• Change in population controls. Each year, the ACS uses the latest independent population estimates of certain demographic characteristics as “controls” for the ACS sample results. These controls align the ACS estimates with official population estimates for those characteristics and enable us to account for the coverage differences of the ACS estimates. The 2024 ACS 1-year estimates use population controls that incorporate an improved methodology for estimating net international migration. As a result, the population estimates more accurately reflect the net increase in migrants to the United States between 2022 and 2024, resulting in a cumulative increase of about 5 million people in the 2024 estimate of the total population, to which the ACS is controlled. This increase may contribute to changes in the 2024 ACS 1-year estimates compared to other years.
• Computer-Assisted Personal Interviewing (CAPI) optimization. After the self-response phase of data collection, a sample of nonresponding addresses is selected for follow-up with a Census Bureau representative, either by phone or in-person; this phase is known as CAPI. The ACS uses administrative data to optimize data collection costs and data quality in the CAPI operation by informing contact strategies—when to decrease the number of visits or when to stop interviewing. In 2024, less than 1% of the monthly ACS sample was removed from the CAPI workload each month from this optimization.
• Nonresponse bias. The ACS has had lower response rates in the 2020s compared to the 2010s, which can increase the risk of nonresponse bias. Nonresponse bias is when survey respondents have significantly different characteristics from nonrespondents, making a survey’s results less representative of the overall population. Census Bureau analysts evaluated the potential for nonresponse bias by examining certain characteristics about the sample population from available administrative records. At the national level, the characteristics of who responded to the ACS in 2024 were not largely different from the characteristics of 2023 ACS respondents. Additional analysis was done to examine changes in nonresponse bias at the state level. The Census Bureau monitors the presence of nonresponse bias in the data and continues to research how to mitigate it.
• Weighting. The ACS weights responses to make the survey estimates more representative of the overall population. The weights are based on the sample design, response and non-response, and the housing unit and population controls. The weighting reduces bias due to nonresponse and coverage. The weights essentially determine how many total people each person in the sample represents. Larger households and households that respond during the in-person follow-up phase of data collection tend to have a larger impact on the estimates. This can sometimes compound changes driven by other factors, such as the population controls.

More information on these potential contributors is available in the 2024 ACS 1-Year Research Note.

Guidance

• Per our standard guidance, we encourage data users to be mindful of margins of error published alongside the ACS estimates. The margins of error provide information on the uncertainty embedded in the estimates related to sampling variability and should guide the use of the estimates.
• We also encourage data users to be mindful of the potential interplay of the factors discussed in this user note. Use caution when interpreting significant changes that may be a result of the methodology.
• Finally, the ACS is designed to measure characteristic distributions. Whenever possible, the Census Bureau recommends that users compare population characteristics such as percentages, means, medians, and rates rather than estimates of population totals.