The FDIC National Survey of Unbanked and Underbanked Households, or the “Household Survey” supports the FDIC’s commitment to expanding Americans' access to safe, secure, and affordable banking services. Conducted biennially since 2009 in partnership with the U.S. Census Bureau, the Household Survey collects information on bank account ownership and other financial products and services that households may use to meet their transaction and credit needs. The nationally representative survey is administered to approximately 30,000 U.S. households and yields findings that are representative for the 50 states and the District of Columbia.
Resources: Household Survey Home | Five-Year Table Tool | Instructions
FAQs
- What are five-year estimates?
- How are five-year estimates different from single-year estimates?
- Can five-year estimates be compared to one another?
- What are the benefits and tradeoffs of five-year estimates?
- Why do Metropolitan Statistical Area (MSA) names differ depending on the chosen five-year range?
- Where can I find additional information about the five-year estimates?
What are five-year estimates?
Five-year estimates are computed by combining data from three consecutive surveys. For example, the 2019-2023 five-year estimates are based on data from the 2019, 2021, and 2023 surveys. Roughly speaking, you can think of five-year estimates as reflecting an average over the five-year time period covered by the three surveys the estimates are based on. The precise interpretation of the five-year estimates is slightly more nuanced because the survey is conducted every two years, rather than each year.
How are five-year estimates different from single-year estimates?
Single-year estimates are based on data from a single survey (e.g., 2023). Single-year estimates of unbanked rates cover the point in time when the survey was administered (e.g., June 2023).
Five-year estimates combine data from three consecutive surveys (e.g., 2019, 2021, and 2023). Five-year estimates of unbanked rates cover the three points in time when each survey was administered (e.g., June 2019, June 2021, and June 2023).
Five-year estimates are also different from multiyear estimates available in the custom data table and custom chart tools. The multiyear option in the custom data table and custom chart tools provides comparisons of single-year estimates over time.
Can five-year estimates be compared to one another?
Five-year estimates from one time period (e.g., 2019-2023) can be compared with other five-year estimates from the same time period. For example, 2019-2023 five-year estimates for Pennsylvania can be compared with 2019-2023 five-year estimates for the Pittsburgh, PA, MSA.
Five-year estimates from one time period should not be compared with five-year estimates from another time period if the time periods overlap. For example, 2019-2023 five-year estimates should not be compared with 2015-2019 five-year estimates because they share the common year, 2019.
In many cases, 2019-2023 five-year estimates can be compared with 2009-2013, 2011-2015, and 2013-2017 five-year estimates because the time periods do not overlap. However, for race and ethnicity, 2019-2023 and 2009-2013 five-year estimates should be compared with caution because of differences in racial and ethnic category definitions. For the 2019 report, racial and ethnic categories were revised to be consistent with the 1997 U.S. Office of Management and Budget (OMB) standards for the classification of race and ethnicity and with CPS tabulations of race and ethnicity (see Appendix 1 of the 2019 report for a description of these racial and ethnic categories). All five-year estimates other than the 2009-2013 five-year estimates use the racial and ethnic categories from the 2019 and later reports, while the 2009-2013 five-year estimates use the racial and ethnic categories from the 2017 and previous reports (see Appendix 1 of the 2017 report for a description of these racial and ethnic categories). Additionally, as described in a later FAQ, all five-year estimates other than the 2009-2013 five-year estimates use February 2013 MSA definitions, while the 2009-2013 five-year estimates use June 2003 MSA definitions. For MSAs that experienced a change in geographic boundaries (e.g., the addition or subtraction of a county), 2019-2023 and 2009-2013 five-year estimates should be compared with caution. For all five-year estimates, other than 2009-2013 five-year estimates, a tilde (~) next to an MSA name indicates that the MSA was affected by a geographic boundary change.
In March 2024, OMB published new standards for the classification of race and ethnicity. As the new standards are implemented in the CPS, please note that future reports may use different racial and ethnic categories.
What are the benefits and tradeoffs of five-year estimates?
Five-year estimates are based on a larger sample than the single-year estimates, which produces more precise estimates. The increased precision permits estimates for more geographies and subpopulations. For example, five-year estimates are available for more MSAs than single-year estimates. Additionally, five-year estimates for subpopulations, such as low-income households or Black households, are more widely available at the national level, state level, and for larger MSAs. One trade-off with the five-year estimates is that they are not as current as the single-year estimates.
Why do Metropolitan Statistical Area (MSA) names differ depending on the chosen five-year range?
MSA definitions are established by OMB. The 2009-2013 five-year estimates use June 2003 MSA definitions and names (see OMB Bulletin Number 03-04 [June 6, 2003]). The remaining five-year estimates use February 2013 MSA definitions and names (see OMB Bulletin Number 13-01 [February 28, 2013]). As discussed in the technical documentation for the June 2015 survey, the Census Bureau phased the 2013 MSA definitions into the CPS (and phased out the 2003 definitions) over the period May 2014 to July 2015.
Where can I find additional information about the five-year estimates?
For additional information, please see the Five-Year Estimates Table Tool Instructions. For technical questions, including more details about how the five-year estimates are computed, you may email the Survey Team at HouseholdSurvey@FDIC.gov.
All PDF files on this page reference Portable Document Format (PDF) files. Adobe Acrobat, a reader available for free on the Internet, is required to display or print PDF files. (PDF Help)