Banking Deserts Dashboard
January 16, 2024
Navigating Banking Deserts
The pace of bank branch closures across the United States has doubled since 2020. While the popularity of online banking has skyrocketed, physical banking still plays an important role for many consumers. For some, including lower-income, rural, older, and disabled individuals, having no or fewer bank branches nearby could disrupt their access to banking services. The same is true for those facing barriers to using online financial services. This lack of access can in turn restrict opportunities to improve their financial health and build wealth. That’s why it’s important to know which communities have few or no bank branches nearby. An analysis that identifies banking deserts is one way to assess the scarcity of physical bank branches.
Defining Banking Deserts
A banking desert is defined as a census tract without a physical bank branch within a certain geographic radius from its population center or within the tract itself. That radius is determined by the type of community within which the census tract lies: 2 miles for urban communities, 5 miles for suburban communities, and 10 miles for rural communities.
A potential banking desert is defined as a census tract that will become a banking desert if one bank branch closes within a certain radius from its population center or within the tract itself. That radius is 2 miles for urban communities, 5 miles for suburban communities, and 10 miles for rural communities. Keep reading for answers to frequently asked questions about banking deserts.
The Banking Deserts Dashboard can help you quickly identify banking deserts and potential banking deserts. It includes census tract-level data of physical bank branch availability within all 50 states from 2019–2023. Knowing where there are few or no bank branches nearby can help inform solutions for better serving the financial and banking needs of residents of these communities.
Banking Deserts Dashboard
The Banking Deserts Dashboard is intended for informational and educational purposes.
Have a question or comment? Contact us.
Banking Desert FAQs
A banking desert is defined as a census tract without a physical bank branch within a certain geographic radius from its population center or within the tract itself. That radius is determined by the type of community: 2 miles for urban communities, 5 miles for suburban communities, and 10 miles for rural communities. This definition is based on “areas with very low branch access” in the Interagency Notice of Proposed Rulemaking to Implement the CRA.
A potential banking desert is defined as a census tract that will become a banking desert if one bank branch closes within a certain radius from its population center or within the tract itself. That radius is 2 miles for urban communities, 5 miles for suburban communities, and 10 miles for rural communities. This definition is based on “areas with low branch access” in the Interagency Notice of Proposed Rulemaking to Implement the CRA.
Research has suggested that bank branches in lower-income communities are correlated with more mortgage originations and lower interest rates, with these effects strengthening the closer a branch is to the neighborhood. Lower access to traditional banking services can also drive consumers to alternative financial services. Bank branch closings may increase the demand for fringe banking services such as payday lenders and check-cashing services, which raise the overall cost of consuming and inhibit opportunities for saving.
Banking deserts likely have the greatest impact on people who struggle with access to transportation, digital devices, or broadband and those who live in areas where things are more geographically spread out, especially if they also struggle with credit access. Additionally, there’s still high demand for banking with a teller in lower-income and rural communities. This continued demand for in-person banking is also true of older adults and people with disabilities.
Most financial institutions have online and mobile banking options. However, in order to take advantage of these options you need three things: a modern digital device, internet access, and the skills to navigate websites and apps. Unfortunately, home broadband adoption remains out of reach for many lower-income, rural, and older populations and people of color. Without high-speed internet, consumers may find it more challenging to fully utilize online banking services. When we talk about digital inclusion at the Federal Reserve, we recognize that people need high-speed internet, devices, and digital skills.
A census tract with limited broadband access is one in which more than 21.5% of households in the census tract did not have access to broadband internet in the 2016-2020 American Community Survey (ACS). These tracts are in the top quartile of tracts for the percent of the tract’s households that do not have access to broadband internet.
A census tract is considered low- and moderate-income (LMI) if its 2016-2020 ACS median family income is below 80 percent of the area median family income and middle- and upper-income (MUI) if the 2016-2020 ACS median family income is greater than or equal to 80 percent. Values are not available for tracts with insufficient income information.
A census tract is urban if its population lies primarily within a metropolitan statistical area (MSA) and a principal city of its MSA. It is suburban if its population lies primarily within an MSA but not within the principal cities of its MSA. The tract is rural if its population does not lie primarily within an MSA. We use US Census 2023 delineation files for principal cities of MSAs.
A census tract has a majority race/ethnicity if more than half of its 2016-2020 ACS population shares the same race or ethnicity. If no race/ethnicity makes up more than half of a tract’s population, it is labeled as having “no predominant race.”
The dashboard data will be updated annually. To receive emails about future updates of the dashboard and other Fed Communities content, sign up for our newsletter.
Fed Communities published an article about mobile bank branches as a strategy used by some communities in response to bank branch closures. The Philadelphia Fed has recently published a report that analyzes this data, and has also recently published a deeper dive of banking deserts in the Federal Reserve Third District states of Pennsylvania, New Jersey, and Delaware.
Please note that the Federal Reserve is not a grantmaking organization. Additionally, the Fed does not maintain grant money or any other type of funds or accounts for individuals.
Banking Desert Dashboard Methodology
We calculate the desert status of census tracts using bank branch locations from S&P Global Market Intelligence SNL US Bank Branch Data Set (“S&P SNL branch data”). The sample of branches used are inclusive of brick-and-mortar, non-in-store, full service, and retail branches. We include all types of banks and credit unions, but not Community Development Financial Institutions (CDFIs).
We also make additional data-cleaning decisions in constructing the sample of bank branches. First, we remove cases where the branch status is “closed” but the closed date is missing. We also remove cases where the branch status is “active,” but the closed date is populated. Second, we identify a small sample of branches likely to be duplicated (3%). These are cases in which the institution company name, first address line, and city are the same as one or more other observations with overlapping open/closed dates. We then decide which observation is most likely accurate within each group of duplicates through a series of decisions that lead us to drop about half of the 3 percent of observations identified as duplicates.
Finally, while approximately 15 percent of the remaining observations are missing open dates, we include those observations to avoid overestimating the number of banking deserts. A cross-examination of Summary of Deposits (SOD) data from the Federal Deposit Insurance Corporation, the main data source used to construct the non-credit union portion of the S&P SNL branch data, revealed that 99.4 percent of bank branch cases with missing open dates that we identify as having stayed open at the end of 2019 were identified and active in the SOD data as of June 2019. Thus, we assume with great confidence that these branches were open in 2019.
We use the latitude and longitude of bank branches in the S&P SNL branch data to geocode each branch to 2020 census tracts. We use the 2016–2020 American Community Survey (ACS) data and 2022 FFIEC Census Flat File for relevant census data, US Census TIGER/Line Shapefiles for geographic boundary computations, US Census 2023 delineation files for principal cities of metropolitan statistical areas, and the Missouri Census Data Center’s 2022 Geocorr application for data on population-based tract centroids and areal overlap between tracts and principal city boundaries.
We define banking deserts and potential banking deserts based on the definitions of “areas with very low branch access” and “areas with low branch access,” respectively, introduced in the Interagency Notice of Proposed Rulemaking to Implement the CRA. However, there are two slight differences between our data and methodologies and the interagency data and methodologies: For one, we use a stricter definition of “active branch,” and two, our data are more recent and have been retroactively updated to capture closures that occurred in 2020 and 2021.