The Great Consolidation of Banks and Acceleration of Branch Closures Across America:
Branch Closure Rate Doubled During the Pandemic
Photo by: ©1599685sv – stock.adobe.com
Jad Edlebi, Senior GIS Specialist, Research
Bruce C. Mitchell, PhD, Senior Research Analyst
Jason Richardson, Senior Director, Research
— Key Takeaways
As the COVID-19 pandemic disrupted local economies, banks took advantage of the crisis to double the pace of branch closures.
Accelerating branch closure rates make it difficult to enforce Community Reinvestment Act rules, which must be modernized.
Small businesses still depend on in-person banking services despite the proliferation of online alternatives, and the shrinking of branch networks threatens local economic activity that is key to wealth-building in marginalized communities.
The Great Consolidation resulted in a few large megabanks dominating the market while small community banks disappeared.
— Key Findings
Two-thirds of banking institutions have disappeared since the early 1980s — declining from nearly 18,000 in 1984 to fewer than 5,000 in 2021.
Small banks suffered the greatest decline. When banking competition was robust in the mid-1990s, 84% of institutions held less than $330 million in deposits. Today, nearly half of the banks that remain hold $1.3 billion or more.
Mergers and acquisitions drive most branch closures, along with a shift to mobile and internet-based transactions.
Nine percent of all branch locations in the US closed between 2017 and 2021 — a loss of about 7,500 brick-and-mortar locations.
The branch closure rate doubled during the pandemic. Banks have closed more than 4,000 branches since March 2020. This doubled the rate of closures from 99 per month during the 10 years prior to the pandemic to 201 closings per month.
Low- to moderate-income and minority neighborhoods were hit hard. One-third of the branches closed from 2017 to 2021 were in a low- to moderate-income and/or a majority-minority neighborhood where access to branches is crucial to ending inequities in access to financial services.
Pre-Pandemic Closure Patterns
Banks have been aggressive in closing physical branch locations since the Great Recession. Many of these losses were caused by mergers and consolidations, but other factors are involved. Financial institutions have increased their reliance on technological services, leading to the proliferation of financial technology (fintech) companies, a shift to online banking, and the automation of administrative labor within banks.
Last year, NCRC released a report that examined bank branch closures up to 2020, finding that more than 13,000 branches closed between 2008 and 2020 — a 14% loss nationally. The impact of these closures was unevenly distributed across the country, with rural communities and low-income urban communities suffering the worst effects. Some of those hardest-hit areas have substantial majority-minority communities.
At the publication of last year’s update, NCRC anticipated a continuation of the branch closure trend. What we did not expect was that there would be an acceleration of closures exacerbated by the pandemic. Larger financial institutions drove that acceleration in 2021, as the biggest banks closed branches more rapidly than before the start of the pandemic. In the ten years from March 2010 through March 2020, there was a net loss of 11,820 branches in the US, an average of 98.5 per month. Since the pandemic reached our shores in March 2020, 4,025 branches have closed — an average of 201 branch closures per month. This was beyond our prediction and calls into question the future of bank branches as the principal conduit for access to financial services.
The Great Consolidation of the Banking Industry
The most visible presence of any bank in any community is the local branch. Neighborhood branches and the access to financial services they provide were foundational to the concept of community reinvestment. Banks take deposits via their local branches and then support local communities by making loans and offering services to small businesses, community development organizations and charitable organizations. The presence of local branch offices provides an opportunity for face-to-face interactions that build both trust and financial literacy for individual borrowers and small businesses.
Originally intended as a corrective to the history of public and private redlining, the Community Reinvestment Act (CRA) obligates banks to serve the communities in which they operate. The focus of CRA is on access to loans and financial services for households and businesses in low- to moderate-income (LMI) neighborhoods.1 Regulators routinely examine banks to assess how well they meet their duty to equitably provide loans and financial services within their market or assessment areas. CRA utilizes a depository institution’s branch location “footprint” in its delineation of assessment areas, by which the success of community access to services is evaluated. The presence of branches in LMI areas is one criterion that federal regulators use to assess a bank’s success in meeting the financial services needs of communities.2
Major shifts in the number of banking institutions and the extent and number of their branch networks in the years after CRA was enacted have complicated these policy mechanisms. Mergers and consolidations have driven the banking industry toward fewer institutions of greater asset size during the past 45 years. When CRA passed in 1977 — well before banking system deregulation removed constraints on national bank mergers — there were 14,411 total banking institutions with 47,519 branch locations. The number of institutions peaked in 1983-84 at 17,811, thereafter steadily declining to 4,897 in 2021 (Figure 1).
The number of bank locations, meanwhile, peaked in 2009 at 92,394 individual physical branches. The financial crisis of 2007 initiated the Great Recession, after which institutional failures and consolidations rippled through the financial system. The local branch banking system that CRA was built upon has been on a downward trajectory since that time: Banks have closed more than 13,089 branches since 2009 (Figure 2).
Mergers have decimated small banks in the US. As the largest banks merged and grew bigger, smaller institutions and their branches were often targeted for acquisition and disappeared in sizable numbers. This hurt rural areas in particular, as NCRC found more than 80 rural banking deserts were created from 2008 through 2016. Meanwhile, an increasing share of the market has been dominated by the 25 largest banks. Figure 3 shows how the proportions of branches for small, large, and the top 25 banks by asset size have changed since 1994.3 The impact on banking consolidation and branch closures has been most acute among small banks — and by 2020, only 9% of branches fell into this asset-size category. The trend of domination by the top 25 banks reached a plateau in 2010, with other large banks increasing their share of this shrinking pie, but — crucially — not increasing the number of branches where customers can interact directly with loan officers who know them and know their neighborhood. This corresponds to a wave of consolidations and branch closures by the major banks which continues to the present.
These shifts have weakened the business structures upon which the Community Reinvestment Act was built. But they do not alter banks’ obligations under the law. Today’s larger, less-local banks are still charged with serving the credit needs of the entire community they serve. Changes in how the public interacts with their bank do not create an exemption to the law.
In 1994, small banks were 84% of all banks in the US. However, due to the rise of interstate banking as deregulation took effect and the Great Consolidation of bank mergers and acquisitions that followed the 2008 Great Recession, the banking landscape of 2021 looks very different. There are far fewer banks, with an almost even split between smaller banks (52%) and larger banks (48%).
Much of the growth within large banks comes from the top 25 institutions, whose combined branch network skyrocketed from 8,740 branches in 1994 to 37,884 branches by 2010 before plateauing. Institutions such as Wells Fargo Bank, JPMorgan Chase Bank and Truist Bank all share this exponential growth compared to their smaller counterparts. The chart below shows the consistent growth of the combined branch network of large banks alongside the steady decline of small banks’ own networks, from 27,386 branches in 1994 to 7,078 branches by 2021 nationwide. The top 25 banks had already been increasing their market share prior to the Great Recession and subsequent mergers and acquisitions boom. In that post-crash Great Consolidation, these larger institutions acquired and closed many of the branches that were previously owned by smaller banks. This business practice cut off physical access to face-to-face financial services in LMI and majority-minority neighborhoods in both urban and rural areas, eventually leading to the banking deserts of today.
The Adverse Impacts of Closures on Communities
Bank branches have played a unique role in community development and financial inclusion for decades. Chartered by the public, branches embodied banks’ obligation to serve the people in the communities where they had offices. Often, the branch was also the main office of the small banks that dominated the American financial landscape.
With the expansion of the American economy after World War II, bank branch networks spread steadily across the country, from a total of 17,198 individual bank branches at the war’s end to 47,519 by 1977. This expansion improved financial access for businesses and homeowners — and made the bank branch a natural venue for assessing quality and equity of service. The authors of CRA recognized that branches are crucial to financial services access and used a bank’s network of branches to define their service area. CRA obligated banks to maintain or improve financial services access in LMI neighborhoods, although these areas had fewer branches than wealthier neighborhoods.
In 1994, however, the Riegle-Neal Interstate Banking and Branching Efficiency Act, allowed banks to open branches across state lines. This sparked a shift within the financial services industry which has resulted in the disappearance of two-thirds of banking institutions by 2020. However, branch networks continued to expand, reaching a peak of 92,000 locations in 2009. The financial crisis and subsequent Great Recession fundamentally changed banks’ marketing strategies away from expanding the footprint of their branch networks, to increased automation. With the advent of internet and mobile banking, banks could attract and service a customer base with decreased reliance on physical branches. However, less tech-savvy and lower-income households with more volatile income streams are still reliant on branches to access financial services. Many branches have closed, and we expect many more to do so over the next several years.
When bank branches close, there are several adverse effects on the surrounding community. Small business lending and activity in the area declines. More people use alternative financial services that open them to unregulated and predatory financial practices. An important commercial tenant and employer are lost.
In LMI communities which often had less access to branches to begin with, even a small number of closures can have an enormous impact on these communities that already have few branches to choose from. Branch closures have been shown to reduce financial access and, consequently, increase the number of unbanked or underbanked LMI and minority individuals and families. Banks have therefore directly hampered these communities’ ability to build wealth — at the same time that core businesses in their neighborhoods become more vulnerable to collapse because the local loan officer they trust is simply gone.
In their bi-annual survey, “How America Banks,” the Federal Deposit Insurance Corporation (FDIC) asks how people use banks and how vital are services such as branch access to consumers. As we noted in our previous report on branch closures:
Mobile and internet banking are often cited as the key drivers of bank branch closures. However, in 2019 the FDIC found that despite the growing use of those amenities, 87% of banking consumers had still visited a branch in the past year. More than 28% of consumers visited a branch at least ten times in that year. Consumers still think a branch matters. 4
While consumers have embraced mobile and internet banking to one degree or another, they clarify that branches matter to them as well, and without branches nearby, they are more likely to be un- or under-banked. Alternative financial providers — unregulated actors such as payday lenders, auto loan title lenders, or check cashing locations that charge exorbitant fees for financial services — often take their place.
2017-2021: Branch Closures and the COVID-19 Pandemic
As previously mentioned, the Great Consolidation has led many banks to close branches as banks merged at a record pace. However, between 2017 and 2021, almost one in ten branches had closed with many of the closings occurring during the pandemic. The downward trajectory of closings has turned into an avalanche over the past few years.
This study makes use of FDIC Summary of Deposits data from 2017 through 2021. The Summary of Deposits (SOD) is annual data released by the FDIC that provides a snapshot of open bank branches as of June 30 of that year. This study examines full-service brick-and-mortar retail branch locations in the 50 states and the District of Columbia. We have also used the FDIC institutional history to combine banks that merged during the study period.
The United States experienced an overall loss of 7,425 branches between 2017 and 2021, which amounts to almost 9% of the total number of branches that existed at the beginning of the study period. Of those closures, one out of every three branches were located in LMI or majority-minority neighborhoods. When considering the generally sparser branch networks in LMI and majority-minority neighborhoods, taking out one branch could have much more adverse impacts on access to financial services in a given neighborhood. For some households that don’t have access to transportation, closing a single branch may mean disconnection from accessing financial services entirely — thereby interrupting their opportunity to establish a credit history.
We identified the 50 metropolitan areas which had the largest number of bank branches in 2017, then sorted them by the share of those branches that have closed in the past four years. No metro area in America has lost as much of its local branch banking capacity as Portland, Oregon. In 2021, there were 421 bank branches in operation across the Portland metro area, a 19.7% reduction from 2017. Baltimore, New York, Hartford, Connecticut and Sacramento, California round out the top five. New York, home to far more branches than any other metro, lost 747 of them during the study period.
Small and intermediate metro areas saw smaller raw-number branch closures but suffered a much larger reduction in branch network capacity than we found in the biggest cities. Jacksonville, Florida lost the same one-in-five share of branches as did Portland, Oregon. In Portland’s Multnomah County, one in four branches closed. In one particular case, US Bank had previously planned to close 62 branches in the Portland area yet the pandemic seemed to have catalyzed those closures to a three month period between October 2020 and January 2021. In a statement, US Bank emphasized that even though they are closing branches, they “are continuing to open and enhance others, as well as rapidly enhancing our digital capabilities.”
Suffolk and Nassau, the two counties that make up Long Island, lost more than 160 branches — almost 20% of their total network. Larger metro areas such as New York, Chicago, Baltimore and Phoenix saw fewer closures than smaller metro areas, yet losses were substantial. Each of these four metro areas lost approximately 13% of its branches. In Chicago and Baltimore, branch closures in suburban counties trended higher than for their respective core cities. In Phoenix, about 15% of branches that existed in LMI and majority-minority communities prior to the pandemic have now closed.
At the other end of the spectrum several cities in Texas experienced very little loss, with Austin, Houston, San Antonio, Dallas and rural non-metro Texas reporting losses ranging from 2.1% (Dallas) to just 0.5% (Austin). The rural parts of Nebraska, Iowa and Kansas also saw few losses. This study is necessarily limited in its analysis of metropolitan or state level causes of branch closures in each case. In the case of Texas, there are several factors which warrant further research. For instance, were these areas already underserved in comparison to the rest of the country? Do small losses have an inordinate impact on rural areas, creating branch deserts? Additionally, there may be broader economic factors, such as lower rent for commercial space, lower labor costs, or localized businesses such as agriculture or mining incenting banks to keep branches open in these areas?
As many communities experienced a loss in access, it is important to note that most closures involve the biggest names in the financial services market. For instance, Wells Fargo had the largest branch network in 2017, but has since closed 993 branches. JPMorgan Chase has supplanted Wells Fargo in maintaining the largest branch network in the country.
When we ranked banks by raw number of branch closures, three of the top five banks on our branch closures list were engaged in substantial mergers or acquisitions during our study period. After SunTrust and BB&T merged in 2019 and rebranded as Truist, management closed 919 branches – the second-largest number of closures on our list. The next two on the list – US Bank (863 closures) and PNC Bank (418 closures) – both purchased competitors during our study window. Wells Fargo and the remaining top five branch closers – Bank of America (379 closures) – aggressively pursued mergers during the Great Consolidation period following the late-2000s financial crisis.
Because these raw numbers only capture part of the story, we also examined closings as a share of the overall branch location footprint of banks. Capital One tops the list by this metric, closing fully half of its branch network from 2017 to 2021. It today operates 320 locations, down from 641 in 2017. Truist and US Bank each closed one in four of their branches, while Santander and The Huntington National Bank closed one in five. Given the relationship between closures and mergers, the pending merger applications of Umpqua and Columbia Bank as well as US Bank and MUFG Union Bank suggests that both will increase their closures in 2022.
Detailing Branch Change by Area (2017 - 2021)
The source for the percent losses by neighborhood characteristics is the annual FDIC SOD file from 2017-2021. The closing and opening data come from the FDIC weekly branch changes file.
We have examined branch closures — both the absolute number and as a share of the capacity of bank branch networks. What was the impact on LMI and minority areas? According to the FDIC annual Summary of Deposits file, the year between July 2020 and July 2021 saw middle- and upper-income census tracts with a majority White population lose 4% of their branches, while LMI tracts that were majority-minority lost 2.3% of theirs.
This apparently proportional closure rate is cited by the banking industry as proof that closures are shared equally across neighborhoods. However, the results are disproportionate if the residents of LMI census tracts already had fewer bank branches accessible to them. Additionally, the tools of financial access through mobile banking and internet access are less available to LMI customers and to communities of color, making seemingly proportional closures have a higher impact on access than in wealthier areas. Equality doesn’t necessarily mean equity.
We know that when a branch closes, there is also a loss of access to credit and wealth building opportunities, which impacts families and small businesses. A 2019 Federal Reserve Bank of Philadelphia study found that rural areas were particularly hard hit by bank branch closures. Rural counties with poorer, less educated, and higher percentages of minority residents were particularly vulnerable.
Future research could address the actual impact on banking needs and services from closures. Very little is known about how branches vary in the services they offer in neighborhoods that are majority people of color or low-to moderate- income. Are there important differences such as services offered, hours of operation, transit connectivity, or other measures that affect how banking services are offered?
Breaking Down Closures During the COVID-19 Pandemic
By reviewing the weekly branch change filings, it is possible to see more detail about the pandemic-era closures hinted at in the Summary of Deposits file. Beginning about three months after the start of the pandemic, most banks responded by dramatically increasing the rate of branches they closed each month.
These pandemic-era changes to branch closure patterns stand out even compared to other major modern convulsions in the industry, our review of weekly data on branch changes found. Closures spiked following the Great Recession with a peak of 873 branches closing and just 339 new openings recorded in the second quarter of 2010 alone.
Branch Changes 2000 - 2021
Branch closings and openings over time. The leap in closings during the pandemic is clearly visible in the chart below. Data retrieved November 15, 2021. Please note that delays in reporting by banks likely mean that more recent time periods undercount the number of branch changes.
Pandemic-era branch closure rates exceeded even the highs reached in the wake of the financial crisis in 2010. There were 1,492 closures and just 330 new openings during the fourth quarter of 2020, leaving 1,162 net closures. This greatly accelerated pace of closures may indicate that banks are using the pandemic to dramatically reduce their physical footprint.
Bank branch closure rates nearly doubled in the wake of the first pandemic lockdowns. The 20 months of post-lockdown FDIC data available show that banks shuttered 4,025 branches from March 2020 through October 2021. That’s about twice the total from the immediate prior 20-month period. Some individual banks have far exceeded even this dramatic acceleration in closures. HSBC, for example, has closed 88 branches since March 2020 after shuttering only four in the prior period, a move the bank attributes to a new long term strategy to retrench their US operations. Such outliers aside, every bank in the dataset seems to have begun closing branches en masse about three months into the pandemic. The data offer no indication the closure wave is slowing.
While the intensity of the pace of branch closures increased nationally, several large metropolitan areas saw banks retreat from branch service at an even faster clip. Banks closed twice as many branches in Los Angeles during the pandemic as they had closed in the prior 20 months, for example, while closures tripled in Washington, DC. In other major cities the rate of closures accelerated even more – Boston had a 1,375% increase, San Francisco with a 617% increase, St. Louis with a 1,100% increase, and Atlanta where the increase was 369%. Smaller metro areas were also hit hard — Grand Rapids, Michigan had the highest increase in the local branch closure rate, 3,600%, while in Columbus, Ohio the rate was 680%. Closure rates more than tripled in each of Baltimore, Las Vegas, San Diego, Pittsburgh and Portland, Oregon. Only Chicago saw the rate of closures decrease during the pandemic.
These metros have little else in common other than branch closure rates. They are geographically distributed cities that vary widely in size. Most have increasing populations. The nature of their economies also show no clear pattern: Both declining rust belt cities and more economically dynamic globally connected cities make the list. This suggests banks are simply rushing to close as many branches as they can, as quickly as they can, anywhere they can.
In the midst of economic uncertainty, high unemployment, and a mounting death toll, banks decided to take advantage of the crisis and economic downturn to rapidly increase the pace of bank closures.
Future research should address the long term impact on communities. How many residents were affected by job loss or lack of banking services? How will future small business lending in these areas change? Was there a public health impact from these closures? Lawmakers should encourage regulators, advocates and banks to work together in order to mitigate the impact of bank closures on nearby communities.
Unfortunately, research on branch closures is complicated by the FDIC data on those locations. As we have previously noted, the location data for branches in the Summary of Deposits file has an error rate in excess of 5%. There are likely errors in the branch change file as well, and for branches that have closed it will be very hard to identify erroneous data. The datasets also lack data dictionaries or published methodologies.
Future policy efforts could address the poor condition of public data on branch locations.
What Will Banking in America Look Like in 2040?
It seems likely that several trends will continue to affect the number of bank branches in operation in the US. Bank mergers, the increasing sophistication of artificial intelligence and use of algorithms to automate banking functions, and the growing comfort level of consumers with online banking make the cost of maintaining extensive networks in fully-staffed locations challenging to justify in pure profit/loss terms.
But robust branch networks are not just a red-ink line item on a balance sheet. They are a key factor in local economic outcomes. Customers still use branches frequently. Physical, staffed bank locations remain critical for supporting small businesses in their communities. Rural areas and low- to moderate-income neighborhoods in urban areas, which already have fewer branches servicing them, are losing access to bank branches at a rapid pace. The nonbank financial services that supplement bank branches often ignore these communities.
As the pandemic appears to have supercharged banks’ plans to close locations, swift action is required to protect affected communities. NCRC calls upon the FDIC and other US federal prudential regulators to engage with banks and get more information on plans to close branches and best practices for transitioning consumers that use branches to other means of support. The Community Reinvestment Act requires banks to serve the communities where their branches are located. As branches become rarer, the power of CRA to support communities and businesses that would otherwise see less investment is eroded. Modernizing CRA to address the declining role of bank branches and expanding it to cover nonbanks offering banking services is the first step in guaranteeing that LMI, BIPOC and rural communities are not left behind.
Recommendations for Policymakers and Industry
Taken together, these findings demonstrate that both policymakers and bank industry leaders need to act to mitigate this rapidly accelerating wave of branch closures. We offer the following specific recommendations:
- Regulators should facilitate discussions between banks, communities and non-profit lenders to mitigate the impact of branch closures on surrounding communities.
- Banks should work with local businesses to ensure a continuity of lending and other services that companies rely on to remain in operation.
- The Community Reinvestment Act should be updated to address a market where bank branch networks no longer sufficiently define bank service areas. Internet and mobile banking have greatly expanded the areas which banks service, without extending their obligation to provide those services equitably.
- FDIC data on branches and institutions is fragmented across several resources, covering different time periods. The FDIC should work with banks to improve both data quality and public access to these online resources.
Data on bank branch openings and closings are maintained by the Federal Deposit Insurance Corporation (FDIC). Banks are required to provide the FDIC with prior notice of branch openings and closings under Section 228 of the Federal Deposit Insurance Corporation Improvement Act of 1991. Two files served as primary data sources for this analysis: the FDIC Summary of Deposits (SOD) and the FDIC Institution Directory. The FDIC SOD provides yearly reports on the number of bank branches open on June 30, while the FDIC Institution Directory is updated weekly. In addition, historical data was retrieved from the FDIC indicating the number of institutions and bank branches open, going back to the inception of record-keeping in 1934.
The FDIC SOD provides locational data, including address and imputed geographic coordinates. The accuracy of this locational data varies, and the June 30, 2021 dataset is shown in Table x below. The least accurate matches are at the US city/state and the US ZIP code levels, which default to assigning the geographic center or centroid of the state or ZIP code. Assignment at these geographic levels is imprecise, and 4.13% of the branch locations are relegated to these categories. Because of geolocation’s imprecision, it is impossible to determine the tract-level location of those bank branches. The remaining categories ranked in order from most precise to the least precise are “percentage assigned,” “US rooftop,” and “US streets.”
Bank branch geolocation precision for the June 30, 2021 FDIC SOD. US City/State and US Zip Code utilize the centroid of the state, city or ZIP code area to locate the bank branch. US streets, US Rooftop and “Assigned” geolocate branches with much greater precision. (Source: FDIC SOD of institutions and branches.)]
- See 75 Fed. Reg. 11642, 11666 (March 11, 2010)
- As of January 1, 2021, FDIC and the Federal Reserve categorized a bank as “large” if it had assets of at least $1.322 billion on December 31 of the previous two calendar years. Any bank with less than $330 million in assets for the same period was labeled “small.”