Credit unions physically cover under half the country's counties; banks are near-universal — and that gap has been stable, even as credit unions' financial share slowly grows

Summary

By physical footprint, banks cover the country almost completely and credit unions cover well under half of it. Bank branches show up in essentially every county nationwide (3,199 distinct county-FIPS codes with a reporting FDIC office), while credit union branches are concentrated in roughly 1,430-1,440 of the country's ~3,143 counties, about 46% — and that share has barely moved since 2017. Credit unions have shed branches at roughly the same steady pace as banks (from ~19,700 CU offices in 2017 to ~17,300-17,500 today, a decline of about 11-12%), so the geographic gap between the two sectors has not closed. Where credit unions HAVE gained ground is financially, not geographically: their share of combined bank+credit-union assets grew from roughly banks-are-13x-larger in 2014 to roughly banks-are-10x-larger today, and membership hit 144.7 million as of Q4 2025 (NCUA), equal to about 42% of the US population (though that overstates unique people, since many Americans belong to more than one credit union).

How 'footprint' was measured

Two distinct footprint measures were computed directly from the warehouse and are reported separately because they answer different questions:

• Physical branch presence by county — computed from banking.ncua_branch_locations (NCUA's quarterly 5300 Call Report branch file, 2017 Q1 through 2026 Q1) and banking.locations (the FDIC's current office-location snapshot). This captures where a member/customer can walk into a branch, which understates true credit union reach because many credit unions participate in shared-branching networks (CO-OP Shared Branch etc.) that let members use a DIFFERENT credit union's physical branch — a channel this branch-ownership data cannot see.
• Financial/membership share — pulled from NCUA and Federal Reserve/FDIC published aggregates (assets, membership), since the warehouse's banking schema does not carry a national credit-union asset or membership time series to query directly; these figures are cited to their web sources, not computed from askengine tables.

The county-coverage gap has not closed

Querying banking.ncua_branch_locations by quarterly cycle: the count of distinct counties with at least one credit union branch office ran 1,428 (2017 Q1) → peaked near 1,442-1,443 (2019-2020) → 1,436 (2024 Q1), a range of essentially 45-46% of the roughly 3,143 US counties and county-equivalents throughout the whole nine-year window. There is no visible trend toward broader OR narrower geographic reach — credit unions have consolidated (fewer institutions, fewer branches, per NCUA) without materially changing how many places have a branch at all. The NCUA source file's county-name field stopped populating reliably after the 2024-12 cycle (0 distinct county values in 2025-2026 cycles), so the most recent five quarters could not be included in the county-coverage series; this is flagged as a data-quality gap in the underlying NCUA source extract, not a real collapse in coverage.

By contrast, banking.locations (FDIC office locations) shows bank branches present in 3,199 distinct county-FIPS codes — more than the ~3,143 mainland counties because the FDIC file also includes territories and some duplicate/independent-city codes — meaning for practical purposes every county in the country has at least one bank branch.

Branch counts are shrinking for both sectors, at a similar pace

Credit union branch offices declined from 19,725 (2017 Q1) to 17,492 (2026 Q1) in the warehouse's NCUA data — down about 11%. Independently, bank branches nationwide fell from 82,461 (2012) to 68,632 (2024), down about 17% over a longer window, with the pace of closures slowing sharply in 2023-2024 (net declines of ~1,500 and ~1,100 branches respectively, versus ~3,100 and ~2,800 in the pandemic years 2021-2022), per Statista/Federal Reserve figures and the ABA's summary of FDIC/NCUA deposit statistics. Both sectors are shrinking their physical networks as digital banking substitutes for branches; nothing in either data source suggests credit unions are either gaining or losing physical ground relative to banks.

Where credit unions ARE gaining: assets and membership, not geography

NCUA membership reached 144.7 million as of Q4 2025 (NCUA press release), against a 2025 US population estimate of 341.8 million (Census Bureau) — about 42%, though this overstates unique individuals since a person can hold memberships at more than one credit union. On the asset side, credit unions held about $2.4 trillion in assets in 2025 versus roughly $24.4 trillion at banks — about 8.9% of the combined total — and that ratio has been narrowing: banks held roughly 13x as many assets as credit unions in 2014, versus roughly 10x in 2025. So the credit-union sector's SHARE of the country's deposits/assets and its membership base have both grown gradually, even while its physical branch FOOTPRINT (measured either as branch count or as counties reached) has not expanded and, if anything, has slightly contracted along with the rest of the industry.

Caveats

• Physical-branch presence undercounts true credit union reach because of shared-branching networks not visible in ownership-level branch data.
• The NCUA county field is unreliable after 2024-12 in the warehouse extract; the county-coverage trend line stops at 2024 Q1 for that reason, not because coverage actually vanished.
• The FDIC's 3,199-county figure modestly exceeds the ~3,143 standard county count due to territories/independent cities in the FDIC file; it is used here only to support 'near-universal coverage,' not as a precise county count.
• Membership and asset figures (144.7M members, $2.4T CU / $24.4T bank assets, the 13x→10x ratio) are pulled from NCUA, Federal Reserve, and secondary compilations (WalletHub/MX) via web search, not computed from an askengine table — the warehouse's banking schema has no national CU asset/membership time series loaded.

Every query behind this report

3 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

SELECT MIN(established_date), MAX(established_date), COUNT(*) FROM banking.locations

SELECT cycle, COUNT(DISTINCT site_id) AS branches, COUNT(DISTINCT cu_number) AS cus, COUNT(DISTINCT state_abbr) AS states, COUNT(DISTINCT county_name) AS counties FROM banking.ncua_branch_locations GROUP BY cycle ORDER BY cycle

SELECT COUNT(*) AS branches, COUNT(DISTINCT cert) AS banks, COUNT(DISTINCT state_abbr) AS states, COUNT(DISTINCT county_fips) AS counties FROM banking.locations