This chart maps banking hubs, bank branches and Post Offices against the FCA’s access to cash criteria. The baseline is a service point within 1 mile in urban areas and 3 miles in rural areas, urban being the built-up extent of a settlement of 10,000 or more people.
The legend identifies each service type and its shaded area shows where that criterion is met. It shows where a qualifying point lies within reach. It does not weight points, measure queues, opening hours or the services actually offered, and it uses straight-line distance rather than travel time.
Whilst this is not a direct proxy for access to banking, which can be done virtually, unlike cash which is solely physical, it provides a helpful view of where there are potential access gaps.
Turn on the Constituencies overlay to see a per-seat breakdown. Estimated against small-area census geography, around 130,000 residents across the UK live beyond the access to cash standard, concentrated in the most rural seats.
| Constituency | % within | Est. beyond | Hubs | Banks | POs |
|---|
The analysis estimates the extent to which the UK population lives within reach of a staffed cash and banking service point, measured against the FCA’s access to cash geographic baseline, and reports the result both nationally and for each of the 650 Westminster constituencies created in 2024.
The FCA access to cash regime, set out in its ATCS sourcebook, frames reasonable access in distance terms. We apply that baseline directly: a location meets the criterion if a qualifying service point lies within one mile in urban areas or three miles in rural areas. Urban is defined as lying within the built-up extent of a settlement of 10,000 or more residents, consistent with the ONS rural-urban basis. All distances are straight-line, not road or travel-time, mirroring how the FCA baseline itself is expressed.
Three layers are treated as qualifying staffed service points: banking hubs, bank and building society branches, and Post Office branches. Hubs are from the Cash Access UK list and include those open and under construction. Branches are the open, fixed-premises branches from the GeoLytix Retail Points (Banks) October 2025 release, excluding mobile units and locations already counted as hubs to avoid double counting. Post Offices are the 2019 Post Office Ltd open-branch list. Including all three reflects the most inclusive qualifying scenario; the data retains the bank versus building society distinction so narrower scenarios can be modelled later.
Constituency populations are the ONS mid-2022 estimates by 2024 constituency. To estimate how that population is distributed within a seat, we use the population-weighted centroids of small-area census geography — Lower Layer Super Output Areas in England and Wales and their equivalents in Scotland and Northern Ireland, roughly 48,000 areas of about 1,500 residents each. Each small area is assigned to a constituency by point-in-polygon against the July 2024 ONS boundaries and carries an equal share of its constituency’s population, appropriate because equal population is the design principle of the geography.
For each small-area centroid we test whether any qualifying service point lies within the applicable distance. Because a single centre point stands in for an area of real extent, a size-scaled tolerance is added: small in dense urban areas, larger in sparse rural ones, derived from the spacing between neighbouring centroids and capped at 1.5 kilometres. This reflects that part of a physically large rural area lies closer to a service point than its centre does. The population of areas still beyond the standard is summed per constituency to give a residents-beyond figure and a percentage within.
The service-point counts per constituency and the constituency populations are actual figures. The within-or-beyond split is a model: it depends on the small-area centroid as a proxy for where people live, on the urban-rural classification, and on the size-scaled tolerance. The headline percentage is an estimate; the population totals beneath it are firm.
Distances are straight-line, so the model flatters hilly terrain, estuaries and islands where road distance is longer. Service points are at town-centre approximations for hubs and reported coordinates for branches and Post Offices, not verified premises. The Post Office data is from 2019 and predates later closures and openings. A small area is judged on its single centre point, so a large rural area can be misjudged at its edges, which the size-scaled tolerance mitigates but does not eliminate. The tolerance cap is a defensible judgement, not an empirical constant, set deliberately conservative; sensitivity to that choice is material. Finally, the analysis measures physical proximity to cash, which is not a direct proxy for access to banking, much of which is now conducted digitally.
Each layer is attributed above with its source and licence: Cash Access UK for hubs, GeoLytix under CC0 for branches, Post Office Ltd under the Open Government Licence for Post Offices, and ONS under the Open Government Licence for boundaries, population and small-area geography. The map is a single self-contained file with the data embedded, so the figures can be regenerated and audited.