Why pubs close: decomposing the 2026 numbers
By Jack Hawkins · Updated
Pub closures can reflect costs, leases, demand and the commercial circumstances of individual sites. The BBPA reported 295 closures from January to September 2026 and forecast around 360 for the year. September survivor trading measures improved in a separate Oxford Partnership/Vianet series. Those observations can coexist; they do not, by themselves, prove a single cause for closures or rule out weaker total market demand.
The cited September report describes higher occupancy, longer dwell and spend per head of £26.97 in its measured sample. It should not be treated as a census of all pub demand across 2026. Surviving venues can improve their averages while the estate shrinks, and higher spend can reflect price or mix as well as consumption. The useful analysis keeps the observation window, sample and cost context visible.
The five cost shocks, with their receipts
Employment costs. Payroll changes can materially affect a pub, with the amount determined by hours, wage rates, employer NI, holiday treatment and pension eligibility. The BBPA's £322m annual estimate covers the combined November 2025 Budget measures across pubs and brewers; it is not an employment-only calculation. Use the actual payroll bridge for a venue and keep the association's combined estimate separately labelled.
Rates. Business rates depend on the applicable valuation, multiplier and relief, not just the venue's current cashflow. England's eligible-pub relief for 2026/27 and the separately announced later relief need their own dates and eligibility rules. Regional closure totals are not closure rates: compare each area's starting estate before claiming one region faces a greater risk. Scotland, Wales and Northern Ireland have their own rates arrangements.
Energy and utilities. Cellar cooling, extraction and dishwashing can create a meaningful base load. Actual tariffs, equipment and opening hours determine the bill; the energy-price shock does not mean every venue still pays a fixed multiple of a 2021 amount. Measure consumption and contract terms separately. Efficiency changes may help, while safe storage and service requirements constrain what can be switched off.
Duty. Product duty depends on category, ABV and any qualifying relief. The rates effective 1 February 2026 increased the relevant duty amounts; qualifying draught and standard rates need to be distinguished. Model the duty included in a specific supplier price or release rather than applying one generic pint tax. A change in duty is one input into cost, not proof that it caused an individual closure.
The lease. A renewal or rent review can change a venue's fixed-cost base, but the terms and timing differ by property. An illustrative 25% rent increase matters according to the original rent and the venue's available contribution. A closure series without coded causes cannot show that lease resets explain every trading-difficulties statement. Model the actual agreement before choosing a negotiation, operating-hours or financing response.
What the geography and the survivors say
The BBPA reports roughly 2,200 net pub closures over five years and 43,656 pubs at the end of 2025. The separate September trading series describes suburban ROS growth of 3.2% and London growth of 2.0%. One possible explanation is demand shifting among surviving venues, but these datasets do not prove that survivors are better-run or that they absorbed all demand from closed sites. Keep population, geography and period alongside the comparison.
For an operator the honest takeaways are unglamorous and controllable: know your break-even coverage monthly and negotiate from above it, not below; treat labour percentage as one adjustable cost line alongside recipe choices, buying, waste and operating hours; and watch spend per head rather than footfall as the demand signal, because the cited September sample shows higher spend and dwell, with total visits and full-year demand remaining separate questions. A venue-level review should test demand and costs together. Market context can inform the questions, but the actual lease, payroll, buying terms, sales and cash position determine the options available to an individual pub.
Read closure statistics with a denominator
An area with more closures may simply have more pubs. To compare risk, divide closures by the relevant starting estate over the same period and use the same closure definition. Permanent closures, temporarily closed sites and net changes after openings answer different questions. The BBPA figures concern pubs; a separate trading sample may include bars and clubs and measures surviving venues. Taken together they provide context, not a controlled experiment proving why a particular pub closed. For an individual operator, combine that context with the actual lease, payroll, cashflow and customer demand before choosing a response.
FAQ
How many pubs are closing in the UK in 2026? 295 closures January-September, ~360 forecast for the year (BBPA, October 2026) — roughly one a day — against 43,656 remaining at end-2025 and ~2,200 net closures over five years.
Why are pubs closing if people are still drinking? A pub can face higher costs or an unfavourable lease even while some trading indicators improve. The BBPA's £322m estimate covers combined Budget measures, while the September ROS and spend figures concern a separate measured sample. Those figures do not identify a single closure cause: examine the individual venue's demand, costs, tenure and cash position together.
Which UK regions are losing the most pubs? 2026-to-September: England 226 (South East 42, North West 39 leading), Scotland 64, Wales 5 — raw regional totals, which need starting-estate denominators and further evidence before they can establish regional risks or causes.
What's the difference between closure counts and net closures? Gross closures count venues that shut; net closures subtract new openings. The BBPA's 295/360 figures are closures; the five-year 2,200 figure is net. The difference remains material: openings can offset some closures, and definitions of permanent closure, licensed premises and operating venues vary. Compare the same population and dates before combining the series.
Method and assumptions
BBPA closure and cost data, October 2026; checked 19 October. Oxford Partnership/Vianet Market Watch September 2026 (published 14 October 2026). IWSR 2025 global volume reset, UK category reports. UK Bar Trading Index provides the dated first-edition context. Analysis is Better Bartender's; underlying figures remain the sources'. Method: data standards.
Edition of 16 October 2026.