UK pub business closures can be measured in more than one way. The Office for National Statistics recorded 4,850 VAT and/or PAYE enterprise deaths in the exact public houses and bars category in 2016, falling to 3,815 in 2020. A broader ONS category, beverage serving activities, recorded 4,190 enterprise deaths in 2024. These measures cover different classifications, so they should be read separately.
Contents
- Exact public-house and bar deaths
- The broader beverage-serving trend
- Active businesses and new enterprise births
- Regional closure counts in 2020
- How many new businesses survived
- Employer deaths
Exact public-house and bar deaths
The most specific figures in the supplied ONS series use UK Standard Industrial Classification code 56302, public houses and bars. The ONS measure is VAT and/or PAYE enterprise deaths: businesses that disappeared from the relevant enterprise population during the stated year. It is not a count of buildings demolished, premises sold, or every individual pub that changed hands.
The ONS figures for 2016 through 2020 were:
| Year | Enterprise deaths |
|---|---|
| 2016 | 4,850 |
| 2017 | 4,495 |
| 2018 | 4,275 |
| 2019 | 4,420 |
| 2020 | 3,815 |
Source: ONS, “Pub closures and their locations between 2017 to 2022”.
The annual count declined from 4,850 in 2016 to 4,495 in 2017 and 4,275 in 2018. It then rose to 4,420 in 2019 before reaching 3,815 in 2020. The 2020 figure is therefore the lowest of the five exact-category observations supplied, but it should not be treated as proof that operating conditions were easier that year. Enterprise-death counts describe exits recorded by the statistical system; they do not explain why each exit occurred.
For a bar manager, the classification matters. A pub that changes ownership may not be represented as a death if the enterprise remains in the statistical population. Conversely, the figures are about enterprises rather than a simple physical count of signs above doors. They are best used as a measure of business exits within SIC 56302.
The broader beverage-serving trend
The latest annual ONS business-demography series uses SIC group 563, beverage serving activities. This is broader than public houses and bars, so the following numbers are not exact pub-only closure totals. ONS recorded the following VAT and/or PAYE enterprise deaths:
| Year | Beverage-serving enterprise deaths |
|---|---|
| 2019 | 4,775 |
| 2020 | 4,120 |
| 2021 | 3,065 |
| 2022 | 3,660 |
| 2023 | 4,440 |
| 2024 | 4,190 |
Source: ONS, “Business demography, UK”.
Within this broader group, deaths were 4,775 in 2019 and 4,120 in 2020. The count fell to 3,065 in 2021, then rose to 3,660 in 2022 and 4,440 in 2023. It was 4,190 in 2024. The latest two years of deaths in this annual business-demography series are provisional under the dataset notes, and ONS controls counts by rounding.
The broader series shows why a single “pub closures” number can be misleading. The 2019 exact 56302 count was 4,420, while the broader group 563 count was 4,775 for the same year. Those values answer different questions: one concerns public houses and bars, while the other includes the wider beverage-serving group. A comparison should identify the SIC level before interpreting movement.
Active businesses and new enterprise births
Business deaths are only one side of the market. ONS counted 42,585 active VAT and/or PAYE enterprises in beverage serving activities in 2019. The active base was 41,080 in 2020, 40,030 in 2021, 40,420 in 2022, 40,645 in 2023 and 40,415 in 2024.
These are active enterprises in broader SIC group 563, not an exact count of pubs. The supplied observations show a base above 42,000 in 2019, followed by figures close to 40,000 from 2021 through 2024. The 2024 count was 40,415.
ONS also recorded enterprise births in the same broader group:
- 2019: 4,795
- 2020: 3,615
- 2021: 3,340
- 2022: 3,505
- 2023: 4,090
- 2024: 4,340
Source for the active-enterprise and birth figures: ONS, “Business demography, UK”.
The birth count was 4,795 in 2019, compared with 3,615 in 2020 and 3,340 in 2021. It increased to 3,505 in 2022, 4,090 in 2023 and 4,340 in 2024. These figures describe enterprises entering the statistical population, not necessarily new premises opening under a completely new concept. A manager assessing competition or local demand should keep that distinction in mind.
The active base and birth series also should not be added together or interpreted as a simple annual profit-and-loss statement. The active count is a stock measured for the group, while births and deaths are flows during a period. They can be read together to describe market churn, but the supplied data does not provide a complete reconciliation of every movement.
Regional closure counts in 2020
The ONS regional breakdown covers exact SIC 56302 public-house-and-bar enterprise deaths in 2020. The United Kingdom total was 3,815. Counts for the named regions were:
| Region | Enterprise deaths |
|---|---|
| North East | 185 |
| North West | 585 |
| Yorkshire and The Humber | 400 |
| East Midlands | 315 |
| West Midlands | 380 |
| East | 300 |
| London | 230 |
| South East | 445 |
| South West | 420 |
Source: ONS, “Pub closures and their locations between 2017 to 2022”.
The largest supplied regional count was 585 in the North West. The South East recorded 445, the South West 420, Yorkshire and The Humber 400, and the West Midlands 380. The East Midlands count was 315, the East was 300, London was 230, and the North East was 185.
These are counts, not regional closure rates. The supplied figures do not include the number of pubs or bars operating in each region, population, floor area, or the size of the local customer market. A higher count therefore does not automatically mean a higher risk for an individual venue. It means more SIC 56302 enterprise deaths were recorded in that regional total during 2020.
How many new businesses survived
The ONS cohort figures follow beverage-serving enterprises born in a given year. They are again for broader SIC group 563, not exact pub-only cohorts. For the 2019 birth cohort, 4,505 enterprises survived one year, equal to 94.0%; 3,685 survived two years, equal to 76.9%; 3,045 survived three years, equal to 63.5%; 2,545 survived four years, equal to 53.1%; and 2,095 survived five years, equal to 43.7%.
The 2020 cohort had 3,385 survivors after one year, or 93.6%; 2,890 after two years, or 79.9%; and 2,280 after three years, or 63.1%. The 2021 cohort had 3,220 survivors after one year, or 96.4%, and 2,680 after two years, or 80.2%. For the 2022 cohort, 3,325 enterprises survived one year, or 94.9%.
Source: ONS, “Business demography, UK”.
| Birth cohort | Survival period | Survivors | Survival rate |
|---|---|---|---|
| 2019 | 1 year | 4,505 | 94.0% |
| 2019 | 5 years | 2,095 | 43.7% |
| 2020 | 3 years | 2,280 | 63.1% |
| 2021 | 2 years | 2,680 | 80.2% |
| 2022 | 1 year | 3,325 | 94.9% |
The complete cohort sequence gives more context than a one-year closure count. In the 2019 cohort, the reported survival rate moved from 94.0% at one year to 43.7% at five years. That is a cohort survival measure, not a forecast for every new pub or bar. It also does not identify causes of closure, differences between regions, or the financial performance of the survivors.
Employer deaths
ONS separately recorded employer deaths in beverage serving activities. In this dataset, an employer death is an enterprise meeting the dataset’s employee-based definition. It is therefore a narrower measure than all enterprise deaths in the broader group.
Employer deaths were 330 in 2019, 275 in 2020, 270 in 2021, 370 in 2022, 465 in 2023 and 430 in 2024. The 2023 count was the highest in the supplied six-year sequence, followed by 430 in 2024.
Source: ONS, “Business demography, UK”.
For operational planning, this distinction prevents two common errors. First, an enterprise-death total should not be described as the number of employer businesses that closed. Second, employer-death figures should not be presented as all pub closures. The exact 56302 series, the broader 563 series, active-enterprise counts, births, cohort survival and employer deaths each measure a different part of business change. ONS counts are rounded, and the latest two years of deaths in the annual business-demography series are provisional under the dataset notes.