HMRC have published their 2026 Measuring Tax Gaps statistics. The total tax gap has ticked up somewhat, but a lot is going on under the surface. Drawing on the discussions at our recent residential conference and a review of the methodology underlying the tax gap, CenTax Directors Professor Arun Advani and Professor Andy Summers ask how much weight the headline figures should carry, and whether a different approach is needed.
HMRC’s 2026 Measuring Tax Gaps publication puts the UK tax gap at £59.2 billion or 6.4% of theoretical liabilities for the 2024-25 tax year, compared with £52.8 billion (6.0%) in 2023-24. Small businesses again account for the largest share at 62%, driven by Corporation Tax.[1]
The tax gap statistic does a lot of work in UK public life. It drives headlines about how much tax goes uncollected. It frames parliamentary scrutiny of HMRC’s performance and shapes the public conversation about who is and is not paying their fair share. For a number carrying that much weight, the foundations of some of its largest components are surprisingly thin.
Here we focus on two areas that stand out for different reasons: the small business tax gap, which is the largest single component of the headline figure, and the offshore tax gap, which is excluded from the annual publication and where the separate published estimate covers only a small slice of the problem. Both were prominent themes at our residential conference, Closing the Tax Gap, held at the University of Warwick in April.[2] We draw on what we learned at the conference in what follows.
The largest single component of the tax gap rests on a tiny sample
The small business tax gap is the largest single share of the published total, at 62% of the headline figure. The biggest driver within it is Corporation Tax for small businesses, at £17.3 billion, almost half (44.6%) of the Corporation Tax these companies should be paying.
Based on HMRC’s published estimates, this small business Corporation Tax gap has grown from £1.2 billion in 2011-12 (9.2%) to £17.3 billion in 2024-25 (44.6%).[3] If real, that would be an astonishing increase, although HMRC warn that figures are not directly comparable over the whole time period.[4] In any case, £17.3 billion is larger than the entire budget for the Ministry of Justice, which is responsible for courts, prisons and probation.[5] Is there really that much potential revenue to be had from clamping down on small businesses?
To assess how much weight we should put on these figures, it is important to understand where they come from. They are reached using a ‘bottom up’ method: HMRC select a sample of taxpayers at random each year, conduct a full compliance check on each, and scale the findings up to the wider population. That is the right method, and what most tax authorities use. However, the size of the sample used for small businesses raises serious questions about the reliability of the numbers that HMRC produce.
HMRC’s Corporation Tax random enquiry programme for small businesses has run at 324 cases for 2019-20, 327 for 2020-21, 359 for 2021-22, and 337 for 2022-23. Results are then scaled up across the whole population of small companies that file a Corporation Tax return. A sample of around 330 audits a year is what underpins the £17.3 billion figure.[6]
To be clear, especially since it may come as a surprise to many: 330 is not the number of non-compliant cases detected; it is the total number of audits undertaken, with non-compliant cases a subset of these. So, an estimate of billions of pounds of missing tax rests on little more than a handful of non-compliant cases, scaled up to the full population. HMRC recognise that this is problematic, since they raised the overall uncertainty rating for the small business Corporation Tax gap from ‘medium’ to ‘high’ this year.[7]
International comparisons highlight how tiny the UK’s random audit programme currently is. In 2023 the IMF reviewed how different countries estimate the small business Corporation Tax gap, and the UK came out at the bottom of the comparison set on sampling rate.[8] Canada uses a sample of 4,500 audits to estimate its tax gap, more than ten times the UK sample. The smallest documented sample size outside the UK was for Denmark, which used a sample of 350 audits, but for a business population that is 10 times smaller than the UK’s. While other countries have around 100 audits per 100,000 of business population, the equivalent figure for the UK is 17.
This low sampling rate explains why HMRC have been unable to provide further details about what is driving changes in the small business tax gap. With such a small sample, it is difficult even to produce a reliable estimate of the total amount of non-compliance within this population.[9] Breaking down these 17 businesses per 100,000 business population into even smaller groups, for example by industry, would not be statistically meaningful, and so rightly HMRC do not attempt it. There is simply no way HMRC can produce the breakdowns that many observers have been asking for, without radically increasing the number of random audits that HMRC conduct.
This does not necessarily mean HMRC’s current estimate of the small business Corporation Tax gap is wrong. However, with a sample size that small, it is not possible to know. Tackling the small business tax gap must start with an expansion of HMRC’s random audit programme to a level where we can begin to assess with confidence how large this gap actually is and what is driving it. A tenfold increase would be required to bring the UK up to its international peers, and with so much tax potentially at stake, serious steps in this direction look obviously worthwhile.
The offshore tax gap remains absent
This year’s publication does not include an updated offshore tax gap. HMRC’s first estimate was published as a separate statistic in October 2024, putting tax lost from undisclosed foreign income at £300 million in 2018-19.[10]
Over the last decade, tax authorities have been exchanging information with HMRC about UK resident individuals, with foreign income, under arrangements known collectively as Automatic Exchange of Information (AEOI). HMRC’s offshore estimate matches that data against Self-Assessment returns, identifies cases where AEOI-reported foreign income doesn’t match what was declared, audits a random sample of 400 individuals with discrepancies, and scales the findings up to the population covered by AEOI.
HMRC rightly emphasise that this estimate does not represent the entire ‘offshore tax gap’. As is well-known, it only captures undisclosed foreign income held in jurisdictions that exchange information with the UK. However, less well-known is that it effectively only captures under-declared interest on savings, and dividends from a small subset of shares (where these are held by an intermediary). Foreign earnings, trading profits, pension income, and most other foreign investment income are not included in the estimate. The estimate also excludes accounts held via companies, trusts and other legal entities controlled by UK residents, and assets such as property and shares held directly. HMRC acknowledge that the figure is “likely to be a lower bound”. The National Audit Office has said the published estimate is too partial to count as a measure of the offshore tax gap.[11]
Consider if we wanted to measure non-compliance among hairdressers by checking each hairdresser’s books against their reported income. No one would think this was a complete measure of the hairdressing tax gap, because hairdressers who want to evade tax are unlikely to put the income through their books in the first place. The published offshore tax gap has the same problem. Only the most amateurish tax evader would fail to declare savings interest on a directly-held bank account in a CRS-reporting jurisdiction, as the data trail to HMRC is automatic. The techniques that deliberate evaders are most likely to use are invisible to this measure.
In its recent Treasury Minute response to the Public Accounts Committee (PAC), the government committed to publish an improved offshore tax gap measure by 2027.[12] That commitment is welcome. Much of the data, methodology and analytical expertise needed to produce a more comprehensive measure sits outside HMRC, in the research community. HMRC should engage actively with researchers in designing and producing this new measure, building on similar collaborations between tax authorities and researchers in the United States and Denmark.[13] The categories the published estimate currently omits, including real estate held offshore, indirectly held assets, and structures involving trusts, are precisely those where the most determined offshore non-compliance is now most likely to sit.
Where next for the tax gap publication?
The two issues we have highlighted are different in nature. The small business Corporation Tax gap is published, well-known, and central to the headline figure. The offshore tax gap is partial, recently published, and not yet an established measure of the issue it is trying to capture. Both are regularly cited in conversations around tax compliance in the UK. However, as we set out here, the evidence that underpins them is questionable.
HMRC’s tax gap publication is often described as world-leading because it is unique in being both comprehensive and annual. But this approach also has drawbacks. The cost of trying to cover the full breadth of the tax system every year is a lack of depth in any particular area. We think that HMRC should be granted the resources needed to continue publishing a comprehensive annual estimate of the tax gap, whilst also shoring up the foundations of key components. But if it is not possible to do both, then it may be better to sacrifice the flagship publication in favour of a rotating programme that targets specific areas of the tax gap on a less than annual cycle.
25 June 2026
About the authors
Arun Advani is Director of the Centre for the Analysis of Taxation (CenTax) and Professor of Economics at the University of Warwick.
Andy Summers is Director of the Centre for the Analysis of Taxation (CenTax) and Professor of Law at the London School of Economics and Political Science.
References
[1] HMRC, Measuring tax gaps 2026 edition, https://www.gov.uk/government/statistics/measuring-tax-gaps This year’s edition revises the 2023-24 gap up from £46.8 billion (5.3%) at the time of the original publication in June 2025.
[2] CenTax, Closing the Tax Gap: CenTax Residential Conference 2026, https://centax.org.uk/closing-the-tax-gap-centax-residential-conference-2026/.
[3] HMRC, Measuring tax gaps 2026, Chapter 5: Corporation Tax.
[4] HMRC now treat the earlier years as likely understated. They state they have become better at finding and recording non-compliance since 2019-20, and warn that figures before and after that point are not directly comparable.
[5] Ministry of Justice, Annual Report and Accounts 2024-25, HC 36-27 (November 2025), https://assets.publishing.service.gov.uk/media/691213e3e9348ac8fb54f4c4/36.27_MOJ_Annual_Report_and_Accounts_2024-2025_WEB.pdf. Total gross expenditure for 2024-25 was £15.5 billion.
[6] HMRC, Measuring tax gaps: Methodological annex, 2026 edition, https://www.gov.uk/government/statistics/measuring-tax-gaps-methodological-annex
[7] HMRC, Methodological annex 2026, Chapter B.
[8] Barra, Hutton and Prokof’yeva, Corporate Income Tax Gap Estimation by Using Bottom-Up Techniques in Selected Countries: Revenue Administration Gap Analysis Program, IMF Technical Notes and Manuals 2023/06 Table 2, https://www.imf.org/-/media/files/publications/tnm/2023/english/tnmea2023006.pdf.
[9] For this reason, HMRC have this year taken the decision to move to reporting the small business tax gap based on a rolling average that combines multiple years of audit data, instead of relying on data from a single year, which is highly volatile. This change is welcome from a statistical point of view, but the underlying problem, which has not been fixed, is the tiny sample size in each year’s random audit programme.
[10] HMRC, Undisclosed foreign income in Self Assessment by UK residents, 2018 to 2019 (24 October 2024), https://www.gov.uk/government/statistics/undisclosed-foreign-income-in-self-assessment-by-uk-residents-2018-to-2019/undisclosed-foreign-income-in-self-assessment-by-uk-residents-2018-to-2019.
[11] National Audit Office, Collecting the right tax from wealthy individuals, HC 876, Session 2024-25 (16 May 2025), https://www.nao.org.uk/wp-content/uploads/2025/05/collecting-the-right-tax-from-wealthy-individuals.pdf.
[12] HM Treasury, Treasury Minutes: Government response to the Committee of Public Accounts on the Fortieth report from Session 2024-25. https://assets.publishing.service.gov.uk/media/6a0dcc939171ee9f72826f01/E03613937-_CP_1587_Treasury_Minutes_Accessible.pdf
[13] See for example Guyton, J., Langetieg, P., Reck, D., Risch, M. and Zucman, G. (2021), Tax Evasion at the Top of the Income Distribution: Theory and Evidence, NBER Working Paper No. 28542, https://www.nber.org/papers/w28542; Boas, H. F., Johannesen, N., Kreiner, C. T., Larsen, L. T. and Zucman, G. (2024), Taxing Capital in a Globalized World: The Effects of Automatic Information Exchange, NBER Working Paper No. 32714, https://www.nber.org/papers/w32714; Johannesen, N., Langetieg, P., Reck, D., Risch, M. and Slemrod, J. (2020), ‘Taxing Hidden Wealth: The Consequences of US Enforcement Initiatives on Evasive Foreign Accounts’, American Economic Journal: Economic Policy 12(3): 312-346, https://www.aeaweb.org/articles?id=10.1257/pol.20180410.