Tracking HMRC's compliance and enforcement expenditure in detail
Public spending transparency has become a familiar expectation in countries with mature democratic oversight. In the United Kingdom, HM Revenue and Customs publishes detailed breakdowns of how it allocates resources to chase unpaid tax, prosecute evasion and support compliant behaviour across the population. For Australians comparing fiscal systems across jurisdictions, the UK dataset offers a useful benchmark, especially as the Australian Taxation Office operates under a similar compliance mandate yet reports through a different legislative lens.
The records now aggregated on PublicSpending.net draw together multi-year accounts of HMRC's tax collection enforcement and compliance program expenditure, standardised against other jurisdictions such as Australia, Greece and the United States. Researchers, journalists and curious citizens can interrogate these figures through a SPARQL endpoint or download them in bulk. What follows is a guide to what those numbers mean, where the money goes, and how to use the published datasets in practice.
| Programme area | Approx. annual spend (GBP millions) | Comparable ATO spend (AUD millions) | Headcount involved |
|---|---|---|---|
| Debt recovery and field force | 1,200 | 850 | 12,400 |
| Criminal investigation and prosecutions | 460 | 220 | 4,100 |
| Customer compliance and risk assessment | 980 | 760 | 9,600 |
| Tax credits and benefits compliance | 740 | 410 | 6,800 |
| Digital and data systems support | 520 | 330 | 3,200 |
What the compliance and enforcement budget actually covers
HMRC's expenditure on tax collection enforcement and compliance is not a single line item but a cluster of interlocking programmes designed to close the tax gap, the difference between what is theoretically owed and what is actually collected. Funding flows into risk-assessment systems that segment taxpayers by behaviour, into field officers who visit businesses and high-net-worth individuals, into specialist teams that pursue offshore avoidance, and into the legal infrastructure that supports prosecution through the Crown Prosecution Service.
The categories cover both preventive work, such as nudge campaigns and real-time PAYE adjustments, and reactive work, such as raids on suspected fraud rings. Investment in analytics has shifted the cost profile considerably over the past decade, with more spend going into data platforms and less into traditional auditing desks. For Australians familiar with the ATO's justified trust programme, the parallel is reasonably direct, though the UK system retains a heavier physical presence in certain regions and historically placed more weight on face-to-face contact for small and medium enterprises.
Year-on-year trends in HMRC enforcement spending
Over the last decade, HMRC's compliance and enforcement budget has fluctuated between roughly £3 billion and £4.5 billion per year once fully loaded costs are included. Real-terms expenditure peaked around 2019, dipped during the pandemic as physical enforcement was paused, and recovered sharply by 2023. The recovery reflects both catch-up activity on deferred cases and an explicit policy to fund additional staff in the wealthy taxpayer and mid-sized business segments.
These movements matter because the yield per pound spent varies by programme. Debt recovery tends to return several pounds for every pound invested, while criminal investigation delivers lower direct yields but produces deterrent effects that are difficult to quantify. Australians looking at the same data can compare year-on-year shifts with ATO annual reports, where similar peaks and troughs are visible but explained through different administrative categories. The ANAO and parliamentary committees in Canberra rely on these published figures when scrutinising enforcement priorities.
The headline totals also mask significant composition changes. Between 2017 and 2023, the share of total compliance spending allocated to digital infrastructure roughly doubled, while the share allocated to traditional casework gradually fell. This pattern mirrors what the Australian Taxation Office has done with its investment in data analytics and automated risk scoring, suggesting a cross-jurisdictional shift away from human-led audit work toward platform-led detection. Researchers should treat the year-on-year numbers as a weighted sum rather than a stable basket of activities.
Where the money goes: people, systems and partnerships
The largest cost component is staff. HMRC employs tens of thousands of compliance officers, customer service representatives, IT specialists and caseworkers, with their salaries, pensions and overheads making up roughly two-thirds of the enforcement and compliance budget. Contracted services, primarily for digital infrastructure, account for most of the remainder, with smaller amounts spent on legal fees, training and accommodation.
Beyond its own workforce, HMRC partners with other UK bodies, including the Department for Work and Pensions and HM Treasury, to share intelligence and pursue tax-credit overpayments. Internationally, it exchanges information under treaties that allow the ATO in Canberra or its equivalent in Wellington to flag suspicious flows. These inter-agency costs are baked into the published totals, which is why the figures on PublicSpending.net are higher than HMRC's own narrow departmental accounts might suggest.
Comparing HMRC and ATO compliance spending
The comparison table earlier gives a snapshot, but a more detailed look reveals structural differences. HMRC spends a larger absolute amount on enforcement, reflecting both the UK population size and the breadth of taxes administered. The ATO, by contrast, allocates a higher proportion of its compliance budget to GST and large-business audit work, areas where Australian revenue dependence is greater relative to income tax. The Australian Taxation Office's administration is governed by the Taxation Administration Act 1953, which shapes its investigative powers in ways that differ from HMRC's enabling legislation.
For a researcher in Melbourne or Sydney, the practical value of having both datasets on a single platform is the ability to test hypotheses about cost-effectiveness across very different fiscal cultures. PublicSpending.net has aligned the categories where possible, though some harmonisation is necessarily approximate. Local considerations matter here too: ATO compliance priorities reflect Australian-specific realities such as the dominance of mining royalties, the spread of small businesses serving the Sydney and Melbourne metros, and the popularity of self-managed superannuation funds that HMRC has no direct analogue for.
How the data is collected and standardised
The numbers published on the site come from HMRC's annual reports and resource accounts, supplemented by Treasury data on allocated spending. Where HMRC reports internal categories that do not map cleanly onto a comparable taxonomy, the project team has either created a derived field or flagged the mismatch. The same approach is used for Australian, Greek and US datasets, which means cross-jurisdiction comparisons should be treated as indicative rather than exact.
Each record carries metadata on the source document, the reporting period and any revisions. The SPARQL endpoint lets users run linked-data queries that join HMRC records with, for example, population statistics or GDP figures, enabling time-series analysis without having to clean the raw files manually. The ontology behind the dataset is openly available, so researchers can verify how each term is interpreted before drawing conclusions in published work.
Several methodological choices deserve attention. Inflation adjustments are applied using country-specific deflators from the OECD, so a pound spent in 2015 is converted to its 2024 equivalent before being compared with current expenditure. Where HMRC has restated earlier figures, the site preserves both the original and restated values, allowing users to detect revisions. Currency conversions to AUD, EUR and USD use average annual exchange rates rather than spot rates, which smooths out short-term volatility and avoids spurious comparisons.
Putting the records to work
For a journalist in Perth preparing a story on offshore tax evasion, the bulk download option removes the manual extraction step entirely. For an academic in Brisbane writing on comparative tax administration, the standardised tables support direct citation. For a citizen in Adelaide who simply wants to know how much HMRC spent on prosecutions last year, the public-facing graphs answer that question in seconds.
Feedback from users, including several Australian researchers who have flagged edge cases, continues to refine the mappings. Each dataset is downloadable under an open licence, and the underlying records are versioned so that historical analyses remain reproducible even as HMRC revises its published figures. The site also publishes source documentation alongside each annual release, which means anyone checking a figure can trace it back to the originating HMRC report.
If you want to start exploring the HMRC compliance and enforcement dataset directly, head to the PublicSpending.net data portal where you can filter by programme, download CSV or JSON files, or open the SPARQL endpoint for custom queries. The bulk archive is updated as new HMRC annual reports are released, and the project team welcomes corrections, additional context, or new jurisdiction proposals from users in Australia, the UK or elsewhere.